LIGANDS FOR ODOR RECEPTORS AND OLFACTORY NEURONS

Abstract

The disclosure provides compounds useful as insect repellents and compositions comprising such repellents. The disclosure further provides insect traps and method for identifying ligands and cognates for biological molecules.

Description

TECHNICAL FIELD

The disclosure provides compounds useful as insect repellents and compositions comprising such repellents. The disclosure further provides compounds useful as insect attractants and compositions comprising such attractants. The disclosure further provides compounds useful as insect traps.

BACKGROUND

Numerous insects are vectors for disease. Mosquitoes in the genus Anopheles are the principle vectors of malaria, a disease caused by protozoa in the genus Trypanosoma. Aedes aegypti is the main vector of the viruses that cause Yellow fever and Dengue. Other viruses, the causal agents of various types of encephalitis, are also carried by Aedes spp. mosquitoes. Wuchereria bancrofti and Brugia malayi, parasitic roundworms that cause filariasis, are usually spread by mosquitoes in the genera Culex, Mansonia, and Anopheles.

Horse flies and deer flies may transmit the bacterial pathogens of tularemia (Pasteurella tularensis) and anthrax (Bacillus anthracia), as well as a parasitic roundworm (Loa loa) that causes loiasis in tropical Africa.

Eye gnats in the genus Hippelates can carry the spirochaete pathogen that causes yaws (Treponema pertenue), and may also spread conjunctivitis (pinkeye). Tsetse flies in the genus Glossina transmit the protozoan pathogens that cause African sleeping sickness (Trypanosoma gambiense and T. rhodesiense). Sand flies in the genus Phlebotomus are vectors of a bacterium (Bartonella bacilliformis) that causes Carrion's disease (oroyo fever) in South America. In parts of Asia and North Africa, they spread a viral agent that causes sand fly fever (pappataci fever) as well as protozoan pathogens (Leishmania spp.) that cause Leishmaniasis.

SUMMARY

The methods of the disclosure provide an odor receptor optimized descriptor-based in silico screen of chemical space. The methods of the disclosure are useful for identifying ligands for odor receptors (Ors), greatly reducing the number of compounds needing to be physically tested through methods such as single-unit electrophysiology or cell imaging. In addition a very large number of odorants can be computationally predicted in a single run of a chemical informatics pipeline, thus enabling one to select the appropriate chemicals to use as ligand for target odor receptor based on other important considerations that can be easily determined such as volatility, solubility, toxicity, costs, environmental safety or other physico-chemical properties. As most approaches to ligand identification require physically testing odorants using expensive assays and purchasing large collections of test chemicals is very expensive, the in silico approaches described herein provides the ability to predict ligands with high accuracy greatly reduces the cost of identifying novel ligands.

The disclosure provides a method of identifying a ligand for a biological molecule comprising: (a) identifying a known ligand or set of known ligands for a biological molecule, or identifying a compound which causes a specific biological activity, (b) identifying a plurality of descriptors for the known ligand or compound, (c) using a Sequential Forward Selection (SFS) descriptor selection algorithm to incrementally create a unique optimized descriptor subsets from the plurality of descriptors for the known ligand or compound, (d) identifying a putative ligand or compound that best-fits the unique optimized descriptor subset, and (e) testing the putative ligand or compound in a biological assay comprising the biological molecule wherein a change in activity of the biological molecule compared to the molecule without the putative ligand is indicative of a ligand the interacts with the biological molecule. The method above can be applied to any number of biological molecules that have a binding cognate. For example, the biological molecule can be a receptor, a ligand gated ion channel or G-protein coupled receptor. In a specific embodiment, the receptor is an odor receptor. In another embodiment, the receptor is expressed in a cell. In any of the foregoing embodiments, the plurality of descriptors are selected from the group consisting of distance metrics, descriptor sets, and activity thresholds. Further, in any of the foregoing embodiments, the distance metrics are selected from the group consisting of Euclidean, Spearman, and Pearson coefficients. In any of the foregoing embodiment, the descriptor sets are selected from Dragon, Cerius2, and a combined Dragon/Cerius2 set. In yet another embodiment, which can be implemented and used with any of the foregoing embodiments, two activity threshold methods are compared. In a further embodiment, the activity threshold comprises spike activity cutoffs and a cluster-based cutoff. In yet another embodiment of any of the foregoing the identifying further comprises selecting a putative ligand or compound with in a desired Euclidian distance of the known ligand or biological compound. For example, the Euclidian distance is about 0.001 to about 6.60 from a known ligand or cluster of ligands in chemical space. In another embodiment, the ligand binds to a CO₂ receptor and wherein the ligand has a Euclidian distance of about 0.001 to 6.60 from a known ligand for a CO₂ receptor. In yet another embodiment, the putative ligand is selected from a compound in Table 9 and 10. In another embodiment of any of the foregoing the descriptors are selected from the descriptors in Table 7 and 8. The methods described above can utilize a known ligand or set of known ligands identified through electrophysiology, imaging assays, or binding assays. The methods above can be used to screen a library of compounds. The method may be fully automated or may output the putative ligand or compound to a user who may then perform a biological assay. The biological assay can use various indicators for determining a ligand (e.g., an agonist or antagonist ligand) including a biological assay measuring a change in spike frequency, florescence intensity, or binding affinity. The odor receptor may be a vertebrate or invertebrate odor receptor. In yet another embodiment of any of the foregoing, the putative ligands or compounds are soluble ligands or compounds and the receptor is a gustatory receptor expressed by an invertebrate species or a gustatory receptor neurons present in an invertebrate. In yet another embodiment of any of the foregoing, the putative ligands or compounds the receptor is a gustatory receptor expressed by an invertebrate species or a gustatory receptor neurons present in an invertebrate. In yet another embodiment of any of the foregoing, the putative ligands or compounds the receptor is a gustatory receptor expressed by an invertebrate species or a gustatory receptor neurons present in an invertebrate. In yet another embodiment of any of the foregoing, the putative ligands or compounds the receptor is a gustatory receptor expressed by an mammal species or a gustatory receptor neurons present in an mammal. In yet another embodiment of any of the foregoing, the putative ligands or compounds the receptor is a gustatory receptor expressed by an mammal species or a gustatory receptor neurons present in an mammal. In yet another embodiment of any of the foregoing, the putative ligands or compounds the receptor is a gustatory receptor expressed by an mammal species or a gustatory receptor neurons present in an mammal.

The disclosure also provides a ligand or compound identified by the method of any of the foregoing claims. In one embodiment, the compound/ligand is set forth in Table 4, 6, 9 and 10. The ligand or compound can be an odor receptor ligand having a desired Euclidian distance from a cluster of known ligands defined by structural-data information wherein the compound reversibly or irrevisibly binds an odor receptor.

The disclosure also provides use of a ligand or compound identified by the methods of the disclosure or a ligand or compound in Table 4, 6, 9 or 10 to lure insect species into traps by virtue of activating odor receptors or odor receptor neurons. In an embodiment, the trap is suction based, light based, electric current based. In another embodiment, the ligand or compound is used the preparation of a topical cream, spray or dust present within or near a trap entrance. The ligand or compound can be used in a vapor emitted from vaporizers, treated mats, treated pods, absorbed material, cylinders, oils, candles, wicked apparatus, fans, within or near trap entrances. The ligand or compound can be used a repellant or attractant. The repellant or attractant can be used in a cream, lotion, spray, dust, vapor emitter, candle, oil, wicked apparatus, fan, or vaporizer. The ligand or compound can be used to affect mating behavior.

The disclosure also provides a composition comprising a ligand or compound of as described above in a cream, oil, lotion, spray, perfume, cologne, fragrance, deodorant, masking agent, candle, vaporizer, and the like.

The methods of the disclosure can also be used to identify food additives of flavorants.

The details of one or more embodiments of the disclosure are set forth in the accompanying drawings and the description below. Other features, objects, and advantages of the disclosure will be apparent from the description and drawings, and from the claims.

DESCRIPTION OF DRAWINGS

FIG. 1 shows a schematic of a method of the disclosure used to identify an optimized descriptor subsets for each Or.

FIG. 2 shows a variety of selection method combinations.

FIGS. 3A-3T show diagrams of compound activity classification through activity clustering. Compounds were clustered based on difference in activity. Compounds below certain squares, indicate cut points. Diagrams are provided for receptors: Or7a (FIG. 3A); Or9a (FIG. 3B); Or22a (FIG. 3C); Or35a (FIG. 3D); Or49b (FIG. 3E); Or59b (FIG. 3F); Or10a (FIG. 3G); Or19a (FIG. 3H); Or43b (FIG. 3I); Or47a (FIG. 3J); Or67a (FIG. 3K); Or67c (FIG. 3L); Or82a (FIG. 3M); Or85a (FIG. 3N); Or85b (FIG. 3O); Or98a (FIG. 3P); Or2a (FIG. 3Q); Or23a (FIG. 3R); Or43a (FIG. 3S); and Or85f (FIG. 3T).

FIG. 4 shows a schematic of selecting highest scoring optimization methods.

FIG. 5 is a graph comparing APoA values.

FIGS. 6A-6T show analyses of APoA for individual Odor receptors. Analyses are provided for receptors: Or7a (FIG. 6A); Or9a (FIG. 6B); Or22a (FIG. 6C); Or35a (FIG. 6D); Or10a (FIG. 6E); Or19a (FIG. 6F); Or43b (FIG. 6G); Or47a (FIG. 6H); Or49b (FIG. 6I); Or59b (FIG. 6J); Or82a (FIG. 6K); Or85a (FIG. 6L); Or67a (FIG. 6M); Or67c (FIG. 6N); Or85b (FIG. 6O); Or98a (FIG. 6P); Or2a (FIG. 6Q); Or23a (FIG. 6R); Or43a (FIG. 6S); and Or85f (FIG. 6T).

FIG. 7 shows a comparison of highest molecular descriptor APoA for each Or.

FIGS. 8A-8T show clustering of drosophila odorants by optimized descriptor subsets. Clustering diagrams are provided for receptors: Or7a (FIG. 8A); Or9a (FIG. 8B); Or22a (FIG. 8C); Or35a (FIG. 8D); Or49b (FIG. 8E); Or59b (FIG. 8F); Or10a (FIG. 8G); Or19a (FIG. 8H); Or43b (FIG. 8I); Or47a (FIG. 8J); Or67a (FIG. 8K); Or67c (FIG. 8L); Or82a (FIG. 8M); Or85a (FIG. 8N); Or85b (FIG. 8O); Or98a (FIG. 8P); Or2a (FIG. 8Q); Or23a (FIG. 8R); Or43a (FIG. 8S); and Or85f (FIG. 8T).

FIG. 9A shows a computational validation of Drosophila optimized descriptor sets.

FIG. 9B shows high-throughput flowchart for in silico screen of each Or with >240,000 compounds.

FIGS. 10A-10I show electrophysiology validations of drosophila in silico screen for the following receptors: Or7a (FIG. 10A); Or10a (FIG. 10B); Or22a (FIG. 10C); Or47a (FIG. 10D); Or49b (FIG. 10E); Or59b (FIG. 10F); Or85a (FIG. 10G); Or85b (FIG. 10H); and Or98a (FIG. 10I).

FIG. 11 shows an electrophysiology testing for drosophila “false negative” rates of prediction.

FIGS. 12A-12F show table 2, drosophila compounds tested for activity: Or2a-Or49b. Compounds tested for activity: Drosophila Or2a-Or49b. Chemical name, a 2-D structural image, and distance measure are listed for each tested compound. All distances are Euclidean and represent the distance between each compound and their closest known active by optimized descriptor values. Known active compounds from the training set are the top 12, 7, 13, 5, 9 and 3 compounds respectively in each column, predicted compounds that were validated as actives are appropriately boxed, inhibitors are appropriately boxed, and inactive compounds are boxed.

FIGS. 13A-13F show table 3 drosophila compounds tested for activity: Or59b-Or98a. Compounds tested for activity: Drosophila Or59b-Or98a. List of compounds that were tested using electrophysiology for each Or. Chemical name, a 2-D structural image, and distance measure are listed for each tested compound. All distances are Euclidean and represent the distance between each compound and their closest known active by optimized descriptor values. Known active compounds from the training set are the top 12, 7, 13, 5, 9 and 3 compounds respectively in each column, predicted compounds that were validated as actives are appropriately boxed, inhibitors are appropriately boxed, and inactive compounds are boxed.

FIG. 14 shows validation accuracy for predicted drosophila ligands.

FIGS. 15A-15C show ligand prediction from neuronal activity. FIG. 15A shows an optimized descriptor based cluster for Or42b. FIG. 15B shows an electrophysiology validation for various compounds. FIG. 15C is a 3-D structural image of a compound from the electrophysiology validation.

FIGS. 16A and 16B depict ligand prediction from narrowly tuned Ors. FIG. 16A shows flow diagrams for (i) predicting initial screening compounds from single strong ligand, (ii) identifying additional activators through electrophysiology validation, and (iii) predicting additional activating compounds from single strong ligand and validated activators. FIG. 16B shows an example for Or82a of (i) an all descriptor cluster, (ii) electrophysiology validation, and (iii) an optimized descriptor based cluster.

FIGS. 17A-17AK show clustering mammalian odorants by optimized descriptor subsets for MOR1-1 (FIG. 17A); MOR106-1 (FIG. 17B); MOR139-1 (FIG. 17C); MOR162-1 (FIG. 17D); MOF189-1 (FIG. 17E); MOR2-1 (FIG. 17F); MOR107-1 (FIG. 17G); MOR129-1 (FIG. 17H); MOR170-1 (FIG. 17I); MOR184-1 (FIG. 17J); MOR203-1 (FIG. 17K); MOR204-6 (FIG. 17L); MOR136-1 (FIG. 17M); MOR223-1 (FIG. 17N); MOR185-1 (FIG. 17O); MOR260-1 (FIG. 17P); MOR207-1 (FIG. 17Q); MOR273-1 (FIG. 17R); MOR250-1 (FIG. 17S); MOR256-17 (FIG. 17T); MOR261-1 (FIG. 17U); MOR268-1 (FIG. 17V); MOR277-1 (FIG. 17W); MOR30-1 (FIG. 17X); MOR258-1 (FIG. 17Y); MOR259-1 (FIG. 17Z); MOR271-1 (FIG. 17AA); MOR272-1 (FIG. 17AB); MOR33-1 (FIG. 17AC); MOR37-1 (FIG. 17AD); MOR40-1 (FIG. 17AE); MOR41-1 (FIG. 17AF); MOR5-1 (FIG. 17AG); OR1A1 (FIG. 17AH); OR2J2 (FIG. 17AI); OR2W1 (FIG. 17AJ); and OR5P3 (FIG. 17AK).

FIGS. 18A-18G show computational validation of mammalian OR compound clustering for MOR41-1 (FIG. 18A); MOR271-1 (FIG. 18B); MOR203-1 (FIG. 18C); MOR272-1 (FIG. 18D); MOR139-1 (FIG. 18E); OR1A1 (FIG. 18F); and OR2W1 (FIG. 18G).

FIG. 19 shows clustering CO2 neuron activating odorants from training set 1 by optimized descriptor subsets.

FIG. 20 shows clustering CO2 neuron activating odorants from training set 2 by optimized descriptor subsets.

FIGS. 21A-21W show accumulated percentage of actives and activity based cluster analysis. FIG. 21A shows a representative example for an Accumulated Percentage of Actives (APoA) calculation. Green box=active, grey box=inactive. To calculate APoA each active compound was iteratively used as a reference active. Compounds are sorted based upon their increasing descriptor based distance from reference active, and the APoA calculated for each of the other compounds as a ratio of the number of actives over the total number of compounds considered from the reference compound. This process was repeated using each active odorant as a reference active. Reference compound APoAs were averaged to a single mean APoA value. The higher the APoA value while considering a fixed number of nearest neighboring compounds, the greater the proportion of active compounds clustered together. FIG. 21B shows a plot of the mean APoA calculated values calculated using each molecular descriptor method, averaged across all 20 Ors for Dragon, Cerius2, MCS and Atom Pair. FIG. 21C shows coloured cells mark the method that clusters active ligands best as determined by the highest Area-Under-Curve (AUC) values. E=Euclidean, S=Spearmans coefficient, and T=Tanimoto coefficient. FIG. 21D-21S show compounds clustered based on activity of Or, such as Or7a (FIG. 21D); Or9a (FIG. 21E); Or22a (FIG. 21F); Or35a (FIG. 21G); Or49b (FIG. 21H); Or59b (FIG. 21I); Or10a (FIG. 21J); Or19a (FIG. 21K); Or43b (FIG. 21L); Or47a (FIG. 21M); Or67a (FIG. 21N); Or67c (FIG. 21O); Or82a (FIG. 21P); Or85a (FIG. 21Q); Or85b (FIG. 21R); and Or98a (FIG. 21S). Activity color scale is indicated. Branches marked with small green squares (either 1 or 2) were considered as actives. FIGS. 21T-21W show dependent cluster analysis for Ors that have only weak ligands as done in FIGS. 21D-21S, including Or2a (FIG. 21T); Or23a (FIG. 21U); Or43a (FIG. 21V); and Or85f (FIG. 21W).

FIG. 22 shows that vapor pressure possibly affects ligand-Odor receptor activation. Vapor pressures and activities (in spikes/sec) were plotted for validated odorant predictions. Compounds are divided into four classes based upon compound activity and vapor pressure values.

FIGS. 23A-23V show predicted breadth of tuning for collected compounds in Odorant receptors, such as Or23a (FIG. 23A); Or82a (FIG. 23B); Or49b (FIG. 23C); Or92a (FIG. 23D); Or42b (FIG. 23E); Or22a (FIG. 23F); Or59b (FIG. 23G); Or43b (FIG. 23H); Or85a (FIG. 23I); Or85f (FIG. 23J); Or19a (FIG. 23K); Or67c (FIG. 23L); Or85b (FIG. 23M); Or7a (FIG. 23N); Or2a (FIG. 23O); Or67a (FIG. 23P); Or43a (FIG. 23Q); Or35a (FIG. 23R); Or98a (FIG. 23S); Or9a (FIG. 23T); Or47a (FIG. 23U); and Or10a (FIG. 23V). Compounds from the collected compound library that have been catalogued as plant, human and total collected volatiles were ranked according to their relative distance from the compound with highest activity. Frequency distribution of compounds within the top 15% is plotted to generate predicted breadth of tuning curves. X-axes are in logarithmic scale.

DETAILED DESCRIPTION

As used herein and in the appended claims, the singular forms “a,” “and,” and “the” include plural referents unless the context clearly dictates otherwise. Thus, for example, reference to “an insect” includes a plurality of such insects and reference to “the compound” includes reference to one or more compounds, and so forth.

Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood to one of ordinary skill in the art to which this disclosure belongs. Although any methods and reagents similar or equivalent to those described herein can be used in the practice of the disclosed methods and compositions, the exemplary methods and materials are now described.

Also, the use of “or” means “and/or” unless stated otherwise. Similarly, “comprise,” “comprises,” “comprising” “include,” “includes,” and “including” are interchangeable and not intended to be limiting.

It is to be further understood that where descriptions of various embodiments use the term “comprising,” those skilled in the art would understand that in some specific instances, an embodiment can be alternatively described using language “consisting essentially of” or “consisting of.”

All publications mentioned herein are incorporated herein by reference in full for the purpose of describing and disclosing the methodologies, which are described in the publications, which might be used in connection with the description herein. The publications discussed above and throughout the text are provided solely for their disclosure prior to the filing date of the present application. Nothing herein is to be construed as an admission that the inventors are not entitled to antedate such disclosure by virtue of prior disclosure.

The methods of the disclosure allows intelligent and rapid screening of untested volatile chemical space by computationally identifying important characteristics shared between known active compounds. Also provided are compounds identified by the methods of the disclosure for use as insect repellents and attractants.

The olfactory system can detect and discriminate amongst an extremely large number of volatile compounds in the environment, and this is critical for important behaviors like finding food, finding mates, and avoiding predators. To detect this wide variety of volatiles, most organisms have evolved extremely large families of receptor genes that typically encode 7-transmembrane proteins that are expressed in the olfactory neurons. Little is known, however, about how small volatile molecules are detected and represented with high levels of specificity and sensitivity by the activities of odor receptor repertoires. The disclosure is able to greatly increase this understanding, and improve the ability to manipulate the olfactory based behavior of an organism. Additionally the computational method can be used to identify novel fragrances for individual odor receptors, which can have use in the fragrance, food, beverage, cleaning and other volatile chemical related industries.

Most blood feeding insects, including mosquitoes, sandflies, Testse flies, use olfactory cues to identify human hosts. This group of hematophagous insects can transmit a wide assortment of deadly human diseases that together cause more suffering and deaths globally than any other disease condition. Diseases transmitted by such insects include malaria, dengue fever, yellow fever, West Nile virus, filariasis, river blindness, epidemic polyarthritis, Leshmaniasis, trypanosomiasis, Japanese encephalitis, St. Louis Encephalitis amongst others.

Traditional vector control methods often involve the heavy use of chemical insecticides that are harmful to the environment and often to human health. Moreover, insects can develop resistance to these chemicals, suggesting that there is a need to identify novel ways of insect control that are effective, cheap, and environmentally friendly. Integrating methods that inhibit vector-human contact, such as vector control and the use of insect repellents, bednets, or traps, may play a complementary and critical role in controlling the spread of these deadly diseases.

In insects host-odor cues, among others, are detected by olfactory receptor neurons (ORNs) that are present on the surface of at least two types of olfactory organs, the antennae and the maxillary palps. The antenna is the main olfactory organ and its surface is covered by hundreds of sensilla, each of which is innervated by the dendrites of 1-5 ORNs. Odor molecules pass through pores on the surface of sensilla and activate odor receptor proteins present on the dendritic membranes of the ORNs.

The odor receptor (Or) gene family in insects was first identified in D. melanogaster. It comprises a highly divergent family of 60 Odor receptor (Or) genes that encode proteins predicted to contain seven trans-membrane regions.

One of the most important host-seeking cues for hematophagous insects is CO₂. The CO₂ receptor was first identified in D. melanogaster. This receptor comprises two proteins, Gr21a and Gr63a, which are encoded by two members of a large Gustatory receptor (Gr) gene family that is distantly related in sequence to the Or genes. Both Gr21a and Gr63a are extremely well conserved in sequence across several insect species. Orthologs for both Gr21a and Gr63a have been identified in An. gambiae and Ae. aegypti. Moreover, both mosquitoes possess a third gene that is closely related to Gr21a. The three An. gambiae homologs AgGr22, AgGr23 and AgGr24 are co-expressed in ORNs of the maxillary palp. Functional expression studies in Drosophila have demonstrated that they are CO₂ receptors as well.

Odor responses of ORNs on the surface of the antennae and maxillary palps have been studied using two separate techniques. Whole organ recordings called electroantennograms (EAGs) and electropalpograms (EPGs) have been used to detect the aggregate electrical activities from a large number of neurons in response to odors. A more sensitive and exact method has also been used to examine the functional properties of olfactory neurons within a single sensillum, and neurons that respond to behaviourally important ligands such as CO₂, ammonia, phenols, 1-octen-3-ol, lactic acid, and carboxylic acids have been identified.

Because mosquitoes rely on their sense of smell to identify human odors, olfactory system function is a prime target to modify host-seeking behaviour. The kairomone CO₂ is used as bait by several mosquito traps that are currently sold on the market. In some instances an additional odor, usually 1-octen-3-ol, is also included to increase the efficiency of mosquito catches. Identification of more potent attractant odors, or more efficacious odor blends are required to further improve the efficiency of these CO₂ traps. Development of cheap CO₂-free traps may be of particular importance since generating CO₂ in a trap is problematic.

In a complementary fashion, blocking of insect odor receptors may be effective in masking human hosts, or may even work as repellents. There has been a great interest to identify novel classes of volatile compounds that can block mosquito receptors that detect kairomones like CO₂.

Volatile chemical space is immense. Odors in the environment that have been catalogued in some plant sources alone number more than two thousand. A very small proportion of chemical space has been systematically tested for the ability to activate or inhibit individual odor receptors, and a very small fraction of odor receptors, whose sequences are known, have been tested for activity. The complete 3-D structures of odor receptor proteins have not yet been determined, thus modeling of odor-protein interactions is not yet possible except in rare instances. Furthermore, were a 3-D receptor structure to become available, application of one odor-receptor interaction to study others may be confounded by the possibility of multiple ligand binding sites in a single receptor, as well as the sequence divergence amongst different odor receptors.

Odor receptor responses to odorants have been tested in vivo in the organism of interest predominately through two separate techniques. One approach involves whole organ recordings called electroantennograms (EAGs), eletropalpograms (EPGs), and electroolfactograms (EOGs) which have been used to detect the aggregate electrical activities from a large number of olfactory neurons in response to odors. This technique does not allow for differentiation between odor receptor neuron responses and thus does not allow for identification of individual odor receptor responses to an odorant. A more sensitive and precise technique called single unit electrophysiology allows for individual odor receptor neuron responses to odors to be quantitatively measured. This technique either requires the odor receptor map to have been previously established by molecular tools or use of an “empty-neuron” system that utilizes a transgenic approach.

In Drosophila melanogaster a mutant antennal neuron called the “empty neuron” has been identified. The system uses a mutant strain of D. melanogaster in which a chromosomal deletion has resulted in the loss of the Or22a gene. The Or22a gene product is usually expressed in an easily identifiable and accessible neuron type in the antenna called ab3A, which now does not express an odor receptor and therefore does not respond to any odors. An exogenous Or gene can then be functionally expressed in this mutant “empty neuron” genetic background using the promoter of Or22a. Responses to a diverse set of odorants can be recorded using single-sensillum electrophysiology. Through iteratively inserting and testing Or genes, electrophysiological responses of 24 Ors to a preliminary set of 110 diverse compounds was determined, as well as 21 additional Or genes to a set of 27 compounds. The compound sets consisted of volatile compounds with varying functional groups and hydrocarbon chain lengths. It has also been demonstrated that expression of functional odor receptors from other organisms is possible in the Drosophila “empty neuron” system. The level of throughput of this system is ˜100s to 1000s of odors in one year.

Additionally, other in vivo techniques have been used involving testing individual odor receptors of interest through transgenic expression in other organisms. Heterologous expression of Odor receptor genes from many species has been performed in Xenopus oocytes and Human Embryonic Kidney (HEK) 293 cells. Exposure of these cells to volatile compounds allows for a quantitative measure of response.

While these systems do provide a means to specifically express an odor receptor and obtain a quantitative measure of activation to a panel of odorants, their use is a very time consuming, expensive, and difficult process. Use of the “empty neuron” system and other heterologous expression approaches require transgenic fly lines to be produced or cDNA expression constructs made for each odor receptor to be tested. It has also been debated whether these expression systems produce wild type responses in all cases, as some cell specific components such as odorant binding proteins (OBPs) may be absent. Additionally all systems require the requirement of purchasing odors, diluting them, and performing the technically challenging testing of odorants.

In previous studies, individual odor receptors have sometimes been found to recognize compounds of similar functional groups containing similar hydrocarbon chain lengths. In addition it has also been shown that many Ors can be responsive to multiple distinct groups of structurally similar compounds. This property of odor receptors recognizing structurally similar compounds provides a framework for using cheminformatic similarity measures to predict novel active odorants.

Molecular descriptors are able to describe the structure of molecules through computationally derived values, which represent zero, one, two, or three-dimensional information of a compound. These descriptor type dimensionalities confer molecular information through classes such as constitutional, structural fragment, topographic, or spatial information, respectively.

Comparison of molecular descriptors to identify commonalities between highly active odorant structures has recently proven to be highly beneficial. In species where a specific behaviour, such as avoidance, has been tested against a panel of odors it is possible to use molecular descriptors to identify novel potential ligands using the known actives as a training set. For instance, the structure of N,N-diethyl-m-toluamide (DEET) was recently used to create a focused structural library, which was computationally ranked using Artificial Neural Networks (ANNs), and used to identify a more potent mosquito repellent. In another study a group analyzed Drosophila ORN responses to odors to identify activation metrics that were used to predict and test ligands from a small set of 21 compounds (Schmuker et al., 2007). The success rate of this strategy, as established by applying a neuronal firing rate cut-off of 50 spikes/sec to categorize activators, was <25%. Most recently a multi species approach was used to identify molecular descriptors that were important in compounds involved in olfaction however predictions were not possible. In another study by the same lab, an electronic nose was trained such that when presented with a novel odor it could predict whether or not the odor would activate an individual Or.

The methods of the disclosure allows intelligent and rapid screening of untested volatile chemical space and chemical libraries by computationally identifying important characteristics shared between known active compounds, circumventing many of the previously described obstacles.

The disclosure provides a chemical informatics method that identifies important structural features shared by ligands such as activating odors for individual odor receptors or olfactory neurons and utilizes these important features to screen large libraries of compounds in silico for novel ligands. These important structural features can also be used to increase understanding of breadth of tuning for each cognate of a ligand such as an odor receptor in chemical space and perform reverse chemical ecology in silico.

Although the methods of the disclosure have been exemplified using odor receptor and volatile chemical species. The method is also predicatable to taste receptors, g-protein coupled receptors, ion gated channels, ligand gated channels and the like.

The disclosure provides methods for identifying and the identified compositions comprising volatile odorants that can modulate the electrophysiological response of neuron in various insect disease vectors including Drosophila melanogaster, Culex quinquefasciatus, An. gambiae and Aedes aegypti mosquitoes. In some embodiment, the odorants can completely inhibit the electrophysiological response of the neuron at very low concentrations.

The odorants of the disclosure provide new and useful compositions for insect repellents, masking agents and traps. The compounds of the disclosure are useful in small quantities, can be delivered in multiple forms like vapors and lotions, are economical, environmentally friendly, and are present in natural sources.

Based upon the data and chemical odorants identified herein, additional odorants can be identified using the structural information of the odorants, in silico modeling and screening and biological assays.

The disclosure provides a group of volatile chemicals that can be used to modify host-seeking behaviour by stimulating or inhibiting odor and taste receptors.

The compounds and compositions of the disclosure can be used as antagonist to mask the chemo attractant activity for a particular odor receptor. Alternatively, the certain compounds may at as agonist in which they activate the receptor and stimulate the neuron. In such instances the compounds and compositions can be used as attractants alone or in combination with other materials depending upon the subject and purpose (e.g. an insecticide, trap, or other mechanical, electrical or chemical that kills the insect or prevents its escape).

An antagonist refers to a compound the can reversibly or irreversibly inhibit that activity of a sensing neuron upon exposure to the compound such that the neuron ORN cannot properly signal upon a change in odor levels.

Structure-based clustering can be used to identify compounds useful in compositions of the disclosure. The algorithm can include linkage clustering to join compounds into similarity groups, where every member in a cluster shares with at least one other member a similarity value above a user-specified threshold.

The identified compounds can then be assayed to identify their biological activity using the electrophysiology measurements described below. For example, a compound can be contacted with a CO2 receptor neuron and changes in the electrical signal measured. Alternatively, the compounds may be screened in a Drosophila CO₂ avoidance chamber.

The disclosure provides chemicals that can be used as insect repellents and/or masking agents by virtue of their property to block a critical component of the host odor cue. The compounds are effective if they are capable of inhibiting the electrophysiological response of the neuron.

The volatile compounds of the disclosure have masking and repellant effects by impairing the ability to find a host via long-range cues emitted from a typical target or subject (e.g., human breath).

The disclosure provides a method of controlling insect attraction to a subject, the method comprising the step of inhibiting receptor activation (e.g., CO₂ sensing gustatory receptors) in the insect or over stimulating the receptor with an antagonist (or a combination of antagonists).

In another embodiment, this disclosure provides a method of inhibiting, preventing or reducing the incidence of insect-borne disease in a subject, the method comprising the step of over stimulating or antagonizing a receptor in an insect with a compounds or combination of compounds, wherein the receptor response is modified and attraction to the subject inhibited, thereby inhibiting, preventing or reducing the incidence of insect-borne disease in a subject.

In one embodiment, the disease is malaria, dengue, yellow fever, river blindness, lymphatic filariasis, sleeping sickness, leishmaniasis, epidemic polyarthritis, West Nile virus disease or Australian encephalitis.

The compounds may be used alone or in combination with other agents. The compounds of the disclosure may be combined with additional active agent, insecticides and the like in traps to reduce the presence of amount of an insect in the environment. For example, compounds of the disclosure may be used in combination with insect traps (e.g., tape, combustibles, electric traps).

In yet a further embodiment, the compounds may be formulated for application to the skin, clothing or other material. The compounds of the disclosure can “mask” the location of a subject by antagonizing the receptor neurons of an insect etc. thereby inhibiting the ability to locate a prey.

For example, the compounds of the disclosure may be used as repellents or in compositions comprising said repellent compounds and the use of such repellent compounds and compositions in controlling pests.

Liquid formulations may be aqueous-based or non-aqueous (e.g., organic solvents), or combinations thereof, and may be employed as lotions, foams, gels, suspensions, emulsions, microemulsions or emulsifiable concentrates or the like. The formulations may be designed to be slowly release from a patch or canister.

The compositions may comprise various combinations of compounds as well as varying concentrations of the compound depending upon the insect to be repelled or masked, the type of surface that the composition will be applied to, or the type of trap to be used. Typically the active ingredient compound of the disclosure will be present in the composition in a concentration of at least about 0.0001% by weight and may be 10, 50, 99 or 100% by weight of the total composition. The repellent carrier may be from 0.1% to 99.9999% by weight of the total composition. The dry formulations will have from about 0.0001-95% by weight of the pesticide while the liquid formulations will generally have from about 0.0001-60% by weight of the solids in the liquid phase.

As mentioned above, the compositions may be formulated for administration to a subject. Such formulations are typically administered to a subject's skin. The composition may also be formulated for administration to garments, belts, collars, or other articles worn or used by the subject from whom insects are to be repelled. The formulation may be applied to bedding, netting, screens, camping gear and the like. It will be recognized that the application of the compositions and compounds of the disclosure do not only include human subjects, but include canines, equines, bovines and other animals subject to biting insects. For topical application, the formulation may take the form of a spray formulation or a lotion formulation.

The compounds according to the disclosure may be employed alone or in mixtures with one another and/or with such solid and/or liquid dispersible carrier vehicles as described herein or as otherwise known in the art, and/or with other known compatible active agents, including, for example, insecticides, acaricides, rodenticides, fungicides, bactericides, nematocides, herbicides, fertilizers, growth-regulating agents, and the like, if desired, in the form of particular dosage preparations for specific application made therefrom, such as solutions, emulsions, suspensions, powders, pastes, and granules as described herein or as otherwise known in the art which are thus ready for use.

The repellent compounds may be administered with other insect control chemicals, for example, the compositions of the invention may employ various chemicals that affect insect behaviour, such as insecticides, attractants and/or repellents, or as otherwise known in the art. The repellent compounds may also be administered with chemosterilants.

In yet another aspect, the volatile compounds of the disclosure may be emitted from vaporizers, treated mats, cylinders, oils, candles, wicked apparatus, fans and the like. Liquid source that can evaporate to form vapors may be used in barns, houses, or patios.

The disclosure also provides chemicals that can be used as bait to lure insects to traps by virtue of activating neurons. An advantage of these odorants will be their ability to be delivered in an economical and convenient form for use with traps. This function can be achieved by applying or locating the chemotractant compound of the disclosure near a suction based, or light based, or electric current based or other forms of trapping apparatus.

The disclosure provides a structural basis of odorant molecule interaction with odor receptors through a novel chemical informatics platform. The disclosure provides a method to identify molecular structural properties that are shared between the activating odorants (actives) for an individual odor receptor. By identifying the molecular features shared by actives, the disclosure provides a system to perform in silico screens of large chemical space (100s of thousands to millions) to predict novel ligands for odor receptors or odor receptor neurons. This method can be applied in virtually any species where a training set of odorant responses is known for individual receptor or cellular level. The disclosure demonstrates this using a single unit electrophysiology to test a subset of the predictions in vivo. The data demonstrate that the method is very successful in predicting novel ligands.

The disclosure demonstrates the method can be modified to be able to predict ligands for narrowly-tuned receptors and neurons that are thought to be highly specialized, like pheromone receptors. In addition olfactory neurons whose response profiles are known, but whose odor receptors have not yet been decoded are provided. The method is also able to predict odorant ligands for two distinctly different classes of odor receptors. Insect odor receptors are proposed to be 7 transmembrane GPCR like proteins with inverse orientation in the membrane that function as either heteromeric ligand gated ion channels or cyclic-nucleotide activated cation channels. Mammalian odor receptors on the other hand are true GPCRs. The method is able to predict ligands for both insect and mammalian odor receptor classes. In addition to predicting ligands the disclosure also allows investigation of the coding of each tested receptor or receptor neuron in chemical space consisting of plant volatiles, fragrances and human volatiles.

The CO₂ receptor is believed to be very important in host seeking behaviour in mosquitoes. There are several commercially available approaches that use CO₂ as a lure to trap insects. However, these current approaches have several drawbacks. Many traps require the use of a CO₂ tank or dry ice to produce the CO₂ lure plume. These CO₂ tanks are large and heavy, making the trap itself cumbersome. Dry ice melts quickly and must be replaced often. A much smaller and longer lasting trapping approach would be advantageous. Identification of odors that could specifically activate this receptor could provide a very effective means of luring mosquitoes into traps. The approach can be used to identify odors that activate individual receptors, such as the CO₂ receptor.

Since different odor receptors can respond to vastly differing compound shapes and sizes it is unlikely that the full collection of molecular descriptors would be optimal for all receptors. Depending upon the unique structural features of active odors certain molecular descriptors may be better suited at describing characteristics of activating compounds for an individual receptor, and such descriptors can be identified from much larger sets by dimensionality reduction. Thus it is possible to greatly improve Or-specific descriptor space by identifying specific molecular descriptors from amongst the large collection that were best suited for each Or.

The disclosure provides a method of computationally screening a vast number of compounds to predict ligands (activators or inhibitors) for individual receptors or receptor expressing cells, wherein a known ligand or set of known ligands for a receptor or receptor expressing cell, either identified through electrophysiology, imaging assays, or binding assays, are used as a training set for selecting optimized molecular descriptors, which can subsequently be used to screen a large collection of untested compounds computationally to identify compounds that are structurally related to the known ligands, outputting the identified putative ligands to a user and exposing a receptor or receptor expressing cell to the putative ligand and determining either a change in spike frequency, florescence intensity, or binding affinity in the receptor or receptor expressing cell, wherein a change compared to baseline is indicative of a ligand for the receptor or receptor expressing cell.

The disclosure also provides a method of computationally screening a vast number of compounds to predict ligands (activators or inhibitors) for individual receptors or receptor expressing cells that have only one known strong activator or inhibitor, either identified through electrophysiology, imaging assays or binding assays, wherein a single known ligand from a receptor or receptor expressing cell is used to identify the structurally closest compounds in a chemical space made using several or all available structural descriptors, outputting the identified putative ligands to a user and exposing a receptor or receptor expressing cell to the putative ligand and determining either a change in spike frequency, florescence intensity, or binding affinity in the receptor or receptor expressing cell, wherein a change compared to baseline is indicative of a ligand for the receptor or receptor expressing neuron. In one embodiment, positives having a desired functional activity are used to further define the structural descriptors along with previously known activating odorants.

The disclosure also provides a method of computationally screening a vast number of compounds to predict compounds which cause a specific behavior (attraction, repellency, mating, aggression, or oviposition), wherein an compound or set of known compounds causing a specific behavior are used as a training set for selecting optimized molecular descriptors, which can subsequently be used to screen a large collection of untested odorants computationally to identify compounds that are structurally related to the known behavior modifying compounds, outputting the identified putative behavior modifying compounds to a user and testing the compounds for behavior modification, wherein a change compared to baseline behavior is indicative of a behavior modifying compound. In various embodiments, compounds are volatile odors and either the receptor is an odor receptor expressed by a specific neuron or cell type in a specific invertebrate species or receptor-expressing cells are odor receptor neurons present in a specific species of invertebrate.

In other embodiment, compounds are soluble ligands and either the receptor is a gustatory receptor expressed by a specific neuron or cell type in a specific invertebrate species or receptor-expressing cells are gustatory receptor neurons present in a specific species of invertebrate. In yet other embodiments, the compounds are volatile ligands and either the receptor is a gustatory receptor expressed by a specific neuron or cell type in a specific invertebrate species or receptor-expressing cells are gustatory receptor neurons present in a specific species of invertebrate. In further embodiments, the compounds are volatile odors and either the receptor is an odor receptor expressed by a specific neuron or cell type in a specific vertebrate species or receptor-expressing cells are odor receptor neurons present in a specific species of mammals. In some embodiments, the compounds are soluble ligands of volatile ligands and either the receptor is a gustatory receptor expressed by a specific neuron or cell type in a specific vertebrate species or receptor-expressing cells are gustatory receptor neurons present in a specific species of mammals.

As mentioned above, the methods of the disclosure can be used to screen ligands for a number of different biological molecules including GPCR. Accordingly, in one embodiment, the compounds are soluble or volatile ligands and either the receptor is a GPCR expressed by a specific neuron or cell type in a specific invertebrate or vertebrate species or receptor-expressing cells are GPCR expressing cells present in a specific species of invertebrate or vertebrate.

In yet other embodiment, the compounds are identified by the method of the disclosure and are identified as compounds for ligand gated ion channels. For example, the compounds can be soluble or volatile ligand and either the receptor is a ligand gated ion channel expressed by a specific neuron or cell type in a specific invertebrate or vertebrate species or receptor-expressing cells are ligand gated ion channel expressing cells present in a specific species of invertebrate or vertebrate.

The disclosure provides a method of identifying a ligand for a biological molecule comprising (a) identifying a known ligand or set of known ligands for a biological molecule, or identifying a compound which causes a specific biological activity, (b) identifying a plurality of descriptors for the known ligand or compound, (c) using a Sequential Forward Selection (SFS) descriptor selection algorithm to incrementally create a unique optimized descriptor subsets from the plurality of descriptors for the known ligand or compound, (d) identifying a putative ligand or compound that best-fits the unique optimized descriptor subset, and (e) testing the putative ligand or compound in a biological assay comprising the biological molecule wherein a change in activity of the biological molecule compared to the molecule without the putative ligand is indicative of a ligand the interacts with the biological molecule.

The disclosure utilizes in one embodiment a Sequential Forward Selection (SFS) descriptor selection method to incrementally create unique optimized descriptor subsets for each odor receptor. For example, starting with the combined group of 3424 descriptors from the full sets of Dragon and Cerius2 descriptors, an initial descriptor was selected whose values for the 109 odors showed the greatest correlation with activity for a specific Or. Additional descriptors were incrementally added to the growing optimized descriptor set based on their ability to further increase the Pearson correlation with activity for a specific Or. Each iteration increased the size of the optimized descriptor set for that Or by one. When a round of descriptor selection failed to increase the correlation between compound distance based upon the descriptor sets and those based upon known compound activity, the selection process was halted. As a result, optimized descriptor sets and their sizes are expected to vary across Ors. Additionally, 6 selection method combinations were used to identify the best statistical method for determining descriptor inclusion in the optimized set (FIG. 2).

In order to identify a method to select optimized descriptors for each Or the method was applied to 18 combinations of distance metrics, descriptor sets, and activity thresholds. Distance metrics included Euclidean, Spearman, and Pearson coefficients. Descriptor sets included Dragon, Cerius2, and a combined Dragon/Cerius2 set from which optimized descriptors would be chosen. Two activity threshold methods were compared for each combination. First, the four (>200, >150, >100, and >50 spikes/second) activity cut-offs were used. Second, a cluster based cut-off method was used to determine actives. For this approach a cluster analysis of the 109 odors for each individual Or was used using compound activity to calculate distances between Ors. The resulting activity trees for each Or were inspected, and active compounds were classified by selecting either one or two branches containing the active clusters (FIGS. 3A-3T).

Compounds are then clustered based on differences in activity. Compounds falling below a cut point are classified as active. Cut point locations can be determined manually. For example, each of the 3 distance metrics (FIG. 2) were applied to the 6 descriptor subsets (FIG. 1) to produce 18 unique descriptor based odor relationship sets. Accumulative Percentage of Actives (APoAs) values were calculated from distances between compounds based on each of the 18 methods and compared by AUC values as has been described previously. The highest-scoring selection method and the resulting optimized molecular descriptor set were identified for each Or.

If the optimized descriptor sets are better than the large collections of non-optimized descriptors, then one would find that they are able to cluster known active ligands closer together in chemical space. In order to determine whether the optimized descriptor sets are better at bringing the active compounds closer together in chemical space 4 non-optimized descriptor methods including Dragon, Cerius2, Maximum Common Substructure (MCS), Atom Pair (AP), were compared to a “selected” descriptor set from a published study that was selected for activation of the olfactory system by all 20 Drosophila Ors and across multiple species. The averaged APoA values for each of the 6 descriptor sets (Or-optimized, all Dragon, all Cerius2, Atom-pair, MCS, previous study) were compared for each of the 20 Ors and the Or-optimized descriptor sets provided APoA values far greater than all other methods, across all numbers of nearest neighbours.

FIGS. 6A-6T show analyses of APoA for individual Odor receptors. Plots of the mean APoA values obtained from various Molecular Descriptor methods demonstrates that optimized descriptor subsets generate highest values. Molecular descriptor methods were compared using 109 compounds.

The highest-scoring selection method and the resulting optimized molecular descriptor set were identified for each Or. Selection method 5 followed by 11, which proved to work the best by virtue of having the highest AUC scores when considered at an individual Or level, used the combined Dragon+Cerius2 descriptor set, activity-cluster threshold method, and either Euclidean distance Or Spearman correlation as a similarity metric. Euclidean distance provided the highest AUC values for 18 of the Ors and Spearman for 2.

To better visualize how well each Or-optimized descriptor set grouped active ligands, the compounds can be clustered by distances calculated using the optimized descriptor sets for each Or. For example, the 109 compounds were clustered by distances calculated using the optimized descriptor sets for each Or. As expected from the APoA values, highly active ligands are seen tightly clustered for each Or. There were some differences in the ability to cluster actives with Or7a, Or9a, Or10a, Or22a, Or35a, Or43b, Or47a, Or59b, Or67a, Or85a, Or85b and Or98a providing the best clusters, while Or2a, Or23a, Or43a and Or85f did not provide as tight a clustering as predicted. A correlation can be observed between APoA values and the number of highly active compounds grouped tightly together by descriptors. The simplest interpretation of these results is that the Or-descriptor selection method and resulting optimized descriptor sets are considerably better at clustering activating odors than previously tested sets.

The poorer performance of Or2a, Or23a, Or43a and Or85f was expected since of the 109 odorants that were tested, very few showed any activity. The simplest interpretation is that “true” ligands for these 4 receptors have not been discovered from within the tested panel. However, the few odors that poorly activate each of these 4 receptors do cluster together in chemical space after identification of Or-optimized sets, albeit not as well as the ones with known strong ligands. This indicates that the Or-descriptor selection method was able to identify common features amongst the weakly activating odors and hence cluster them together, suggesting the possibility that stronger ligands may be identified from a larger chemical space using this information. From this point onwards these 4 Ors are referred to as “Semi-orphan” Ors.

Using the principles above, an in silico method of compound identification and clustering was used to characterize potential receptor ligands. Since the Or-optimized descriptors can group highly active compounds tightly together in chemical space for each Or, this method can be used to rank untested compounds according to their distance from known actives. This allowed us to computationally screen a vast area of chemical space of potential volatiles in a very efficient and accurate manner. In total close to 5,000,000 interactions were systematically tested between 20 Ors and >240,000 different putative volatile compounds. This would be entirely unfeasible using current assay technology. With electrophysiology the largest screen so far has tested <3000 different interactions, which is ˜0.06% the size of the in silico screen. Moreover traditional high-throughput plate-based assays, as used for GPCRs that detect ligands in solution, are not appropriate for odor receptors since volatile ligands are largely (if not completely) absent from soluble plate-based combinatorial chemical libraries available.

A large collection of potential volatile compounds were identified by using criteria from known odors, such as molecular weight>200 and atom types limited to C, O, N, S, and H. Using these criteria over 240,000 compounds were selected from Pubchem and their structures were obtained. The distances in chemical space was then calculated for each of the >240,000 compounds based on the Or-optimized descriptor sets for each of the 20 Ors. In this fashion the unknown compounds were sorted by distances from each of the compounds considered as active from the 109 tested compounds. Euclidean distance or Spearman correlation, depending on which had previously been determined to be optimal for the corresponding Or, was used as similarity measures. Using this system the untested compounds in the 240,000 compound library were ranked according to their closeness to the known active ligands. The top 500 (0.2%) of hits in this large chemical space for each Or is listed below. Since each Or-optimized descriptor set was unique, unknowns were ranked independently for each receptor. Compounds were ranked systematically as actives for each of the 20 Ors using the Or-optimized descriptor sets and similarity measures to computationally rank all 240,000 compounds. These predictions could prove to be extremely valuable, not only do they provide an incredibly rich array of information regarding the coding of information by the peripheral olfactory system, it also provides an extremely large number of putative novel ligands for each of these 20 Or genes in Drosophila.

The results of the in silico screen are provocative. However in order to verify whether these predictions were meaningful, functional evidence was obtained. In order to validate the success of the in silico predictions the responses of 9 Odor receptors using single-sensillum electrophysiology directly on the Drosophila melanogaster antenna were analyzed. For each Odor receptor several odorants were tested from the top 500 predicted hits. A sampling of ˜192 novel odorants were tested with ˜11-21 novel odorants tested for each receptor, which were scattered somewhat randomly within the top 500 predictions for each receptor, providing a relatively unbiased set of chemical structures.

To test identified compounds any number of biological assays can be used to measure ORN activity in the presence of a putative ligand/compounds. For example, to demonstrate the activity of the compounds identified above, a single-unit electrophysiology test was performed on D. melanogaster antenna for each predicted compound, resulting in a quantitative value of activation. For the purpose of testing each of these volatile compounds the compounds were diluted to ˜10⁻² in paraffin oil or distilled water. The 9 Ors tested are expressed in well-defined olfactory receptor neurons (ORNs) housed within the large and small basiconic sensilla (ab1-ab7) on the antenna. A previously identified diagnostic panel of odorants was used to distinguish individual classes of sensilla (ab1-ab7) and therefore identified the sensilla that contained the target Or expressing ORN.

FIGS. 10A-10I show the firing rates of odorants that were not predicted to be actives were tested using single unit electrophysiology. This demonstrates the specificity of the invention. Bars indicate the strength of response (spikes/s). All values have been corrected for spontaneous firing rate. Spontaneous activity of neuron was subtracted. All odorants were tested at a concentration of 10⁻². N=3. Error bars=s.e.m.

FIGS. 11 and 12A-12F provide a list of exemplary compounds. Chemical name, a 2-D structural image, and distance measure are listed for each tested compound. All distances are Euclidean and represent the distance between each compound and their closest known active by optimized descriptor values. Known active compounds from the training set are in yellow boxes, predicted compounds that were validated as actives are green, inhibitors are red, and inactive compounds are white.

As can be seen a majority of the predicted actives evoked responses from the target ORNs; ˜71% evoked either activation (>50 spikes/sec above the spontaneous activity) or inhibition (>50% reduction in spontaneous activity). The success rates for different Ors varied from 100% for Or98a, to 27% for Or49b. Extrapolation of these values to the entire in silico screen suggests that between 500 and 135 novel ligands were identified for each of the 19 Ors.

The data demonstrate that >61% of the predicted compounds elicited >50 spikes per second, and >40% evoked strong responses of >100 spikes per second. In a few instances volatiles were identified that could activate the odor receptors extremely strongly (>250 spikes/sec); e.g. isopropyl acetate (Or59b, ab2A) and prenyl acetate (Or98a, ab7A). (see, e.g., FIG. 14).

The top 500 out of 240,000 compounds are an arbitrarily selected criteria and it is possible that compounds beyond the top 500 may also activate the receptors. Further examples were tested using two receptors Or22a and Or85b to extend the analysis to the top 1000 compounds. An additional 4 compounds were selected that are ranked between 500-1000 in the predictions and tested them using electrophysiology. Approximately 100% of these compounds were ligands, suggesting that the total number of new ligands identified by using the top 500 cut-off is underestimated.

Taken together these results demonstrate that the Or-optimized descriptor set based in silico screening of chemical space is extremely efficient at identifying volatile ligands for odor receptors.

The disclosure provides a chemical informatics method that identifies important structural features shared by activating odors for individual odor receptors or olfactory neurons and utilizes these important features to screen large libraries of compounds in silico for novel ligands. These important structural features can also be used to increase understanding of breadth of tuning for each Or in chemical space and perform reverse chemical ecology in silico.

The examples are illustrative. It will be recognized the use of specific odor receptors in the examples below can be substituted with any biological molecule that is capable or binds to a cognate/ligand. Such ligands can be small or large molecule organic molecules. The tables below are also illustrative. Each molecule in the table can be used independently in formulations, compositions or devices or may be used in combination. To described each and every combination would be redundant to the general descriptions herein and one of skill in the art will recognize that the various individual compositions, the various receptors can be utilized by the methods and compositions of the disclosure.

The following examples are intended to illustrate but not limit the disclosure. While they are typical of those that might be used, other procedures known to those skilled in the art may alternatively be used.

Examples

Chemical Informatics

Maximum Common Substructures, Atom Pairs, Cerius2 (Accelerys), Dragon (Talete) were used to compute distances. Energy minimized 3-D structures for Dragon were generated using Omega2 software (OpenEye). Optimized descriptor subsets were identified based on the correlation between descriptor distances with the distances between compounds based upon activity. The process is iteratively used to search for additional descriptors leading to further increases in correlation and stopped when increase stops.

Actives were classified either by thresholds of (>200, >150, >100, and >50 spikes/second), or using cluster analysis of receptor activity to compounds to select branch with strongest actives. The Accumulative Percentage of Actives (APoA) calculated for each descriptor set individually using a method used previously. The Area Under the Curve (AUC) scores from APoA values for each of the combinations were calculated by approximation of the integral under each plotted APoA line.

For each Odor receptor, the “optimized descriptor set” was used to calculate a distance metric that could be used for rank 240,000 compounds according to their closest distance to each known active compound. Compound distances were converted into a relative percentage distances based on the maximum possible compound distance for each Or individually.

Cluster Analysis of Ors.

Euclidean distance matrixes were used to create clusters using hierarchical clustering and complete linkage for three cases. The first 20 descriptors selected for each Or were used to create an identity matrix. The top 500 predicted compounds were used to create an identity matrix for all Ors. The responses of each of the Ors to a panel of 109 compounds⁶ were converted into an Or-by-Or Euclidean distance matrix.

Calculation of Descriptors.

Commercially available software packages Cerius2 (200 individual descriptors) and Dragon (3224 individual descriptors) from Accelerys and Talete were used to calculate molecular descriptors. Prior to inputting compounds into Dragon, 3-Dimensional structures were predicted for compounds through use of the Omega2 software. Descriptor values were normalized across compounds to standard scores by subtracting the mean value for each descriptor type and dividing by the standard deviation. Molecular descriptors that did not show variation across all compounds were removed. Maximum Common Substructures were determined using an existing algorithm. Atom Pairs were computed from the version implemented in Chemmine®.

Classification of Active Compounds.

In Drosophila actives were classified using two methods. In method one four different thresholds were based on the activation of action potentials by the compounds on the odour receptor (>200, >150, >100, and >50 spikes/second) as done in the electrophysiology study. For each odour receptor, APoA values were calculated using odorants falling within each of the four thresholds. The average APoA values for each threshold were then averaged, providing a relatively unbiased representation for which method best brought active odours closer together. In the second method cluster analysis was performed for the 109 compounds for each receptor based on activity in spikes/sec. Active compounds present in a single branch, or two branches, were selected manually as actives.

In mammals actives were classified through cluster analysis. EC₅₀ values obtained were converted to positive values by subtraction from 0 and used directly as measures of compound activity. Converted values ranged from 0 (inactive) to 7.242 (Strongest Activator). Activating compounds for each receptor were clustered by distances in activity. Active compounds present in a single branch, or two branches, were selected manually as actives.

Determination of Optimized Drosophila Descriptor Subsets.

A compound-by-compound activity distance matrix was calculated from activity data available for each of the Ors that have been tested for activity to 109 odours. Separate 3424 compound-by-compound descriptor distance matrices were calculated using values from Dragon and Cerius2. Active compounds for each Or were identified individually through activity thresholds. The correlation between the compound-by-compound activity and compound-by-compound descriptor distance matrices were compared for each actively classified compound, considering their distances to all other compounds. The goal was to identify the descriptor that calculates distance between compounds that most closely correlates with the distances between compounds based upon activity. The descriptor that correlates best is retained and the process iteratively used to search for additional descriptors leading to further increases in correlation. In this manner the size of the optimized descriptor set increases by one in each iteration as the best descriptor set from the previous step is combined with all possible descriptors to find the next best descriptor. This process is halted when all possible descriptor additions in iteration fail to improve the correlation value from the previous step. This whole process is repeated once for each Or resulting in unique descriptor sets that are optimized for each Or.

Determination of Optimized Mammalian Descriptor Subsets.

Mammalian descriptor set optimization was performed the same as for drosophila. The only difference for mammalian is that actives were classified only by cluster analysis.

Calculation of Accumulative Percentage of Actives (APoA). The accumulative percentage of actives is calculated for each descriptor set individually using a method used previously. The “optimized descriptor set” for a given odour receptor is used to calculate distances (Euclidean or Spearman) between the 109 compounds of known activity and the compounds are ranked according to their distance from each known active, resulting in one set of ranked compound distances for each active. Moving down the list for each of these rankings, ratios are calculated for the number of active compounds observed divided by the total number of compounds inspected, or the APoA. APoA values are averaged across all active compound rankings, creating a single set of mean values representing the APoA for a single Or and descriptor set. Using this approach ApoA mean values are calculated for each of the 24 Odour receptors, separately for each of the descriptor sets used, optimized set, all Dragon, all Cerius2, Atom Pair, Maximum Common Substructure. The Area Under the Curve (AUC) scores from APoA values for each of the combinations were calculated by approximation of the integral under each plotted APoA line.

Ranking Untested Putative Volatile Compounds.

A large collection of >240,000 untested compound structures were obtained from Pubchem using the following criteria. Compounds had molecular weights between 32 and 200 and were limited to H, C, N, O, or S atom types. Compound structures were converted into 3-Dimensional models using Omega2. Cerius2 and Dragon descriptors were calculated for each compound followed by the standard normalization of values through subtraction of the mean and division by standard deviation. For each Odour receptor, the previously determined “optimized descriptor set” was used to calculate a distance metric that could be used for ranking. The known active compounds for each Or were used individually to rank the set of greater than 240,000 compounds according to their closest distance to each known active compound, resulting in a matrix of dimensions 240,000 by the number of actives for the particular Or. Using this matrix each of the 240,000 compound structures were ranked according to their closest distance to any known active compound.

Clustering Ors by Most Common Descriptors.

The first 20 descriptors selected by the optimized descriptor selection algorithm for each Or were used to create an identity matrix. Each row representing an Or and column a specific descriptor. Ors that share common descriptors contain is in the same column. This matrix was then converted into an Or by Or Euclidean distance matrix and clustered using hierarchical clustering and complete linkage.

Clustering Compounds by Activity of or.

The responses of each of the Ors that had previously been tested against a panel of compounds were converted into an Or-by-Or Euclidean distance matrix. Ors were clustered using hierarchical clustering and complete linkage. Specifically, this was achieved by creating a compound-by-compound distance matrix using the differences in activity between compounds tested on a single Or. Hierarchical clustering using each Or distance matrix and then manually identifying the sub cluster which contained the most compact group of highly active compounds resulted in each Or's actively classified compounds.

Calculation of Pharmacophores.

Pharmacophore calculation was performed by Ligand Scout. Tightly clustering validated compounds for each Drosophila Or were aligned by shared pharmacophore features.

Clustering Ors by Predicted Ligand Space.

Percentages of overlapping predictions within the top 500 predicted compounds were calculated pair-wise for all Ors. Euclidean distances were calculated from the similarity between Ors

Calculation of or Tuning Using Pubchem and Collected Datasets.

Initially all extreme outliers were manually removed from the dataset for each Or. On average 5.82 compounds were manually removed for each Or resulting in a mean dataset reduction of 0.0024%. Next all compounds whose distance was greater than 3 standard deviations from the strongest activating compound were removed to reduce outliers. Distance-densities were produced for each Or. The large majority of these densities follow a Gaussian distribution with the exception of Or10, which appears bimodal. All remaining compound distances were converted into a relative percentage distances based on the maximum possible compound distance for each Or individually. The numbers of compounds within the top 15 percent of relative distance were plotted on a logarithmic scale for each Or to generate computationally derived tuning curves. The same maximum distance value for each Or was also used to calculate and plot the top 15 percent of collected compound relative distance.

Collected Volatile Compound Library.

A subset of 3197 volatile compounds were assembled from acknowledged origins including plants, humans, and a fragrance collection (Sigma flavours & fragrances, 2003 and 2007) that may have additional fruit and floral volatiles.

Calculation of Breadth of or Tuning Across Datasets.

From each of the three datasets (Hallem, Collected, Collected+Pubchem) an Or by Compound binary identity matrix was created. For the Hallem plot all compounds known to activate at least one Or at greater than 50 spikes/sec and any Or for which at least one activating compound was known were considered. Using these criteria the identity matrix was created and filled for each case of Or activation. For both the Collected and Pubchem+Collected datasets the top 500 predicted compounds for each Or for which predictions were made were used to fill binary identity matrices. All matrices were sorted in decreasing order of the percent of either known or predicted cross activation and plotted.

Computational Validation of Drosophila Optimized Descriptor Sets.

A 5-fold cross-validation was performed by dividing the dataset into 5 equal sized partitions containing roughly 22 compounds each. During each run, one of the partitions is selected for testing, and the remaining 4 sets are used for training. The training process is repeated 5 times with each unique odorant set being used as the test set exactly once. For each training iteration a unique set of descriptors was calculated from the training compound set. These descriptors were then used to calculate minimum distances from the test set compounds to the closest active exactly as used to predict ligands in a ligand discovery pipeline. Once test set compounds have been ranked by distance from closest to furthest to a known active in the training set, a receiver operating characteristics (ROC) analysis is used to analyze the performance of the computational ligand prediction approach. This analysis was performed on 12 Ors that were activated strongly by at least five odors (>100 spikes/sec) and very strongly by at least one odor (>150 spikes/sec) and were considered to have sufficient known ligands for this type of validation (Or7a, Or9a, Or10a, Or22a, Or35a, Or43b, Or47a, Or59b, Or67a, Or67c, Or85b, Or98a). A single average ROC curve for all 12 Ors was calculated and plotted (FIG. 9A).

Computational Validation of Mammalian OR Compound Clustering.

A 5-fold cross-validation was performed by dividing the dataset into 5 equal sized partitions containing 12 compounds each. During each run, one of the partitions is selected for testing, and the remaining 4 sets are used for training. The training process is repeated 5 times with each unique odorant set being used as the test set exactly once. For each training iteration a unique set of descriptors was calculated from the training compound set. These descriptors were then used to calculate minimum distances from the test set compounds to the closest active exactly as used to predict ligands in the ligand discovery pipeline. Once test set compounds have been ranked by distance from closest to furthest to a known active in the training set, a receiver operating characteristics (ROC) analysis is used to analyze the performance of the computational ligand prediction approach. Using ROC one can determine the predictive ability for 7 of the most broadly tuned receptors (Or2W1, MOr271-1, MOr203-1, Or1A1, MOr272-1, MOr139-1, and MOr41-1). To retain as many active compounds for each test set division as possible, the activity threshold was reduced for each of the Ors to the lowest level. All compounds with a recorded activation in the previous study were considered “active”. ROC curve averages for all of the compounds were calculated and plotted (FIGS. 18A-18G).

Or-Ligand Interaction Map.

The Or-ligand interaction map was developed using Cytoscape. Each predicted Or-ligand interaction from the top 500 predicted ligands for all of the Ors listed Table 4 were used to calculate the map. All predicted interactions are labelled in grey. In addition all interactions identified in this study, previous study and interactions for ab1A and ab1B from another study were included and labelled in black. All compounds are represented as small black circles and Ors are represented as large coloured circles. Or names are provided on the upper right corner of each Or.

Electrophysiology.

Extracellular single-sensillum electrophysiology was performed as before with a few modifications. 50 □l odor at 10⁻² dilution in paraffin oil was applied to cotton wool in odor cartridge. Odor stimulus flow=12 ml/second. Due to variability in temporal kinetics of response across various odors, the counting window was shortened to 250 milliseconds from the start of odor stimulus. A diagnostic panel of odorants to distinguish individual classes of sensilla (ab1-ab7) and therefore unequivocally identified the target ORN.

Since the structure of receptor protein complexes is not known odor-receptor interactions were analyzed by applying the similarity property principle, which reasons that structurally similar molecules (e.g. activating odorants) are more likely to have similar properties. To identify a method that describes common structural features shared by receptor actives in a quantitative fashion tractable for computational analysis four types of vastly differing molecular descriptor systems were tested: Cerius2 (Accelrys Software Inc), Dragon (Talete), Maximum-Common-Substructure, and Atom-Pair, to construct a chemical space for 109 odors that had previously been tested against 24 odor receptors from Drosophila melanogaster. These represent virtually all of the Or genes expressed in the Drosophila antenna. The four descriptor methods and associated similarity measures varied in their ability to group actives close together in descriptor space as measured for each Or using Accumulative Percentage of Actives (APoA) and value of Area Under the Curve (AUC).

Individual Ors are Tuned to Overlapping but Distinct Subsets of Ligands.

It was reasoned that cherry-picked subsets of molecular descriptors that are suited to cluster actives for an individual Or may be more effective at defining Or-specific chemical space, rather than the entire descriptor set that likely includes a number of features irrelevant for that Or. Using a Sequential-Forward-Selection method similar to previously used approaches unique optimized descriptor subsets were incrementally created for each Or from an initial set of 3424 Dragon and Cerius2 descriptors, which had performed better than Atom Pair and MCS (FIG. 1). 18 combinations of distance metrics, descriptor sets, and activity thresholds, were tested to identify the optimal selection method for each Or (FIGS. 2 and 21A-21W). Not surprisingly, the composition of the optimized descriptor sets varied greatly for individual Ors. There is an overwhelming preference for 3-D and 2-D descriptors compared to 1-D and 0-D descriptors, which suggests that structural features rather than the chemical properties of odorants are more important for receptor-odor interactions. The Or-optimized descriptor sets were far superior to non-optimized methods, and to a previous method that did not perform receptor-specific optimization (FIGS. 5 and 21A-21W).

Distances calculated by each Or-optimized descriptor set clustered the highly active compounds (˜70%) close together (FIGS. 3A-3T and 8A-8T). In a few cases, such as for Or35a and Or98a, not all the highly active compounds are clustered, suggesting the possibility of multiple or flexible binding sites, or imperfect selection of descriptors. Or2a, Or23a, Or43a and Or85f do not have strong actives, however the few weak actives of each of these 4 receptors do cluster together (FIGS. 3A-3T and 8A-8T). Actives of an Or have similar structures and pharmacophore features (FIGS. 3A-3T and 8A-8T).

Since Or-optimized descriptors can group highly active compounds in chemical space, There were used to rank untested compounds according to their distance from known actives. Approximately 4,500,000 odor-receptors interactions were systematically screened in silico, representing 19 Ors and >240,000 putative volatile compounds, a scale >1500 times that achieved in previous electrophysiology studies of odor-receptor interactions. This represents a significant achievement since high-throughput plate-based assays are not appropriate for screening volatile Or ligands, which are largely absent from the soluble combinatorial chemical libraries available for such methods. The top 500 (0.2%) hits from this vast chemical library for each of the 19 Ors were generated a fraction of which are presented in Table 4.

To validate the in silico screen several untested odorants were obtained (192; ˜11-25/Or) belonging to the top 500 predicted ligands for 9 different Ors (Tables 3 and 4). They were systematically tested with each predicted receptor-odor combination using single-unit electrophysiology to record from the olfactory receptor neurons (ORNs) to which these 9 Ors have been previously mapped in the D. melanogaster antenna (FIGS. 17A-17AK and 18A-18G). A majority of the predicted actives evoked responses from the target ORNs (FIG. 18); ˜75% evoked either activation (>50 spikes/sec above the spontaneous activity) or inhibition (>50% reduction in spontaneous activity) (FIG. 14). The success rate varied between Ors (27%-100%). A number of predicted actives that do not evoke a response (16/44) are compounds with very low volatility, raising the possibility that they may not be delivered at adequate levels to the ORNs. Taken together the physiological analysis provides the most important validation of the Or-optimized descriptor-based in silico screen of chemical space to identify volatile ligands for Ors. Previous studies have not performed well (<25% success for >50 spikes/sec) in evaluating novel odorants.

Approximately 10% of the predicted compounds showed a strong inhibitory effect (FIGS. 10A-10I, 14, 15A-15C and 16A-16B). Interestingly, inhibitors for 3 receptors, Or22a, Or47a and Or59b, were identified for which there are no previously reported inhibitors. Compounds that inhibit Ors were identified by virtue of structural similarity to Or activators. Thus the approach may provide a high-throughput method to identify putative competitive inhibitors and provide tools to investigate mechanisms of Or inhibition and their consequences in blocking specific behaviors.

Although an increasing number of insect Ors are being decoded using various methods like the Drosophila “empty neuron” system, and heterologous expression in Xenopus oocytes or cells, the process is extremely tedious and expensive. However, information on odor response profiles of single ORNs is available for several species of insects and vertebrates and relatively easy to obtain using single cell recording and/or imaging techniques. In most cases, individual ORNs ensure expression of a single Or gene and the response specificity of an ORN is imparted primarily by this associated Or. One can perform descriptor optimization using the odor response profile of the ORN directly. Or92a and Or42b have not been decoded however their corresponding antennal ORNs (ab1A and ab1B) have been tested with a panel of 47 odors. ORN-optimized descriptor sets (FIG. 12A-12F) that were efficient at clustering actives close together in chemical space were used (FIG. 15A-B). The ORN-optimized descriptor sets for ab1A and ab1B were used to screen the >240,000 library and predicted 500 novel ligands as before (Table 3, 4). Approximately 20 novel compounds were tested for each ORN, which revealed a high degree of success: >68% for ab1A and >94% for ab1B (FIGS. 15A and 15B).

Or82a was intractable to the selection of Or-optimized descriptors because it is activated strongly by a single compound, geranyl acetate, a pheromone-like long-chain hydrocarbon compound. Or82a activity is reminiscent of known insect pheromone receptors, which are often responsive only to single compounds and present an extreme challenge to understanding receptor-odor interactions. To identify novel ligands for the narrowly tuned Or82a, three additional activators of Or82a were identified from approximate predictions made using all 3424 Dragon and Cerius2 descriptors to calculate distances of >240,000 compounds in the library from geranyl acetate (FIGS. 16A-16B). The new set of four activating ligands was used to identify an Or82a-optimized descriptor set, which was successful in clustering the actives close together in chemical space (FIGS. 16A-16B). As described above, ligands were predicted from the library (Table 4), suggesting that this 3-step process can be used to predict novel ligands for narrowly tuned odor receptors, such as pheromone receptors.

The rate of false negative predictions was examined for each Or using electrophysiology to systematically test ligands of each Or against other non-target receptors. Of >640 non-target receptor-odor interactions tested, only 10.8% evoked a response >50 spikes/sec and 4.3% evoked a response >100 spikes/sec. Considering that the Or-optimized descriptor method did not incorporate any additional computational screening to rule out non-target activators, it is quite specific in its predictive ability.

Drosophila Or proteins are considered to be 7-transmembrane proteins that have a non-traditional inside-out membrane orientation, active as heteromeric ligand-gated ion channels with an obligate partner Or83b. Mammalian odor receptors on the other hand are G-protein coupled receptors with a traditional outside-in 7-transmembrane orientation. Mammals have far larger families of odor receptors (˜1000 in mice, ˜350 in humans) and thus pose a greater challenge to examine odor coding. In order to test whether the chemical informatics platform would be as successful with mammalian odor receptors a similar analysis on 33 odor receptors from mouse and 4 odor receptors from humans was performed, for which responses to a panel of 60 odorants have been determined in heterologous cells and >2 actives have been identified.

Optimized descriptor subsets for each OR were selected from an initial set of 3424 Dragon and Cerius2 descriptors as before (Table 5). The ApoA and the AUC values were comparable, if not better, than the Drosophila Ors suggesting that the descriptors were able to efficiently cluster actives together (FIGS. 17A-17AK). Since the experimental tests of predictions for mammalian receptors are beyond the scope of the analysis, a well-established computational approach to validate the in silico predictions was used. ORs with >15 known ligands were selected and for each OR 20% of the compounds (12/60) were excluded as a test set, while the remaining were used as a training set to generate the optimized descriptors. Distances of all 60 compounds from each of the known actives were calculated in chemical space and classified as active based on activity threshold. This operation was repeated five times for each receptor, each trial performed by excluding a different subset of 20% of the compounds. Average Receiver Operating Characteristic (ROC) curves were generated and AUC values were calculated, which show that optimized-descriptors generated using the training sets could accurately identify actives from the test sets (FIGS. 18A-18G).

The OR-optimized descriptors were the used to systematically screen ˜8,880,000 odor-receptor interactions in silico, representing 33 mouse ORs, 4 human ORs, and >240,000 putative volatile compounds. The top 500 hits for each receptor represent several potential novel ligands for each receptor from various natural plant and animal sources, fragrances and artificial compounds (Table 6).

Since receptor-optimized descriptor sets and the predicted ligand space they define are a function of shared molecular features that a receptor may employ to recognize ligands, it was important to determine how these characteristics correlate with receptor properties, such as their known activity profiles and amino acid sequences. Hierarchical cluster analysis was used to create trees that represent the various receptors based on: shared descriptors selected; known activity-based relationship; degree of overlap of predicted ligands; and amino acid sequence. In Drosophila, the known activity and the predicted cross-activity trees overlap to a lesser extent to each other than they do to the descriptor tree (˜67% Ors present in common subgroups). In contrast, a similar analysis for the mammalian dataset reveals a greater degree of common relationships across the known activity, predicted cross-activity and descriptor trees (˜77% ORs present in common subgroups). Similarly, the Drosophila Or-phylogenetic tree has sparser subgroup relationships conserved with each of the other trees (<45%), as opposed to the mammalian ORs where the majority of subgroups in the phylogenetic tree (>56%) are conserved across the various trees. This difference may reflect the much greater amino-acid similarity across the mammalian receptors (47%) as compared to Drosophila (23%).

Coding of odors in a large volatile space (>240,000) by a receptor repertoire is virtually impossible to determine experimentally. Based on the Or-optimized descriptor sets tuning curves were computationally derived for the 22 Drosophila Ors and 36 mammalian receptors in this large chemical space. Substantial variation in the width of the predicted tuning curves for the different receptors was demonstrated. The predicted response profiles suggest that the olfactory system can potentially detect tens of thousands of volatile chemicals, many of which the organism may never have encountered in its chemical environment.

To analyze breadth of tuning and coding potential of the antennal repertoire of Drosophila Ors to natural odors, tuning curves were calculated to an assembled set of 3197 volatile compounds from plants, humans, and a fragrance collection. Plant volatiles constituted an overwhelming majority of compounds that are predicted to be ligands for Drosophila Ors, consistent with its chemical ecology. To further analyze odor source representation odors were classified that belong to top 500 prediction lists according to their source, if known and find that Ors are not specialized for odors from a single source.

To study the predicted ensemble activation patterns of odors across all Ors, the across-receptor activation patterns of the collected compounds were analyzed for each receptor listed in Table 4. Surprisingly only a small fraction (<25%) of the collected odors are predicted to activate multiple Ors. Inclusion of all the top 500 predicted actives for each receptor further reduces the proportion of across-receptor activating compounds. Consistent with this prediction it was demonstrated that cross-activation by ligands evaluated in this study (870 receptor-odor interactions for 10 receptor neurons from FIG. 14) is lower than that reported previously using ligands of comparable strength. These data suggest that a significant number of natural odors may in fact be detected by only one or few receptors, particularly at physiologically relevant concentrations. This concept contrasts with the current model of combinatorial coding in which a majority of volatile chemicals, with the exception of pheromones and CO₂, are detected by combinations of various odor receptors. One possible explanation for this disparity is that previously tested subsets of odors were typically chosen on the basis of strong responses in electroantennograms and behavior assays, which could bias towards selection of cross activating odors. The observations that complex fruit odors activate fewer Ors than the number activated by single odors at comparable concentrations such as pentyl acetate, hexanol etc. from a typical test panel, and complex stimuli such as apple-cider-vinegar activate no more than 4-6 glomeruli lend support to this notion. The architecture of the olfactory code therefore appears to integrate two different models. On the one hand, most odors are detected by one or few Ors from the repertoire, which may enhance the specificity and efficiency of the olfactory system for detection of a large number of odors. On the other hand, 15-20% of odors are predicted to activate combinations of Ors (up to 50%), which may serve to increase the resolving capacity of the system in discriminating the defining properties of an odor stimulus.

To create a more generalized metric to quantify odorant similarity all Drosophila Or-specific molecular descriptors were concantonated and used to compute a 322-dimensional space. By visualizing the space in 2-dimensions using the two principle components, the map of the >240K chemical library overlaps well with the 3197 collected-compound volatile library, except for high molecular weight specialized flavor structures. The new ligands identified (+) overlap with previously tested compounds, and odorants distribute according to size and functional group (colors and shapes).

A network view of peripheral odor coding in the Drosophila antenna was created by mapping all predicted and tested odor-receptor combinations as has been done previously for mapping drug-target networks. The ability to decode odor receptors in silico offers a powerful approach to study the chemical ecology of an organism by potentially matching most known odors from a specific environmental source to large repertoires of target receptors or ORNs to engender a systems level view of olfactory system activation. Databases of predicted ligands will provide an invaluable tool for further studies of olfactory systems. The search for novel flavor and fragrance compounds for human beings can also be greatly assisted by a rational prioritization using such a cheminformatics approach. An emerging area of research is the identification of odors that can modify host-seeking behavior in insect disease vectors, either by virtue of their ability to inhibit ORNs that detect host-seeking cues, or by activating ORNs that cause avoidance behavior, or by confounding the pheromone detection pathway and cause mating disruption. In silico screens can provide a rational foundation for identification of novel insect repellents and lures that are environmentally safe and can aid in the fight against insect-borne diseases.

TABLE 1
Optimized descriptor sets for each Drosophila Or. Optimized descriptors occurrences, symbol, brief description,
class, and dimensionality are listed. Descriptors are listed in ascending order of when they were selected into the
optimized set. Weights indicate the number of times a descriptor was selected in an optimized descriptor set. A summary
of the total number of descriptors selected for the receptor repertoire is provided as the beginning.
Drosophila Descriptor Lists
Descriptor Class Type Counts for all Ors
	3D-MoRSE descriptors	84
	GETAWAY descriptors	84
	functional group counts	51
	2D autocorrelations	49
	edge adjacency indices	49
	2D binary fingerprints	48
	atom-centred fragments	41
	WHIM descriptors	40
	topological charge indices	26
	atomtypes (Cerius2)	25
	molecular properties	24
	Burden eigenvalues	23
	topological descriptors	22
	geometrical descriptors	18
	2D frequency fingerprints	11
	RDF descriptors	8
	walk and path counts	6
	information indices	6
	topological (Cerius2)	5
	connectivity indices	5
	constitutional descriptors	4
	structural (Cerius2)	3
	Randic molecular profiles	2
	eigenvalue-based indices	0
	charge descriptors	0
Dimensionality Counts (Weights Included)
	Num zero dimensional descriptors:	7
	Num one dimensional descriptors:	140
	Num two dimensional descriptors:	250
	Num three dimensional descriptors:	236
Origin (Weights Included)
	Num Dragon descriptors:	601
	Num Cerius2 descriptors:	33
Dimensionality Counts (Weights Excluded)
	Num zero dimensional descriptors:	6
	Num one dimensional descriptors:	50
	Num two dimensional descriptors:	130
	Num three dimensional descriptors:	145
Origin (Weights Excluded)
	Num unique Dragon descriptors:	315
	Num unique Cerius2 descriptors:	17
Number of Descriptors Per Or
	Mean (Weights Included):	41.7
	Mean (Weights Excluded):	27
Weights
	Mean:	1.5
	SD:	1.2
	Median:	1
	Mode:	1
Descriptor					Dimen-
(#Unique)	Weight	Symbol	Description	Class	sionality
Or2a (18)	1	Mor18p	3D-MoRSE - signal 18/weighted by atomic polarizabilities	3D-MoRSE descriptors	3
	1	Mor17e	3D-MoRSE - signal 17/weighted by atomic Sanderson electronegativities	3D-MoRSE descriptors	3
	1	Mor28u	3D-MoRSE - signal 28/unweighted	3D-MoRSE descriptors	3
	1	J3D	3D-Balaban index	geometrical descriptors	3
	2	O-057	phenol/enol/carboxyl OH	atom-centred	1
				fragments
	1	SIC2	structural information content (neighborhood symmetry of 2-order)	information indices	2
	1	EEig10x	Eigenvalue 10 from edge adj. matrix weighted by edge degrees	edge adjacency indices	2
	1	MATS5e	Moran autocorrelation-lag 5/weighted by atomic Sanderson	2D autocorrelations	2
			electronegativities
	1	F05[C—O]	frequency of C—O at topological distance 05	2D frequency	2
				fingerprints
	1	HNar	Narumi harmonic topological index	topological descriptors	2
	1	MATS8m	Moran autocorrelation - lag 8/weighted by atomic masses	2D autocorrelations	2
	1	G3s	3st component symmetry directional WHIM index/weighted by atomic	WHIM descriptors	3
			electrotopological states
	1	Mor27m	3D-MoRSE - signal 27/weighted by atomic masses	3D-MoRSE descriptors	3
	1	B04[C—O]	presence/absence of C—O at topological distance 04	2D binary fingerprints	2
	1	H8v	H autocorrelation of lag 8/weighted by atomic van der Waals volumes	GETAWAY descriptors	3
	1	Mor10v	3D-MoRSE - signal 10/weighted by atomic van der Waals volumes	3D-MoRSE descriptors	3
	1	Mor18v	3D-MoRSE - signal 18/weighted by atomic van der Waals volumes	3D-MoRSE descriptors	3
	2	R8p+	R maximal autocorrelation of lag 8/weighted by atomic polarizabilities	GETAWAY descriptors	3
Or7a (31)	1	MAXDP	maximal electrotopological positive variation	topological descriptors	2
	1	MAXDN	maximal electrotopological negative variation	topological descriptors	2
	1	B06[C—C]	presence/absence of C—C at topological distance 06	2D binary fingerprints	2
	2	HATS1v	leverage-weighted autocorrelation of lag 1/weighted by atomic van der	GETAWAY descriptors	3
			Waals volumes
	3	Hy	hydrophilic factor	molecular properties	1
	1	S_ssO	S_ssO	atomtypes (Cerius2)	1
	1	JGT	global topological charge index	topological charge	2
				indices
	2	H-051	H attached to alpha-C	atom-centred	1
				fragments
	2	EEig10d	Eigenvalue 10 from edge adj. matrix weighted by dipole moments	edge adjacency indices	2
	1	O-057	phenol/enol/carboxyl OH	atom-centred	1
				fragments
	5	HATS8u	leverage-weighted autocorrelation of lag 8/unweighted	GETAWAY descriptors	3
	1	G2s	2st component symmetry directional WHIM index/weighted by atomic	WHIM descriptors	3
			electrotopological states
	2	Mor16u	3D-MoRSE - signal 16/unweighted	3D-MoRSE descriptors	3
	4	B02[O—O]	presence/absence of O—O at topological distance 02	2D binary fingerprints	2
	1	R5p+	R maximal autocorrelation of lag 5/weighted by atomic polarizabilities	GETAWAY descriptors	3
	1	EEig08d	Eigenvalue 08 from edge adj. matrix weighted by dipole moments	edge adjacency indices	2
	1	DISPp	d COMMA2 value/weighted by atomic polarizabilities	geometrical descriptors	3
	2	C-008	CHR2X	atom-centred	1
				fragments
	1	R4e+	R maximal autocorrelation of lag 4/weighted by atomic Sanderson	GETAWAY descriptors	3
			electronegativities
	1	EEig09d	Eigenvalue 09 from edge adj. matrix weighted by dipole moments	edge adjacency indices	2
	1	nArOH	number of aromatic hydroxyls	functional group	1
				counts
	1	R2m+	R maximal autocorrelation of lag 2/weighted by atomic masses	GETAWAY descriptors	3
	1	nRCOOR	number of esters (aliphatic)	functional group	1
				counts
	1	B02[C—O]	presence/absence of C—O at topological distance 02	2D binary fingerprints	2
	1	GATS7m	Geary autocorrelation - lag 7/weighted by atomic masses	2D autocorrelations	2
	1	E2s	2nd component accessibility directional WHIM index/weighted by atomic	WHIM descriptors	3
			electrotopological states
	1	nRCO	number of ketones (aliphatic)	functional group	1
				counts
	1	Mor03m	3D-MoRSE - signal 03/weighted by atomic masses	3D-MoRSE descriptors	3
	1	MATS8m	Moran autocorrelation - lag 8/weighted by atomic masses	2D autocorrelations	2
	1	CIC5	complementary information content (neighborhood symmetry of 5-order)	information indices	2
	1	D/Dr06	distance/detour ring index of order 6	topological descriptors	2
Or9a (29)	1	BEHp8	highest eigenvalue n. 8 of Burden matrix/weighted by atomic	Burden eigenvalues	2
			polarizabilities
	1	BELv1	lowest eigenvalue n. 1 of Burden matrix/weighted by atomic van der	Burden eigenvalues	2
			Waals volumes
	1	DISPe	d COMMA2 value/weighted by atomic Sanderson electronegativities	geometrical descriptors	3
	2	EEig09d	Eigenvalue 09 from edge adj. matrix weighted by dipole moments	edge adjacency indices	2
	2	BEHp5	highest eigenvalue n. 5 of Burden matrix/weighted by atomic	Burden eigenvalues	2
			polarizabilities
	1	E2e	2nd component accessibility directional WHIM index/weighted by atomic	WHIM descriptors	3
			Sanderson electronegativities
	1	Mor25m	3D-MoRSE - signal 25/weighted by atomic masses	3D-MoRSE descriptors	3
	1	B03[C—C]	presence/absence of C—C at topological distance 03	2D binary fingerprints	2
	3	B07[C—C]	presence/absence of C—C at topological distance 07	2D binary fingerprints	2
	1	B01[C—O]	presence/absence of C—O at topological distance 01	2D binary fingerprints	2
	1	Atype_H_49	Number of Hydrogen Type 49	atomtypes (Cerius2)	1
	1	Infective-80	Ghose-Viswanadhan-Wendoloski antiinfective-like index at 80%	molecular properties	1
	3	O-057	phenol/enol/carboxyl OH	atom-centred	1
				fragments
	1	Mor22m	3D-MoRSE - signal 22/weighted by atomic masses	3D-MoRSE descriptors	3
	1	EEig10d	Eigenvalue 10 from edge adj. matrix weighted by dipole moments	edge adjacency indices	2
	1	R1u+	R maximal autocorrelation of lag 1/unweighted	GETAWAY descriptors	3
	1	GATS7m	Geary autocorrelation - lag 7/weighted by atomic masses	2D autocorrelations	2
	1	MATS4v	Moran autocorrelation - lag 4/weighted by atomic van der Waals volumes	2D autocorrelations	2
	1	R4e+	R maximal autocorrelation of lag 4/weighted by atomic Sanderson	GETAWAY descriptors	3
			electronegativities
	1	G3p	3st component symmetry directional WHIM index/weighted by atomic	WHIM descriptors	3
			polarizabilities
	1	Hy	hydrophilic factor	molecular properties	1
	1	S_dssC	S_dssC	atomtypes (Cerius2)	1
	1	nRCHO	number of aldehydes (aliphatic)	functional group	1
				counts
	1	B08[C—C]	presence/absence of C—C at topological distance 08	2D binary fingerprints	2
	1	R2m	R autocorrelation of lag 2/weighted by atomic masses	GETAWAY descriptors	3
	1	HATS5e	leverage-weighted autocorrelation of lag 5/weighted by atomic Sanderson	GETAWAY descriptors	3
			electronegativities
	1	D/Dr06	distance/detour ring index of order 6	topological descriptors	2
	1	RDF030m	Radial Distribution Function - 3.0/weighted by atomic masses	RDF descriptors	3
	2	Jhetv	Balaban-type index from van der Waals weighted distance matrix	topological descriptors	2
Or10a	3	S_dO	S_dO	atomtypes (Cerius2)	1
(11)	1	BEHm7	highest eigenvalue n. 7 of Burden matrix/weighted by atomic masses	Burden eigenvalues	2
	1	E2u	2nd component accessibility directional WHIM index/unweighted	WHIM descriptors	3
	1	HATS8m	leverage-weighted autocorrelation of lag 8/weighted by atomic masses	GETAWAY descriptors	3
	1	BELe4	lowest eigenvalue n. 4 of Burden matrix/weighted by atomic Sanderson	Burden eigenvalues	2
			electronegativities
	1	Mor25e	3D-MoRSE - signal 25/weighted by atomic Sanderson electronegativities	3D-MoRSE descriptors	3
	1	B08[C—C]	presence/absence of C—C at topological distance 08	2D binary fingerprints	2
	1	JGI3	mean topological charge index of order3	topological charge	2
				indices
	1	ESpm03u	Spectral moment 03 from edge adj. matrix	edge adjacency indices	2
	1	nR = Ct	number of aliphatic tertiary C(sp2)	functional group	1
				counts
	2	E2e	2nd component accessibility directional WHIM index/weighted by atomic	WHIM descriptors	3
			Sanderson electronegativities
Or19a	1	Mor31p	3D-MoRSE - signal 31/weighted by atomic polarizabilities	3D-MoRSE descriptors	3
(25)	1	H2m	H autocorrelation of lag 2/weighted by atomic masses	GETAWAY descriptors	3
	1	L1m	1st component size directional WHIM index/weighted by atomic masses	WHIM descriptors	3
	1	R1m+	R maximal autocorrelation of lag 1/weighted by atomic masses	GETAWAY descriptors	3
	1	Mor27u	3D-MoRSE - signal 27/unweighted	3D-MoRSE descriptors	3
	1	HATS6u	leverage-weighted autocorrelation of lag 6/unweighted	GETAWAY descriptors	3
	3	GGI7	topological charge index of order 7	topological charge	2
				indices
	1	Gs	G total symmetry index/weighted by atomic electrotopological states	WHIM descriptors	3
	1	O-057	phenol/enol/carboxyl OH	atom-centred	1
				fragments
	1	H-049	H attached to C3(sp3)/C2(sp2)/C3(sp2)/C3(sp)	atom-centred	1
				fragments
	1	piPC08	molecular multiple path count of order 08	walk and path counts	2
	2	R7u+	R maximal autocorrelation of lag 7/unweighted	GETAWAY descriptors	3
	2	G3s	3st component symmetry directional WHIM index/weighted by atomic	WHIM descriptors	3
			electrotopological states
	1	R4m+	R maximal autocorrelation of lag 4/weighted by atomic masses	GETAWAY descriptors	3
	1	MATS7p	Moran autocorrelation - lag 7/weighted by atomic polarizabilities	2D autocorrelations	2
	1	R6u+	R maximal autocorrelation of lag 6/unweighted	GETAWAY descriptors	3
	1	Hy	hydrophilic factor	molecular properties	1
	1	ARR	aromatic ratio	constitutional	0
				descriptors
	1	BEHp7	highest eigenvalue n. 7 of Burden matrix/weighted by atomic	Burden eigenvalues	2
			polarizabilities
	1	RDF050v	Radial Distribution Function-5.0/weighted by atomic van der Waals	RDF descriptors	3
			volumes
	1	C-005	CH3X	atom-centred	1
				fragments
	1	nRCHO	number of aldehydes (aliphatic)	functional group	1
				counts
	1	nRCOOH	number of carboxylic acids (aliphatic)	functional group	1
				counts
	1	R5m+	R maximal autocorrelation of lag 5/weighted by atomic masses	GETAWAY descriptors	3
	2	C-002	CH2R2	atom-centred	1
				fragments
Or22a	1	Mor29v	3D-MoRSE - signal 29/weighted by atomic van der Waals volumes	3D-MoRSE descriptors	3
(43)	1	MAXDN	maximal electrotopological negative variation	topological descriptors	2
	1	piPC04	molecular multiple path count of order 04	walk and path counts	2
	1	Mor10e	3D-MoRSE - signal 10/weighted by atomic Sanderson electronegativities	3D-MoRSE descriptors	3
	1	Mor27m	3D-MoRSE - signal 27/weighted by atomic masses	3D-MoRSE descriptors	3
	1	R7p+	R maximal autocorrelation of lag 7/weighted by atomic polarizabilities	GETAWAY descriptors	3
	1	S_sCH3	S_sCH3	atomtypes (Cerius2)	1
	2	EEig12r	Eigenvalue 12 from edge adj. matrix weighted by resonance integrals	edge adjacency indices	2
	1	nRCOOR	number of esters (aliphatic)	functional group	1
				counts
	4	R6u+	R maximal autocorrelation of lag 6/unweighted	GETAWAY descriptors	3
	1	Mor32p	3D-MoRSE - signal 32/weighted by atomic polarizabilities	3D-MoRSE descriptors	3
	1	AlogP98	AlogP98 value	structural (Cerius2)	0
	4	O-057	phenol/enol/carboxyl OH	atom-centred	1
				fragments
	1	L3s	3rd component size directional WHIM index/weighted by atomic	WHIM descriptors	3
			electrotopological states
	1	R1v+	R maximal autocorrelation of lag 1/weighted by atomic van der Waals	GETAWAY descriptors	3
			volumes
	2	nHDon	number of donor atoms for H-bonds (N and O)	functional group	1
				counts
	2	B10[C—C]	presence/absence of C—C at topological distance 10	2D binary fingerprints	2
	1	Mor18m	3D-MoRSE - signal 18/weighted by atomic masses	3D-MoRSE descriptors	3
	1	B04[C—O]	presence/absence of C—O at topological distance 04	2D binary fingerprints	2
	2	Jhetp	Balaban-type index from polarizability weighted distance matrix	topological descriptors	2
	1	STN	spanning tree number (log)	topological descriptors	2
	2	ESpm15u	Spectral moment 15 from edge adj. matrix	edge adjacency indices	2
	1	GATS1v	Geary autocorrelation - lag 1/weighted by atomic van der Waals volumes	2D autocorrelations	2
	1	F03[O—O]	frequency of O—O at topological distance 03	2D frequency	2
				fingerprints
	1	GATS8m	Geary autocorrelation - lag 8/weighted by atomic masses	2D autocorrelations	2
	2	HATS5e	leverage-weighted autocorrelation of lag 5/weighted by atomic Sanderson	GETAWAY descriptors	3
			electronegativities
	1	DISPv	d COMMA2 value/weighted by atomic van der Waals volumes	geometrical descriptors	3
	1	R3v+	R maximal autocorrelation of lag 3/weighted by atomic van der Waals	GETAWAY descriptors	3
			volumes
	1	E2e	2nd component accessibility directional WHIM index/weighted by atomic	WHIM descriptors	3
			Sanderson electronegativities
	1	Mor32u	3D-MoRSE - signal 32/unweighted	3D-MoRSE descriptors	3
	2	B02[O—O]	presence/absence of O—O at topological distance 02	2D binary fingerprints	2
	1	G3e	3st component symmetry directional WHIM index/weighted by atomic	WHIM descriptors	3
			Sanderson electronegativities
	1	nCrs	number of ring secondary C(sp3)	functional group	1
				counts
	2	HOMT	HOMA total	geometrical descriptors	3
	1	B05[C-C]	presence/absence of C-C at topological distance 05	2D binary fingerprints	2
	1	MATS7m	Moran autocorrelation - lag 7/weighted by atomic masses	2D autocorrelations	2
	1	RDF030m	Radial Distribution Function-3.0/weighted by atomic masses	RDF descriptors	3
	1	EEig12x	Eigenvalue 12 from edge adj. matrix weighted by edge degrees	edge adjacency indices	2
	1	R1m+	R maximal autocorrelation of lag 1/weighted by atomic masses	GETAWAY descriptors	3
	1	MATS4p	Moran autocorrelation - lag 4/weighted by atomic polarizabilities	2D autocorrelations	2
	1	B09[C—O]	presence/absence of C—O at topological distance 09	2D binary fingerprints	2
	1	Mor15p	3D-MoRSE - signal 15/weighted by atomic polarizabilities	3D-MoRSE descriptors	3
	2	S_sOH	S_sOH	atomtypes (Cerius2)	1
Or23a	1	ATS3p	Broto-Moreau autocorrelation of a topological structure - lag 3/weighted	2D autocorrelations	2
(37)			by atomic polarizabilities
	2	O-056	alcohol	atom-centred	1
				fragments
	1	J3D	3D-Balaban index	geometrical descriptors	3
	1	BELm5	lowest eigenvalue n. 5 of Burden matrix/weighted by atomic masses	Burden eigenvalues	2
	1	TPSA(Tot)	topological polar surface area using N, O, S, P polar contributions	molecular properties	1
	1	B08[C—O]	presence/absence of C—O at topological distance 08	2D binary fingerprints	2
	2	Mor27v	3D-MoRSE - signal 27/weighted by atomic van der Waals volumes	3D-MoRSE descriptors	3
	2	R6u+	R maximal autocorrelation of lag 6/unweighted	GETAWAY descriptors	3
	1	DISPe	d COMMA2 value/weighted by atomic Sanderson electronegativities	geometrical descriptors	3
	1	ESpm12d	Spectral moment 12 from edge adj. matrix weighted by dipole moments	edge adjacency indices	2
	1	Mor17m	3D-MoRSE - signal 17/weighted by atomic masses	3D-MoRSE descriptors	3
	2	EEig09d	Eigenvalue 09 from edge adj. matrix weighted by dipole moments	edge adjacency indices	2
	1	Hy	hydrophilic factor	molecular properties	1
	2	GATS3e	Geary autocorrelation - lag 3/weighted by atomic Sanderson	2D autocorrelations	2
			electronegativities
	1	GATS8m	Geary autocorrelation - lag 8/weighted by atomic masses	2D autocorrelations	2
	1	R4e+	R maximal autocorrelation of lag 4/weighted by atomic Sanderson	GETAWAY descriptors	3
			electronegativities
	1	Mor18m	3D-MoRSE - signal 18/weighted by atomic masses	3D-MoRSE descriptors	3
	2	nRCOOH	number of carboxylic acids (aliphatic)	functional group	1
				counts
	1	S_sOH	S_sOH	atomtypes (Cerius2)	1
	1	E3m	3rd component accessibility directional WHIM index/weighted by atomic	WHIM descriptors	3
			masses
	1	G3s	3st component symmetry directional WHIM index/weighted by atomic	WHIM descriptors	3
			electrotopological states
	2	BELm6	lowest eigenvalue n. 6 of Burden matrix/weighted by atomic masses	Burden eigenvalues	2
	1	GATS1m	Geary autocorrelation - lag 1/weighted by atomic masses	2D autocorrelations	2
	2	EEig08d	Eigenvalue 08 from edge adj. matrix weighted by dipole moments	edge adjacency indices	2
	1	F05[C—O]	frequency of C—O at topological distance 05	2D frequency	2
				fingerprints
	2	nHDon	number of donor atoms for H-bonds (N and O)	functional group	1
				counts
	1	EEig10d	Eigenvalue 10 from edge adj. matrix weighted by dipole moments	edge adjacency indices	2
	1	R5p+	R maximal autocorrelation of lag 5/weighted by atomic polarizabilities	GETAWAY descriptors	3
	1	BIC	BIC	topological (Cerius2)	2
	2	Infective-80	Ghose-Viswanadhan-Wendoloski antiinfective-like index at 80%	molecular properties	1
	1	GATS4p	Geary autocorrelation - lag 4/weighted by atomic polarizabilities	2D autocorrelations	2
	1	DISPp	d COMMA2 value/weighted by atomic polarizabilities	geometrical descriptors	3
	1	O-057	phenol/enol/carboxyl OH	atom-centred	1
				fragments
	1	Atype_H_49	Number of Hydrogen Type 49	atomtypes (Cerius2)	1
	1	GATS5m	Geary autocorrelation - lag 5/weighted by atomic masses	2D autocorrelations	2
	1	B02[O—O]	presence/absence of O—O at topological distance 02	2D binary fingerprints	2
	2	JGI5	mean topological charge index of order5	topological charge	2
				indices
Or33b	6	O-057	phenol/enol/carboxyl OH	atom-centred	1
(32)				fragments
	2	EEig08x	Eigenvalue 08 from edge adj. matrix weighted by edge degrees	edge adjacency indices	2
	1	DISPv	d COMMA2 value/weighted by atomic van der Waals volumes	geometrical descriptors	3
	1	TPSA(NO)	topological polar surface area using N, O polar contributions	molecular properties	1
	5	B06[C—C]	presence/absence of C—C at topological distance 06	2D binary fingerprints	2
	4	Atype_H_49	Number of Hydrogen Type 49	atomtypes (Cerius2)	1
	2	R3v+	R maximal autocorrelation of lag 3/weighted by atomic van der Waals	GETAWAY descriptors	3
			volumes
	1	G1e	1st component symmetry directional WHIM index/weighted by atomic	WHIM descriptors	3
			Sanderson electronegativities
	1	R2m+	R maximal autocorrelation of lag 2/weighted by atomic masses	GETAWAY descriptors	3
	4	B05[C—O]	presence/absence of C—O at topological distance 05	2D binary fingerprints	2
	1	C-006	CH2RX	atom-centred	1
				fragments
	2	TPSA(Tot)	topological polar surface area using N, O, S, P polar contributions	molecular properties	1
	1	L/Bw	length-to-breadth ratio by WHIM	geometrical descriptors	3
	1	EEig08d	Eigenvalue 08 from edge adj. matrix weighted by dipole moments	edge adjacency indices	2
	3	F04[C—O]	frequency of C—O at topological distance 04	2D frequency	2
				fingerprints
	1	BEHv5	highest eigenvalue n. 5 of Burden matrix/weighted by atomic van der	Burden eigenvalues	2
			Waals volumes
	1	Mor30p	3D-MoRSE - signal 30/weighted by atomic polarizabilities	3D-MoRSE descriptors	3
	1	nArCO	number of ketones (aromatic)	functional group	1
				counts
	1	nRCO	number of ketones (aliphatic)	functional group	1
				counts
	1	R1p+	R maximal autocorrelation of lag 1/weighted by atomic polarizabilities	GETAWAY descriptors	3
	1	MATS4p	Moran autocorrelation - lag 4/weighted by atomic polarizabilities	2D autocorrelations	2
	1	nN	number of Nitrogen atoms	constitutional	0
				descriptors
	1	B07[C—C]	presence/absence of C—C at topological distance 07	2D binary fingerprints	2
	2	JGI4	mean topological charge index of order4	topological charge	2
				indices
	1	nRCOOH	number of carboxylic acids (aliphatic)	functional group	1
				counts
	1	nCconj	number of non-aromatic conjugated C(sp2)	functional group	1
				counts
	1	C-005	CH3X	atom-centred	1
				fragments
	1	JGI3	mean topological charge index of order3	topological charge	2
				indices
	1	HATS3p	leverage-weighted autocorrelation of lag 3/weighted by atomic	GETAWAY descriptors	3
			polarizabilities
	1	HATS8u	leverage-weighted autocorrelation of lag 8/unweighted	GETAWAY descriptors	3
	1	E2u	2nd component accessibility directional WHIM index/unweighted	WHIM descriptors	3
	2	H-051	H attached to alpha-C	atom-centred	1
				fragments
Or35a	1	ATS4e	Broto-Moreau autocorrelation of a topological structure - lag 4/weighted	2D autocorrelations	2
(51)			by atomic Sanderson electronegativities
	2	TPSA(NO)	topological polar surface area using N, O polar contributions	molecular properties	1
	1	Mor27p	3D-MoRSE - signal 27/weighted by atomic polarizabilities	3D-MoRSE descriptors	3
	8	R6p+	R maximal autocorrelation of lag 6/weighted by atomic polarizabilities	GETAWAY descriptors	3
	6	nRCOOH	number of carboxylic acids (aliphatic)	functional group	1
				counts
	3	EEig10d	Eigenvalue 10 from edge adj. matrix weighted by dipole moments	edge adjacency indices	2
	2	Gs	G total symmetry index/weighted by atomic electrotopological states	WHIM descriptors	3
	9	JGI2	mean topological charge index of order2	topological charge	2
				indices
	3	EEig12r	Eigenvalue 12 from edge adj. matrix weighted by resonance integrals	edge adjacency indices	2
	7	R4e+	R maximal autocorrelation of lag 4/weighted by atomic Sanderson	GETAWAY descriptors	3
			electronegativities
	7	Mor28e	3D-MoRSE - signal 28/weighted by atomic Sanderson electronegativities	3D-MoRSE descriptors	3
	5	MATS7p	Moran autocorrelation - lag 7/weighted by atomic polarizabilities	2D autocorrelations	2
	2	L3s	3rd component size directional WHIM index/weighted by atomic	WHIM descriptors	3
			electrotopological states
	6	Mor25v	3D-MoRSE - signal 25/weighted by atomic van der Waals volumes	3D-MoRSE descriptors	3
	4	Mor30e	3D-MoRSE - signal 30/weighted by atomic Sanderson electronegativities	3D-MoRSE descriptors	3
	5	HATS8u	leverage-weighted autocorrelation of lag 8/unweighted	GETAWAY descriptors	3
	7	O-057	phenol/enol/carboxyl OH	atom-centred	1
				fragments
	3	HATS5m	leverage-weighted autocorrelation of lag 5/weighted by atomic masses	GETAWAY descriptors	3
	3	Jhetp	Balaban-type index from polarizability weighted distance matrix	topological descriptors	2
	4	JGI8	mean topological charge index of order8	topological charge	2
				indices
	3	Mor04m	3D-MoRSE - signal 04/weighted by atomic masses	3D-MoRSE descriptors	3
	1	S_dssC	S_dssC	atomtypes (Cerius2)	1
	2	E1m	1st component accessibility directional WHIM index/weighted by atomic	WHIM descriptors	3
			masses
	2	nHDon	number of donor atoms for H-bonds (N and O)	functional group	1
				counts
	2	RDF135u	Radial Distribution Function-13.5/unweighted	RDF descriptors	3
	2	D/Dr06	distance/detour ring index of order 6	topological descriptors	2
	3	E2s	2nd component accessibility directional WHIM index/weighted by atomic	WHIM descriptors	3
			electrotopological states
	2	EEig10r	Eigenvalue 10 from edge adj. matrix weighted by resonance integrals	edge adjacency indices	2
	1	G2s	2st component symmetry directional WHIM index/weighted by atomic	WHIM descriptors	3
			electrotopological states
	3	GATS3p	Geary autocorrelation - lag 3/weighted by atomic polarizabilities	2D autocorrelations	2
	2	GGI1	topological charge index of order 1	topological charge	2
				indices
	2	Atype_C_18	Number of Carbon Type 18	atomtypes (Cerius2)	1
	1	nRCO	number of ketones (aliphatic)	functional group	1
				counts
	1	C-005	CH3X	atom-centred	1
				fragments
	1	Mor27u	3D-MoRSE - signal 27/unweighted	3D-MoRSE descriptors	3
	2	F08[C—O]	frequency of C—O at topological distance 08	2D frequency	2
				fingerprints
	3	G3s	3st component symmetry directional WHIM index/weighted by atomic	WHIM descriptors	3
			electrotopological states
	3	SIC5	structural information content (neighborhood symmetry of 5-order)	information indices	2
	1	G(N . . . N)	sum of geometrical distances between N . . . N	geometrical descriptors	3
	2	nR = Ct	number of aliphatic tertiary C(sp2)	functional group	1
				counts
	2	E3m	3rd component accessibility directional WHIM index/weighted by atomic	WHIM descriptors	3
			masses
	1	nArCOOR	number of esters (aromatic)	functional group	1
				counts
	1	HATS6m	leverage-weighted autocorrelation of lag 6/weighted by atomic masses	GETAWAY descriptors	3
	1	nArCO	number of ketones (aromatic)	functional group	1
				counts
	1	Jhete	Balaban-type index from electronegativity weighted distance matrix	topological descriptors	2
	1	G(O . . . O)	sum of geometrical distances between O . . . O	geometrical descriptors	3
	1	nCt	number of total tertiary C(sp3)	functional group	1
				counts
	1	H-051	H attached to alpha-C	atom-centred	1
				fragments
	1	nN	number of Nitrogen atoms	constitutional	0
				descriptors
	1	P2s	2nd component shape directional WHIM index/weighted by atomic	WHIM descriptors	3
			electrotopological states
	1	C-025	R—CR—R	atom-centred	1
				fragments
Or42b	14	R3m+	R autocorrelation of lag 3/weighted by atomic masses	GETAWAY descriptors	3
(ab1B)	1	HATS3m	leverage-weighted autocorrelation of lag 3/weighted by atomic masses	GETAWAY descriptors	3
(13)	1	S_dO	S_dO	atomtypes (Cerius2)	1
	4	Mor15m	3D-MoRSE - signal 15/weighted by atomic masses	3D-MoRSE descriptors	3
	2	nDB	number of double bonds	constitutional	0
				descriptors
	4	nRCO	number of ketones (aliphatic)	functional group	1
				counts
	1	EEig08d	Eigenvalue 08 from edge adj. matrix weighted by dipole moments	edge adjacency indices	2
	3	nROH	number of hydroxyl groups	functional group	1
				counts
	2	Ks	K global shape index/weighted by atomic electrotopological states	WHIM descriptors	3
	2	B07[C—C]	presence/absence of C—C at topological distance 07	2D binary fingerprints	2
	2	E3v	3rd component accessibility directional WHIM index/weighted by atomic	WHIM descriptors	3
			van der Waals volumes
	1	P2s	2nd component shape directional WHIM index/weighted by atomic	WHIM descriptors	3
			electrotopological states
	1	R2u+	R autocorrelation of lag 2/unweighted	GETAWAY descriptors	3
	1	ESpm15u	Spectral moment 15 from edge adj. matrix	edge adjacency indices	2
	1	Mor27e	3D-MoRSE - signal 27/weighted by atomic Sanderson electronegativities	3D-MoRSE descriptors	3
	1	nArCO	number of ketones (aromatic)	functional group	1
				counts
	1	B01[C—N]	presence/absence of C—N at topological distance 01	2D binary fingerprints	2
	1	O-057	phenol/enol/carboxyl OH	atom-centred	1
				fragments
	1	HATS0p	leverage-weighted autocorrelation of lag 0/weighted by atomic	GETAWAY descriptors	3
			polarizabilities
	1	EEig08r	Eigenvalue 08 from edge adj. matrix weighted by resonance integrals	edge adjacency indices	2
	1	nR-Cs	number of aliphatic secondary C(sp2)	functional group	1
				counts
	1	R4m+	R autocorrelation of lag 4/weighted by atomic masses	GETAWAY descriptors	3
Or43a	2	O-056	alcohol	atom-centred	1
(27)				fragments
	1	BELm5	lowest eigenvalue n. 5 of Burden matrix/weighted by atomic masses	Burden eigenvalues	2
	1	B07[C—O]	presence/absence of C—O at topological distance 07	2D binary fingerprints	2
	1	R5e	R autocorrelation of lag 5/weighted by atomic Sanderson	GETAWAY descriptors	3
			electronegativities
	1	TPSA(Tot)	topological polar surface area using N,O,S,P polar contributions	molecular properties	1
	1	R6e+	R maximal autocorrelation of lag 6/weighted by atomic Sanderson	GETAWAY descriptors	3
			electronegativities
	2	JGI7	mean topological charge index of order7	topological charge	2
				indices
	3	B04[C—C]	presence/absence of C—C at topological distance 04	2D binary fingerprints	2
	1	EEig10d	Eigenvalue 10 from edge adj. matrix weighted by dipole moments	edge adjacency indices	2
	5	B02[O—O]	presence/absence of O—O at topological distance 02	2D binary fingerprints	2
	3	Mor13m	3D-MoRSE - signal 13/weighted by atomic masses	3D-MoRSE descriptors	3
	3	nHDon	number of donor atoms for H-bonds (N and O)	functional group	1
				counts
	1	Mor21m	3D-MoRSE - signal 21/weighted by atomic masses	3D-MoRSE descriptors	3
	1	JX	JX	topological (Cerius2)	2
	1	R1m+	R maximal autocorrelation of lag 1/weighted by atomic masses	GETAWAY descriptors	3
	2	GATS7m	Geary autocorrelation - lag 7/weighted by atomic masses	2D autocorrelations	2
	1	BELm6	lowest eigenvalue n. 6 of Burden matrix/weighted by atomic masses	Burden eigenvalues	2
	1	E3m	3rd component accessibility directional WHIM index/weighted by atomic	WHIM descriptors	3
			masses
	2	MATS3e	Moran autocorrelation - lag 3/weighted by atomic Sanderson	2D autocorrelations	2
			electronegativities
	1	F04[C—O]	frequency of C—O at topological distance 04	2D frequency	2
				fingerprints
	1	nRCHO	number of aldehydes (aliphatic)	functional group	1
				counts
	1	Infective-80	Ghose-Viswanadhan-Wendoloski antiinfective-like index at 80%	molecular properties	1
	1	EEig09x	Eigenvalue 09 from edge adj. matrix weighted by edge degrees	edge adjacency indices	2
	1	GATS1m	Geary autocorrelation - lag 1/weighted by atomic masses	2D autocorrelations	2
	1	CIC2	complementary information content (neighborhood symmetry of 2-order)	information indices	2
	1	EEig01d	Eigenvalue 01 from edge adj. matrix weighted by dipole moments	edge adjacency indices	2
	1	HATS6u	leverage-weighted autocorrelation of lag 6/unweighted	GETAWAY descriptors	3
Or43b	1	EEig04x	Eigenvalue 04 from edge adj. matrix weighted by edge degrees	edge adjacency indices	2
(29)	1	BEHv4	highest eigenvalue n. 4 of Burden matrix/weighted by atomic van der	Burden eigenvalues	2
			Waals volumes
	1	Mor25e	3D-MoRSE - signal 25/weighted by atomic Sanderson electronegativities	3D-MoRSE descriptors	3
	2	EEig09d	Eigenvalue 09 from edge adj. matrix weighted by dipole moments	edge adjacency indices	2
	1	E1p	1st component accessibility directional WHIM index/weighted by atomic	WHIM descriptors	3
			polarizabilities
	1	BEHe8	highest eigenvalue n. 8 of Burden matrix/weighted by atomic Sanderson	Burden eigenvalues	2
			electronegativities
	1	R1m+	R maximal autocorrelation of lag 1/weighted by atomic masses	GETAWAY descriptors	3
	2	B07[C—C]	presence/absence of C—C at topological distance 07	2D binary fingerprints	2
	1	MAXDN	maximal electrotopological negative variation	topological descriptors	2
	1	O-057	phenol/enol/carboxyl OH	atom-centred	1
				fragments
	1	Infective-80	Ghose-Viswanadhan-Wendoloski antiinfective-like index at 80%	molecular properties	1
	3	B04[CvC]	presence/absence of C—C at topological distance 04	2D binary fingerprints	2
	1	MATS5e	Moran autocorrelation - lag 5/weighted by atomic Sanderson	2D autocorrelations	2
			electronegativities
	1	Mor24v	3D-MoRSE - signal 24/weighted by atomic van der Waals volumes	3D-MoRSE descriptors	3
	1	Mor25v	3D-MoRSE - signal 25/weighted by atomic van der Waals volumes	3D-MoRSE descriptors	3
	1	BEHp4	highest eigenvalue n. 4 of Burden matrix/weighted by atomic	Burden eigenvalues	2
			polarizabilities
	1	S_sCH3	S_sCH3	atomtypes (Cerius2)	1
	1	HATS3p	leverage-weighted autocorrelation of lag 3/weighted by atomic	GETAWAY descriptors	3
			polarizabilities
	1	H7m	H autocorrelation of lag 7/weighted by atomic masses	GETAWAY descriptors	3
	1	JGI7	mean topological charge index of order7	topological charge	2
				indices
	1	STN	spanning tree number (log)	topological descriptors	2
	1	nRCOOH	number of carboxylic acids (aliphatic)	functional group	1
				counts
	1	MATS6m	Moran autocorrelation - lag 6/weighted by atomic masses	2D autocorrelations	2
	1	HATS1u	leverage-weighted autocorrelation of lag 1/unweighted	GETAWAY descriptors	3
	1	EEig10d	Eigenvalue 10 from edge adj. matrix weighted by dipole moments	edge adjacency indices	2
	1	Atype_H_49	Number of Hydrogen Type 49	atomtypes (Cerius2)	1
	1	EEig08d	Eigenvalue 08 from edge adj. matrix weighted by dipole moments	edge adjacency indices	2
	1	nCrs	number of ring secondary C(sp3)	functional group	1
				counts
	2	H-047	H attached to C1(sp3)/C0(sp2)	atom-centred	1
				fragments
Or47a	1	piPC04	molecular multiple path count of order 04	walk and path counts	2
(21)	2	DISPm	d COMMA2 value/weighted by atomic masses	geometrical descriptors	3
	1	R7e+	R maximal autocorrelation of lag 7/weighted by atomic Sanderson	GETAWAY descriptors	3
			electronegativities
	1	Mor10p	3D-MoRSE - signal 10/weighted by atomic polarizabilities	3D-MoRSE descriptors	3
	1	Mor20u	3D-MoRSE - signal 20/unweighted	3D-MoRSE descriptors	3
	1	IC1	information content index (neighborhood symmetry of 1-order)	information indices	2
	1	nRCOOH	number of carboxylic acids (aliphatic)	functional group	1
				counts
	1	EEig01d	Eigenvalue 01 from edge adj. matrix weighted by dipole moments	edge adjacency indices	2
	2	Infective-80	Ghose-Viswanadhan-Wendoloski antiinfective-like index at 80%	molecular properties	1
	1	MATS4m	Moran autocorrelation - lag 4/weighted-by atomic masses	2D autocorrelations	2
	1	GATS5p	Geary autocorrelation - lag 5/weighted by atomic polarizabilities	2D autocorrelations	2
	1	PW4	path/walk 4-Randic shape index	topological descriptors	2
	1	Mor32p	3D-MoRSE - signal 32/weighted by atomic polarizabilities	3D-MoRSE descriptors	3
	1	Mor09e	3D-MoRSE - signal 09/weighted by atomic Sanderson electronegativities	3D-MoRSE descriptors	3
	1	TPSA(NO)	topological polar surface area using N, O polar contributions	molecular properties	1
	1	B04[C—C]	presence/absence of C—C at topological distance 04	2D binary fingerprints	2
	1	O-057	phenol/enol/carboxyl OH	atom-centred	1
				fragments
	1	Atype_H_49	Number of Hydrogen Type 49	atomtypes (Cerius2)	1
	1	ESpm01d	Spectral moment 01 from edge adj. matrix weighted by dipole moments	edge adjacency indices	2
	1	EEig10d	Eigenvalue 10 from edge adj. matrix weighted by dipole moments	edge adjacency indices	2
	1	P2m	2nd component shape directional WHIM index/weighted by atomic masses	WHIM descriptors	3
	2	Mor06e	3D-MoRSE - signal 06/weighted by atomic Sanderson electronegativities	3D-MoRSE descriptors	3
Or47b	3	EEig02d	Eigenvalue 02 from edge adj. matrix weighted by dipole moments	edge adjacency indices	2
(14)	5	ESpm03d	Spectral moment 03 from edge adj. matrix weighted by dipole moments	edge adjacency indices	2
	1	nHBonds	number of intramolecular H-bonds (with N, O, F)	functional group	1
				counts
	4	X5A	average connectivity index chi-5	connectivity indices	2
	1	EEig08x	Eigenvalue 08 from edge adj. matrix weighted by edge degrees	edge adjacency indices	2
	1	C-006	CH2RX	atom-centred	1
				fragments
	1	nRCHO	number of aldehydes (aliphatic)	functional group	1
				counts
	2	nRCOOR	number of esters (aliphatic)	functional group	1
				counts
	1	nRCOOH	number of carboxylic acids (aliphatic)	functional group	1
				counts
	1	EEig08d	Eigenvalue 08 from edge adj. matrix weighted by dipole moments	edge adjacency indices	2
	1	X4Av	average valence connectivity index chi-4	connectivity indices	2
	1	GATS6m	Geary autocorrelation - lag 6/weighted by atomic masses	2D autocorrelations	2
	1	EEig07r	Eigenvalue 07 from edge adj. matrix weighted by resonance integrals	edge adjacency indices	2
	1	R2m	R autocorrelation of lag 2/weighted by atomic masses	GETAWAY descriptors	3
Or49b	2	nCb-	number of substituted benzene C(sp2)	functional group	1
(37)				counts
	1	BEHm6	highest eigenvalue n. 6 of Burden matrix/weighted by atomic masses	Burden eigenvalues	2
	2	F04[C—O]	frequency of C—O at topological distance 04	2D frequency	2
				fingerprints
	1	D/Dr06	distance/detour ring index of order 6	topological descriptors	2
	1	BEHp6	highest eigenvalue n. 6 of Burden matrix/weighted by atomic	Burden eigenvalues	2
			polarizabilities
	3	H-047	H attached to C1(sp3)/C0(sp2)	atom-centred	1
				fragments
	1	GATS1m	Geary autocorrelation - lag 1/weighted by atomic masses	2D autocorrelations	2
	3	HATS8p	leverage-weighted autocorrelation of lag 8/weighted by atomic	GETAWAY descriptors	3
			polarizabilities
	2	ISH	standardized information content on the leverage equality	GETAWAY descriptors	3
	1	Mor16e	3D-MoRSE - signal 16/weighted by atomic Sanderson electronegativities	3D-MoRSE descriptors	3
	1	JGI5	mean topological charge index of order5	topological charge	2
				indices
	1	R8e+	R maximal autocorrelation of lag 8/weighted by atomic Sanderson	GETAWAY descriptors	3
			electronegativities
	1	Mor25e	3D-MoRSE - signal 25/weighted by atomic Sanderson electronegativities	3D-MoRSE descriptors	3
	2	EEig10d	Eigenvalue 10 from edge adj. matrix weighted by dipole moments	edge adjacency indices	2
	1	Mor16p	3D-MoRSE - signal 16/weighted by atomic polarizabilities	3D-MoRSE descriptors	3
	1	JGI4	mean topological charge index of order4	topological charge	2
				indices
	1	MATS3p	Moran autocorrelation - lag 3/weighted by atomic polarizabilities	2D autocorrelations	2
	3	CIC	CIC	topological (Cerius2)	2
	1	P2m	2nd component shape directional WHIM index/weighted by atomic masses	WHIM descriptors	3
	1	nHDon	number of donor atoms for H-bonds (N and O)	functional group	1
				counts
	1	Mor03m	3D-MoRSE - signal 03/weighted by atomic masses	3D-MoRSE descriptors	3
	2	JGI7	mean topological charge index of order7	topological charge	2
				indices
	1	Mor23v	3D-MoRSE - signal 23/weighted by atomic van der Waals volumes	3D-MoRSE descriptors	3
	1	Mor30e	3D-MoRSE - signal 30/weighted by atomic Sanderson electronegativities	3D-MoRSE descriptors	3
	1	IC	IC	topological (Cerius2)	2
	1	Mor21m	3D-MoRSE - signal 21/weighted by atomic masses	3D-MoRSE descriptors	3
	1	Mor13m	3D-MoRSE - signal 13/weighted by atomic masses	3D-MoRSE descriptors	3
	1	R7v+	R maximal autocorrelation of lag 7/weighted by atomic van der Waals	GETAWAY descriptors	3
			volumes
	1	piPC07	molecular multiple path count of order 07	walk and path counts	2
	1	nArOH	number of aromatic hydroxyls	functional group	1
				counts
	1	Mor25v	3D-MoRSE - signal 25/weighted by atomic van der Waals volumes	3D-MoRSE descriptors	3
	1	Mor08v	3D-MoRSE - signal 08/weighted by atomic van der Waals volumes	3D-MoRSE descriptors	3
	1	R6e+	R maximal autocorrelation of lag 6/weighted by atomic Sanderson	GETAWAY descriptors	3
			electronegativities
	1	EEig06x	Eigenvalue 06 from edge adj. matrix weighted by edge degrees	edge adjacency indices	2
	1	C-001	CH3R/CH4	atom-centred	1
				fragments
	1	Mor07m	3D-MoRSE - signal 07/weighted by atomic masses	3D-MoRSE descriptors	3
	1	DISPe	d COMMA2 value/weighted by atomic Sanderson electronegativities	geometrical descriptors	3
	1	nR05	number of 5-membered rings	constitutional	0
				descriptors
	1	Mor07e	3D-MoRSE - signal 07/weighted by atomic Sanderson electronegativities	3D-MoRSE descriptors	3
	1	EEig09x	Eigenvalue 09 from edge adj. matrix weighted by edge degrees	edge adjacency indices	2
	1	B05[C—O]	presence/absence of C—O at topological distance 05	2D binary fingerprints	2
	1	X5Av	average valence connectivity index chi-5	connectivity indices	2
	1	HATS3p	leverage-weighted autocorrelation of lag 3/weighted by atomic	GETAWAY descriptors	3
			polarizabilities
	1	R8u+	R maximal autocorrelation of lag 8/unweighted	GETAWAY descriptors	3
	1	O-060	Al—O—Ar/Ar—O—Ar/R . . . O . . . R/R—O—C═X	atom-centred	1
				fragments
	2	B04[C—O]	presence/absence of C—O at topological distance 04	2D binary fingerprints	2
Or59b	1	piPC06	molecular multiple path count of order 06	walk and path counts	2
(23)	1	R3u	R autocorrelation of lag 3/unweighted	GETAWAY descriptors	3
	1	S_sCH3	S_sCH3	atomtypes (Cerius2)	1
	4	B06[C—C]	presence/absence of C—C at topological distance 06	2D binary fingerprints	2
	1	R1e+	R maximal autocorrelation of lag 1/weighted by atomic Sanderson	GETAWAY descriptors	3
			electronegativities
	1	ESpm03u	Spectral moment 03 from edge adj. matrix	edge adjacency indices	2
	1	EEig10r	Eigenvalue 10 from edge adj. matrix weighted by resonance integrals	edge adjacency indices	2
	1	EEig08d	Eigenvalue 08 from edge adj. matrix weighted by dipole moments	edge adjacency indices	2
	1	E1u	1st component accessibility directional WHIM index/unweighted	WHIM descriptors	3
	1	nCconj	number of non-aromatic conjugated C(sp2)	functional group	1
				counts
	1	SP13	shape profile no. 13	Randic molecular	3
				profiles
	2	S_dO	S_dO	atomtypes (Cerius2)	1
	2	Atype_H_49	Number of Hydrogen Type 49	atomtypes (Cerius2)	1
	1	EEig10d	Eigenvalue 10 from edge adj. matrix weighted by dipole moments	edge adjacency indices	2
	1	nHDon	number of donor atoms for H-bonds (N and O)	functional group	1
				counts
	1	R8u+	R maximal autocorrelation of lag 8/unweighted	GETAWAY descriptors	3
	2	O-057	phenol/enol/carboxyl OH	atom-centred	1
				fragments
	1	Mor10v	3D-MoRSE - signal 10/weighted by atomic van der Waals volumes	3D-MoRSE descriptors	3
	1	R5m+	R maximal autocorrelation of lag 5/weighted by atomic masses	GETAWAY descriptors	3
	1	Mor09e	3D-MoRSE - signal 09/weighted by atomic Sanderson electronegativities	3D-MoRSE descriptors	3
	1	nOHp	number of primary alcohols	functional group	1
				counts
	1	EEig09d	Eigenvalue 09 from edge adj. matrix weighted by dipole moments	edge adjacency indices	2
	1	nCrs	number of ring secondary C(sp3)	functional group	1
				counts
	1	ESpm01d	Spectral moment 01 from edge adj. matrix weighted by dipole moments	edge adjacency indices	2
Or65a	1	F04[O—O]	frequency of O—O at topological distance 04	2D frequency	2
(14)				fingerprints
	2	Mor30m	3D-MoRSE - signal 30/weighted by atomic masses	3D-MoRSE descriptors	3
	4	Atype_H_51	Number of Hydrogen Type 51	atomtypes (Cerius2)	1
	1	EEig08d	Eigenvalue 08 from edge adj. matrix weighted by dipole moments	edge adjacency indices	2
	2	nArOH	number of aromatic hydroxyls	functional group	1
				counts
	2	JGI7	mean topological charge index of order7	topological charge	2
				indices
	1	nHBonds	number of intramolecular H-bonds (with N, O, F)	functional group	1
				counts
	1	Mor13p	3D-MoRSE - signal 13/weighted by atomic polarizabilities	3D-MoRSE descriptors	3
	1	EEig07d	Eigenvalue 07 from edge adj. matrix weighted by dipole moments	edge adjacency indices	2
	1	B06[C—O]	presence/absence of C—O at topological distance 06	2D binary fingerprints	2
	1	C-008	CHR2X	atom-centred	1
				fragments
	1	EEig08r	Eigenvalue 08 from edge adj. matrix weighted by resonance integrals	edge adjacency indices	2
	1	B01[C—O]	presence/absence of C—O at topological distance 01	2D binary fingerprints	2
	2	Mor32e	3D-MoRSE - signal 32/weighted by atomic Sanderson electronegativities	3D-MoRSE descriptors	3
Or67a	2	AlogP98	AlogP98 value	structural (Cerius2)	0
(37)	8	B04[C—O]	presence/absence of C—O at topological distance 04	2D binary fingerprints	2
	6	F08[C—O]	frequency of C—O at topological distance 08	2D frequency	2
				fingerprints
	1	GGI4	topological charge index of order 4	topological charge	2
				indices
	3	E2u	2nd component accessibility directional WHIM index/unweighted	WHIM descriptors	3
	2	O-057	phenol/enol/carboxyl OH	atom-centred	1
				fragments
	1	Mor03v	3D-MoRSE - signal 03/weighted by atomic van der Waals volumes	3D-MoRSE descriptors	3
	4	X5A	average connectivity index chi-5	connectivity indices	2
	3	Mor10v	3D-MoRSE - signal 10/weighted by atomic van der Waals volumes	3D-MoRSE descriptors	3
	1	B03[C—O]	presence/absence of C—O at topological distance 03	2D binary fingerprints	2
	3	X4A	average connectivity index chi-4	connectivity indices	2
	3	nCt	number of total tertiary C(sp3)	functional group	1
				counts
	1	C-026	R—CX—R	atom-centred	1
				fragments	1
	3	RDF075m	Radial Distribution Function-7.5/weighted by atomic masses	RDF descriptors	3
	2	C-008	CHR2X	atom-centred	1
				fragments
	2	B03[C—C]	presence/absence of C—C at topological distance 03	2D binary fingerprints	2
	1	B01[C—O]	presence/absence of C—O at topological distance 01	2D binary fingerprints	2
	1	nRCHO	number of aldehydes (aliphatic)	functional group	1
				counts
	1	Jhetv	Balaban-type index from van der Waals weighted distance matrix	topological descriptors	2
	1	L1s	1st component size directional WHIM index/weighted by atomic	WHIM descriptors	3
			electrotopological states
	1	Hy	hydrophilic factor	molecular properties	1
	2	C-003	CHR3	atom-centred	1
				fragments
	1	GATS7m	Geary autocorrelation - lag 7/weighted by atomic masses	2D autocorrelations	2
	1	Mor16e	3D-MoRSE - signal 16/weighted by atomic Sanderson electronegativities	3D-MoRSE descriptors	3
	1	Mor06u	3D-MoRSE - signal 06/unweighted	3D-MoRSE descriptors	3
	1	RDF030m	Radial Distribution Function-3.0/weighted by atomic masses	RDF descriptors	3
	1	Atype_C_18	Number of Carbon Type 18	atomtypes (Cerius2)	1
	1	F03[O—O]	frequency of O—O at topological distance 03	2D frequency	2
				fingerprints
	1	nCrs	number of ring secondary C(sp3)	functional group	1
				counts
	2	nArOH	number of aromatic hydroxyls	functional group	1
				counts
	1	GATS8m	Geary autocorrelation - lag 8/weighted by atomic masses	2D autocorrelations	2
	1	Jhete	Balaban-type index from electronegativity weighted distance matrix	topological descriptors	2
	1	EEig13x	Eigenvalue 13 from edge adj. matrix weighted by edge degrees	edge adjacency indices	2
	1	DISPm	d COMMA2 value/weighted by atomic masses	geometrical descriptors	3
	1	X3A	average connectivity index chi-3	connectivity indices	2
	1	G(N . . . N)	sum of geometrical distances between N . . . N	geometrical descriptors	3
	1	Mor32u	3D-MoRSE - signal 32/unweighted	3D-MoRSE descriptors	3
Or67c	1	BEHe8	highest eigenvalue n. 8 of Burden matrix/weighted by atomic Sanderson	Burden eigenvalues	2
(24)			electronegativities
	1	O-056	alcohol	atom-centred	1
				fragments
	1	Mor25m	3D-MoRSE - signal 25/weighted by atomic masses	3D-MoRSE descriptors	3
	1	BELv4	lowest eigenvalue n. 4 of Burden matrix/weighted by atomic van der	Burden eigenvalues	2
			Waals volumes
	3	B07[C—C]	presence/absence of C—C at topological distance 07	2D binary fingerprints	2
	1	TPSA(Tot)	topological polar surface area using N, O, S, P polar contributions	molecular properties	1
	1	DISPm	d COMMA2 value/weighted by atomic masses	geometrical descriptors	3
	4	HATS6u	leverage-weighted autocorrelation of lag 6/unweighted	GETAWAY descriptors	3
	2	EEig08d	Eigenvalue 08 from edge adj. matrix weighted by dipole moments	edge adjacency indices	2
	2	EEig10d	Eigenvalue 10 from edge adj. matrix weighted by dipole moments	edge adjacency indices	2
	1	Gs	G total symmetry index/weighted by atomic electrotopological states	WHIM descriptors	3
	3	O-057	phenol/enol/carboxyl OH	atom-centred	1
				fragments
	1	B08[C—C]	presence/absence of C—C at topological distance 08	2D binary fingerprints	2
	1	R1m+	R maximal autocorrelation of lag 1/weighted by atomic masses	GETAWAY descriptors	3
	1	BELm5	lowest eigenvalue n. 5 of Burden matrix/weighted by atomic masses	Burden eigenvalues	2
	1	F03[O—O]	frequency of O—O at topological distance 03	2D frequency	2
				fingerprints
	1	STN	spanning tree number (log)	topological descriptors	2
	1	Atype_H_49	Number of Hydrogen Type 49	atomtypes (Cerius2)	1
	1	H-051	H attached to alpha-C	atom-centred	1
				fragments
	1	B01[C—O]	presence/absence of C—O at topological distance 01	2D binary fingerprints	2
	1	Infective-80	Ghose-Viswanadhan-Wendoloski antiinfective-like index at 80%	molecular properties	1
	1	Hy	hydrophilic factor	molecular properties	1
	1	Mor22m	3D-MoRSE - signal 22/weighted by atomic masses	3D-MoRSE descriptors	3
	1	JGI7	mean topological charge index of order7	topological charge	2
				indices
Or82a	1	GGI9	topological charge index of order 9	topological charge	2
(31)				indices
	1	Mor02e	3D-MoRSE - signal 02/weighted by atomic Sanderson electronegativities	3D-MoRSE descriptors	3
	1	Mor30v	3D-MoRSE - signal 30/weighted by atomic van der Waals volumes	3D-MoRSE descriptors	3
	1	Mor02v	3D-MoRSE - signal 02/weighted by atomic van der Waals volumes	3D-MoRSE descriptors	3
	1	Mor30u	3D-MoRSE - signal 30/unweighted	3D-MoRSE descriptors	3
	2	BLTD48	Verhaar model of Daphnia base-line toxicity from MLOGP (mmol/l)	molecular properties	1
	2	Mor10v	3D-MoRSE - signal 10/weighted by atomic van der Waals volumes	3D-MoRSE descriptors	3
	2	Atype_H_53	Number of Hydrogen Type 53	atomtypes (Cerius2)	1
	1	O-058	═O	atom-centred	1
				fragments
	1	B02[C—O]	presence/absence of C—O at topological distance 02	2D binary fingerprints	2
	2	R5u+	R maximal autocorrelation of lag 5/unweighted	GETAWAY descriptors	3
	1	H6e	H autocorrelation of lag 6/weighted by atomic Sanderson	GETAWAY descriptors	3
			electronegativities
	1	MATS7p	Moran autocorrelation - lag 7/weighted by atomic polarizabilities	2D autocorrelations	2
	1	GATS3p	Geary autocorrelation - lag 3/weighted by atomic polarizabilities	2D autocorrelations	2
	1	Mor18m	3D-MoRSE - signal 18/weighted by atomic masses	3D-MoRSE descriptors	3
	1	H-051	H attached to alpha-C	atom-centred	1
				fragments
	2	Mor13p	3D-MoRSE - signal 13/weighted by atomic polarizabilities	3D-MoRSE descriptors	3
	1	SIC2	structural information content (neighborhood symmetry of 2-order)	information indices	2
	1	Mor32u	3D-MoRSE - signal 32/unweighted	3D-MoRSE descriptors	3
	1	Mor10m	3D-MoRSE - signal 10/weighted by atomic masses	3D-MoRSE descriptors	3
	1	nR = Cp	number of terminal primary C(sp2)	functional group	1
				counts
	1	Mor25p	3D-MoRSE - signal 25/weighted by atomic polarizabilities	3D-MoRSE descriptors	3
	1	GATS8m	Geary autocorrelation - lag 8/weighted by atomic masses	2D autocorrelations	2
	1	JGI1	mean topological charge index of order 1	topological charge	2
				indices
	1	E-ADJ-mag	E-ADJ-mag	topological (cerius2)	2
	1	EEig11x	Eigenvalue 11 from edge adj. matrix weighted by edge degrees	edge adjacency indices	2
	1	B03[O—O]	presence/absence of O—O at topological distance 03	2D binary fingerprints	2
	1	Mor30e	3D-MoRSE - signal 30/weighted by atomic Sanderson electronegativities	3D-MoRSE descriptors	3
	1	Rotlbonds	Number of rotatable bonds	structural (Cerius2)	0
	1	EEig09d	Eigenvalue 09 from edge adj. matrix weighted by dipole moments	edge adjacency indices	2
	2	GATS7m	Geary autocorrelation - lag 7/weighted by atomic masses	2D autocorrelations	2
Or85a	1	EEig04r	Eigenvalue 04 from edge adj. matrix weighted by resonance integrals	edge adjacency indices	2
(15)	2	C-006	CH2RX	atom-centred	1
				fragments
	3	ATS6e	Broto-Moreau autocorrelation of a topological structure - lag 6/weighted	2D autocorrelations	2
			by atomic Sanderson electronegativities
	3	JGI5	mean topological charge index of order5	topological charge	2
				indices
	2	B07[C—C]	presence/absence of C—C at topological distance 07	2D binary fingerprints	2
	1	nCp	number of terminal primary C(sp3)	functional group	1
				counts
	2	DISPm	d COMMA2 value/weighted by atomic masses	geometrical descriptors	3
	2	GATS4m	Geary autocorrelation - lag 4/weighted by atomic masses	2D autocorrelations	2
	1	Mor25p	3D-MoRSE - signal 25/weighted by atomic polarizabilities	3D-MoRSE descriptors	3
	1	nHDon	number of donor atoms for H-bonds (N and O)	functional group	1
				counts
	1	EEig09d	Eigenvalue 09 from edge adj. matrix weighted by dipole moments	edge adjacency indices	2
	1	R2m+	R maximal autocorrelation of lag 2/weighted by atomic masses	GETAWAY descriptors	3
	1	JGI4	mean topological charge index of order4	topological charge	2
				indices
	1	Mor11e	3D-MoRSE - signal 11/weighted by atomic Sanderson electronegativities	3D-MoRSE descriptors	3
	2	HATS7m	leverage-weighted autocorrelation of lag 7/weighted by atomic masses	GETAWAY descriptors	3
Or85b	1	piPC05	molecular multiple path count of order 05	walk and path counts	2
(26)	1	BLTF96	Verhaar model of Fish base-line toxicity from MLOGP (mmol/l)	molecular properties	1
	2	GATS4p	Geary autocorrelation - lag 4/weighted by atomic polarizabilities	2D autocorrelations	2
	1	GGI7	topological charge index of order 7	topological charge	2
				indices
	3	B05[C—O]	presence/absence of C—O at topological distance 05	2D binary fingerprints	2
	2	O-057	phenol/enol/carboxyl OH	atom-centred	1
				fragments
	1	Mor27v	3D-MoRSE - signal 27/weighted by atomic van der Waals volumes	3D-MoRSE descriptors	3
	1	HATS4v	leverage-weighted autocorrelation of lag 4/weighted by atomic van der	GETAWAY descriptors	3
			Waals volumes
	1	Gs	G total symmetry index/weighted by atomic electrotopological states	WHIM descriptors	3
	2	Infective-80	Ghose-Viswanadhan-Wendoloski antiinfective-like index at 80%	molecular properties	1
	2	R7u+	R maximal autocorrelation of lag 7/unweighted	GETAWAY descriptors	3
	2	nCbH	number of unsubstituted benzene C(sp2)	functional group	1
				counts
	1	B04[C—O]	presence/absence of C—O at topological distance 04	2D binary fingerprints	2
	2	JGI7	mean topological charge index of order7	topological charge	2
				indices
	2	DISPe	d COMMA2 value/weighted by atomic Sanderson electronegativities	geometrical descriptors	3
	1	R4p+	R maximal autocorrelation of lag 4/weighted by atomic polarizabilities	GETAWAY descriptors	3
	1	EEig12x	Eigenvalue 12 from edge adj. matrix weighted by edge degrees	edge adjacency indices	2
	1	B06[C—O]	presence/absence of C—O at topological distance 06	2D binary fingerprints	2
	1	MATS5e	Moran autocorrelation - lag 5/weighted by atomic Sanderson	2D autocorrelations	2
			electronegativities
	1	HATS4m	leverage-weighted autocorrelation of lag 4/weighted by atomic masses	GETAWAY descriptors	3
	1	HATS6u	leverage-weighted autocorrelation of lag 6/unweighted	GETAWAY descriptors	3
	1	GATS4m	Geary autocorrelation - lag 4/weighted by atomic masses	2D autocorrelations	2
	1	F03[O—O]	frequency of O—O at topological distance 03	2D frequency	2
				fingerprints
	1	H8v	H autocorrelation of lag 8/weighted by atomic van der Waals volumes	GETAWAY descriptors	3
	1	EEig09d	Eigenvalue 09 from edge adj. matrix weighted by dipole moments	edge adjacency indices	2
	2	Mor16e	3D-MoRSE - signal 16/weighted by atomic Sanderson electronegativities	3D-MoRSE descriptors	3
Or85f	1	BEHp8	highest eigenvalue n. 8 of Burden matrix/weighted by atomic	Burden eigenvalues	2
(53)			polarizabilities
	5	F05[C—O]	frequency of C—O at topological distance 05	2D frequency	2
				fingerprints
	4	BELm4	lowest eigenvalue n. 4 of Burden matrix/weighted by atomic masses	Burden eigenvalues	2
	1	HATS8m	leverage-weighted autocorrelation of lag 8/weighted by atomic masses	GETAWAY descriptors	3
	2	B04[C—O]	presence/absence of C—O at topological distance 04	2D binary fingerprints	2
	6	O-057	phenol/enol/carboxyl OH	atom-centred	1
				fragments
	1	RDF030v	Radial Distribution Function-3.0/weighted by atomic van der Waals	RDF descriptors	3
			volumes
	1	GGI7	topological charge index of order 7	topological charge	2
				indices
	1	Gs	G total symmetry index/weighted by atomic electrotopological states	WHIM descriptors	3
	4	B07[C—C]	presence/absence of C—C at topological distance 07	2D binary fingerprints	2
	1	E2e	2nd component accessibility directional WHIM index/weighted by atomic	WHIM descriptors	3
			Sanderson electronegativities
	1	MATS2m	Moran autocorrelation - lag 2/weighted by atomic masses	2D autocorrelations	2
	2	Mor28u	3D-MoRSE - signal 28/unweighted	3D-MoRSE descriptors	3
	3	BEHp5	highest eigenvalue n. 5 of Burden matrix/weighted by atomic	Burden eigenvalues	2
			polarizabilities
	2	Infective-80	Ghose-Viswanadhan-Wendoloski antiinfective-like index at 80%	molecular properties	1
	1	HATS4e	leverage-weighted autocorrelation of lag 4/weighted by atomic Sanderson	GETAWAY descriptors	3
			electronegativities
	3	JGI6	mean topological charge index of order6	topological charge	2
				indices
	6	B05[C—O]	presence/absence of C—O at topological distance 05	2D binary fingerprints	2
	2	JGI7	mean topological charge index of order7	topological charge	2
				indices
	2	DISPm	d COMMA2 value/weighted by atomic masses	geometrical descriptors	3
	5	RDF030m	Radial Distribution Function-3.0/weighted by atomic masses	RDF descriptors	3
	1	R1e+	R maximal autocorrelation of lag 1/weighted by atomic Sanderson	GETAWAY descriptors	3
			electronegativities
	1	HATS8p	leverage-weighted autocorrelation of lag 8/weighted by atomic	GETAWAY descriptors	3
			polarizabilities
	1	Atype_H_49	Number of Hydrogen Type 49	atomtypes (Cerius2)	1
	2	Hy	hydrophilic factor	molecular properties	1
	1	Jhetp	Balaban-type index from polarizability weighted distance matrix	topological descriptors	2
	1	H8v	H autocorrelation of lag 8/weighted by atomic van der Waals volumes	GETAWAY descriptors	3
	2	EEig11d	Eigenvalue 11 from edge adj. matrix weighted by dipole moments	edge adjacency indices	2
	1	MATS8m	Moran autocorrelation - lag 8/weighted by atomic masses	2D autocorrelations	2
	1	MATS2p	Moran autocorrelation - lag 2/weighted by atomic polarizabilities	2D autocorrelations	2
	4	B08[C—C]	presence/absence of C—C at topological distance 08	2D binary fingerprints	2
	1	S_sCH3	S_sCH3	atomtypes (Cerius2)	1
	2	HATS1e	leverage-weighted autocorrelation of lag 1/weighted by atomic Sanderson	GETAWAY descriptors	3
			electronegativities
	1	nCconj	number of non-aromatic conjugated C(sp2)	functional group	1
				counts
	1	B04[C—C]	presence/absence of C—C at topological distance 04	2D binary fingerprints	2
	1	S_aasC	S_aasC	atomtypes (cerius2)	1
	1	R8m+	R maximal autocorrelation of lag 8/weighted by atomic masses	GETAWAY descriptors	3
	1	nRCOOH	number of carboxylic acids (aliphatic)	fundtional group	1
				counts
	1	S_sOH	S_sOH	atomtypes (Cerius2)	1
	1	BELe3	lowest eigenvalue n. 3 of Burden matrix/weighted by atomic Sanderson	Burden eigenvalues	2
			electronegativities
	1	GATS8m	Geary autocorrelation - lag 8/weighted by atomic masses	2D autocorrelations	2
	1	BEHp4	highest eigenvalue n. 4 of Burden matrix/weighted by atomic	Burden eigenvalues	2
			polarizabilities
	2	MATS5e	Moran autocorrelation - lag 5/weighted by atomic Sanderson	2D autocorrelations	2
			electronegativities
	1	E3s	3rd component accessibility directional WHIM index/weighted by atomic	WHIM descriptors	3
			electrotopological states
	2	Jhetv	Balaban-type index from van der Waals weighted distance matrix	topological descriptors	2
	1	nR = Ct	number of aliphatic tertiary C(sp2)	functional group	1
				counts
	1	nRCHO	number of aldehydes (aliphatic)	functional group	1
				counts
	1	HATS8v	leverage-weighted autocorrelation of lag 8/weighted by atomic van der	GETAWAY descriptors	3
			Waals volumes
	1	Mor28p	3D-MoRSE - signal 28/weighted by atomic polarizabilities	3D-MoRSE descriptors	3
	1	C-003	CHR3	atom-centred	1
				fragments
	1	GATS7m	Geary autocorrelation - lag 7/weighted by atomic masses	2D autocorrelations	2
	1	JGI9	mean topological charge index of order9	topological charge	2
				indices
	1	B03[C—C]	presence/absence of C—C at topological distance 03	2D binary fingerprints	2
Or88a	3	nHBonds	number of intramolecular H-bonds (with N, O, F)	functional group	1
(19)				counts
	2	nRCO	number of ketones (aliphatic)	functional group	1
				counts
	3	GATS6m	Geary autocorrelation - lag 6/weighted by atomic masses	2D autocorrelations	2
	2	EEig08x	Eigenvalue 08 from edge adj. matrix weighted by edge degrees	edge adjacency indices	2
	1	nFuranes	number of Furanes	functional group	1
				counts
	1	nArCO	number of ketones (aromatic)	functional group	1
				counts
	1	ESpm15d	Spectral moment 15 from edge adj. matrix weighted by dipole moments	edge adjacency indices	2
	1	C-005	CH3X	atom-centred	1
				fragments
	1	O-057	phenol/enol/carboxyl OH	atom-centred	1
				fragments
	1	L/Bw	length-to-breadth ratio by WHIM	geometrical descriptors	3
	1	nArCOOR	number of esters (aromatic)	functional group	1
				counts
	1	ESpm15u	Spectral moment 15 from edge adj. matrix	edge adjacency indices	2
	1	E2u	2nd component accessibility directional WHIM index/unweighted	WHIM descriptors	3
	1	EEig08d	Eigenvalue 08 from edge adj. matrix weighted by dipole moments	edge adjacency indices	2
	1	H-051	H attached to alpha-C	atom-centred	1
				fragments
	1	ESpm14d	Spectral moment 14 from edge adj. matrix weighted by dipole moments	edge adjacency indices	2
	1	GATS7m	Geary autocorrelation - lag 7/weighted by atomic masses	2D autocorrelations	2
	1	PJI3	3D Petitjean shape index	geometrical descriptors	3
	2	X3A	average connectivity index chi-3	connectivity indices	2
Or92a	2	nRCOOR	number of esters (aliphatic)	functional group	1
(ab1A)				counts
(22)	1	Mor10u	3D-MoRSE - signal 10/unweighted	3D-MoRSE descriptors	3
	1	Mor04m	3D-MoRSE - signal 04/weighted by atomic masses	3D-MoRSE descriptors	3
	1	R1e+	R autocorrelation of lag 1/weighted by atomic Sanderson	GETAWAY descriptors	3
			electronegativities
	1	Mor27m	3D-MoRSE - signal 27/weighted by atomic masses	3D-MoRSE descriptors	3
	1	nHAcc	number of acceptor atoms for H-bonds (N, O, F)	functional group	1
				counts
	1	Elm	1st component accessibility directional WHIM index/weighted by atomic	WHIM descriptors	3
			masses
	1	GATS5m	Geary autocorrelation - lag 5/weighted by atomic masses	2D autocorrelations	2
	1	nROH	number of hydroxyl groups	functional group	1
				counts
	1	R5v	R autocorrelation of lag 5/weighted by atomic van der Waals volumes	GETAWAY descriptors	3
	1	Mor10p	3D-MoRSE - signal 10/weighted by atomic polarizabilities	3D-MoRSE descriptors	3
	1	C-006	CH2RX	atom-centred	1
				fragments
	2	Mor11e	3D-MoRSE - signal 11/weighted by atomic Sanderson electronegativities	3D-MoRSE descriptors	3
Or98a	1	Lop	Lopping centric index	topological descriptors	2
(20)	4	O-057	phenol/enol/carboxyl OH	atom-centred	1
				fragments
	2	B04[C—O]	presence/absence of C—O at topological distance 04	2D binary fingerprints	2
	1	GVWAI-80	Ghose-Viswanadhan-Wendoloski drug-like index at 80%	molecular properties	1
	1	HATS7p	leverage-weighted autocorrelation of lag 7/weighted by atomic	GETAWAY descriptors	3
			polarizabilities
	1	HATS5v	leverage-weighted autocorrelation of lag 5/weighted by atomic van der	GETAWAY descriptors	3
			Waals volumes
	1	MLOGP2	Squared Moriguchi octanol-water partition coeff. (logP{circumflex over ( )}2)	molecular properties	1
	2	GATS5e	Geary autocorrelation - lag 5/weighted by atomic Sanderson	2D autocorrelations	2
			electronegativities
	1	H-049	H attached to C3(sp3)/C2(sp2)/C3(sp2)/C3(sp)	atom-centred	1
				fragments
	1	MATS8m	Moran autocorrelation - lag 8/weighted by atomic masses	2D autocorrelations	2
	1	nCrs	number of ring secondary C(sp3)	functional group	1
				counts
	3	HATS3p	leverage-weighted autocorrelation of lag 3/weighted by atomic	GETAWAY descriptors	3
			polarizabilities
	1	G1s	1st component symmetry directional WHIM index/weighted by atomic	WHIM descriptors	3
			electrotopological states
	1	S_aasC	S_aasC	atomtypes (Cerius2)	1
	1	SP18	shape profile no. 18	Randic molecular	3
				profiles
	1	B05[C—C]	presence/absence of C—C at topological distance 05	2D binary fingerprints	2
	1	JGI2	mean topological charge index of order2	topological charge	2
				indices
	1	JGI8	mean topological charge index of order8	topological charge	2
				indices
	1	X4A	average connectivity index chi-4	connectivity indices	2
	1	H5e	H autocorrelation of lag 5/weighted by atomic Sanderson	GETAWAY descriptors	3
			electronegativities

TABLE 4
Top 25 predicted compounds for each Drosophila Or.
Tables contain SMILES strings, and distances, of the top
25 predicted compounds for each Or. All distances represent
the minimum distance based on optimized descriptors to
an active compound for that particular Or.
SMILES                    Dist
Or2a
CCSC(C)OC(C)O             0.06547077
CC(C)CCOC(═O)N            0.07017575
CC(C)CC═CC(═O)C           0.08148948
C(CCC)OC(═O)C═C           0.08191658
CCCCSCC(C)O               0.08378222
CCCOC(═O)C(═C)C           0.083826
CC(C)CCOC(═O)C═C          0.0868181
CC(C)CCOC(═O)CS           0.09010645
CCCCOC(═O)CS              0.0962103
CCC(═O)OCCC(C)C           0.09663025
CCCCOC(═O)NC              0.09927085
CC(C)CCOC═O               0.1000939
CCCCOC(═O)N               0.1060414
CCCCOC(═O)CC              0.1064284
CCCNOC(═O)CC              0.1068854
CCCC(═O)CCC#CC            0.1072292
CC(═CC(═C)OC(═O)C)C       0.1073104
CC(═C)C(═O)OCCCN          0.1084215
CCCOC(═O)C(═CC)C          0.1113651
CCC(C)C(═O)OCC═C          0.1126422
CCCCC(═O)C═C(C)C          0.1134776
CCOC(C)OCC#C              0.1143103
CCCCOC(═O)CN              0.1163705
CCC#CC(CC(C)C)O           0.1186069
CCCOC(═O)C═C(C)C          0.1191027
Or7a
CCC═CC═O                  0.06287397
CC(═CC)CO                 0.08745256
CC(═CCCO)C                0.092048
C1═C(ON═C1)C═O            0.112377
CCC(C)(C)C═NO             0.1149527
CC═CC═CC═O                0.1158738
C1CC(═CC═C1)C═O           0.1165349
C1═COC═C1C═O              0.1266509
C1═C(OC═N1)C═O            0.1277235
CCC(C)CC═O                0.1310643
CCCN═CC═O                 0.1384452
CCC═CCO                   0.1388489
CCC(═C═CCO)C              0.139234
CCC(═C═CC═O)C             0.1407318
C═C(CO)CC                 0.1424594
CC(═C)CCCO                0.1441704
CC(═CCCC═O)C              0.1441953
C═CC(═C)CCC═O             0.1466388
CCCC(═C)CCO               0.1482145
C1CC1CCO                  0.148306
C1CC1CO                   0.1521086
C1═CC═NC(═C1)C═O          0.153194
CC═CCCCO                  0.1548523
CC═CCO                    0.1553325
CCCC#CCO                  0.1555892
Or9a
CC(CCC═C)O                0.09391671
CC(CCCO)C                 0.1227934
CC(C)NC(═O)C              0.1291849
CCC(CCC)O                 0.1337314
CC(C)C(C)O                0.1449655
CC═CCOC(═O)C              0.1529503
CCC(═O)NCC#C              0.165569
COCCC═C═N                 0.1686873
CC(C)NC(═O)C═C            0.1706617
CC(CCCC)O                 0.1771611
C═CC(CCC)O                0.1824546
CCC#CCCO                  0.18493
CCC(CC═C)O                0.1958346
CCC(═O)NCC(═O)C           0.2001053
CC(CCOC(═O)C)O            0.2011426
CCC(═O)NCC#N              0.2057452
CC(CC(C)O)C               0.2065425
COC(═O)CC(O)C             0.209219
CC(CC(═O)C)O              0.2110131
CCC(═O)OC(C)(C)O          0.2110216
C═CC(C═C)O                0.2127051
CC(C)C(═O)NC═C            0.2141241
CC(═O)CNCC(═O)C           0.2158649
CCC(═O)NCC                0.2169106
CCC(CC#C)O                0.2179049
Or10a
C1═CC═C(C═C1)C═S═O        0
CC1═C(C(═CC═C1)C)N═S═O    0
CC1═CC(═C(C═C1)OC)C═O     0
CN(C═O)C1═CC═CC═N1        0
CN1C═CC═CC1C═O            0
C1═CC═C(C═C1)C(C═O)C#N    0
C1═CC═C(C(═C1)NC═O)O      0
C1CCC(C1)C(═O)NN          0
C1═CC═C(C═C1)C═C═O        0
CNC(═O)C1═CC═NC═C1        0
CC(═O)NCN(C)C             0
CC(C1═CC═CC═C1)N═C═O      0
CC(═O)C#CC1═CC═CC═C1      0
CC(C)CC(C)C═O             0
C1CCCN(CCC1)N═O           0
C1CCC(CC1)C═O             0
C═CC(═O)C1═CC═CC═C1       0
C1CCN(CC1)N═O             0
C1CCC(CC1)NN═O            0
C═CC(═O)C(═O)C1═CC═CC═C1  0
C1═CC═C(C═C1)CN(C═O)O     0
C1═CC═C(C═C1)C(═O)CS      0
CCN(C═O)C1═CC═CC═N1       0
C1═CC═C(C═C1)C(═S═O)C#N   0
COC(═O)C1═CC═CC═C1N═O     0
OR19a
CC(CCC(CC)O)CC            0.1773688
C═CCCCCC                  0.1795821
CCCCCC(═N)C               0.1837474
CCCCC(═O)CC               0.1983858
CCCCCC1CO1                0.2055113
CCCCOC1CC1                0.2097268
CCCCCC1(CO1)C             0.2192309
CCCCCC(═C)C               0.2327693
CCCC(CCC═C)O              0.233932
CCCCCC(C)(C#C)O           0.2437784
CCC(CCCC═C)O              0.2460947
CCCCCC(═O)C═C             0.2539876
CCCCCC(C)S                0.2540855
CCCCC(═O)C(═C)CC          0.2541479
CC(CC)CCCC                0.2605213
CCCC(═O)CCC═C             0.268571
CCCCC(═O)OC(═O)C          0.2821946
CCCC(═O)CCC               0.2839333
CCCCCC(C)(C)O             0.2848814
OC(C(═O)C)CCCCC           0.2870452
CCCCC(CC(═C)C)O           0.2945198
CCCCCC(C)O                0.3036086
CCCCCOC═C                 0.304017
CCCCCC(C═C)S              0.3054288
CCCCCC(C)(C═C)O           0.3074979
Or22a
CCCCOC(═O)CCC             0.2657116
CCC#CCCOC(═O)C            0.2798496
CCCCC(═O)OC               0.2807561
CCC(═O)OCC═CC             0.3192234
CCCCC(═O)OCC              0.3386281
C(CC)OC(═O)CCCC           0.3432405
CCCC═CC(═O)OCC            0.3470114
COC(═O)C═CC               0.3564047
CCCOC(═O)CC               0.3620649
CCCC(═O)OCCC              0.3642294
CCCCOC(═O)CC              0.408598
CC(═C)CCCC(═O)OC          0.4087118
CCCCCC(═O)OCCC            0.4096699
CCC═CCCOC(═O)C            0.4280228
CCC═CCC(═O)OCC            0.4509044
COC(═O)CCC═CCCC           0.4515538
COC(═O)C═CCC              0.4606538
CCC#CC(═O)OC              0.4635536
CCCCCOC(═O)CC             0.465345
CC(═CCOC(═O)C)C           0.4677529
COC(═O)CC═CC              0.4684388
CCC═CC(═O)OCC             0.4687615
CC(═C)CCCOC(═O)C          0.4696284
CC(═COC(═O)C)C            0.4710929
C(C)OC(═O)CCCCCC          0.4714722
Or23a
CCCCC═CCO                 0.4489215
C(CCC═CCC)O               0.4645003
C═CCCCCCO                 0.4966429
CC═CC═CCO                 0.5369311
CC(CCC═C)O                0.5705127
C#CCCCCO                  0.653074
CCCCCOO                   0.6743714
C(CCCCCC)O                0.679884
C#CCCCCCO                 0.6854802
COC═CCCCO                 0.6878323
C(C═CC═CCC)O              0.730026
CC═CC═CCCO                0.7327627
CC(CCCC═C)O               0.7331066
CCCC(C)CCO                0.7638355
CCCC#CC#CO                0.7642293
CCCCOO                    0.8340461
CCC#CCCCO                 0.8383432
CCCC═CCCO                 0.8559539
CC(CCC#C)O                0.8633463
CCCCCCOO                  0.8935004
CCCCCC#CO                 0.895056
CC(CCC═C═C)O              0.913551
CCC═CCO                   0.9216458
CC#CCCCCO                 0.9630825
CCC#CCCO                  0.9669537
Or35a
CCC#CCCO                  0.150652
CCC═CCCOC(═O)C            0.153849
C═CCCCCO                  0.1577711
C(CCCCCC═C)O              0.1896431
CCC═CC═O                  0.1996007
CCC#CCCOC(═O)C            0.203439
CCCC#CCO                  0.2051737
CC(═O)OCCCCC═C            0.2169564
CC═CCCCO                  0.2327925
C(CCC═CC)OC(═O)C          0.2366404
C(CCC═C)O                 0.253342
CC(═O)OCCCC═C             0.2575001
C#CCCCCO                  0.258166
C(CCCCCCCC)O              0.262205
C(CCO)CCS                 0.2659421
CC═CC═CCOC(═O)C           0.277348
C(CC═C═O)CS               0.2857512
CC(═O)OCCCC═C═C           0.2908757
CCC═CCO                   0.2957919
CC(═O)OCCCCC#C            0.3021282
C#CCCCCCCO                0.3034524
CCCC#CC═O                 0.3056329
C═CCCCCCCCO               0.3066109
C(CC═CC)O                 0.3164314
CCC═CCCCCO                0.3186713
Or43a
CCCCCC(C#C)O              0.00332052
C1CC1CCCCO                0.00699056
CCC#CCC(CC)O              0.01572877
CCCCC(CC)O                0.01572877
CCCCCC(CC)O               0.01642782
CCCCC(C)CO                0.01782593
CCC#CC(CC)O               0.0180007
C═CCC(C═C)O               0.01835023
CCC#CCCO                  0.01887452
CC(═C)CCCO                0.01904928
C═CC(CCCC)O               0.02114645
CCCC#CCO                  0.0223698
CCC#CC(C)O                0.0223698
CCCCCC(C)O                0.0223698
CC═CC═CCO                 0.02341839
CCC═CCO                   0.02376791
CC1(CC1)CCCO              0.02499126
CC#CCCO                   0.02603985
CC═C═CC(C)O               0.02638937
C═CCCCC(C═C)O             0.0267389
CC(C)CCCCO                0.02743796
CCC(CCC(C)C)O             0.02743796
CCC(C#CC)O                0.02761272
CCCCC(C#C)O               0.02778749
CCCCC(C═CC)O              0.02778749
Or43b
CCCONC(═O)C               0.0959588
CCN(C(═O)C)O              0.1130635
CCOC(═O)SCC               0.1132685
CCNNC(═O)C                0.1183047
CC#CC(NC)O                0.1231371
CC═C═CC(C)(C)O            0.1294547
CCNCNC(═O)C               0.1317853
CCCC(O)OCC                0.1391272
CCC(═O)NCC                0.1435372
CCCC(═O)NCC               0.1476275
CCC(O)OCC                 0.1478191
CCNC(═O)OCC               0.1489972
CC#CC(═O)NC               0.1490445
CC(CN)NC(═O)C             0.1502223
CCNC(═O)C                 0.1502272
CCCC(═O)NC                0.1678103
CCOC(C)ON                 0.1752449
CC(C)OC(C)C#N             0.1764415
CC(═O)C═CCC               0.1765129
CC(CNC(═O)C)O             0.1766655
CC(C)C═CC(═O)C            0.1781594
CC═CC(═O)C                0.181488
CCOC(C)O                  0.1828469
CCOC(C)OC#N               0.1867745
CCNC(═O)NCC               0.1871594
Or47a
CCCCCCC(═O)C              0.00616096
CCC(CC)CCC(═O)C           0.00770119
CCCSCO                    0.00847131
CCCOC#N                   0.00924143
CCOCC#N                   0.00924143
CCCCC(═O)COC              0.01001155
CCCCCCC(═O)C═O            0.01078167
CCCCC(═O)CC(═O)C          0.01078167
CCCCC(═O)OC               0.01078167
CC(C)C(C)COC(═O)C         0.01155179
CC(C)CCC(═O)OC            0.01155179
CCCCCOC(═O)C═C            0.01232191
CCCCCOC(═O)CC             0.01232191
CC(C)COC#N                0.01309203
CCCCCCC(═O)OC             0.01309203
CCCCCCSSC                 0.01386215
CCCCC(C)CC(═O)C           0.01386215
CCCCCC(═O)C═C             0.01386215
CCCCCCC(═O)CO             0.01386215
CC(C)OCC#N                0.01386215
CCCCCC(═O)C═C             0.01386215
CC(═O)OCC(C)(C)C          0.01463227
CCCCSS(═O)C               0.01463227
CCCCCSSC                  0.01540239
CCCCCC(═O)OC═C            0.01540239
Or49b
C1═CC═C(C═C1)N═O          0.01680465
CC1═CC(═CC═C1)O           0.02191376
C1═CC═C(C(═C1)O)S         0.05712351
Cc1ccc(cc1)O              0.08002216
C1═CC(═CC═C1N)O           0.08291527
C1═CC═C(C═C1)N═C═O        0.08630082
C1═CC═C(C(═C1)N)O         0.08697793
C1═CC(═CC(═C1)O)N         0.08747038
C1═CC═C(C(═C1)O)C1        0.08876304
c1(ccccc1)NC═O            0.09744237
C1═CC(═CC═C1O)S           0.09756548
C1═CC═C(C═C1)C#CO         0.1029824
C1═CC═C(C═C1)C═C═O        0.1066141
CC1═C(C═CC═C1O)N          0.1069219
CC1═CC(═C(C═C1)C)O        0.1087686
CC1═C(C(═CC═C1)O)C        0.1101844
C═C(C1═CC═CC═C1)O         0.1105537
CC1═C(C═CC═C1S)O          0.1107999
C1═CC═C(C═C1)NN═O         0.1138777
CC1═C(C═C(C═C1)N)O        0.1143086
C1═CC(═CC(═C1)O)O         0.1145548
C1═CC═C(C═C1)C═CO         0.1148626
CC1═C(C(═CC═C1)O)S        0.1148626
C1═CC═C(C═C1)NO           0.1188637
C1═CC(═CC═C1O)O           0.1207719
Or59b
CCC(═O)OC(C)O             0.08309379
CCC(═O)OC                 0.08527063
C(C)OC(═O)CC              0.09857435
CCC(═O)COC                0.1112024
CCCOC(═O)C                0.1141674
CCCOC(═O)C                0.1141674
CCC(O)OC(═O)C             0.1244704
CC(COC(═O)C)O             0.1292618
CCC(C(═O)OC)O             0.1352768
CCC(N═C═O)OC              0.1459781
COC(═O)CC(O)C             0.1504875
CCCS(═O)OC                0.1531444
CCC(═O)C(O)OC             0.1567203
CCCC(═O)N(C)O             0.1589646
CC(CC(═O)C)O              0.1612506
CCC(═O)CC                 0.1613363
CC(N═C═O)OC               0.1654712
OC(C)C(═O)CC              0.165828
CC(O)S(═O)C               0.1659486
COCC(═O)C                 0.1665356
CCN(C)N═O                 0.1718486
CC(═O)CCOC                0.1721226
CCC#CC(═O)C               0.1737079
CCS(═O)OC                 0.1760367
CCC(═O)OOC                0.1770918
Or67a
C═CC(═O)C1═CC═CC═N1       0.3008233
C1═CC═C(C(═C1)CC#N)C═O    0.3080015
CCOC(═N)C1═CC═CC═C1       0.312236
COC(═N)C1═CC═CC═C1        0.3311703
CCC1═COC(═N1)CC           0.3703241
CCC(═O)C1═CC═CC═N1        0.3768891
C1═CC═C(C(═C1)C═O)C═O     0.3797241
CCOC(═O)N1C═CC═C1         0.3857737
C1═CC(═CC(═C1)C═O)C═O     0.3905579
CC1═CC(═CC═C1)CO          0.3917814
CCOCC1═CC═CC═C1           0.399528
COC(═O)N1C═CC═C1          0.4010939
COC(c1ccccc1)O            0.4035766
C1═CC═C(C═C1)C2═CC═NO2    0.4060794
c1(ccccc1)COC             0.4097667
COC1═NN═CC2═CC═CC═C21     0.4106803
COC(═O)C1═CC═CC═C1C#N     0.4139384
CC1═NOCC2═CC═CC═C12       0.4173282
COC1═NC═C(C═C1)C═C        0.419944
CC(═CCOC)C                0.4208605
CCC═CCC(═O)C              0.4243553
CC1═CC2═CC═CC═C2CO1       0.4298624
COC1═NN═C2N1C3═CC═CC═C3O2 0.4389819
CC1═CC═C(C═C1)C(═O)OC     0.4404134
C1═CC(═O)C═CC2═C1C═CO2    0.4413276
Or67c
CC(CC#C)O                 0.07067509
C═CCC(C═C═C)O             0.0775118
CC(C)(CC═C)O              0.0885166
C═CCCCCO                  0.09353587
CC(═C)CCCO                0.1018462
CC(CCCC)O                 0.1056086
CCC(CC═C)O                0.1068447
CCC#CCCCO                 0.1081803
CCC(CC#C)O                0.1259778
CC═C(C)C(C)O              0.1262036
CC═CCCCO                  0.1270505
CC(CC═C)CO                0.1274556
CCC(CCC)O                 0.1279088
CC(C)(CC#C)O              0.1294606
CC(CCO)C═C                0.1341464
CC═C═CCCCO                0.1372657
CC(C)C(C#C)O              0.1429075
CCC#CCC(C)O               0.1430763
CCC(═C)C(C)O              0.1438052
CCC(C)(CC═C)O             0.1477914
CCCCC(═C)CO               0.1527974
CCC(═CC)CO                0.1538365
CC(CC═C)O                 0.1609561
CCC(C(C)C)O               0.1618278
CCCC(C)CO                 0.1653694
Or85a
CCC#CCCO                  0.09486577
CC═CCCCO                  0.1241049
CCCC(═N)OCC               0.1455693
CCCCNC(═C)C               0.1695939
CCCC#CCO                  0.1791638
CCCC(═O)CCO               0.1893542
CC═C═CCCCO                0.1938411
CCN═C(C)CC(═C)O           0.1971383
CC═CC═CCCO                0.1971396
CCOC#CC(C)O               0.2069597
CCC#CCCCO                 0.2311878
CCCC(O)OCC                0.2512179
CCNC(C)OC═C               0.255045
CCC(═O)NCC(═O)C           0.2887675
CC(CCCO)C═C               0.289174
CC(C)(CCCOC)O             0.2891912
CCOC(═O)C(C)OC            0.294917
CCOC(C)OC(═C)O            0.297801
CCCNC(C)C═O               0.3026955
CC(C(C)OCC═C)O            0.3104233
CC═CC═CCO                 0.312425
OCCCCC(═O)C               0.3180066
CCCCC(═C)CO               0.3214983
CCC(═O)NCC(═C)C           0.331915
CCOC(C)C(O)OC             0.3407239
Or85b
CC(CCCC═C)CO              0.04010449
CCCCCCC═O                 0.0541304
CCC(═O)C═CCCC             0.05661388
CCCCC═CC(═O)C             0.05802127
C(CCCCCC═C)O              0.06257155
CCCCCC(═O)CC              0.06590403
CC(C(═O)C)CCCC            0.0741716
CN(CCCC═C)O               0.08071376
CC(CCC═C═C)C═O            0.08460999
CCCCC(═C)C═O              0.08656348
CCC(CCC#CC)O              0.0872067
CN(CCCCC═C)O              0.08897917
CCCCC1CCOC1═O             0.09145651
C═CCCCCO                  0.09536294
CCCCC═CCO                 0.09564121
CCC(CCC═C═C)O             0.0958081
CC(C)CCCCOC               0.09698922
CCCCCC(CC)O               0.0993378
CC(═CC)CCC(═O)C           0.1024115
CCCCC(C)C(C)O             0.1036823
CCCCC(C)C(C)O             0.1036823
CCCCC#CC(C)O              0.1038561
CC(C)CC═CC(═O)C           0.1081452
CCC(C)CCC(C)O             0.1082218
CC(CCC═C)N(C)O            0.1085992
Or85f
CC(CCC═C)O                0.3215251
CCCC#CCO                  0.3977383
CCC#CCCO                  0.4721775
CCCC(═O)OC(C)O            0.5291351
COC(═O)CC(O)C             0.5396708
CCCC(COC)O                0.5401751
CC(CCCC═C)O               0.574608
CC(CC(═O)OC═C)O           0.5760439
CC═CCCCO                  0.5830891
CCCC(═O)N(C)O             0.5926106
CN(CCCCO)N═O              0.6121783
CC(C)(C1C(O1)C#C)O        0.6193232
C(CC═CC)O                 0.6211374
CC(CCC═C═C)O              0.6297574
CCC(C(═O)OCC)O            0.6342998
CC(═O)OC(COC)O            0.6391873
CCCC1C(O1)CO              0.6497865
CCOC(═O)CC(C)O            0.6566422
CC(C#COC═C)O              0.663147
CN(CCCC═C)O               0.6794308
C═CC(CCCC)O               0.6991725
CCC(COC═C)O               0.7054714
CCCCCN(C)O                0.7137121
CC(COCCC#N)O              0.7174705
CC(CCO)C═C                0.7219598
Or98a
CC(CCCCO)C═C              0
CCCC(═O)OCNO              0
CC(CC═C═C(C)C)O           0
CNCC(═O)OCCO              0
CC(C)COCC(C)O             0.0006135
CC(═O)OC═CC═C             0.0006135
C═CCCCC(═C)CO             0.0006135
C═CCOCC(C═C)O             0.0006135
C(CN)C(═O)OCCO            0.0006135
C═CC(COCC#C)O             0.0006135
CCOCCC(C)O                0.0006135
CC(CC(═O)OCCO)O           0.0006135
CCOC(═O)CC(═O)C           0.0006135
CCOC#CC(C)O               0.00122699
CC(C)(C#CCN(C)C)O         0.00122699
CCC(═O)COCC               0.00122699
CC(═O)CC(═O)OCCO          0.00122699
CC(C)OC(═O)NCO            0.00122699
CCOC(═O)CS(═O)C           0.00122699
CC(CO)OCCC═C              0.00122699
CCCNOC(═O)CC              0.00122699
CCCC(═O)OC(═O)C           0.00122699
CC(═C)COCCOC              0.00122699
CC(═O)OC(═O)CC═C          0.00122699
CCOCOC(═O)C               0.00184049
ab1A
CC(═O)C(═O)C              0.02090025
CC(OCC(C)C)═O             0.05702
CC(═O)OCCC#C              0.07507874
CCC(═O)OOCC               0.07770847
CCOC(═O)OC                0.07784295
CCCOC(═O)CC               0.08788487
CCCOC(═O)OC               0.09563759
CCOC(═O)CC#C              0.09593777
CCC(═O)OCC#C              0.1027616
CC1CC(═O)OC1              0.1040388
COCCC(═O)OC               0.1130966
CCCC(OC)═O                0.11606
CCSCC(C)C═O               0.116945
CC1COC(═O)O1              0.1170738
CCCOC(═O)C                0.1177654
C(C)OC(═O)OCC             0.1181321
CC(C)OC═O                 0.1209444
CC(C)CCS(═O)C             0.1236193
CCC(C)OCC═O               0.1263356
CCOCS(═O)C                0.1272632
CCOC(═O)ONN               0.1288552
CCC(═O)OC                 0.1298152
CC1OCC(═O)O1              0.1309472
CC(═O)OCC#C               0.1313023
COC(═O)CCCS               0.132371

TABLE 5
Optimized descriptor sets for each Mammalian OR. Optimized descriptors occurrences, symbol, brief description, class,
and dimensionality are listed. Descriptors are listed in ascending order of when they were selected into the optimized
set. Weights indicate the number of times a descriptor was selected in an optimized descriptor set. A summary of
the total number of descriptors selected for the receptor repertoire is provided as the beginning.
Mammalian Descriptor Lists
Descriptor Class Type Counts for all Org
	GETAWAY descriptors	109
	atom-centred fragments	49
	2D autocorrelations	48
	RDF descriptors	48
	3D-MoRSE descriptors	46
	WHIM descriptors	43
	functional group counts	33
	2D binary fingerprints	26
	Burden eigenvalues	23
	edge adjacency indices	21
	geometrical descriptors	21
	topological descriptors	14
	2D frequency fingerprints	13
	topological charge indices	12
	atomtypes (Cerius2)	11
	molecular properties	11
	walk and path counts	7
	constitutional descriptors	6
	Randic molecular profiles	5
	topological (Cerius2)	4
	information indices	4
	connectivity indices	3
	structural (Cerius2)	1
	eigenvalue-based indices	1
	charge descriptors	0
Dimensionality Counts (Weights Included)
	Num zero dimensional descriptors:	7
	Num one dimensional descriptors:	104
	Num two dimensional descriptors:	176
	Num three dimensional descriptors:	272
Origin (Weights Included)
	Num Dragon descriptors:	546
	Num Cerius2 descriptors:	13
Dimensionality Counts (Weights Excluded)
	Num zero dimensional descriptors:	7
	Num one dimensional descriptors:	37
	Num two dimensional descriptors:	93
	Num three dimensional descriptors:	155
Origin (Weights Excluded)
	Num unique Dragon descriptors:	284
	Num unique Cerius2 descriptors:	8
MOR1.1
844	2	Mor17m	3D-MoRSE - signal 17/weighted by atomic masses	3D-MoRSE descriptors	3
1300	8	H-051	H attached to alpha-C	atom-centred fragments	1
1248	2	R6p+	R maximal autocorrelation of lag 6/weighted by atomic	GETAWAY descriptors	3
			polarizabilities
914	4	Mor23e	3D-MoRSE - signal 23/weighted by atomic Sanderson	3D-MoRSE descriptors	3
			electronegativities
857	5	Mor30m	3D-MoRSE - signal 30/weighted by atomic masses	3D-MoRSE descriptors	3
1211	4	R5v+	R maximal autocorrelation of lag 5/weighted by atomic van der	GETAWAY descriptors	3
			Waals volumes
923	1	Mor32e	3D-MoRSE - signal 32/weighted by atomic Sanderson	3D-MoRSE descriptors	3
			electronegativities
519	1	JGI7	mean topological charge index of order7	topological charge indices	2
1019	2	E1s	1st component accessibility directional WHIM index/weighted by	WHIM descriptors	3
			atomic electrotopological states
1254	1	nCt	number of total tertiary C(sp3)	functional group counts	1
1270	1	nArCO	number of ketones (aromatic)	functional group counts	1
1304	1	O-058	#NAME?	atom-centred fragments	1
1344	1	B07[C—O]	presence/absence of C—O at topological distance 07	2D binary fingerprints	2
302	2	GATS2m	Geary autocorrelation - lag 2/weighted by atomic masses	2D autocorrelations	2
756	1	RDF110e	Radial Distribution Function - 11.0/weighted by atomic Sanderson	RDF descriptors	3
			electronegativities
1262	1	nCconj	number of non-aromatic conjugated C(sp2)	functional group counts	1
1282	1	C-006	CH2RX	atom-centred fragments	1
1256	1	nCrs	number of ring secondary C(sp3)	functional group counts	1
307	1	GATS7m	Geary autocorrelation - lag 7/weighted by atomic masses	2D autocorrelations	2
1280	1	C-003	CHR3	atom-centred fragments	1
276	1	MATS8m	Moran autocorrelation - lag 8/weighted by atomic masses	2D autocorrelations	2
MOR106.1
948	1	Mor25p	3D-MoRSE - signal 25/weighted by atomic polarizabilities	3D-MoRSE descriptors	3
476	1	BEHe6	highest eigenvalue n. 6 of Burden matrix/weighted by atomic	Burden eigenvalues	2
			Sanderson electronegativities
212	1	IC1	information content index (neighborhood symmetry of 1-order)	information indices	2
1282	2	C-006	CH2RX	atom-centred fragments	1
1233	2	RTe+	R maximal index/weighted by atomic Sanderson	GETAWAY descriptors	3
			electronegativities
147	1	piPC07	molecular multiple path count of order 07	walk and path counts	2
743	1	RDF045e	Radial Distribution Function - 4.5/weighted by atomic Sanderson	RDF descriptors	3
			electronegativities
1266	1	nRCOOH	number of carboxylic acids (aliphatic)	functional group counts	1
1213	1	R7v+	R maximal autocorrelation of lag 7/weighted by atomic van der	GETAWAY descriptors	3
			Waals volumes
630	1	HOMT	HOMA total	geometrical descriptors	3
683	1	RDF035m	Radial Distribution Function - 3.5/weighted by atomic masses	RDF descriptors	3
1298	1	H-049	H attached to C3(sp3)/C2(sp2)/C3(sp2)/C3(sp)	atom-centred fragments	1
608	1	SHP2	average shape profile index of order 2	Randic molecular profiles	3
MOR107.1
1255	1	nCq	number of total quaternary C(sp3)	functional group counts	1
	5
866	1	Mor07v	3D-MoRSE - signal 07/weighted by atomic van der Waals	3D-MoRSE descriptors	3
			volumes
465	1	BELv3	lowest eigenvalue n. 3 of Burden matrix/weighted by atomic van	Burden eigenvalues	2
			der Waals volumes
1246	1	R4p+	R maximal autocorrelation of lag 4/weighted by atomic	GETAWAY descriptors	3
			polarizabilities
964	1	E1u	1st component accessibility directional WHIM index/unweighted	WHIM descriptors	3
516	1	JGI4	mean topological charge index of order4	topological charge indices	2
635	1	DISPv	d COMMA2 value/weighted by atomic van der Waals volumes	geometrical descriptors	3
29	1	nR06	number of 6-membered rings	constitutional descriptors	0
684	1	RDF040m	Radial Distribution Function - 4.0/weighted by atomic masses	RDF descriptors	3
1232	1	R8e+	R maximal autocorrelation of lag 8/weighted by atomic	GETAWAY descriptors	3
			Sanderson electronegativities
147	1	piPC07	molecular multiple path count of order 07	walk and path counts	2
1012	1	L2s	2nd component size directional WHIM index/weighted by atomic	WHIM descriptors	3
			electrotopological states
148	1	piPC08	molecular multiple path count of order 08	walk and path counts	2
22	1	nDB	number of double bonds	constitutional descriptors	0
1300	1	H-051	H attached to alpha-C	atom-centred fragments	1
975	1	E1m	1st component accessibility directional WHIM index/weighted by	WHIM descriptors	3
			atomic masses
1338	1	B05[C—O]	presence/absence of C—O at topological distance 05	2D binary fingerprints	2
497	1	BELp3	lowest eigenvalue n. 3 of Burden matrix/weighted by atomic	Burden eigenvalues	2
			polarizabilities
MOR129.1
1045	1	Dv	D total accessibility index/weighted by atomic van der Waals	WHIM descriptors	3
			volumes
1136	2	HATS7e	leverage-weighted autocorrelation of lag 7/weighted by atomic	GETAWAY descriptors	3
			Sanderson electronegativities
1355	1	F01[C—C]	frequency of C—C at topological distance 01	2D frequency fingerprints	2
805	1	Mor10u	3D-MoRSE - signal 10/unweighted	3D-MoRSE descriptors	3
1094	1	HATS5m	leverage-weighted autocorrelation of lag 5/weighted by atomic	GETAWAY descriptors	3
			masses
1100	2	H1v	H autocorrelation of lag 1/weighted by atomic van der Waals	GETAWAY descriptors	3
			volumes
870	1	Mor11v	3D-MoRSE - signal 11/weighted by atomic van der Waals	3D-MoRSE descriptors	3
			volumes
1337	1	B05[C—C]	presence/absence of C—C at topological distance 05	2D binary fingerprints	2
751	1	RDF085e	Radial Distribution Function - 8.5/weighted by atomic Sanderson	RDF descriptors	3
			electronegativities
1044	1	Dm	D total accessibility index/weighted by atomic masses	WHIM descriptors	3
1079	1	H0m	H autocorrelation of lag 0/weighted by atomic masses	GETAWAY descriptors	3
901	1	Mor10e	3D-MoRSE - signal 10/weighted by atomic Sanderson	3D-MoRSE descriptors	3
			electronegativities
107	1	D/Dr06	distance/detour ring index of order 6	topological descriptors	2
1095	1	HATS6m	leverage-weighted autocorrelation of lag 6/weighted by atomic	GETAWAY descriptors	3
			masses
297	1	MATS5p	Moran autocorrelation - lag 5/weighted by atomic polarizabilities	2D autocorrelations	2
683	1	RDF035m	Radial Distribution Function - 3.5/weighted by atomic masses	RDF descriptors	3
1126	1	H7e	H autocorrelation of lag 7/weighted by atomic Sanderson	GETAWAY descriptors	3
			electronegativities
1099	1	H0v	H autocorrelation of lag 0/weighted by atomic van der Waals	GETAWAY descriptors	3
			volumes
1184	1	R5m	R autocorrelation of lag 5/weighted by atomic masses	GETAWAY descriptors	3
MOR136.1
682	1	RDF030m	Radial Distribution Function - 3.0/weighted by atomic masses	RDF descriptors	3
1466	1	S_dssC	S_dssC	atomtypes (cerius2)	1
832	1	Mor05m	3D-MoRSE - signal 05/weighted by atomic masses	3D-MoRSE descriptors	3
479	1	BELe1	lowest eigenvalue n. 1 of Burden matrix/weighted by atomic	Burden eigenvalues	2
			Sanderson electronegativities
1175	1	R5u+	R maximal autocorrelation of lag 5/unweighted	GETAWAY descriptors	3
608	1	SHP2	average shape profile index of order 2	Randic molecular profiles	3
MOR139.1
1100	2	H1v	H autocorrelation of lag 1/weighted by atomic van der Waals	GETAWAY descriptors	3
			volumes
1070	1	HATS1u	leverage-weighted autocorrelation of lag 1/unweighted	GETAWAY descriptors	3
1310	1	TPSA(NO)	topological polar surface area using N, O polar contributions	molecular properties	1
146	1	piPC06	molecular multiple path count of order 06	walk and path counts	2
1087	1	H8m	H autocorrelation of lag 8/weighted by atomic masses	GETAWAY descriptors	3
1316	1	GVWAI-80	Ghose-Viswanadhan-Wendoloski drug-like index at 80%	molecular properties	1
1198	1	R1v	R autocorrelation of lag 1/weighted by atomic van der Waals	GETAWAY descriptors	3
			volumes
302	1	GATS2m	Geary autocorrelation - lag 2/weighted by atomic masses	2D autocorrelations	2
915	1	Mor24e	3D-MoRSE - signal 24/weighted by atomic Sanderson	3D-MoRSE descriptors	3
			electronegativities
358	1	EEig09d	Eigenvalue 09 from edge adj. matrix weighted by dipole moments	edge adjacency indices	2
MOR162.1
627	1	HOMA	Harmonic Oscillator Model of Aromaticity index	geometrical descriptors	3
1094	1	HATS5m	leverage-weighted autocorrelation of lag 5/weighted by atomic	GETAWAY descriptors	3
			masses
998	1	E2e	2nd component accessibility directional WHIM index/weighted by	WHIM descriptors	3
			atomic Sanderson electronegativities
1121	1	H2e	H autocorrelation of lag 2/weighted by atomic Sanderson	GETAWAY descriptors	3
			electronegativities
1212	1	R6v+	R maximal autocorrelation of lag 6/weighted by atomic van der	GETAWAY descriptors	3
			Waals volumes
993	1	P2e	2nd component shape directional WHIM index/weighted by	WHIM descriptors	3
			atomic Sanderson electronegativities
297	1	MATS5p	Moran autocorrelation - lag 5/weighted by atomic polarizabilities	2D autocorrelations	2
628	1	RCI	Jug RC index	geometrical descriptors	3
1095	1	HATS6m	leverage-weighted autocorrelation of lag 6/weighted by atomic	GETAWAY descriptors	3
			masses
683	1	RDF035m	Radial Distribution Function - 3.5/weighted by atomic masses	RDF descriptors	3
1120	1	H1e	H autocorrelation of lag 1/weighted by atomic Sanderson	GETAWAY descriptors	3
			electronegativities
MOR170.1
1290	3	C-025	R—CR—R	atom-centred fragments	1
1371	1	F05[C—O]	frequency of C—O at topological distance 05	2D frequency fingerprints	2
1212	1	R6v+	R maximal autocorrelation of lag 6/weighted by atomic van der	GETAWAY descriptors	3
			Waals volumes
998	2	E2e	2nd component accessibility directional WHIM index/weighted by	WHIM descriptors	3
			atomic Sanderson electronegativities
1464	1	S_aaCH	S_aaCH	atomtypes (cerius2)	1
1233	2	RTe+	R maximal index/weighted by atomic Sanderson	GETAWAY descriptors	3
			electronegativities
1178	1	R8u+	R maximal autocorrelation of lag 8/unweighted	GETAWAY descriptors	3
262	1	ATS2p	Broto-Moreau autocorrelation of a topological structure - lag 2/	2D autocorrelations	2
			weighted by atomic polarizabilities
297	1	MATS5p	Moran autocorrelation - lag 5/weighted by atomic polarizabilities	2D autocorrelations	2
714	1	RDF045v	Radial Distribution Function - 4.5/weighted by atomic van der	RDF descriptors	3
			Waals volumes
1004	1	P2p	2nd component shape directional WHIM index/weighted by	WHIM descriptors	3
			atomic polarizabilities
1249	1	R7p+	R maximal autocorrelation of lag 7/weighted by atomic	GETAWAY descriptors	3
			polarizabilities
1184	1	R5m	R autocorrelation of lag 5/weighted by atomic masses	GETAWAY descriptors	3
627	1	HOMA	Harmonic Oscillator Model of Aromaticity index	geometrical descriptors	3
MOR184.1
1461	1	S_dCH2	S_dCH2	atomtypes (cerius2)	1
301	1	GATS1m	Geary autocorrelation - lag 1/weighted by atomic masses	2D autocorrelations	2
1297	1	H-047	H attached to C1(sp3)/C0(sp2)	atom-centred fragments	1
37	1	Qindex	Quadratic index	topological descriptors	2
635	1	DISPv	d COMMA2 value/weighted by atomic van der Waals volumes	geometrical descriptors	3
979	1	L2v	2nd component size directional WHIM index/weighted by atomic	WHIM descriptors	3
			van der Waals volumes
18	1	nCIC	number of rings	constitutional descriptors	0
1111	1	HATS2v	leverage-weighted autocorrelation of lag 2/weighted by atomic	GETAWAY descriptors	3
			van der Waals volumes
802	1	Mor07u	3D-MoRSE - signal 07/unweighted	3D-MoRSE descriptors	3
1222	1	R7e	R autocorrelation of lag 7/weighted by atomic Sanderson	GETAWAY descriptors	3
			electronegativities
136	1	MPC06	molecular path count of order 06	walk and path counts	2
373	1	EEig09r	Eigenvalue 09 from edge adj. matrix weighted by resonance	edge adjacency indices	2
			integrals
19	1	nCIR	number of circuits	constitutional descriptors	0
685	1	RDF045m	Radial Distribution Function - 4.5/weighted by atomic masses	RDF descriptors	3
497	1	BELp3	lowest eigenvalue n. 3 of Burden matrix/weighted by atomic	Burden eigenvalues	2
			polarizabilities
358	1	EEig09d	Eigenvalue 09 from edge adj. matrix weighted by dipole moments	edge adjacency indices	2
1001	1	L2p	2nd component size directional WHIM index/weighted by atomic	WHIM descriptors	3
			polarizabilities
1156	1	HATS7p	leverage-weighted autocorrelation of lag 7/weighted by atomic	GETAWAY descriptors	3
			polarizabilities
1246	1	R4p+	R maximal autocorrelation of lag 4/weighted by atomic	GETAWAY descriptors	3
			polarizabilities
837	1	Mor10m	3D-MoRSE - signal 10/weighted by atomic masses	3D-MoRSE descriptors	3
MOR185.1
103	1	BAC	Balaban centric index	topological descriptors	2
1091	1	HATS2m	leverage-weighted autocorrelation of lag 2/weighted by atomic	GETAWAY descriptors	3
			masses
1178	1	R8u+	R maximal autocorrelation of lag 8/unweighted	GETAWAY descriptors	3
168	1	X5A	average connectivity index chi-5	connectivity indices	2
997	1	E1e	1st component accessibility directional WHIM index/weighted by	WHIM descriptors	3
			atomic Sanderson electronegativities
1233	1	RTe+	R maximal index/weighted by atomic Sanderson	GETAWAY descriptors	3
			electronegativities
998	1	E2e	2nd component accessibility directional WHIM index/weighted by	WHIM descriptors	3
			atomic Sanderson electronegativities
302	1	GATS2m	Geary autocorrelation - lag 2/weighted by atomic masses	2D autocorrelations	2
1140	1	H1p	H autocorrelation of lag 1/weighted by atomic polarizabilities	GETAWAY descriptors	3
1156	1	HATS7p	leverage-weighted autocorrelation of lag 7/weighted by atomic	GETAWAY descriptors	3
			polarizabilities
683	1	RDF035m	Radial Distribution Function - 3.5/weighted by atomic masses	RDF descriptors	3
608	1	SHP2	average shape profile index of order 2	Randic molecular profiles	3
1244	1	R2p+	R maximal autocorrelation of lag 2/weighted by atomic	GETAWAY descriptors	3
			polarizabilities
MOR189.1
1256	1	nCrs	number of ring secondary C(sp3)	functional group counts	1
1457	1	V-DIST-	V-DIST-mag	topological (cerius2)	2
		mag
610	1	J3D	3D-Balaban index	geometrical descriptors	3
1413	2	Atype_C_40	Number of Carbon Type 40	atomtypes (Cerius2)	1
375	1	EEig11r	Eigenvalue 11 from edge adj. matrix weighted by resonance	edge adjacency indices	2
			integrals
1183	1	R4m	R autocorrelation of lag 4/weighted by atomic masses	GETAWAY descriptors	3
930	1	Mor07p	3D-MoRSE - signal 07/weighted by atomic polarizabilities	3D-MoRSE descriptors	3
1316	1	GVWAI-80	Ghose-Viswanadhan-Wendoloski drug-like index at 80%	molecular properties	1
681	1	RDF025m	Radial Distribution Function - 2.5/weighted by atomic masses	RDF descriptors	3
1343	1	B07[C—C]	presence/absence of C—C at topological distance 07	2D binary fingerprints	2
1174	1	R4u+	R maximal autocorrelation of lag 4/unweighted	GETAWAY descriptors	3
913	1	Mor22e	3D-MoRSE - signal 22/weighted by atomic Sanderson	3D-MoRSE descriptors	3
			electronegativities
1304	1	O-058	#NAME?	atom-centred fragments	1
356	1	EEig07d	Eigenvalue 07 from edge adj. matrix weighted by dipole moments	edge adjacency indices	2
360	1	EEig11d	Eigenvalue 11 from edge adj. matrix weighted by dipole moments	edge adjacency indices	2
MOR2.1
685	1	RDF045m	Radial Distribution Function - 4.5/weighted by atomic masses	RDF descriptors	3
1316	2	GVWAI-80	Ghose-Viswanadhan-Wendoloski drug-like index at 80%	molecular properties	1
485	1	BELe7	lowest eigenvalue n. 7 of Burden matrix/weighted by atomic	Burden eigenvalues	2
			Sanderson electronegativities
686	1	RDF050m	Radial Distribution Function - 5.0/weighted by atomic masses	RDF descriptors	3
905	1	Mor14e	3D-MoRSE - signal 14/weighted by atomic Sanderson	3D-MoRSE descriptors	3
			electronegativities
346	1	EEig12x	Eigenvalue 12 from edge adj. matrix weighted by edge degrees	edge adjacency indices	2
843	1	Mor16m	3D-MoRSE - signal 16/weighted by atomic masses	3D-MoRSE descriptors	3
376	2	EEig12r	Eigenvalue 12 from edge adj. matrix weighted by resonance	edge adjacency indices	2
			integrals
949	1	Mor26p	3D-MoRSE - signal 26/weighted by atomic polarizabilities	3D-MoRSE descriptors	3
804	1	Mor09u	3D-MoRSE - signal 09/unweighted	3D-MoRSE descriptors	3
1262	1	nCconj	number of non-aromatic conjugated C(sp2)	functional group counts	1
845	1	Mor18m	3D-MoRSE - signal 18/weighted by atomic masses	3D-MoRSE descriptors	3
1173	1	R3u+	R maximal autocorrelation of lag 3/unweighted	GETAWAY descriptors	3
1344	1	B07[C—O]	presence/absence of C—O at topological distance 07	2D binary fingerprints	2
1358	1	F02[C—C]	frequency of C—C at topological distance 02	2D frequency fingerprints	2
MOR203.1
1340	6	B06[C—C]	presence/absence of C—C at topological distance 06	2D binary fingerprints	2
1298	1	H-049	H attached to C3(sp3)/C2(sp2)/C3(sp2)/C3(sp)	atom-centred fragments	1
931	1	Mor08p	3D-MoRSE - signal 08/weighted by atomic polarizabilities	3D-MoRSE descriptors	3
1390	1	Hbond	Number of Hydrogen bond acceptors	structural (Cerius2)	0
		acceptor
661	1	RDF075u	Radial Distribution Function - 7.5/unweighted	RDF descriptors	3
1203	1	R6v	R autocorrelation of lag 6/weighted by atomic van der Waals	GETAWAY descriptors	3
			volumes
1268	3	nRCHO	number of aldehydes (aliphatic)	functional group counts	1
1266	2	nRCOOH	number of carboxylic acids (aliphatic)	functional group counts	1
272	1	MATS4m	Moran autocorrelation - lag 4/weighted by atomic masses	2D autocorrelations	2
1018	1	G3s	3st component symmetry directional WHIM index/weighted by	WHIM descriptors	3
			atomic electrotopological states
106	1	D/Dr05	distance/detour ring index of order 5	topological descriptors	2
1270	1	nArCO	number of ketones (aromatic)	functional group counts	1
1352	1	B10[C-C]	presence/absence of C-C at topological distance 10	2D binary fingerprints	2
274	1	MATS6m	Moran autocorrelation - lag 6/weighted by atomic masses	2D autocorrelations	2
445	1	BEHm7	highest eigenvalue n. 7 of Burden matrix/weighted by atomic	Burden eigenvalues	2
			masses
80	1	MAXDN	maximal electrotopological negative variation	topological descriptors	2
1012	1	L2s	2nd component size directional WHIM index/weighted by atomic	WHIM descriptors	3
			electrotopological states
481	1	BELe3	lowest eigenvalue n. 3 of Burden matrix/weighted by atomic	Burden eigenvalues	2
			Sanderson electronegativities
665	1	RDF095u	Radial Distribution Function - 9.5/unweighted	RDF descriptors	3
MOR204.6
1262	1	nCconj	number of non-aromatic conjugated C(sp2)	functional group counts	1
1463	1	S_dsCH	S_dsCH	atomtypes (cerius2)	1
1092	2	HATS3m	leverage-weighted autocorrelation of lag 3/weighted by atomic	GETAWAY descriptors	3
			masses
635	1	DISPv	d COMMA2 value/weighted by atomic van der Waals volumes	geometrical descriptors	3
1174	1	R4u+	R maximal autocorrelation of lag 4/unweighted	GETAWAY descriptors	3
107	2	D/Dr06	distance/detour ring index of order 6	topological descriptors	2
1185	1	R6m	R autocorrelation of lag 6/weighted by atomic masses	GETAWAY descriptors	3
837	1	Mor10m	3D-MoRSE - signal 10/weighted by atomic masses	3D-MoRSE descriptors	3
373	1	EEig09r	Eigenvalue 09 from edge adj. matrix weighted by resonance	edge adjacency indices	2
			integrals
1222	1	R7e	R autocorrelation of lag 7/weighted by atomic Sanderson	GETAWAY descriptors	3
			electronegativities
1173	1	R3u+	R maximal autocorrelation of lag 3/unweighted	GETAWAY descriptors	3
1199	1	R2v	R autocorrelation of lag 2/weighted by atomic van der Waals	GETAWAY descriptors	3
			volumes
1371	1	F05[C—O]	frequency of C—O at topological distance 05	2D frequency fingerprints	2
1136	1	HATS7e	leverage-weighted autocorrelation of lag 7/weighted by atomic	GETAWAY descriptors	3
			Sanderson electronegativities
MOR207.1
1290	3	C-025	R—CR—R	atom-centred fragments	1
1371	1	F05[C—O]	frequency of C—O at topological distance 05	2D frequency fingerprints	2
1212	1	R6v+	R maximal autocorrelation of lag 6/weighted by atomic van der	GETAWAY descriptors	3
			Waals volumes
998	2	E2e	2nd component accessibility directional WHIM index/weighted by	WHIM descriptors	3
			atomic Sanderson electronegativities
1464	1	S_aaCH	S_aaCH	atomtypes (cerius2)	1
1233	2	RTe+	R maximal index/weighted by atomic Sanderson	GETAWAY descriptors	3
			electronegativities
1178	1	R8u+	R maximal autocorrelation of lag 8/unweighted	GETAWAY descriptors	3
262	1	ATS2p	Broto-Moreau autocorrelation of a topological structure - lag 2/	2D autocorrelations	2
			weighted by atomic polarizabilities
297	1	MATS5p	Moran autocorrelation - lag 5/weighted by atomic polarizabilities	2D autocorrelations	2
714	1	RDF045v	Radial Distribution Function - 4.5/weighted by atomic van der	RDF descriptors	3
			Waals volumes
1004	1	P2p	2nd component shape directional WHIM index/weighted by	WHIM descriptors	3
			atomic polarizabilities
1249	1	R7p+	R maximal autocorrelation of lag 7/weighted by atomic	GETAWAY descriptors	3
			polarizabilities
1184	1	R5m	R autocorrelation of lag 5/weighted by atomic masses	GETAWAY descriptors	3
627	1	HOMA	Harmonic Oscillator Model of Aromaticity index	geometrical descriptors	3
1213	1	R7v+	R maximal autocorrelation of lag 7/weighted by atomic van der	GETAWAY descriptors	3
			Waals volumes
1140	1	H1p	H autocorrelation of lag 1/weighted by atomic polarizabilities	GETAWAY descriptors	3
MOR273.1
1015	1	P2s	2nd component shape directional WHIM index/weighted by	WHIM descriptors	3
			atomic electrotopological states
77	2	Jhetv	Balaban-type index from van der Waals weighted distance matrix	topological descriptors	2
305	1	GATS5m	Geary autocorrelation - lag 5/weighted by atomic masses	2D autocorrelations	2
1070	1	HATS1u	leverage-weighted autocorrelation of lag 1/unweighted	GETAWAY descriptors	3
815	1	Mor20u	3D-MoRSE - signal 20/unweighted	3D-MoRSE descriptors	3
518	1	JGI6	mean topological charge index of order6	topological charge indices	2
1216	1	R1e	R autocorrelation of lag 1/weighted by atomic Sanderson	GETAWAY descriptors	3
			electronegativities
827	1	Mor32u	3D-MoRSE - signal 32/unweighted	3D-MoRSE descriptors	3
372	1	EEig08r	Eigenvalue 08 from edge adj. matrix weighted by resonance	edge adjacency indices	2
			integrals
441	1	BEHm3	highest eigenvalue n. 3 of Burden matrix/weighted by atomic	Burden eigenvalues	2
			masses
MOR250.1
1045	1	Dv	D total accessibility index/weighted by atomic van der Waals	WHIM descriptors	3
			volumes
1297	6	H-047	H attached to C1(sp3)/C0(sp2)	atom-centred fragments	1
443	1	BEHm5	highest eigenvalue n. 5 of Burden matrix/weighted by atomic	Burden eigenvalues	2
			masses
1282	3	C-006	CH2RX	atom-centred fragments	1
297	2	MATS5p	Moran autocorrelation - lag 5/weighted by atomic polarizabilities	2D autocorrelations	2
1303	1	O-057	phenol/enol/carboxyl OH	atom-centred fragments	1
107	2	D/Dr06	distance/detour ring index of order 6	topological descriptors	2
947	2	Mor24p	3D-MoRSE - signal 24/weighted by atomic polarizabilities	3D-MoRSE descriptors	3
1014	3	P1s	1st component shape directional WHIM index/weighted by atomic	WHIM descriptors	3
			electrotopological states
356	1	EEig07d	Eigenvalue 07 from edge adj. matrix weighted by dipole moments	edge adjacency indices	2
1249	1	R7p+	R maximal autocorrelation of lag 7/weighted by atomic	GETAWAY descriptors	3
			polarizabilities
986	3	E1v	1st component accessibility directional WHIM index/weighted by	WHIM descriptors	3
			atomic van der Waals volumes
1012	2	L2s	2nd component size directional WHIM index/weighted by atomic	WHIM descriptors	3
			electrotopological states
901	2	Mor10e	3D-MoRSE - signal 10/weighted by atomic Sanderson	3D-MoRSE descriptors	3
			electronegativities
1100	1	H1v	H autocorrelation of lag 1/weighted by atomic van der Waals	GETAWAY descriptors	3
			volumes
1183	4	R4m	R autocorrelation of lag 4/weighted by atomic masses	GETAWAY descriptors	3
683	1	RDF035m	Radial Distribution Function - 3.5/weighted by atomic masses	RDF descriptors	3
447	1	BELm1	lowest eigenvalue n. 1 of Burden matrix/weighted by atomic	Burden eigenvalues	2
			masses
1096	2	HATS7m	leverage-weighted autocorrelation of lag 7/weighted by atomic	GETAWAY descriptors	3
			masses
1367	1	F04[C—O]	frequency of C—O at topological distance 04	2D frequency fingerprints	2
1336	1	B04[O—O]	presence/absence of O—O at topological distance 04	2D binary fingerprints	2
1337	1	B05[C—C]	presence/absence of C—C at topological distance 05	2D binary fingerprints	2
1280	2	C-003	CHR3	atom-centred fragments	1
1140	3	H1p	H autocorrelation of lag 1/weighted by atomic polarizabilities	GETAWAY descriptors	3
838	2	Mor11m	3D-MoRSE - signal 11/weighted by atomic masses	3D-MoRSE descriptors	3
341	1	EEig07x	Eigenvalue 07 from edge adj. matrix weighted by edge degrees	edge adjacency indices	2
1316	3	GVWAI-80	Ghose-Viswanadhan-Wendoloski drug-like index at 80%	molecular properties	1
519	2	JGI7	mean topological charge index of order7	topological charge indices	2
147	3	piPC07	molecular multiple path count of order 07	walk and path counts	2
30	1	nR09	number of 9-membered rings	constitutional descriptors	0
776	1	RDF060p	Radial Distribution Function - 6.0/weighted by atomic	RDF descriptors	3
			polarizabilities
1266	1	nRCOOH	number of carboxylic acids (aliphatic)	functional group counts	1
837	1	Mor10m	3D-MoRSE - signal 10/weighted by atomic masses	3D-MoRSE descriptors	3
302	1	GATS2m	Geary autocorrelation - lag 2/weighted by atomic masses	2D autocorrelations	2
479	2	BELe1	lowest eigenvalue n. 1 of Burden matrix/weighted by atomic	Burden eigenvalues	2
			Sanderson electronegativities
212	1	IC1	information content index (neighborhood symmetry of 1-order)	information indices	2
272	1	MATS4m	Moran autocorrelation - lag 4/weighted by atomic masses	2D autocorrelations	2
1274	1	nArOR	number of ethers (aromatic)	functional group counts	1
106	1	D/Dr05	distance/detour ring index of order 5	topological descriptors	2
658	1	RDF060u	Radial Distribution Function - 6.0/unweighted	RDF descriptors	3
MOR256.17
1452	7	BIC	BIC	topological (cerius2)	2
335	1	EEig01x	Eigenvalue 01 from edge adj. matrix weighted by edge degrees	edge adjacency indices	2
1095	6	HATS6m	leverage-weighted autocorrelation of lag 6/weighted by atomic	GETAWAY descriptors	3
			masses
1272	5	nOHp	number of primary alcohols	functional group counts	1
1465	3	S_sssCH	S_sssCH	atomtypes (cerius2)	1
1270	3	nArCO	number of ketones (aromatic)	functional group counts	1
1298	3	H-049	H attached to C3(sp3)/C2(sp2)/C3(sp2)/C3(sp)	atom-centred fragments	1
1265	3	nR = Ct	number of aliphatic tertiary C(sp2)	functional group counts	1
1088	3	HTm	H total index/weighted by atomic masses	GETAWAY descriptors	3
889	1	Mor30v	3D-MoRSE - signal 30/weighted by atomic van der Waals	3D-MoRSE descriptors	3
			volumes
306	2	GATS6m	Geary autocorrelation - lag 6/weighted by atomic masses	2D autocorrelations	2
702	1	RDF130m	Radial Distribution Function - 13.0/weighted by atomic masses	RDF descriptors	3
742	2	RDF040e	Radial Distribution Function - 4.0/weighted by atomic Sanderson	RDF descriptors	3
			electronegativities
31	1	nR10	number of 10-membered rings	constitutional descriptors	0
1351	1	B09[C—S]	presence/absence of C—S at topological distance 09	2D binary fingerprints	2
1283	1	C-008	CHR2X	atom-centred fragments	1
168	1	X5A	average connectivity index chi-5	connectivity indices	2
275	1	MATS7m	Moran autocorrelation - lag 7/weighted by atomic masses	2D autocorrelations	2
883	1	Mor24v	3D-MoRSE - signal 24/weighted by atomic van der Waals	3D-MoRSE descriptors	3
			volumes
918	1	Mor27e	3D-MoRSE - signal 27/weighted by atomic Sanderson	3D-MoRSE descriptors	3
			electronegativities
358	1	EEig09d	Eigenvalue 09 from edge adj. matrix weighted by dipole moments	edge adjacency indices	2
MOR258.1
1198	3	R1v	R autocorrelation of lag 1/weighted by atomic van der Waals	GETAWAY descriptors	3
			volumes
448	1	BELm2	lowest eigenvalue n. 2 of Burden matrix/weighted by atomic	Burden eigenvalues	2
			masses
1140	2	H1p	H autocorrelation of lag 1/weighted by atomic polarizabilities	GETAWAY descriptors	3
964	1	E1u	1st component accessibility directional WHIM index/unweighted	WHIM descriptors	3
1091	1	HATS2m	leverage-weighted autocorrelation of lag 2/weighted by atomic	GETAWAY descriptors	3
			masses
514	1	JGI2	mean topological charge index of order2	topological charge indices	2
1234	1	R1p	R autocorrelation of lag 1/weighted by atomic polarizabilities	GETAWAY descriptors	3
1340	1	B06[C—C]	presence/absence of C—C at topological distance 06	2D binary fingerprints	2
1012	1	L2s	2nd component size directional WHIM index/weighted by atomic	WHIM descriptors	3
			electrotopological states
631	1	DISPm	d COMMA2 value/weighted by atomic masses	geometrical descriptors	3
608	1	SHP2	average shape profile index of order 2	Randic molecular profiles	3
1060	1	H1u	H autocorrelation of lag 1/unweighted	GETAWAY descriptors	3
1015	1	P2s	2nd component shape directional WHIM index/weighted by	WHIM descriptors	3
			atomic electrotopological states
MOR259.1
1261	1	nCb−	number of substituted benzene C(sp2)	functional group counts	1
1018	1	G3s	3st component symmetry directional WHIM index/weighted by	WHIM descriptors	3
			atomic electrotopological states
1183	1	R4m	R autocorrelation of lag 4/weighted by atomic masses	GETAWAY descriptors	3
136	1	MPC06	molecular path count of order 06	walk and path counts	2
635	1	DISPv	d COMMA2 value/weighted by atomic van der Waals volumes	geometrical descriptors	3
1234	1	R1p	R autocorrelation of lag 1/weighted by atomic polarizabilities	GETAWAY descriptors	3
1371	1	F05[C—O]	frequency of C—O at topological distance 05	2D frequency fingerprints	2
1208	1	R2v+	R maximal autocorrelation of lag 2/weighted by atomic van der	GETAWAY descriptors	3
			Waals volumes
964	1	E1u	1st component accessibility directional WHIM index/unweighted	WHIM descriptors	3
302	1	GATS2m	Geary autocorrelation - lag 2/weighted by atomic masses	2D autocorrelations	2
998	1	E2e	2nd component accessibility directional WHIM index/weighted by	WHIM descriptors	3
			atomic Sanderson electronegativities
1060	1	H1u	H autocorrelation of lag 1/unweighted	GETAWAY descriptors	3
MOR260.1
727	1	RDF110v	Radial Distribution Function - 11.0/weighted by atomic van der	RDF descriptors	3
			Waals volumes
1190	2	R2m+	R maximal autocorrelation of lag 2/weighted by atomic masses	GETAWAY descriptors	3
520	1	JGI8	mean topological charge index of order8	topological charge indices	2
1308	1	Hy	hydrophilic factor	molecular properties	1
1302	1	O-056	alcohol	atom-centred fragments	1
1299	1	H-050	H attached to heteroatom	atom-centred fragments	1
276	1	MATS8m	Moran autocorrelation - lag 8/weighted by atomic masses	2D autocorrelations	2
750	1	RDF080e	Radial Distribution Function - 8.0/weighted by atomic Sanderson	RDF descriptors	3
			electronegativities
1095	1	HATS6m	leverage-weighted autocorrelation of lag 6/weighted by atomic	GETAWAY descriptors	3
			masses
MOR261.1
756	1	RDF110e	Radial Distribution Function - 11.0/weighted by atomic Sanderson	RDF descriptors	3
			electronegativities
1282	2	C-006	CH2RX	atom-centred fragments	1
720	1	RDF075v	Radial Distribution Function - 7.5/weighted by atomic van der	RDF descriptors	3
			Waals volumes
665	1	RDF095u	Radial Distribution Function - 9.5/unweighted	RDF descriptors	3
631	1	DISPm	d COMMA2 value/weighted by atomic masses	geometrical descriptors	3
1278	1	C-001	CH3R/CH4	atom-centred fragments	1
446	1	BEHm8	highest eigenvalue n. 8 of Burden matrix/weighted by atomic	Burden eigenvalues	2
			masses
727	1	RDF110v	Radial Distribution Function - 11.0/weighted by atomic van der	RDF descriptors	3
			Waals volumes
MOR268.1
260	2	ATS8e	Broto-Moreau autocorrelation of a topological structure - lag 8/	2D autocorrelations	2
			weighted by atomic Sanderson electronegativities
1282	1	C-006	CH2RX	atom-centred fragments	1
83	1	TIE	E-state topological parameter	topological descriptors	2
686	1	RDF050m	Radial Distribution Function - 5.0/weighted by atomic masses	RDF descriptors	3
1350	3	B09[C—O]	presence/absence of C—O at topological distance 09	2D binary fingerprints	2
1343	5	B07[C—C]	presence/absence of C—C at topological distance 07	2D binary fingerprints	2
1300	4	H-051	H attached to alpha-C	atom-centred fragments	1
1465	3	S_sssCH	S_sssCH	atomtypes (cerius2)	1
274	1	MATS6m	Moran autocorrelation - lag 6/weighted by atomic masses	2D autocorrelations	2
1006	2	G2p	2st component symmetry directional WHIM index/weighted by	WHIM descriptors	3
			atomic polarizabilities
757	1	RDF115e	Radial Distribution Function - 11.5/weighted by atomic Sanderson	RDF descriptors	3
			electronegativities
1298	1	H-049	H attached to C3(sp3)/C2(sp2)/C3(sp2)/C3(sp)	atom-centred fragments	1
1303	1	O-057	phenol/enol/carboxyl OH	atom-centred fragments	1
672	1	RDF130u	Radial Distribution Function - 13.0/unweighted	RDF descriptors	3
963	1	G3u	3st component symmetry directional WHIM index/unweighted	WHIM descriptors	3
1268	1	nRCHO	number of aldehydes (aliphatic)	functional group counts	1
1270	1	nArCO	number of ketones (aromatic)	functional group counts	1
1266	1	nRCOOH	number of carboxylic acids (aliphatic)	functional group counts	1
301	1	GATS1m	Geary autocorrelation - lag 1/weighted by atomic masses	2D autocorrelations	2
1262	1	nCconj	number of non-aromatic conjugated C(sp2)	functional group counts	1
297	1	MATS5p	Moran autocorrelation - lag 5/weighted by atomic polarizabilities	2D autocorrelations	2
MOR271.1
1299	1	H-050	H attached to heteroatom	atom-centred fragments	1
88	1	PHI	Kier flexibility index	topological descriptors	2
518	3	JGI6	mean topological charge index of order6	topological charge indices	2
691	2	RDF075m	Radial Distribution Function - 7.5/weighted by atomic masses	RDF descriptors	3
1298	2	H-049	H attached to C3(sp3)/C2(sp2)/C3(sp2)/C3(sp)	atom-centred fragments	1
621	1	SPH	spherosity	geometrical descriptors	3
1308	2	Hy	hydrophilic factor	molecular properties	1
343	1	EEig09x	Eigenvalue 09 from edge adj. matrix weighted by edge degrees	edge adjacency indices	2
1179	1	RTu+	R maximal index/unweighted	GETAWAY descriptors	3
308	1	GATS8m	Geary autocorrelation - lag 8/weighted by atomic masses	2D autocorrelations	2
1266	1	nRCOOH	number of carboxylic acids (aliphatic)	functional group counts	1
786	1	RDF110p	Radial Distribution Function - 11.0/weighted by atomic	RDF descriptors	3
			polarizabilities
304	1	GATS4m	Geary autocorrelation - lag 4/weighted by atomic masses	2D autocorrelations	2
297	1	MATS5p	Moran autocorrelation - lag 5/weighted by atomic polarizabilities	2D autocorrelations	2
1196	1	R8m+	R maximal autocorrelation of lag 8/weighted by atomic masses	GETAWAY descriptors	3
MOR272.1
1322	1	BLTF96	Verhaar model of Fish base-line toxicity from MLOGP (mmol/l)	molecular properties	1
639	1	DISPe	d COMMA2 value/weighted by atomic Sanderson	geometrical descriptors	3
			electronegativities
1347	1	B08[C—O]	presence/absence of C—O at topological distance 08	2D binary fingerprints	2
1155	1	HATS6p	leverage-weighted autocorrelation of lag 6/weighted by atomic	GETAWAY descriptors	3
			polarizabilities
274	1	MATS6m	Moran autocorrelation - lag 6/weighted by atomic masses	2D autocorrelations	2
727	1	RDF110v	Radial Distribution Function - 11.0/weighted by atomic van der	RDF descriptors	3
			Waals volumes
1018	1	G3s	3st component symmetry directional WHIM index/weighted by	WHIM descriptors	3
			atomic electrotopological states
1298	1	H-049	H attached to C3(sp3)/C2(sp2)/C3(sp2)/C3(sp)	atom-centred fragments	1
1299	1	H-050	H attached to heteroatom	atom-centred fragments	1
1190	1	R2m+	R maximal autocorrelation of lag 2/weighted by atomic masses	GETAWAY descriptors	3
308	1	GATS8m	Geary autocorrelation - lag 8/weighted by atomic masses	2D autocorrelations	2
1134	1	HATS5e	leverage-weighted autocorrelation of lag 5/weighted by atomic	GETAWAY descriptors	3
			Sanderson electronegativities
1082	1	H3m	H autocorrelation of lag 3/weighted by atomic masses	GETAWAY descriptors	3
441	1	BEHm3	highest eigenvalue n. 3 of Burden matrix/weighted by atomic	Burden eigenvalues	2
			masses
MOR273.1
1015	1	P2s	2nd component shape directional WHIM index/weighted by	WHIM descriptors	3
			atomic electrotopological states
77	2	Jhetv	Balaban-type index from van der Waals weighted distance matrix	topological descriptors	2
305	1	GATS5m	Geary autocorrelation - lag 5/weighted by atomic masses	2D autocorrelations	2
1070	1	HATS1u	leverage-weighted autocorrelation of lag 1/unweighted	GETAWAY descriptors	3
815	1	Mor20u	3D-MoRSE - signal 20/unweighted	3D-MoRSE descriptors	3
518	1	JGI6	mean topological charge index of order6	topological charge indices	2
1216	1	R1e	R autocorrelation of lag 1/weighted by atomic Sanderson	GETAWAY descriptors	3
			electronegativities
827	1	Mor32u	3D-MoRSE - signal 32/unweighted	3D-MoRSE descriptors	3
372	1	EEig08r	Eigenvalue 08 from edge adj. matrix weighted by resonance	edge adjacency indices	2
			integrals
441	1	BEHm3	highest eigenvalue n. 3 of Burden matrix/weighted by atomic	Burden eigenvalues	2
			masses
MOR277.1
1112	1	HATS3v	leverage-weighted autocorrelation of lag 3/weighted by atomic	GETAWAY descriptors	3
			van der Waals volumes
997	4	E1e	1st component accessibility directional WHIM index/weighted by	WHIM descriptors	3
			atomic Sanderson electronegativities
273	1	MATS5m	Moran autocorrelation - lag 5/weighted by atomic masses	2D autocorrelations	2
1009	1	E2p	2nd component accessibility directional WHIM index/weighted by	WHIM descriptors	3
			atomic polarizabilities
683	2	RDF035m	Radial Distribution Function - 3.5/weighted by atomic masses	RDF descriptors	3
1190	2	R2m+	R maximal autocorrelation of lag 2/weighted by atomic masses	GETAWAY descriptors	3
1232	1	R8e+	R maximal autocorrelation of lag 8/weighted by atomic	GETAWAY descriptors	3
			Sanderson electronegativities
608	3	SHP2	average shape profile index of order 2	Randic molecular profiles	3
306	1	GATS6m	Geary autocorrelation - lag 6/weighted by atomic masses	2D autocorrelations	2
497	1	BELp3	lowest eigenvalue n. 3 of Burden matrix/weighted by atomic	Burden eigenvalues	2
			polarizabilities
79	1	Jhetp	Balaban-type index from polarizability weighted distance matrix	topological descriptors	2
1300	1	H-051	H attached to alpha-C	atom-centred fragments	1
373	1	EEig09r	Eigenvalue 09 from edge adj. matrix weighted by resonance	edge adjacency indices	2
			integrals
481	1	BELe3	lowest eigenvalue n. 3 of Burden matrix/weighted by atomic	Burden eigenvalues	2
			Sanderson electronegativities
1267	1	nRCOOR	number of esters (aliphatic)	functional group counts	1
965	1	E2u	2nd component accessibility directional WHIM index/unweighted	WHIM descriptors	3
517	1	JGI5	mean topological charge index of order5	topological charge indices	2
303	1	GATS3m	Geary autocorrelation - lag 3/weighted by atomic masses	2D autocorrelations	2
957	1	L2u	2nd component size directional WHIM index/unweighted	WHIM descriptors	3
1466	1	S_dssC	S_dssC	atomtypes (cerius2)	1
996	1	G3e	3st component symmetry directional WHIM index/weighted by	WHIM descriptors	3
			atomic Sanderson electronegativities
1340	1	B06[C—C]	presence/absence of C—C at topological distance 06	2D binary fingerprints	2
1001	1	L2p	2nd component size directional WHIM index/weighted by atomic	WHIM descriptors	3
			polarizabilities
MOR30.1
1350	1	B09[C—O]	presence/absence of C—O at topological distance 09	2D binary fingerprints	2
1302	1	O-056	alcohol	atom-centred fragments	1
1344	2	B07[C—O]	presence/absence of C—O at topological distance 07	2D binary fingerprints	2
722	1	RDF085v	Radial Distribution Function - 8.5/weighted by atomic van der	RDF descriptors	3
			Waals volumes
1300	5	H-051	H attached to alpha-C	atom-centred fragments	1
691	1	RDF075m	Radial Distribution Function - 7.5/weighted by atomic masses	RDF descriptors	3
1282	3	C-006	CH2RX	atom-centred fragments	1
625	1	L/Bw	length-to-breadth ratio by WHIM	geometrical descriptors	3
356	1	EEig07d	Eigenvalue 07 from edge adj. matrix weighted by dipole moments	edge adjacency indices	2
724	1	RDF095v	Radial Distribution Function - 9.5/weighted by atomic van der	RDF descriptors	3
			Waals volumes
1009	1	E2p	2nd component accessibility directional WHIM index/weighted by	WHIM descriptors	3
			atomic polarizabilities
307	2	GATS7m	Geary autocorrelation - lag 7/weighted by atomic masses	2D autocorrelations	2
857	1	Mor30m	3D-MoRSE - signal 30/weighted by atomic masses	3D-MoRSE descriptors	3
804	1	Mor09u	3D-MoRSE - signal 09/unweighted	3D-MoRSE descriptors	3
355	1	EEig06d	Eigenvalue 06 from edge adj. matrix weighted by dipole moments	edge adjacency indices	2
308	1	GATS8m	Geary autocorrelation - lag 8/weighted by atomic masses	2D autocorrelations	2
1321	1	Infective-80	Ghose-Viswanadhan-Wendoloski antiinfective-like index at 80%	molecular properties	1
1230	1	R6e+	R maximal autocorrelation of lag 6/weighted by atomic	GETAWAY descriptors	3
			Sanderson electronegativities
302	1	GATS2m	Geary autocorrelation - lag 2/weighted by atomic masses	2D autocorrelations	2
743	1	RDF045e	Radial Distribution Function - 4.5/weighted by atomic Sanderson	RDF descriptors	3
			electronegativities
MOR33.1
1377	1	F07[C—O]	frequency of C—O at topological distance 07	2D frequency fingerprints	2
1266	1	nRCOOH	number of carboxylic acids (aliphatic)	functional group counts	1
635	1	DISPv	d COMMA2 value/weighted by atomic van der Waals volumes	geometrical descriptors	3
1367	1	F04[C—O]	frequency of C—O at topological distance 04	2D frequency fingerprints	2
908	2	Mor17e	3D-MoRSE - signal 17/weighted by atomic Sanderson	3D-MoRSE descriptors	3
			electronegativities
1300	1	H-051	H attached to alpha-C	atom-centred fragments	1
1282	1	C-006	CH2RX	atom-centred fragments	1
307	1	GATS7m	Geary autocorrelation - lag 7/weighted by atomic masses	2D autocorrelations	2
1299	1	H-050	H attached to heteroatom	atom-centred fragments	1
MOR37.1
1350	1	B09[C—O]	presence/absence of C—O at topological distance 09	2D binary fingerprints	2
1302	1	O-056	alcohol	atom-centred fragments	1
1347	1	B08[C—O]	presence/absence of C—O at topological distance 08	2D binary fingerprints	2
MOR40.1
727	2	RDF110v	Radial Distribution Function - 11.0/weighted by atomic van der	RDF descriptors	3
			Waals volumes
1300	1	H-051	H attached to alpha-C	atom-centred fragments	1
908	1	Mor17e	3D-MoRSE - signal 17/weighted by atomic Sanderson	3D-MoRSE descriptors	3
			electronegativities
1282	1	C-006	CH2RX	atom-centred fragments	1
307	1	GATS7m	Geary autocorrelation - lag 7/weighted by atomic masses	2D autocorrelations	2
MOR41.1
201	1	HVcpx	graph vertex complexity index	information indices	2
1443	1	Kappa-3	Kappa-3	topological (cerius2)	2
303	4	GATS3m	Geary autocorrelation - lag 3/weighted by atomic masses	2D autocorrelations	2
1266	8	nRCOOH	number of carboxylic acids (aliphatic)	functional group counts	1
1298	3	H-049	H attached to C3(sp3)/C2(sp2)/C3(sp2)/C3(sp)	atom-centred fragments	1
869	2	Mor10v	3D-MoRSE - signal 10/weighted by atomic van der Waals	3D-MoRSE descriptors	3
			volumes
781	2	RDF085p	Radial Distribution Function - 8.5/weighted by atomic	RDF descriptors	3
			polarizabilities
372	3	EEig08r	Eigenvalue 08 from edge adj. matrix weighted by resonance	edge adjacency indices	2
			integrals
1452	4	BIC	BIC	topological (cerius2)	2
308	5	GATS8m	Geary autocorrelation - lag 8/weighted by atomic masses	2D autocorrelations	2
1085	3	H6m	H autocorrelation of lag 6/weighted by atomic masses	GETAWAY descriptors	3
489	1	BEHp3	highest eigenvalue n. 3 of Burden matrix/weighted by atomic	Burden eigenvalues	2
			polarizabilities
515	1	JGI3	mean topological charge index of order3	topological charge indices	2
663	4	RDF085u	Radial Distribution Function - 8.5/unweighted	RDF descriptors	3
302	1	GATS2m	Geary autocorrelation - lag 2/weighted by atomic masses	2D autocorrelations	2
913	4	Mor22e	3D-MoRSE - signal 22/weighted by atomic Sanderson	3D-MoRSE descriptors	3
			electronegativities
1255	3	nCq	number of total quaternary C(sp3)	functional group counts	1
1008	1	E1p	1st component accessibility directional WHIM index/weighted by	WHIM descriptors	3
			atomic polarizabilities
715	1	RDF050v	Radial Distribution Function - 5.0/weighted by atomic van der	RDF descriptors	3
			Waals volumes
91	2	PW3	path/walk 3-Randic shape index	topological descriptors	2
1316	1	GVWAI-80	Ghose-Viswanadhan-Wendoloski drug-like index at 80%	molecular properties	1
1283	1	C-008	CHR2X	atom-centred fragments	1
1105	1	H6v	H autocorrelation of lag 6/weighted by atomic van der Waals	GETAWAY descriptors	3
			volumes
271	1	MATS3m	Moran autocorrelation - lag 3/weighted by atomic masses	2D autocorrelations	2
1405	1	Atype_C_18	Number of Carbon Type 18	atomtypes (Cerius2)	1
457	1	BEHv3	highest eigenvalue n. 3 of Burden matrix/weighted by atomic van	Burden eigenvalues	2
			der Waals volumes
672	1	RDF130u	Radial Distribution Function - 13.0/unweighted	RDF descriptors	3
1268	1	nRCHO	number of aldehydes (aliphatic)	functional group counts	1
1338	1	B05[C—O]	presence/absence of C—O at topological distance 05	2D binary fingerprints	2
620	1	MEcc	molecular eccentricity	geometrical descriptors	3
165	1	X2A	average connectivity index chi-2	connectivity indices	2
MOR5.1
1266	1	nRCOOH	number of carboxylic acids (aliphatic)	functional group counts	1
1377	1	F07[C—O]	frequency of C—O at topological distance 07	2D frequency fingerprints	2
1367	1	F04[C—O]	frequency of C—O at topological distance 04	2D frequency fingerprints	2
1303	1	O-057	phenol/enol/carboxyl OH	atom-centred fragments	1
908	1	Mor17e	3D-MoRSE - signal 17/weighted by atomic Sanderson	3D-MoRSE descriptors	3
			electronegativities
OR1A1
1077	2	HATS8u	leverage-weighted autocorrelation of lag 8/unweighted	GETAWAY descriptors	3
1019	1	E1s	1st component accessibility directional WHIM index/weighted by	WHIM descriptors	3
			atomic electrotopological states
1211	2	R5v+	R maximal autocorrelation of lag 5/weighted by atomic van der	GETAWAY descriptors	3
			Waals volumes
925	1	Mor02p	3D-MoRSE - signal 02/weighted by atomic polarizabilities	3D-MoRSE descriptors	3
639	1	DISPe	d COMMA2 value/weighted by atomic Sanderson	geometrical descriptors	3
			electronegativities
1340	3	B06[C—C]	presence/absence of C—C at topological distance 06	2D binary fingerprints	2
1268	2	nRCHO	number of aldehydes (aliphatic)	functional group counts	1
944	1	Mor21p	3D-MoRSE - signal 21/weighted by atomic. polarizabilities	3D-MoRSE descriptors	3
515	1	JGI3	mean topological charge index of order3	topological charge indices	2
1303	1	O-057	phenol/enol/carboxyl OH	atom-centred fragments	1
696	1	RDF100m	Radial Distribution Function - 10.0/weighted by atomic masses	RDF descriptors	3
273	1	MATS5m	Moran autocorrelation - lag 5/weighted by atomic masses	2D autocorrelations	2
1194	1	R6m+	R maximal autocorrelation of lag 6/weighted by atomic masses	GETAWAY descriptors	3
665	1	RDF095u	Radial Distribution Function - 9.5/unweighted	RDF descriptors	3
1266	1	nRCOOH	number of carboxylic acids (aliphatic)	functional group counts	1
414	1	ESpm06d	Spectral moment 06 from edge adj. matrix weighted by dipole	edge adjacency indices	2
			moments
451	1	BELm5	lowest eigenvalue n. 5 of Burden matrix/weighted by atomic	Burden eigenvalues	2
			masses
OR2J2
1019	3	E1s	1st component accessibility directional WHIM index/weighted by	WHIM descriptors	3
			atomic electrotopological states
1374	1	F06[C—O]	frequency of C—O at topological distance 06	2D frequency fingerprints	2
635	1	DISPv	d COMMA2 value/weighted by atomic van der Weals volumes	geometrical descriptors	3
517	1	JGI5	mean topological charge index of order5	topological charge indices	2
1300	3	H-051	H attached to alpha-C	atom-centred fragments	1
1060	1	H1u	H autocorrelation of lag 1/unweighted	GETAWAY descriptors	3
631	4	DISPm	d COMMA2 value/weighted by atomic masses	geometrical descriptors	3
462	1	BEHv8	highest eigenvalue n. 8 of Burden matrix/weighted by atomic van	Burden eigenvalues	2
			der Weals volumes
1298	2	H-049	H attached to C3(sp3)/C2(sp2)/C3(sp2)/C3(sp)	atom-centred fragments	1
1341	1	B06[C—O]	presence/absence of C—O at topological distance 06	2D binary fingerprints	2
1303	1	O-057	phenol/enol/carboxyl OH	atom-centred fragments	1
805	1	Mor10u	3D-MoRSE - signal 10/unweighted	3D-MoRSE descriptors	3
1087	1	H8m	H autocorrelation of lag 8/weighted by atomic masses	GETAWAY descriptors	3
1355	2	F01[C—C]	frequency of C—C at topological distance 01	2D frequency fingerprints	2
1154	2	HATS5p	leverage-weighted autocorrelation of lag 5/weighted by atomic	GETAWAY descriptors	3
			polarizabilities
297	1	MATS5p	Moran autocorrelation - lag 5/weighted by atomic polarizabilities	2D autocorrelations	2
1085	1	H6m	H autocorrelation of lag 6/weighted by atomic masses	GETAWAY descriptors	3
1466	1	S_dssC	S_dssC	atomtypes (cerius2)	1
1129	1	HATS0e	leverage-weighted autocorrelation of lag 0/weighted by atomic	GETAWAY descriptors	3
			Sanderson electronegativities
1249	1	R7p+	R maximal autocorrelation of lag 7/weighted by atomic	GETAWAY descriptors	3
			polarizabilities
541	1	VEA2	average eigenvector coefficient sum from adjacency matrix	eigenvalue-based indices	2
OR2W1
1337	2	B05[C—C]	presence/absence of C—C at topological distance 05	2D binary fingerprints	2
1155	2	HATS6p	leverage-weighted autocorrelation of lag 6/weighted by atomic	GETAWAY descriptors	3
			polarizabilities
698	1	RDF110m	Radial Distribution Function - 11.0/weighted by atomic masses	RDF descriptors	3
1190	1	R2m+	R maximal autocorrelation of lag 2/weighted by atomic masses	GETAWAY descriptors	3
297	1	MATS5p	Moran autocorrelation - lag 5/weighted by atomic polarizabilities	2D autocorrelations	2
OR5P3
1262	2	nCconj	number of non-aromatic conjugated C(sp2)	functional group counts	1
1092	1	HATS3m	leverage-weighted autocorrelation of lag 3/weighted by atomic	GETAWAY descriptors	3
			masses
1222	3	R7e	R autocorrelation of lag 7/weighted by atomic Sanderson	GETAWAY descriptors	3
			electronegativities
206	1	Yindex	Balaban Y index	information indices	2
1231	1	R7e+	R maximal autocorrelation of lag 7/weighted by atomic	GETAWAY descriptors	3
			Sanderson electronegativities
1323	2	BLTD48	Verhaar model of Daphnia base-line toxicity from MLOGP (mmol/l)	molecular properties	1
1185	1	R6m	R autocorrelation of lag 6/weighted by atomic masses	GETAWAY descriptors	3
1297	1	H-047	H attached to C1(sp3)/C0(sp2)	atom-centred fragments	1
1183	1	R4m	R autocorrelation of lag 4/weighted by atomic masses	GETAWAY descriptors	3
302	3	GATS2m	Geary autocorrelation - lag 2/weighted by atomic masses	2D autocorrelations	2
631	1	DISPm	d COMMA2 value/weighted by atomic masses	geometrical descriptors	3
805	2	Mor10u	3D-MoRSE - signal 10/unweighted	3D-MoRSE descriptors	3
774	1	RDF050p	Radial Distribution Function - 5.0/weighted by atomic	RDF descriptors	3
			polarizabilities
1336	1	B04[O—O]	presence/absence of O—O at topological distance 04	2D binary fingerprints	2
447	1	BELm1	lowest eigenvalue n. 1 of Burden matrix/weighted by atomic	Burden eigenvalues	2
			masses
870	1	Mor11v	3D-MoRSE - signal 11/weighted by atomic van der Waals	3D-MoRSE descriptors	3
			volumes
1136	2	HATS7e	leverage-weighted autocorrelation of lag 7/weighted by atomic	GETAWAY descriptors	3
			Sanderson electronegativities
1337	1	B05[C—C]	presence/absence of C—C at topological distance 05	2D binary fingerprints	2
1298	1	H-049	H attached to C3(sp3)/C2(sp2)/C3(sp2)/C3(sp)	atom-centred fragments	1
1343	2	B07[C—C]	presence/absence of C—C at topological distance 07	2D binary fingerprints	2
1266	1	nRCOOH	number of carboxylic acids (aliphatic)	functional group counts	1
941	1	Mor18p	3D-MoRSE - signal 18/weighted by atomic polarizabilities	3D-MoRSE descriptors	3
1111	1	HATS2v	leverage-weighted autocorrelation of lag 2/weighted by atomic	GETAWAY descriptors	3
			van der Waals volumes

TABLE 6
Top ~25 predicted compounds for each Mammalian OR. Tables contain
SMILES strings, and distances, of the top ~25 predicted compounds for
each Or. All distances represent the minimum distance based on optimized
descriptors to an active compound listed in gray cells for that particular Or.
SMILES                           Distance
Mor1-1
CC1═CC2═C(C═C1)OC(═O)C2          0.04917087
CC1═CC2═C(CC(═O)O2)C═C1          0.06445035
CC1═CC═CC2═C1OC(═O)C2            0.06478577
CC(CCCC(═O)O)N                   0.0766186
CC1═CC(═C2C(═C1)CC(═O)O2)C       0.09134395
CC1═C(C2═C(CC(═O)O2)C═C1)C       0.09749545
CC1═CC(═C2CC(═O)OC2═C1)C         0.1021952
CC(C)(C)C(CCC(═O)O)O             0.1026351
C1C2═C(C(═CC═C2)N)OC1═O          0.1122016
C1C2═CC═CC═C2NC1═O               0.1200522
C1C2═C(C═CC═C2OC1═O)N            0.1221153
C═CCCCC(═O)O                     0.1245319
CC(C)(C)CCCC(═O)N                0.1311838
C(CC(N)N)CC(═O)O                 0.1339592
CC(C1═C2CC(═O)NC2═CC═C1)N        0.1356993
CC1═C2CC(═O)OC2═CC═C1            0.136907
CC1═CC═CC2═C1NC(═O)C2            0.1403033
C(═O)CCCC                        0.1404621
C1C2═C(C═C(C═C2)O)OC1═O          0.1431714
CCC(C)(C)CC(═O)N                 0.1436566
CC(CCCC(═O)O)O                   0.1452177
C1C2═C(C═C(C═C2)N)NC1═O          0.146264
CC(C)(C)C1═CCC(═O)O1             0.1538394
CCCCCC(═O)N                      0.1546248
CC(C)(C)CC(CC(═O)O)O             0.156154
C1═CNC═C1CCC(═O)N                0.1571392
Mor106-1
CC═COC1═CC═CC═C1                 0.08891955
CC(S)SC1═CC═CC═C1                0.1534956
C═C(C1═CC═CC═C1)C                0.1583203
CC(═C)C1═CC═CC═C1                0.1583203
CC1═C2C(═CC═C1)N2C               0.1611889
CC1═CC═CC═C1C═C                  0.1622158
CC1C(N1)C2═CC═CC═C2              0.1766318
CSC(═O)C1═CC═CC═C1               0.1913783
CSC(C1═CC═CC═C1)S                0.1981317
C═COC1═CC═CC(═C1)O               0.1996824
C═CC1═CC(═C(C═C1)S)S             0.2060344
C═CC1═CC(═CC═C1)N═C═S            0.2098719
C#COC1═CC═CC═C1                  0.2130395
C1CC1CC2═CC═CC═C2                0.2152684
COC═CC1═CC═CC═C1                 0.2170264
CC1═C(C2═C(O2)C═C1)C             0.2178679
c1(ccccc1)CC#N                   0.2180725
CC1═CC═CC═C1OC#C                 0.2181859
C═COC1═CC═CC═C1                  0.2188387
CSC1═CC═CC2═C1C═C2               0.219095
C1═CC═C(C═C1)C(═O)NO             0.2207775
C1═CC═C(C═C1)N(C(═S)S)O          0.2218528
C1═CC═C2C(═C1)C(═O)SC2═O         0.2247712
C1═CC═C(C═C1)S(═O)(═O)N═C═S      0.226246
CCC1═C2C(═CC═C1)N2               0.2282767
C1═CC═C(C═C1)C2OS2(═O)═O         0.2323088
Mor107-1
CC1(C2(CCC1(CC2═O)N)C)C          0.2473219
CC1C2(CCC1(C(═O)C2)C)C           0.2764237
CC1(C2(CCC1(CC2═O)O)C)C          0.3186515
CC1(C2CC(C1(C(═O)C2)C)O)C        0.3204451
CC1(C2CC(═O)C1(CC2O)C)C          0.3482935
CC1(C2CCC1(C(═O)C2)CO)C          0.366073
CC1(C2CCC1(C(═O)C2)CS)C          0.4426886
CCC12CCC(C1(C)C)CC2═O            0.4550365
CC12CCC(C1(C)CO)CC2═O            0.4566952
CC1(C2CCC1(CC2═O)C)C             0.4653999
CC1(C2CCC1(C(C2═O)O)C)C          0.4703974
CC1(C2CC(C1(CC2═O)C)O)C          0.5505192
CC(C)C12CCC(C1)(CC2═O)C          0.5732307
CC1(C2CCC1(C(═NC)C2)C)C          0.5802225
CC1(C(CC2C1(C2)C)CC═O)C          0.6171489
CC1C2CCC(C1═O)(C2(C)C)C          0.629438
CC1(C2CCC1(C(═O)C2)C═C)C         0.6401904
CC1(C2(CCC1(CC2═O)OC)C)C         0.6463336
CCCC1(CCC(C(═O)C1)(C)C)C         0.6494132
CC1(CCCC12CC═NC2)C               0.6903515
CC1(C2CCC1(C(═O)C2)C═O)C         0.7002376
CC1(C(C1(C)C)C(═O)NC2CC2)C       0.7104875
CC12CC3C1(C(═O)CC2C3)C           0.7142688
CCOC(═O)C1C2(C13CC3)CC2          0.7263507
CC12CCCC(═O)C1(COC2)C            0.7281732
CCC12CCC1C(CC2═O)(C)C            0.7489101
Mor129-1
C1C(═O)CNC2═CC═CC═C21            0.1220788
CC1═CCCC(C1)(C)C(═O)C            0.131721
C1C2C(═CC═CO2)C═CC1═O            0.1324226
CC1CC(CC═C1C#N)(C)C              0.1401867
CC1═CCC(CC1)C(C)C═O              0.1440427
C1CCC2C(C1)CCC(═O)O2             0.1447183
CC(C)(CO)C1═CC═CC═C1             0.1488985
C1═CC═C2C(═C1)C═COC2═O           0.1541709
COC12CCC(CC1)NC(═O)C2            0.1556671
C1═CC═C2C(═C1)C═CC(═O)N2         0.1562217
CC(═O)C1═CC═CC═C1N               0.1588058
CC1CC(═CC(C1CO)C)C               0.1637881
CC1CCC(═CC1═O)C(C)C              0.1638838
C1═CC═C2C(═C1)C(═O)C═CN2         0.1644921
CC1CC(CC═C1C═O)(C)C              0.1653725
C1═CNC2═CC(═O)C═CC2═C1           0.168101
C1═CC═C2C(═C1)C(═O)C═CO2         0.1688119
C1C═CC2═C(C1O)N═CC═C2            0.1727835
CC1═CCC(CC1O)C(═C)C              0.1737252
C1C2═CC═CC═C2C(═O)CN1            0.1787667
CC1CCC(CC1C)(C)C═O               0.1795582
C1CC(CC═C1)CC#N                  0.1819388
CC(C)(C)C1═CC═C(CC1)O            0.1839154
c1(ccccc1)C(C)O                  0.1866107
C1C(═O)C═C2C═CC═CC2═N1           0.1886765
C1CC═CC(C1)CC═O                  0.1891273
Mor136-1
CCC1(CC(OC1═N)(C)C)CC            0.05816986
C1═CC(═C2C(═C1)S2)C(═S)N         0.06587855
CC(C)(C)C1CC(═O)C2C1C2           0.06816311
CCC1(CCCC1═O)CC                  0.0729801
CCCCCC(CC)C(═O)C1CC1             0.07530886
CC(C)C12CCC(C1)(CC2═O)C          0.07590504
CC(═O)C1CCC2C1CCCC2              0.08637492
CC1CCC2(CC1)C═CC(C2═O)C          0.08638542
CC1(C2CC(═O)C1(CC2═O)C)C         0.08683412
CCN1C2CCC1CC(═O)C2               0.0869081
CC1C═CC2(C1═O)CCCCC2             0.08745782
C1CCC(═O)NCCCC(═O)C1             0.0884375
CCC1CCCC(═O)CCC1CC               0.09197357
CC(C)OC1═NC═CC═CN1               0.09294477
C1CC2CC(C1)CC(═O)C2              0.09388228
C1CC2COCC(C1)C2═O                0.1089249
CC1(C(═O)CC23C1(CCC2)CCC3)C      0.1093057
C1CCC(═O)C2CCCC(C1)C2            0.1097336
CCCCCC1(CCCC1═O)CC═C             0.1102119
CN(C)C(═NS(═O)O)N(C)C            0.1104801
CC(C)(C═C)C1CCCC1═O              0.1145671
C1═CC═C(OC═C1)NCO                0.116571
C1C2CC3CC1C(C3═O)C═C2            0.1167317
COC1CCC(═O)C12CCCC2              0.1170204
CC1C2CCCN1CC2═O                  0.1171827
Mor139-1
C1CCC2═C(C1)CCCC2═O              0.04565114
CCC(C)C1CCC(═O)CC1               0.04807124
C1CC2CC═CCC2C(═O)C1              0.04894259
CC(C)C1═CC(═O)CCC1               0.04953565
CCCC(C)C1═CCCC1═O                0.05030901
CC(C)CCC1═CCCC1═O                0.05046165
CC1CCCC2═C1CCC2═O                0.05373275
CC(C)CC1═CC(═O)CC1               0.05429959
CN(C)CCC1═CC═NC═C1               0.0645497
CC1═C2CSCC2CC1═O                 0.0662884
C1CCC2(CCC2)C(═O)C1              0.06641502
C1C(═O)COC2═CC═CC═C21            0.06771594
CC1CCCC2═C1C(═O)CC2              0.06775099
CCCCC1CCC(═O)C1═C                0.07151075
CC1(C2C1C(═O)C(═C)CC2)C          0.07206624
CC1═CCCC(═C(C)C)C1═O             0.07226345
CC1CC(═O)C2═C1CCCC2              0.07233297
C1CCC(═C2CCCC2═O)C1              0.0727393
C1CC2CCC(═O)C═C2C1               0.07644502
CC(═CC1═CC(═O)CCC1)C             0.07737008
CC1CC2C(C2(C)C)CC1═O             0.07785543
CCCCCC1═CCCC1═O                  0.07816834
CC1═CC(═O)C(CC1)C(═C)C           0.07862932
CCC(C)CC1═CCCC1═O                0.0793086
C1CCC2(CC1)CCC2═O                0.08011041
Mor162-1
CC1NC(═O)C2═CC═CC═C2O1           0.03923089
C1═CC═C2C(═C1)C═C(NO2)C═O        0.05289857
CC1═CN═C(C(═N1)C)C═O             0.06707111
C1C═C(C2═C(O1)N═CC═C2)O          0.06713544
C1═CC═C2C(═C1)C═C(C(═O)O2)N      0.06748666
C1CC2═CC3═C(C(═O)N2C1)NC═C3      0.06865916
CC1═CC(═C(C═C1)C═O)C             0.07170324
C1═CC(═CC(═C1)NC(═O)O)C═O        0.07340692
C1═CC(═C(C═C1C(═O)CO)O)O         0.07503177
CC1═CC(═C(C(═C1)O)C═O)O          0.07530568
C1COC(═N1)C2═CN═CC═C2            0.07562215
CC(═O)C1═CC2═CC═CC═C2C1          0.07665453
CC1═C(NC═C1C(═O)C)C              0.07707266
C1═C(NC(═C1)C2═CN═CO2)C═O        0.0776176
C1═CC═C2C(═C1)C(═O)C(═CO2)N      0.07804614
C1═CC2═C(NC(═O)C═C2)N═C1         0.07999517
C1═CC2═C(C═CC(═C2O)C═O)C═C1O     0.08200772
C1═CC═C2C(═C1)C═C(C═N2)O         0.08304375
CC1═C(NC2═C1C═C(C═C2)C═C)C       0.08360804
C1═CC(═CC═C1C═O)C2═NN═CO2        0.08368816
CC(═O)C1═CC═C(C═C1)N             0.08371712
CC(═O)C1═CN═CC═C1                0.08373728
C1═CC═C2C═C(C═CC2═C1)C═NO        0.0837472
COC(═O)C1═CC═C(C═C1)O            0.08383989
C1C═CC2═CC═CC═C2SS1              0.08392745
Mor170-1
C1═CC═C2C(═C1)C═C(NO2)C═O        0.06619393
CC1═NC(═O)C2═CC═CC═C2N1          0.0745505
CN1C═NC2═CC═CC═C2C1═O            0.08105557
CC1NC(═O)C2═CC═CC═C2O1           0.08323528
COC(═N)C1═CC═CC═C1               0.09673947
C1═CC═C2C(═C1)C(═O)N═CC═N2       0.1024509
C1C2═CC═CC═C2ON═C1C═O            0.1030488
C1═CC═C(C═C1)C(═O)CCN            0.1099358
CNCC(═O)C1═CC═CC═C1              0.1100806
CC1COC2═CC═CC═C2C1═O             0.1146739
C═CN1C═NC2═CC═CC═C2C1═O          0.1150157
C1═CNN(N═C1)C2═CC═C(C═C2)C═O     0.1202606
CC1═NC2═CC═CC═C2C(═O)N1C         0.1237537
C═NNC(═O)C1═CC═CC═C1             0.1254566
CN1NC(═O)C2═CC═CC═C2O1           0.1294035
CC1═NC2═CC═CC═C2C(═O)N1N         0.1318522
CNNC(═O)C1═CC═CC═C1              0.1335452
C1NC(═O)C2═CC═CC═C2S1            0.1342063
CC1═CC(═O)C2═CC═CC═C2O1          0.134347
CN1C(═O)C2═CC═CC═C2N═N1          0.1366252
C1═CC═C2C(═C1)C═C(C(═O)O2)C#N    0.1383896
C1═CC═C2C(═C1)C(═O)C═NS2         0.1389853
CC(═O)NN═CC1═CC═CC═C1            0.1393494
C1═CC═C2C(═C1)C═CC(═O)N2         0.1397765
CN1C(═O)C2═CC═CC═C2N═C1N         0.1422149
Mor184-1
CC1CCC(C(═C1)O)C(═C)C            0.2379638
CC(═C)C1CC═C(C(═O)C1)O           0.3728224
CC1═CCC(CC1═O)C(═C)C═O           0.3772072
CC1(CCCCC1═O)CC═C                0.3977689
CC1CC(═O)C(C(═O)C1)CC═C          0.4097808
CC1═CC(═O)C(CC1)C(═C)C           0.4227823
CC(═C)CC1(CCCCC1)O               0.4271719
C═CCC1CCC(═O)NC1═O               0.4313385
CC(═C)C1CCC(CC1)C═O              0.4430801
CC1CCC(C(═O)C1)C(═C)C            0.4515747
CC1C(CCC1C(═O)O)C═C              0.4557655
CC1═CCC(CC1O)C(═C)C              0.4625187
CC(═C)C1CCC(═CC1)C(O)O           0.4628184
CC(═C)C1CCC(═CC1)C(═O)N          0.472155
C═CCCC1C(═O)CCCC1═O              0.4769287
CC(CC═C)C1CCCCC1═O               0.4788082
CC1CCCC1(CC═C)CC═O               0.48704
C═CCCC(═O)C1(CCCCC1)O            0.4917173
C═CCCC1(CCCCC1)C═O               0.4941705
CC(═C)CC1═CCCCC1═O               0.4981998
C═CC1CCC(CC1)C(═O)OO             0.5022362
CC(═O)C(═C)C1CCCCC1              0.5079129
CC(═C)CC1(CCCCC1═O)C             0.5103795
CCCCC(═C)C1(CCCCC1)O             0.5108372
CC1CCC(C(C1)O)C(═C)C             0.5122089
Mor185-1
C1CC2═COC═C2CC1═O                0.02606647
C1═CC2═C(C═CC(═O)N2)N═C1         0.03118136
C1CC(═O)C2═C1C═CC═N2             0.0323404
C1CCC2═C(C1)CCC(═O)O2            0.03555811
CCC1CCCC(C1)N                    0.0418893
CC(═C)C1═COC═C1                  0.04258475
CCC1═CCC(CC1)O                   0.04279004
CCC1CCC(CC1)N                    0.04383469
C1CC(═O)C2(C1)CC2                0.04396265
C1CC(═O)C2═C1N═CC═C2             0.04402903
C1CN2CC(═O)OCC2CN1               0.04424295
C1C(CC2C1CNC2)N                  0.046368
C1CCC2C(C1)CC2═O                 0.04657459
C1CN2CCOC(═O)C2CN1               0.04905777
C1C2C═CC═C2C(CN1)O               0.04945669
C1CCC2(CC1)CC2O                  0.04991706
C1CCC2═C(C1)C(═O)CN2             0.05012532
CCC1═C(CCCC1)O                   0.0513685
CC(═O)C1═CCCC═C1                 0.05278158
C1CC(═O)C2═CN═CN═C21             0.05282474
C1═CC2═C(NC═CC2═O)N═C1           0.05299349
CONCCN1CCCC1                     0.05360856
CCC1CCCCC(C1)O                   0.0536915
CC1═CC(═CCC1)OC                  0.05373938
CCC1═CC═CC═C1N                   0.05481358
Mor189-1
CC1(C2CCC1(CC2═O)C)C             0.04668916
CC(═C)C1CC═C(C(═O)C1)O           0.06676451
CC1═CC(═O)C(CC1)C(═C)C           0.1183577
C═C(C)C(CCC1C)═CC1═O             0.1347063
CC1CCC(═CC1═O)C(C)C              0.1680434
CC1═CCCC(C1═O)(C)C#C             0.1763185
CC1═CC(═O)CC(C1)C(═C)C           0.1962587
CC1═CC(CCC1═O)C(C)C              0.1976955
CC1C2(CCC1(C(═O)C2)C)C           0.2061102
CC1CC═C(C(═O)C1)C(C)C            0.2072536
O═C1C2C(CCC(C2)C1C)═C            0.2078362
CC1(CC(C(═O)C1)CC═C)C            0.2092703
CC(C)(C12CCC(═O)C1C2)O           0.2097237
CCC1═C(C(═O)C(CC1)C)C            0.2140007
CC1═CCCC(C1═O)(C)C═C             0.2180427
CCCC1(CCC(═O)C═C1)C              0.2194855
CC(C1C═CCCN1)C(═O)C              0.2209319
CC1═CCC(CC1)C(═C)C═O             0.2250367
CC1(CCC(═O)C═C1C═C)C             0.2310372
C1CN(CCC1CCN)C═O                 0.2406093
O═C1C═C(C)CCC1C(C)C              0.241701
C═C1C═CCCC1CCC═O                 0.2420684
CCC1═CC(═O)CCC1(C)C              0.2445011
CC1C(═C)C2CCC(C2)C1═O            0.2468526
CCCC1CCC(═O)C(═C1)C              0.2519408
Mor2-1
CC1C(═O)N(C1═O)C2═CC═CC═C2       0.1582261
CCCOC(═O)CC1═CC═CC═C1            0.1927919
C#CCOC(═O)CC1═CC═CC═C1           0.2135794
CC1CCC2═C(C1)SC(═N2)CC#N         0.2261997
C1CC1(C2═CC═CC═C2)OCCS           0.2307089
C═CCOC(═O)CC1═CC═C(C═C1)O        0.2685227
COC(═O)Cc1ccc(cc1)OC             0.2720388
CCC(C1═C(C═NC═C1)CO)OC           0.2840643
COC(═O)CC1═CC═CC(═C1)C#N         0.2858069
C1CCS(═O)(═O)C2═CC═CC═C2C1       0.2878607
C1CC1(C2═CC═CC═C2CO)O            0.2910175
CC(C)OC(═O)CC1═CC═CC═C1          0.2962586
C1═C(C(═C(N1)CN)CC#N)CCC#N       0.2973288
CC(CC1═CC═CC═C1)N═C═S            0.2995831
CCC(COCC1═CC═CC═C1)S             0.2998371
C1CC2═CC═CC═C2NC(═O)C1           0.3024153
C1CC(C2═C(C1)C═CS2)NC(═O)N       0.3030766
CC1(CN(C2═C1SC═C2)C═O)C          0.3041082
CCC(C1═CC═CC═C1N)C(═O)O          0.3045982
C1CC(C2═CC═CC═C2SC1)O            0.3073687
C1CC(═O)CCC2═C(C1)NC═C2          0.3082988
C1CCC(C(═O)CC1)C2═NC═CN2         0.3085543
CC(C1═CC═CC═C1N)C(═O)OC          0.3109102
C1═CC(═CC(═C1)CN)CC(═O)O         0.3115366
CCOC1═CC═C(C═C1)CC(═O)O          0.3116775
Mor203-1
CCCCC(═O)CCC                     0.09870324
CCC(═O)CCCC(C)C                  0.1234426
CCC1CCC(CC1)C(═O)CC              0.1310437
CC(CCCCCO)C                      0.1429018
CCC1CC(C1)C(═O)C                 0.1522912
CCC(═O)C═CCCC                    0.1569658
CCCCCC(═O)CCC                    0.1636356
CC═CC═CCCO                       0.1710732
CC(═O)CC1CCCC═C1                 0.1711094
C═CC1═CC═C(C═C1)CCCO             0.1778983
CCCCCCC(CCC)O                    0.1785371
C#CC1═CCC(CC1)CCO                0.1786504
CC(CCCC═C)O                      0.1865765
CC(═C)CCCC(═O)C                  0.1901121
CCCCCC1CC(═O)C1                  0.1926162
CCC(CCC═C═C)O                    0.1948856
CC(C)CCCC(═O)C═C                 0.1949919
CCCC═C═CCO                       0.1959073
CCCCC(═O)NC                      0.196173
C1CC(═CC═C1)CCCO                 0.1965939
C(C═CC═CCC═CCC)O                 0.1966562
CC(C)CCNC(═O)C                   0.1977104
CC(C)C(CCCC═C)O                  0.1995084
C1CC1═CCCCCO                     0.2000587
CC(C)CC═CC(═O)C                  0.2005665
Mor204-6
CC1(CCCC═C1C(═O)O)C              0.1592494
CC1CCC(═CC1═O)C(C)C              0.1767846
CC1CCC(═CC1═O)C(C)C              0.1767846
C1═CC═C2C(═C1)C═COC2═O           0.1926777
C1═CC(═O)NC2═C1C═NC═C2           0.2034395
CC1═CC(═O)C(CC1)CCO              0.2121746
C═CC1═CC═C(C═C1)C(═O)O           0.2283565
COC(═O)C1═CCCC(C1)O              0.2532418
C1═CC2═C(C(═C1)O)C(═O)NC═C2      0.2600318
C1═CC2═C(NC(═O)C═C2)N═C1         0.2607831
C1═CC(═O)NC2═NC═NC═C21           0.2758744
CN1C2C1C(CC(═C2)C(═O)OC)O        0.2813197
C1═CC2═C3C(═C1)NC═C3C(═O)C═C2    0.2828428
C═CC1═CN═C(C═C1)C(═O)O           0.2854501
CC(═C)CC1═CCCCC1═O               0.31266
CC1═CCC(CC1═NO)C(═C)C            0.3135219
C1C(═CC1═O)C2═CC═CC═C2           0.3180221
COC(═O)C1═CCC(CC1)SC             0.3191871
C1CC(═O)C2CC1C(═O)C═C2           0.3195927
CCC(═O)C1═CCCC(S1)C              0.3235687
C1CCC2(C1)CCC═C(C2═O)O           0.3305075
C1C(═CC(═O)N1)C2═CC═CC═C2        0.3309083
C1═CC═C2C(═C1)C═CC(═O)N2         0.3342322
CC1CC═C(C(O1)C)C(═O)O            0.3351069
CC1CCCC═C1C(═O)O                 0.337837
Mor207-1
C1═CC═C2C(═C1)C═C(NO2)C═O        0.06933906
CC1═NC(═O)C2═CC═CC═C2N1          0.08035855
CC1NC(═O)C2═CC═CC═C2O1           0.08366296
CN1C═NC2═CC═CC═C2C1═O            0.09183173
COC(═N)C1═CC═CC═C1               0.1088847
CNCC(═O)C1═CC═CC═C1              0.1134785
CC1COC2═CC═CC═C2C1═O             0.1147292
C1═CC═C(C═C1)C(═O)CCN            0.1167073
C1C2═CC═CC═C2ON═C1C═O            0.1201505
C═CN1C═NC2═CC═CC═C2C1═O          0.124919
CC1═NC2═CC═CC═C2C(═O)N1C         0.1258814
C1═CC═C2C(═C1)C(═O)N═CC═N2       0.1260327
CN1NC(═O)C2═CC═CC═C2O1           0.1322961
CNNC(═O)C1═CC═CC═C1              0.1338931
CC1═CC(═O)C2═CC═CC═C2O1          0.1355292
CC1═NC2═CC═CC═C2C(═O)N1N         0.1374166
C1═CC═C2C(═C1)C═CC(═O)N2         0.1397979
CC(═O)NN═CC1═CC═CC═C1            0.1408284
C1═CC═C2C(═C1)C═C(C(═O)O2)C#N    0.1409447
CCOC(═N)C1═CC═CC═C1              0.1460253
CN1C(═O)C2═CC═CC═C2N═N1          0.1465203
C═NNC(═O)C1═CC═CC═C1             0.1469559
C1NC(═O)C2═CC═CC═C2S1            0.1485208
C1═CC═C2C(═C1)C(═O)SN2           0.1501398
C1C(C(═O)C2═CC═CC═C2O1)N         0.1512307
Mor223-1
C#CCOC(═O)CC1═CC═CC═C1           0.09948513
CCOC(═O)CC1═CC═CC═C1             0.1235873
CCCOC(═O)CC1═CC═CC═C1            0.1371426
C═CC(═O)OCC1═CC═CC═C1            0.160449
COC(═O)CC1═CC═CC═C1              0.2023979
CCC(═O)OCC1═CC═CC═C1             0.2040253
C1═CC═C(C═C1)CCCOC═O             0.2240541
C═C═CC(═O)OCC1═CC═CC═C1          0.228408
CCC(═O)OC1═CC═CC═C1              0.2532228
CC═C═CC(═O)OCC1═CC═CC═C1         0.2709531
C═CC(═O)OC1═CC═CC═C1             0.2744787
CC(═O)OCCCC1═CC═CC═C1            0.2756416
C═CCC(═O)OCC1═CC═CC═C1           0.2810861
C#CC(═O)OCC1═CC═CC═C1            0.2843047
C1═CC═C(C═C1)CCCCOC═O            0.2902787
CCCC(═O)OC1═CC═CC═C1             0.299426
CC#CCOC(═O)CC1═CC═CC═C1          0.3071297
CCCCC(═O)OC1═CC═CC═C1            0.3087009
COC(═O)CCC1═CC═CC═C1             0.3129654
COC(═O)CCCCC1═CC═CC═C1           0.3463383
CCCCOC(═O)CC1═CC═CC═C1           0.3527429
CC(═O)OC═CC1═CC═CC═C1            0.3589161
COC(═O)CCCC1═CC═CC═C1            0.3614484
C1═CC═C(C═C1)C═CCOC═O            0.3668489
C═COC(═O)CCC1═CC═CC═C1           0.3705043
Mor250-1
CCC(═O)OC1═CC═CC═C1              0.3793798
CCC(═O)OC1═CC═CC═C1              0.3793798
C1═CC═C2C(═C1)C═CC(═O)N2         0.384888
C1═CC═C(C═C1)OC(═O)N             0.4700359
CCC(═O)C1═CC═CC═C1               0.4719471
C1C(═O)C═CC2═CC═CC═C21           0.4828623
CC(═O)NC1═CC═CC═C1               0.4926594
C1C═CC2═CC═CC(═O)C2═C1           0.5147033
C1═CC═C(C═C1)NOC═O               0.534839
C1═CC2═CC═CC(═O)N2C═C1           0.5353036
C1C═C2C═CC═CC2OC1═O              0.5594665
C═C1C(═O)OC2═CC═CC═C2N1          0.572254
C1═CC═C(C═C1)ON═C═O              0.5755313
CC(═O)C═C1C═CCC═C1               0.5820786
C1═CC═C(C═C1)N(C═O)N             0.5914498
C1═CC═C(C═C1)NC(C═O)O            0.5925996
CC1═CC═C(C═C1)C(C)O              0.5991708
C1═CC═C(C═C1)ONC═O               0.6023895
C1═CC═C(C═C1)OC(═O)NS            0.6065903
c1(ccccc1)NC═O                   0.6157112
C1═CC═C(C═C1)NC(═O)N             0.620347
CCC(C1═CC═C(C═C1)N)O             0.6529591
CC1═CC═CC═C1C(C)O                0.659726
C1═CC═C(C═C1)OC#N                0.668126
CC(C1═CC═C(C═C1)N)O              0.670643
Mor256-17
CCCCCCCCNC(═O)C                  0.1277119
CCCCCCC(═O)OCCCC                 0.1383745
C(CCCCN)CCCC(═O)O                0.1551848
CC1(CCCSC1)O                     0.1810965
CCCCCCCCNC(═O)N                  0.201845
CCCCCCOOC(═O)C                   0.2354623
CCCCCCCC(═O)OCCC                 0.2373345
CCCCCC(═O)OOCCCC                 0.2398072
CCCCCCCCCC(═O)N                  0.2537211
C(CCCC(═O)O)CCCN                 0.256625
C(C)OC(═O)CCCCCCCC               0.2626695
C(CCCCC)OC(═O)CCCC               0.2654262
C1═CC═C(C═C1)CCCCCCN             0.2666345
CC(═O)NCCCCCCN                   0.2891708
CCCCCCCOC(═O)CCC                 0.2895618
CC1(NCCCN1)S                     0.2910104
CCCCCCCCOC(═O)N                  0.2941937
CCCCCCCCOC(═C)C                  0.2947368
C1═CC═C2C(═C1)NC(═O)O2           0.2981597
CCCCCCNCC(═O)O                   0.2988469
C1CCC(CC1)(O)S                   0.3023335
C1═CC═C(C═C1)CCCCCNO             0.3041483
CC(═O)CCCCOC(═O)C                0.3058697
C1═CC═C2C(═C1)NC(═O)N2           0.3085566
C1═CC2═CNON2C═C1                 0.3135061
Mor258-1
C1═CC═C2C(═C1)NC(═O)C═N2         0.04626341
CC1═NC2═CN═CN2C═C1               0.06964932
C1═CC2═C(N═CN2C═C1)C═O           0.07254368
CC1═CC2═C(C═C1)C═C(S2)N          0.07338813
C1CC(═CC═C1)C═C═O                0.08079227
CC1CC═NC═N1                      0.08086793
C1═CSC(═C1)C2═NN(C═C2)O          0.08123408
C1CNC(C═C1)C═O                   0.08133689
CC1═NC2═C(C═C1)NC═N2             0.08243389
CNC1═CC═NC═C1                    0.08274814
CC1═NC(═NC═C1C#N)C               0.08415395
CC1═CCC(═O)CC1                   0.08472221
C1═COC(═C1)C2═NN═NC═C2           0.08486507
CC1═CC2═NC═NN2C═C1               0.08532597
CC(═O)OCSC                       0.08674062
C1COC2═C1C═C(C═C2)C(═O)O         0.08754144
C1═CC═C2C(═C1)C═CN2N═O           0.08813593
C1═C2N═C(N═CN2N═C1)C#N           0.08876163
C1═CC2═C(N═C1)SC(═C2)C═O         0.088843
C1CC2═C(C═C1)NNC(═O)C2           0.08899985
C1CSN(N1)COC(═O)N                0.0890455
C1CC(SC1)CO                      0.09011572
CC1═NN(CS1)C                     0.09045361
C1CC2═C(C1)SC(═N2)N              0.09100104
C1C═CC2═CN═NC2═C1                0.09216601
Mor259-1
C1═CC═C2C(═C1)NC(═O)C═N2         0.04197753
C1═CC═C2C(═C1)NC(═N2)C═O         0.09240541
C1═CC2═C(NN═C2C═C1)C═O           0.0990186
C1═CC═C2C(═C1)C(═CC2═O)O         0.10483
CC(═O)NC1═CC═CC(═C1)C(═N)N       0.1129747
C1═CC2═COC(═C2C═C1)C═O           0.1146544
C1═CC═C2C(═C1)NC(═O)O2           0.11715
C1C2═CC═CC═C2ON1C═O              0.1209336
C1═CC═C2C(═C1)NC(═O)NN2          0.1401864
C1═COC(═C1)C2═CC═C(C═C2)O        0.1428006
C═C1C2═CC═CC═C2OC1═O             0.149781
C1═CC(═CC═C1C(═O)O)NCCC#N        0.1550444
CN═CC1═CC═C(C═C1)C(═O)OC         0.1553915
C1═CC═C2C(═C1)C═COC2═O           0.1566442
C═CC(═O)NC1═CC═C(C═C1)C(═N)N     0.1586167
C1═CC═C2C(═C1)C═NN2C═O           0.1601506
CN═NNC1═CC═C(C═C1)C(═O)OC        0.1621242
C1═CC═C2C(═C1)C(═CO2)N═O         0.1633878
COC1═CC═C(C═C1)NC(═O)C═C         0.1651134
CC(═O)NC1═CC═C(C═C1)N(C)C        0.1651899
CCC(═O)C1═CC═C(C═C1)OC           0.1667638
C1═CC═C2C(═C1)C═C(C═N2)O         0.1670003
CCNC1═CC═C(C═C1)N(O)O            0.1673194
CCC(═O)C1═CC═C(C═C1)NO           0.1704537
C1═CC═C2C(═C1)OS2(═O)═O          0.1718171
Mor260-1
COC(═O)CCNC1═CC═CC═C1            0.04581804
C═CC(═O)CCCCCCCC(═O)O            0.04753727
C#CCCCCCC═CC(═O)O                0.04908247
CCOC1═CC═C(C═C1)OC(═O)NC         0.05223202
CCCCC(CC)COCC(═O)O               0.06334015
C1CCC(C1)C(═O)CCCCC(═O)O         0.06629472
CCCCNC1═CC═C(C═C1)OC             0.06633087
CCCCNC(C)CC(═O)OC                0.0667401
CNC(═O)COC1═CC═C(C═C1)C#N        0.07077628
CCCCCCCCC(═O)C(═O)O              0.07088963
CCCCCCCCCOC(═O)O                 0.07169935
CCCCCC1CCC(CC1)C(═O)O            0.07283703
CCNCCC1═CC2═C(C═C1)OCO2          0.07466881
CC(═O)CCC(═O)NC1═CCCCC1          0.07735112
CCCC1CCC(CC1)CCC(═O)O            0.07825345
COC(═O)C═CCNC1CCCCC1             0.0817986
CCOC(═O)CCNC(═O)CC═C             0.08188413
CNCC1═CC═C(C═C1)N2CCCC2          0.0845592
CCCCCCOC(C)CS                    0.08794286
CCCCCC(═O)OC(C)S                 0.08888077
CCCCCCOC(═O)CC(═O)O              0.08920102
CCNCC1═CC═C(C═C1)C(═O)OC         0.08954888
CCCCOC1═CN═C(C═C1)C(═O)O         0.08968592
COC(═O)CNCCC1═CC═CC═C1           0.09056786
CCCCCCCC(═O)NCC                  0.09156643
Mor261-1
CCCCCC(CCO)O                     0.1725635
CC(═O)SCC1═CC═C(C═C1)C═C         0.2079287
CCCCC(CCCCO)O                    0.2205779
CCNC1═CN(N═C1)C2═CC═CC═C2        0.2643034
CCCCCCCCNN                       0.2799805
CC1═CC═C(C═C1)NCC(═O)NN          0.2822771
CC1═CC═C(C═C1)N2C═C(N═C2)CO      0.2832535
CCCOC(═O)C1═C(C═C(C═C1)N)O       0.2852718
CCCCN1N═C2C═CC═CC2═N1            0.2984866
CCCCNC(═O)OC1═CC═CC═C1           0.3016491
CCCC(═O)CCCCCO                   0.314819
CCCCCOC1═CCCC═C1                 0.323052
CCCCSC1═CC═C(C═C1)N              0.3336667
CCOC(═O)CC1═CC(═C(C═C1)N)O       0.3378758
CC1═CN═C(C═C1)NCCC(═O)O          0.3419532
CCOC1═CC═C(C═C1)NC(═O)NN         0.349067
CCCCC═CC═CCO                     0.3594656
CCCCCCC(CCO)O                    0.3622535
CCOC(═N)CCCCC#N                  0.3642063
CCCOC(═O)SC1═CC═CCC1             0.3687816
CCCOC1═CC═C(C═C1)CC(═N)N         0.3695899
CCCCCSC1CCCC1                    0.3784109
CCOC1═CC(═NO1)C2═CC═CC═C2        0.3805914
CC1═CC═C(S1)CNC2CCCC2            0.3893983
CC1═CN(N═C1)CCCC(═O)O            0.3914392
Mor268-1
C═C(CCO)CC1═CC═CC═C1             0.08680266
C1═CC═C(C(═C1)CCCCO)S            0.09776434
C1═CC═C(C═C1)CC2═C(OC═C2)CO      0.1173957
CC1═CC═CC═C1C2═CC(═NO2)CO        0.1177446
C═CCC1═CC═CC═C1C#CCO             0.12333
CCCC1═CC(═C(C═C1)CN)N            0.124973
CC1═CC═CC═C1CCCCO                0.1258311
C1═CC═C(C(═C1)CCCCN)N            0.132615
CCCCC1═CC═CC═C1CCN               0.1390437
C1═CC═C(C═C1)N2C═C(C(═N2)N)CO    0.1412993
CCCCCCN(C)C(═S)N                 0.1478017
C1CCCCCCCCC                      0.1517669
C═C═CCC1═CC═CC═C1CCO             0.1577928
CN1C(═CC(═N1)C2═CC═CC═C2)CO      0.1577991
C1═CC═C(C═C1)NNC(═O)CN           0.1627573
C1═CC═C(C(═C1)CCCCO)N            0.1671111
CC1N═C(NN1C2═CC═CC═C2)CO         0.1706577
CCCCCCC1(CNC1═O)C                0.1736143
C1CCC(═CCCCO)CC1                 0.1749776
CC(═CCCC(═CCN)C)C                0.1790916
C1═CC═C(C(═C1)C#CCCO)N           0.1844991
CC(═C)CCCCCCO                    0.1861627
C1═CC═C(C(═C1)CCCCN)S            0.192109
C(CCCCCC═C)O                     0.1949365
CC1═C(C═C(C═C1)CCCN)C            0.1985303
Mor271-1
CC(═O)CCC(═O)C                   0.00696499
CC(═C(C)N═NC)C                   0.02227481
CC═C(C)C(═O)OC                   0.02303129
CCC(═C)C(═O)OC                   0.03826216
CN(C)C(═O)CC═C                   0.03853879
COC(═C)C(═C)OC                   0.05227276
CC(C═C)C(═O)OC                   0.05905649
C1CCC(CC1)C2C═CCOO2              0.05928864
CCCC1SCCS1                       0.059703
CCOCC(═O)C(═C)C                  0.06042718
COC(═C(C#N)C#N)C1═CC═CO1         0.06070587
CCCC1(C(O1)(C)C)C(═O)OCC         0.06170662
CCN1C(═C(C(═C(C1═O)C)C)C)C       0.06559493
C═CC(═O)OC1CCCCCC1               0.06979654
CCN═C1N2CCCC2CCS1                0.07527722
CC(C)CC(═O)SC                    0.07652444
CC(C(═O)C═C)OC                   0.07826366
CSC1CCCCO1                       0.0787251
CC#CC(═O)N(C)C                   0.0789155
C═CCC(CC1CCCCC1)C#N              0.07892005
COC(═O)C(═C1CCCCC1)C#N           0.079376
C═C═CCN1C(COC1═O)C#C             0.08053451
CCC(═O)C1C(═O)CC(CC1═O)(C)C      0.08254048
CC1C═CN(C═CC1═C)C(═O)C           0.08378967
CC1═CC═CC(═C1C#N)C(═O)OC         0.08382356
Mor272-1
CC1CCC(═O)C1CCC#C                0.03189391
CC1═C(C/C═C\C)C(CC1)═O           0.0385644
CC1COC(O1)C(C)(C)C               0.03930837
CC(C)(C)C1OCCCO1                 0.04139548
CCCCCC(═O)C═C                    0.04630508
CCC(═C)COC(═O)C                  0.04645563
CC1═NC(OC1)C(C)C                 0.04785742
CCCC(═O)CCC                      0.04868388
CC(═O)CCCCCC                     0.05098426
CCCCCCC(═O)C                     0.05098426
CCC1═NC(CO1)(C)C                 0.05112795
CC1(C(═O)C(S1)(C)C)C             0.05313001
C(C)OC(═O)C(═C)C                 0.05386116
CCCC1C═C(C(═O)O1)C               0.05455699
CC(C)(C)C1OCC(═C)O1              0.05557511
C═C1CC(CO1)C2═CC═CC═C2           0.05582976
CC(C)(C)C(═O)N1C═CC═C1           0.05717043
CC1═NCCN1CC2═CC═CC═C2            0.05790857
CC(C)C(═O)OC═C                   0.05836228
CC(═C)C(═O)OC(═C)C               0.05855966
CC(═O)CCCCC═C                    0.05864676
CC(═O)OC1CCCC═C1                 0.05871441
CCOC(═O)C1(CC1)C                 0.06032142
CC(═CC)COC(═O)C                  0.06154408
CCC(C)COC(═O)C                   0.06223368
Mor273-1
CC1(C2CCC(C2)(C1═O)N)C           0.05501643
CCC1(C2CCC(C2)C1═C)CC            0.0736611
CC1CC2═CC═CC═C2N1                0.07383829
CCCCC1═C(OC═C1C)C                0.07413368
CC1CNC2═CC═CC═C12                0.07595554
CC(C)C(═O)OC═C                   0.0776473
CCC(C)NC                         0.08056587
CCSNC(C)C                        0.08292255
CCCCC1═C(CCCC1═O)O               0.08494866
CCCC(C)OC                        0.08782356
CC12CCCC1═CC═CC2                 0.08803756
CC1CC1C2═CC═CC═C2                0.08885662
CC1CCC(S1)C                      0.08906697
CN(C)C(═O)COC                    0.08911168
CC1C2═CCC(C2)C1(C)C              0.08918327
CC1CC═C(C1)C(═O)C(C)C            0.09082242
CC1CC2═C(C(═C(C12)C)C)C          0.09087068
CC1CC═CC12CCCC2═O                0.09092754
CC1CCCC═C1C(═O)N(C)C             0.09201093
CC1═C2C(═NC═NC2═NN1)N            0.09262628
CC1═CC(C(CC1)C(C)(C)O)O          0.09308545
CN1C═NC2═C1C(═NC═N2)N            0.0936681
CC(═NOC(═O)C(C)(C)C)C            0.09476478
CC(C)C1CCCCC1═O                  0.09521714
CC1═C(C(═O)N(C1═O)NC)C           0.09626298
Mor277-1
CC1(C2CC(═O)C1(C═C2)C)C          0.1480439
CC1C2(CCC1(C(═O)C2)C)C           0.1501646
C1CC(CC(═O)C1)N                  0.173522
C1C(C2═CC(═O)C(═O)CC2═N1)O       0.1843483
CC1CCCC(═O)C1                    0.1871075
C1CC(═O)C2CC═CC1S2               0.2020606
CCC12CCC(C1(C)C)CC2═O            0.2177276
C1CNC2═C(C1═O)C═CC═N2            0.2429412
CC1COCC(C1═O)C(C)(C)C            0.2469828
C1CC(═O)C2═C(NC1)N═CC═C2         0.2562849
C1CC(═O)C2═CC═CC═C2OC1           0.2580454
CC(C)C12CCC(C1)(CC2═O)C          0.2605051
CC1(C2(CCC1(C(═O)C2)C)C)C        0.2605627
C1C2CC3C═CC2CC1C3═O              0.2618416
CCC1(C(═O)CC12CC2)CC             0.2665882
C1CC(CC(═O)C1)S                  0.2680335
CC1CCCC12CCC2═O                  0.268307
CC1CCCC(═O)C1(C)C(C)C            0.2696565
CC1CCCC(═O)C1(C)C                0.2719524
CC1(CC(═O)CC2(C1O2)C)C           0.2734493
C1C═CNCC1═O                      0.2802712
CC(C)C1CC(═O)C═CN1               0.2822056
C1(═O)C═CCCC1                    0.2823245
CC(C)(C)C1CC(═O)C═CN1            0.2876126
CC1CC(═O)C═CN1                   0.2899973
Mor30-1
C(═O)(CCCCCCCCCC)O               0.3122994
C#CCCCCCCC(═O)O                  0.3287775
C═CCCCCCCCCC(═O)O                0.3395462
CC#CCCCCCC(═O)O                  0.3545879
C(═O)CCCCCCCCCC                  0.3841893
C═CCCCCCCC(═O)O                  0.4070303
C(═O)CCCCCCCCCCC                 0.4113624
CCCC#CCCCC(═O)O                  0.4176973
CCCCCC#CCC(═O)O                  0.4183262
CC#CCCCCCCCCC═O                  0.4361534
CCCCCCCCCCCCC═O                  0.4486903
C#CC#CCCCCCC(═O)O                0.4608185
CC(C)CCCCCCC═O                   0.4667896
CCC#CCCCCC(═O)O                  0.4774187
CC(═C)CCCCCCC(═O)O               0.478635
CCCCCCC1CC(═O)O1                 0.4816979
C═CCCCCCCC═O                     0.4847203
C#CCCCCCCC═O                     0.4893661
C═CCCCCCCCCCC═O                  0.4941443
C1C═C1CCCCCCC(═O)O               0.4958194
C═CCCCCC1CC(═O)O1                0.4999771
C(═O)(CCCCCCCC═C)O               0.5103754
C1C═C1CCCCCCCC(═O)O              0.5167787
CC(C)CCCCCCCCCC═O                0.5233348
C(═O)CCC═CCCCC                   0.5329278
Mor33-1
C1C═C1CCCCCCC(═O)O               0.03288254
C(═O)(CCC═CCCCCC)O               0.1022495
C(═O)(CC═CCCCCCC)O               0.1268822
CCCCCCC═CCCCC(═O)O               0.1272125
C(C)C(CCC(═O)O)CCCC              0.143634
CC(C)N1C═C(C═N1)CCC(═O)O         0.149314
C═C(CCCCC#N)C(═O)O               0.1541986
CC#CCCCCCCCC(═O)O                0.1626916
CC(CCC(═O)O)N═NC(C)(C)C#N        0.1652641
CC(C)CCCCCC(═O)O                 0.181324
CCCC(═C)CCC(═O)O                 0.1901408
CC(C)C1CCC(═CC1)CCC(═O)O         0.1912924
CCCCC═CCC(═O)O                   0.1939872
CCCCC#CC#CCCCC(═O)O              0.2014499
C1C═CC═CC1CCC(═O)O               0.2309548
C1C═C1CCCCCCCC(═O)O              0.2314075
C(═O)(CCCC═CCCCCCCCC)O           0.2344658
CC#CCCCCC(═O)O                   0.2381936
CCC1═CC(═C(C═C1)CCC(═O)O)C       0.247278
CCC1═CC═C(C═C1)CCC(═O)O          0.253656
CN(C)C1═CC═C(C═C1)CCC(═O)O       0.2546323
CC1(CCC(═CC1)CCC(═O)O)C          0.255894
CCC(C)OC(═O)CC(═O)O              0.2640391
CCCC(CC(C)C(═O)O)C#N             0.2650706
C#CCCCCCCC(═O)O                  0.2760099
Mor37-1
CCCCCCCCCCCCCC(OC(C)C)═O         0
CCCCCCCCC═O                      0
CCCCCCCCCC═O                     0
O═C(C)CC/C═C(CC/C═C(C)/C)\C      0
C(═O)(CCCCCCCC)O                 0
C(═O)(CCCCCCCCC)O                0
C(═O)(CCCCCCCCCCC)O              0
C(═O)(CCCCCCCCCCCC)O             0
C(═O)(CCCCCCCCCCCCC)O            0
C(═O)(CCCCCCCCCCCCCCC)O          0
CC1═C(C═C(C═C1)[C@H]C)CCC═C(C)C)O 0
CC(OCCCC/C═C\CCCC)═O             0
CC/C═C/CCCCCCCCCC([H])═O         0
CCCCCCCCCC[C@H](OC(C)═O)[C@@](O1)([H])CCCC1═O 0
O═C(CCCCCCC)N1CCC(C2═CC═CC═C2)CC1 0
O═C(CCCCCCCC)N1C(CC)CCCC1        0
O═C(CCCCCCCCC)N1C(C)CCCC1        0
O═C(N1CCC(C)CC1)CCCCCCCCC        0
O═C(CCCCCCCCC═C)N1CCCCC1         0
O═C(CCCCCCCCC═C)N1C(CC)CCCC1     0
O═C(CCCCCCCCC═C)N1CCC(C2═CC═CC═C2)CC1 0
O═C(CCCCCCCCC═C)N1CCC(C)CC1      0
O═C(CCCCCCCCCC)N1CCCCC1          0
O═C(CCCCCCCCCCC)N1C(C)CCCC1      0
O═C(CCCCCCCCCCC)N1CC(C)CCC1      0
Mor40-1
C1═CC═C(C═C1)CCC2═NN═C(C═C2)N    0.02341789
C(═O)(CCCCCCCC═C)O               0.03694565
COC1═CC═C(C═C1)CCCCC#N           0.04434623
C═CCCCCCCCC(═O)S                 0.04698026
CNC1═CC═C(C═C1)CCC(═O)OC         0.04732122
C(CCCC(O)O)CCCC(═O)O             0.04909072
CCCCC(═O)C1═CC═C(C═C1)OC         0.05102408
CCCCCC(═O)C1═CC═C(S1)O           0.05385012
CCCCCCC(CC(═O)O)O                0.0581458
CCCCCCCC(═O)OS                   0.0582002
CCCCC#CC(═O)CCCC                 0.05901835
CC(═C)CCCCCCC(═O)O               0.06104174
CCCCCCCC(═O)C#N                  0.06314685
CCCCCCCC(═O)C═C═C                0.0637365
COC(═O)CCCCCC═CC═C               0.06446715
CCCCCCC(═O)OC(C)S                0.06476652
COC(═O)C1═CC═C(C═C1)CCC═O        0.06633289
CCCCCCCC(═O)NO                   0.06904818
COC(═O)CCC1═CC═C(C═C1)C═C        0.07307333
CN(C(═O)CCCC1═CC═CC═C1)O         0.0740539
CCCCCCCC(═O)N(C)N═O              0.0767582
CCCC(═O)C═C═CC1═CC═CC═C1         0.07690233
CCCC(CCCC(═O)C═C)O               0.07774642
COC(═O)CCCCCCC#C                 0.07783589
COC(═O)CCC1═CC(═CC═C1)NN         0.07920649
Mor41-1
C1C(═O)C2═CC═CC═C2ON1            0.1690913
C1═CC═C2C(═C1)C═CC(═O)N2         0.19021
C1CC2═C(C═C1)OC(═O)C═C2          0.195202
CCCCC1═C(NNC1═O)C                0.1966665
CC═C1CCC(═O)CC1                  0.2004243
C1C═CC2═C(C1═O)C═CC(═O)O2        0.2108212
CCCCC1═C(OCC1═O)O                0.2185848
C1NC(═O)C2═CC═CC═C2S1            0.2228905
C1C(═O)C2═CC═CC═C2OS1            0.2251387
C1═CC2═NNN═C2C═C1C(═O)N          0.2380402
CC1CC1(C2═CC═CC═C2)O             0.2385417
CC1CCC(═CC1═O)C(C)C              0.2447182
C1═CC═C(C═C1)S(═O)N              0.2462735
C1CNC2═CC═CC═C2C1═O              0.2512114
CC1(C(═O)O1)C2═CC═CC═C2          0.252103
C1═CC═C2C(═C1)C(═O)NNN2          0.255054
C1CSC2═CC═CC═C2C1═O              0.2633634
C1═CC2═NSN═C2C═C1C(═O)N          0.2647783
CS(═O)C1═CC═CC═C1                0.2651372
C1C(═O)C2═CC(═C(C═C2S1)O)O       0.2656982
C1C(C(═O)N1)C2═CC═CC═C2          0.2658363
C1═CC2═NON═C2C═C1C(═O)N          0.2685111
CC(═C1CCCCC1)O                   0.2698948
C1═CC(═C(C═C1O)C(═O)NO)O         0.2742529
CN(C1═CC(═O)CCC1)O               0.2797649
Mor5-1
CCCCC(C)CC(C)C(═O)O              0.00047992
CCCN1CNC═C1CC(C(═O)O)N           0.00047992
CCCCC═CCC(═O)O                   0.00047992
CC1═CC═C(C═C1)CC(C(═O)O)N        0.00047992
C1CCC2═C(C1)CCNC2CC(═O)O         0.00047992
CCC(C)(C)C(C)C(CC(═O)O)N         0.00095984
C1CCC(C1)C═C═CCCC(═O)O           0.00095984
CCSCCCC(═C)C(═O)O                0.00095984
C1CCC2═C(C1)C═NC2CC(═O)O         0.00095984
CCCCN═C(C)CC(═O)O                0.00095984
C(CC(C(═O)O)N)CNCC(═N)N          0.00143976
C1CC(CC═C1)CCC(═O)O              0.00143976
C1═CN(C═N1)CCCCCCC(═O)O          0.00143976
CC(═C═C1CCCCC1)C(═O)O            0.00143976
CCCCCCC═CC(═O)O                  0.00143976
CC1═C(C═NN1C(C)C)CCC(═O)O        0.00143976
C1═CC═C(C(═C1)CCC(═O)O)CN        0.00143976
CCN1C(═C(C(═N1)C)CCC(═O)O)C      0.00191969
CCC(CCCCC(═O)O)S                 0.00191969
CCC1═CN═C(C═C1)CCC(═O)O          0.00191969
C(CCCC(═O)O)CCCS                 0.00191969
CCC1═CC═C(C═C1)N(CC(═O)O)N       0.00191969
CSCCCC(C(═O)O)N                  0.00191969
C1═CC(═CN═C1)CC(CN)C(═O)O        0.00191969
CCCCC1NC(CS1)C(═O)O              0.00191969
Or1A1
CCCCC(═O)CCC                     0.06049592
C═CCCC(═NO)C1═CC═CC═C1           0.07078234
CCCOCCC(═O)C═C                   0.0717614
C1CCCN(CC1)C(═O)N2CC2            0.07369093
C1CCC2═C(C1)C═CC═C2CC(═O)N       0.07424287
CCC(═NO)C1═CC═C(C═C1)C           0.07771264
CC1CC═C(NC1═O)C(C)C              0.07826973
C1CC(═O)N(N═C1)CC2═CC═CC═C2      0.07849599
CC1(OCC(CO1)C(C═C)O)C            0.07977273
CCCC(CC1═CC═CC═C1)C(═O)N         0.08055253
CC1═C2C═(C═CC2═NC═C1)C(═O)N      0.08126738
C1CC(C2═CC═CC═C2C1)CC(═O)N       0.08161454
COCC1═CC═CC2═C1CCCC2═O           0.08240595
CC(CC═C)OC(═O)C1═CC═CC═C1        0.08262245
CC(═O)C1═CC2═C(C═C1)OCCNC2       0.08372653
C1C═CC═C(C1═C═O)CC2═CC═CC═C2     0.0853223
CCOC(═O)CC(C)C1═CC═NC═C1         0.08749861
CC1═NC2═CC═CC═C2C═C1C(═O)N       0.08760062
CCC(C(═O)C)NC1═CC═CC═C1          0.08768825
CCCNC(═O)C1CCCN1C                0.0881405
C1C(NC(═O)CO1)C2═CC═C(C═C2)N     0.0887641
CC1═C(N═NC2═CC═CC═C12)C(═O)C     0.08925951
CCCCC(C1═CC═CC═C1)(O)O           0.08964993
CC1(OCCO1)CC2═CC═CC═C2           0.0898149
CC1(CCCNN1)C2═CC═C(C═C2)N        0.09071153
Or2J2
CCCCC(CCCCO)O                    0.1878163
CC(C)CCCCCCO                     0.2001934
C(CCCCN)CCCCON                   0.2449217
C═CCCCCCC1CO1                    0.2829139
CCCCCCNCC(CC)O                   0.2913393
CCCCCCCCN(C)O                    0.3036347
CC(C)CC1COC(N1)CCO               0.3482781
CCCCCCCCONC                      0.3506971
CCCCCCC(C)NCCO                   0.35406
CCCCCN1CCC(C1)CO                 0.3679643
CCCCCOC(C)CCO                    0.3976565
CCCCCCCC(C)CO                    0.4018389
CCCCCCCCNOC                      0.4019279
C(CCCCNCCO)CCCN                  0.4055974
CC(C)CCCCNCCO                    0.4075006
CCCC1CC(C1O)O                    0.4174384
CCCCC(C)(CC(C)O)O                0.4285046
CCCC1CCN(CC1)CCO                 0.4289913
CCCCCCCCCCNO                     0.4353896
C═CCCCC#CCO                      0.4379203
CCCCCCNC(C)CO                    0.4486621
CCC(C)CCCCCCO                    0.4517683
CCCCCCC1CCNO1                    0.4523566
CCCCC(CC(C)(C)O)O                0.459996
C1═CC═C(C═C1)C(═O)NCN═C═O        0.462984
Or2W1
CCCC(CCC═C(C)C)O                 0.00049109
CC(C)CC(═O)CC(C)C═C              0.00069469
C1CCC(CC1)C═NC(═O)CO             0.00069469
C1C═CNNC2═CC═CC═C21              0.00069469
CCCOC1═CC═C(C═C1)C               0.00085074
CCCCCC1CCC═CO1                   0.00085074
CC(C)NCCCNCCCN                   0.00098218
CC(═CCCC(C═C)C═O)C               0.00098218
C═C═CCCCCO                       0.00098218
CCCCCOCC#N                       0.00118032
CC(C)CCCC(C)C1CO1                0.00120303
CC(═O)C1CCC(═C)C(═C)C1           0.00120303
CC(C)(C)C(═O)OCCCO               0.00120303
CC(C)C(═C)C(═O)NC1═CC═CC═C1      0.00127841
CC(C)NC(═O)CC(CCN)N              0.00129418
COCCOCCOCC1═CC1                  0.00129719
CCCCC(═C)N                       0.00138938
CCCCC═C═CC(C)(C)O                0.00138938
COC(═O)CC1═CC═C(C═C1)C═C         0.00143135
CC(C)NCC1═CC═C(C═C1)NC           0.00145282
CN(C)CC#CCCCC#C                  0.00147327
C═CCCCCO                         0.00147361
CC#CC(═O)C1CCCCC1                0.00147361
CCCN1C═C(C═N1)C(C)NC             0.00147361
CC1(CC1C(═O)NC2═CC═CC═C2)C       0.00153553
Or5P3
C═CC(═O)C1CCCCC1                 0.3110759
CC1═CC(═O)C(CC1)C(═C)C           0.3586025
C1═CC═C2C(═C1)C═CC(═NO)O2        0.3721933
C1═CC═C2C(═C1)C═CNC2═S           0.3723315
C1═CC═C2C(═C1)C═COC2═O           0.3945556
C═C1C2═CC═CC═C2ONC1═O            0.3977094
C1═CC═C2C(═C1)C═CC(═O)N2         0.4009083
CC1═C(C(═O)CC1)CC═C              0.4052338
C1═CC(═CC2═C1C═CC(═O)O2)S        0.4053821
CC1═CCC(CC1═NO)C(═C)C            0.4125247
CC1(C2C1C(═O)C(═C)CC2)C          0.4507506
CC1C═CC(═O)C12CCCCC2             0.452584
CCC12CCC(═O)C═C1CC(C2)O          0.4690015
C1═CC═C2C(═C1)C═C(C(═O)O2)O      0.476153
C1CCC2(CC1)CCC═CC2═O             0.48111
C═C1CC2CCCCC2C1═O                0.4838933
C1═CC═C2C(═C1)C(═S)C═CN2O        0.4840491
C1CCC2(CC1)CC═CC(═O)C2           0.4924455
CC(═C)C(═O)CCC#C                 0.4929412
C═C1CC2(C1═O)CCCCC2              0.5054476
CC1═CC2CCCC(═O)C2═CC1            0.5090888
C1CC═CC2(C1)CC═CC(═O)NC2         0.5100904
CC1═C2C(═C)C(═O)NC2═CC═C1        0.5151608
C1═CC2═C(C(═C1)O)OC(═O)C═C2      0.516556
CC(═O)C1═CCC2(C1)C(═C)CCC2═O     0.5272902
CCC1═CC2CCCCN2C1═O               0.5776696
C1═CC═C2C(═C1)NC(═O)C═CS2        0.5831019
CC1═C2C═CC(═O)NC2═CC═C1          0.5841847
C1CC(═O)C2═C(C═C1)C═CC(═C2)O     0.5848835
C1═CC2═CN═C(C(═O)N═C2C═C1)N      0.5849539
C1═CC═C2C(═C1)C═CN3C2═NNC3═O     0.5850339
C═C1CC2═CC═CC═C2OC1═O            0.5862692
CCC(═C)CC1═CCCC1═O               0.5878639
CC1═CC(═O)OC2═C1C(═C(C═C2)O)N    0.5881438
CC12CC═CCC1CC(═O)C═C2            0.5900818
C═CCC1CCC(═O)C═C1                0.5902211
C1CC2(CCC═CC2═O)CC═C1            0.5931621
CC1═CC(═O)NC2═C1C(═CC═C2)N       0.5952647
CC1═C(CCC1(C)C)C(═O)OC           0.5982892
CC1(CC1C(═O)C═C)C                0.5994571
C1C2═CC═CC═C2C═C(C1═O)O          0.6030376
CC1═CCC(CC1═O)C2(CO2)C           0.6032878
CC1C═C(C(═O)O1)C2═CC═CS2         0.6051044
CC1═CC(═O)CCCC1CC═C              0.6074293
CC(C)C1CCC═C1C(═O)C              0.6076714
CC1═CC(═O)OC1C2═CC═CC═C2         0.6080884
CC1═CC(═S)C2═CC═CC═C2O1          0.6085333
C1C2═CC═CC═C2C(═O)C1═CN          0.608787
C1═CC2═C(C═CC(═O)O2)C═C1N        0.6088277
C1CC2(CC3CC2C═C3)C═CC1═O         0.6095803

The approach described herein was also used to predict activators of neurons that are responsive to CO₂. In order to train the platform to predict CO₂ neuron activators a large panel of odors was assembled that have previously been tested against CO₂ responsive neurons in several species. The panel comprises 108 odors, which have been tested against one or more of the following species: Anopheles Gambiae, Culex Pipiens, Aedes Aegypti, Drosophila Melanogaster. The panel consists of a broad collection of functional groups including alcohols, esters, acids, ketones, alkanes, aromatics, terpenes, and heterocycles. The activities of these odors were normalized from 100 to −100 representing the range from the strongest observed activator to the most inhibitory, respectively. Upon normalizing, it was observed that the strongest activators were heterocycles and some moderate activators were non-aromatic cyclic compounds. These distinct structural differences would likely drastically alter the outcome of the predictive platform. Due to this, the dataset was divided odors into two distinct sets. The first set focuses on activating odors with aromatic structures that look very structurally distinct from inhibitors. This set does not include non aromatic activators, activators which share structural characteristics with inhibitory odors, or odors which inhibit the receptor at greater than 30 percent of maximum. The second set is broader in scope and consists of odors both aromatic and non-aromatic structures as well as all inhibitory odors.

	Final	Training	Training
Odor Name	Activity	Set 1	Set 2
butanal	−85	No	Yes
pentanal	−51	No	Yes
hexanal	−32	No	Yes
heptanal	−21	Yes	Yes
octanal	−20	Yes	Yes
butanol	−25	Yes	Yes
pentanol	−41	No	Yes
hexanol	−70	No	Yes
heptanol	−38	No	Yes
octanol	−35	No	Yes
butanone	−25	Yes	Yes
pentanone	−28	Yes	Yes
hexanone	−19	Yes	Yes
heptanone	−12	Yes	Yes
octanone	−18	Yes	Yes
butyl acetate	−28	Yes	Yes
pentyl acetate	−15	Yes	Yes
hexyl acetate	−12	Yes	Yes
heptyl acetate	−8	Yes	Yes
octyl acetate	−15	Yes	Yes
butyric acid	−94	No	Yes
pentanoic acid	−26	Yes	Yes
hexanoic acid	−16	Yes	Yes
heptanoic acid	−14	Yes	Yes
octanoic acid	−21	Yes	Yes
pentane	−31	No	Yes
hexane	−25	Yes	Yes
heptane	−29	Yes	Yes
octane	−34	No	Yes
2,3-butanedione	−99	No	Yes
1-octen-3-ol	−27	Yes	Yes
Ethanol	−16	Yes	Yes
3-octanol	−14	Yes	Yes
Methanol	−14	Yes	Yes
Nonanol	−12	Yes	Yes
Eugenol Methyl Ether	−9	Yes	Yes
Acetic Acid	−7	Yes	Yes
gamma-valerolactone	−5	Yes	Yes
Fenchone	−2	Yes	Yes
Isoamyl Acetate	−2	Yes	Yes
Limonene	−2	Yes	Yes
Menthol	−2	Yes	Yes
(E)2-hexenal	0	Yes	Yes
Geranyl Acetate	0	Yes	Yes
Methional	0	Yes	Yes
Eugenol	1	Yes	Yes
4-methylphenol	3	Yes	Yes
Isopropyl Alcohol	3	Yes	Yes
Carvone	4	Yes	Yes
Phenylethanone	5	Yes	Yes
Anisole	6	Yes	Yes
Benzaldehyde	6	Yes	Yes
Benzophenone	8	Yes	Yes
Citronellal	8	Yes	Yes
Geraniol	8	Yes	Yes
Ethyl Acetate	8	Yes	Yes
Methylsalicylate	13	No	Yes
Thymol	15	No	Yes
Cyclohexanone	48	No	Yes
Indole	21	Yes	Yes
2-methylphenol	24	No	Yes
methyl pyruvate	−100	No	Yes
propionyl bromide	−88	No	Yes
propionyl chloride	−73	No	Yes
propionaldehyde	−68	No	Yes
2,3-pentanedione	−55	No	Yes
2-heptanol	−39	No	Yes
2-(propylamino)-ethanol	−39	No	Yes
butyryl chloride	−39	No	Yes
propionic acid	−32	No	Yes
2-methyl-3-heptanone	−26	Yes	Yes
3-heptanol	−16	Yes	Yes
4-(methylthio)-1-butanal	−15	Yes	Yes
4-hydroxy-2-butanone	−11	Yes	Yes
2,5-dimethylthiophene	−9	Yes	Yes
6-methyl-5-hepten-2-ol	0	Yes	Yes
1,5-pentanediol	0	Yes	Yes
1-hepten-3-ol	0	Yes	Yes
3-decanone	1	Yes	Yes
pyruvic acid	2	Yes	Yes
3-nonanone	2	Yes	Yes
4-heptanone	2	Yes	Yes
2-hexanol	2	Yes	Yes
1-bromohexane	3	Yes	Yes
1-hexanethiol	3	Yes	Yes
hexylsilane	3	Yes	Yes
phenylacetaldehyde	3	Yes	Yes
1-iodohexane	3	Yes	Yes
2,4,5-trimethylthiazole	5	Yes	Yes
ethyl valerate	5	Yes	Yes
cis-2-hexene	5	Yes	Yes
3-methyl-2-pentene	5	Yes	Yes
methoxyacetone	6	Yes	Yes
1-chlorohexane	8	Yes	Yes
cis-3-hexen-1-ol	10	Yes	Yes
fluoroacetone	10	Yes	Yes
acetophenone	15	No	Yes
2-acetylthiophene	31	No	Yes
pyridine	99	Yes	Yes
thiazole	100	Yes	Yes
2-ethyl-3,5-dimethylpyrazine	8	Yes	Yes
2,5-dimethylpyrazine	26	Yes	Yes
pyrazine	−8	Yes	Yes
naphthalene	−14	Yes	Yes

Optimized descriptors were calculated from the CO₂ neuron activity dataset 1. As activities for the odors have been averaged across the top 2 responders of the 4 species, only a single set of descriptors were optimized representing CO₂ responsive neuron activity. Molecular descriptors for this class of neuron was optimized using the same method as described above. To better visualize how well each Or-optimized descriptor set grouped CO₂ responsive neuron activators, all 78 compounds were clustered by distances calculated using the optimized descriptor sets. As seen in previous examples, highly active ligands clustered tightly for each Or. (See, e.g., FIG. 19).

Optimized descriptors were calculated from the CO₂ neuron activity dataset 2. As activities for the odors have been averaged across the top 2 responders of the 4 species, only a single set of descriptors were optimized representing CO₂ responsive neuron activity. Molecular descriptors for this class of neuron was optimized using the same method as described in above. To better visualize how well each Or-optimized descriptor set grouped CO₂ responsive neuron activators, all 104 compounds were clustered by distances calculated using the optimized descriptor sets. As seen in previous examples, highly active ligands clustered tightly for each Or. (see, e.g., FIG. 20).

Table 7 shows optimized descriptor sets calculated for CO2 activator set 1. The table shows the optimized descriptor subset calculated from activator dataset 1 as described in FIGS. 1-4 and 21. Optimized descriptor occurrences, symbol, brief description, class, and dimensionality are listed. Descriptors are listed in ascending order of when they were selected into the optimized set. Weights indicate the number of times a descriptor was selected in an optimized descriptor set.

symbol	breif description	class	dimensionality	occurrence
HNar	Narumi harmonic topological index	topological descriptors	2	1
R3v+	R maximal autocorrelation of lag 3/weighted by	GETAWAY descriptors	3	4
	atomic van der Waals volumes
HATS3m	leverage-weighted autocorrelation of lag 3/weighted	GETAWAY descriptors	3	1
	by atomic masses
Mor13p	3D-MoRSE - signal 13/weighted by atomic	3D-MoRSE descriptors	3	1
	polarizabilities
ISH	standardized information content on the leverage	GETAWAY descriptors	3	2
	equality
P1s	1st component shape directional WHIM index/	WHIM descriptors	3	1
	weighted by atomic electrotopological states
R4e+	R maximal autocorrelation of lag 4/weighted by	GETAWAY descriptors	3	1
	atomic Sanderson electronegativities
nRCHO	number of aldehydes (aliphatic)	functional group counts	1	2
JGI2	mean topological charge index of order2	topological charge indices	2	2
E1u	1st component accessibility directional WHIM index/	WHIM descriptors	3	2
	unweighted
MATS5m	Moran autocorrelation - lag 5/weighted by atomic	2D autocorrelations	2	1
	masses
STN	spanning tree number (log)	topological descriptors	2	2
DISPe	d COMMA2 value/weighted by atomic Sanderson	geometrical descriptors	3	1
	electronegativities
B06.C.O.	presence/absence of C—O at topological distance 06	2D binary fingerprints	2	1
X4A	average connectivity index chi-4	connectivity indices	2	4
JGI3	mean topological charge index of order3	topological charge indices	2	1
De	D total accessibility index/weighted by atomic	WHIM descriptors	3	2
	Sanderson electronegativities
Mor25u	3D-MoRSE - signal 25/unweighted	3D-MoRSE descriptors	3	1
nRCOX	number of acyl halogenides (aliphatic)	functional group counts	1	1
B03.O.O.	presence/absence of O—O at topological distance 03	2D binary fingerprints	2	1
nHDon	number of donor atoms for H-bonds (N and O)	functional group counts	1	1
MATS3e	Moran autocorrelation-lag 3/weighted by atomic	2D autocorrelations	2	1
	Sanderson electronegativities
RBF	rotatable bond fraction	constitutional descriptors	1	1
GATS5m	Geary autocorrelation - lag 5/weighted by atomic	2D autocorrelations	2	1
	masses
C.008	CHR2X	atom-centred fragments	2	1
Mor13v	3D-MoRSE - signal 13/weighted by atomic van der	3D-MoRSE descriptors	3	1
	Waals volumes
R6u.	R maximal autocorrelation of lag 6/unweighted	GETAWAY descriptors	3	1

Table 8 shows optimized descriptor sets calculated for CO2 activator set 2. The optimized descriptor subset calculated from activator dataset 2 as described in FIGS. 1-4 and 21. Optimized descriptor occurrences, symbol, brief description, class, and dimensionality are listed. Descriptors are listed in ascending order of when they were selected into the optimized set. Weights indicate the number of times a descriptor was selected in an optimized descriptor set.

symbol	breif description	class	dimensionality	occurrence
N.075	R—N—R/R—N—X	atom-centred fragments	2	1
R3v.	R maximal autocorrelation of lag 3/weighted by atomic	GETAWAY descriptors	3	1
	van der Waals volumes
H.049	H attached to C3(sp3)/C2(sp2)/C3(sp2)/C3(sp)	atom-centred fragments	2	1
nRCHO	number of aldehydes (aliphatic)	functional group counts	1	1
		constitutional
nN	number of Nitrogen atoms	descriptors	1	1
ISH	standardized information content on the leverage	GETAWAY descriptors	3	1
	equality
EEig07d	Eigenvalue 07 from edge adj. matrix weighted by dipole	edge adjacency indices	2	1
	moments
piPC04	molecular multiple path count of order 04	walk and path counts	2	1
MATS4e	Moran autocorrelation - lag 4/weighted by atomic	2D autocorrelations	2	1
	Sanderson electronegativities
ESpm14d	Spectral moment 14 from edge adj. matrix weighted by	edge adjacency indices	2	1
	dipole moments
Mor12m	3D-MoRSE - signal 12/weighted by atomic masses	3D-MoRSE descriptors	3	1

Table 9 shows the top 500 predicted compounds for CO2 activator set 1. The top 500 predicted compounds for predictions made from activator dataset 1.

SMILES Structures	Distance	SMILES Structures	Distance
c1ccncn1	2.033459	CN(C)CCc1cccnc1	3.487232
Cn1cncc1	2.170379	OCCCCCNCc1ccncc1	3.491708
C1═NC═CN1	2.297704	OCCC(C)c1ccncc1	3.499793
c1ncc[nH]1	2.297704	Cc1cncc(c1)c1ccccc1	3.508292
Cc1cnc[nH]1	2.409222	C═COCCNCCc1ccccn1	3.512242
OCCCCc1ccncc1	2.551873	CCOCCc1ccccn1	3.515488
Cn1cccn1	2.646606	O1CCC(CC1)c1ccncc1	3.517108
CCCCCc1ccncc1	2.66955	C#Cc1ccncc1	3.522205
Cn1cncn1	2.683402	NNCc1cccnc1	3.522389
N1N═CC═C1	2.768945	Cc1cncc(C)c1	3.522667
c1ccn[nH]1	2.768945	C1CCC(CC1)c1ccncc1	3.522853
c1ccnnc1	2.785637	CCCOCCc1ccccn1	3.526678
CC(O)/C═C\c1cccnc1	2.801027	CCCn1cncc1	3.533624
CC(C)CCCc1ccncc1	2.810277	OCCNCc1ccccn1	3.533863
Cc1ncc[nH]1	2.819446	CCCCCC1═NC═CC═C1	3.534064
Cc1c[nH]cn1	2.821426	CCCCCc1ccccn1	3.534064
CCCCC/C═C/c1cccnc1	2.848973	CCCCC(CC)CCc1ccncc1	3.541254
C═COCCNCCc1ccncc1	2.856722	CCCC(C)Nc1ccccn1	3.547778
Cc1c[nH]nc1	2.864148	n1ccc(cc1)CNC1CCCC1	3.549867
CCCc1ccncc1	2.867075	c1ccc(CNCc2ccc[nH]2)cn1	3.555656
CCCCc1ccncc1	2.88623	c1ccc(cc1)\C═C/c1ccncn1	3.55859
CCCCCCCCc1ccncc1	2.889588	CN(C)CCc1ccncc1	3.561808
c1cnn[nH]1	2.890173	c1ccc(cc1)c1cnc[nH]1	3.574979
CC(C)CNCc1ccncc1	2.891618	CC1OCCC(C1)c1ccncc1	3.577056
Cc1ccn[nH]1	2.894352	CNCCc1cccnc1	3.577696
CNCC/C═C/c1cccnc1	2.904657	N1CCC(CC1)Cc1ccncc1	3.581065
NCCCc1ccncc1	2.919508	CCOCCCNCc1ccncc1	3.582788
CCCCCCNCc1ccncc1	2.930239	OCCC(N)c1ccncc1	3.58383
OCCCc1ccncc1	2.963816	C#CC/N═C\c1ccccc1	3.584109
CC(O)CCCc1ccncc1	2.972694	CC(N)Cc1cccnc1	3.586069
C/C═C/CCc1ccccn1	2.992266	NNc1ccncc1	3.589658
CCCNCc1ccncc1	2.995958	c1ccc(cc1)Cc1ccncc1	3.589705
CCCCCCCCn1ncnc1	2.997064	C1CCC(CN1)Cc1cccnc1	3.59285
CC1═CSC═N1	3.028959	NCCCc1ccccn1	3.592971
Cc1cscn1	3.028959	C═Cn1cncc1	3.595285
NCCNCCc1ccncc1	3.055465	OCCCCCNCc1ccccn1	3.598167
Cc1n[nH]cn1	3.057359	OC1CCN(CC1)Cc1ccncc1	3.599777
C1CCC(CC1)Cc1ccncc1	3.067562	NNc1cccnc1	3.608748
OCCNc1ccncc1	3.097835	c1ccc(cn1)c1ccccc1	3.609145
N1CCC(CC1)Cc1cccnc1	3.125454	c1ccc(cc1)Cc1cccnc1	3.611841
CCCCCCc1cccnc1	3.13767	COc1cncc(OC)n1	3.612227
OCCCCCNCc1cccnc1	3.138696	COCCc1nccc(C)c1	3.615322
OCCNCCc1ccncc1	3.151658	CCCCCCn1cncc1	3.618375
CCCc1cccnc1	3.161792	CCCCOc1ccncc1	3.618896
CCCCNCc1ccncc1	3.163311	CC(N)CCn1cccn1	3.624274
Nc1cncs1	3.164302	C1CCC(CC1)CCCc1ccncc1	3.630432
CCCCc1cccnc1	3.17431	c1ccc(cc1)CCCc1ccncc1	3.635746
OCCNCc1cccnc1	3.190109	C═CCc1ccccn1	3.6393
Cc1csnc1	3.190481	CN(N)c1ccncc1	3.642521
Cc1nccs1	3.19845	CCCCCCCCn1ccnc1	3.647362
CNCCCNCc1ccncc1	3.22583	C1CCC(CC1)NCc1ccncc1	3.652448
CNCCc1ccncc1	3.232084	C1CCC(CC1)Nc1cccnc1	3.660288
OCCNCCNCc1ccncc1	3.2344	CCCCCOc1cccnc1	3.663809
CCC(CO)NCc1ccncc1	3.238621	OCCc1cccnc1	3.665454
C═CCNCc1ccncc1	3.239061	n1ccc(cc1)C1CNCC1	3.668676
CCC(C)NCc1ccncc1	3.249082	CCCCCc1cccnc1	3.669409
CCCCN(C)Cc1ccncc1	3.262013	CC(C)CCNCc1cccnc1	3.675901
CC(C)Cc1ccncc1	3.276179	CC(C)NCc1ccncc1	3.682412
C═CCNc1cccnc1	3.284046	Oc1cncc(O)c1	3.684342
OCCCNCc1ccncc1	3.28714	CCCCNCc1ccccn1	3.685337
CC(C)Cc1cccnc1	3.290287	c1ccc(cc1)\C═C/c1ncccn1	3.688202
CCc1cncc(C)c1	3.296279	C1CNC(C1)Cc1cccnc1	3.693127
C1CCC(NC1)Cc1ccncc1	3.299132	CNc1ccccn1	3.69627
OCCCc1cccnc1	3.300138	CCn1cncc1	3.697594
CCC(CO)NCc1cccnc1	3.301803	CCNCc1ccccn1	3.69802
n1ccc(cc1)CCn1cccc1	3.309288	Cc1ccc(cn1)c1ccccc1	3.701697
CN(C)/N═C/c1ccccn1	3.309846	CCCCC(CCCC)c1ccccn1	3.705387
CCOCCc1ccncc1	3.313444	COc1cc(O)cnc1	3.705406
OCCOCCNCc1cccnc1	3.324462	CC(C)OCCCNCc1ccncc1	3.706372
CCNCc1ccncc1	3.332404	CNc1ncccn1	3.706648
OCCNCc1ccncc1	3.355513	NCCc1cccnc1	3.710193
CC(N)Cc1ccncc1	3.358047	CCc1ccccn1	3.71041
C═Cc1ccncc1	3.364881	CCCCc1cnc(N)nc1	3.710859
Nc1nccs1	3.369255	c1ccc(cc1)CCc1ccncc1	3.713706
OCCC(CCO)c1ccncc1	3.38942	Cc1ccc(cc1)c1ccncc1	3.71477
NCCc1ccncc1	3.396504	C/N═C/c1ccccc1	3.716001
C1COC═N1	3.400156	CCO/C═N/c1ccccc1	3.717426
CC(C)CCCc1ccccn1	3.404779	CCO/C═N\c1ccccc1	3.720444
C═Nc1ccccc1	3.40591	COCCNCc1cccnc1	3.721211
CC(O)/C═C/c1ccncc1	3.408198	CN1C═CC═C1	3.721854
c1ccns1	3.419366	Cn1cccc1	3.721854
CN(C)CCNCc1cccnc1	3.422927	Nn1cccc1	3.722507
COCCCNCc1ccncc1	3.423191	CN(C)CCN(C)Cc1cccnc1	3.726622
CN(C)CCCc1ccncc1	3.429515	CCCCC(N)c1cccnc1	3.728514
CN(C)CCCNCc1ccncc1	3.43432	CC(O)/C═C/c1ccccn1	3.728985
Nc1ccn[nH]1	3.442882	n1ccc(cc1)CNC1CC1	3.733392
C#CCNCc1ccncc1	3.448445	Nc1cncc(N)c1	3.735731
n1ccc(CCNC2CC2)cc1	3.449431	c1ccc(cc1)CNc1cccnc1	3.739868
CNCc1ccncc1	3.449774	CC(C)CNc1ncccn1	3.74062
CC(C)Nc1cccnc1	3.450399	C/C═C/CC(C/C═C/C)c1ccncc1	3.743814
N1CCC(CC1)c1ccncc1	3.45071	c1cnc(nc1)c1ccccc1	3.749999
OCCNCCNCc1cccnc1	3.461336	CCCCCn1cncc1	3.751592
OCCOCCNCc1ccncc1	3.462592	COc1ncccn1	3.752011
CCCCNc1ncccn1	3.463639	CCCCC(CCCC)c1ccncc1	3.753138
OCCC(═C)c1ccncc1	3.464044	Nc1ccccc1CCc1cccnc1	3.75703
C═CCCCCCCCC/C═C/c1cccnc1	3.46455	CCC1═NC(═CN═C1)C	3.758666
c1ccc(cc1)\C═C/c1ccncc1	3.469761	NCCc1ccncn1	3.763006
c1nnn[nH]1	3.470708	CC(N)Cc1ccccn1	3.763389
Cc1ccc(\C═C/c2ccncc2)cc1	3.478825	[nH]1nc[nH]nc1	3.767631
c1ccc(nc1)CCc1ccccn1	3.4792	OCCCn1cncn1	3.770534
Nc1cccc(\C═C/c2ccncc2)c1	3.772142	Cc1ncc(N)nc1	3.921502
Cc1cc[nH]n1	3.772838	c1ccc(cc1)c1ccccn1	3.92194
C═C/C═C\CCCCCCCCO	3.778543	C1CCC(NC1)CCn1cncc1	3.922694
CC1CCCCN1Cc1ccncc1	3.779824	Nc1ncc(s1)c1ccccc1	3.924187
C1CCCC(CCC1)NCc1ccncc1	3.781268	CCCCOCCCNCc1ccncc1	3.924987
CCCNCc1ccccn1	3.781866	CCCCCCCCOCn1cncn1	3.92499
CC(C)CCc1nccnc1	3.785656	CCCCCCCCNCc1cccnc1	3.927536
CCOCn1cncn1	3.7873	NCCCC(N)c1ccncc1	3.928649
c1ccc(nc1)Cn1cccc1	3.788948	NCCNCCNCCc1ccncc1	3.929749
CCN(CC)Cc1ccccn1	3.789854	N1CCC═CC1	3.929961
COCCNCc1ccccn1	3.792512	OCCNCCNCc1ccccn1	3.930056
n1ccc(nc1)c1cscc1	3.804935	OCCC(CC)c1ccncc1	3.933383
NCCCc1cccnc1	3.805856	Oc1cnc(nc1)c1ccccc1	3.934528
C═Cn1cncn1	3.807139	Cn1cnnn1	3.93691
Nc1cnccc1c1ccccc1	3.812096	Cn1nncn1	3.939852
Cc1ccc(CNCc2cccnc2)s1	3.812097	C1CNC(CO1)c1cccnc1	3.940647
CCC(N)c1ccncc1	3.812101	CCCCCNCc1ccncc1	3.942016
CC1═CN═CC(C)═N1	3.815402	Cc1ncnc(N)c1	3.94231
Cc1cncc(C)n1	3.815402	CCCCOc1ccccn1	3.94276
CCn1cncn1	3.815701	CCCCC(O)Cc1ccccn1	3.943205
CCC1OCCC(C1)c1ccncc1	3.816733	C═CCn1cccn1	3.943451
C1═CC═CS1	3.817309	OCCCCCCCCCCCc1ccccn1	3.943807
c1cccs1	3.817309	C1CCCC(CCC1)Nc1ncccn1	3.947406
CCCCCCCNCc1ccccn1	3.821691	N1CCC(CC1)Cn1cncc1	3.948581
Cc1ccc(\C═C/c2ccccn2)cc1	3.822084	NCCCC(N)c1cccnc1	3.949274
CCC(CC)c1ccncc1	3.824604	C1NCC(C1)Cc1cccnc1	3.952809
Cc1nccnc1CCO	3.829076	CC(C)/N═C/c1ccccc1	3.953581
n1ccc(cc1)C1CCCN1	3.82945	CCCCCCCCCCCCCc1ccncc1	3.954487
COCCCNCc1ccccn1	3.829925	COCCNCc1cnn(C)c1	3.956375
CC1═NC═CN═C1CC	3.82998	O═CNc1ccccn1	3.95891
CCc1nccnc1C	3.82998	c1ccc(cn1)c1ccc[nH]1	3.961611
OCCCc1cnn(C)c1	3.831839	c1ccc(cn1)Cn1cccc1	3.962207
NNCCc1ccncc1	3.836858	Cc1cc(CC═C)ncc1	3.964769
CC1OC(C)CC(C1)c1ccncc1	3.837242	C═Cc1ccccn1	3.966734
OCCc1cncc(C)c1	3.837984	c1ccc(cc1)c1cncs1	3.967009
Nc1c[nH]cn1	3.838209	C═CCn1cncc1	3.967641
N1CCC(CC1)CCn1cncc1	3.838228	C1CCCC(CCC1)NCc1cccnc1	3.96989
Cc1ccc(cc1)c1nccs1	3.838297	Nc1ccc(cc1)Cc1cccnc1	3.971588
Nc1ncc[nH]1	3.838354	Cc1nccnc1CCCC	3.972325
Cc1cnc(nc1)c1ccccc1	3.839282	COc1ccccn1	3.972372
CCOc1ccncc1N	3.839786	CCCCc1ccccn1	3.972673
Nc1ccc(CCc2ccncc2)cc1	3.840154	Cc1ccc2cnccc2c1	3.974412
CCCCNCc1cccnc1	3.840485	Cc1ccnc(c1)c1ccccc1	3.975783
CCc1ccc(C═C)nc1	3.844342	c1csc(c1)\C═N/N═C/c1cccs1	3.977461
n1ccc(cc1)C1CC1	3.846694	Cc1cccnc1	3.97977
CC(C)CNCc1ccccn1	3.847544	NCCCCc1ccccn1	3.980221
c1ccc(cc1)CCCc1cccnc1	3.848581	OCCNCc1cnn(CC)c1	3.98241
NCCC(CCO)c1ccncc1	3.849054	C1CCC(CN1)COc1cccnc1	3.982921
C1CCC(CC1)Nc1nccs1	3.851878	c1ccc(cc1)\C═C/c1ccccn1	3.983757
CCC1═NC═CN═C1CC	3.853768	CN(C)/C═N/c1ccccc1	3.984609
CCc1nccnc1CC	3.853768	OCC(C)Cn1cccn1	3.985181
CCCCn1cncc1	3.855689	C═Cc1ccc(C)nc1	3.98793
c1ccc(nc1)c1ccccn1	3.857047	Cc1cncs1	3.988318
CCCc1ccccn1	3.858559	COCn1cncc1	3.989514
n1ccc(cc1)CCNc1ccccc1	3.858864	CCC(C)n1cccn1	3.990983
n1ccc(cc1)C1CCC═CC1	3.859078	CCCCOc1ccc(CNCc2cccnc2)cc1	3.99149
Oc1cccs1	3.861787	OCn1cncc1	3.9916
C1CCNC(CC1)c1ccncc1	3.862012	CCC(C)NCc1cccnc1	3.992449
C1CCC(CNCCCn2cncc2)CC1	3.862902	CC(C)NCc1ccccn1	3.994445
c1ccc(nc1)Nc1ccccn1	3.864257	c1ccc(cc1)COCc1cncs1	3.994736
Nc1nccc(c1)c1ccccc1	3.86608	CCCN(CC1CC1)Cc1ccncc1	3.995188
n1ccc(cc1)Cn1cccc1	3.867767	Cc1ccccc1Cc1cccnc1	3.996041
NCCCNc1cccnc1	3.869999	CC(C)c1ccccn1	4.00194
OCCc1ccncn1	3.872425	CCCCCCNCc1cccnc1	4.002367
CCCc1ccc(O)cn1	3.873872	COc1ncc(N)cn1	4.00385
Nc1ccc(cn1)c1ccccc1	3.87394	C1CCc2nccnc2C1	4.004779
Nc1nncs1	3.876929	c1ccc(cc1)NCc1ccncc1	4.005823
c1ccc(cc1)c1c[nH]cn1	3.877242	C═Cc1nccnc1C	4.006881
c1ccc(cc1)Cc1ccccn1	3.881093	O═Cc1cnc(s1)c1ccccc1	4.007171
OCCCc1cnn(CC)c1	3.882615	CC(N)c1ccncc1	4.009614
c1ccc(cc1)c1ncc[nH]1	3.883167	CCCC(CCC)c1ccccn1	4.012339
c1ccc(cn1)c1n[nH]cc1	3.885425	OCC1CCN(CC1)Cc1ccncc1	4.012772
NCCCCn1cncc1	3.885956	Cc1cc(C═O)cnn1	4.013137
c1ccc(nc1)\C═C/c1cccnc1	3.886201	CNCCCC(O)c1cccnc1	4.013331
COCCc1ccccn1	3.886334	NCCCCCc1cnc[nH]1	4.014127
C12═CC═CC═C1N═CC═N2	3.889099	COCC(NC)c1ccccn1	4.016933
c1ccc2nccnc2c1	3.889099	CCCCc1cnccn1	4.017648
c1ccc(cc1)c1ccn[nH]1	3.890727	CCCCCCC(C)Cc1ccncc1	4.017949
c1ccc(nc1)NC1CCCC1	3.892768	O1CCN(CC1)c1ccncc1	4.018166
OCCNCc1cnn(C)c1	3.893737	C1N═Cc2ccccc2C═C1	4.018324
CCOc1ccccn1	3.901646	CCCN1CCC(CC1)NCc1ccncc1	4.01907
Cc1ccnc(C)c1	3.902494	NCCNc1cccnc1	4.019396
COCc1ccccn1	3.907315	CCCCCn1ccnc1C	4.020339
NCCc1cncs1	3.909061	OCCc1ccncc1	4.023571
Nc1cccc(c1)c1ccncc1	3.90929	C═CCn1cncn1	4.024219
CCCCCCCCCCCCn1ccnc1	3.910022	n1ccc(cc1)C1CCCCN1	4.024864
CCC1CCC(CC1)NCc1ccncc1	3.910398	OCCNc1ccncc1N	4.025311
CCCCCCn1cncc1C	3.911132	CCCCN(CCCC)c1ccncc1	4.025433
n1ccc(cc1)\C═C/c1ccccn1	3.91129	Cc1ncc(C)cn1	4.027625
Oc1cnsn1	3.912538	CC(═N)NCCc1ccncc1	4.028513
CCC1═NC═CN═C1	3.914359	CC(C)Cn1cccn1	4.028995
CCc1cnccn1	3.914359	c1ccc(nc1)NCc1cccs1	4.030544
NCCNc1ccccn1	3.914758	C═Cc1cccnc1	4.031618
c1cnc(nc1)NC1CCCC1	3.915571	CCC/N═C/c1ccccc1	4.032101
CCC(CC)c1ccccn1	3.915951	COc1ncc(O)cn1	4.032331
Cc1ccc(CNc2cccnc2)cc1	3.917527	Nc1ccc(cc1)Cc1ccncc1	4.032768
OCCC(CCO)c1ccnc(C)c1	3.919641	CC1OC(C)CN(C1)c1ccncc1	4.034999
Oc1ccccc1\C═N/c1nccs1	3.919723	CCN(CC)CCc1ccccn1	4.035536
COCCNCc1ccncc1	3.920298	n1ccc(cc1)CNCc1cccs1	4.037681
OCc1cncn1CC	3.920528	CC1═NC═CN═C1OC	4.038081
O═CNc1nnc(CC(C)C)s1	4.075545	COc1nccnc1C	4.038081
Cc1cccnc1c1ccccc1	4.075705	CCN(CC)CCc1ccncc1	4.038201
C1CCC(CN1)Oc1ccncc1	4.075903	OCC1CCN(CC1)Cc1cccnc1	4.038372
N1CCC(CC1)CCc1ccccn1	4.077268	COC1═NC═CN═C1CC	4.03881
c1ccc(nc1)c1ccncc1	4.078491	COc1nccnc1CC	4.03881
CCCCc1ccc(C═O)cc1	4.07867	CCn1cccn1	4.039322
c1ccc(nc1)C1CCOCC1	4.080083	CCOC(OCC)Cn1cccn1	4.039711
c1ccn2nccc2n1	4.081331	N1CCC(CC1)Cc1ccccn1	4.0407
CC(CC)n1cncn1	4.082857	N1═CNCCN═CNCC1	4.040829
OC1CCN(CC1)Cc1cccnc1	4.083942	NCCCn1cncn1	4.041459
Cc1nncc(c1)c1ccccc1	4.084906	Cc1ccc(cc1)c1cncnc1	4.043691
CCC/C═N/Nc1ccccc1	4.087764	CCOc1cnccc1OCC	4.044495
C/C═C/CC(C/C═C/C)c1ccccn1	4.089417	NCC(C)c1ccncc1	4.044646
COCC(N)c1ccccn1	4.0913	CN(C)Cc1ccccn1	4.045576
CC/C═C\C/C═C\C/C═C\CO	4.092416	OCCCNC(C)c1ccncc1	4.047575
CCC(N)Cn1cncc1	4.092784	C═CC1═CN═CC(C)═N1	4.048223
CC(N)Cn1cncc1	4.093284	NCCCOc1cccnc1	4.048408
OCCCNCc1ccccn1	4.094216	Cc1ccc(cc1)c1ccccn1	4.048665
C/N═C/c1ccc(cc1)C(C)C	4.094231	Cc1ccnc(c1)c1nccc(C)c1	4.049527
C[C@H](O)c1ccncc1	4.094662	CCCCC(CCCC)c1cccnc1	4.049925
CC(O)c1ccncc1	4.094662	C1CNC(C1)Cn1cccn1	4.052425
CC(═C)Cc1ccccn1	4.095187	CC(C)CNc1ccccn1	4.053681
CN1CCN(CC1)Cc1cccnc1	4.096565	Nc1cnc(nc1)c1ccccc1	4.053727
c1csc(c1)\C═N/N═C\c1cccs1	4.097744	N1CCCN(CC1)Cc1ccncc1	4.053805
C═CCNc1nccs1	4.098084	CC(C)Nc1ccncc1N	4.055925
CC(C)CC(C)c1ccccn1	4.098869	n1ccc(nc1)c1cccs1	4.056535
CCCCNc1ccncc1N	4.100089	CCOc1ccc(CNCc2cccnc2)cc1	4.056591
OC1CCCN(C1)Cc1cccnc1	4.100969	NCCCCCC(N)c1ccccn1	4.056898
OCc1ccc(CO)cn1	4.104339	CCC(NC)c1ccccn1	4.056979
COCCn1cncc1CO	4.104345	NCc1ccc(s1)c1ccncc1	4.058566
OCc1cnc[nH]1	4.104405	Cc1cccc(c1)c1cncnc1	4.058849
OCCC/N═C\c1cccs1	4.106196	CC(C)Oc1nccnc1C	4.059391
C1CCN(C1)c1nccs1	4.109143	Nc1n[nH]cn1	4.059558
C1CCC(NC1)CCc1ccccn1	4.11042	Nc1cncc(O)c1	4.061363
CCCn1cc(N)cn1	4.111561	NC1CCCN(C1)Cc1cccnc1	4.061742
Cc1c[nH]cc1	4.111597	COCn1cncn1	4.062793
Nc1ccc(CCc2cccnc2)cc1	4.113066	Nc1ccc(cc1)c1ncsc1	4.064229
C1C═CC═C1	4.113191	CCOc1cnc(C)cn1	4.06474
OCC(N)c1cccnc1	4.113658	Cc1nccnc1CCC	4.066969
c1ccc2ccncc2c1	4.113947	Nc1cc(CO)cnc1	4.067234
c1ccc(CNCc2cccs2)cn1	4.114262	OCCCc1ccnc(C)c1	4.068903
CC(N)Cn1cncn1	4.115272	c1ccc(cc1)CNc1ccccn1	4.069396
CCCCn1nccc1C	4.115424	CCCCNc1nccs1	4.072269
CCC(CO)NCc1ccccn1	4.118634	Cc1ccc[nH]1	4.072803
OCCn1nccc1N	4.119024	Nc1cnc(C)nc1	4.073371
CCCCn1ccnc1C	4.119754	CCCCCCCNCc1ccncc1	4.073913
Nc1ccccc1c1ccncc1	4.121111	Nc1ccc(nc1)c1ccccc1	4.073915
O═Cc1ccnn1CC	4.121672	CC1═COC═C1	4.074861
OCCCCCCc1cccnc1	4.122515	c1cnc2cccnc2c1	4.074967
CC1═NC═C(CC)C═C1	4.124326	C1CNC(C1)Cn1cncn1	4.075254

TABLE 10
Top 500 predicted compounds for CO2 activator set 2. The top 500 predicted
compounds for predictions made from activator dataset 2.
SMILES Structures	Distance	SMILES Structures	Distance
O═C1CCNCC1	1.66855	O═C1CNCC1	5.644213
O═C1CCCCCN1	2.03737	C1COC═N1	5.670562
O═C1CCCCCN1	2.03737	S═C(NCCc1ccccc1)NC(C)(C)C	5.700199
O═C1NCCCC1	2.16724	N#CC(═C1CCCCCCCCCCC1)C#N	5.701419
O═C1CCCCN1	2.16724	COC1═CC═CCC1	5.71922
O═S1CCNCC1	2.17659	CC(C)(C)NCc1ccccc1	5.720564
O═C1NCCOCC1	2.43129	CC(C)CC(═O)c1cccnc1	5.7212
O═C1CCNCCC1	2.50405	NC1CCCCC1	5.72578
O═C1CCC═CCC1	2.67564	CCc1cncc(C)c1	5.744045
Oc1ccncn1	2.97767	CNC1CCCC1	5.745127
C1═CC═CS1	2.99768	OC1CCNCC1	5.749396
c1cccs1	2.99768	CC(C)═C[C@H]1C[C@@H](C)CCO1	5.751913
O═C1CCOCC1	3.06572	COC1CCC═C1	5.761583
CC1CCOCO1	3.1695	CC(C)OC(═O)C1CCCCC1	5.765544
O═S1CCCCC1	3.34159	CC(═CCOCC1CCCNC1)C	5.773864
O═C1CCCCCC1	3.37411	C1CCCSC1	5.776573
Cc1ccncc1	3.44412	C1CCSCN1	5.78245
O═C1CCCNC1	3.53322	O═Cc1c[nH]cn1	5.783256
O═C1CCCCO1	3.6966	CCOC(═O)N1CCCCCC1	5.783974
NC1CCOCC1	3.73016	C1COC═CC1	5.794384
O═CN1CCCC1	3.75356	COC1═CCCCC1	5.795498
O═C1CCCCC═C1	3.83862	O═C(OC(C)(C)C)c1cccnc1	5.795791
NC1CCSCC1	3.88042	OC1CCCCO1	5.807573
CC1═CC═NC(═O)C1	3.88476	OC1CCCCCC1	5.823095
O═C1C═CCCC1	3.97735	CCCCCC#Cc1ncccc1C#N	5.8237
O═C1CCCC═C1	3.97735	CC(═O)CC(═O)N1CCCCCC1	5.82382
OC1CCSCC1	3.9899	O═C(CC(C)(C)C)c1ccccc1	5.828928
C1CCOCO1	4.04859	CC(C)(C)OCc1ccccc1	5.830675
O═C1CCC═CC1	4.05289	CC1(O)CCCCCC1	5.846164
N1CCC═CC1	4.09099	O═c1cc[nH]cc1	5.850777
C1C═CC═C1	4.09349	C/C═C/C1OCCO1	5.852582
O═S1(═O)CCNCC1	4.25706	CS(═O)(═O)N1CCCCCC1	5.852941
CC1CCOC(═O)C1	4.31492	Cc1ncc(C)cn1	5.872505
O═S1(═O)C═CCCC═C1	4.35802	CC(C)(C)NCC#Cc1ccccc1	5.872605
O═c1cccn[nH]1	4.38874	CCCOc1cccc(C)c1	5.874069
CC1CCNCC1	4.39789	C1NNC═C1	5.876098
CC1CCNCC1	4.39789	CC(═CC(═O)N1CCCCCC1)C	5.892267
c1ccncn1	4.44564	CC(C)CCOC1CCNCC1	5.894852
C1CCC═COC1	4.46091	CCCN1CCCCC1	5.899168
c1ccnnc1	4.50924	O═Cc1cccs1	5.902546
O═C1CCCC(═O)N1	4.53618	CC(═C)Cc1ncccc1C	5.90473
O═S1(═O)CCCCC1	4.55931	CC1(C)CCNCC1	5.91625
O═C1CNCCN1	4.60068	O═C1CCCCCCCCC(═O)OCCCCO1	5.927301
C1═NC═CN1	4.61974	O═C1CCNC(═O)N1	5.933254
c1ncc[nH]1	4.61974	OC1CCNCCC1	5.937124
O═C1CNCCCN1	4.64001	O═C1COCCN1	5.944039
CN1CCOCC1	4.6583	O═C(NC(C)(C)C)c1ccncc1	5.947731
O═C1NCCCO1	4.721	CC(O)CC#CCN1CCCCC1	5.948158
O═S1(═O)CCOCC1	4.79345	O═C1CCCC1	5.94834
O═C1CCCCC(═O)N1	4.79537	O═C1CCCC1	5.94834
CN1CCC═CC1	4.91248	C1CC═CC1═O	5.95686
O═C1C═CC═CO1	4.9195	OCCC#CCN1CCCCC1	5.957687
N1N═CC═C1	4.94377	C═CC1OCCO1	5.959976
c1ccn[nH]1	4.94377	CN1CCCCCC1	5.962768
C1CCCOC1	4.97412	Cc1cc[nH]n1	5.96375
O═C1CCCC(═O)C1	5.01506	O═C1NCCNCC1	5.967599
Cc1ccncn1	5.03254	CC(C)═CC1C═C(C)CCO1	5.971966
c1nnn[nH]1	5.0834	CC(═C═CSc1ccccc1)C	5.975644
CC1═NNCC1	5.12159	CCOCCCNC(═O)n1cncc1	5.979802
N1CCCCCC1	5.12723	OCC1CCCC1	5.980761
OCc1cnn[nH]1	5.15293	C1CNCCNC1	5.984318
CC(C)CC(═O)N1CCNCCC1	5.1552	CCOC1CCCO1	5.986821
CN1CCCCC1	5.16702	CCCCOc1cccc(N)c1	5.987676
CN1CCCCC1	5.16702	CC1═CCCN1	5.994846
NCc1nnn[nH]1	5.23459	CC(═O)CC(═O)NC1CCCCC1	5.995065
c1ccns1	5.24232	O═C1CC(C)(O)CCO1	5.996197
C1CNCCOC1	5.24713	O═C1OCCC(C)(O)C1	5.996197
C1CCNOC1	5.25085	CCCNC1CCCCCC1	5.996813
C1CCCS1	5.27768	OCC1COCC1	5.99948
c1cnn[nH]1	5.28134	CCCn1ccc(═N)cc1	6.007298
O═c1cn[nH]c(═O)[nH]1	5.29523	OC[C@H]1CNCC1	6.00856
O1CCCCCC1	5.29737	CC(═CC(═O)N1CCNCC1)C	6.009864
O═C1CCCC(═O)O1	5.29845	O═C1OCC[C@@](C)(O)C1	6.011353
O═c1ccnc[nH]1	5.29899	O═C1CC2CCC1CC2	6.017842
C1OCC═CCO1	5.33288	CC(C)CC(═O)N1CCCCCCC1	6.01785
O═S1(═O)CCCCO1	5.35017	CC1CC(C)CC(═O)C1	6.018673
C/C═C/C(═O)N1CCCCCC1	5.3502	O═Cc1ccc(OCCC(C)C)cc1	6.022296
CC(C)CC(═O)C1CCCCC1	5.38026	CCOC1═CCCCC1	6.030238
CNc1nnn[nH]1	5.38777	CC1CNCC(C)C1	6.031252
O═C1NCCCN1	5.41177	O═S1(═O)C═CC═CC═C1	6.03423
CC/C═C/N1CCCCC1	5.41882	CN(C)NC(═O)c1ccccc1	6.034587
n1ccnnc1	5.42336	NC1CCNCC1	6.039425
C═CCON═C1CCCCC1	5.43999	CC(C)c1nnn[nH]1	6.044467
O═C(NC(C)(C)C)c1ccccc1	5.44125	CCCCCCCc1ccc(C#N)cc1	6.044688
C[C@H]1CNCC1	5.45426	C1CC═CC1	6.044933
CC1CNCC1	5.46662	CN1CCCC1	6.046603
NC1═NCCO1	5.469	CC(C)(C)NSc1ccccc1	6.046834
C[C@@H]1CNCC1	5.48036	CN(C)CCCNC(═O)n1cncc1	6.05255
O═C1CCCCCO1	5.51226	N#CC1═CCCCCCCCCCC1	6.058942
CCON═C1CCCCC1	5.53578	CN(C)CCC1CCNCC1	6.062633
CC(C)NC(═O)n1cncc1	5.5454	O═C1CC═CC1	6.06268
CNC1CCCCCC1	5.56246	CCC(═O)NCCCn1cncc1	6.068374
CN1CCCN(C)C1	5.60836	CCc1cnc(N)nc1	6.07402
CC(C)═CC1CC(C)═CCO1	5.60984	CC(C)═CC1OCCC(C1)C	6.075926
CC1COCC1	5.61068	CC1CCOC(C1)C═C(C)C	6.075926
O═Cc1cc[nH]n1	5.61618	CC(C)OC(═O)CCNC(═O)n1ccnc1	6.076841
CC(═C)Cc1ccccn1	5.61981	CCCn1cccc1	6.081147
CC(C)(C)/N═C/c1ccccc1	5.62608	C1Cc2[nH]ncc2C1	6.084577
O═C1CCCS(═O)(═O)CC1	5.62707	CC(C)CCNc1ccccc1	6.088477
O═C1OCCOCC1	5.62874	CCCCCCN1CCCCCC1═O	6.088796
CCCCCCCCn1ccc(═N)cc1	6.09256	CSc1ccc(cc1)C(═O)NC(C)C	6.328161
CNc1cccc(C)c1	6.098	CCCCCCCCCn1ccc(═N)cc1	6.328322
CC(═O)C1CCCCCCCCCCC1═O	6.09886	CCCNCC1CCCCC1	6.33169
CC1═CCCO1	6.1036	S═C(N/N═C/c1cccnc1)NC(C)(C)C	6.332076
CCCCCn1ccc(═N)cc1	6.11175	N#CC(═CNc1ccncc1)C#N	6.332338
CCCCc1ccc(N)cc1	6.11758	CCCCNc1ccccc1	6.3364
N#CCCOc1cc(C)cc(C)c1	6.11793	CC(C)COC1CCNCC1	6.336875
CCNC(═O)C1CCNCC1	6.12143	C1C═CCCC═C1	6.337419
Cc1cccc(NN═C(C)C)c1	6.12331	CCOC1CCCCO1	6.338426
CN(C)CCCNC(═O)c1cccs1	6.12848	CCOC(═O)C#Cc1cccc(C#CC(═O)OCC)c1	6.33956
CCC(═O)NCCC1═CCCCC1	6.12936	COC(═O)\C(═C/C#CC1CCCCC1)/C	6.340497
CC(═C)Cc1cccc(C)n1	6.13039	CC(═C)COCC1CCCNC1	6.341306
C═CCc1ccccn1	6.14194	Cc1cccc(CC═C)n1	6.342212
OC1CCCCC1	6.14367	O═C(NCCCn1cccn1)C1CC1	6.342747
OC1CCCCC1	6.14367	CCCC(═O)NC1CCN(C)CC1	6.345418
CCN(CC)C(═O)Sc1ccccc1	6.15327	OC1CCCCN1	6.345647
CCCC(═O)C1CCNCC1	6.15506	CCN(CC)CCCNC(═O)c1cccs1	6.347501
CC1CCCS1	6.15615	CC(═O)NCCCNC(═O)c1cccnc1	6.348592
C1CCN═CN1	6.1566	CCC1NCC═C1	6.348743
CCCCc1ccc(C#N)cc1	6.15777	CCC1CCCCO1	6.35009
CC(C)CCNC(═O)c1cccs1	6.16447	CN(C)/C═N/c1ccccc1	6.350651
CS(═O)(═O)N1CCNCCC1	6.16796	CSCCC(═O)Nc1nccs1	6.353907
CC1CCCN(C)C1	6.16806	CN1CCNCCC1	6.355103
O═C1CCCCC(═O)O1	6.16947	O═c1cc[nH]c(═O)[nH]1	6.3581
NN1CCCCC1	6.16961	CC(N)CCn1cccn1	6.359583
CCOC(═O)C1CCNCC1	6.16969	CCOC(═O)N1CCCCC1	6.360332
OCC#CCCOCc1ccc(OC)cc1	6.17176	O═C(Nc1ccncc1)CC(C)(C)C	6.360498
NCC1COCC1	6.17315	Cc1scc(c1)C(═O)Nc1ccncc1	6.361459
O═C(Nc1cccnc1)NC(C)(C)C	6.17625	CC(═CCOC1CCCNC1)C	6.366281
CSc1ccccc1C(═O)NCCCN1CCCCCC1	6.17718	CCCC(═O)NC1CCCCCC1	6.370017
CC(C)C1NCC═C1	6.18084	CC(C)CNCc1ccc(C)cc1	6.372029
OC(═O)C1═CCCCCCCCCC1	6.18133	CCCCOc1ccccc1C	6.373041
O═C(Nc1ccccc1)CC(C)(C)C	6.18333	CCOC1OCCCO1	6.373474
CC(═O)CC(═O)NCCCc1ccccc1	6.1847	CCC(CC)C(═O)NCCCN1CCOCC1	6.374652
CCCC(═O)N1CCSCC1	6.1868	CC1═NCCC1	6.380214
CCc1ccc[nH]1	6.18755	CCC(═O)C1CCCCC1	6.380337
O═C(NCCCn1cccn1)C1CCCCC1	6.1899	COC1═NCCCCC1	6.382579
Cc1cc(CC═C)ncc1	6.19214	COC(═O)NCCC1═CCCCC1	6.382998
O═C1CCCCC(═O)C1	6.19215	CNC(═O)NCc1ccccn1	6.383801
O═C1NC═NC(═O)C1	6.19497	CC(C)NC(═O)c1cscc1	6.38503
CC(═O)CCC═C1CCCCC1	6.1999	CNCCc1c[nH]cn1	6.389962
C═CC(C)CSc1ccccc1	6.20114	CCNC(═O)NC1CCCCC1	6.391185
CC(N)c1nnn[nH]1	6.20863	O═C1CC(═CC(═O)O1)C	6.391193
CC(C)CCNC(═O)n1cncc1	6.21052	C═CCSC(═O)NC(═O)Cc1ccccc1	6.393044
CCCC1CCCCNC1═O	6.21079	CCCC(═O)NCCCc1ccccc1	6.39395
CC(═O)NCCCNC(═O)c1cccs1	6.21297	O═S1OCCCCO1	6.393961
CC(C)CC(═O)NCCc1ccncc1	6.21313	NCCC(═O)N1CCNCC1	6.394319
CC(═O)CC(═O)NC1CCCCCC1	6.21501	CCCCN(C)Cc1ccc(OC)cc1	6.394897
CCc1nnn[nH]1	6.21826	Cn1cncc1C═O	6.395174
S1SCCC1	6.22029	CCCc1ccc(CN)cc1	6.397567
C═CC#CCN1CCCC1	6.2212	CC(═CC(═O)CCN1CCCCC1)C	6.398219
C1NCCS1	6.22176	CCOC(═O)CC1CCCNC1	6.399213
CCN(CC)C(═O)N1CCCCC1	6.22256	CC(C)(C)NNc1ccccc1	6.40262
ON1CCCCC1	6.22438	CCNC(═O)CCSc1nc(C)cc(C)n1	6.403173
CCCCCc1ccc(NC)cc1	6.23282	CC(C)NC(═S)N/N═C/c1cccnc1	6.404253
CNCC(═O)N1CCCCC1	6.23427	CC(C)CC(═O)NC1CCCCCC1	6.405964
O═C(NCC#Cc1ccccc1)NCCc1cccs1	6.23686	OC1(C)C═CCCC1	6.407874
Cc1c[nH]cc1	6.24026	CC1CCCO1	6.410532
CC(═O)NCCCNC(═O)c1ccncc1	6.24118	CCCCCCn1ccc(═CC═C(C#N)C#N)cc1	6.4135
CC(C)CC1CCCCCN1	6.24936	CCCCC1═CCCOC1	6.416425
CCN(CC)C(═S)NCCc1ccccc1	6.24969	CC(C)CCNC(═O)c1cccnc1	6.420616
CC(═O)CCCN1CCCCC1	6.25029	CC(C)/N═C(/C)\c1ccccc1	6.421906
O═C1C═CCCC═C1	6.25035	Oc1ccncc1	6.427433
CC(═CCOC1CCNCC1)C	6.25104	CCS/C═C/C#CC1(O)CCCCC1	6.428407
CCc1cccc(C)n1	6.25801	CCC(CC)C1CCCCCN1	6.428797
N#Cc1ccc(CSC2═NCCCN2)cc1	6.25841	CC1CCCC(O)C1	6.429721
O═C(O/N═C/c1ccccc1)N1CCOCC1	6.26173	Cc1c[nH]cn1	6.431754
CCCC(C)NC(═O)n1cncc1	6.26938	CCc1ccccc1OCC1CCCNC1	6.432372
N#Cc1ccc(\C═N/c2ccccc2)cc1	6.27228	CNC1CCCCC1	6.433502
CCCCCC/C═C1\CCCCC/1═O	6.27628	CCCCn1cccc1	6.434268
CNCCCOc1ccccc1C	6.28193	O═C(CCCc1ccccc1)NC(C)(C)C	6.434314
Oc1cccc(O)n1	6.2826	CC(C)CCCc1ccccn1	6.435371
OC[C@@H]1CNCC1	6.28635	CC1(C)COCCN1	6.435727
OCC1CNCC1	6.28726	S═C(NCCC1═CCCCC1)N(C)C	6.43606
O═C(NCCCc1ccccc1)C(C)(C)C	6.2907	N#CCCNc1ccccc1C	6.436422
CCc1nnc(N)nn1	6.29176	CN(C)CCCNC(═O)N1CCOCC1	6.437192
N#CC1(N)CCCC1	6.29315	CC(═C)Cc1nccc(C)c1	6.437725
CC(C)CCSc1ccc(C)cc1	6.29428	NCCOC1CCCCO1	6.437952
Cc1ccccc1OCCCCCN1CCCCC1	6.2943	CSc1ccccc1C(═O)NCCCN1CCCC1	6.43854
CC(C)CSc1ccccc1	6.2943	NCCCN1CCCCC1	6.44011
C/C═C/CCc1ccccn1	6.29577	C═COc1cccc(C)c1	6.442179
NC1CCCCCC1	6.29817	Nc1ccncc1	6.4422
O═C(NCC#Cc1ccccc1)c1cccs1	6.29973	C[C@H]1CCC[C@@H](O)C1	6.442564
CCCCCCCc1ccc(N)cc1	6.30274	CC(C)CCOc1ccc(N)cc1	6.445304
O═C(NCCCn1cncc1)C1CCNCC1	6.30333	NCCCN1CCSCC1	6.446523
Cc1ccc(C)cn1	6.30334	CN(C)c1ccccc1	6.448354
CCCCCCc1ccc(N)cc1	6.30632	CCCc1cccc(N)n1	6.448834
CCCCOCCCNC(═O)N1CCOCC1	6.30824	CCc1ccc(C═C)nc1	6.449475
CC(C)(C)COc1cncnn1	6.30882	CC(C)CNCc1cscc1	6.450718
COc1ccc(cc1)CN1CCCCCC1	6.31102	CC(C)CCCc1ccncc1	6.451931
COc1cccc(C)n1	6.31161	CC(C)NC(═O)N1CCNCC1	6.453215
C1C═CCS1	6.31378	Cc1cc(C)nc(Sc2ccc(cc2)C(═O)O)n1	6.454888
O═C(NCCCN1CCCC1)C1CCNCC1	6.31572	CC(C)(C)c1ccccn1	6.456365
Cc1ncccc1OCC1CCCNC1	6.31862	CCN(CC)C(═O)NC1CCCCC1	6.457056
CCCC(═O)NCc1ccccn1	6.32015	CCC1═NCCCN1	6.457151
CCCC1OCCS1	6.32076	CN(C)Cc1ccccn1	6.457247
CCOC(═O)\C═C\Cc1ccccc1	6.32096	CCCCSC1(C)CCCCC1	6.459166
CC(C)(C)COCC1CCCNC1	6.32318	COc1ccc(CNC(═O)CC(C)C)cc1	6.459845
NCCCOc1cccc(C)c1	6.32451	N#Cc1ccc(C#Cc2ccccc2)cc1	6.463152
O═C(NCCc1ccccc1)OC(C)(C)C	6.3266	CC(C)CCNC(═O)Cc1cccs1	6.465874
CC1(C)CCCNC1	6.46693	CCNCc1ccccc1	6.513302
Cc1n[nH]cn1	6.46712	CCCC1NCC═C1	6.514383
CCCC1═NCCO1	6.46827	CCCCCCn1ccc(═N)cc1	6.515571
COC1CNCC1	6.47129	CNC(C)Cc1ccccn1	6.515612
OC1(C)CCCC1	6.47168	CNCC(═O)N1CCCCCC1	6.51581
CC1(O)CCCC1	6.47168	O═C(NCCCN1CCCCC1)c1ccccc1	6.516288
O═C(NCCCN1CCSCC1)c1ccccc1	6.47174	CN(C)C1CCCC1	6.517077
CCC(CC)C(═O)NCCCc1ccccc1	6.47308	C1CCC(CN1)Oc1ncccn1	6.52029
CC(C)(C)NCCCc1ccccc1	6.47342	NCc1csnc1	6.521403
CCCCN1CCC(C)CC1	6.47395	CCC(═O)NCC#Cc1ccccc1	6.521463
O═C(NCCc1ccccc1)C(C)C	6.47507	CCCC1CCCS1	6.52177
O═C(NCCCn1cccn1)C1CCCC1	6.47677	OC1CCOCC1	6.522021
CCc1ccccc1Oc1ncccn1	6.4771	O═S(═O)(NCCCN1CCCCCC1)c1ccccc1	6.523599
CC(═O)CC(═O)NCCc1ccccc1	6.47888	CCN(CC)C(═O)NCc1ccccn1	6.523717
CCCCc1ccc(C═O)cc1	6.48035	O═C1C/C═C\CCCCCCCCO1	6.523937
CCCCNC(═O)N1CCOCC1	6.48123	O═C(OCCSC(═S)NC(C)(C)C)Nc1ccccc1	6.524631
N#CCCNC(═O)C1CCNCC1	6.48516	CC(C)CNC(═O)c1cscc1	6.52518
NCCCOc1ccccc1C	6.4853	O═C(NCCC1═CCCCC1)C1CC1	6.525225
CCCCNC(═O)Nc1nccs1	6.48555	CC(N)CCN1CCCCCC1	6.526511
O═S(═O)(NCCCN1CCCCC1)c1ccccc1	6.48645	CCNC(═O)N(C)Cc1ccccc1	6.528375
CCc1ccc(CNCC(C)C)cc1	6.48768	CN(C)C1CCCCC1	6.529209
CC(C)(C)NCc1cccnc1	6.49017	CC(C)CNCc1ccccc1C	6.53036
CCC(C)NC(═O)c1cscc1	6.49085	CC(C)(C)NCc1ccncc1	6.5317
CCCc1ccc(N)cc1	6.49096	CCCCNCc1ccc(CC)cc1	6.533015
CCCCN1CCCCC1	6.49113	CCCCCn1cccc1	6.533323
CC(C)OC1CCCCO1	6.49161	O═C(OC(C)(C)C)n1cncc1	6.533548
CN(C)C(═O)NC1CCCCC1	6.49217	CCNC(═S)Nc1ccc(CC)cc1	6.53399
O═C(NCC#Cc1ccccc1)c1cscc1	6.49234	OC[C@@H]1CCCN1	6.534219
CN1CCN(CC1)C(═O)CC(C)C	6.49243	CC(═C)CN1CCCCCC1═O	6.53494
CC(C)COc1cccc(N)c1	6.49256	O═C(NCCCc1ccccc1)C(C)C	6.534992
O═C(NCCCn1cncc1)n1cncc1	6.49626	OCC1CCCN1	6.535179
CC(C)C1OCC═CCO1	6.49662	CC(C)(C)N═C═NC(C)(C)C	6.535355
CCC(═O)N1CCNCC1	6.4982	CN1CCCN(CC1)C(═O)CSc1ccccc1	6.535507
CC(C)NC(═O)C1CCNCC1	6.49827	COc1ccc(cc1)C(═O)NC(C)(C)C	6.536675
CCCCN1CCC(N)CC1	6.49841	CCCS(═O)c1ccccc1	6.539592
O═C1NCCN(CC1)C(═O)OC(C)(C)C	6.49861	CC(O)CC#CCN1CCOCC1	6.539593
CC(C)(C)NCCCOc1ccccc1	6.49939	CC(═CCCC(═CCCC1═COC═C1)C)C	6.54101
O═C(Nc1cccnc1)N1CCCCC1	6.50169	N#CCCCCCC1CCCC(═O)C1	6.541903
NN1CCOCC1	6.50235	Cc1ccsc1CNCc1ccncc1	6.542274
CC(C)(C)N═C═NCc1ccccc1	6.5031	CC(═CCCC#CCCOC1CCCCO1)C	6.546339
CCCC1OCCCO1	6.5042	CCCCn1cncc1	6.547629
O═S1(═O)CCOCCS(═O)(═O)NCCOCCN1	6.50421	OC1CCCC1	6.550298
CNC1CCCCNC1	6.50462	CCC(═O)Nc1nccs1	6.552381
CC(C)C1CCCCCN1	6.50482	O═CNc1nnn[nH]1	6.552698
O═Cc1cnc(nc1)NC(C)C	6.50527	CC1(C)CCOCO1	6.554363
CN1CCCN(CC1)C(═O)CC(C)C	6.50672	CCNC(═S)NCCN1CCCCC1	6.555586
Cc1ccc(OCC2CNCC2)cc1	6.50714	CCOc1cccc(N)c1	6.555999
COC(═O)NCCCn1cncc1	6.50724	CN(C)CCCNCc1ccncc1	6.556703
Cc1ccsc1CNCc1cccnc1	6.51228	CC(═O)NCC1CCNCC1	6.556707
C1CCSCS1	6.55782	O═C(NCCCN1CCCC1═O)Cc1cccs1	6.556977

A number of embodiments of the disclosure have been described. Nevertheless, it will be understood that various modifications may be made without departing from the spirit and scope of the disclosure. Accordingly, other embodiments are within the scope of the following claims.

Claims

1-41. (canceled)

42. A method of attracting one or more insect species comprising the use of a composition comprising 2-ethylpyrazine.

43. The method of claim 42, wherein said method is effected by activating odor receptors or odor receptor neurons.

44. The method of claim 42, wherein 2-ethylpyrazine is present in said composition at a concentration of from about 10% to about 99% by weight.

45. The method of claim 44, wherein 2-ethylpyrazine is present in said composition at a concentration of from about 50% to about 99% by weight.

46. The method of claim 42, wherein said composition further comprises a carrier.

47. The method of claim 42, wherein said composition further comprises a suitable solvent.

48. The method of claim 47, wherein said solvent is an organic solvent.

49. The method of claim 42, wherein said composition is present in a wicked apparatus.

50. The method of claim 43, wherein said odor receptors comprise one or more members of the Gustatory receptor (Gr) family.

51. The method of claim 50, wherein said one or more members of the Gr family comprise CO2 receptors.

52. The method of claim 51, wherein said CO2 receptors comprise Gr21a, Gr63a, AgGr22, AgGr23 and/or AgGr24 proteins.

53. The method of claim 42, wherein said insect species is selected from the group consisting of mosquitoes, sandflies and Testse flies.

54. The method of claim 53, wherein said insect species comprises mosquitoes selected from the group consisting of Anopheles gambiae, Culex pipiens and Aedes aegypti mosquitoes.

55. A method of luring one or more insect species to a trap comprising the use of a composition comprising 2-ethylpyrazine.

56. The method of claim 55, wherein said method of luring is carried out by virtue of activating odor receptors or odor receptor neurons in said one or more insect species

57. The method of claim 55, wherein 2-ethylpyrazine is present in said composition at a concentration of from about 50% to about 99% by weight.

58. The method of claim 55, wherein said composition further comprises a carrier.

59. The method of claim 55, wherein said composition further comprises a suitable solvent.

60. The method of claim 59, wherein said solvent is an organic solvent.

61. The method of claim 55, wherein said composition is present in a wicked apparatus.

62. The method of claim 56, wherein said odor receptors comprise one or more members of the Gustatory receptor (Gr) family.

63. The method of claim 62, wherein said one or more members of the Gr family comprise CO2 receptors.

64. The method of claim 63, wherein said CO2 receptors comprise Gr21a, Gr63a, AgGr22, AgGr23 and/or AgGr24 proteins.

65. The method of claim 55, wherein said insect species is selected from the group consisting of mosquitoes, sandflies and Testse flies.

66. The method of claim 65, wherein said insect species comprises mosquitoes selected from the group consisting of Anopheles gambiae, Culex pipiens and Aedes aegypti mosquitoes.

67. The method of claim 55, wherein said trap is selected from the group consisting of a tape and a mechanical trap.

68. The method of claim 67, wherein said mechanical trap is selected from the group consisting of a suction-based trap, a light-based trap and an electric current-based trap.

69. A method of inhibiting, preventing or reducing the incidence of insect-borne disease in a subject, said method comprising luring one or more insect species to a trap comprising the use of a composition comprising 2-ethylpyrazine, thereby reducing or eliminating said one or more insect species and thereby inhibiting, preventing or reducing the incidence of insect-born disease in said subject.

70. The method of claim 69, wherein said luring is carried out by virtue of activating odor receptors or odor receptor neurons in said one or more insect species.

71. The method of claim 69, wherein 2-ethylpyrazine is present in said composition at a concentration of from about 50% to about 99% by weight.

72. The method of claim 69, wherein said composition further comprises a carrier.

73. The method of claim 69, wherein said composition further comprises a suitable solvent.

74. The method of claim 73, wherein said solvent is an organic solvent.

75. The method of claim 69, wherein said composition is present in a wicked apparatus.

76. The method of claim 69, wherein said insect-borne disease is selected from the group consisting of malaria, dengue, yellow fever, river blindness, lymphatic filariasis, sleeping sickness, leishmaniasis, epidemic polyarthritis, West Nile virus disease and Australian encephalitis.