Patent Application
20180127804
The methods and compositions provided herein generally relate to the fields of molecular biology and genetic engineering.
Synthetic biologists routinely assemble well-characterized DNA parts into larger constructs and introduce those DNA assemblies into host organisms to achieve desired phenotypes. See Weenink and Ellis (2013) Methods Mol. Biol. 1073: 51-60; Polizzi (2013) Methods Mol. Biol. 1073: 3-6; Munnelly (2013) ACS Synth Biol. 2: 213-215; Stephanopoulos (2012) ACS Synth. Biol. 1: 514-525. This is often a trial-and-error process that requires building and testing tens to thousands of DNA assemblies. For example, a comprehensive combinatorial exploration of five genes each expressed at five levels would require 3125 DNA assemblies. At synthetic biology companies, it is common to build many constructs to test diverse hypotheses or to optimize a multi-gene pathway using iterative design-build-test-learn cycles similar to strategies described previously. See Gardner et al. (U.S. Pat. Nos. 8,859,261; 8,415,136); Du et al. (2014) ACS Chem. Biol. 9: 2748-2754; Ajikumar et al. (2010) Science 330: 70-74. At this scale, quality control (QC) of large numbers of DNA assemblies creates logistical and economic challenges.
High-throughput strain engineering facilities routinely use automated workflows to assemble thousands of DNA constructs ranging in size from 3-30 kb and containing 2-12 DNA parts. The DNA assemblies must hence undergo rigorous QC to avoid building and testing incorrectly engineered strains, which could lead to erroneous conclusions regarding genotype-phenotype relationships. Because no assembly method is perfect, finding a correct assembly requires QC analysis to be performed on multiple clones. Until recently, this involved comparing the observed restriction endonuclease fragment sizes to those computationally predicted for four colonies, followed by Sanger sequencing of the chosen clone. To achieve 2× coverage across a 10 kb assembly using Sanger sequencing requires at least 24 reads spaced appropriately across the assembly and costs at least $72 at present day value. This is too expensive and logistically onerous for a high throughput operation.
Next-generation sequencing (NGS) technology has greatly reduced the cost of sequencing whole genomes, but its application for the simultaneous sequencing of multiple plasmid constructs or other smaller size DNA constructs has been limited. Thus, there remains a need for high-throughput, low-cost sequencing methods for less than genome-scale applications.
Provided herein are methods, compositions, and kits for preparing and simultaneously sequencing a plurality of polynucleotides (e.g., plasmids comprising DNA assemblies) in a single sequencing run of a sequencing instrument. In certain embodiments, a next-generation sequencing platform is combined with an acoustic liquid handling instrument to provide a rigorous, low-cost QC method that enables complete sequencing of almost every DNA assembly built by a high throughput operation. Embodiments of the present invention increase the efficiency of sequencing operations by simplifying workflow and reducing cost and hands-on time to perform experiments, as compared to known sequencing methods. The Illumina MiSeq sequencer can provide about 5 gigabases (GB) of data in a 24 hour run using the 300-cycle v2 kit (Perkins et al. (2013) PLoS One 8: e67539; Loman et al. (2012) Nat. Biotechnol. 30: 434-439), theoretically allowing 25,000 plasmids of 10 kb average size to be sequenced. However, there were several obstacles to overcome before even a fraction of this high level of multiplexing can be achieved.
The Illumina Nextera method for preparing sequencing libraries is convenient and robust (Caruccio (2011) Methods Mol. Biol. 733: 241-255). However, cost-effective sequencing of plasmids in the 3 to 30 kb range requires hundreds of barcode primers and a significant reduction in the use of the expensive Nextera reagents. A recent report described a Nextera workflow in which reaction volumes were reduced eight-fold relative to the Illumina protocol (Lamble (2013) BMC Biotechnol. 13: 104). Here, in addition to showing that the volume of the tagmentation reaction can be reduced 100-fold using acoustic droplet ejection, it has been demonstrated that thousands of uniquely barcoded samples can be handled with the appropriate automation infrastructure. It has also been demonstrated that over 4000 plasmids with an average size of 8 kb (largest about 20 kb) can be simultaneously sequenced at a consumables cost of less than $3 per plasmid. Furthermore, embodiments of the present invention include systems and software to track the samples and associated sequence data and to rapidly identify correctly assembled constructs having the fewest defects. This NGS quality control (QC) process should be of value to any group operating a high-throughput molecular biology pipeline.
Thus, in one aspect, provided herein is a method of preparing a plurality of polynucleotides for simultaneous sequencing. The method comprises, for each input polynucleotide of a plurality of input polynucleotides, (a) amplifying the input polynucleotide by rolling circle amplification (RCA) in an RCA solution to generate a target polynucleotide; (b) diluting the RCA solution comprising the target polynucleotide by a standard dilution factor; (c) generating a reaction mixture having a volume of about 0.005 μL to about 2 μL and comprising tagged polynucleotide fragments by contacting the diluted RCA solution comprising the target polynucleotide with transposases pre-loaded with transposon end sequences to fragment and tag the target polynucleotide; (d) removing the transposases from the tagged polynucleotide fragments, thereby generating a reaction solution; and (e) performing a polymerase chain reaction (PCR) with the reaction solution comprising the tagged polynucleotide fragments, wherein the PCR utilizes adapter primers comprising barcode sequences that are capable of hybridizing to the tagged polynucleotide fragments to generate barcoded polynucleotide fragments.
In one embodiment, the method further comprises: (f) combining the barcoded polynucleotide fragments generated for each input polynucleotide of the plurality of input polynucleotides; (g) sequencing the combined barcoded polynucleotide fragments in step (f) in a single sequencing run to generate sequence reads; (h) sorting the sequence reads from the sequencing run using the barcode sequences associated with each input polynucleotide; and (i) aligning and assembling the sequence reads for each input polynucleotide to generate a consensus sequence of the input polynucleotide.
In another embodiment, the barcode sequences are selected from the group consisting of SEQ ID NO: 1 through SEQ ID NO: 192.
In another embodiment, the plurality of input polynucleotides is at least 1000, at least 2000, at least 3000, or at least 4000.
In another embodiment, the input polynucleotide is a plasmid DNA.
In another embodiment, the input polynucleotide comprises a DNA assembly of a plurality of DNA components.
In another embodiment, the input polynucleotide is a plasmid and the combined barcoded polynucleotide fragments are generated from at least 1000 plasmids.
In another embodiment, the input polynucleotide is a plasmid and the combined barcoded polynucleotide fragments are generated from at least 4000 plasmids.
In another embodiment, less than 2 percent of the plasmids had less than 15 times average sequencing coverage.
In another embodiment, the reaction mixture has a volume of about 0.5 μL. In another embodiment, the reaction mixture has a volume of less than about 1 μL. In another embodiment, the reaction mixture has a volume of less than about 2 μL.
In another embodiment, the standard dilution factor is determined by: (a) measuring a concentration of the target polynucleotide in the RCA solution for at least a portion of the plurality of input polynucleotides; (b) determining an average concentration of the target polynucleotides in the RCA solution for the at least the portion of the plurality of input polynucleotides; and (c) calculating the standard dilution factor by dividing the average concentration by 5 ng/μL.
In another embodiment, the diluted RCA solution comprises the target polynucleotide at a concentration between about 3 ng/μL and about 10 ng/μL.
In another embodiment, the transposases are removed from the tagged polynucleotide fragments by treating the reaction mixture from step (c) under a dissociation condition.
In another embodiment, the treating the reaction mixture from step (c) under the dissociation condition comprises adding a dissociation solution to the reaction mixture.
In another embodiment, the dissociation solution comprises sodium dodecyl sulfate (SDS). In another embodiment, a concentration of the SDS in the reaction solution is between about 0.05% to about 0.3%.
In another embodiment, the dissociation solution comprises sodium dodecyl sulfate (SDS) and a concentration of the SDS in the reaction solution is about 0.1%.
In another embodiment, the method further comprises diluting the reaction solution by at least 10-fold with an aqueous solution prior performing the PCR.
In another embodiment, the transposases are removed from the tagged polynucleotide fragments without using solid phase extraction or centrifugation.
In another embodiment, the method further comprises, after the PCR, (f) removing small polynucleotide fragments from PCR products; (g) quantifying a concentration of the barcoded polynucleotide fragments from step (f) for each input polynucleotide; and (h) determining a volume of the barcoded polynucleotide fragments in step (f) to add to a pool assuming an average polynucleotide fragment size of 500 base pairs and normalizing for a length of the input polynucleotide.
In another embodiment, the method further comprises filtering the combined barcoded polynucleotide fragments to remove small fragments having a size less than about 300 base pairs.
In another aspect, provided herein is a method of preparing a plurality of polynucleotides for sequencing, the method comprising: (a) generating a reaction mixture having a volume of about 0.005 μL to about 2 μL and comprising tagged polynucleotide fragments by contacting a target polynucleotide with transposases pre-loaded with transposon end sequences to fragment and tag the target polynucleotide; and (b) performing a polymerase chain reaction (PCR) with a reaction solution comprising the reaction mixture comprising the tagged polynucleotide fragments and adapter primers comprising barcode sequences capable of hybridizing to the tagged polynucleotide fragments to generate barcoded polynucleotide fragments.
In one embodiment, the method further comprises: (c) repeating steps (a) and (b) described above to generate barcoded polynucleotide fragments from a plurality of target polynucleotides, wherein the barcoded polynucleotide fragments from each of the plurality of target polynucleotides comprise a unique barcode sequence; (d) combining the barcoded polynucleotide fragments generated from the plurality of target polynucleotides; and (e) sequencing the combined barcoded polynucleotide fragments in a single sequencing run to generate sequence reads.
In another aspect, provided herein is a method of preparing a plurality of polynucleotides for sequencing, the method comprising: for each input polynucleotide of a plurality of input polynucleotides, (a) amplifying the input polynucleotide by rolling circle amplification (RCA) in an RCA solution to generate a target polynucleotide; (b) diluting the RCA solution comprising the target polynucleotide by a standard dilution factor; (c) generating a reaction mixture having a volume of about 0.005 μL to about 2 μL and comprising tagged polynucleotide fragments by contacting the diluted RCA solution comprising the target polynucleotide with transposases pre-loaded with transposon end sequences to fragment and tag the target polynucleotide; (d) adding a dissociation solution to the reaction mixture to remove the transposases from the tagged polynucleotide fragments, thereby generating a reaction solution; (e) diluting the reaction solution with an aqueous solution; (f) adding to the diluted reaction solution a pair of adapter primers comprising barcode sequences capable of hybridizing to the tagged polynucleotide fragments; (g) performing a polymerase chain reaction (PCR) with the diluted reaction solution and terminal primers to generate barcoded polynucleotide fragments, wherein the terminal primers are capable of hybridizing to the barcoded polynucleotide fragments; (h) combining the barcoded polynucleotide fragments generated in step (g) for each input polynucleotide of the plurality of input polynucleotides; (i) sequencing the combined barcoded polynucleotide fragments of step (h) in a single sequencing run to generate sequence reads; (j) sorting the sequence reads from the sequencing using the barcode sequences associated with each input polynucleotide to assign each of the sequence reads to each input polynucleotide; and (k) aligning and assembling the sorted sequence reads for each of the input polynucleotide to generate a consensus sequence of each input polynucleotide.
In certain embodiments, the reaction mixture is generated using an acoustic liquid handling instrument.
In another aspect, provided herein is a kit comprising: (a) a plurality of barcoded adapter primers produced by the method described herein; and (b) reagents to perform polymerase chain reaction. In certain embodiments, the kit comprises at least 10, at least 20, at least 30, at least 40, at least 50, at least 60, at least 70, at least 80, at least 90, at least 100, at least 110, at least 120, at least 130, at least 140, at least 150, at least 160, at least 170, at least 180, or at least 190 different adapter primers.
In an embodiment, the barcode sequences may be selected from the group consisting of SEQ ID NO: 1 to SEQ ID NO: 192.
In another embodiment, the barcoded polynucleotide fragments comprise combined barcoded polynucleotide fragments generated from a plurality of target polynucleotides, and wherein the barcoded polynucleotide fragments from each of the plurality of target polynucleotides comprise a first barcode sequence selected from the group consisting of SEQ ID NO: 1-96 and a second barcode sequence selected from the group consisting of SEQ ID NO: 97-192.
In another aspect, provided herein is a composition comprising a library of barcoded polynucleotide fragments comprising a barcode sequence produced by the method described herein. In an embodiment, the barcode sequences may be selected from the group consisting of SEQ ID NO: 1 to SEQ ID NO: 192. In certain embodiments, the plurality of target polynucleotides are generated from at least 1000, at least 2000, at least 3000, or at least 4000 samples of plasmid DNA.
FIGS. 7B1 through 7B3 illustrate superimposed fragment analyzer traces of samples treated with the Zymo kit, with 0.2% SDS final concentration, or with 0.1% SDS final concentration. All samples were incubated at room temperature. DNA fragment size is shown along the horizontal axis and DNA concentration is shown along the vertical axis (RFU=Relative Fluorescence Units). Zymo-treated samples have the majority of fragments (by moles) below 600 base pairs. SDS-treated samples have the majority of fragments (by moles) above 600 base pairs.
The rapid growth in the field of synthetic biology over the last decade has been driven in large part by advances in the synthesis and sequencing of DNA sequences. A decade ago, synthesizing DNA, such as simple oligonucleotides, was tedious and could cost hundreds of dollars, but today these DNA parts are ordered automatically and delivered next-day for tens of dollars. The DNA sequencing technology has also progressed, particularly through the extensive automation and scaling of Sanger sequencing technology. However, the progress in DNA sequencing technology has lagged behind DNA synthesis technology and has become cost-limiting for many researchers in this field.
Recent commercialization of so-called next-generation sequencing technologies promise to overcome this lag and dramatically increase the amount of DNA read per dollar. Next-generation sequencing technologies include instruments capable of parallelizing the sequencing process, producing thousands or millions of sequence reads concurrently per instrument run. For genome-size DNA templates, this promise of increasing the amount of DNA read per dollar has been fulfilled by commercially available kits. For smaller size DNA samples, such as plasmid DNA, no workflow has yet been developed that can reap the cost benefits of next-generation sequencing.
The methods, compositions, and kits provided herein improve the efficiency of next-generation sequencing process for samples with input polynucleotides having a small size (e.g., 3-30 kb range) by increasing sample throughput, simplifying workflow, and decreasing the cost. The compositions and methods described herein bridges the power of next-generation sequencing to the plasmid libraries and other smaller size DNAs used in gene synthesis, DNA assembly, enzyme engineering, amplicon sequencing, library deconvolution, and the like. Here, the efficiency of sequencing workflow has improved dramatically, in part, due to reducing sample reaction volumes and reducing the amount of key reagents for each reaction. As a result, the cost of sample preparation is significantly reduced. Furthermore, by increasing the number of samples combined into a single sequencing run, the throughput of sample processing is significantly increased. In particular, there are three main aspects of the present invention that contribute to low-cost, high-throughput processing of thousands of samples.
In one aspect, methods and compositions described herein can provide at least 100-fold reduction in reaction volume for a standard DNA tagmentation reaction. By using an acoustic liquid transfer system, a reaction usually performed at a volume of 50 μL can be reduced down to a volume of 2 μL or less, or even to a volume of about 0.5 μL. The second and third aspects of the invention have been developed to further accommodate this small reaction volume.
In another aspect, the methods and compositions described herein provide concomitant reduction in volume of both target polynucleotide derived from a sample and tagmentation enzyme to reduce overall cost of the reaction. The decreased polynucleotide concentration can be compensated for by increasing the number of cycles in the subsequent PCR step. Although a shift in the size distribution of DNA fragments is observed with increasing PCR cycles, no significant change in sequence quality was observed due to the reduction in a reaction volume during tagmentation.
In another aspect, the methods and compositions described herein provide novel barcode sequences, which increase the number of samples that can be combined together into a single sequencing run. These barcode sequences also decrease the sequencing cost and provide higher throughput, as fewer sequencing runs are required to sequence a large number of samples.
By utilizing the above described and other features of methods and compositions described herein, a workflow has been developed so that a high-quality sequence coverage can be provided for thousands of samples per week. Such high quality sequence coverage can be provided at a reasonable cost, for example, less than $3 per plasmid at present day value. This cost represents more than a 25-fold reduction over the alternative Sanger sequencing technology. The compositions and methods provided herein provide many advantages in the field of synthetic biology as well as other technical areas. These and other aspects of the present invention are described more fully throughout the specification below.
As used herein, the term “transposon” refers to a nucleic acid segment, which is recognized by a transposase and which is a component of a functional nucleic acid-protein complex (i.e., a transposome or transposition complex) capable of transposition.
As used herein, the term “transposase” or “fragmentation and labeling enzyme” refers to an enzyme, which is a component of a functional nucleic acid-protein complex capable of transposition and which is mediating transposition.
As used herein, the term “transposon end” or “transposon end sequence” refers to a double stranded DNA that exhibits nucleotide sequences that are necessary to form the complex with the transposase enzyme that is functional in an in vitro transposition reaction. The transposon end sequences are responsible for identifying the transposon for transposition. A transposon end forms a transposome or transposition complex with a transposase to perform transposition reaction. In certain embodiments, the transposon end sequence may further include additional sequences such as primer binding sites or other functional sequences.
As used herein, the term “transposome” or “transposition complexes” refers to the formation between a transposase enzyme and a fragment of double stranded DNA that contains a specific binding sequence of the enzyme, termed “transposon end.” The complex formed between a transposase enzyme and transposon end capable of mediating transposition and fragmentation of a target polynucleotide is also referred to as transposases “pre-loaded” with transposon end sequences.
As used herein, the term “rolling circle amplification” refers to nucleic acid amplification reactions where a circular nucleic acid template is replicated in a single long strand with tandem repeats of the sequence of the circular template. This first, directly produced tandem repeat strand is referred to as tandem sequence DNA and its production is referred to as rolling circle replication. Rolling circle amplification refers to both to rolling circle replication and to processes involving both rolling circle replication and additional forms of amplification.
As used herein, the term “amplification” refers to a method or process that increases the representation of a population of specific nucleotide sequences in a sample.
As used herein, the term “standard dilution factor” refers to a number that is used to uniformly dilute all solutions comprising target polynucleotides to be simultaneously sequenced. For example, all solutions comprising target polynucleotides may be diluted by a “standard dilution factor” of 1:5 by adding 20 μL of water to 5 μL of each of the solutions, regardless of the concentration of DNA in each solution.
The terms “nucleic acid” or “polynucleotide” refers to a polymeric form of nucleotides of any length, either ribonucleotides or deoxynucleotides. Thus, this term includes, but is not limited to, single-, double-, or multi-stranded DNA or RNA, genomic DNA, cDNA, DNA-RNA hybrids, or a polymer comprising purine and pyrimidine bases or other natural, chemically, or biochemically modified, non-natural, or derivatized nucleotide bases.
As used herein, the term “input polynucleotide” can refer to a nucleic acid molecule from a sample of interest and/or a known nucleic acid sequence, and it may be a source material for generating a target polynucleotide.
As used herein, the terms “target polynucleotide” or “target DNA” may be used to refer to nucleic acid molecules that are derived from an input polynucleotide. The target polynucleotide or target DNA may be subject to fragmentation and/or tagging with adapters and/or barcode sequences. The target polynucleotide may be essentially any nucleic acid of known or unknown sequence. For example, the target polynucleotide may be prepared from a plasmid containing a DNA assembly of known genes and other functional elements. If rolling circle amplification is used to prepare a sample, then the target polynucleotide may include tandem repeats of the sequence of the circular template, such as a plasmid. In some embodiments, a target polynucleotide may include sequences of a vector and a polynucleotide insert (e.g., a DNA assembly).
In an embodiment, an input polynucleotide and a target polynucleotide may be the same. For example, if a plasmid mini-preparation procedure is used to amplify and isolate plasmid DNA, then an input polynucleotide (i.e., a plasmid) and target polynucleotide (i.e., a plasmid) generated from the mini-preparation may be the same. In another embodiment, an input polynucleotide and a target polynucleotide may be different. For example, if a plasmid DNA is subject to rolling circle amplification to generate a concatemer of a plasmid DNA, then the initial plasmid DNA may be referred to as an input polynucleotide, and the concatemer of the plasmid DNA, which is subject to fragmentation and tagging, is referred to a target polynucleotide.
As used herein, the term “sample” generally refers to anything capable of being analyzed by the methods provided herein that contains an input polynucleotide, a target polynucleotide, or any fragments thereof. In an embodiment, a sample may refer to a source for a particular input polynucleotide and/or target polynucleotide. For example, two plasmids comprising two different DNA assemblies may be referred to as two different samples. In some embodiments, replicates or clones comprising the same plasmid DNA may be referred to as separate samples.
As used herein, the term “consensus sequence” is a sequence determined after alignment of sequence reads associated with an input polynucleotide or a target polynucleotide generated from a sequencer by determining the base which is the most commonly found at each position in the compared, aligned sequence reads.
As used herein, the term “tagged DNA fragment,” “tagmented DNA fragment,” “tagged polynucleotide,” or “tagmented polynucleotide” refers to a piece of DNA or polynucleotide which has been fragmented and tagged or appended with one or more additional components, such as a transposon end sequence. In an embodiment, the tagged DNA fragment or tagged polynucleotide fragment may be generated during a tagmentation reaction while incubating a target DNA or a target polynucleotide with transposomes or transposition complexes.
As used herein, the term “tagmentation reaction” refers to incubation of a target polynucleotide with transposomes or transposition complexes to tag and fragment the target polynucleotide with transposon ends.
As used herein, the term “tagmentation reaction mixture” refers to a reaction mixture that includes a mixture of tagged polynucleotide fragments, transposases, unreacted components of a tagmentation reaction, and other components generated from a tagmentation reaction. The term “reaction mixture” is also used herein to refer to a “tagmentation reaction mixture,” and any discussions related to a tagmentation reaction mixture provided herein also applies to a reaction mixture.
As used herein, the term “tagmentation reaction solution” refers to a reaction solution comprising the tagmentation reaction mixture that has been treated under a dissociation condition to remove transposases from tagged polynucleotide fragments. The term “reaction solution” is also used herein to refer to a “tagmentation reaction solution,” and any discussions related to a tagmentation reaction solution provided herein also applies to a reaction solution.
As used herein, the term “dissociation condition” refers to a condition that can be used to treat the tagmentation reaction mixture to dissociate or remove transposases from tagged polynucleotide fragments generated from a tagmentation reaction. The dissociation condition can include, for example, treatment with heat or adding a solution, such as a dissociation or denaturing solution comprising a surfactant, which promote transposases to become unbound from tagged polynucleotide fragments.
As used herein, the term “primer” refers to a polynucleotide sequence that is capable of specifically hybridizing to a polynucleotide template sequence, e.g., a primer binding segment, and is capable of providing a point of initiation for synthesis of a complementary polynucleotide under conditions suitable for synthesis, i.e., in the presence of nucleotides and an agent that catalyzes the synthesis reaction (e.g., a DNA polymerase). The primer is complementary to the polynucleotide template sequence, but it need not be an exact complement of the polynucleotide template sequence. For example, a primer can be at least about 80, 85, 90, 95, 96, 97, 98, or 99% identical to the complement of the polynucleotide template sequence.
As used herein, the term “adapter” refers to a non-target nucleic acid component, generally DNA, which is joined to a target polynucleotide fragment and serves a function in subsequent analysis of the target polynucleotide fragment. In an embodiment, an adapter may include a nucleotide sequence that permits identification, recognition, and/or molecular or biochemical manipulation of the polynucleotide to which the adapter is attached. For example, an adapter may include a sequence which may be used as a primer binding site to read the sequence of the polynucleotide fragments. In another example, an adapter may include a barcode sequence which allows barcoded polynucleotide fragments to be identified.
As used herein, the term “adapter primer” refers to a primer that is capable of specifically hybridizing to a portion of a tagged polynucleotide fragment (e.g., to its primer binding segment, which may include a transposon end sequence), and is capable of providing a point of initiation for synthesis of a complementary polynucleotide under conditions suitable for synthesis. The adapter primer may be used in embodiments of the invention to append an adapter to a tagged polynucleotide fragment to generate a barcoded polynucleotide fragment.
As used herein, the term “barcode sequence” (also referred to as index) may be a known sequence used to associate a polynucleotide fragment with the input polynucleotide or target polynucleotide from which it is produced. It can be a sequence of synthetic nucleotides or natural nucleotides. In some embodiment, a barcode sequence is contained within adapter sequences such that the barcode sequence is contained in the sequencing reads. Each barcode sequence may include at least 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, or more nucleotides in length. In an embodiment, a barcode sequence may include 8 nucleotides in length. Generally, barcode sequences are of sufficient length and sufficiently different from one another to allow the identification of samples based on barcode sequences with which they are associated.
As used herein, “a sample specific barcode sequence” may refer to a barcode sequence specifically used for a particular sample and is different from barcode sequences used for other samples. A sample specific barcode sequence allows the identification of polynucleotide fragments derived from a particular sample (e.g., input or target polynucleotide) from another. In an embodiment, barcoded polynucleotide fragments from each sample may receive a unique combination of two barcode sequences so that sequence reads generated by a sequencer can be assigned to the correct samples (i.e., input polynucleotides) based on the combination of barcode sequences.
As used herein, the term “barcoded adapter primer” refers to an adapter primer which comprises a barcode sequence.
As used herein, the term “tagged polynucleotide fragment” refers to a polynucleotide fragment resulting from a tagmentation reaction. The tagged polynucleotide fragment is “tagged” with transposon end sequences during tagmentation and may further include additional sequences added during extension during a few cycles of PCR.
As used herein, the term “barcoded polynucleotide fragment” refers to a polynucleotide fragment which comprises a barcode sequence. The barcoded polynucleotide fragment may be appended with one or more barcode sequences. The barcoded polynucleotide fragment may be appended with one or more adapters which include barcode sequences.
As used herein, the term polynucleotide “fragment” refers to a polynucleotide including part but not all of the polynucleotide from which it is derived. For example, a polynucleotide fragment may include a piece of a target polynucleotide which is tagmented, cut, or sheared. In some embodiments, a polynucleotide fragment may be generated by amplifying a particular target region from a genome or other sequences.
As used herein, the term “library” refers to a plurality of nucleic acids, and may be used to refer to nucleic acids derived from the same input polynucleotide, target polynucleotide and/or same sample.
As used herein, the term “sequencing run” refers to any step or portion of a sequencing experiment performed to determine some information related to at least one nucleic acid molecule.
As used herein, the term “next-generation sequencing” is a method for sequencing nucleic acid sequences at high speed and at low cost than the previously used Sanger sequencing. The term “next-generation sequencing” platform refers to massive parallel sequencing platforms that allow millions of nucleic acid molecules to be sequenced simultaneously.
A “next-generation sequencer” refers to a sequencer which is capable of next-generation sequencing. A next-generation sequencer can include a number of different sequencers based on different technologies, such as Illumina (Solexa) sequencing, Roche 454 sequencing, Ion torrent sequencing, SOLiD sequencing, and the like.
As used herein, the term “sequence reads” refers to a sequence or data representing a sequence of nucleotide bases, in other words, the order of monomers in a polynucleotide, which is determined by a sequencer.
As used herein, “depth (coverage)” in DNA sequencing refers to the number of times a nucleotide is read during the sequencing process. Deep sequencing indicates that the total number of reads is many times larger than the length of the sequence under study.
As used herein, “average coverage” refers to an average or median of all the per base coverage values. For example, a plasmid with 30× coverage will have an average of 30 reads spanning any given position within the plasmid. Some regions will have higher coverage, and some will have lower coverage. In an embodiment, an average coverage of 15× is set as a threshold to determine the quality of a consensus sequence generated from the sequence reads.
The term “simultaneously” or “concurrently” as used herein refers to any two or more processes that are occurring more or less at the same time. It is not intended that each process begins and ends precisely together, but only that their respective durations may overlap.
The singular forms “a,” “an,” and “the” include plural references unless the context clearly dictates otherwise. Thus, for example, reference to “an adapter primer” includes a single adapter primer as well as a plurality of adapter primers.
In one aspect of the invention, provided herein is a method of preparing polynucleotides and generating polynucleotide fragments for highly multiplexed sequencing. The present invention is particularly useful for simultaneously sequencing small-sized input polynucleotides (e.g., about 3 kb to 30 kb range) from hundreds to thousands of samples. The small sized input polynucleotide includes, for example, a plasmid DNA, PCR amplicons, and 16 rRNA. In one embodiment, an input polynucleotide in a sample may be a plasmid DNA comprising an assembled polynucleotide produced by stitching several DNA components. In some embodiments, the assembled polynucleotide in a plasmid may be produced using compositions and methods described in U.S. Pat. Nos. 8,546,136, 8,221,982, and 8,110,360, each of which is incorporated herein by reference in its entirety.
The plurality of input polynucleotides can be processed, combined, and sequenced together in a single sequencing run of a sequencing instrument in a cost effective and time efficient manner. In an embodiment, polynucleotides from many samples (e.g., 400, 500, 600, 700, 800, 900, 1000, 1100, 1200, 1300, 1400, 1500, 1600, 1700, 1800, 1900, 2000, 2100, 2200, 2300, 2400, 2500, 2600, 2700, 2800, 2900, 3000, 3100, 3200, 3300, 3400, 3500, 3600, 3700, 3800, 3900, 4000, 4100, 4200, 4300, 4400, 4500, 4600, 4700, 4800, 4900, 5000, 5100, 5200, 5300, 5400, 5500, 5600, 5700, 5800, 5900, 6000, 6100, 6200, 6300, 6400, 6500, 6600, 6700, 6800, 6900, 7000, 7100, 7200, 7300, 7400, 7500, 7600, 7700, 7800, 7900, 8000, 8100, 8200, 8300, 8400, 8500, 8600, 8700, 8800, 8900, 9000, 9100, 9200, 9300, 9400, 9500, 9600, 9700, 9800, 9900, 10000, 10100, 10200, 10300, 10400, 10500, 10600, 10700, 10800, 10900, 11000, 11100, 11200, 11300, 11400, 11500, 11600, 11700, 11800, 11900, 12000, 12100, 12200, 12300, 12400, 12500, 12600, 12700, 12800, 12900, 13000, 13100, 13200, 13300, 13400, 13500, 13600, 13700, 13800, 13900, 14000, 14100, 14200, 14300, 14400, 14500, 14600, 14700, 14800, 14900, 15000, 15100, 15200, 15300, 15400, 15500, 15600, 15700, 15800, 15900, 16000, 16100, 16200, 16300, 16400, 16500, 16600, 16700, 16800, 16900, 17000, 17100, 17200, 17300, 17400, 17500, 17600, 17700, 7800, 17900, 18000, 18100, 18200, 18300, 18400, 18500, 18600, 18700, 18800, 18900, 19000, 19100, 19200, 19300, 19400, 19500, 19600, 19700, 19800, 19900, 20000, or more) can be prepared to generate target polynucleotides which are then fragmented and tagged with unique barcode sequences. Thereafter, the barcoded polynucleotide fragments from different samples can be combined together and sequenced in a single sequencing run. The sequence reads generated from the sequencer can then be sorted according to the unique barcode sequences associated with each sample (i.e., input polynucleotide).
In embodiments of the present invention, any suitable methods can be used to tag target polynucleotides with barcode sequences. In one embodiment, target polynucleotides may be initially fragmented because a next-generation sequencer can typically read only about 10 to 1,000 base pairs. Generally, fragmentation can include enzymatic, chemical, or mechanical methods which are well known and available in the art. For example, polynucleotides can be fragmented by acoustic shearing, nebulization, sonication, restriction enzymes, or transposomes. See, e.g., U.S. Patent Application Publication Nos. 2010/0120098 and 2012/0264228. Thereafter, polynucleotide fragments can be appended with one or more adapters at their 5′ and/or 3′ ends, each adapter comprising a unique barcode sequence as well as additional functional sequences. The functional sequences, such as primer binding sites, may be used during subsequent library amplification and sequencing.
Adapters comprising barcode sequences may be attached to polynucleotide fragments using a variety of standard techniques known and available in the art. For example, adapters can be attached to polynucleotide fragments by a ligase or a polymerase. The ligase may be any enzyme capable of ligating an adapter sequence or any oligonucleotide to polynucleotides. Suitable ligases include T4 DNA ligase, which is commercially available. See, e.g., New England Biolas (Ipswich, Mass.). Methods for using ligases are also well known in the art. Exemplary methods are described in, for example, Bentley et al., Nature 456:49-51 (2008); WO 2008/023179; U.S. Pat. No. 7,115,400; and U.S. Patent Application Publication Nos. 2007/0128624; 2009/0226975; 2005/0100900; 2005/0059048; 2007/0110638; and 2007/0128624, each of which is incorporated herein by reference in its entirety.
Alternatively, target polynucleotides derived from a sample may be fragmented and adapters may be added to the 5′ and 3′ ends using tagmentation or transposition reactions. The methods for tagmentation or transposition reactions are well-known and available in the art. Exemplary methods are described in, for example, U.S. Publication Application No. 2010/0120098, which is incorporated herein by reference in its entirety. This technology is illustrated in
As shown in
Step (a) of
The previous tagmentation step leaves a short single stranded sequence gap in the tagged polynucleotide fragments. As shown in step (c), fragmented ends of the tagged polynucleotide fragment 113 are repaired and extended with a strand-displacing DNA polymerase. These extended fragments are also referred to as the tagged polynucleotide fragments in embodiments of the present invention. As shown in step (d), limited-cycle PCR can be performed with four primers: a terminal primer 114, a barcoded adapter primer 115, a terminal primer 116, and a barcoded adapter primer 117. This limited-cycle PCR reaction adds the barcoded adapters 125 and 127 to the tagged polynucleotide fragment 113.
As shown in
The terms, i5 and i7, shown in
The primers in the Illumina Nextera sample preparation kit have the following sequences:
In the i5 and i7 index primers shown above, the positions of the barcode sequences are shown as [i5] and [i7], respectively. As shown in
After PCR amplification in step (e), barcoded polynucleotide fragments 123 are generated. As shown in
In the flowchart illustrated in
In the exemplary workflow shown in
In the exemplary embodiment shown in
Referring to
When RCA prepared target polynucleotides are used in tagmentation reactions, it was discovered by the present inventors that the size distributions of RCA prepared target polynucleotides that had been normalized before tagmentation were very similar to those that had not been normalized. See, e.g.,
It was also discovered by the present inventors that when the polynucleotide concentration in the RCA solution is diluted to about 3 ng/μL to about 10 ng/μL (e.g., average of about 5 ng/μL) prior to the tagmentation step, then the quality of sequencing improves for pooled samples. See, e.g.,
Referring to step (202) of
A suitable standard dilution factor may be determined in a number of different ways. In one embodiment, a standard dilution factor may be determined by quantifying target polynucleotides in at least a portion of a plurality of RCA solutions. For example, if there are 4000 RCA solutions comprising target polynucleotides, then the polynucleotide concentration may be quantified for each of 4000 RCA solutions. In some embodiments, the polynucleotide concentration in a portion of the samples (e.g., a single 384-well plate instead of all plates) may be measured since RCA provides a relatively consistent final concentration of target polynucleotides. Based on the measured concentration of target polynucleotide in each RCA solution, an average concentration of target polynucleotides in all or at least a portion of RCA solutions may be calculated. The standard dilution factor to dilute each RCA solution can then be determined by dividing the average concentration by any number selected from 3 ng/μL to 10 ng/μL, as this range was found to provide relatively consistent sequencing coverage and less variability during sequencing. In an embodiment, a number in the middle of the range (e.g., 5, 6, or 7 ng/μL) can be selected for determining a standard dilution factor. In an embodiment, the standard dilution factor is calculated by dividing the average concentration by 5 ng/μL. Thus, in certain embodiments, an average of about 1.5 ng to about 5 ng of polynucleotides is used in a tagmentation reaction volume of 0.5 μL. In another embodiment, an average of about 3 ng to about 10 ng of polynucleotides is used in a tagmentation reaction volume of 1 μL. In another embodiment, an average of 6 ng to 20 ng of polynucleotides is used in a tagmentation reaction volume of 2 μL.
In another embodiment, a standard dilution factor may be determined by measuring a concentration of target polynucleotides in a mixed RCA solution. For example, an equal volume of RCA solutions derived from all samples (or at least a portion thereof) can be mixed together, thereby generating a mixed RCA solution comprising target polynucleotides. Thereafter, an average concentration of target polynucleotides in the mixed RCA solution can be determined. This requires quantification of only a single “mixed” RCA solution. Based on the concentration of polynucleotides in the mixed RCA solution, a suitable standard dilution factor may be determined.
In step (202), any suitable methods can be used to quantify a concentration of polynucleotides in a solution. For example, a fluorescent dye, PicoGreen dsDNA quantitation reagent (Quant-iT PicoGreen dsDNA assay kit, Life Technologies, Foster City), may be used. The method utilizes the increased fluorescent intensity that is observed when PicoGreen binds to dsDNA. The fluorescent intensity of the PicoGreen dye is measured with a spectrofluorometer capable of producing the excitation wavelength of about 480 nm and recording at the emission wavelength of about 520 nm.
While steps (201) and (202) in
Referring to
Any suitable transposomes or transposition complexes may be used in the present method. Some of them are known in the art and available as commercially available kits. For example, the Ez-Tn™ hyperactive Tn5 Transposase and the HyperMu™ Hyperactive MuA Transposase are available from Epicentre Technologies, Madison, Wis. See, also, U.S. Patent Application Publication No. 2010/0120098, which is incorporated herein by reference in its entirety. In an embodiment, the transposition complexes may include transposases such as Tn5 or MuA and their respective transposon terminal end sequences. See, e.g., Goryshin and Reznikoff, J. Biol. Chem., 237: 7367, 1998; and Mizuuchi, Cell, 35: 785, 1983; Savilahti et al., EMBO J., 14: 4893, 1995; which are incorporated by reference in their entireties. Other transposition complexes including transposases, such as Tn552, Ty1, Tn7, and Tn3, may be used in some embodiments of the present invention. Transposomes or transposition complexes are also commercially available as kits and can be purchased from, for example, Illumina Inc. (Nextera DNA library preparation kit), KAPA Biosystems (Kapa DNA library preparation kits), Molecular Cloning Laboratories (Next DNA sample kit), New England Laboratory (NEB Next kits), and the like.
A suitable ratio of transposomes to target polynucleotides for tagmentation reaction can be determined based on knowledge in the art and the present disclosure. Generally, it is desirable to have a relatively precise transposomes to target polynucleotide ratio during tagmentation. The ratio can affect the quality of tagmentation as well as coverage during sequencing. The extent of the fragmentation and/or the size of fragments can be controlled using appropriate reaction conditions such as by using the suitable concentration of transposomes and controlling the temperature and time of incubation. In an embodiment, suitable reaction conditions can be obtained using known amounts of a test library of nucleic acids and titrating the transposomes and time to build a standard curve for actual sample libraries. Exemplary tagmentation reaction conditions are also described in detail in the Examples section.
In an embodiment, any suitable tagmentation reaction volumes may be selected to fragment and tag target polynucleotides. In some embodiments, a suitable tagmentation reaction volume may include 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0.5, 0.1, 0.01, 0.005 μL or any number in between these numbers. For highly multiplexed sequencing, tagmentation reactions are generally performed in a small volume. A small tagmentation volume requires a reduced amount of transposases and other tagmentation reagents, which can save cost. Furthermore, if an acoustic liquid transfer system (e.g., Echo 550, Labcyte, Sunnyvale, Calif.) is used, it does not require pipettes for liquid transfer, reducing potential contamination between samples. In some embodiments, a suitable tagmentation reaction volume may include between about 0.005 μL to about 2 μL. In certain embodiments, the tagmentation reaction is performed at a volume of about 2 μL or less, typically about 1 μL or less, and more typically at about 0.5 μL. For a small reaction volume of 0.5 μL, typically 200 nL of DNA (having a concentration between about 3 ng/uL to about 10 ng/uL, typically about 5 ng/μL) can be added to 300 nL of a tagmentation enzyme solution which includes transposition complexes and reagents. In other words, about 0.6 ng to about 2 ng (typically about 1 ng) of target polynucleotide is generally used in a tagmentation reaction having a volume of about 0.5 μL.
In some embodiments as shown in the Examples section, the tagmentation reaction is performed at 0.5 μL, which is 100-fold less than the tagmentation reaction volume required in the Illumina Nextera kit. It was discovered by the present inventors that the 100-fold reduction in tagmentation volume does not change the quality of sequencing coverage or variability. For example, as shown in
Referring to
In an embodiment, a dissociation solution may comprise an ionic surfactant, such as sodium dodecyl sulfate (SDS). For example, a dissociation solution comprising SDS at a final concentration of about 0.05% to about 0.3%, more typically about 0.1% (weight per volume percent) may be used to remove transposases. The final concentration of SDS may refer to the concentration of SDS when the solution comprising SDS is added to a tagmentation reaction mixture (containing tagged polynucleotide fragments, transposases, and other components used in the tagmentation reaction). For example, 125 nL of 0.5% SDS in TE can be added to 500 nL of the tagmentation mixture, which results in a final SDS concentration of 0.1%. In some embodiments, the dissociation solution consists of SDS as a dissociation or denaturing agent in TE (or other suitable buffers). In some embodiments, other dissociation agents may be used alone or in combination with SDS. For example, Triton X-100 may be used in combination with SDS. In some embodiments, a dissociation solution may comprise 1% Triton X-100 and 0.3% SDS.
While there are advantages to using a dissociation condition without column spins or other solid phase extraction, embodiments of the present invention are not limited to using specific transposase removal methods. Any suitable removal methods, column spin or DNA binding matrix beads, may be used to separate transposases from polynucleotide fragments prior to PCR. For example, commercially available kits, such as Zymo kit (Illumina, San Diego, Calif.), may be used.
Referring to
In an embodiment, the barcode sequence can be a sequence of synthetic nucleotides or natural nucleotides that allow for easy identification of the polynucleotide fragments to which it is attached in a collection of other polynucleotide fragments. Generally, barcode sequences are of sufficient length and comprise sequences that are sufficiently different from one another. For example, each barcode sequence may include at least 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, or more nucleotides in length. In an embodiment, a barcode sequence may include 8 nucleotides in length. The barcode sequences generated by the present method (see section 6.3 below) can be used to uniquely tag polynucleotide fragments from each sample (i.e., input polynucleotide). In some embodiments, the barcode sequences designed according to the present method can be incorporated into any suitable adapter primers. For example, the present barcode sequences can be incorporated into Illumina i5 and i7 index primers if the Illumina MiSeq or other sequence platform is used for sequencing. In this embodiment, any one of barcode sequences SEQ ID NO: 1 through 192 may be inserted into positions [i5] and [i7] of adapter primers having SEQ ID NO: 195 and SEQ ID NO: 196, respectively.
In an embodiment, a pair of unique barcode sequences may be introduced to each polynucleotide fragment. After introducing a pair of barcode sequences into polynucleotide fragments and dually indexing them, a suitable sequencing instrument can be used to read both barcode sequences to identify the source of the polynucleotide fragments (e.g., input polynucleotide from a sample). Through dual indexing, sample misidentification inaccuracies can be reduced. For sequencing a smaller number of samples, however, a single barcode sequence may be used if desired.
In step (204) of
The PCR reaction can be initiated in a reaction chamber comprising a PCR master mix and a tagmentation reaction solution that includes tagged polynucleotides and adapter primers under a suitable thermocycling condition (205). A PCR master mix may include a solution that contains water, 10× Thermopol buffer, MgSO4, DNA polymerase, dNTPs, MgCl2, deoxynucleotide triphosphates, terminal primers, and a DNA polymerase at their optimal concentrations for efficient amplification of template DNA by PCR. As shown in
In an embodiment, the PCR master mix may include a large amount of water or other suitable aqueous solution to dilute the tagmentation reaction solution generated in the previous step (203). The large dilution prevents transposases in the solution from interfering with the PCR reaction. For example, if the tagmentation reaction is performed at a volume of 0.5 μL, then 20.275 μL of water may be added together with other PCR reagents to bring the final volume of PCR reaction to 25 μL. While this exemplary dilution illustrates a 50-fold dilution of the tagmentation mixture (i.e., 0.5 μL diluted to 25 μL), any suitable dilution ratio may be used to prevent transposases from interfering with PCR. For example, the tagmentation mixture may be diluted by at least 5, 10, 15, 20, 25, 30, 35, 40, 45, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, 100, 105, 110, 115, 120, 125, 130, 135, 140, 145, 150, 155, 160, 165, 170, 175, 180, 185, 190, 195, 200, or more. The reduced amount of template polynucleotide during PCR can be compensated by adjusting the number of PCR cycles. In an embodiment, 8 to 24 cycles of PCR, more typically about 12 cycles, may be used to generate and amplify barcoded polynucleotide fragments.
While
Referring to
In some embodiments, a “double-sided” solid reverse immobilization (DSPRI) purification protocol can be used to clean the libraries of PCR products. Polynucleotide fragments that have a high proportion of larger fragments (e.g., greater than 1000 base pairs) can result in a lower average depth coverage during sequencing. During the DSPRI, a first set of beads may be added to the polynucleotide fragments at a low volume to remove large fragments (e.g., greater than 1000 base pairs), and the supernatant is then collected. A second set of beads can then be added to the supernatant to remove small fragments (e.g., less than 300 base pairs). The DSPRI protocol may enrich DNA fragments having a length between 300 and 800 base pairs, which is desirable for next-generation sequencing. By removing populations of both small fragments and large fragments prior to sequencing, the average depth of sequencing may be improved.
After cleaning the libraries of barcoded polynucleotide fragments by removing undesired fragment sizes, the polynucleotide fragments in the libraries can be quantified if desired (207). To achieve the highest quality of data on sequencing platforms, the barcoded polynucleotide fragments from each sample can be accurately quantified so that they can be combined at equal molar ratios with barcoded polynucleotide fragments from other samples. This process can improve even depth of coverage across the combined pool of polynucleotide fragments. The DNA quantification of libraries can be performed using any suitable methods, such as PicoGreen assay. The details of an exemplary protocol for the PicoGreen assay are further described in the Examples section. In some embodiments, other dsDNA-specific fluorescent dye method, such as Qubit, may be used to quantify the library.
Each of steps (201) through (207) shown in
Referring to
Furthermore, in step (207), the libraries can be normalized for the input polynucleotide length prior to pooling in certain embodiments. As an illustration, if all the libraries are derived from a plasmid having the same length, then all the libraries are pooled together at an equal volume (assuming that the libraries have the same concentration of DNA). On the other hand, if the first library is derived from a plasmid which has twice the length as the second library, then the volume of the first library added into a pool will be twice as large as the second library (assuming that both libraries have the same DNA concentration). This way, the entire length of both plasmids will be equally presented to a sequencer for even coverage of all the libraries.
While steps (207) and (208) can improve the depth of sequencing coverage across the combined pool of polynucleotide fragments, these steps are optional and can be omitted for expediency without greatly reducing the quality of sequence data.
Referring to
In certain embodiments, the filtered pool of polynucleotide fragments can then be further characterized before sequencing in step (209). For example, the distribution of fragment sizes of the pooled polynucleotide fragments can be measured using a Bioanalyzer, Fragment Analyzer, or by integrating the signal intensity along an agarose gel. The molar concentration of the pooled DNA sample can be calculated using PicoGreen value and the measured average fragment size as further described in the Examples section. For example, the molar concentration of the pooled polynucleotide fragments can be calculated as follows:
Molar concentration (nM)=PicoGreen value (ng/μL)×1,000,000/(660×avg fragment size)
Any suitable sequencer (e.g., MiSeq) can be used to load a combined pool of barcoded polynucleotide fragments at a suitable molar concentration (e.g., 12 pM) as recommended by the sequencer. The sequence reads generated from the sequencer can be sorted or demultiplexed based on the barcode sequences using the software provided with the sequencer.
The workflow shown in
It should be appreciated that the specific steps illustrated in
In another aspect, provided herein are barcode sequences, adapter primers comprising barcode sequences, and methods of generating these sequences suitable for highly multiplexed sequencing. In some embodiments, unique barcode sequences can be incorporated into adapters, which are appended to polynucleotide fragments to generate barcoded polynucleotide fragments for sequencing. In some embodiments, unique barcode sequences may be appended or ligated directly to the tagged polynucleotide fragments. The specific sequence or “index” used as a barcode sequence is unrestricted. It can be any suitable length, such as 6, 7, 8, 9, 10, 11, 12, or the like. Generally, barcode sequences are of sufficient length and comprise sequences that are sufficiently different from other barcode sequences to allow the identification of samples to which they are associated.
In
If the candidate barcode sequence does not include a homopolymer run of 3 base pairs or more, it is determined, using the computer processor, whether every candidate barcode sequence has a Hamming distance of three or more from all other candidate barcode sequences (1120). By definition, the Hamming distance between two strings of equal length is the number of positions at which the corresponding symbols are different. In other words, it is the number of substitutions required to transform one string into another. For example, in the context of a nucleic acid sequence, the Hamming distance between AAGGTTCG (SEQ ID NO: 198) and AAGGCCCG (SEQ ID NO: 199) is 2 since “TT” in the first sequence needs to be replaced with “CC” to transform it into the second sequence. One of these two candidate barcode sequences will be eliminated since they have a Hamming distance of less than three.
The method of generating barcode sequences further includes determining whether every candidate has a Hamming distance of three or more from every eight base segment of the conserved regions of adapter primers. For example, if adapter primers, SEQ ID NOS: 195 and 196, shown below were selected as adapter primer sequences for amplifying tagged polynucleotides, then every candidate must have a Hamming distance of three or more from every eight base segment shown in SEQ ID NOS: 195 and 196.
As an example, if a candidate barcode has a sequence of TTTGATA in step (1125), then this candidate will be eliminated as a potential barcode sequence because it has a Hamming distance of 2 with the first 8 bases (AATGATA) (SEQ ID NO: 200) of the N′ terminal end of SEQ ID NO: 195.
Based on the above steps (1110) through (1125), a novel set of 826 8-base pair candidate indices have been identified. To further optimize the quality of barcode sequences in the context of adapter primers, each of the candidate barcode sequences is inserted into the barcode position of the adapter primers to be used during PCR. For example, if adapter primers shown in SEQ ID NO: 195 and 196 are to be used during PCR (e.g., step (205) of
In the next few steps, candidate barcoded adapter primers are further analyzed. For example, candidate barcoded adapter primers generated in step (1130) are filtered out if they have mononucleotide runs longer than two bases or a GC content outside of 35% to 65% (1135). The “GC content” refers to the ratio of the number of guanine and cytosine to the total number of all bases in nucleic acids or deoxyribonucleic acids. Then, sequences differing by at least three bases from all other barcoded adapter primers in the set, or from sequences complementary to all 8-base sequences present within the conserved regions of the adapter primers are then selected (1140).
The candidate barcode sequences selected through step (1140) are further filtered by placing them into the context of the full-length adapter primers. For example, each candidate barcode sequence is inserted into position [i5] of SEQ ID NO: 195 and position [i7] of SEQ ID NO: 196. The resulting barcoded adapter primers are analyzed to determine their melting profile. For this step, any suitable DNA melting prediction software, such as DINAMelt, may be used (1145). See Nicholas R. Markham at Rensselaer Polytechnic Institute, which is downloadable from the DINAMelt web site. See, also, Nuc. Acids Res. 2005, vol. 33, W577-W581. The DNA melting prediction software can be used to simulate oligonucleotide melting, and to select those with the lowest predicted tendency to form inter- or intra-molecular duplexes. For example, an oligonucleotide that satisfies a threshold Gibbs free energy may be selected as a final set of barcoded adapter primers (1150). Generally, oligonucleotides that have a more negative Gibbs free energy tend to form inter- or intra-molecular duplexes. Therefore, the stability (Gibbs free energy) may be set at any suitable threshold level (e.g., ΔG=−5) under a typical PCR reaction and salt conditions to filter out unstable barcoded adapter primer candidates.
Using the steps shown in the flowchart of
The barcode sequences or barcoded adapter primers generated using the method shown in
It should be appreciated that the specific steps illustrated in
In another aspect of the invention, a kit for generating a sequencing library is provided. A kit may comprise a pair of barcoded adapter primers that includes one or more barcoding sequences generated according to embodiments of the present invention. See section 6.3 above. In some embodiments, the barcoded adapter primers may include barcode sequences of SEQ ID NO: 1 through SEQ ID NO: 192. In another embodiment, these barcode sequences can be inserted into adapter primers of SEQ ID NO: 195 and SEQ ID NO: 196 at position [i5] or [i7] to generate barcoded adapter primers. Each of these barcode sequences and barcoded adapter primers is optimally designed to be distinguishable during sequencing using the Illumina or other sequencing platform. Kit embodiments may also include other additional adapter primer sequences which are generated using the method described with reference to
In some embodiments, the kits may further include reagents that can be used with the present barcoded adapter primers. These kit embodiments may comprise a PCR master mix including one or more standard dNTPs, a DNA polymerase (e.g., Vent polymerase), terminal primers, buffers, and the like. Some kit embodiments may further include reagents for DNA sample preparation, a tagmentation reaction mix, and a transposase removal agent. The kit can further include instructions for the sample preparation, tagmentation reaction and removal of transposases, PCR reactions, sequencing, and the like.
Some kits may further comprise software for processing sequence data. For example, the software may include sorting sequence reads and assigning them to their source (e.g., sample) using the barcode sequences, and aligning and assembling the sorted sequence reads for each sample to generate a consensus sequence of the template polynucleotide in the sample. The software may further include modules to align the sequence reads and/or the consensus sequence to a reference sequence to identify sequence differences (e.g., deletions, indels, mutations, sequencing errors, etc.). The software may further include modules to correct sequencing errors based on the alignment.
In another aspect, the barcoded polynucleotide fragments prepared and generated in accordance with the present invention can be sequenced using any suitable methods. In an embodiment, a next-generation sequencer can be used to sequence millions of nucleic acid molecules simultaneously. Some platforms rely on sequencing-by-synthesis approach, while other platforms may use sequencing-by-ligation or other approach.
An example of a sequencing technology that can be used in the present methods is the Illumina platform. The Illumina platform is based on amplification of DNA on a solid surface (e.g., flow cell) using fold-back PCR and anchored primers (e.g., capture oligonucleotides). For sequencing with the Illumina platform, DNA is fragmented, and adapters are added to both terminal ends of the fragments. DNA fragments are attached to the surface of flow cell channels by capturing oligonucleotides which are capable of hybridizing to the adapter ends of the fragments. The DNA fragments are then extended and bridge amplified. After multiple cycles of solid-phase amplification followed by denaturation, an array of millions of spatially immobilized nucleic acid clusters or colonies of single-stranded nucleic acids are generated. Each cluster may include approximately hundreds to a thousand copies of single-stranded DNA molecules of the same template. The Illumina platform uses a sequencing-by-synthesis method where sequencing nucleotides comprising detectable labels (e.g., fluorophores) are added successively to a free 3′hydroxyl group. After nucleotide incorporation, a laser light of a wavelength specific for the labeled nucleotides can be used to excite the labels. An image is captured and the identity of the nucleotide base is recorded. These steps can be repeated to sequence the rest of the bases. Sequencing according to this technology is described in, for example, U.S. Patent Publication Application Nos. 2011/0009278, 2007/0014362, 2006/0024681, 2006/0292611, and U.S. Pat. Nos. 7,960,120, 7,835,871, 7,232,656, and 7,115,200, each of which is incorporated herein by reference in its entirety.
In some embodiments, paired end reads may be obtained on nucleic acid clusters on the substrate, where each immobilized polynucleotide is sequenced from both ends of the fragment. Paired end runs read from one end to the other end, and then start another round of reading from the opposite end. In other words, the sequences of the paired reads are read towards each other on opposite strands. When they are aligned against the genome or reference sequence, one read should align to the forward strand, and the other should align to the reverse strand, at a higher base pair position so that they are pointed towards one another. Paired end sequencing runs can provide additional positioning information about the DNA template. Methods for obtaining paired end reads are described in WO/2007/010252 and WO/2007/091077, each of which is incorporated herein by reference.
Another example of a DNA sequencing technology that can be used with the methods of the present invention is SOLiD technology by Applied Biosystems from Life Technologies Corporation (Carlsbad, Calif.). In SOLiD sequencing, DNA may be sheared into fragments, and adapters may be attached to the terminal ends of the fragments to generate a library. Clonal bead populations may be prepared in microreactors containing template, PCR reaction components, beads, and primers. After PCR, the templates can be denatured, and bead enrichment can be performed to separate beads with extended primers. Templates on the selected beads undergo a 3′ modification to allow covalent attachment to the slide. The sequence can be determined by sequential hybridization and ligation with several primers. A set of four fluorescently labeled di-base probes compete for ligation to the sequencing primer. Multiple cycles of ligation, detection, and cleavage are performed with the number of cycles determining the eventual read length.
Another example of a DNA sequencing technology that can be used with the methods of the present invention is Ion Torrent sequencing. In this technology, DNA is sheared into fragments, and oligonucleotide adapters are then ligated to the terminal ends of the fragments. The fragments are then attached to a surface, and each base in the fragments is resolvable by measuring the H+ ions released during base incorporation. This technology is described in, for example, U.S. Patent Publication Application Nos. 2009/0026082, 2009/0127589, 2010/0035252, 2010/0137143, and 2010/0188073, each of which is incorporated herein by reference in its entirety.
While three different sequencing technologies are described above, other sequencing platforms and processes can be easily implemented for use with the methods, compositions, and kits described herein.
In another aspect, provided herein is a method of analyzing sequence reads generated by a sequencer using a set of computer-readable instructions or codes (i.e., software). After the sequencer has generated sequenced reads and assigned them to the proper sample, each batch of reads can be aligned to its template (e.g., a digital reference sequence stored in a database). While these functions can be performed by a sequence analyzer module of a sequencer (e.g., Miseq), in some embodiments, these and other functions can be programmed as separate software and performed by a separate computer apparatus dedicated to a sequencer, a user computer and/or a server computer as shown in
In an embodiment, a computer apparatus or system with a user interface may be provided to upload a sample sheet (e.g., csv file) that includes sample and barcode information for each sequencing run on a sequencer. The sequencer assigns each run to the correct sample based on the barcode sequences, and collects the sequence reads in files in a suitable file format (e.g., FASTQ). In the method shown in
Referring to
If all DNA components of the DNA assembly are present, then the method further includes analyzing assembled read sequences and the digital reference sequences for smaller differences, for example, single nucleotide polymorphism (SNPs) or indels (e.g., deletions or insertions) (1170). If all of the DNA components are present, then it can be either delivered to a customer who requested the DNA assembly and/or stored in the bank (e.g., freezer) (1172). If there are only small differences between the sequence reads and the digital reference sequence, then the algorithm determines if those differences are in a portion of the plasmid that may affect the function or expression of the genes in the construct (1174). For example, if a change is observed in a linker (e.g., a region of untranslated DNA between two parts), the plasmid containing the DNA assembly may be considered “safe” and may be delivered to the customer or stored in the bank. However, if the variant (e.g., SNPs or indels) is likely to disrupt the intended function (e.g., a premature stop codon in the coding part), it may be flagged as fatal, and the plasmid may be discarded and/or not delivered to the customer.
In some embodiments, a sequence data plot for a plasmid DNA can be generated and displayed on a user interface of a computer for each sample (1176). In a sequence data plot, the x-axis may represent the nucleotide position of the plasmid DNA, and the y-axis may represent the depth of coverage for each nucleotide position. Exemplary sequence data plots are illustrated in
In some embodiments, for plasmids containing DNA assemblies stitched from several DNA components, it may be desirable to sequence replicates (e.g., multiple clones) of the plasmid as part of quality control. In these embodiments, the sequence reads from each replicate can be compared against its reference sequence stored in a database. The aligned sequences for each of the replicates can then be compared, and the best replicate (e.g., with read sequences with no deletions, mutations, or substitutions, or the like compared to the reference sequence) may be determined. The method shown in
In an embodiment, the method shown in
It should be appreciated that the specific steps illustrated in
Various methods of the present invention can be performed using one or more computer apparatuses in a computer system. An exemplary computer system 1200 is shown in
All the computer apparatuses shown in
While some of the components of the computer apparatuses are shown in
The sequencer 1220, in addition to sequence data receiver module 1221 may include sequence analysis module 1222 in memory 1224, a processor 1223, and input/output module 1225. The sequencer data receiver module 1221 may receive a sample sheet (e.g., in csv file) that contains information related to a sample, barcode sequences, and other relevant information for sequence analysis through input/output module 1225 and communication medium 1260. The sequence analysis module 1222 may analyze sequence reads and sort the sequence reads using the barcode sequences and other sample information received in the sequencer data receiver module 1221. The analyzed sequence information may be transmitted to the server computer 1260 and/or the user computer 1250 through the communication medium 1260 for further analysis. Although
The oligonucleotide synthesizer 1230, in addition to the oligonucleotide data receiver 1231, may include a synthesis module 1232 in memory 1234, a processor 1233, and input/output module 1235. The oligonucleotide synthesizer 1230 may receive a request to synthesize a barcode sequence, a primer, an adapter, or other nucleotide sequences through the input/output module 1235 and communication medium 1260. The synthesis module 1232 may include software to execute the synthesis of requested oligonucleotides.
The server computer 1240 may include a processor 1241, memory 1242, data storage 1243, and input/output module 1244. The server computer 1240 may interact with other computer apparatuses of the system 1200 and may be used to store data, obtain data, process data, or to output processed and analyzed data to the user computer 1250, sequencer 1220 and/or oligonucleotide synthesizer 1230. For example, reference sequences stored in the data storage 1243 may be retrieved by the user computer 1250 or the sequencer 1220 to compare the digitally stored reference sequences against sequence reads generated by the sequencer 1220.
The user computer 1250 may also include a processor 1251, memory 1252, data storage 1253, and input output device 1256 which may include input/output module 1254 and user interface 1255. The user of the user computer 1250 can communicate with any computer apparatuses of the computer system 1200 via the communication medium 1260. The user of the user computer 1250 may request data or receive data through input/output module 1255 and communication medium 1260. The data, such as sequence alignment and/or sequence coverage data may be analyzed by the server computer 1240 or the user computer 1250, and the analyzed data may be displayed on the user interface 1255 on the user computer. For example, the user computer 1250 may compare sequence reads against a reference sequence for a sample and display sequence data plots as shown in
Any of the software components or functions described in this application may be implemented as software code to be executed by a processor using any suitable language, such as, for example, Java, C++, or F#. The software code may be stored in a series of instructions, or commands on a computer readable medium, such as random access memory (RAM), a read only memory (ROM), a magnetic medium, such as a hard-drive, or an optical medium such as a CD-ROM. Any such computer readable medium may reside on or within a single computer apparatuses, or may be present on or within different computer apparatuses within a system or network.
Liquid transfers were carried out on Biomek FX or NX robots (Beckman Coulter, Brea, Calif.) for volumes greater than 2 μL or on an Echo 550 plus Access robotics (Labcyte, Sunnyvale, Calif.) for volumes less than 2 μL. Sequencing was done on a MiSeq (Illumina, Inc., San Diego, Calif.). Fluorescence was read on an M5 plate reader (Molecular Devices, LLC, Sunnyvale, Calif.). DNA fragment size profiles were determined using either a Bioanalyzer 2100 (Agilent Technologies, Inc., Santa Clara, Calif.) or a Fragment Analyzer (Advanced Analytical Technologies, Inc., Ames, Iowa).
DNA parts with specific linker sequences at each end were assembled in a shuttle vector using yeast homologous recombination, followed by shuttling into Escherichia coli for isolation of DNA, as previously described (Dharmadi et al. (2014) Nucleic Acids Res 42: e22). DNA assemblies built using the ligase cycling reaction (LCR) (de Kok et al. (2014) ACS Synth. Biol. 3: 97-106) were also used in some experiments. Plasmid DNA was prepared by alkaline lysis and silica gel binding (Dharmadi et al., supra) or was amplified using an Illustra Templiphi kit (GE Healthcare Life Sciences, Piscataway, N.J.). DNA concentration was measured using Quant-iT PicoGreen reagent (Life Technologies, Foster City, Calif.) in Costar 3658 or 3677 black 384-well plates (Corning, Inc., Corning, N.Y.). The PicoGreen reagent was diluted with TE (10 mM Tris-HCl, pH 8, 0.5 mM EDTA) containing 0.05% Tween 20.
As described above,
Adapters for the Illumina sequencing process, including 8-base barcodes, were attached to each tagmented DNA sample using 12 cycles of PCR. All primers were obtained from IDT (Integrated DNA Technologies, Inc., Coralville, Iowa) with standard desalting. The barcodes inserted into the Illumina i5 and i7 adapter primer sequences are listed in Table 2. Using the Echo, each sample well received 125 nL of a forward barcode primer and 125 nL of a reverse barcode primer (each at 100 μM). A PCR master mix (24.5 μL) was then added using a Biomek robot. The master mix contained 0.2 units/μL of Vent DNA polymerase (New England Biolabs, Ipswich, Mass.), 1× Thermopol buffer (NEB), 2 mM MgSO4, 200 μM of each deoxynucleotide triphosphate, and 200 nM of each terminal primer (to mitigate the fact that long oligonucleotides have 5′-end truncations). The thermocycler program was 3 minutes at 72° C., then 12 cycles of 10 seconds at 98° C., 30 seconds at 63° C. and 60 seconds at 72° C. Small fragments and unincorporated primers were removed from the resulting PCR products using 0.6 volume of Ampure XP paramagnetic bead suspension (A63880, Beckman Coulter, Indianapolis, Ind.) per volume of PCR reaction according to the manufacturer's instructions.
Libraries were pooled and normalized based on DNA concentration, and the size of the DNA assembly from which the library was generated. The goal of normalization is to achieve equal molar amounts of the DNA representing each plasmid (see Results and Discussion). The pool was filtered and concentrated using a Microcon Fast-Flow filter unit (EMD Millipore, Billerica, Mass.). The DNA concentration and average fragment size of the pool were determined by Picogreen fluorescence and a high sensitivity DNA chip on a Bioanalyzer 2100, respectively. After diluting the filtered pool to 1.11 nM with water, 18 μL was denatured by adding 2 μL 1N NaOH. After 5 minutes at room temperature, 980 μL ice-cold Illumina Hybridization Buffer was added, followed by 2 μL 1N HCl. The denatured pool was loaded on the MiSeq at 12 pM, which was empirically determined to give the optimum cluster density when following this protocol.
A web-based sequencing tracking system was created to manage the many samples and the large amounts of data generated. It facilitates the creation of runs, generation of sample sheets required by the MiSeq, and analysis of multiple data types, including the NGS QC data described here. Reads were demultiplexed using the embedded MiSeq Reporter software. For large numbers of multiplexed samples (greater than 1000), the “File Copy Timeout” setting was increased to avoid premature interruption of the demultiplexing process, which can take several extra hours after a highly multiplexed run appears to have completed. When a sequencing run completes, the system automatically retrieves the FASTQ files from the MiSeqOutput folder. Read mapping to the intended assembly sequences uses BWA v0.6.232 and the “sample” method with default settings. See Li and Durbin (2009) Bioinformatics 25: 1754-1760. Alignments are stored in BAM file format using SAMTOOLS v0.1.19. See Ramirez-Gonzalez et al. (2012) Source Code Biol. Med. 7: 6; Li et al. (2009) Bioinformatics 25: 2078-2079. Mapping statistics are obtained using the SAMTOOLS flagstat utility. A pileup file is generated using SAMTOOLS mpileup with default options to obtain read coverage along the reference sequence.
Table 1 provides an exemplary schematic workflow of next-generation sample preparation. The sample preparation typically has three main phases. In the first phase, tagmentation samples are all normalized to a uniform concentration (1a) and then treated with a fragmentation and labeling enzyme, such as Tn5 transposase pre-loaded with DNA that will flank all template fragments (1b). Once the reaction is complete, the DNA (e.g., tagged polynucleotide fragments) is separated from the tightly-bound transposase in such a way that the template is still competent for PCR (1c). In the second phase, samples are amplified using limited-cycle PCR with primers that contain unique barcodes (2a, b). Once PCR is complete, small high-molarity DNAs that would compete for binding sites on the sequencing surface are removed (2c). In the third phase, the sample concentration and fragment size distribution can be measured and used to normalize the molarity of sequenceable molecules across all samples in certain embodiments (3a).
Tagmentation is like transposon insertion (Reznikoff (2008) Annu Rev. Genet. 42: 269-286), except the transposome cuts the target DNA and appends tags (transposon terminal sequences) to the resulting fragments as shown in
The total amount used per sample can be decreased by scaling down the tagmentation reaction from 50 μL to 0.5 μL. The reduction in volume was performed in a stepwise fashion by modifying other protocol steps as necessary to adjust for reduced samples volume and reduced total mass of DNA.
Since conventional liquid handlers have unacceptable accuracy for handling liquids having a volume of less than 2 μL, a reaction volume of 5 μL (2 μL of DNA and 3 μL of a 1:5 mix of enzyme with 2× reaction buffer) was performed initially. As a first step in reaching this volume, it was determined that dilution of the Tn5 enzyme into 2× reaction buffer prior to addition to the DNA did not significantly affect the sequencing quality. The tagmentation reactions were also performed at a volume of 50 μL, 20 μL, and 10 μL, and no significant difference in sequence quality was observed due to reduction in the tagmentation reaction volume.
As an alternative strategy to overcome the pipetting inaccuracy of conventional liquid handlers (for volumes less than 2 μL), an acoustic liquid handling instrument designed to handle transfers in the nanoliter range was used for the next experiment. Using an acoustic transfer instrument, the tagmentation reaction was performed at 0.5 μL scale.
Early experiments showed that the tagmentation reagents could be used as a master mix and that 5 μL reactions gave sequence data quality equivalent to that obtained using the Nextera kit according to Illumina's protocol (50 μL tagmentation). This remained true upon further reduction of the reaction volume to 0.5 μL using the Echo acoustic liquid dispensing system (Labcyte, Sunnyvale, Calif.).
After tagmentation, the transposase remains tightly-bound to the DNA (Reznikoff et al. (2008) Annu. Rev. Genet. 42: 269-286) and can inhibit the initial strand-displacing extension required for the PCR. In the Illumina protocol, the tagmented DNA is purified away from the transposase using Zymo Clean and Concentrate columns, but this is impractical for a high throughput process. Thus, other dissociation conditions for removing transposases from nucleic acids were explored. Tagmented DNA fragments or a control reagent (PCR products with ends identical to tagmented fragments after end repair) were subjected to various treatments, and the efficiency of PCR amplification was compared to that using Zymo column purification.
Five treatment possibilities were explored: 1) dilution with TE buffer; 2) dilution with TE buffer and heat; 3) SDS and Triton; 4) high pH and neutralization; and 5) chaotropic salts+dilution. These treatments were compared to Zymo treated samples using a simple experimental system, which compared the post-PCR yield of either plasmid DNA that had been fragmented by Tn5 protein or linear DNA that was not exposed to Tn5 protein but was still flanked by the same terminal primer binding sites.
In the first two treatments, the following conditions were compared with the Zymo kit: 1) dilution with TE buffer; and 2) dilution with TE buffer and heat. Pooled tagmentation reactions were split between the three treatments. The Zymo samples were prepared according the Zymo kit protocol. Samples for the dilution treatments were diluted by adding 90 μL of TE to 10 μL of tagmentation reaction. Samples for the first treatment stayed at room temperature (25-27° C.) for 10 minutes while samples for the second treatment were incubated at 68° C. for 10 minutes. All samples were used in 10-cycle PCR reactions with a common pair of barcode primers and, after cleaning up PCR reaction products with Ampure beads to remove small DNA fragments, the cleaned up PCR reaction products were compared on an Agilent Bioanalyzer.
The results indicated that none of these treatments inhibited the PCR reaction, and the Zymo kit treatment produced the highest PCR yield. Amplification of the linear DNA, which tested for inhibition of the PCR reaction, was statistically indistinguishable for the three conditions (lowest P=0.07): 1) dilution of the tagmentation reaction mixture with TE yielded 0.80 times as much DNA as the Zymo kit; and 2) dilution of the tagmentation reaction mixture with TE and heat yielded 0.92 times as much as the Zymo kit (Data not shown). Amplification of the tagmented plasmid DNA (which tested removal of the Tn5 protein) revealed a doubling in DNA yield for each treatment from the worst treatment to the best treatment: 1) the dilution of the tagmentation reaction mixture with TE resulted in a DNA yield which is 0.28 times as much as that of the Zymo kit; and 2) the dilution of the tagmentation reaction mixture with TE and heat resulted in a DNA yield which is 0.53 times as much as that of the Zymo kit (Zymo kit=1±0.04X). While a simple treatment such as diluting the tagmentation reaction mixture with TE and heat provided 50% as much DNA as the Zymo kit, the better treatment conditions that can yield higher DNA yields were explored in the next set of experiments.
The third treatment explored was the addition of SDS to remove protein followed by addition of Triton X-100 (triton) to sequester the SDS. As before, pooled tagmentation reactions were split between different Tn5-removal treatments. A matrix of 24 SDS/triton treatments was prepared, where each sample received one of 6 different SDS solutions and one of 4 different triton solutions. The Zymo kit samples were processed according to the manufacturer's protocol. Non-Zymo reactions were incubated at 75° C. for 10 minutes after addition of SDS, amended with triton in TE, and mechanically shaken. All reactions were then used in identical PCR reactions and compared by Fragment Analyzer.
The experimental results of the third treatment are illustrated in
For the linear DNA (data not shown), the recovery of DNA increased slightly with lower concentrations of SDS: at 0% SDS, the DNA yield was 0.96 times as much as the Zymo treated sample; at 0.1% SDS, the DNA yield was 1.1 times as much as the Zymo treated sample; at 0.2% SDS, however, the DNA yield dropped to 0.1 times as much as the Zymo treatment sample, indicating PCR inhibition. The addition of triton after the SDS treatment ameliorated the inhibition of the PCR reaction even when the SDS concentrations were as high as 0.3%.
For the tagmented plasmid (
The fourth and fifth treatment conditions, high pH and guanidine isothiocyanate, also resulted in a reasonable amount of DNA recovery. These treatment conditions, however, did not improve recovery of DNA as compared to the SDS treatment. The fourth and fifth treatment conditions were not further explored as they may add operational challenges in some circumstances. As a note, it was discovered that samples incubated with guanidine isothiocyanate at room temperature had statistically indistinguishable recovery of DNA compared to samples incubated at a temperature of 68° C. This result indicated that heating samples, an operationally challenging step, was not necessary. As noted above, it was also later discovered that heating was unnecessary for the SDS treatment conditions for the maximum recovery of DNA.
After completing the five different treatment conditions, the treatment conditions with SDS were further explored. Experimental conditions were designed to further increase DNA recovery. In the designed experiments, a number of different conditions were varied: the SDS concentration was varied; the incubation temperature was varied; the sample was diluted to 50 μL instead of 100 μL to add twice as much DNA to the PCR reaction. The only sample that showed the reduced PCR efficiency was the one containing the highest amount of SDS (0.02% in the PCR). No adverse effect was found from the SDS concentration or dilution in any other samples. However, a large effect was found from the incubation temperature: Incubation at 75° C. returned, as before, 0.53 time as much as the Zymo treatment; incubation at 50° C. returned 0.87 times as much as the Zymo treatment; and incubation at 25° C. returned an average of 0.98 times as much as the Zymo treatment. Therefore, the following conditions were selected as optimum treatment conditions: 0.1% SDS and 25° C.
To verify that this modified sample preparation protocol resulted in high-quality sequence data, a set of 32 plasmids was treated three ways: 1) by Zymo kit; 2) with 0.1% SDS (final concentration); or 3) with 0.2% SDS (final concentration). Samples from all three treatments were uniquely barcoded but otherwise put through identical PCR reactions, purified, analyzed by Fragment Analyzer, normalized, pooled, and sequenced.
It was first verified that samples prepared with these new SDS-based conditions returned as much DNA after barcoding PCR reactions as samples prepared with the Zymo kit. The tagmented SDS-treated plasmid samples in this experiment (n=15) returned an average of 1501±169 ng while the average DNA returned for Zymo column samples (n=16) was 1412±206 ng.
As a note, it was discovered that the distribution of fragment sizes was significantly different between samples treated with SDS and with the Zymo kit. This is illustrated in FIGS. 7B1 through 7B3. FIGS. 7B1 through 7B3 show superimposed fragment analyzer traces of samples treated with 1) Zymo kit; 2) 0.2% SDS (final concentration); 3) 0.1% SDS (final concentration). All samples were incubated at room temperature. The DNA fragment size is shown along the horizontal axis, and the DNA concentration is shown along the vertical axis (RUF=relative fluorescence units). The DNA treated with the Zymo kit was broadly distributed between roughly 400 base pairs and 2000 base pairs (FIG. 7B1). The DNA samples treated with SDS had less than 25% of their DNA mass below 600 base pairs, and the majority in a large peak centered around 2000 base pairs (FIG. 7B3). Because the sequencing process favors molecules in the 300-800 base pair range, it was found that this altered distribution may necessitate adjusting the PCR extension time to favor smaller fragments as well as revising the normalization and dilution calculations so that the same number of sequenceable DNA fragments reaches the sequencer regardless of the shape of the distribution.
The sequence data revealed two groups of statistically significant differences between Zymo-treated and SDS-treated samples. The first group of results is rooted in the insert size. The Zymo-treated samples contained, on average, a larger fraction of fragments that were smaller than 150 base pairs. Because these small fragments are informatically discarded, the final sequence metrics are strongly affected. The second group of results related to how evenly sequence data is distributed across the plasmids. Surprisingly, it was discovered that coverage was significantly more evenly distributed across SDS-treated samples than across Zymo-treated samples (P<0.0001). Specifically, the coefficient of variation (CV) of sequence depth was 25% for Zymo-treated samples but 20% and 18% for the 0.2% and 0.1% SDS-treated samples, respectively. This unexpected difference is valuable because it will allow increased plexity; the reduced variability will in turn decrease the average coverage required to meet the sequence quality specification. Thus, while other dissociation conditions can be used to remove transposases from DNA, the addition of SDS to a final concentration of 0.1% was found to be most effective at removing the transposase without interfering with the subsequent PCR. This discovery and other suitable treatment conditions led to elimination of the cost-prohibitive column spin step during sample preparation for sequencing in certain embodiments.
Unique barcodes can be added to every DNA fragment at one or both ends. The specific sequence or “index” used as a barcode sequence is unrestricted, though the field has established a precedent of 8-bp indices. Each index can be used for either of the two ends, which have slightly different sequences added by the Tn5 protein and are referred to as the i5 and i7 ends.
To enable the required level of multiplexing, a set of barcode adapter primers was designed using previously described algorithms (Bystrykh (2012) PLoS One 7: e36852; Frank (2009) BMC Bioinformatics 10: 362). The structure of the i5 and i7 index primers was maintained, but in order to reach higher plexity, a novel set of 826 8-base pair candidate indices were identified using the following criteria: (1) no index contained a homopolymer run of 3 base pairs or more; (2) every candidate index has a Hamming distance of three or more from all other indices; and (3) every candidate has a Hamming distance of three or more from every eight base segment of the conserved sections of the i5 and i7 sequence. These candidate indices were then used to generate the corresponding candidate i5 and i7 barcode primers. From all possible 8-base sequences generated, those with mononucleotide runs longer than two bases or GC content outside the range of 35% to 65% were removed. The following sets of sequences were then selected: sequences differing by at least three bases from all other barcodes in the set, or from sequences complementary to all 8-base sequences present within the conserved regions of the i5 and i7 adapter primers. These sequences (approximately 800) were then placed into the context of the full-length Illumina adapter primer, and the resulting adapter primers were analyzed using DINAMelt (Markham (2005) Nucleic Acids Res. 33: W577-581) to predict the stability (Gibbs free energy) of each folded polynucleotide. In other words, the resulting adapter primers were examined to find those with the lowest predicted tendency to form inter- or intra-molecular duplexes.
Table 2 lists the set of barcode sequences generated by the method described above. These barcode sequences were custom ordered from Integrated DNA Technologies, and were used in highly multiplexed sequencing experiments.
For the experiments shown in
The performance of Vent-based master mix according to the present invention was compared to the Illumina Nextera PCR Mastermix (NPM). It was found that there were two differences. The first difference was that NPM samples tend to have a larger fraction of DNA smaller than 400 base pairs while Vent samples tend to have a larger fraction between 500 base pairs and 1000 bp (P=0.025). The second difference was that NPM samples had roughly double the DNA concentration of Vent samples. (Data not shown). A two-fold difference after 8 cycles suggests that, in each cycle, NPM is 10% more efficient than Vent (i.e., 1.18=2.1). Further experiments showed that this difference in DNA yield could be ameliorated by adding one or two PCR cycles to reactions using Vent polymerase.
It was also found that the concentration of barcode primer also had a large effect on the DNA yield for Vent-based master mix. Experiments that used Vent-based master mix and 0.1 μM barcode primer yielded less than 5% as much as the equivalent NPM reaction (data not shown). When barcode primers were used at or above 0.5 μM, the DNA yield of Vent-based master mix reached a plateau of 45% as much as the equivalent NPM reaction. The yield of NPM reactions remained unchanged across this concentration range (data not shown). It was found that there was no statistical difference in DNA yield or in the fragment size distribution between NPM reactions using the Illumina barcode primers and NPM reactions using the barcode primers according to the present invention.
It was tested whether the Vent-based PCR master mix would adversely affect sequence quality by preparing and sequencing a set of 42 recently-constructed plasmids using either NPM or Vent and using both presently designed and Illumina-provided barcode primers. Because of the difference in polymerase efficiency, NPM samples were given 8 cycles of PCR, and Vent samples were given 10 cycles of PCR. No statistically significant difference was found in any of the sequence quality metrics, including the number or quality of mutations identified, between samples prepared with NPM and sample prepared with the Vent-based master mix. Similarly, the origin of barcode primer resulted in no statistically significant difference in any sequence quality metric. Based on this data, it was concluded that the Vent-based master mix according to the present invention performs at least as well as a commercially available alternative, Illumina NPM, as long as additional PCR cycles compensate for the lower DNA yield.
For preparing plasmid DNA, rolling circle amplification (RCA) takes less than a third the hands-on time and produces more consistent final DNA concentrations compared to plasmid minipreps (Dean et al. (2001) Genome Res 11: 1095-1099). In particular, rolling circle amplification (RCA) of plasmids using Phi29 polymerase generates large amounts of linear high molecular weight concatamers of the plasmid. This is a much less labor intensive way to obtain DNA than plasmid minipreps, which involve multiple centrifugation steps. Furthermore, RCA gives good Sanger sequence data (Dean et al. (2001) Genome Res 11: 1095-1099), good restriction digest banding (Dharmadi et al. (2014) Nucleic Acids Res. 42: e22), and whole genome-amplified DNA provides good Illumina sequence data (Indap et al. (2013) BMC Genomics 14: 468).
A set of 384 DNA assemblies ranging in size from 4 kb to 20 kb were used to prepare both RCA DNA and plasmid DNA, and the 768 DNA samples were used to prepare a pool of 768 Nextera libraries for the MiSeq.
Although the average depth of coverage for the 768 samples spanned over three orders of magnitude and displayed wide statistical variation (
Since tagmentation reaction involves combining the DNA template with the Tn5 enzyme at a relatively precise protein to DNA ratio, the Echo acoustic liquid transfer system was considered for diluting the RCA preps to 2.5 ng/μl. However, since normalizing DNA concentration for each sample individually for many samples is time and labor intensive, other options were explored for this step. After quantifying RCA DNA using PicoGreen, the BiomekFX robot was used to normalize DNA. This normalization process took about an hour for 4 plates. The normalized DNA was then used on the Echo to set up our tagmentation reactions. In parallel, one of the four plates was taken, and the DNA was uniformly diluted to the same volume (e.g., 5 μL of DNA to 35 μL water) across all samples on the plate. This method was chosen because the DNA generated by RCA tends to be relatively constant in concentration, more so than DNA prepared by minipreps. From the calculations of how much DNA was to be added to water using the BiomekFX robot, the ratio of 5 μL DNA to 35 μL water was the average dilution required for that plate in some implementations.
For a robust QC process, the samples should receive similar average read coverage and few should have less than 15× coverage. To achieve this, each sample in the pool should have a similar molar concentration of sequenceable fragments such that each forms a similar number of clusters on the MiSeq flow cell. When the same pool of Nextera libraries derived from the same set of plasmid constructs was sequenced in separate MiSeq runs, coverage was highly correlated between the runs (
The large deviation in average coverage across the sample population in
Samples at the edges of a plate sometimes had low concentrations, which were thought to be due to droplets veering to the sides such that reagents were not completely mixed at the bottom of wells. To mitigate this, plates were centrifuged at 1,000 g immediately after dispensing on the Echo in some implementations. Also, the entire volume of any sample with a low concentration was decided to be added to the pool, because such samples then had a chance of receiving coverage without significantly affecting the coverage of other samples.
The protocol changes discussed above were implemented for the parallel sequencing of 4078 plasmids.
In the above QC of 4078 plasmids, the consumables cost was $2.68 at present value per MiSeq sample, which breaks down as shown in Table 3.
Although this is almost $11 per assembly at present day value (because four replicates of each are sequenced), achieving only 1× coverage by Sanger sequencing of this same set of DNA assemblies would be about 10-fold more expensive and would include the need to order and track many primers to distribute the reads across the assemblies appropriately.
Aligning reads to a digital reference and choosing the best replicate of an assembly is conceptually simple, but requires rapid, parallel analysis of many datasets. The SAMTOOLS and BCFTOOLS (Ramirez-Gonzalesz et al. (2012) Source Code Biol. Med. 7:6) were initially tested to identify single-nucleotide polymorphism (SNPs) and indels, but it was difficult to find appropriate settings to reliably call all mutations found in the plasmids. A possible cause for this could be the high read coverage seen in some samples (approaching 1000×), which may hinder some part of the mutation calling algorithm. Subsampling the sequencing data in these cases would not be ideal as this reduces resolution of SNP frequency and complicates base calling in regions of low coverage. Another possible cause is that the DNA samples may be mixed populations that do not resemble the diploid genomic samples against which these algorithms and tool sets were developed. For example, a SNP at 10% frequency does not match a heterozygous or homozygous situation. Interestingly, it was found that the features were identified correctly at the level of read alignment but sometimes missed by the calling algorithms.
Given the small size of the plasmids that were sequencing (compared to genomes), in certain embodiments of the present invention, a simple feature detection method was implemented based on the pileup file. Software was written in F# (fsharp.org) to call mutations and assign severity scores to features (e.g., SNPs and indels) based on their sequence context (e.g., part type and the probability that they could impair function). The software ranks the replicates of each assembly based on the number of mutations and their severity and reports which replicate best matches the digital template. In addition, the software stores all sequence variants found, along with other relevant information, in a postgreSQL database.
Finally, the software generates a graphic for each sample (
In the run with 4078 samples described herein, 4056 were four replicates of 1014 constructs assembled by yeast homologous recombination. The remaining 22 samples were internal process controls, which were not used for data analysis. Table 4 shows the statistics for the sequence differences between the samples and the digital reference sequences.
The importance of replicates is highlighted by the fact that although 5.8% of the samples were misassembled, only 1% of the constructs had no correctly assembled replicate.
When a SNP or indel is present in only one replicate of a construct, this is likely due to errors in the primers or errors by the polymerase during PCR amplification of parts. Alternatively, errors may arise during RCA for MiSeq sample preparation. The frequency of this type of mutation appears consistent with the known fidelity of the polymerases (McInerney et al. (2014) Mol. Biol. Int. 2014: 287430), or with the reported frequency of errors in oligonucleotide primers (Hecker and Rill (1998) Biotechniques 24: 256-260). Many indels were located at homopolymers, which are known to be susceptible to contraction during replication and are also prone to sequencing artefacts even on the Illumina platform. When the same SNPs or indels are present in all four replicates, or in the same part in different constructs, they are most likely due to errors in either the digital reference sequence (i.e. data entry) or the template used for PCR amplification of the part. Several errors were due to the use of a physical part for the PCR template that was not the same as the part specified in the digital request. The frequency of this type of mutation was higher than anticipated, which can be further reduced. Since the run with 4078 samples described here, this NGS QC process has been used in more than ten assembly cycles, thus accumulating a large amount of NGS QC data. A comprehensive analysis of this data can be used to identify how the assembly process generates the different types of mutations, which can illustrate areas of improvement for the DNA assemblies.
All liquid transfers are accomplished using automation. All transfers less than 2 μL were accomplished using the Echo and all transfers greater than 2 μL were accomplished using a BiomekFX or NX.
It should be appreciated that the specific steps illustrated in the exemplary protocol provides a particular method of preparing plasmids. Other sequences of steps may be performed according to alternative embodiments. For example, alternative embodiments of the present invention may perform the steps outlined above as multiple sub-steps as appropriate to the individual step. Furthermore, additional steps may be added or removed depending on the particular applications. For example, step 9) of quantifying DNA concentration using PicoGreen assay can be omitted. In another example, the DNA samples can be pooled without normalizing the concentration in step 10).
One or more features from any embodiment described herein may be combined with one or more features of any other embodiment without departing from the scope of the invention.
All publications, patents and patent applications cited in this specification are incorporated herein by reference as if each individual publication or patent application were specifically and individually indicated to be incorporated by reference. Although the foregoing invention has been described in some detail by way of illustration and example for purposes of clarity of understanding, it will be readily apparent to those of ordinary skill in the art in light of the teachings of this invention that certain changes and modifications may be made thereto without departing from the spirit or scope of the appended claims.
This application claims priority to U.S. Provisional Patent Application No. 62/088,416 filed Dec. 5, 2014 and U.S. Provisional Patent Application No. 62/144,174, filed Apr. 7, 2015, which are incorporated herein by reference.
This invention was made with Government support under Agreement HR0011-12-3-0006, awarded by DARPA. The Government has certain rights in the invention.
| Filing Document | Filing Date | Country | Kind |
|---|---|---|---|
| PCT/US2015/064029 | 12/4/2015 | WO | 00 |
| Number | Date | Country | |
|---|---|---|---|
| 62088416 | Dec 2014 | US | |
| 62144174 | Apr 2015 | US |
