NSF Award
9630339
This project will develop a data management system to support a variety of data of importance to molecular biologists interested in groups of related proteins. The system will be distributed by the San Diego Supercomputer Center (SDSC) as part of an existing biological data resource supporting, among other things, data on the protein kinase family of enzymes (http://www.sdsc.edu/kinases) and temporal 3-dimensional structure data derived using X-ray crystallography. The system will permit queries integrating sequence homology, structure homology, function, and disease in a way not possible with the current generation of molecular biology resources. The overall goal is to provide to the community a non-proprietary system that is efficient, extensible and, for the molecular biologist, simple to use for loading their own data. These data could be subsets of data available from existing sources and/or their own private data. The system, will be coded in Java (or language of equivalent functionality), will run on a local computer or in a client-server environment using a local area network or the Internet. Specific deliverable are: (I) compact binary data storage, object oriented in-memory representation of macromolecular features, and query methodology suitable for analyzing comparative features found in proteins. This is an extension of technology already found in the software WPDB (http://www.sdsc.edu/CCMS/Packages/wpdb.html) and MOOSE (http://www.sdsc.edu/moose). (II) A client-server environment which is beyond the current stateless access provided by the World Wide Web. (III) Porting of existing methods written in C++ to Java and the creation of new methods providing an applet library for use in molecular biology. (IV) A low-level graphics library written in Java specifically for use in molecular biology. (V) A molecular biology viewer (i.e., a domain-specific browser, editor and interpreter). This viewer will be built from the graphics library and provide th e features of the current generalized stateless browsers, but also respond to specific types of data and methods peculiar to molecular biology. The interoperability between display objects, used successfully in other software packages (WPDB and PDBtool, http://www.cse.ogi.edu/DlSC/PDBTool/) will provide the appropriate level of intuitive interaction. The proposed data management system addresses limitations perceived in the current generation of World Wide Web-based servers supporting molecular biology and many other disciplines. Notable limitations are: (i) the difficulty of keeping servers current; (ii) the information content is broad but not deep; (iii) tools for visualizing results of Web-based query are limited and hampered by the stateless nature of the interaction and the disproportional utilization of computing resources between client and server; (iv) tools for the Web-based comparative analysis of proteins could be expanded; (v) there are few generic tools for the molecular biologist to use in building and maintaining this type of resource for either local or global use.
