The major challenge of the post genomic biology is to gain knowledge on the biological function of gene products and the mechanisms of the cellular processes in which they are involved. An important first step in this endeavor is to derive the circuitry of the cellular systems: the networks of physical, biochemical and regulatory interactions, as well as the information processing (signal transduction) pathways that occur in cells. Large efforts are therefore devoted world-wide to collecting genome scale data on these interactions and pathways, using experimental methods, bioinformatics inference procedures, and text mining algorithms. As a result, impressive bodies of valuable but noisy data are being produced. But consolidating and validating these data, and more importantly, deriving meaningful biological information from them poses major challenges.
Our contribution towards addressing these challenges has been to develop a unified conceptual model for representing information on biochemical pathways (metabolic, regulatory, transport, signal transduction). This conceptual model has been implemented in the aMAZE and related databases BioMaze, TransMaze, and has served as an inspiration for the BioPax standard that is now being adopted by a number of specialized databases storing information on metabolic and other types of pathways. To find out more see:
-
14582519
An overview of data models for the analysis of biochemical pathways.Briefing in Bioinformatics, 4, 246-259
More recently, we developed methods for inferring metabolic pathways from information on the universal metabolic network stored in the KEGG database, which comprises all the chemical reactions and enzymes believed to occur in genomes sequenced to date:
-
16337962
Inferring meaningful pathways in weighted metabolic networks.J Mol Biol. Feb 10;356(1):222-36.
Currently we are developing a workbench for displaying and analyzing metabolic and other types of networks using Cytoscape. These include efficient means for interactively mapping onto these networks information on protein sequences, domains and 3D structures, as well as data on drug targets, and genes associated with human diseases. Presently, we use such mappings to investigate the relations between local network properties, orthology relationships and protein/gene function.
We are also developing chemoinformatics tools for the quantitative analysis of chemical reactions and for representing ligand and metabolite chemistry and 3D conformations.
Our longer-term aim is to produce a versatile environment enabling complexity reduction through quantitative analyses of biochemical networks, as well as ready access to information on these networks available from multiple sources.
Researcher and Co-op students contributing to these projects:
- Brian Turner
- Shuye Pu
- Steve Constable
- Zahrah Khalid
- Karen Ronan
