Monday, July 25, 2011
Grand Ballroom, 5th fl (Sheraton New Orleans)
Extracting knowledge from increasingly large quantities of metabolic data requires novel computational tools that enable automated discovery and analysis of biologically relevant metabolic pathways. Manually navigating metabolic data is not highly scalable and often inhibits the ability to find novel or non-standard pathways. Identifying these pathways are important in applications such as metabolic engineering, understanding the metabolic scope of multi-species communities, and identifying redundant metabolic pathways. To help automate metabolic pathway discovery, we have developed and validated new graph-based search algorithms for finding linear and branched metabolic pathways in multi-genome scale data. Our results demonstrate the algorithms are able to find biologically relevant metabolic pathways by explicitly tracking atoms through the metabolic network. We have made our algorithms easily available through a web server located at <http://atommetanetweb.kavrakilab.org/>http://atommetanetweb.kavrakilab.org/. The basic query only requires a start compound and a target compound. A set of metabolic pathways, which try to maximize the number of carbons from the start compound to the target compound, are returned. The default setting is to rank the pathways first by the number of carbons they conserve and second by the number of reactions they contain. The pathways are drawn in the web browser and the user can interact with them to obtain more information about the compounds and reactions. A number of other options, such as setting the number of carbons to conserve or other ranking criteria, are provided for more advanced queries. We welcome feedback on how to make the tools provided more useful to the community.