Intelligent worm gene function prediction with GeneMANIA

The ability to collect functional data, like physical and genetic interactions and co-expression, about every gene in the genome is expanding the possibilities of biological research. However, navigating through these data can be a difficult task for those without specialized computational expertise. GeneMANIA uses many large publicly available datasets to analyze gene lists and identify highthroughput data sources in which the query genes are highly associated. GeneMANIA is useful for generating hypotheses about gene function and planning mediumscale phenotypic screens. GeneMANIA will extend your gene list with more genes that are functionally associated with those in the list. The genes returned by GeneMANIA often represent promising candidates for functional assays (Mostafavi et al., 2008). It is designed to adapt to your query, so if your query list consists of members of a protein complex, GeneMANIA will retrieve more potential members of the complex. If members of your gene list function in a particular biological process (like neuronal development), GeneMANIA will return more putative contributors to that process. Using the software is easy just choose worm, enter your set of favorite genes and press Go. Example gene lists are provided. GeneMANIA currently supports standard gene symbols (from WormBase) and Entrez, Ensembl, Uniprot/SwissProt and RefSeq database identifiers. The advanced options panel can be used to choose additional networks to search, change the network weighting method, and increase the number of genes to retrieve. You can choose from among many different networks to be used in the analysis, although the default networks are reasonable for most queries. GeneMANIA currently supports 76 networks for C. elegans, including 10 co-expression, 8 physical interaction, 4 genetic interaction, 1 co-localization, 2 shared protein domain and 50 predicted networks. The predicted networks largely consist of WormNet (Lee et al., submitted) and I2D (Brown and Jurisica, 2007) data. GeneMANIA displays results as an interactive network, illustrating the functional relatedness of the query and retrieved genes. The links represent datasets in which genes are associated. The tabbed browser on the right side of the page can display information about the datasets or the genes in the network. Genes can be highlighted and moved around. The network visualization menu provides a range of options, including the ability to save the analysis. A complete report of the analysis can be generated, including a publication quality image of the network. GeneMANIA is free, web-based, open source and still a work in progress. We welcome any comments, suggestions and collaborations.