MotifNetwork: Genome-Wide Domain Analysis using Grid-enabled Workflows

It has been recently demonstrated that the understanding of gene interactions, gene function and ultimately gene and organism evolutionary processes, is facilitated by determination of relationships among the constituent "domains" and "motifs." This understanding, however, comes at the expense of substantial computation necessary to support the required data mining. The detailed computations behind a typical domain and motif analysis are described in our recently developed grid-enabled environment called the MotifNetwork. In this work, the MotifNetwork is applied, for the first time, to large genome-wide association analysis of the entire Apis mellifera (honeybee) genome. Several impediments to this computation are identified including weaknesses in workflow creation and enactment capabilities and limitations in hardware capacity. Nonetheless, though generated in a non-optimum fashion, preliminary results of the honeybee analysis are presented. Lastly, suggestions for future work are summarized.