G-BLAST: a Grid-Based Solution of mpiBLAST on Computational Grids

Abstract

Research in the area of bioinformatics has grown with each passing day in recent years as demands for more computing power increased. The solutions to these demands usually involve using parallelism techniques. For example, mpiBLAST is a parallel program that executes jobs in parallel in clusters with MPI. Although Cluster environments can reduce the execution time and increase alignment efficiency, they may not be a good solution when aligning very large genomic databases. Grid Computing can coordinate the resources of distributed virtual organizations and satisfy a great many computational demands. In addition to integrating distributed resources, Grid Computing can reduce server idle times via management of integrated computing resources. This study proposes an approach that integrates mpiBLAST in a Grid Computing environment called BioGrid, and implementation of a Grid service called G-BLAST. In here, the G-BLAST is compliant with the Web Service Resource Framework (WSRF) standard.