A Hybrid Grid and Its Application to Orthologous Groups Clustering

Abstract

AbstractOrthologous groups are useful in the genome annotation, studies on gene evolution, and comparative genomics. However, the construction of orthologous groups is difficult to automate and takes so much time as the number of genome sequences increases. Furthermore, it is not easy to guarantee the accuracy of the automatically constructed orthologous groups. We propose an automatic orthologous group construction system for a large number of genomes. A hybrid grid computer system, consisting of 40 PCs, has been devised for fast construction of the orthologous groups from large number of genome sequences. The grid system constructs orthologous groups for 89 complete prokaryotes genomes just in a week (it takes 8 months on a single computer system). Furthermore, the system provides good extensibility for adopting new genomes in the existing orthologous groups. In the real experiment of the orthologous group constructions, more than 85% of the constructed orthologous groups coincide with those of KO (KEGG Ortholog) and COGs (Clusters of Orthologous Group of Proteins). Note that KO and COGs have been constructed manually or semi-automatically at the sacrifice of the extensibility for newly completed genomes.KeywordsGrid ComputingOrthologous GroupMaster NodeOrthologous ClusterPropose Cluster AlgorithmThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.