Plant & Animal Genome VII Conference

S14

WHAT IS SO SPECIAL ABOUT PLANTS: ANALYSIS OF PROTEIN SEQUENCES ENCODED BY ARABIDOPSIS AND BY COMPLETELY SEQUENCED GENOMES

Department of Computational Biology, Axys Pharmaceuticals, Inc., 11099 N. Torrey Pines Rd., La Jolla, CA 92037 USA

Putative proteins encoded by contigs that have been sequenced in the course of Arabidopsis genome project were analysed using iterative sequence similarity search, feature prediction, and domain dissection. Arabidopsis proteins are highly conserved in evolution, with ~70% of proteins matching sequences from taxa other than higher plants. The annotation framework based on natural taxonomy of proteins was applied to Arabidopsis proteins, allowing completely automated and highly accurate assignment of function for ~35% of these matches. For almost equal number of matches, at least general biochemical function could be predicted by interactive analysis. Small number of exon misassemblies was detected as well as relatively large number of misannotations (errors of filtering, errors of domain assignment, and errors of ortholog definition). Taxonomic distribution of the plant-related sequences in the database was analysed. Of particular interest is the comparison to Synechocystis sp. Among the Arabidopsis proteins that have a bacterial ortholog, about 50% have a Synechocystis sequence as the best bacterial match. A fraction of those has no established relation to the chloroplast function, and may represent displacement of pivotal cellular functions by endosymbiont-derived genes. Interestingly, high similarities between plant and Synechocystis include such quintessential signal transduction domains as serine-threonine protein kinases and Toll-like motifs. Protein families that distinguish plants from animals and fungi, including those truly unique to plants, as well as dramatically expanded in plants as compared to other phylae of life, were analyzed and will be discussed in more detail.