A Global Gene Family Classification Resource for Plants and Its Utility for Comparative Genomics, Genome Annotation, and Gene Family Studies

As diverse plant genomes and transcriptomes are sequenced at a rapidly increasing pace, there are exciting new challenges and opportunities for increasingly comprehensive comparative genomic analyses. To meet the needs of an expanding plant genomics community, we have developed an objective gene family classification that leverages evolutionary diversity across land plants. This classification has been constructed using the predicted protein sequences from 22 land plants selected for their phylogenetic representation, annotation quality, and characteristics of their evolutionary history, such as substitution rates and polyploidy. Our primary classification circumscribes deeply conserved, but narrowly defined gene lineages, on top of which we have layered a hierarchical higher-level classification using multiple stringencies and a domain-based functional annotation. We have extensively validated the quality of this classification and have extended its utility by providing pipelines to automatically sort new genome or transcriptome sequences into the classification and compute alignments and phylogenetic trees. Our database and analysis pipelines have broad utility within the plant biology community and I will highlight a few case studies leveraging this framework. We have identified conserved single copy genes useful in species tree inference and have reconstructed the ancestral gene content in seed plants and angiosperms. We have also used this classification to identify and correct systematic errors in plant genome annotations. Together with iPlant we are using this classification to sort genes from the One Thousand Plant Transcriptomes (1KP) project into gene families for phylogenomic analysis and large scale studies of gene family evolution. We are working with iPlant to make our gene family classification database and sorting pipelines broadly accessible to the Plant Science community through web services and the iPlant Discovery Environment.