Curation is forever: comparative genomics approaches to functional annotation

Abstract

The process of deriving accurate annotations for genes in newly sequenced genomes has up to now been based largely on projection of results in one organism to genes in other organisms based on sequence similarity. We are now entering a period in which metabolic reconstructions will play an increasingly significant role. These reconstructions will support a gradual refinement of estimates of function, leading to more consistent and precise understanding. They will become key components in establishing which functional roles have not yet been connected to genes and which aspects of an organism's physiology can be understood in terms of existing functional assignments. As each refinement is made, we incrementally approach the goal of establishing a qualitative understanding of an organism in terms of its inventory of genes.