The Use of Phylogenetic Profiles for Gene Predictions Revisited

Determining gene functions from genomic sequences is a central goal of bioinformatics. Traditionally computational approaches to this problem are based on the detection of genes with homologous sequences. With the completion of fully sequenced genomes alternative approaches have become feasible. One such method is that of phylogenetic profiles. In this method a gene is described by its phylogenetic profile, i.e. a string that encodes the presence or absence of a homologous gene in other genomes. This string is then used to search for other genes with similar profiles. In this paper we briefly review the field, including extensions to the original method. We also discuss variations on this theme including inverse phylogenetic profiles and non-exact profiles using phylogenetic trees. In conclusion our work indicates that phylogenetic profiles might be useful for some, but not all functional annotations. Functional annotation of genomes remains an important problem in genomics when no close homologs exist.