SIFTER search: a web server for accurate phylogeny-based protein function prediction.
Bottom Line: As only a minuscule fraction of these have been experimentally characterized, computational methods are widely used for automated annotation.The SIFTER web server thus provides access to precomputed predictions on 16 863 537 proteins from 232 403 species.Users can explore SIFTER predictions with queries for proteins, species, functions, and homologs of sequences not in the precomputed prediction set.
Affiliation: Department of Plant and Microbial Biology, University of California, Berkeley, CA 94720, USA.Show MeSH
Mentions: In this option, users can submit the sequences of proteins not yet in our precomputed prediction set, and the web server will show the predictions for the closest homologs in our prediction set (found based on top BLAST hits obtained from the NCBI-BLAST website (33)). The output includes the statistics of the significance of the hits along with the SIFTER predictions for each of the hits to the given input sequence. For each search, users can enter up to 10 protein sequences. Figure 4 illustrates sample output for a queried sequence. While this option embodies limitations of BLAST, often there is such a close homolog with a precomputed SIFTER prediction that the errors are limited. By contrast, the distance to the closest experimentally annotated protein is often so far as to introduce significant errors.
Affiliation: Department of Plant and Microbial Biology, University of California, Berkeley, CA 94720, USA.