Effective function annotation through catalytic residue conservation

Abstract

Because of the extreme impact of genome sequencing projects, protein sequences without accompanying experimental data now dominate public databases. Homology searches, by providing an opportunity to transfer functional information between related proteins, have become the de facto way to address this. Although a single, well annotated, close relationship will often facilitate sufficient annotation, this situation is not always the case, particularly if mutations are present in important functional residues. When only distant relationships are available, the transfer of function information is more tenuous, and the likelihood of encountering several well annotated proteins with different functions is increased. The consequence for a researcher is a range of candidate functions with little way of knowing which, if any, are correct. Here, we address the problem directly by introducing a computational approach to accurately identify and segregate related proteins into those with a functional similarity and those where function differs. This approach should find a wide range of applications, including the interpretation of genomics͞proteomics data and the prioritization of targets for high-throughput structure determination. The method is generic, but here we concentrate on enzymes and apply high-quality catalytic site data. In addition to providing a series of comprehensive benchmarks to show the overall performance of our approach, we illustrate its utility with specific examples that include the correct identification of haptoglobin as a nonenzymatic relative of trypsin, discrimination of acid-D-amino acid ligases from a much larger ligase pool, and the successful annotation of BioH, a structural genomics target. enzymes ͉ function prediction ͉ EC ͉ PSI-BLASTA ssigning function to protein sequences continues to be of key importance (1). Currently, most approaches to protein function prediction rely on searching sequence databases to identify homologous sequences with prior annotation. The most widely used search tools are BLAST and PSI-BLAST (2); at the National Center for Biotechnology Information alone, Ͼ70,000 BLAST searches are performed each day for the general public. It is certainly no coincidence that the BLAST algorithm was the most highly cited paper of the last decade, surpassing all biology publications (3). PSI-BLAST is an iterative method that uses results from a BLAST search to create a profile (position-specific scoring matrix). The profile is used to search the database for additional homologues, and these results can be used to further improve the profile. A profile captures family-specific information, including functionally and structurally important residue positions, and can therefore identify distant homologues not recognized by alignment to a single sequence.PSI-BLAST and powerful fold recognition methods such as GENTHREADER (4) are good at identifying distantly related proteins, even down to 10% sequence identity. However, recent studies have shown that, simply on the basis of overall simi...

Local Structure Comparison of Proteins

“…As expressed by George and his coauthors [26], correct function prediction can significantly simplify and decrease the time needed for experimental validation. However incorrect assignments may mislead experimental design and waste resources.…”

Protein sequences insight into heavy metal tolerance in Cronobacter sakazakii BAA-894 encoded by plasmid pESA3

“…Consequently, the detection of functional residues in proteins is important in functional annotation (George et al 2005;Livingston and Barton 1996;Ouzounis et al 1998;Pupko et al 2002). In recent years, the computational studies have been used to predict metal binding motifs and associated genes, along with the binding motif responsible for metals and/or heavy metals interactions (Thilakaraj et al 2007).…”

Functional site prediction by exploiting correlations between labels of interacting residues

“…Though this method holds well in theory, it doesn't perform well in practice. One of the main disadvantages of this technique is that, even after selecting related sequences, which by itself is a herculean task, the conservation method returns too many false positives [9]. Even though various methods have been developed in the past for addressing the problem, they struggle to achieve substantial performance.…”