Protein Misprediction Revealed by New Technique

A new bioinformatics tool is capable of identifying and correcting abnormal, incomplete, and mispredicted protein annotations in public databases. The MisPred tool currently utilizes five principles to identify suspect proteins that are likely to be abnormal or mispredicted.

Dr. Laszlo Patthy developed this new approach with colleagues from the Institute of Enzymology of the Hungarian Academy of Sciences (Budapest, Hungary). He explained how necessary it is, "Recent studies have shown that a significant proportion of eukaryotic genes are mispredicted at the transcript level. As the MisPred routines are able to detect many of these errors, and may aid in their correction, we suggest that it may significantly improve the quality of protein sequence data based on gene predictions.” The MisPred technique has the potential to save much time and effort that would otherwise be spent in further investigation of erroneously identified genes.

The MisPred approach rates annotations according to five principals: (1) extracellular or transmembrane proteins must have appropriate secretory signals. (2) A protein with intra- and extra-cellular parts must have a transmembrane segment. (3) Extracellular and nuclear domains must not occur in a single protein. (4) The number of amino acid residues in closely related members of a globular domain family must fall into a comparatively narrow range. (5) A protein must be encoded by exons located on a single chromosome.

There are some exceptions to these rules, as pointed out by Dr. Patthy, "Some secreted proteins may truly lack secretory signal peptides since they are subject to leaderless protein secretion. Similarly, it cannot be excluded at present that transchromosomal chimeras can be formed and may have normal physiological functions. Nevertheless, the fact that MisPred analyses of protein sequences of the Swiss-Prot database identified very few such exceptions indicates that the rules of MisPred are generally valid.”

The investigators discovered that the absence of expected signal peptides and violation of domain integrity account for the majority of mispredictions. The researchers noted that "Interestingly, even the manually curated UniProtKB/Swiss-Prot dataset is contaminated with mispredicted or abnormal proteins, although to a much lesser extent than UniProtKB/TrEMBL or the EnsEMBL- or GNOMON-predicted entries.”

The study was described in August 2008 in the open access journal BMC Bioinformatics.

Related Links:
Institute of Enzymology of the Hungarian Academy of Sciences