SIB-BLAST (Simple Is Beautiful) is a novel algorithm developed to overcome the model corruption problem that occurs frequently in the later iterations of PSI-BLAST searches.The algorithm compares resultant hits from iteration two and the final iteration of a PSI-BLAST search, calculates the figure of merit for each “overlapped” hit and re-ranks the hits according to their figure of merit. The premise of the algorithm is based on the observation that the profile, namely, the position specific scoring matrix (PSSM), in the first two rounds of a PSI-BLAST search, is the least corrupted since it is comprised mostly of close homologs. These profiles are used to search for more distant homologs, which are used to generate subsequent PSSMs. As more distant homologs are incorporated into the PSSM, non-homologous sequences frequently get included also, thus leading to model corruption. Hence, “benchmarking” hits from later iteration against earlier round when the model is least corrupted should improve the accuracy of a PSI-BLAST search.
:: MORE INFORMATION
M.M. Lee, M. Chan, and R. Bundschuh,
“Simple is beautiful: a straightforward approach to improve the delineation of true and false positives in PSI-BLAST searches“,
Bioinformatics 24 (2008) 1339-1343.