DMCA
Research Project for MMath The Use of Vector Seeds to Improve PSI-BLAST Sensitivity by
BibTeX
@MISC{Zhang_researchproject,
author = {Fengkai Zhang and Supervisor Prof and Daniel Brown},
title = {Research Project for MMath The Use of Vector Seeds to Improve PSI-BLAST Sensitivity by},
year = {}
}
Share
 |
 |
 |
 |
||
OpenURL
Abstract
PSI-BLAST [5] is widely used for searching protein databases for sequence similarities, especially distant homologies. Position specific score matrices are constructed during its running. These give the program the power to capture remote relations of a query sequence. PSI-BLAST needs multiple iterations in most circumstances and is time-consuming. Here, we modified the vector seed optimizing algorithm introduced by Brejova et al. [7] to apply to position dependent scoring systems. Our algorithm successfully identifies some vector seeds which are significantly better than the default seed used in PSI-BLAST. Our finding suggests that PSI-BLAST may be improved by using new seeds, thereby reducing its number of iterations. ii
