IGLOSS: iterative gapless local similarity search

Abstract

Searching for local sequence patterns is one of the basic tasks in bioinformatics. Sequence patterns might have structural, functional or some other relevance, and numerous methods have been developed to detect and analyze them. These methods often depend on the wealth of information already collected. The explosion in the number of newly available sequences calls for novel methods to explore local sequence similarity. We have developed a high sensitivity web-based iterative local similarity scanner, that finds sequence patterns similar to a submitted query. Availability: The IGLOSS web server is available at http://compbioserv.math.hr/igloss/

A Clique-Based Method for Improving Motif Scanning Accuracy

“…In order to test the method, we applied it to responses from three iterative motif scanners -PSI-BLAST (PB) ( [1]), JackHMMER (JH) ( [6]) and IGLOSS (IG) ( [9]) -and compared the maximal clique with the original response. As in ( [9]), scanners were applied to five plant proteomes -Arabidopsis thaliana (AT, v. TAIR9), Oryza sativa (OS, v. MSU v7), Solanum tuberosum (ST, v. ITAG1), Solanum lycopersicum (SL, v. ITAG2.3) and Beta vulgaris (BV, v. KWS2320) -where we searched for members of an extensively studied, motif characterized protein family -GDSL lipases. GDSL lipases belong to lipid hydrolyzing enzymes that exhibit a GDSL motif.…”

“…Block I contains the main characteristic motif (PROSITE:PS01098) ( [10]) from which the main search query of 10 amino acids was constructed. As in [9], the condition positive set was determined by processing the information from GoMapMan resource [12].…”

“…Both steps can be a source of errors -an inaccurate ranking scheme, with a low threshold, will generate a huge response, causing a large type I-like error, whereas a high threshold, while testing positive on strongest examples, might miss the candidates with a slightly weaker signal. Various applications deal with these problems in various ways -for example, JackHMMer ( [6]) uses -effectively -several scoring functions, each with its own ranking and threshold, while IGLOSS ( [9]) uses detailed parameter estimation to minimize both issues.…”