TY - JOUR
T1 - In silico identification of functional regions in proteins
AU - Nimrod, Guy
AU - Glaser, Fabian
AU - Steinberg, David
AU - Ben-Tal, Nir
AU - Pupko, Tal
N1 - Funding Information:
We thank Itay Mayrose and Sarel Fleishman for helpful discussions and comments on the manuscript. This work was supported by the Israel Cancer Association grant to N.B.-T. The research was initiated when T.P. was doing his Postdoctorate with Dr Dave Swofford. He thanks Dave Swofford for his support and for helpful discussion. T.P. was supported by a grant in Complexity Science from the Yeshaia Horvitz Association.
PY - 2005/6
Y1 - 2005/6
N2 - Motivation: In silico prediction of functional regions on protein surfaces, i.e. sites of interaction with DNA, ligands, substrates and other proteins, is of utmost importance in various applications in the emerging fields of proteomics and structural genomics. When a sufficient number of homologs is found, powerful prediction schemes can be based on the observation that evolutionarily conserved regions are often functionally important, typically, only the principal functionally important region of the protein is detected, while secondary functional regions with weaker conservation signals are overlooked. Moreover, it is challenging to unambiguously identify the boundaries of the functional regions. Methods: We present a new methodology, called PatchFinder, that automatically identifies patches of conserved residues that are located in close proximity to each other on the protein surface. PatchFinder is based on the following steps: (1) Assignment of conservation scores to each amino acid position on the protein surface. (2) Assignment of a score to each putative patch, based on its likelihood to be functionally important. The patch of maximum likelihood is considered to be the main functionally important region, and the search is continued for non-overlapping patches of secondary importance. Results: We examined the accuracy of the method using the IGPS enzyme, the SH2 domain and a benchmark set of 112 proteins. These examples demonstrated that PatchFinder is capable of identifying both the main and secondary functional patches.
AB - Motivation: In silico prediction of functional regions on protein surfaces, i.e. sites of interaction with DNA, ligands, substrates and other proteins, is of utmost importance in various applications in the emerging fields of proteomics and structural genomics. When a sufficient number of homologs is found, powerful prediction schemes can be based on the observation that evolutionarily conserved regions are often functionally important, typically, only the principal functionally important region of the protein is detected, while secondary functional regions with weaker conservation signals are overlooked. Moreover, it is challenging to unambiguously identify the boundaries of the functional regions. Methods: We present a new methodology, called PatchFinder, that automatically identifies patches of conserved residues that are located in close proximity to each other on the protein surface. PatchFinder is based on the following steps: (1) Assignment of conservation scores to each amino acid position on the protein surface. (2) Assignment of a score to each putative patch, based on its likelihood to be functionally important. The patch of maximum likelihood is considered to be the main functionally important region, and the search is continued for non-overlapping patches of secondary importance. Results: We examined the accuracy of the method using the IGPS enzyme, the SH2 domain and a benchmark set of 112 proteins. These examples demonstrated that PatchFinder is capable of identifying both the main and secondary functional patches.
UR - http://www.scopus.com/inward/record.url?scp=29144462537&partnerID=8YFLogxK
U2 - 10.1093/bioinformatics/bti1023
DO - 10.1093/bioinformatics/bti1023
M3 - ???researchoutput.researchoutputtypes.contributiontojournal.article???
AN - SCOPUS:29144462537
SN - 1367-4803
VL - 21
SP - i328-i337
JO - Bioinformatics
JF - Bioinformatics
IS - SUPPL. 1
ER -