Premium
Studies on the inference of protein binding regions across fold space based on structural similarities
Author(s) -
Teyra Joan,
Hawkins John,
Zhu Hongbo,
Pisabarro M. Teresa
Publication year - 2011
Publication title -
proteins: structure, function, and bioinformatics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.699
H-Index - 191
eISSN - 1097-0134
pISSN - 0887-3585
DOI - 10.1002/prot.22897
Subject(s) - protein data bank (rcsb pdb) , computational biology , inference , structural similarity , binding site , protein structure , hierarchy , protein family , identification (biology) , biology , computer science , genetics , artificial intelligence , biochemistry , economics , gene , market economy , botany
The emerging picture of a continuous protein fold space highlights the existence of non obvious structural similarities between proteins with apparent different topologies. The identification of structure resemblances across fold space and the analysis of similar recognition regions may be a valuable source of information towards protein structure‐based functional characterization. In this work, we use non‐sequential structural alignment methods (ns‐SAs) to identify structural similarities between protein pairs independently of their SCOP hierarchy, and we calculate the significance of binding region conservation using the interacting residues overlap in the ns‐SA. We cluster the binding inferences for each family to distinguish already known family binding regions from putative new ones. Our methodology exploits the enormous amount of data available in the PDB to identify binding region similarities within protein families and to propose putative binding regions. Our results indicate that there is a plethora of structurally common binding regions among proteins, independently of current fold classifications. We obtain a 6‐ to 8‐fold enrichment of novel binding regions, and identify binding inferences for 728 protein families that so far lack binding information in the PDB. We explore binding mode analogies between ligands from commonly clustered binding regions to investigate the utility of our methodology. A comprehensive analysis of the obtained binding inferences may help in the functional characterization of protein recognition and assist rational engineering. The data obtained in this work is available in the download link at www.scowlp.org . Proteins 2011. © 2010 Wiley‐Liss, Inc.