Premium
Analysis of topological and nontopological structural similarities in the PDB: New examples with old structures
Author(s) -
Alexandrov Nickolai N.,
Fischer Daniel
Publication year - 1996
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/(sici)1097-0134(199607)25:3<354::aid-prot7>3.0.co;2-f
Subject(s) - protein data bank (rcsb pdb) , similarity (geometry) , protein data bank , structural similarity , pairwise comparison , sequence (biology) , structural alignment , structural classification of proteins database , computational biology , computer science , relevance (law) , trace (psycholinguistics) , protein structure , data mining , sequence alignment , biology , chemistry , peptide sequence , artificial intelligence , stereochemistry , genetics , biochemistry , gene , political science , law , image (mathematics) , linguistics , philosophy
We have developed a new method and program, SARF2, for fast comparison of protein structures, which can detect topological as well as nontopological similarities. The method searches for large ensembles of secondary structure elements, which are mutually compatible in two proteins. These ensembles consist of small fragments of C α ‐trace, similarly arranged in three‐dimensional space in two proteins, but not necessarily equally‐ordered along the polypeptide chains. The program SARF2 is available for everyone through the World‐Wide Web (WWW). We have performed an exhaustive pairwise comparison of all the entries from a recent issue of the Protein Data Bank (PDB) and report here on the results of an automated hierarchical cluster analysis. In addition, we report on several new cases of significant structural resemblance between proteins. To this end, a new definition of the significance of structural similarity is introduced, which effectively distinguishes the biologically meaningful equivalences from those occurring by chance. Analyzing the distribution of sequence similarity in significant structural matches, we show that sequence similarity as low as 20% in structurally‐prealigned proteins can be a strong indication for the biological relevance of structural similarity. © 1996 Wiley‐Liss, Inc.