Premium
Protein fragment clustering and canonical local shapes
Author(s) -
Hunter Cornelius G.,
Subramaniam Shankar
Publication year - 2003
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.10309
Subject(s) - centroid , cluster analysis , basis (linear algebra) , fragment (logic) , cluster (spacecraft) , pattern recognition (psychology) , resolution (logic) , set (abstract data type) , computer science , artificial intelligence , mathematics , data mining , algorithm , geometry , programming language
A novel clustering method is used to cluster protein fragments by shape. The centroids (mean fragments from each cluster) form a basis set of structural motifs. A database of 156,643 seven‐residue fragments is used, and eight different basis sets with varying levels of resolution are generated. Coarse basis sets contain tens of centroids and provide meaningful local shapes, which are more detailed than the traditional secondary structure categories. High‐resolution basis sets contain thousands of centroids and can be used to model tertiary structure of longer segments. The basis sets generated fit nontraining set proteins with the expected accuracy. Proteins 2003;50:580–588. © 2003 Wiley‐Liss, Inc.