Similarity searching in databases of flexible 3D structures using autocorrelation vectors derived from smoothed bounded distance matrices

Nicholas Rhodes; David E Clark; Peter Willett

doi:10.1021/ci0503863

Similarity searching in databases of flexible 3D structures using autocorrelation vectors derived from smoothed bounded distance matrices

J Chem Inf Model. 2006 Mar-Apr;46(2):615-9. doi: 10.1021/ci0503863.

Authors

Nicholas Rhodes¹, David E Clark, Peter Willett

Affiliation

¹ Department of Information Studies, The University of Sheffield, Regent Court, 211 Portobello Street, Sheffield, South Yorkshire, S1 4DP, United Kingdom.

PMID: 16562990
DOI: 10.1021/ci0503863

Abstract

This paper presents an exploratory study of a novel method for flexible 3-D similarity searching based on autocorrelation vectors and smoothed bounded distance matrices. Although the new approach is unable to outperform an existing 2-D similarity searching in terms of enrichment factors, it is able to retrieve different compounds at a given percentage of the hit-list and so may be a useful adjunct to other similarity searching methods.

Publication types

Comparative Study

MeSH terms

Computing Methodologies*
Databases as Topic*
Enzyme Inhibitors / chemistry
Enzyme Inhibitors / classification
Image Processing, Computer-Assisted / methods*
Imaging, Three-Dimensional
Molecular Structure
Pharmaceutical Preparations / chemistry*
Structure-Activity Relationship*

Substances

Enzyme Inhibitors
Pharmaceutical Preparations