Premium
A Novel MHDV Descriptor for Dipeptide QSAR Studies
Author(s) -
Liu Shushen,
Yin Chunsheng,
Cai Shaoxi,
Li Zhiliang
Publication year - 2001
Publication title -
journal of the chinese chemical society
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.329
H-Index - 45
eISSN - 2192-6549
pISSN - 0009-4536
DOI - 10.1002/jccs.200100041
Subject(s) - chemistry , dipeptide , quantitative structure–activity relationship , principal component analysis , molecular descriptor , molecule , biological system , correlation coefficient , matrix (chemical analysis) , hydrogen bond , peptide , computational chemistry , artificial intelligence , stereochemistry , chromatography , machine learning , organic chemistry , biochemistry , computer science , biology
Abstract A novel molecular holographic distance vector (MHDV) is proposed to characterize the structures of the peptide molecules and employed to relate to the biological activities of the peptides by means of principal component regression (PCR) method. For two pan els of dipeptides, the correlation coefficient ( R ) between the estimated and the observed activities are respectively 0.9370 and 0.9585 and the R obtained by cross‐validation method are respectively 0.8676 and 0.9295, which is the best result to date for the two sets of dipeptides. The novel MHDV descriptor only depends on distance matrix and various atomic types of non‐hydrogen atoms in a molecule and requires no 3D structural information, so, it is a very simple and easy to use descriptor.