Premium
Predicting the Activity of Peptides Based on Amino Acid Information
Author(s) -
Wang XiaoYu,
Wang Juan,
Hu Yong,
Lin Yong,
Shu Mao,
Wang Li,
Cheng XiaoMing,
Lin ZhiHua
Publication year - 2011
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.201190139
Subject(s) - chemistry , partial least squares regression , quantitative structure–activity relationship , stepwise regression , peptide , oxytocin , regression analysis , amino acid , regression , computational biology , biochemistry , stereochemistry , machine learning , statistics , medicine , computer science , biology , mathematics
A set of amino acid descriptors including hydrophobic, stereo and electrical properties were applied to construct quantitative structure‐activity relationships (QSARs) models of three peptides datasets (angiotensin‐converting enzyme inhibitor dipeptides, bactericidal peptides and oxytocin peptides) with stepwise multiple regression combined partial least squares regression (SMR‐PLS). The results of QASRs models are very robust, with multiple correlation coefficients (R 2 ), and cross validation (Q 2 ) equal to 0.687, 0.671; 0.977, 0.890 and 0.950, 0.802 respectively. The robust models show the descriptors can be further expanded for polypeptides and serve as a useful quantitative tool for the rational drug design and discovery.