A quantitative structure–activity relationship (QSAR) study of peptide drugs based on a new descriptor of amino acids

Quantitative structure-activity relationships (QSAR) approach is used for finding the relationship between molecular structures and the activity of peptide drugs. In this work, stepwise multiple regression, was employed to select optimal subset of descriptors that have significant contribution to the drug activity of 21 oxytocin analogues, 48 bitter tasting threshold, and 58 angiotensin-converting enzyme inhibitors. A new set of descriptor, SVWGM, was used for the prediction of the drug activity of peptide drugs and then were used to build the model by partial least squares method, for it’s estimation stability and generalization ability was strictly analyzed by both internal and external validations, with cross-validation correlation coefficient, correlation coefficient and correlation coefficient of external validation