QSAR Study of Angiotensin I-Converting Enzyme Inhibitory Dipeptides Based on PRIN Descriptors by Support Vector Machine

The support vector machine (SVM) was employed to construct dipeptides Quantitative Structure-Activity Relationship (QSAR) model. Amino acid descriptors PRIN parameter has been introduced in bioactive peptides QSAR Study in the article at firstly. Outliers of 168 Angiotensin I-Converting Enzyme (ACE) Inhibitory Dipeptides are deleted and QSAR model is constructed by support vector regression (SVR) leave-one-sample-out cross-validation (LOSOCV). The importance of each parameter or property at each position in peptides is estimated by the value of the model PRESS obtained using leave-one-parameter-out (LOPO) approach. Amino acid residues with bulky side chains as well as hydrophobic side chains were preferred for dipeptides by using leave one parameter out SVR, Which is similar to the literature. This will be provided with certain guidance meaning to design and exploit ACE Inhibitor. The results indicate that SVM can be used as an alternative powerful modeling tool for peptide QSAR studies, and give a advice (LOPO) about evaluating the importance of parameter in SVR model. Moreover, it also offered an idea about nonlinear relation between bioactive of peptides and their structural descriptors PRIN. The establishment of such methods will be a very meaning work to peptide bioactive investigation in peptide analogue drug design.