mACPpred: A Support Vector Machine-Based Meta-Predictor for Identification of Anticancer Peptides

Anticancer peptides (ACPs) are promising therapeutic agents for targeting and killing cancer cells. The accurate prediction of ACPs from given peptide sequences remains as an open problem in the field of immunoinformatics. Recently, machine learning algorithms have emerged as a promising tool for helping experimental scientists predict ACPs. However, the performance of existing methods still needs to be improved. In this study, we present a novel approach for the accurate prediction of ACPs, which involves the following two steps: (i) We applied a two-step feature selection protocol on seven feature encodings that cover various aspects of sequence information (composition-based, physicochemical properties and profiles) and obtained their corresponding optimal feature-based models. The resultant predicted probabilities of ACPs were further utilized as feature vectors. (ii) The predicted probability feature vectors were in turn used as an input to support vector machine to develop the final prediction model called mACPpred. Cross-validation analysis showed that the proposed predictor performs significantly better than individual feature encodings. Furthermore, mACPpred significantly outperformed the existing methods compared in this study when objectively evaluated on an independent dataset.