DeepACPpred: A Novel Hybrid CNN-RNN Architecture for Predicting Anti-Cancer Peptides

Abstract

Anti-cancer peptides (ACPs) are a promising alternative to traditional chemotherapy. To aid wet-lab and clinical research, there is a growing interest in using machine learning techniques to help identify good ACP candidates computationally. In this paper, we describe DeepACPpred, a novel deep learning model composed of a hybrid CNN-RNN architecture for predicting ACPs. Using several gold-standard ACP datasets, we demonstrate that DeepACPpred is highly effective compared to state-of-the-art ACP prediction models.