AntiCP 2.0: an updated model for predicting anticancer peptides

Welcome to the official repository for AntiCP 2.0, an updated and improved computational method for predicting and designing anticancer peptides (ACPs) from amino acid sequences. This resource is designed to support researchers in peptide therapeutics, cancer biology, and computational drug discovery. Web Server: https://webs.iiitd.edu.in/raghava/anticp2/ Standalone (GitHub): https://github.com/raghavagps/anticp2/ Docker Container: https://webs.iiitd.edu.in/gpsrdocker/ Citation Agrawal, P., Bhagat, D., Mahalwal, M., Sharma, N., & Raghava, G. P. S. (2021). AntiCP 2.0: an updated model for predicting anticancer peptides. Briefings in Bioinformatics, 22(3), bbaa153. https://doi.org/10.1093/bib/bbaa153 About the Tool AntiCP 2.0 is an updated version of the original AntiCP method, developed to predict and design anticancer peptides using multiple machine learning classifiers trained on the largest available dataset. It consolidates sequence-level features — composition, binary profiles, terminus patterns, and motifs — into a unified prediction framework, enabling systematic identification and design of ACPs from raw amino acid sequences. The tool integrates data from: CancerPPD database (anticancer peptides) ACP-DL, ACPP, ACPred-FL, AntiCP and iACP datasets SwissProt (for random peptide generation in alternate dataset) Key Features Two Curated Datasets Main dataset: 861 ACPs vs. 861 AMPs (non-ACPs) Alternate dataset: 970 ACPs vs. 970 random peptides 80/20 split for training and validation Rich Feature Extraction Amino Acid Composition (AAC) — 20-dimensional vector Dipeptide Composition (DPC) — 400-dimensional vector Terminus Composition — N5, N10, N15, C5, C10, C15, and combined Binary Profile — captures residue order (not just composition) Hybrid Features — composition + binary profile + motif Multiple Machine Learning Classifiers Support Vector Machine (SVM) Random Forest (RF) Extra Trees (ETree) K-Nearest Neighbors (KNN) Artificial Neural Network / MLP Ridge Classifier