AntiAngioPred: A Server for Prediction of Anti-Angiogenic Peptides

Welcome to the official repository for AntiAngioPred, a machine learning-based platform developed for predicting anti-angiogenic peptides involved in inhibiting blood vessel formation associated with tumor growth and cancer progression. Web Server: https://webs.iiitd.edu.in/raghava/antiangiopred/ Mirror Server: http://crdd.osdd.net/raghava/antiangiopred/ Brief Description Angiogenesis is the biological process responsible for the formation of new blood vessels and plays a major role in tumor development and cancer metastasis. Anti-angiogenic peptides can suppress this process and are therefore considered promising therapeutic agents for cancer treatment. AntiAngioPred was developed to computationally identify anti-angiogenic peptides using sequence-based machine learning approaches. The study analyzed experimentally validated anti-angiogenic peptides and identified characteristic residue preferences, motifs, and positional patterns associated with angiogenesis inhibition. Multiple predictive models were developed using amino acid composition, dipeptide composition, and binary profile patterns combined with machine learning classifiers such as Support Vector Machines (SVM), Random Forest, Naïve Bayes, and Neural Networks. The platform provides prediction, peptide screening, and peptide mutation analysis modules through a user-friendly web interface for researchers working in peptide therapeutics and computational oncology. Citation Ramaprasad, A. S. E., Singh, S., Raghava, G. P. S., & Venkatesan, S. (2015). AntiAngioPred: A Server for Prediction of Anti-Angiogenic Peptides. PLOS ONE, 10(9), e0136990. https://doi.org/10.1371/journal.pone.0136990 About the Platform AntiAngioPred is an in silico prediction system developed for identifying anti-angiogenic peptides from peptide sequences using machine learning techniques. The platform performs: Anti-angiogenic peptide prediction Multiple peptide screening Mutational analysis of peptides Peptide analog generation The prediction models were trained on experimentally validated peptides collected from: Published literature Patents Swiss-Prot derived negative datasets