Advancing Peptide Toxicity Prediction with the Structure-Aware Deep Learning Model, tAMPer

We are pleased to announce the publication of tAMPer, a cutting-edge deep learning model for predicting peptide toxicity. Published as “Structure-Aware Deep Learning Model for Peptide Toxicity Prediction” in the journal Protein Science, tAMPer integrates amino acid sequence composition with ColabFold-predicted peptide structures through graph and recurrent neural networks. This model aims to expedite antimicrobial peptide (AMP) discovery by reducing reliance on costly toxicity screening experiments. tAMPer is poised to advance antimicrobial research and accelerate the development of new peptide/protein-based biologics in our global fight against multi-drug-resistant pathogens. Discover the full potential of tAMPer and its impact on antimicrobial research by visiting our tAMPer GitHub project page and reading the tAMPer manuscript!