Enhancing Antigenic Peptide Discovery: Improved MHC-I Binding Prediction and Methodology

This work analyses weaknesses in existing MHC-I presentation benchmarks, including train–test overlap and limited tests of generalisation to unseen peptides and alleles. It introduces a stricter benchmark and HLABERT, a pretrained Transformer model for pan-specific MHC-I presentation prediction.

An extended abstract was presented at the Machine Learning for Drug Discovery Workshop at ICLR 2023. The canonical peer-reviewed publication is the final Methods article.

Citation

Giziński, S., Preibisch, G., Kucharski, P., Tyrolski, M., Rembalski, M., Grzegorczyk, P., & Gambin, A. (2024). Enhancing antigenic peptide discovery: Improved MHC-I binding prediction and methodology. Methods, 224, 1–9. https://doi.org/10.1016/j.ymeth.2024.01.016