Dataset for SiaScoreNet: A Siamese neural network-based model integrating prediction scores for HLA-peptide interaction prediction

Objective. Cancer immunotherapy uses the immune system to recognize and eliminate tumor cells by presenting tumor antigens through Human Leukocyte Antigen (HLA) molecules. These molecules are crucial in immune surveillance, presenting peptides to T cells. Accurate prediction of HLA–peptide interactions is essential for personalized immunotherapy development. Allele-specific models achieve high accuracy and handle variable peptide lengths but require separate training for each allele, limiting scalability to rare or unseen HLAs. Pan-specific models generalize across multiple alleles and match or surpass allele-specific methods. Ensemble methods improve prediction by combining outputs from multiple predictors, often via linear combinations, though nonlinear strategies may better capture HLA–peptide complexities.Methods. We propose SiaScoreNet, a three-step predictive pipeline enhancing HLA–peptide interaction prediction. First, ESM, a pretrained transformer-based protein language model, embeds HLA and peptide sequences into fixed length representations, accommodating varying sequence lengths. Second, we integrate predicted scores from state of-the-art models into a comprehensive feature vector. Third, a nonlinear ensemble strategy combines all features, capturing complex dependencies and boosting predictive performance.Results. Benchmark evaluations show SiaScoreNet outperforms existing models in accuracy, comparable to TransPHLA, BigMHC, and CapHLA. Recent models prioritize recall over precision, valuable for identifying potential binders but resource-intensive. SiaScoreNet offers improved predictive performance and runtime efficiency compared to these models, evaluated against HPV viruses for HLA–peptide interaction prediction. SiaScoreNet strong performance supports personalized cancer immunotherapy advancements.