Foundation Protein Language Models for Influenza A Virus T-Cell Epitope Prediction: A Transformer-Based...

Area of Science:

• Computational biology
• Immunoinformatics
• Virology

Background:

• Influenza A virus poses a global health challenge due to rapid evolution and antigenic variability.
• T-cell immunity offers broader protection against conserved viral regions, making T-cell epitope identification crucial for vaccine design.
• Existing computational methods often rely on limited handcrafted features, hindering the capture of complex sequence dependencies.

Purpose of the Study:

• To develop a novel transformer-based viroinformatics framework for predicting Influenza A T-cell epitopes (TCEs).
• To leverage protein language models (PLMs) for automated feature extraction from amino acid sequences.
• To improve the accuracy and interpretability of computational TCE prediction for vaccine research.

Main Methods:

• Utilized a pretrained Evolutionary Scale Modeling-2 (ESM-2) protein language model to generate contextualized embeddings from peptide sequences.
• Employed an attention-based transformer classifier to learn epitope-specific patterns from ESM-2 embeddings.
• Incorporated Monte Carlo dropout for uncertainty-aware predictions and attention-based interpretability for biological insights.

Main Results:

• Achieved high predictive performance with approximately 97% accuracy and an AUC close to 0.99 under cross-validation.
• Demonstrated that PLM representations and self-attention significantly outperform classical machine learning baselines.
• Identified residue-level contributions to model predictions, providing biologically meaningful insights into epitope recognition.

Conclusions:

• The proposed PLM-enhanced transformer framework offers an effective and interpretable computational approach for Influenza A TCE discovery.
• This method advances epitope-based vaccine design by enabling more accurate identification of targets for durable T-cell immunity.
• The framework provides a promising tool for immunological research and the development of next-generation influenza vaccines.