MHC-Fine: Fine-tuned AlphaFold for Precise MHC-Peptide Complex Prediction.

Area of Science:

• Computational immunology
• Structural biology
• Vaccine development

Background:

• Accurate prediction of Major Histocompatibility Complex (MHC)-peptide complex structures is crucial for understanding T-cell mediated immunity.
• Existing generalist models like AlphaFold lack the specialized precision required for MHC-peptide interactions.
• Advancing vaccine design and immunotherapy necessitates high-resolution structural predictions of these complexes.

Approach:

• Fine-tuned AlphaFold using a curated dataset of high-resolution MHC-peptide crystal structures.
• Developed a specialized model to overcome the limitations of generalist protein structure prediction tools.
• Compared performance against established methods, including Pandora and the standard AlphaFold multimer model.

Key Points:

• The fine-tuned model achieved superior performance with a median RMSD of 0.65 Å.
• Enhanced predicted lDDT scores indicate more reliable structural predictions.
• Demonstrated significant improvement over existing homology modeling and general AlphaFold approaches.

Conclusions:

• Specialized fine-tuning of AlphaFold significantly enhances MHC-peptide complex structure prediction accuracy.
• This improved precision offers a powerful computational tool for drug discovery and vaccine development.
• Advances in predicting MHC-peptide interactions will accelerate the design of targeted immunotherapies.