NetMHCII

NetMHCII predicts peptide binding to MHC class II alleles, including HLA-DR, HLA-DQ, HLA-DP and mouse MHC class II molecules, to support analysis of antigen presentation and adaptive immune responses.

Key Features:

• NN-align artificial neural network: Uses an artificial neural network-based NN-align approach to simultaneously identify peptide-binding cores and predict binding affinities to MHC class II molecules.
• Bias correction in training data: Employs a training algorithm that corrects for biases from redundant representation of binding cores to improve prediction accuracy.
• Flanking residue integration: Incorporates information from residues flanking the peptide-binding core to enhance prediction performance.
• Extensive benchmarking: Validated on large-scale benchmarks including six independent datasets covering 14 human MHC class II alleles.
• Allele coverage: Supports predictions for HLA-DR, HLA-DQ, HLA-DP and mouse MHC class II alleles.

Scientific Applications:

• Adaptive immune response analysis: Prediction of peptide–MHC class II binding to study cellular and humoral immunity mediated by MHC class II molecules.
• Host–pathogen interaction studies: Identification of peptides likely presented during infections to elucidate host–pathogen interactions.
• Vaccine development: Prioritization of candidate CD4+ T cell epitopes for vaccine antigen selection.
• Immunotherapy design: Selection of peptide targets for immunotherapeutic strategies requiring MHC class II presentation.
• Translational and mouse-model studies: Comparative prediction across human and mouse MHC class II alleles for experimental setups.

Methodology:

NN-align artificial neural network for simultaneous binding-core identification and affinity prediction; training algorithm with bias correction for redundant binding-core representation; incorporation of flanking residue information into predictions; benchmarking on six independent datasets covering 14 human MHC class II alleles.