Abstract:Abstract Background MHC class II binding predictions are widely used to identify epitope candidates in infectious agents, allergens, cancer and autoantigens. The vast majority of prediction algorithms for human MHC class II to date have targeted HLA molecules encoded in the DR locus. This reflects a significant gap in knowledge as HLA DP and DQ molecules are presumably equally important, and have only been studied less because they are more difficult to handle experimentally. Results In this study, we aimed to narrow this gap by providing a large scale dataset of over 17,000 HLA-peptide binding affinities for a set of 11 HLA DP and DQ alleles. We also expanded our dataset for HLA DR alleles resulting in a total of 40,000 MHC class II binding affinities covering 26 allelic variants. Utilizing this dataset, we generated prediction tools utilizing several machine learning algorithms and evaluated their performance. Conclusion We found that 1) prediction methodologies developed for HLA DR molecules perform equally well for DP or DQ molecules. 2) Prediction performances were significantly increased compared to previous reports due to the larger amounts of training data available. 3) The presence of homologous peptides between training and testing datasets should be avoided to give real-world estimates of prediction performance metrics, but the relative ranking of different predictors is largely unaffected by the presence of homologous peptides, and predictors intended for end-user applications should include all training data for maximum performance. 4) The recently developed NN-align prediction method significantly outperformed all other algorithms, including a naïve consensus based on all prediction methods. A new consensus method dropping the comparably weak ARB prediction method could outperform the NN-align method, but further research into how to best combine MHC class II binding predictions is required.

HLA class II DQ epitopes.

HLA-DQ-Specific Recombinant Human Monoclonal Antibodies Allow for In-Depth Analysis of HLA-DQ Epitopes

Identification of Antibodies to DQβ:DRα Interisotypic Heterodimers in Human Sera

Characterisation of the HLA‐DQ alpha eplet 52SK targeting the DQA1*01 alleles family

HLA class I peptide polymorphisms contribute to class II DQβ0603:DQα0103 antibody specificity

Separating the Wheat from the Chaff among HLA-DQ Eplets

Peptide binding motifs and specificities for HLA-DQ molecules

Identification of Three Novel HLA‐DQA1 Alleles: DQA101:08, DQA101:09 and DQA1*03:03:02

Combined effect of glutamine at position 70 of HLA-DRB1 and alanine at position 57 of HLA-DQB1 in type 1 diabetes: An epitope analysis

A Short History of B-Cell HLA Epitopes

Calcium-dependent HLA-DQ epitope revealed by EDTA mediated inhibition of antibody reactions in the Luminex single antigen bead assay

A new strategy for systematically classifying HLA alleles into serological specificities: Update and refinement

Conformational and structural characteristics of peptides binding to HLA-DR molecules.

Antibody verification of HLA class I and class II eplets by human monoclonal HLA antibodies

Microsatellite Single Nucleotide Polymorphisms in the HLA-DQ Region.

The antibodies 3D12 and 4D12 recognise distinct epitopes and conformations of HLA-E

Peptide binding predictions for HLA DR, DP and DQ molecules

Similarity between segments in protein conformational epitopes and MHC II peptides.

Predominant naturally processed peptides bound to HLA-DR1 are derived from MHC-related molecules and are heterogeneous in size

Characterisation of Four Novel HLA Alleles: ‐A02:01:222, ‐B07:02:104, ‐C07:02:156 and ‐DQB106:03:51

DQ Microsatellite Association Studies in Three Ethnic Groups.