Fig. 3: SWING captures pMHC biology.

a, Hydrophobic scale for SWING pMHC prediction. b, Hydrophobicity score class I model SCV performance with permutation testing defined by AUROC (left). Blue, validation curve; red, permuted mean; green, perfect classifier and gray, random classifier. Hydrophobicity score class I model cross-prediction performance on three alleles in the validation set as defined by the AUROC (right). Blue, HLA-A02:02; orange, HLA-B40:02; magenta, HLA-C05:01; green, perfect classifier and gray, random classifier. c, Hydrophobicity score class II model SCV performance with permutation testing defined by AUROC (left). Blue, validation curve; red, permuted mean; green, perfect classifier and gray, random classifier. Hydrophobicity score class II model cross-prediction performance on two alleles in the validation set as defined by the AUROC (right). Blue, DRB1_0102; orange, DRB1_0404; green, perfect classifier and gray, random classifier. d, Sequence length modification for SWING pMHC prediction. e, Full sequence class I model SCV performance with permutation testing defined by AUROC (left). Blue, validation curve; red, permuted mean; green, perfect classifier and gray, random classifier. Full sequence class I model cross-prediction performance on three alleles in the validation set as defined by AUROC (right). Blue, HLA-A02:02; orange, HLA-B40:02; magenta, HLA-C05:01; green, perfect classifier and gray, random classifier. f, Full sequence class II model SCV performance with permutation testing defined by AUROC (left). Blue, validation curve; red, permuted mean; green, perfect classifier and gray, random classifier. Full sequence class II model cross-prediction performance two alleles in the validation set as defined by AUROC (right). Blue, DRB1_0102; orange, DRB1_0404; green, perfect classifier and gray, random classifier. g, Peptide length distribution of the interacting peptides in the class II datasets defined by percentage. Magenta, training set; blue, DRB1_0102 validation set and orange, DRB1_0404 validation set. h, Peptide length truncation in training and test datasets affects the predictive power of the SWING class II model as defined by the AUC for each cutoff size in two class II datasets (left). Blue, DRB1_0102 and orange, DRB1_0404. i, Stratification of the model performance using four truncation cutoffs for cross predictions on DRB1_0102 (left) and DRB1_0404 (right) defined by AUROC. Sea green, full-length peptides; purple, 20-AA truncation; magenta, 16-AA truncation and yellow, 12-AA truncation. Mean AUROC ± 2 × standard deviation. Panels a and d created using BioRender.com.