{"architecture":{"family":"NNAlign_MA neural network","input":"peptide (12\u201325 aa) + MHC class II pseudosequence","notes_en":"Class II grooves are open at both ends, so the peptide overhangs; finding the correct register inside a long peptide is the hard part, and the deconvolution machinery is what made multi-allelic class II data usable.","output":"%Rank and predicted 9-mer binding core"},"article":"/blog/modele-prezentacji-antygenu","cancer_slugs":["pan-cancer"],"category":"immunopeptidomics","confidence":"medium","datasets":[{"name":"IEDB \u2014 Immune Epitope Database","note":"","role":"training","slug":"iedb"}],"developer":"Health Tech, Technical University of Denmark (Nielsen lab)","evaluation":[{"benchmark":"class II ligand and epitope prediction vs NetMHCIIpan-3.2","external":true,"metric":"AUC / PPV","source":"https://academic.oup.com/nar/article/48/W1/W449/5837056","value":"improved (paper)"}],"hf":null,"kind":"task-model","license":"free for academic use; commercial licence required (DTU)","limitations":"- Class II data are scarcer and noisier than class I; register uncertainty is intrinsic.\n- Same assumed-negative and MS-bias caveats as NetMHCpan.","links":{"demo":null,"docs":"https://services.healthtech.dtu.dk/services/NetMHCIIpan-4.0/","doi":"10.1093/nar/gkaa379","github":null,"huggingface":null,"paper":"https://academic.oup.com/nar/article/48/W1/W449/5837056","pmid":null},"modalities":["protein-sequence","immunopeptidomics"],"name":"NetMHCIIpan-4.0","notable_uses":"","openness":"api-only","regulatory":{"intended_use_en":"Research software.","intended_use_pl":"Oprogramowanie badawcze.","source_url":"https://services.healthtech.dtu.dk/services/NetMHCIIpan-4.0/","status":"research-only"},"regulatory_status":"research-only","release_date":"2020-05-22","run_snippet":"","settings":["basic-research","vaccine-design","immunotherapy"],"slug":"netmhciipan","sources":[{"label":"Reynisson B et al. Nucleic Acids Res 2020","url":"https://academic.oup.com/nar/article/48/W1/W449/5837056"},{"label":"cancer3.ai \u2014 Trzydzie\u015bci cztery litery zamka","url":"https://cancer3.ai/blog/modele-prezentacji-antygenu"}],"summary":"The class II counterpart of NetMHCpan, released in the same paper: predicts presentation by HLA-DR, -DQ and -DP, whose groove is open at both ends, so peptides are longer and the binding core has to be found inside a longer sequence.","tasks":["antigen-presentation","peptide-mhc-binding","neoantigen-prioritisation"],"training":{"summary_en":"Binding affinity plus MS-eluted class II ligands, with the same assumed-negative construction as class I.","summary_pl":"Powinowactwa i ligandy klasy II ze spektrometru, z t\u0105 sam\u0105 konstrukcj\u0105 przyj\u0119tych negatyw\u00f3w co w klasie I."},"updated_at":"2026-09-05T22:26:00.816287","url":"/ai-oncology/models/netmhciipan","usage":{"api":"https://services.healthtech.dtu.dk/services/NetMHCIIpan-4.0/","library":"standalone binary / web server"},"verified_at":"2026-09-05T22:26:00.814125","verified_by":"editorial","version":"4.0","what_it_does":"Same NNAlign_MA machinery as class I \u2014 pan-allele pseudosequence input, motif deconvolution of multi-allelic data \u2014 applied to CD4+ T-cell epitopes, which matter for vaccine helper responses."}