EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC 4.0

MultiMHC: a multiscale learning framework integrating sequence and residue contact topology for peptide–MHC binding prediction

Wenfan Chen, Ying Chi iD, Yihao Wang, Jingcheng Wu iD, Mengmeng Wei, Wenyi Zhao iD, Yu Kang iD, Xueli Bai, Shuqing Chen iD, Zhan Zhou iD, Ruhong Zhou iD

Briefings in Bioinformatics · 02.11.2026 · 27 (6)

DOI: 10.1093/bib/bbag466 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Briefings in Bioinformatics, online erschienen am 06.11.2026. Verfasst von Wenfan Chen und 10 weiteren Autor:innen. Beteiligte Einrichtung: Institute of Quantitative Biology, College of Life Sciences, Zhejiang University, 866 Yuhangtang Road, Xihu District, Hangzhou 310058, China. Hauptthemen laut MEDLINE: Peptides, Major Histocompatibility Complex, Deep Learning. Die Arbeit nennt 50 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0). Gefördert von: Fundamental Research Funds for the Central Universities, Key Project under the Zhejiang Provincial and Ministry Co-construction Plan, Zhejiang Provincial Natural Science Foundation of China.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Accurate prediction of peptide–major histocompatibility complex (pMHC) binding is central to understanding antigen presentation and identifying T-cell epitopes in cancer immunotherapy. Although recent methods have achieved notable progress, most approaches rely predominantly on sequence-based features and only partially capture structural determinants of pMHC recognition. In this study, we present MultiMHC, a multiscale deep learning framework that integrates complementary sequential and biologically-informed residue contact topology for pMHC binding prediction. MultiMHC combines a recurrent neural network to model global sequence context with a graph-based representation to encode residue contact patterns derived from known pMHC binding interfaces. Features from these two pathways are subsequently fused to generate binding predictions across peptides of varying lengths. Benchmark evaluations show that MultiMHC achieves performance comparable to existing state-of-the-art methods, with particularly strong performance in ranking-based assessments. Moreover, evaluation on the Tumor Neoantigen Selection Alliance dataset showed that MultiMHC binding scores can enrich experimentally validated T-cell epitopes among top-ranked candidates, supporting its potential utility as an upstream binding-based filtering step in neoantigen discovery workflows. Together, these results highlight the value of incorporating residue interaction patterns and lightweight structural information into pMHC modeling and establish MultiMHC as a robust and extensive framework for antigen presentation analysis.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

PeptidesMajor Histocompatibility ComplexDeep LearningHumansProtein BindingImmunoinformaticsEpitopes, T-LymphocytePrediction AlgorithmsComputational Biology

Stichwörter der Autor:innen

deep learninggraph neural networkmulti-modalneoantigenpeptide–MHC binding prediction

Substanzen

PeptidesEpitopes, T-Lymphocyte

Autor:innen und Institutionen

  • Institute of Quantitative Biology, College of Life Sciences, Zhejiang University, 866 Yuhangtang Road, Xihu District, Hangzhou 310058, China.
  • State Key Laboratory of Advanced Drug Delivery and Release Systems & Innovation Institute for Artificial Intelligence in Medicine, School of Pharmacy, Zhejiang University, 866 Yuhangtang Road, Xihu District, Hangzhou 31…
  • ZJU-UoE Institute, Zhejiang University, 718 East Haizhou Road, Haining 314400, China.
  • Zhejiang Provincial Key Laboratory of Anti-Cancer Drug Research & MOE Engineering Research Center of Innovative Anticancer Drugs, Zhejiang University, 866 Yuhangtang Road, Xihu District, Hangzhou 310058, China.
  • Key Laboratory of Clinical Cancer Pharmacology and Toxicology Research of Zhejiang Province, Hangzhou First People's Hospital, Westlake University School of Medicine, 261 Huansha Road, Shangcheng District, Hangzhou 3100…
  • Department of Hepatobiliary and Pancreatic Surgery, the First Affiliated Hospital, Zhejiang University School of Medicine, 79 Qingchun Road, Shangcheng District, Hangzhou 310003, China.

ORCID-Profile: Ying Chi, Jingcheng Wu, Wenyi Zhao, Yu Kang, Shuqing Chen, Zhan Zhou, Ruhong Zhou

Förderung

  • Fundamental Research Funds for the Central Universities
  • Key Project under the Zhejiang Provincial and Ministry Co-construction Plan
  • Zhejiang Provincial Natural Science Foundation of China
  • "Pioneer and Leading Goose + X" S&T Program of Zhejiang
  • Starry Night Science Fund of Zhejiang University Shanghai Institute for Advanced Study
  • Shanghai Artificial Intelligence Lab
  • National Center of Technology Innovation for Biopharmaceuticals
  • National Independent Innovation Demonstration Zone Shanghai Zhangjiang Major Projects

Zitieren

APA

Chen, W., Chi, Y., Wang, Y., Wu, J., Wei, M., Zhao, W., Kang, Y., Bai, X., Chen, S., Zhou, Z., & Zhou, R. (2026). MultiMHC: a multiscale learning framework integrating sequence and residue contact topology for peptide–MHC binding prediction. Briefings in Bioinformatics, 27(6). https://doi.org/10.1093/bib/bbag466

Vancouver

Chen W, Chi Y, Wang Y, Wu J, Wei M, Zhao W, et al. MultiMHC: a multiscale learning framework integrating sequence and residue contact topology for peptide–MHC binding prediction. Briefings in Bioinformatics. 2026;27(6). doi:10.1093/bib/bbag466

BibTeX
@article{Chen2026,
  author = {Chen, Wenfan and Chi, Ying and Wang, Yihao and Wu, Jingcheng and Wei, Mengmeng and Zhao, Wenyi and Kang, Yu and Bai, Xueli and Chen, Shuqing and Zhou, Zhan and Zhou, Ruhong},
  title = {MultiMHC: a multiscale learning framework integrating sequence and residue contact topology for peptide–MHC binding prediction},
  journal = {Briefings in Bioinformatics},
  year = {2026},
  volume = {27},
  number = {6},
  doi = {10.1093/bib/bbag466}
}
Kostenlos starten →