Benchmarking unsupervised methods for inferring TCR specificity
NAR Genomics and Bioinformatics · 07.10.2025 · 7 (4)
DOI: 10.1093/nargab/lqaf150 ↗Kurz gefasst
Übersichtsarbeit in der Zeitschrift NAR Genomics and Bioinformatics, online erschienen am 19.11.2025. Verfasst von Charline Jouannet und 4 weiteren Autor:innen. Beteiligte Einrichtung: Sorbonne Université, INSERM, Immunoregulation-Immunopathology-Immunotherapy (i3), 75005, Paris, France. Hauptthemen laut MEDLINE: Receptors, Antigen, T-Cell, Computational Biology. Die Arbeit wurde bisher 17-mal zitiert (Crossref) und nennt 52 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: European Research Area Network-Cardiovascular Diseases, AIR-MI Consortium, Agence Nationale de la Recherche.
Abstract (Originalsprache)
Identifying T-cell receptor (TCR) specificity is crucial for advancing the understanding of adaptive immunity. Despite the development of computational methods to infer TCR specificity, their clustering behavior has not been thoroughly compared. We addressed this by curating a unified database of 190 670 human TCRs with known specificities for 2313 epitopes across 121 organisms, combining data from IEDB, McPAS-TCR, and VDJdb. We asked whether widely used TCR clustering methods produce comparable results on the same high-confidence dataset. We hypothesized that shared assumptions about conserved CDR3 motifs would yield similar patterns, with differences reflecting algorithmic design. Nine methods for clustering TCRs based on similarity were benchmarked against this dataset. DeepTCR demonstrated the best retention, while ClusTCR, TCRMatch, and GLIPH2 excelled in cluster purity but had lower retention. GLIPH2, Levenshtein distance, Hamming distance, and clusTCR generated large clusters in contrast to TCRMatch and DeepTCR. Smaller, antigen-specific clusters were produced by GIANA and iSMART. DeepTCR was the most sensitive in capturing antigen-specific TCRs. We confirmed these observations using a larger dataset from 10X Genomics containing antigen-specific labeled TCRs as well non-labeled cells. This study offers a unified TCR database and a benchmark of specificity inference methods, guiding researchers in selecting appropriate tools.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Substanzen
Autor:innen und Institutionen
- Sorbonne Université, INSERM, Immunoregulation-Immunopathology-Immunotherapy (i3), 75005, Paris, France.
- AP-HP, Hôpital Pitié-Salpêtrière, Biotherapy UF, CIC-Paris Est, 75013, Paris, France.
- Institut Universitaire de France, France.
ORCID-Profile: Hélène Vantomme, Encarnita Mariotti-Ferrandiz
Förderung
- European Research Area Network-Cardiovascular Diseases
- AIR-MI Consortium
- Agence Nationale de la Recherche
- Global Platform for the Prevention of Autoimmune Diabetes
- Institut Universitaire de France
- European Research Council
- Sorbonne Université
Zitieren
APA
Jouannet, C., Vantomme, H., Gouge, K. L., Klatzmann, D., & Mariotti-Ferrandiz, E. (2025). Benchmarking unsupervised methods for inferring TCR specificity. NAR Genomics and Bioinformatics, 7(4). https://doi.org/10.1093/nargab/lqaf150
Vancouver
Jouannet C, Vantomme H, Gouge KL, Klatzmann D, Mariotti-Ferrandiz E. Benchmarking unsupervised methods for inferring TCR specificity. NAR Genomics and Bioinformatics. 2025;7(4). doi:10.1093/nargab/lqaf150
BibTeX
@article{Jouannet2025,
author = {Jouannet, Charline and Vantomme, Hélène and Gouge, Kenz Le and Klatzmann, David and Mariotti-Ferrandiz, Encarnita},
title = {Benchmarking unsupervised methods for inferring TCR specificity},
journal = {NAR Genomics and Bioinformatics},
year = {2025},
volume = {7},
number = {4},
doi = {10.1093/nargab/lqaf150}
}
EUVIMED