STAPLER: Efficient learning of TCR-peptide specificity prediction from full-length TCR-peptide data

Kwee, Bjorn P. Y.; Messemaker, Marius; Marcus, Eric; Oliveira, Giacomo; Scheper, Wouter; Wu, Catherine J.; Teuwen, Jonas; Schumacher, Ton N.

doi:10.1101/2023.04.25.538237

Cited by 11 publications

(5 citation statements)

References 43 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Decoding the predictive rules of TCR-pMHC specificity is a formidable challenge, largely owing to the extreme sparsity of available training data relative to the diversity of sequences that need to be interrogated in meaningful investigation. A majority of approaches ( 11 , 36 , 37 ) take a complementary approach to RACER-m by training on TCR and/or peptide primary sequence data alone. One recent method achieves training by relaxing a common requirement of having paired CDR3α/β sequences ( 36 ).…”

Section: Discussionmentioning

confidence: 99%

RACER-m leverages structural features for sparse T cell specificity prediction

Wang,

Lin,

Chau

et al. 2024

Sci. Adv.

View full text Add to dashboard Cite

Reliable prediction of T cell specificity against antigenic signatures is a formidable task, complicated by the immense diversity of T cell receptor and antigen sequence space and the resulting limited availability of training sets for inferential models. Recent modeling efforts have demonstrated the advantage of incorporating structural information to overcome the need for extensive training sequence data, yet disentangling the heterogeneous TCR-antigen interface to accurately predict MHC-allele-restricted TCR-peptide interactions has remained challenging. Here, we present RACER-m, a coarse-grained structural model leveraging key biophysical information from the diversity of publicly available TCR-antigen crystal structures. Explicit inclusion of structural content substantially reduces the required number of training examples and maintains reliable predictions of TCR-recognition specificity and sensitivity across diverse biological contexts. Our model capably identifies biophysically meaningful point-mutant peptides that affect binding affinity, distinguishing its ability in predicting TCR specificity of point-mutants from alternative sequence-based methods. Its application is broadly applicable to studies involving both closely related and structurally diverse TCR-peptide pairs.

show abstract

Section: Discussionmentioning

confidence: 99%

RACER-m leverages structural features for sparse T cell specificity prediction

Wang,

Lin,

Chau

et al. 2024

Sci. Adv.

View full text Add to dashboard Cite

show abstract

“…Concurrently, the development of large language models for proteins has led to important advancements in protein structure, function and evolution predictions [23,41,42]. While several studies have applied TCR-specific pLMs to TCR prediction [38,43], extensive research on the application of general pLMs, such as ESM, is lacking.…”

Section: Discussionmentioning

confidence: 99%

TCR clustering by contrastive learning on antigen specificity

Pertseva,

Follonier,

Scarcella

et al. 2024

Preprint

View full text Add to dashboard Cite

Effective clustering of T-cell receptor (TCR) sequences could be used to predict their antigen-specificities. TCRs with highly dissimilar sequences can bind to the same antigen, thus making their clustering into a common antigen group a central challenge. Here, we develop TouCAN, a method that relies on contrastive learning and pre-trained protein language models to perform TCR sequence clustering and antigen-specificity predictions. Following training, TouCAN demonstrates the ability to cluster highly dissimilar TCRs into common antigen groups. Additionally, TouCAN demonstrates TCR clustering performance and antigen-specificity predictions comparable to other leading methods in the field.

show abstract

“…Schumacher furthermore introduced STAPLER (Shared TCR And Peptide Language bidirectional Encoder Representations from transformers), a transformer language model to predict reactivity from full-sequence TCR and peptide-MHC (pMHC) input. 28 STAPLER training data leveraged large available databases on TCRs with unknown reactivity and pMHC complexes with unknown TCR matches to teach the algorithm what these molecules look like and which patterns they entail. In a second step, smaller data sets of matching TCR-pMHC pairs were used to achieve accurate prediction of TCR reactivity but only for those antigens for which training data are available.…”

Section: Keynote Lecturementioning

confidence: 99%

CIMT 2024: Report on the 21st Annual Meeting of the Association for Cancer Immunotherapy

Ahrberg,

Dallmann,

Freitag

et al. 2024

Human Vaccines & Immunotherapeutics

View full text Add to dashboard Cite

The 21 st Association for Cancer Immunotherapy (CIMT) Annual Meeting took place from May 15 th to May 17 th in Mainz, Germany, and was attended by a total of 855 academic and clinical professionals hailing from 33 different countries. The conference served as a platform for these experts to convene and discuss the latest breakthroughs in cancer immunology and immunotherapy research. Dedicated sessions covering advancements in artificial intelligence tools for cancer immunotherapy research, as well as the landscape of cancer care and cancer immunotherapy trials on the African continent, prompted lively and informative discussions among the attendees. This report aims to provide an overview of the most noteworthy highlights and key takeaways from CIMT2024.

show abstract

STAPLER: Efficient learning of TCR-peptide specificity prediction from full-length TCR-peptide data

Cited by 11 publications

References 43 publications

RACER-m leverages structural features for sparse T cell specificity prediction

RACER-m leverages structural features for sparse T cell specificity prediction

TCR clustering by contrastive learning on antigen specificity

CIMT 2024: Report on the 21st Annual Meeting of the Association for Cancer Immunotherapy

Contact Info

Product

Resources

About