Meta learning for mutant HLA class I epitope immunogenicity prediction to accelerate cancer clinical immunotherapy DOI Creative Commons
Long Xu, Qiang Yang, Weihe Dong

и другие.

Briefings in Bioinformatics, Год журнала: 2024, Номер 26(1)

Опубликована: Ноя. 22, 2024

Abstract Accurate prediction of binding between human leukocyte antigen (HLA) class I molecules and antigenic peptide segments is a challenging task key bottleneck in personalized immunotherapy for cancer. Although existing tools have demonstrated significant results using established datasets, most can only predict the affinity peptides to HLA do not enable immunogenic interpretation new epitopes. This limitation from training data computational models relying heavily on large amount peptide-HLA (pHLA) eluting ligand data, which candidate epitopes lack immunogenicity. Here, we propose an adaptive immunogenicity model, named MHLAPre, trained large-scale MS-derived eluted ligandome (mostly presented by epitopes) that are immunogenic. Allele-specific pan-allelic also provided endogenous presentation. Using meta-learning strategy, MHLAPre rapidly assessed affinities across whole pHLA pairs accurately identified tumor-associated antigens. During process immune response T-cells, pHLA-specific presentation pre-task CD8+ T-cell recognition. The factor activating interaction complexes receptors (TCRs). Therefore, performed transfer learning model pHLA-TCR dataset. In task, improvement identifying neoepitope compared with five state-of-the-art models, proving its effectiveness robustness. After exhibited relatively superior performance revealing mechanism immunotherapy. powerful tool identify neoepitopes interact TCR induce responses. We believe proposed method will greatly contribute clinical immunotherapy, such as anti-tumor immunity, tumor-specific engineering, tumor vaccine.

Язык: Английский

T-cell receptor binding prediction: A machine learning revolution DOI Creative Commons
Anna Weber, Aurélien Pélissier, María Rodríguez Martínez

и другие.

ImmunoInformatics, Год журнала: 2024, Номер 15, С. 100040 - 100040

Опубликована: Июль 22, 2024

Язык: Английский

Процитировано

5
Lessons learned from the IMMREP23 TCR-epitope prediction challenge DOI Creative Commons
Morten Nielsen, Anne Eugster, Mathias Fynbo Jensen

и другие.

ImmunoInformatics, Год журнала: 2024, Номер 16, С. 100045 - 100045

Опубликована: Сен. 28, 2024

Язык: Английский

Процитировано

4
Contrastive learning of T cell receptor representations DOI Creative Commons
Yuta Nagano,

Andrew G. T. Pyo,

Martina Milighetti

и другие.

Cell Systems, Год журнала: 2025, Номер unknown, С. 101165 - 101165

Опубликована: Янв. 1, 2025

Язык: Английский

Процитировано

0
Origins of T‐cell‐mediated autoimmunity in acquired aplastic anaemia DOI Creative Commons

Aura Enache,

Shannon A. Carty, Daria V. Babushok

и другие.

British Journal of Haematology, Год журнала: 2025, Номер unknown

Опубликована: Янв. 21, 2025

Summary Acquired aplastic anaemia (AA) is an autoimmune bone marrow failure disease resulting from a cytotoxic T‐cell‐mediated attack on haematopoietic stem and progenitor cells (HSPCs). Despite significant progress in understanding the T‐cell repertoire alterations AA, identifying specific pathogenic T AA patients has remained elusive, primarily due to unknown antigenic targets of attack. In this review, we will synthesize findings several decades research critically evaluate current knowledge repertoires AA. We highlight new insights gained recent vitro studies candidate autoreactive isolated discuss efforts identify shared clonotypes Finally, emerging evidence potential cross‐reactivity between HSPC common viral epitopes that may contribute development some patients. conclude by highlighting areas consensus limitations, as well ongoing uncertainties, promising directions for future field.

Язык: Английский

Процитировано

0
Bridging peptide presentation and T cell recognition with multi-task learning DOI
Li Su, Duolin Wang, Dong Xu

и другие.

Nature Machine Intelligence, Год журнала: 2025, Номер unknown

Опубликована: Фев. 20, 2025

Язык: Английский

Процитировано

0
EpicPred: Predicting phenotypes driven by epitope binding TCRs using attention-based multiple instance learning DOI Creative Commons

Jaemin Jeon,

Suwan Yu,

Sangam Lee

и другие.

Bioinformatics, Год журнала: 2025, Номер 41(3)

Опубликована: Фев. 21, 2025

Correctly identifying epitope-binding T-cell receptors (TCRs) is important to both understand their underlying biological mechanism in association some phenotype and accordingly develop mediated immunotherapy treatments. Although the importance of CDR3 region TCRs for epitope recognition well recognized, methods profiling interactions a certain disease or remains less studied. We developed EpicPred identify phenotype-specific TCR-epitope interactions. first predicts removes unlikely reduce false positives using Open-set Recognition (OSR). Subsequently, multiple instance learning was used specific cancer type severity levels COVID-19 infected patients. From six public TCR databases, 244 552 sequences 105 unique epitopes were predict filter out non-epitope-binding OSR method. The predicted further groups two four TCR-seq datasets bulk single-cell resolution. outperformed competing predicting phenotypes, achieving an average AUROC 0.80 ± 0.07. Software available at https://github.com/jaeminjj/EpicPred.

Язык: Английский

Процитировано

0
Tαβ-Bphg: A Dual-Branch Heterogeneous Graph Neural Network for Tcrαβ-Peptide Binding Prediction DOI

Yun Xie,

Lin Tang,

Guifei Zhou

и другие.

Опубликована: Янв. 1, 2025

Язык: Английский

Процитировано

0
T-cell receptor dynamics in digestive system cancers: a multi-layer machine learning approach for tumor diagnosis and staging DOI Creative Commons

Changjin Yuan,

Bin Wang, Hong Wang

и другие.

Frontiers in Immunology, Год журнала: 2025, Номер 16

Опубликована: Апрель 8, 2025

T-cell receptor (TCR) repertoires provide insights into tumor immunology, yet their variations across digestive system cancers are not well understood. Characterizing TCR differences between colorectal cancer (CRC) and gastric (GC), as developing machine learning models to distinguish types, metastatic status, disease stages crucial for guiding clinical practices. A cohort study of 143 patients (96 CRC, 47 GC) was conducted. High-throughput sequencing performed capture beta (TRB), delta (TRD), gamma (TRG) chain data. Tissue-specific patterns in repertoire features, such V-J gene recombination, complementarity-determining region 3 (CDR3) sequences, motif distributions, were analyzed. Multi-layer learning-based diagnostic developed by leveraging motif-based feature deep extraction using ProteinBERT from the 100 most abundant CDR3 sequences per sample. These used differentiate CRC GC, primary lesions, predict CRC. observed Distinct recombination identified, with showing enrichment TRBV*-TRBJ* combinations, while GC exhibited higher levels γδT-cell-related recombination. Primary lesions displayed distinct preferences (e.g., TRBV7-9/TRBJ2-1 metastatic; TRBV20-1/TRBJ1-2 primary) sequence differences, having shorter TRG lengths (p-value = 0.019). Across stages, later (III-IV) showed clonal diversity < 0.05) stage-specific patterns, alongside amino acid at N-terminal (positions 1-2) central positions 5-12). Multi-dimensional demonstrated exceptional performance all classification tasks. For distinguishing model achieved an accuracy 97.9% area under curve (AUC) 0.996. differentiating 100% AUC 1.000. In predicting attained 96.9% 0.993. Extensive validation simulated publicly available datasets, confirmed robustness reliability models, demonstrating consistent diverse datasets experimental conditions. Our investigation provides novel tumors, highlight potential immune features powerful tools understanding progression potentially improving decision-making.

Язык: Английский

Процитировано

0
Sequence-Based TCR-Peptide Representations Using Cross-Epitope Contrastive Fine-Tuning of Protein Language Models DOI
Chiho Im, Ryan Zhao, Scott D. Boyd

и другие.

Lecture notes in computer science, Год журнала: 2025, Номер unknown, С. 34 - 48

Опубликована: Янв. 1, 2025

Язык: Английский

Процитировано

0
Structural, evolutionary, and immunogenicity insight into the application of microalgal Spirulina maxima asparaginase in medicinal and industrial purposes DOI Creative Commons

Leila Sarafan Soleimanzadeh,

Maryam Azimzadeh Irani

Deleted Journal, Год журнала: 2025, Номер 7(5)

Опубликована: Апрель 25, 2025

Язык: Английский

Процитировано

0