Rapid synthesis and screening of natively paired antibodies against influenza hemagglutinin stem via oPool+display DOI Creative Commons
Wenhao O. Ouyang, Huibin Lv, W. Liu

и другие.

bioRxiv (Cold Spring Harbor Laboratory), Год журнала: 2024, Номер unknown

Опубликована: Авг. 30, 2024

Antibody discovery is crucial for developing therapeutics and vaccines as well understanding adaptive immunity. However, the lack of approaches to synthesize antibodies with defined sequences in a high-throughput manner represents major bottleneck antibody discovery. Here, we presented oPool + display, cell-free platform that combined oligo pool synthesis mRNA display rapidly construct characterize many natively paired parallel. As proof-of-concept, applied probe binding specificity >300 uncommon influenza hemagglutinin (HA) against 9 HA variants through 16 different screens. Over 5,000 tests were performed 3-5 days further scaling potential. Follow-up structural analysis two stem revealed previously unknown versatility IGHD3-3 gene segment recognizing stem. Overall, this study established an experimental not only accelerate characterization, but also enable unbiased molecular signatures.

Язык: Английский

Improving antibody language models with native pairing DOI Creative Commons
Sarah Burbach, Bryan Briney

Patterns, Год журнала: 2024, Номер 5(5), С. 100967 - 100967

Опубликована: Апрель 4, 2024

Existing antibody language models are limited by their use of unpaired sequence data. A recently published dataset ∼1.6 × 10

Язык: Английский

Процитировано

22
Disease diagnostics using machine learning of B cell and T cell receptor sequences DOI
Maxim Zaslavsky, Erin Craig, Jackson Michuda

и другие.

Science, Год журнала: 2025, Номер 387(6736)

Опубликована: Фев. 20, 2025

Clinical diagnosis typically incorporates physical examination, patient history, various laboratory tests, and imaging studies but makes limited use of the human immune system's own record antigen exposures encoded by receptors on B cells T cells. We analyzed receptor datasets from 593 individuals to develop MAchine Learning for Immunological Diagnosis, an interpretive framework screen multiple illnesses simultaneously or precisely test one condition. This approach detects specific infections, autoimmune disorders, vaccine responses, disease severity differences. Human-interpretable features model recapitulate known responses severe acute respiratory syndrome coronavirus 2, influenza, immunodeficiency virus, highlight antigen-specific receptors, reveal distinct characteristics systemic lupus erythematosus type-1 diabetes autoreactivity. analysis has broad potential scientific clinical interpretation responses.

Язык: Английский

Процитировано

3
Targeting neuraminidase: the next frontier for broadly protective influenza vaccines DOI
Nicholas C. Wu, Ali H. Ellebedy

Trends in Immunology, Год журнала: 2023, Номер 45(1), С. 11 - 19

Опубликована: Дек. 15, 2023

Язык: Английский

Процитировано

16
Reading the repertoire: Progress in adaptive immune receptor analysis using machine learning DOI Creative Commons

Timothy J O'Donnell,

Chakravarthi Kanduri, Giulio Isacchini

и другие.

Cell Systems, Год журнала: 2024, Номер 15(12), С. 1168 - 1189

Опубликована: Дек. 1, 2024

Язык: Английский

Процитировано

4
T-cell receptor specificity landscape revealed through de novo peptide design DOI Creative Commons
Gian Marco Visani,

Michael N. Pun,

Anastasia A. Minervina

и другие.

Опубликована: Март 6, 2025

T-cells play a key role in adaptive immunity by mounting specific responses against diverse pathogens. An effective binding between T-cell receptors (TCRs) and pathogen-derived peptides presented on Major Histocompatibility Complexes (MHCs) mediate an immune response. However, predicting these interactions remains challenging due to limited functional data reactivities. Here, we introduce computational approach predict TCR with MHC class I alleles, design novel immunogenic for specified TCR-MHC complexes. Our method leverages HERMES, structure-based, physics-guided machine learning model trained the protein universe amino acid preferences based local structural environments. Despite no direct training TCR-pMHC data, implicit physical reasoning HERMES enables us make accurate predictions of both affinities activities across viral epitopes cancer neoantigens, achieving up 72% correlation experimental data. Leveraging our recognition model, develop protocol de novo peptides. Through validation three systems targeting peptides, demonstrate that designs—with five substitutions from native sequence—activate at success rates 50%. Lastly, use generative framework quantify diversity peptide landscape various complexes, offering insights into specificity humans mice. provides platform neoantigen design, opening new paths vaccine development viruses cancer.

Язык: Английский

Процитировано

0
Leveraging Large Language Models to Predict Antibody Biological Activity Against Influenza A Hemagglutinin DOI Creative Commons
Ella Barkan,

I. Siddiqui,

Kevin J. Cheng

и другие.

Computational and Structural Biotechnology Journal, Год журнала: 2025, Номер unknown

Опубликована: Март 1, 2025

Язык: Английский

Процитировано

0
Supervised fine-tuning of pre-trained antibody language models improves antigen specificity prediction DOI Creative Commons
Meng Wang,

Jonathan Patsenker,

Henry Li

и другие.

bioRxiv (Cold Spring Harbor Laboratory), Год журнала: 2024, Номер unknown

Опубликована: Май 13, 2024

Abstract Antibodies play a crucial role in adaptive immune responses by determining B cell specificity to antigens and focusing function on target pathogens. Accurate prediction of antibody-antigen directly from antibody sequencing data would be great aid understanding responses, guiding vaccine design, developing antibody-based therapeutics. In this study, we present method supervised fine-tuning for language models, which improves previous results binding SARS-CoV-2 spike protein influenza hemagglutinin. We perform four pre-trained models predict these demonstrate that fine-tuned model classifiers exhibit enhanced predictive accuracy compared trained pretrained embeddings. The change attention activations after suggested performance was driven an increased focus the complementarity regions (CDRs). Application BCR repertoire demonstrated could recognize specific elicited vaccination. Overall, our study highlights benefits as mechanism improve antigen prediction. Author Summary are vigilant sentinels system bind targets foreign pathogens, known antigens. This interaction between is highly specific, akin fitting lock key mechanism, ensure each precisely its intended antigen. Recent advancements modeling have led development decode information sequences antibodies. introduce based fine-tuning, enhances predicting interactions. By training large datasets sequences, can better antibodies will important such those found surface viruses like influenza. Moreover, demonstrates potential “read” ongoing offering new insights into how bodies respond These findings significant implications accurate guide more effective vaccines.

Язык: Английский

Процитировано

2
Focused learning by antibody language models using preferential masking of non-templated regions DOI Creative Commons
Kok Mun Ng, Bryan Briney

bioRxiv (Cold Spring Harbor Laboratory), Год журнала: 2024, Номер unknown

Опубликована: Окт. 28, 2024

Existing antibody language models (

Язык: Английский

Процитировано

2
Rapid synthesis and screening of natively paired antibodies against influenza hemagglutinin stem via oPool+display DOI Creative Commons
Wenhao O. Ouyang, Huibin Lv, W. Liu

и другие.

bioRxiv (Cold Spring Harbor Laboratory), Год журнала: 2024, Номер unknown

Опубликована: Авг. 30, 2024

Antibody discovery is crucial for developing therapeutics and vaccines as well understanding adaptive immunity. However, the lack of approaches to synthesize antibodies with defined sequences in a high-throughput manner represents major bottleneck antibody discovery. Here, we presented oPool + display, cell-free platform that combined oligo pool synthesis mRNA display rapidly construct characterize many natively paired parallel. As proof-of-concept, applied probe binding specificity >300 uncommon influenza hemagglutinin (HA) against 9 HA variants through 16 different screens. Over 5,000 tests were performed 3-5 days further scaling potential. Follow-up structural analysis two stem revealed previously unknown versatility IGHD3-3 gene segment recognizing stem. Overall, this study established an experimental not only accelerate characterization, but also enable unbiased molecular signatures.

Язык: Английский

Процитировано

0