Generation of antigen-specific paired-chain antibodies using large language models

The traditional process of antibody discovery is limited by inefficiency, high costs, and low success rates. Recent approaches employing artificial intelligence (AI) have been developed to optimize existing antibodies and generate antibody sequences in a target-agnostic manner. In this work, we present MAGE (monoclonal antibody generator), a sequence-based protein language model (PLM) fine-tuned for the task of generating paired human variable heavy- and light-chain antibody sequences against targets of interest. We show that MAGE can generate novel and diverse antibody sequences with experimentally validated binding specificity against SARS-CoV-2, an emerging avian influenza H5N1, and respiratory syncytial virus A (RSV-A). MAGE represents a first-in-class model capable of designing human antibodies against multiple targets with no starting template.

Authors

Institutions

Publication Details

Journal
Cell
Published
2025-11-04
DOI
https://doi.org/10.1016/j.cell.2025.10.006
Citations
7
Primary Topic
Monoclonal and Polyclonal Antibodies Research
Type
article
Field-Weighted Citation Impact
9.52
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

Generation of antigen-specific paired-chain antibodies using large language models

Gwen Jordaan, Jason S. McLellan, Clinton Holt, Giuseppe A. Sautto et al.
7 citations
Cell
Monoclonal and Polyclonal Antibodies Research
9.52
article

Generation of antigen-specific paired-chain antibodies using large language models

Gwen Jordaan, Jason S. McLellan, Clinton Holt, Giuseppe A. Sautto, Daniel J. Sheward, Matthew J. Vukovich, Ted M. Ross, Jennifer K. Logue, Sarah F. Andrews, Perry T. Wasdin, Suzanne Held, Grant S. Hansman, Masaru Kanekiyo, Nicole V. Johnson, Jeong-Ryeol Kim, Alexis K. Janke, Fani Pantouli, Rebecca A. Gillespie, Ivelin S. Georgiev, Toma M. Marinov, Olivia Powers, Alexandra A. Abu-Shmais, Helen Y. Chu, Léna Vandenabeele
article en
7 citations

Abstract

The traditional process of antibody discovery is limited by inefficiency, high costs, and low success rates. Recent approaches employing artificial intelligence (AI) have been developed to optimize existing antibodies and generate antibody sequences in a target-agnostic manner. In this work, we present MAGE (monoclonal antibody generator), a sequence-based protein language model (PLM) fine-tuned for the task of generating paired human variable heavy- and light-chain antibody sequences against targets of interest. We show that MAGE can generate novel and diverse antibody sequences with experimentally validated binding specificity against SARS-CoV-2, an emerging avian influenza H5N1, and respiratory syncytial virus A (RSV-A). MAGE represents a first-in-class model capable of designing human antibodies against multiple targets with no starting template.

CellVol. 188(25)
Griffith University (AU), National Institutes of Health (US), University of Washington (US), Karolinska Institutet (SE), Cleveland Clinic Florida (US), National Institute of Allergy and Infectious Diseases (US), The University of Texas at Austin (US), Vanderbilt University Medical Center (US)
Openalex Percentile: Top 2%
Monoclonal and Polyclonal Antibodies Research
9.52
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.