@article{46a6b09818dd493f89359f42548c0ebc,
title = "An explainable language model for antibody specificity prediction using curated influenza hemagglutinin antibodies",
abstract = "Despite decades of antibody research, it remains challenging to predict the specificity of an antibody solely based on its sequence. Two major obstacles are the lack of appropriate models and the inaccessibility of datasets for model training. In this study, we curated >5,000 influenza hemagglutinin (HA) antibodies by mining research publications and patents, which revealed many distinct sequence features between antibodies to HA head and stem domains. We then leveraged this dataset to develop a lightweight memory B cell language model (mBLM) for sequence-based antibody specificity prediction. Model explainability analysis showed that mBLM could identify key sequence features of HA stem antibodies. Additionally, by applying mBLM to HA antibodies with unknown epitopes, we discovered and experimentally validated many HA stem antibodies. Overall, this study not only advances our molecular understanding of the antibody response to the influenza virus but also provides a valuable resource for applying deep learning to antibody research.",
keywords = "antibody, data mining, deep learning, hemagglutinin, influenza virus, language model, somatic hypermutations",
author = "Yiquan Wang and Huibin Lv and Teo, {Qi Wen} and Ruipeng Lei and Gopal, {Akshita B.} and Ouyang, {Wenhao O.} and Yeung, {Yuen Hei} and Tan, {Timothy J.C.} and Danbi Choi and Shen, {Ivana R.} and Xin Chen and Graham, {Claire S.} and Wu, {Nicholas C.}",
note = "This work was supported by National Institutes of Health (NIH) DP2 AT011966 (N.C.W.), R01 AI167910 (N.C.W.), the Michelson Prizes for Human Immunology and Vaccine Research (N.C.W.), the Searle Scholars Program (N.C.W.), and Carl R. Woese Institute for Genomic Biology Postdoctoral Fellowship (H.L.). We thank Kristen Flatt at the UIUC Materials Research Laboratory Central Research Facilities for assistance with cryo-EM experiments, as well as Meng Yuan and Zongjun Mou for helpful discussions. All authors conceived and designed the study. Y.W. H.L. and N.C.W. assembled the dataset. Y.W. Y.-H.Y. and N.C.W. performed data analysis. H.L. Q.W.T. A.B.G. W.O.O. T.J.C.T. D.C. and I.R.S. performed the antibody-binding experiments. R.L. C.S.G. and X.C. purified the proteins and performed the cryo-EM analysis. Y.W. H.L. R.L. and N.C.W. wrote the paper and all authors reviewed and/or edited the paper. N.C.W. consults for HeliXon. This work was supported by National Institutes of Health (NIH) DP2 AT011966 (N.C.W.), R01 AI167910 (N.C.W.), the Michelson Prizes for Human Immunology and Vaccine Research (N.C.W.), the Searle Scholars Program (N.C.W.), and Carl R. Woese Institute for Genomic Biology Postdoctoral Fellowship (H.L.). We thank Kristen Flatt at the UIUC Materials Research Laboratory Central Research Facilities for assistance with cryo-EM experiments, as well as Meng Yuan and Zongjun Mou for helpful discussions.",
year = "2024",
month = oct,
day = "8",
doi = "10.1016/j.immuni.2024.07.022",
language = "English (US)",
volume = "57",
pages = "2453--2465.e7",
journal = "Immunity",
issn = "1074-7613",
publisher = "Cell Press",
number = "10",
}