A web-based predictor for SARS-CoV-2 neutralizing antibody escape from genome sequences

Lanxin Zou; Yixuan Miao; Rui Liu; Daihai He; Lanxin Zou; Yixuan Miao; Rui Liu; Daihai He

doi:10.3934/bdia.2026017

Big Data and Information Analytics

2026, Volume 10, 338-349. doi: 10.3934/bdia.2026017

Previous Article Next Article

Research article

A web-based predictor for SARS-CoV-2 neutralizing antibody escape from genome sequences

1.
Department of Applied Mathematics, The Hong Kong Polytechnic University, Hong Kong SAR, China
2.
School of Mathematics, South China University of Technology, Guangzhou 510640, China

Received: 03 October 2026 Accepted: 13 October 2026 Published: 04 November 2026

SARS-CoV-2 variants can carry spike mutations that reduce the activity of therapeutic antibodies. Laboratory neutralization assays provide direct measurements, but they are too slow and resource-intensive for rapid screening of every newly sequenced genome. We developed a web application that accepts FASTA genomes, checks SARS-CoV-2 identity, assigns one of six major variant categories, detects predefined spike substitutions, and predicts IC₅₀ for Casirivimab and Imdevimab. Two ridge regression models were trained on an expanded CoV-DRDB dataset, and the arithmetic mean of their outputs is reported as the cocktail estimate. Spike mutations are called by protein alignment to Wuhan-Hu-1, with offset voting used as a fallback. On a held-out pipeline test set of seven NCBI genomes, both major-category identification and key-mutation recall were 100%. Identification was also 100% on seven external NCBI genomes that were not part of the website reference panel. For ten single-antibody measurements from genomes not used in training, excluding assay-ceiling values, the mean absolute error (MAE) on log$ _{10} $ fold-change was 0.45. On a separate machine-learning test set containing 40 antibody measurements from 20 spike isoforms, the overall MAE was 0.50 and Pearson's $ r $ was 0.87. The application is intended for research and teaching, not clinical decision-making.
- SARS-CoV-2,
- monoclonal antibody,
- IC₅₀ prediction,
- spike mutations,
- ridge regression
Citation: Lanxin Zou, Yixuan Miao, Rui Liu, Daihai He. A web-based predictor for SARS-CoV-2 neutralizing antibody escape from genome sequences[J]. Big Data and Information Analytics, 2026, 10: 338-349. doi: 10.3934/bdia.2026017

Related Papers:

Abstract

SARS-CoV-2 variants can carry spike mutations that reduce the activity of therapeutic antibodies. Laboratory neutralization assays provide direct measurements, but they are too slow and resource-intensive for rapid screening of every newly sequenced genome. We developed a web application that accepts FASTA genomes, checks SARS-CoV-2 identity, assigns one of six major variant categories, detects predefined spike substitutions, and predicts IC₅₀ for Casirivimab and Imdevimab. Two ridge regression models were trained on an expanded CoV-DRDB dataset, and the arithmetic mean of their outputs is reported as the cocktail estimate. Spike mutations are called by protein alignment to Wuhan-Hu-1, with offset voting used as a fallback. On a held-out pipeline test set of seven NCBI genomes, both major-category identification and key-mutation recall were 100%. Identification was also 100% on seven external NCBI genomes that were not part of the website reference panel. For ten single-antibody measurements from genomes not used in training, excluding assay-ceiling values, the mean absolute error (MAE) on log$ _{10} $ fold-change was 0.45. On a separate machine-learning test set containing 40 antibody measurements from 20 spike isoforms, the overall MAE was 0.50 and Pearson's $ r $ was 0.87. The application is intended for research and teaching, not clinical decision-making.

References

[1]	Wu F, Zhao S, Yu B, Chen YM, Wang W, Song ZG, et al. (2020) A new coronavirus associated with human respiratory disease in China. Nature 579: 265–269. https://doi.org/10.1038/s41586-020-2008-3 doi: 10.1038/s41586-020-2008-3
[2]	Harvey WT, Carabelli AM, Jackson B, Gupta RK, Thomson EC, Harrison EM, et al. (2021) SARS-CoV-2 variants, spike mutations and immune escape. Nat Rev Microbiol 19: 409–424. https://doi.org/10.1038/s41579-021-00573-0 doi: 10.1038/s41579-021-00573-0
[3]	Korber B, Fischer WM, Gnanakaran S, Yoon H, Theiler J, Abfalterer W, et al. (2020) Tracking changes in SARS-CoV-2 Spike: Evidence that D614G increases infectivity of the COVID-19 virus. Cell 182: 812–827.e19. https://doi.org/10.1016/j.cell.2020.06.043 doi: 10.1016/j.cell.2020.06.043
[4]	Viana R, Moyo S, Amoako DG, Tegally H, Scheepers C, Althaus CL, et al. (2022) Rapid epidemic expansion of the SARS-CoV-2 Omicron variant in southern Africa. Nature 603: 679–686. https://doi.org/10.1038/s41586-022-04411-y doi: 10.1038/s41586-022-04411-y
[5]	Greaney AJ, Starr TN, Gilchuk P, Zost SJ, Binshtein E, Loes AN, et al. (2021) Complete mapping of mutations to the SARS-CoV-2 spike receptor-binding domain that escape antibody recognition. Cell Host Microbe 29: 44–57.e9. https://doi.org/10.1016/j.chom.2020.11.007 doi: 10.1016/j.chom.2020.11.007
[6]	Sayers EW, Cavanaugh M, Clark K, Pruitt KD, Schoch CL, Sherry ST, et al. (2022) GenBank. Nucleic Acids Res 50: D161–D164. https://doi.org/10.1093/nar/gkab1135 doi: 10.1093/nar/gkab1135
[7]	Tzou PL, Tao K, Pond SLK, Shafer RW.(2022) Coronavirus Resistance Database (CoV-RDB): SARS-CoV-2 susceptibility to monoclonal antibodies, convalescent plasma, and plasma from vaccinated persons. PLoS ONE 17: e0261045. https://doi.org/10.1371/journal.pone.0261045 doi: 10.1371/journal.pone.0261045
[8]	Cock PJA, Antao T, Chang JT, Chapman BA, Cox CJ, Dalke A, et al. (2009) Biopython: Freely available Python tools for computational molecular biology and bioinformatics. Bioinformatics 25: 1422–1423. https://doi.org/10.1093/bioinformatics/btp163 doi: 10.1093/bioinformatics/btp163
[9]	Pedregosa F, Varoquaux G, Gramfort A, Michel V, Thirion B, Grisel O, et al. Scikit-learn: Machine learning in Python. J Mach Learn Res 12: 2825–2830. https://jmlr.org/papers/v12/pedregosa11a.html
[10]	Hoerl AE, Kennard RW, (1970) Ridge regression: biased estimation for nonorthogonal problems. Technometrics 12: 55–67. https://doi.org/10.1080/00401706.1970.10488634 doi: 10.1080/00401706.1970.10488634
[11]	Gotoh O, (1982) An improved algorithm for matching biological sequences. J Mol Biol, 162: 705–708. https://doi.org/10.1016/0022-2836(82)90398-9 doi: 10.1016/0022-2836(82)90398-9
[12]	Wang P, Nair MS, Liu L, Iketani S, Luo Y, Guo Y, et al. (2021) Antibody resistance of SARS-CoV-2 variants B.1.351 and B.1.1.7. Nature 593: 130–135. https://doi.org/10.1038/s41586-021-03398-2 doi: 10.1038/s41586-021-03398-2
[13]	Cao Y, J Wang, F Jian, T Xiao, W Song, A Yisimayi, et al. (2022) Omicron escapes the majority of existing SARS-CoV-2 neutralizing antibodies. Nature 602: 657–663. https://doi.org/10.1038/s41586-021-04385-3 doi: 10.1038/s41586-021-04385-3
[14]	Hansen J, Baum A, Pascal KE, Russo V, Giordano S, Wloga E, et al. (2020) Studies in humanized mice and convalescent humans yield a SARS-CoV-2 antibody cocktail. Science 369: 1010–1014. https://doi.org/10.1126/science.abd0827 doi: 10.1126/science.abd0827
[15]	Baum A, Fulton BO, Wloga E, Copin R, Pascal KE, Russo V, et al. (2020) Antibody cocktail to SARS-CoV-2 spike protein prevents rapid mutational escape seen with individual antibodies. Science 369: 1014–1018. https://doi.org/10.1126/science.abd0831 doi: 10.1126/science.abd0831
[16]	Planas D, et al. (2021) Reduced sensitivity of SARS-CoV-2 variant Delta to antibody neutralization. Nature 596 : 276–280. https://doi.org/10.1038/s41586-021-03777-9 doi: 10.1038/s41586-021-03777-9
[17]	VanBlargan LA, Errico JM, Halfmann PJ, Zost SJ, Crowe Jr JE, Purcell LA, et al. (2022) An infectious SARS-CoV-2 B.1.1.529 Omicron virus escapes neutralization by therapeutic monoclonal antibodies. Nat Med 28: 490–495. https://doi.org/10.1038/s41591-021-01678-y doi: 10.1038/s41591-021-01678-y

Reader Comments

Your name:*

Email:*
© 2026 the Author(s), licensee AIMS Press. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0)