Shimagaki Kai S, Kher Gargi, Lynch Rebecca M, Barton John P
Department of Computational and Systems Biology, University of Pittsburgh School of Medicine, USA.
Department of Physics and Astronomy, University of Pittsburgh, USA.
bioRxiv. 2025 Aug 11:2025.08.08.669352. doi: 10.1101/2025.08.08.669352.
For genetically variable pathogens such as human immunodeficiency virus (HIV)-1, individual viral isolates can differ dramatically in their sensitivity to antibodies. The ability to predict which viruses will be sensitive and which will be resistant to a specific antibody could aid in the design of antibody therapies and help illuminate resistance evolution. Due to the enormous number of possible combinations, it is not possible to experimentally measure neutralization values for all pairs of viruses and antibodies. Here, we developed a simple and interpretable method called grouped neutralization learning (GNL) to predict neutralization values by leveraging viral genetic sequences and similarities in neutralization profiles between antibodies. Our method compares favorably to state-of-the-art approaches and is robust to model parameter assumptions. GNL can predict neutralization values for viruses with no observed data, an essential capability for evaluating novel viral strains. We also demonstrate that GNL can successfully transfer knowledge between independent data sets, allowing rapid estimates of viral sensitivity based on prior knowledge.
对于像人类免疫缺陷病毒1型（HIV-1）这样基因可变的病原体，单个病毒分离株对抗体的敏感性可能有很大差异。预测哪些病毒对特定抗体敏感以及哪些病毒具有抗性的能力，有助于抗体疗法的设计，并有助于阐明抗性进化。由于可能的组合数量巨大，不可能通过实验测量所有病毒与抗体对的中和值。在此，我们开发了一种简单且可解释的方法，称为分组中和学习（GNL），通过利用病毒基因序列和抗体之间中和谱的相似性来预测中和值。我们的方法优于现有技术方法，并且对模型参数假设具有鲁棒性。GNL可以预测没有观测数据的病毒的中和值，这是评估新型病毒株的一项基本能力。我们还证明，GNL能够在独立数据集之间成功转移知识，从而基于先验知识快速估计病毒敏感性。
