BioGeometry, Beijing, China.
Mila-Québec AI Institute, Montréal, QC, Canada.
Nat Commun. 2024 Sep 6;15(1):7785. doi: 10.1038/s41467-024-51563-8.
Increasing the binding affinity of an antibody to its target antigen is a crucial task in antibody therapeutics development. This paper presents a pretrainable geometric graph neural network, GearBind, and explores its potential in in silico affinity maturation. Leveraging multi-relational graph construction, multi-level geometric message passing and contrastive pretraining on mass-scale, unlabeled protein structural data, GearBind outperforms previous state-of-the-art approaches on SKEMPI and an independent test set. A powerful ensemble model based on GearBind is then derived and used to successfully enhance the binding of two antibodies with distinct formats and target antigens. ELISA EC values of the designed antibody mutants are decreased by up to 17 fold, and K values by up to 6.1 fold. These promising results underscore the utility of geometric deep learning and effective pretraining in macromolecule interaction modeling tasks.
提高抗体与其靶抗原的结合亲和力是抗体治疗药物开发中的关键任务。本文提出了一种可预训练的几何图神经网络 GearBind，并探讨了其在计算亲和力成熟中的潜力。利用多关系图构建、多层次几何消息传递和基于大规模、无标签蛋白质结构数据的对比预训练，GearBind 在 SKEMPI 和独立测试集上的表现优于以前的最先进方法。然后，基于 GearBind 衍生出一个强大的集成模型，并成功地增强了两种具有不同形式和靶抗原的抗体的结合。设计的抗体突变体的 ELISA EC 值降低了多达 17 倍，K 值降低了多达 6.1 倍。这些有希望的结果突出了几何深度学习和有效预训练在大分子相互作用建模任务中的实用性。
