Wang Liyao, Tučs Andrejs, Ding Songting, Tsuda Koji, Sljoka Adnan
Graduate School of Frontier Sciences, The University of Tokyo, Kashiwa 277-8561, Japan.
RIKEN Center for Advanced Intelligence Project, RIKEN, Tokyo 103-0027, Japan.
J Chem Theory Comput. 2025 Jul 22;21(14):7173-7187. doi: 10.1021/acs.jctc.5c00175. Epub 2025 May 14.
Accurate modeling of protein-protein complex structures is essential for understanding biological mechanisms. Hydrogen-deuterium exchange (HDX) experiments provide valuable insights into binding interfaces. Incorporating HDX data into protein complex modeling workflows offers a promising approach to improve prediction accuracy. Here, we developed HDXRank, a graph neural network (GNN)-based framework for candidate structure ranking utilizing alignment with HDX experimental data. Trained on a newly curated HDX data set, HDXRank captures nuanced local structural features critical for accurate HDX profile prediction. This versatile framework can be integrated with a variety of protein complex modeling tools, transforming the HDX profile alignment into a model quality metric. HDXRank demonstrates effectiveness at ranking models generated by rigid docking or AlphaFold, successfully prioritizing functionally relevant models and improving prediction quality across all tested protein targets. These findings underscore HDXRank's potential to become a pivotal tool for understanding molecular recognition in complex biological systems.
准确模拟蛋白质-蛋白质复合物结构对于理解生物学机制至关重要。氢-氘交换（HDX）实验为结合界面提供了有价值的见解。将HDX数据纳入蛋白质复合物建模工作流程为提高预测准确性提供了一种有前景的方法。在此，我们开发了HDXRank，这是一个基于图神经网络（GNN）的框架，用于利用与HDX实验数据的比对对候选结构进行排名。HDXRank在一个新整理的HDX数据集上进行训练，捕捉对准确预测HDX谱至关重要的细微局部结构特征。这个通用框架可以与各种蛋白质复合物建模工具集成，将HDX谱比对转化为模型质量指标。HDXRank在对刚性对接或AlphaFold生成的模型进行排名时显示出有效性，成功地对功能相关模型进行了优先排序，并提高了所有测试蛋白质靶点的预测质量。这些发现强调了HDXRank成为理解复杂生物系统中分子识别的关键工具的潜力。
