Xin Minghui, Wang Zechen, Wang Zhihao, Qu Yuanyuan, Yang Yanmei, Li Yong-Qiang, Zhao Mingwen, Zheng Liangzhen, Mu Yuguang, Li Weifeng
School of Physics, Shandong University, Jinan 250100, China.
College of Chemistry, Chemical Engineering and Materials Science, Collaborative Innovation Centre of Functionalized Probes for Chemical Imaging in Universities of Shandong, Key Laboratory of Molecular and Nano Probes, Ministry of Education, Shandong Normal University, Jinan 250014, China.
J Chem Inf Model. 2025 Jan 13;65(1):41-49. doi: 10.1021/acs.jcim.4c01096. Epub 2024 Dec 26.
In recent years, the deep learning (DL) technique has rapidly developed and shown great success in scoring the protein-ligand binding affinities. The protein-ligand conformation optimization based on DL-derived scoring functions holds broad application prospects, for instance, drug design and enzyme engineering. In this study, we evaluated the robustness of a DL-based ligand conformation optimization protocol (DeepRMSD+Vina) for optimizing structures with input perturbations by examining the predicted ligand binding poses and scoring. Our results clearly indicated that compared to traditional optimization algorithms (such as Prime MM-GBSA and Vina optimization), DeepRMSD+Vina exhibits higher performance when treating diverse protein-ligand cases. The DeepRMSD+Vina is robust and can always generate the correct binding structures even if perturbations (up to 3 Å) are introduced to the input structure. The success rate is 62% for perturbation with a RMSD within 2-3 Å. However, the success rate dramatically drops to 11% for large perturbations, with RMSD extending to 3-4 Å. Furthermore, compared to the widely used optimization protocol of AutoDock Vina, the DL-generated conformation shows a balanced performance for all of the systems under examination. Overall, the DL-based DeepRMSD+Vina is unarguably more reliable than the traditional methods, which is attributed to the physically inspired design of the neural networks in DeepRMSD+Vina where the distance-transformed features describing the atomic interactions between the protein and the ligand have been explicitly considered and modeled. The outstanding robustness of the DL-based ligand conformational optimization algorithm further validates its superiority in the field of conformational optimization.
近年来，深度学习（DL）技术迅速发展，并在预测蛋白质 - 配体结合亲和力方面取得了巨大成功。基于深度学习衍生评分函数的蛋白质 - 配体构象优化具有广阔的应用前景，例如药物设计和酶工程。在本研究中，我们通过检查预测的配体结合姿势和评分，评估了基于深度学习的配体构象优化协议（DeepRMSD + Vina）在处理输入扰动时优化结构的稳健性。我们的结果清楚地表明，与传统优化算法（如Prime MM - GBSA和Vina优化）相比，DeepRMSD + Vina在处理各种蛋白质 - 配体情况时表现出更高的性能。DeepRMSD + Vina具有稳健性，即使对输入结构引入扰动（高达3 Å），也总能生成正确的结合结构。对于均方根偏差（RMSD）在2 - 3 Å范围内的扰动，成功率为62%。然而，对于RMSD扩展到3 - 4 Å的大扰动，成功率急剧下降至11%。此外，与广泛使用的AutoDock Vina优化协议相比，深度学习生成的构象在所有受检系统中表现出平衡的性能。总体而言，基于深度学习的DeepRMSD + Vina无疑比传统方法更可靠，这归因于DeepRMSD + Vina中神经网络的物理启发式设计，其中明确考虑并建模了描述蛋白质和配体之间原子相互作用的距离变换特征。基于深度学习的配体构象优化算法的出色稳健性进一步验证了其在构象优化领域的优越性。
