Center for Computational Biology and Department of Molecular Biosciences, University of Kansas, Lawrence, Kansas 66047, United States.
J Phys Chem Lett. 2023 Jun 1;14(21):4970-4982. doi: 10.1021/acs.jpclett.3c00926. Epub 2023 May 23.
We have developed a new deep boosted molecular dynamics (DBMD) method. Probabilistic Bayesian neural network models were implemented to construct boost potentials that exhibit Gaussian distribution with minimized anharmonicity, thereby allowing for accurate energetic reweighting and enhanced sampling of molecular simulations. DBMD was demonstrated on model systems of alanine dipeptide and the fast-folding protein and RNA structures. For alanine dipeptide, 30 ns DBMD simulations captured up to 83-125 times more backbone dihedral transitions than 1 μs conventional molecular dynamics (cMD) simulations and were able to accurately reproduce the original free energy profiles. Moreover, DBMD sampled multiple folding and unfolding events within 300 ns simulations of the chignolin model protein and identified low-energy conformational states comparable to previous simulation findings. Finally, DBMD captured a general folding pathway of three hairpin RNAs with the GCAA, GAAA, and UUCG tetraloops. Based on a deep learning neural network, DBMD provides a powerful and generally applicable approach to boosting biomolecular simulations. DBMD is available with open source in OpenMM at https://github.com/MiaoLab20/DBMD/.
我们开发了一种新的深度增强分子动力学（DBMD）方法。我们实现了概率贝叶斯神经网络模型来构建增强势，这些增强势表现出最小非谐性的高斯分布，从而能够实现分子模拟的精确能量重加权和增强采样。我们在丙氨酸二肽和快速折叠蛋白和 RNA 结构的模型系统上验证了 DBMD。对于丙氨酸二肽，30 ns 的 DBMD 模拟比 1 μs 的传统分子动力学（cMD）模拟多捕捉到 83-125 倍的主链二面角转变，并且能够准确地再现原始的自由能曲线。此外，DBMD 在 300 ns 的 chignolin 模型蛋白模拟中采样了多个折叠和展开事件，并识别出与之前模拟结果相媲美的低能构象状态。最后，DBMD 捕获了具有 GCAA、GAAA 和 UUCG 四核苷酸环的三个发夹 RNA 的一般折叠途径。基于深度学习神经网络，DBMD 为增强生物分子模拟提供了一种强大且普遍适用的方法。DBMD 可在 OpenMM 中通过 https://github.com/MiaoLab20/DBMD/ 获得开源。
