Utilizing Machine Learning to Greatly Expand the Range and Accuracy of Bottom-Up Coarse-Grained Models through Virtual Particles
Utilizing Machine Learning to Greatly Expand the Range and Accuracy of Bottom-Up Coarse-Grained Models through Virtual Particles
复制标题
DOI：
10.1021/acs.jctc.2c01183
复制
发表时间：
2023-02-20
影响因子：
5.5
通讯作者：
Voth，Gregory A.
中科院分区：
文献类型：
--
作者：
Sahrmann，Patrick G.;Loose，Timothy D.;Voth，Gregory A.
使用原子参考数据参数化的粗粒度 (CG) 模型，即“自下而上”的 CG 模型，已被证明在生物分子和其他软物质的研究中非常有用。然而，构建高精度、低分辨率的生物分子 CG 模型仍然具有挑战性。我们在这项工作中演示了如何在相对熵最小化 (REM) 的背景下将虚拟粒子（没有原子对应关系的 CG 位置）作为潜在变量纳入 CG 模型中。所提出的方法，变分导数相对熵最小化（VD-REM），可以通过机器学习辅助的梯度下降算法来优化虚拟粒子相互作用。我们将此方法应用于 1,2-二油酰-sn-甘油-3-磷酸胆碱 (DOPC) 脂质双层的无溶剂 CG 模型的挑战性案例，并证明虚拟粒子的引入捕获了溶剂介导的行为和高阶相关性，而仅靠 REM 无法在仅基于原子集合到 CG 位点的映射的更标准的 CG 模型中捕获这些行为和高阶相关性。
Coarse-grained (CG) models parametrized using atomistic reference data, i.e., “bottom up” CG models, have proven useful in the study of biomolecules and other soft matter. However, the construction of highly accurate, low resolution CG models of biomolecules remains challenging. We demonstrate in this work how virtual particles, CG sites with no atomistic correspondence, can be incorporated into CG models within the context of relative entropy minimization (REM) as latent variables. The methodology presented, variational derivative relative entropy minimization (VD-REM), enables optimization of virtual particle interactions through a gradient descent algorithm aided by machine learning. We apply this methodology to the challenging case of a solvent-free CG model of a 1,2-dioleoyl-sn-glycero-3-phosphocholine (DOPC) lipid bilayer and demonstrate that introduction of virtual particles captures solvent-mediated behavior and higher-order correlations which REM alone cannot capture in a more standard CG model based only on the mapping of collections of atoms to the CG sites.