核酸
力场(虚构)
分子间力
极化率
偶极子
缩放比例
化学
化学物理
计算化学
拓扑(电路)
电荷(物理)
生物系统
可扩展性
量子
电负性
领域(数学)
叠加原理
量子化学
统计物理学
静电学
生物分子
分子动力学
极化(电化学)
分子
计算机科学
计算模型
物理
势能
参数化(大气建模)
量子化学
价(化学)
电荷密度
计算
作者
Qingyan Sun,Xue-Lu Jiang,Xin Guo,Hong Huang,Jian Fei Zhao,Dong‐Xia Zhao,Zhong‐Zhi Yang
标识
DOI:10.1021/acs.jpca.6c03817
摘要
Accurate characterization of the complex electrostatic and polarization environments in nucleic acid-based systems is essential for reliable biomolecular simulations, where atomic charges play a central role. In this work, charges derived from the Kohn-Sham one-electron potential (KSpot), characterized by well-defined real-space topological boundaries, are incorporated into the atom-bond electronegativity equalization method (ABEEM) framework to balance quantum chemical accuracy and computational efficiency. Through systematic parametrization using fragment molecules representing diverse chemical environments, we developed a charge model for nucleic acid systems, termed KS-ABEEM. Validation results demonstrate that KS-ABEEM accurately reproduces KSpot real-space topological charges, yielding linear correlation coefficients of 0.9862 and 0.9885 for the training and test sets, respectively. The model also reliably predicts molecular dipole moments and electrostatically dominated intermolecular interaction energies, showing good agreement with quantum chemical calculations. Furthermore, KS-ABEEM avoids the high computational cost associated with conventional real-space topological integration methods. Compared with traditional topological approaches, the proposed KS-ABEEM achieves computational speedups of 102-104 while maintaining favorable computational scaling for large systems. Overall, this work provides an efficient route for extending quantum chemical topology-based charge models toward nucleic acid systems. Validation is presented for representative fragment molecules, hydrated clusters, and base-pair systems, while the computational efficiency and scalability of the approach for large all-atom nucleic acid structures are demonstrated through timing benchmarks, highlighting the potential of real-space topological descriptors for biomolecular simulations and polarizable force field development.
科研通智能强力驱动
Strongly Powered by AbleSci AI