Institute of Systems and Physical Biology, Shenzhen Bay Laboratory, Shenzhen 518055, People's Republic of China.
Beijing Kein Research Center for Natural Sciences, Beijing 100022, People's Republic of China.
Phys Chem Chem Phys. 2021 Sep 22;23(36):20323-20328. doi: 10.1039/d1cp02805g.
The evaluation of molecular electrostatic potential (ESP) is a performance bottleneck for many computational chemical tasks like restrained ESP charge fitting or quantum mechanics/molecular mechanics simulations. In this paper, an efficient algorithm for the evaluation of ESP is proposed. It regroups the expression in terms of primitive Gaussian type orbitals (GTOs) with identical angular momentum types and nuclei centers. Each term is calculated using a computerized optimized code. This algorithm was integrated into the wavefunction analysis program Multiwfn and was tested on several large systems. In the cases of dopamine and remdesivir, the performance of this algorithm was comparable to or better than some popular state-of-the-art codes. For meta1-organic framework-5, where the number of GTOs and ESP points is 4840 and 259 262, respectively, our code could finish the evaluation in 1874 seconds on ordinary hardware. It also exhibits good parallelization scaling. The source code of this algorithm is freely available and can become a useful tool for computational chemists.
分子静电势（ESP）的评估是许多计算化学任务的性能瓶颈，例如受约束的 ESP 电荷拟合或量子力学/分子力学模拟。在本文中，提出了一种用于 ESP 评估的有效算法。它根据具有相同角动量类型和核中心的原始高斯型轨道（GTO）重新组合表达式。每个术语都使用计算机优化代码进行计算。该算法已集成到波函数分析程序 Multiwfn 中，并在几个大型系统上进行了测试。在多巴胺和瑞德西韦的情况下，该算法的性能可与一些流行的最先进代码相媲美或优于它们。对于元 1-有机骨架-5，其 GTO 数量和 ESP 点数分别为 4840 和 259262，我们的代码可以在普通硬件上在 1874 秒内完成评估。它还具有良好的并行扩展能力。该算法的源代码是免费提供的，它可以成为计算化学家的有用工具。
