Suppr超能文献

使用副本交换包络分布采样实现具有量子力学/分子力学精度的高效多态自由能计算。

Efficient Multistate Free-Energy Calculations with QM/MM Accuracy Using Replica-Exchange Enveloping Distribution Sampling.

作者信息

Pregeljc Domen, Hügli Ramon J R, Riniker Sereina

机构信息

Department of Chemistry and Applied Biosciences, ETH Zürich, Vladimir-Prelog-Weg 2, 8093 Zürich, Switzerland.

出版信息

J Phys Chem B. 2025 Jun 19;129(24):5948-5960. doi: 10.1021/acs.jpcb.5c02086. Epub 2025 Jun 6.

Abstract

Calculating free-energy differences using molecular dynamics (MD) simulations is an important task in computational chemistry. In practice, the accuracy of the results is limited by model approximations and insufficient phase-space sampling due to limited computational resources. In the present work, we address these challenges by integrating the quantum-mechanical/molecular-mechanical (QM/MM) scheme with replica-exchange enveloping distribution sampling (RE-EDS) to obtain a multistate and multiscale free-energy method with high computational efficiency. The performance of QM/MM RE-EDS is showcased by calculating hydration free energies for three data sets using semiempirical methods for the QM zone. We highlight the importance of the choice of QM Hamiltonian and the effect of the compatibility between the QM and MM models. Especially the choice of semiempirical method has a substantial effect on the accuracy compared to experiment, but also the choice of MM water model is non-negligible. Our findings indicate that RE-EDS is an efficient approach for calculating free-energy differences with a QM/MM scheme, and lays the foundation for future developments and applications.

摘要

利用分子动力学(MD)模拟计算自由能差是计算化学中的一项重要任务。在实际应用中,由于计算资源有限,结果的准确性受到模型近似和相空间采样不足的限制。在本工作中,我们通过将量子力学/分子力学(QM/MM)方案与副本交换包络分布采样(RE-EDS)相结合来应对这些挑战,从而获得一种具有高计算效率的多态多尺度自由能方法。通过使用半经验方法对QM区域计算三个数据集的水化自由能,展示了QM/MM RE-EDS的性能。我们强调了QM哈密顿量选择的重要性以及QM和MM模型之间兼容性的影响。特别是与实验相比,半经验方法的选择对准确性有很大影响,而且MM水模型的选择也不可忽视。我们的研究结果表明,RE-EDS是一种使用QM/MM方案计算自由能差的有效方法,并为未来的发展和应用奠定了基础。

https://cdn.ncbi.nlm.nih.gov/pmc/blobs/c474/12183765/bbb2376c023f/jp5c02086_0001.jpg

文献AI研究员

20分钟写一篇综述,助力文献阅读效率提升50倍。

立即体验

用中文搜PubMed

大模型驱动的PubMed中文搜索引擎

马上搜索

文档翻译

学术文献翻译模型,支持多种主流文档格式。

立即体验