Data from "Signatures of a liquid-liquid transition in an ab initio deep neural network model for water"

The DNN model for water was generated using the DP-GEN methodology (https://github.com/deepmodeling/dpgen) and DeepMD-kit v1.0 (https://github.com/deepmodeling/deepmd-kit) software. Simulations were performed using the LAMMPS molecular simulation software v7Aug19 (https://lammps.sandia.gov/) patched with the Plumed 2.0 advanced sampling software (https://www.plumed.org/). Detailed methods are as described in Gartner et al, (https://doi.org/10.1073/pnas.2015440117). Detailed installation instructions for all software are available at the above linked webpages. Please download the data sets from Dropbox using https://www.dropbox.com/sh/o883p0xjgcz92y6/AAD4zmPAP-5h77npA4kKcBFva?dl=0

本研究中用于水体系的深度神经网络（Deep Neural Network, DNN）模型通过DP-GEN方法（https://github.com/deepmodeling/dpgen）与DeepMD-kit v1.0软件（https://github.com/deepmodeling/deepmd-kit）构建生成。模拟计算采用搭载了Plumed 2.0高级采样软件（https://www.plumed.org/）补丁的LAMMPS v7Aug19分子模拟软件（https://lammps.sandia.gov/）完成。详细的研究方法详见Gartner等人的相关研究（https://doi.org/10.1073/pnas.2015440117）。所有软件的详细安装说明均可通过上述对应链接网页获取。请通过以下链接从Dropbox平台下载本数据集：https://www.dropbox.com/sh/o883p0xjgcz92y6/AAD4zmPAP-5h77npA4kKcBFva?dl=0