Are you sure you want to delete this task? Once this task is deleted, it cannot be recovered.
Li Qingguo 3ee595d1ff | 9 months ago | |
---|---|---|
.. | ||
README.md | 3 months ago | |
case1.pdb | 3 months ago | |
case1_addH.pdb | 3 months ago | |
tutorial_c01.ipynb | 3 months ago | |
tutorial_c01.py | 3 months ago | |
tutorial_c02.ipynb | 3 months ago | |
tutorial_c02.py | 3 months ago | |
tutorial_c03.ipynb | 3 months ago | |
tutorial_c03.py | 3 months ago | |
tutorial_c04.ipynb | 3 months ago | |
tutorial_c04.py | 3 months ago | |
tutorial_p01.ipynb | 3 months ago | |
tutorial_p01.py | 3 months ago | |
tutorial_p02.ipynb | 3 months ago | |
tutorial_p02.py | 3 months ago | |
tutorial_p03.ipynb | 3 months ago | |
tutorial_p03.py | 3 months ago | |
tutorial_p04.ipynb | 3 months ago | |
tutorial_p04.py | 3 months ago |
MindSONGE 是一款基于MindSpore开发的模块化、高通量、端到端可微的下一代智能分子模拟程序库。由深圳湾实验室、华为MindSpore开发团队、北京大学和昌平实验室共同开发。
通过template或者pdb、mol2格式的输入文件定义一个system,基本单元为Molecule()。
通过ForceField()或者加载一个神经网络训练力场,对system进行建模,得到一个potential。在potential的基础上,可以增加bias_potential,应用于增强采样/软约束。
可以使用UpdaterMD()定义一个动力学模拟过程,或者使用MindSpore内部的optimizer,如Adam()等等,来对system进行演化。
将system,potential,optimizer包装成一个Sponge()实例,就可以开始运行模拟了。
最后就是使用各种callback来定义输出格式,RunInfo()可以在屏幕输出每一步的结果,WriteH5MD可以将每一步的轨迹保存到一个h5md格式的文件中,并且支持VMD的可视化插件。
tutorial_c01.py:定义2个SPCE模型的真空水分子,并做能量极小化。
tutorial_c02.py:定义64个TIP3P模型的真空水分子,和一个球形谐振子势,在该约束下进行NVT模拟。
tutorial_c03.py:定义125个TIP3P模型的周期性水盒子,并做能量极小化,以及NVT和NPT模拟。
tutorial_c04.py:定义1个氨基酸链和水分子的混合体系,做能量极小化,以及SETTLE约束下的NVT模拟。
tutorial_p01.py:读取1个pdb蛋白质分子,在真空条件下做能量极小化。
tutorial_p02.py:读取1个pdb蛋白质分子,构建周期性水盒子,做能量极小化。
tutorial_p03.py:读取p02保存的水盒子,进行蛋白模拟流程示例:NVT模拟—NPT模拟—成品模拟。
tutorial_p04.py:读取1个pdb蛋白质分子,在真空条件下做能量极小化,定义反应坐标CV,然后进行MetaD增强采样。
.ipynb
格式的文件可以在jupyter notebook中运行,.py
格式的文件可以通过下面的命令运行:
# -e 运行平台:Ascend/GPU,默认:GPU
# -id device id,默认:0
python tutroial_xxx.py -e Ascend -id 0
因为一些案例依赖上一步的结果,所以请按照编号的顺序执行案例
MindScience is scientific computing kits for various industries based on the converged MindSpore framework.
Jupyter Notebook Python Unity3D Asset Pickle nesC other
Dear OpenI User
Thank you for your continuous support to the Openl Qizhi Community AI Collaboration Platform. In order to protect your usage rights and ensure network security, we updated the Openl Qizhi Community AI Collaboration Platform Usage Agreement in January 2024. The updated agreement specifies that users are prohibited from using intranet penetration tools. After you click "Agree and continue", you can continue to use our services. Thank you for your cooperation and understanding.
For more agreement content, please refer to the《Openl Qizhi Community AI Collaboration Platform Usage Agreement》