AMOFMS
An Automated Mapping and Optimization Framework for Molecular Simulation
Overview
AMOFMS (Automated Mapping and Optimization Framework for Molecular Simulation) is a versatile tool designed to streamline the creation and optimization of coarsed grained force field (CGFF) for molecular simulations. With its automated mapping function, AMOFMS simplifies the process of generating coarse-grained models by mapping fine-grained structures, reducing manual effort and increasing accuracy. Built on top of powerful libraries like MDAnalysis, NumPy and SciPy and conducting simulations on GROMACS, the tool offers comprehensive optimization capabilities, allowing users to fine-tune a wide range of molecular parameters for precise simulations and improved force fields. Moreover, AMOFMS provides flexibility and customization options, enabling users to define custom objective functions and parameter equivalences to tailor the optimization process to their specific research needs. Through its user-friendly interface and powerful functionality, AMOFMS empowers researchers to accelerate the development of force field models and advance molecular simulation studies.
Installation
The easiest way to install AMOFMS from the Python Package Index. You can do this using pip:
pip install amofms
Alternatively, you can also install the in-development version with:
pip install https://github.com/JiangGroup/amofms/archive/main.zip
Citing
If you use AMOFMS in your research, please cite it as follows:
Zhixuan Zhong, Lifeng Xu, and Jian Jiang*. A Neural-Network-Based Mapping and Optimization Framework for High-Precision Coarse-Grained Simulation. Journal of Chemical Theory and Computation Article ASAP.
BibTeX entry:
@article{doi:10.1021/acs.jctc.4c01466,
author = {Zhong, Zhixuan and Xu, Lifeng and Jiang, Jian},
title = {A Neural-Network-Based Mapping and Optimization Framework for High-Precision Coarse-Grained Simulation},
journal = {Journal of Chemical Theory and Computation},
volume = {0},
number = {0},
pages = {null},
year = {0},
doi = {10.1021/acs.jctc.4c01466},
note ={PMID: 39782000}
}
Please also cite MDAnalysis, on which AMOFMS is built.
We extend our gratitude to the authors of DSGPM for providing the original source code, which served as a foundation for the development of DSGPM-TP. Users of DSGPM-TP are kindly requested to cite both the original DSGPM publication and the DSGPM-TP publication to acknowledge the contributions of both works.
@Article{D0SC02458A,
author ="Li, Zhiheng and Wellawatte, Geemi P. and Chakraborty, Maghesree and Gandhi, Heta A. and Xu, Chenliang and White, Andrew D.",
title ="Graph neural network based coarse-grained mapping prediction",
journal ="Chem. Sci.",
year ="2020",
pages ="-",
publisher ="The Royal Society of Chemistry",
doi ="10.1039/D0SC02458A",
url ="http://dx.doi.org/10.1039/D0SC02458A",
abstract ="The selection of coarse-grained (CG) mapping operators is a critical step for CG molecular dynamics (MD) simulation. It is still an open question about what is optimal for this choice and there is a need for theory. The current state-of-the art method is mapping operators manually selected by experts. In this work{,} we demonstrate an automated approach by viewing this problem as supervised learning where we seek to reproduce the mapping operators produced by experts. We present a graph neural network based CG mapping predictor called Deep Supervised Graph Partitioning Model (DSGPM) that treats mapping operators as a graph segmentation problem. DSGPM is trained on a novel dataset{,} Human-annotated Mappings (HAM){,} consisting of 1180 molecules with expert annotated mapping operators. HAM can be used to facilitate further research in this area. Our model uses a novel metric learning objective to produce high-quality atomic features that are used in spectral clustering. The results show that the DSGPM outperforms state-of-the-art methods in the field of graph segmentation. Finally{,} we find that predicted CG mapping operators indeed result in good CG MD models when used in simulation."}