(360h) The Molecular Simulation Design Framework (MoSDeF): New Capabilities

The Molecular Simulation Design Framework (MoSDeF)¹ is a collection of python libraries designed to assist with the preparation of chemical, materials, and biomolecular systems for simulation using various open-source engines. The project was motivated by the irreproducibility crisis that exists in the science community, including the field of computational simulation. The software suite consists of three main components to reproducibly encapsulate different steps of the process of system preparation, namely, mBuild², Foyer³ and GMSO (General Molecular Simulation Object).⁴ mBuild² enables programmatic construction of system configurations using the concept of generative modeling that allows complex systems to be constructed from smaller, interchangeable pieces that can be connected together; mBuild allows for the creation of custom initialization scripts that can be shared and used as plugins. The Foyer³ package performs atom-typing, i.e., the application of a force field to a given system; in Foyer, forcefield parameters are stored within XML files along with usage rules encoded as SMARTS strings, that are both human and machine readable, enabling the rules and parameters used to set up a simulation to be captured and disseminated. The GMSO⁴ library is designed to store the fully parameterized system information, which can then be written out to syntactically correct input files for various simulation engines, including Monte Carlo engines (Cassandra, GOMC, and MCCCS) and molecular dynamics simulation engines (HOOMD-Blue, LAMMPS, GROMACS, and NAMD). The project has become a collaborative effort between multiple universities, where the collaborative development team has been working to add new features to these libraries, increasing their capabilities to support a wider range of chemical systems, configurations, simulation engines, and to ensure the interoperability with other existing and developing software packages. These efforts allow us to move closer to the goal of supporting Transferable, Reproducible, Usable by others, and Extensible (TRUE) simulation research.⁵

References

1. “MoSDeF” [Online]. Available: https://mosdef.org
2. Klein, J. Sallai, T. J. Jones, C. R. Iacovella, C. McCabe, and P. T. Cummings, “A Hierarchical, Component Based Approach to Screening Properties of Soft Matter”, Foundations of Molecular Modeling and Simulation, 2016, pp. 79-92.
3. Klein, A. Z. Summers, M. W. Thompson, J. B. Gilmer, C. McCabe, P. T. Cummings, J. Sallai and C. R. Iacovella, Formalizing atom-typing and the dissemination of force fields with foyer, Comput. Mater. Sci., 2019, 167, 215–227.
4. General Molecular Simulation Object (GMSO) [Online]. Available: https://github.com/mosdef-hub/gmso.
5. W. Thompson, J. B. Gilmer, R. A. Matsumoto, C. D. Quach, P. Shamaprasad, A. H. Yang, C. R. Iacovella, C. McCabe and P. T. Cummings, Towards molecular simulations that are transparent, reproducible, usable by others, and extensible (TRUE), Mol. Phys., 2020, 0, e1742938.