Martignac: Computational Workflows for Reproducible, Traceable, and Composable Coarse-Grained Martini Simulations
Tristan Bereau1, Luis J Walter1, Joseph F Rudzinski2
1Institute for Theoretical Physics, Heidelberg University, 69120 Heidelberg, Germany.
Journal of Chemical Information and Modeling
|December 2, 2024
Summary
Martignac computational workflows enhance molecular dynamics (MD) simulations using the Martini force field. This system improves simulation traceability and reproducibility by structuring data as a graph and connecting to the NOMAD database.
Area of Science:
- Computational chemistry
- Biophysics
- Materials science
Background:
- Molecular dynamics (MD) simulations are crucial but lack traceability and reproducibility.
- The coarse-grained (CG) Martini force field is widely used in various scientific domains.
Purpose of the Study:
- Introduce Martignac, a computational workflow system for Martini CG MD simulations.
- Enhance the traceability, reproducibility, and FAIR data principles of MD simulations.
Main Methods:
- Martignac models Martini CG MD simulations as acyclic directed graphs.
- Workflows cover system generation (liquids, bilayers) and free-energy calculations (solvation, permeation).
- Integration with the NOMAD database for automatic data normalization and storage.
Main Results:
- Demonstrated prototypical workflows for system generation and property calculations.
- Ensured automatic data normalization and storage adhering to FAIR principles via NOMAD.
- Established a framework for improved sustainability and reproducibility in molecular simulations.
Conclusions:
- Martignac significantly improves the traceability and reproducibility of Martini CG MD simulations.
- The system promotes FAIR data principles through seamless integration with the NOMAD database.
- Martignac offers a robust solution for complex simulation tasks, advancing scientific discovery.
