Simulation Foundry: Automated and F.A.I.R. Molecular Modeling

Area of Science:

• Computational chemistry and materials science.
• Development of automated workflows for scientific simulations.

Background:

• Molecular modeling (MM) is crucial for predicting properties of complex systems.
• Current MM data creation can be challenging to repeat, replicate, and share effectively.
• Need for standardized, automated workflows to manage increasing simulation demands.

Purpose of the Study:

• To introduce the Simulation Foundry (SF), a modular workflow for automated MM data generation.
• To enhance the reproducibility, replicability, and F.A.I.R. principles of MM data.
• To facilitate high-throughput simulations and data integration for multicomponent systems.

Main Methods:

• Developed a modular workflow based on widely used scripting languages (bash, Python).
• Implemented standardized data structures and file naming conventions for cross-platform compatibility.
• Ensured reusability and re-editability of the workflow for expert users.

Main Results:

• Demonstrated SF's usability through simulations of thermophysical properties of binary mixtures.
• Established a standardized data exchange format for integrating simulated and experimental data.
• Provided complete documentation for result provenance and transparency.

Conclusions:

• The SF enhances the automation, modularity, and efficiency of molecular modeling.
• SF promotes reproducible, transparent, and F.A.I.R. data creation within the scientific community.
• SF serves as a platform for integrating new methods and fostering collaborative research.