A community data repository and curation service for Molecular Mechanics (MM) and Molecular Dynamics (MD) related computations for research into the SARS-CoV-2 / COVID-19 / ”Coronavirus” pandemic maintained by The Molecular Sciences Software Institute (MolSSI) and BioExcel. This virus has spurred the entire scientific world into action looking for treatments and answers to help minimize its spread, its symptoms, and its fatalities. To help researchers and the community quickly provide their resources and expertise, this repository is meant to serve as a curated central repository of MM and MD input files, trajectories, analysis scripts, well-validated algorithms, and related data for immediate community access.
This project may serve as a prototype into a future, more general Molecular Mechanics Exchange of Data project. There may come a time as this project evolves that it naturally does become more general (or change to another specific focus), however, this project focuses on the SARS-CoV-2 pandemic for now and future evolutions can be dealt with as they arise.
The nature of this pandemic requires rapid, and flexible response. This repository, and the data within are designed to get the information out quickly. The repo seeks to provide the following information in a centralized location:
- Provide Simulation input files (structures, configurations, scripts, Jupyter notebooks) to the community in an organized structure. Sorting done by Research Group, Target (spike protein, protease, etc), Simulation Program, Hardware/platform (cluster, cloud, local, etc).
- Provide Trajectories to the community for analysis. Due to the nature of the trajectories’ sizes, we may only be able to provide pointers to the trajectories where possible. As the data, repository, and available resources evolve, MolSSI may be able to provide a central location for full trajectories in the future.
- Provide analysis files and quantitative numbers associated with inputs and trajectories.
- Pointers to preprint servers such as arXiv, bioRxiv, and ChemRxiv on biomolecular simulation research in regards SARS-CoV-2.
- Where appropriate, well-validated and easily deployable algorithms such as functional Monte Carlo, machine learning methods, and heuristic property calculators.
This exchange is a community resource as well as community driven tool. Suggestions as to changes for the folder structure, types of content accepted, regulation of the content, etc. are all welcome and encouraged through issues and pull requests.
The following classes of data are currently accepted in the Exchange (Note: This is a flexible and broad list. New suggestions can be made through issues). Please note that all data in each category should have the subsequent information associated with it to be accepted. All submissions should also have contributing information in the form of a README.md
- Structures
- Provenance of structure including any modifications made. Will accept manually written provenance. If modifications made by program, please include script along with pre-modified files and list of required programs to re-generate.
- Simulation Configuration Files
- Programs required to run (minimal subset which gets all dependencies, with minimal versions),
- Basic script and/or working instructions to execute these configuration files,
- Structure files (required for non-internet accessible structures, e.g. not required for things on the Protein Data Bank)
- Program Structure Files - Structures formatted uniquely for a specific simulation program or run
- Same requirements as Simulation Configuration Files
- Analysis Data
- Complete set of any of the previous data and exact details for replication
- Analysis scripts with placeholders for output trajectory files.
- Pointers to the trajectories if possible
- Trajectories - Not accepted in their entirety to the GitHub repository,
- Pointers will be accepted and noted. (contact information, where it’s stored, how to request, URLs, etc)
- Program used to generate this trajectory.
- Structures (or links there to) which were used to initiate the simulation which yielded this trajectory.
- Config files and or scripts which were used to run the simulation that generate the trajectory (if possible)
- Papers - as links
- Links to preprints and publications
- Inclusion of the files for the paper if they fall into the other categories above.
- Notation of any simulation package or software used to get the data in the paper
- Papers from multiple groups should be added in multiple folders (see Contributing below)
- Movies - hyperlinks only, please do not include full movie files.
- Links to long term-hosted movies (e.g. YouTube).
- Structures (or links there to) which were used to initiate this movie.
- Config files and or scripts which were used to run the simulation that generate the trajectory (if possible)
- The trajectory which this movie references (if possible, see rules on Trajectories above)
- Algorithms - Well-validated and easily deployable algorithms such as functional Monte Carlo and machine learning methods and heuristic property calculators
- Code Snippets
- Programing Language (if restricted)
- List of required software and minimal dependencies (if any)
Due to the rapidly evolving nature of this exchange, organization and requirements may change with time. However, no changes to the repository will remove user-submitted data. The maintainers may reach out to the original contributor to ask for additional information as needed.
The exchange gladly accepts any of the above types of data (and others if you propose an organization). Simply open a PR with the data you wish to submit with the following folder structure (omit any directories for types you are not submitting):
sorted_by_group
└──YOUR_GROUP_NAME
├── README.md
└── TARGET
├── algorithms
├── analysis
├── movies
├── papers
├── program_specific_structures
│ └── SIMULATION_SOFTWARE
├── simulation_configs
│ └── SIMULATION_SOFTWARE
├── structures
└── trajectories
└── SIMULATION_SOFTWARE
Create a folder for your group within the sorted_by_group directory. This will serve as the main storage location for all of your submissions. Within this folder, create a README.md file and include information about your contributing group(s):
- Author: The user(s) who produced the data associated with the submission.
- Group: The group(s) of the author(s) of the submission.
- Organization: The organization(s) of the author(s).
- Reference_Information: If desired, include websites for the associated groups.
- Contact_Information: If desired, include ways that the author(s) can be contacted about their submission.
Note that for submissions by collaborations of multiple groups, create a folder within the sorted_by_group folder for your collaboration. Within the README.md include all authors and groups associated with this collaboration.
For each new submission made to the repository, please create a folder, within your groups folder, with a unique SUBMISSION_ID as its name. Ideally, submission names should represent the order of submissions, allowing each submission to be processed in the order it is received.
Each submission should include a folder representing the TARGET of the submission and a README.md for the submission. The TARGET folder should be named to represent the target type of your submission. We ask that you include all fields under Contributor Information above along with information requested for the type of data in your submission within the README.md file.
Within the TARGET folder, create a subfolder for each type of data within this submission. These folders should contain all relevant files to the data types along with any meta data, such as provenance information, for the data.
This structure is subject to change, but all changes will be made by a maintainer and the instructions updated accordingly.
MolSSI and BioExcel are working together on this project to de-duplicate efforts and to reach the broadest community possible. We are activley looking for other institutes, collaborators, consortia, and anyone who wants to be an active contributor to join us!
Reach out to any of the maintainers below and lets work together to make this as helpful as possible!
- The Molecular Sciences Software Institute (MolSSI)
- BioExcel
- Levi N. Naden (MolSSI)
- Sam Ellis (MolSSI)
- Andrew Abi-Mansour (MolSSI)
MolSSI is funded by the NSF through ACI-1547580