About Metabolic Atlas

Metabolic Atlas is a web platform integrating open-source genome scale metabolic models (GEMs) for easy browsing and analysis. The goal is to collect many curated GEMs, and to bring these models closer to FAIR principles. The website provides visualisations and comparisons of the GEMs, and links to resources, algorithms, other databases, and more general software applications. In short, the vision is to create a one-stop-shop for everything metabolism related. Read more about the project on the About page.

Cite us

If you use Metabolic Atlas in your scientific work, please cite:

Version 3.0: Li, F., et al, 2022. GotEnzymes: an extensive database of enzyme parameter predictions. NAR, gkac831 doi:10.1093/nar/gkac831

To cite previous versions or other adjacent work, see the Citation page on the website.

Contributing

If you would like to contribute to the project, for example if you have ideas for new features, discovered a bug or if you would like to improve the code base, please have a look at CONTRIBUTING.md. All ideas and contributions are highly appreciated.

If you discover any potential vulnerabilities associated with the project, please reach out to us at [email protected].

Get started

The front-end uses Vue.js, with help of Vue CLI. The backend uses ExpressJS and Neo4j as the database.

Docker and docker-compose are used to manage the dependencies of this project. Start by installing these if they are not present on the system.

If you want to try out the latest features of MetabolicAtlas, change the branch to main.

Apart from the current repository, two additional repositories are required in order to deploy Metabolic Atlas locally, they are

• data-generation: for generating neo4j database
• data-files: contains all the data necessary data (integrated models, maps, FTP repository) using Git LFS

Clone the three required repositories by

git clone https://github.com/MetabolicAtlas/MetabolicAtlas
git clone https://github.com/MetabolicAtlas/data-generation
git clone https://github.com/MetabolicAtlas/data-files && pushd data-files; git lfs pull; popd

Go to the repository data-generation and follow the instructions on how to generate the data files required by Metabolic Atlas.

In the folder MetabolicAtlas that has been cloned, add a env-local.env file based on the env-local.env.sample file:

cp env-local.env.sample env-local.env

and modify this env-local.env file.

Make sure the paths for DATA_FILES_PATH and DATA_GENERATOR_PATH are correct for your setup, eg. the paths to where you have downloaded the repositories data-files and data-generation.

To load the list of helper commands run:

source proj.sh

Run the helper command that creates the database and builds the Docker images of the project:

build-stack

Finally, start the Docker containers with

start-stack

Given successful deployment, the frontend should be accessible at: http://localhost/. If you encounter any problems try looking at the logs logs api / logs frontend.

To deploy the stack to a remote server, create another ENV file, e.g. env-dev.env, and modify it accordingly:

cp env-local.env.sample env-dev.env

Use the tag of the file (dev in the previous example) as an argument to the deploy-stack command to pass the ENV file to the Docker command and the Docker compose file:

deploy-stack dev

GotEnzymes

To reconstruct the database for GotEnzymes on the local (development) machine. Run the following. This should take ~10 minutes.

 ma-exec pg psql -f scripts/init.sql -U postgres

For remote servers, the init script is configured to run automatically if the database has not been initialized. To reconstruct the database, delete the mounted volume for the database on the remote server (located at /var/lib/docker-volumes/pg/postgres-data) and deploy again.

Description of helper commands

• To bootstrap the project: build-stack

• To run the project: start-stack

• To stop the project: stop-stack

• To run a command inside a docker container: ma-exec <CONTAINER_NAME> <COMMAND>. Below is a list of useful container specific commands.

  • Run API data validation tests: ma-exec api yarn test
  • To run compartment related tests only: ma-exec api npx jest --testNamePattern=compartments
  • Use prettier to format frontend (also works for api) files: ma-exec frontend yarn format
  • Run eslint for frontend (also works for api): ma-exec frontend yarn lint
  • Create production build and stats.html : ma-exec frontend yarn build

• To clean the project (delete containers and volumes): clean-stack

• To display real-time logs: logs [container-name: frontend/api/nginx/neo4j/ftp]

• To deploy the project: deploy-stack

• To (re-)import the Neo4j database: import-db

A note to Unix/Linux users

When rebuilding the stack, you might have to change the ownership of the directory neo4j:

sudo chown -R <user> neo4j

Replace <user> with your user name. The ownership will be automatically reset when running the project, so you will have to repeat this step for every rebuild.

Licenses