BacMet

Getting started

To get started running this project you need to get through the following steps:

• Install docker and docker compose or other container runtime alternatives
• Perform Data import
• Start the app
• View the app on http://localhost:5000/

Data import

The original data should be fetched from the server at Chalmers and be put in a directory called data-import, replicating the directory structure on the Chalmers server (see below).

The data is imported into the persistent SQLite database /data/database.db by copying the data-import directory into the running app containers persistent volume. This can be done at any time, and copying updated data (or the same old data a further time) will cause the already imported data to be discarded.

This shows how to explicitly throw the old data away and reload it from scratch. In reality, the down -v step is not actually needed:

docker compose down -v
docker compose up -d

docker compose cp data-import app:/home/bacmet/data/

By observing the container log (docker compose logs -f), you should be able to follow the service noticing that new data is available, importing it (in several steps), and then finally restarting the service.

Fetching updated data

Fetching the data is easiest done using rsync:

rsync --progress --stats -h -ia --delete \
    user@server:/remote-path/NBIS/ data-import/

Expected layout of data-import directory

data-import
├── 1-Experimentally_validated
│   ├── 1-BacMet_experimentally_validated_database_3.0.csv
│   ├── 2-BacMet_compound_CAS_numbers.csv
│   └── Experimentally_validated_PDB_files.zip
├── 2-Predicted_database
│   ├── 1-Predicted_unique_homologues.zip
│   ├── 2-Predicted_unique_homologous_sequences.zip     # (not used)
│   └── 3-Predicted_groups.zip
└── 3-Sensitivity_distributions
    ├── MIC_Biocides.zip
    ├── MIC_metals.zip
    └── MIC_other_compounds.zip

4 directories, 9 files

Start the app

The app can be started in a production-like environment or in an environment tuned for convenient development. When you are switching between environments it is important to remember to rebuild the container so either use the build command or add the flag --build.

Start production-like environment

The production-like environment will copy all necessary app related code and assets into the container in order to create a self contained deployable container.

docker compose up --build

Start development environment

The development environment will mount the app directory and automatically reload code when it is changed.

docker compose -f docker-compose.yml -f docker-compose.dev.yml up --build

Run tests

Tests can only be run in the development docker container since the HTML validation requires openjdk. The tests can be run by using the following command:

docker compose -f docker-compose.yml -f docker-compose.dev.yml run --rm app pytest -p no:cacheprovider app/tests