COBRAdb

COBRAdb loads genome-scale metabolic models and genome annotations into a relational database. It powers BiGGr Models, and it is available under the MIT license. The bin/load_db script is the entrypoint for building a database. The cobradb python library is also used by the biggr_models python package.

Installation

It is recommended to run the program using Docker compose.

1. Install and set up Docker.

2. Set up your configuration. Copy the file settings.ini.docker and rename it settings.ini. Change the postgres_password property to a password of choice. Change other options (like the number of processes to use) as desired.

3. Similarly, copy the env.docker file to .env and set the same postgres password here. (Make sure to neither put the settings.ini or .env file under source control or share them in any other way.)

4. Create the following directories:

   • /data/biggr-assemblies
   • /data/biggr-input-models
   • /data/biggr-chebi-cache
   • /data/biggr-datafiles

5. Populate these directories with the required datafiles: /data/biggr-chebi-cache can be left empty. For /data/biggr-assemblies and /data/biggr-input-models, see the biggr_models_data documentation. /data/biggr-datafiles files are available internally on the biggr storage account.

6. Build the docker container using docker compose build and run using docker compose up.