sbp-workflow-configs

The centralised repository to manage all Structural Biology Platform workflow configurations

Offline execution of workflows in workflow queue

When running workflows offline in workflow queue with nextflow/25.10.x via Seqera, the following steps can be used to ensure that the workflow runs successfully without external network access.

1. Clone the workflow repo to a local directory on Gadi, under a folder named after the GitHub organisation (e.g. AustralianBioCommons). For example:

   cd /path/to/local/directory
   mkdir AustralianBioCommons
   cd AustralianBioCommons
   git clone https://github.com/AustralianBioCommons/sbp-proteinfold.git

2. Within the Seqera pre-run script, set the NXF_ASSETS environment variable to point to the parent directory of the GitHub organisation folder, while also setting NXF_OFFLINE to true. For example:

   module load nextflow/25.10.4
   export NXF_OFFLINE=true
   export NXF_ASSETS=/path/to/local/directory

3. Run the workflow via Seqera, using the Git repo as the "Pipeline to launch" (e.g. https://github.com/AustralianBioCommons/sbp-proteinfold.git)

Using mini or prod databases with protienfold and wisps

To use different databases across different SBP instances (e.g. mini databases on dev and prod databases on staging/prod), set the PF_DB_BASE_DIR environment variable to the appropriate database base directory before running the protienfold or wisps workflows. On Seqera, this can be set in the pre-run script.

To use mini databases:

export PF_DB_BASE_DIR="/g/data/if89/proteinfold_dbs/proteinfold_minidbs"

To use prod databases:

export PF_DB_BASE_DIR="/g/data/li87/proteinfold_dbs"