ood-proteinfold (OnDemand App)

ood-proteinfold is an Open OnDemand interactive app for computational protein structure prediction using Proteinfold (AlphaFold2, Boltz, ColabFold, ESMFold, and RoseTTAFold-All-Atom) originally built for UNSW's Katana HPC.

Features

• Predict protein structures from amino acid sequences, FASTA files, or samplesheets.
• Supports multiple prediction methods found within nf-core/proteinfold.
• Customisable run parameters via web form.
• Results accessible via web interface.

Usage

1. Access: Launch the app via the Open OnDemand portal.
2. Input: Provide a sequence, FASTA directory, or samplesheet.
3. Configure: Select prediction method, mode, and database options.
4. Submit: Start the job; monitor progress and download results when complete.

Key Files

• form.yml.erb: Defines the web form and input parameters.
• template/script.sh.erb: Main job script for running predictions.
• submit.yml.erb: Job submission configuration.
• info.html.erb: Displays result links after job completion.
• .github/workflows/: CI/CD deployment workflows.

Installation / Implementation

To deploy ProteinFold on your own Open OnDemand instance:

1. Clone this repository into your OnDemand apps directory:

   git clone https://github.com/Australian-Structural-Biology-Computing/ood-proteinfold.git /var/www/ood/apps/sys/kod-proteinfold

2. Edit files to match institutional settings:

   • Update submit.yml.erb with your facility-specific settings.
   • Change the cluster in form.yml.erb to match your HPC.
   • Create kod_proteinfold-${RUN_ENVIRONMENT}.config files for proteinfold-specific nextflow configurations. Or remove this line from script.sh.erb if not needed.
   • Create a template/.env file to change any of these default parameters:

# Example .env for ood-proteinfold

# Base output directory for all runs (job-specific OUT_DIR set at runtime)
BASE_OUT_DIR=/srv/scratch/${USER}/proteinfold_output

# Environment name (e.g., prod, dev)
RUN_ENVIRONMENT=prod

# Project root: location of Nextflow configs and samplesheet-utils environment
PROJECT_ROOT=/srv/scratch/sbf-pipelines/proteinfold

# Database path for structure prediction modules
DB_PATH=/srv/scratch/sbf-pipelines/proteinfold/dbs

# Git branch to use for proteinfold
BRANCH=master

# Repository path for using alternative versions of proteinfold
REPOSITORY=Australian-Structural-Biology-Computing/proteinfold

# Debug group for permissions on debug files
DEBUGGROUP=sbf-pipelines

# Base debug directory (job-specific DEBUGDIR set at runtime)
BASE_DEBUGDIR=/srv/scratch/sbf/debug

# Nextflow work directory base (job-specific NXF_WORK set at runtime)
BASE_NXF_WORK=/srv/scratch/${USER}/.proteinfold/work

# Apptainer/Singularity blob cache directories (set here or via your institutional Nextflow config, e.g. https://github.com/Australian-Structural-Biology-Computing/unsw_katana)
# APPTAINER_CACHEDIR=/srv/scratch/${USER}/.images
# SINGULARITY_CACHEDIR=/srv/scratch/${USER}/.images

# Nextflow Apptainer/Singularity image cache/library directories (set here or via institutional Nextflow config, e.g. https://github.com/nf-core/configs/blob/master/conf/pipeline/proteinfold/unsw_katana.config)
# NXF_APPTAINER_CACHEDIR=/srv/scratch/${USER}/.images
# NXF_SINGULARITY_CACHEDIR=/srv/scratch/${USER}/.images
# NXF_APPTAINER_LIBRARYDIR=/srv/scratch/sbf-pipelines/proteinfold/singularity
# NXF_SINGULARITY_LIBRARYDIR=/srv/scratch/sbf-pipelines/proteinfold/singularity

Python Environment for samplesheet-utils

samplesheet-utils is required for generating and validating input samplesheets. You must create a Python virtual environment and install the correct version:

python3 -m venv ${PROJECT_ROOT}/${RUN_ENVIRONMENT}/venv
source ${PROJECT_ROOT}/${RUN_ENVIRONMENT}/venv/bin/activate
pip install samplesheet-utils==1.1.2

Update the paths in .env if your environment is elsewhere.

Citation

Please cite doi.org/10.26190/4KQF-M552 and acknowledge the Structural Biology Facility at UNSW in publications.

---

For support, please open an issue.