L2P Galaxy Tool

A Galaxy tool for performing List to Pathways (L2P) enrichment analysis on differential expression data. This tool generates both bar plots and bubble plots for upregulated and downregulated gene sets.

Features

• Automated pathway enrichment analysis using GO, REACTOME, and KEGG databases
• Dual plot types: Bar plots and bubble plots for visualization
• Separate analysis for upregulated and downregulated genes
• Flexible gene selection: By rank (e.g., t-statistic) or by thresholds (p-value and fold-change)
• Publication-ready plots with proper formatting (50pt top margins, wrapped labels, optimized dimensions)
• Container-based deployment via GitHub Container Registry (ghcr.io)

Example Outputs

Container Image

The Docker container is automatically built and published to GitHub Container Registry via CI/CD:

ghcr.io/NIDAP-Community/l2p-galaxy-tool:latest

Available Tags

• latest - Latest build from the main branch
• v1.1.0 - Specific version tags
• main-<sha> - Commit-specific builds

Usage in Galaxy

Installation

1. Add the tool to your Galaxy instance by copying l2p_single.xml to your tools directory
2. Update your tool_conf.xml to include:

<section id="ccbr_tools" name="CCBR Tools">
    <tool file="path/to/l2p_single.xml" />
</section>

3. Ensure Docker is configured in your Galaxy job configuration:

runners:
  local:
    load: galaxy.jobs.runners.local:LocalJobRunner
    workers: 4

execution:
  default: docker_local
  environments:
    docker_local:
      runner: local
      docker_enabled: true
      docker_sudo: false

4. Update the container reference in l2p_single.xml:

<container type="docker">ghcr.io/NIDAP-Community/l2p-galaxy-tool:latest</container>

Input Requirements

The tool requires a tabular (tab-delimited) differential expression file with:

• Gene identifier column (e.g., gene symbols)
• Ranking column (e.g., t-statistic) - for rank-based selection
• Significance column (e.g., p-value)
• Fold-change column (linear scale, not log2)

Example format:

symbol	PSA_high-PSA_low_tstat	PSA_high-PSA_low_pval	PSA_high-PSA_low_FC
GENE1	5.2	0.001	2.3
GENE2	-4.1	0.003	0.4

Parameters

Parameter	Description	Default
DEG table	Tab-delimited differential expression results	Required
Gene name column	Column containing gene identifiers	symbol
Species	Human or Mouse	Human
Select by rank	Use ranking instead of thresholds	Yes
Ranking column	Column for ranking genes (e.g., t-stat)	groupA_vs_groupB_tstat
Significance column	P-value column	groupA_vs_groupB_pval
Fold-change column	Linear fold-change column	groupA_vs_groupB_FC
Significance threshold	P-value cutoff	0.05
Fold-change threshold	Linear FC cutoff	1.2

Outputs

1. Pathways CSV - Enriched pathways for both up and down genes
2. Bar Plot (Upregulated) - Bar chart of top enriched pathways (upregulated genes)
3. Bar Plot (Downregulated) - Bar chart of top enriched pathways (downregulated genes)
4. Bubble Plot (Upregulated) - Bubble plot showing pathway enrichment (upregulated genes)
5. Bubble Plot (Downregulated) - Bubble plot showing pathway enrichment (downregulated genes)

All plots are generated as PNG files (1500×1200 pixels, 150 DPI) with optimized formatting:

• 50pt top margin to prevent title cutoff
• Wrapped pathway labels (30 characters max)
• Proper axis scaling and spacing

Development

Building Locally

docker build -t l2p-galaxy-tool .

Testing Locally

docker run --rm -v $(pwd)/test_data:/data l2p-galaxy-tool \
  Rscript /opt/l2p_single/run_L2P_Single.R \
  /data/deg_results.csv \
  /data/pathways.csv \
  /data/bar_up.png \
  /data/bar_down.png \
  /data/bubble_up.png \
  /data/bubble_down.png \
  symbol \
  Human \
  TRUE \
  PSA_high-PSA_low_tstat \
  PSA_high-PSA_low_pval \
  PSA_high-PSA_low_FC \
  0.05 \
  1.2

Repository Structure

l2p-galaxy-tool/
├── .github/
│   └── workflows/
│       └── docker-publish.yml    # CI/CD workflow
├── Dockerfile                     # Container definition
├── L2P_Single.R                   # Core L2P analysis functions
├── run_L2P_Single.R              # Command-line wrapper
├── l2p_single.xml                # Galaxy tool definition
├── LICENSE                        # License file
└── README.md                      # This file

CI/CD Pipeline

The GitHub Actions workflow automatically:

1. Builds the Docker image on every push to main
2. Tags images appropriately:
   • latest for main branch
   • v* for version tags
   • main-<sha> for commit tracking
3. Pushes to GitHub Container Registry
4. Caches build layers for faster subsequent builds

Triggering a Release

Create a version tag:

git tag v1.1.0
git push origin v1.1.0

This will build and publish with tags: v1.1.0, v1.1, v1, and latest

Credits

Based on the L2P package from CCBR/l2p.

Plot formatting improvements:

• Fixed title cutoff with increased top margins
• Added pathway label wrapping for readability
• Corrected x-axis range calculations
• Optimized plot dimensions for publication quality

License

MIT License - See LICENSE file for details

Troubleshooting

Galaxy Issues

Container not found:

• Ensure the container reference in XML matches your published image
• Check Docker is enabled in Galaxy job configuration
• Verify the container is public or GitHub token is configured

Plots have cut-off titles:

• Update to the latest container version (includes 50pt top margin fix)
• Check container is pulling the latest image (not cached)

Worker crashes:

• Set preload: false in galaxy.yml
• Disable mulled container resolvers in container_resolvers.yml

Docker Issues

Build fails:

• Check Docker has internet access for downloading R packages
• Verify base image rocker/r-ver:4.1.3 is accessible
• Review GitHub Actions logs for specific errors

Support

For issues and questions:

• Open an issue on GitHub
• Check existing issues for solutions
• Review Galaxy logs at database/gravity/log/gunicorn.log