Docking pipeline

Run nextflow run main.nf to start pipeline.

Configurations

Use or create new configurations on config/

config/5TBM-GPU-ACCELERATED.config <= Using quickvina-gpu
config/5TBM.config <= Using vina
...

Packages

If you want to execute this pipeline, i suggest you to use container images;

# QuickVina-GPU, access with QuickVina-W-GPU-2-1 on commandline
docker pull ghcr.io/jsphu/docking_pipeline/quickvina-gpu:latest
# Downloader used for downloading from links
docker pull ghcr.io/jsphu/docking_pipeline/downloader:latest

Post-Docking Analysis & Filtering

After running the pipeline, you can filter the results and prepare them for the next phase (e.g., SwissADME, MD Simulations).

1. Filter Best Ligands

Filter ligands based on scripts/rules.txt (MW < 400, LogP <= 5, LE >= 0.3, PSA < 60, etc.) and apply PAINS filters to remove false positives.

python3 scripts/filter_ligands.py
python3 scripts/pains_filter.py

Output: data/filtered_ligands.csv

2. Prepare for SwissADME

Standardize SMILES strings (fix radical issues and valency) for compatibility with medicinal chemistry tools.

python3 scripts/prepare_swissadme.py

Output: data/filtered_ligands.smi (Ready to upload to SwissADME)

3. Organize Lead PDBQTs

Collect the docking pose files of the filtered leads into a single directory for visualization.

python3 scripts/collect_leads.py

Output: data/best_leads_pdbqt/

4. Generate Summary Report

View a ranked list of the top candidates by combined efficiency.

python3 scripts/summary_report.py