Automated Docking of Ligands to the BK Channel

Overview

This directory contains a bash script that automates molecular docking of one, two, and four ligand molecules, taken from a list, to the human BK channel using AutoDock Vina and Vinardo scoring functions. Vinardo was selected as an additional scoring function as it is well suited for ion channel targets. This script follows the guidelines from the AutoDock Vina manual, available at: AutoDock Vina documentation guidelines.

Directory structure

├── dock.sh # Main docking script - see Script WorkFlow section for details
├── CAS_IDs.txt # List of CID numbers for ligands of interest
├── protein.pdbqt # BK Channel for docking
├── config.txt # Box grid parameters for docking
├── environment_docking.yml # — YAML file listing all dependencies and package versions for the conda environment used in this project
├── fetched_sdf/ # Directory with ligand structures in SDf format, with and without added hydrogens
├── vina_results/ # Directory with obtained docking poses for each ligand in PDBQT format and TXT file containing information about obtained docking poses using AutoDock Vina scoring function.
└── vinardo_results/ # Directory with obtained docking poses for each ligand in PDBQT format and TXT file containing information about obtained docking poses using Vinardo scoring function.

Installing dependencies

environment_docking.yml contains all dependencies and packages needed for dock.sh to work. Set up the environment by running:

  conda env create -f environment_docking.yml

Two additional Python scripts are required that are not available via conda:

• scrub.py — from molscrub, used for protonation and 3D coordinate generation
• mk_prepare_ligand.py — from Meeko, used to convert ligands to PDBQT format for AutoDock Vina

Install both into your environment from source:

  git clone https://github.com/forlilab/molscrub.git && cd molscrub && git checkout develop && pip install -e . && cd ..
  git clone https://github.com/forlilab/Meeko.git && cd Meeko && pip install -e . && cd ..

Protein Preparation

The receptor file protein.pdbqt was prepared from the BK channel crystal structure using AutoDock Tools (ADT) v1.5.7.

Steps performed in ADT:

1. Load structure

   • File → Read Molecule — load the cleaned PDB file

2. Remove water molecules and potassium ions

   • Edit → Delete Water
   • manually select and delete potassium ions

3. Add polar hydrogens

   • Edit → Hydrogens → Add → Polar Only

4. Add Kollman charges

   • Edit → Charges → Add Kollman Charges

5. Merge non-polar hydrogens

   • Edit → Hydrogens → Merge Non-polar

6. Assign AutoDock 4 atom types

   • Edit → Atoms → Assign AD4 typ

7. Save receptor as PDBQT

   • Grid → Macromolecule → Choose → select protein → save as protein.pdbqt
   • only ATOM records are saved (CONECT, CRYST1, END excluded)

8. Define the docking grid box

   • Grid → Grid Box — center and dimensions set to cover the BK channel binding site
   • grid parameters saved to config.txt

Script Workflow

1. Logging All terminal output is echoed to run.log via tee

2. Create necessary directories The script creates the following directories if they don't exist:

   • fetched_sdf — stores ligand structures with and without hydrogen in SDF format and prepared PDBQT files
   • vina_results — stores obtained docking poses for each ligand in PDBQT format and TXT file containing information about obtained docking poses using AutoDock Vina scoring function
   • vinardo_results — stores obtained docking poses for each ligand in PDBQT format and TXT file containing information about obtained docking poses using Vinardo scoring function

3. Fetch molecules by CAS IDs For each CAS number in CAS_IDs.txt:

   • Obtains CID from CAS number by querying PubChem
   • Fetches the SDF file for the CID

4. Ligand Preparation

   • For each SDF file runs mk_prepare_ligand.py to prep ligand for the docking

5. Molecular Docking For each prepped ligand, performs docking against a BK channel (protein.pdbqt) with --exhaustiveness=32 using two scoring functions and three ligand copy configurations:

   • Vina docking with 1, 2 and 4 ligand copies
   • Vinardo docking with 1, 2 and 4 ligand copies

6. Clean-up Moves all docking outputs to appropriate folders.

Usage

Perform the docking by running:

./dock.sh

• The list of docked ligands can be updated by modifying the CAS_IDs.txt file.
• This script uses protein.pdbqt file, which contains a BK channel structure but it can be replaced by any protein of interest - in that case grid box parameters have to be updated in config.txt file to match the binding site of the new protein.

Cheers! :)