Collaborators: Lige Zhang (DKU '26), Bruce Zhou (DKU '24), Prof. Gaoyang Li, Prof. Huansheng Cao.
All scripts and configuration files can be found in the docking/ folder.
The screening library was obtained from Enamine Bioactive Compounds Collection, which consists of annotated compounds and FDA-approved drugs. It can be accessed at: https://enamine.net/compound-libraries/bioactive-libraries.
The screening library was originally prepared in 2D SDF format. We used the Open Babel 3.0.1 on Linux to generate the 3D coordinates for each compound. The quality was set to "slowest", which performs "Force field cleanup (500 cycles) + Slow rotor search".
We prepared the PDBQT files needed for docking with AutoDock Vina using the Meeko package. Specifically, we used the mk_prepare_ligand.py script to turn 3D SDF files into the PDBQT files needed for docking.
The receptors were prepared using the AutoDockTools GUI following the standard workflow. A sample configuration files required for AutoDock Vina was uploaded as config.txt.
The PDBQT files are stored in the folder pdbqtLigands. The vina_docking.sh script will perform docking using AutoDock Vina and automatially collects results using the extract_energy.sh script.
A blind docking strategy was adopted. For each receptor, the docking grid size was determined manually to make sure that the search covers the whole protein.
The maximum number of binding modes to generate was set to 50, and the exhaustiveness was set to 16 to increase search coverage.
The GROMACS 2023.1 version was used to perform the molecular dynamics simulations.
The CHARMM36 all-atom force field (July 2022 version) was utilized, chosen for its optimization in protein-ligand interactions.
- Protein Topology: Generated using GROMACS's built-in commands.
- Ligand Topology: Created via the CGenFF server.
- Simulation Environment: Protein-ligand complexes were embedded in dodecahedral unit cells and solvated using the SPC216 solvent model.
- System Neutralization: Sodium or chloride ions were added to neutralize the system.
- Conducted to ensure system stability before simulations.
- Two stages:
- NVT (constant Number of particles, Volume, and Temperature) at 300K for 100 ps.
- NPT (constant Number of particles, Pressure, and Temperature) at 300K for 100 ps.
- Duration: 100 ns.
- Step Size: 2 ps.
- Total Steps: 50,000,000.
- Stability of protein-ligand complexes was evaluated using RMSD (root-mean-squared deviation) to confirm consistent fluctuations and equilibrium over the 100 ns period.