DiffDock Protocol For Biochem course

Original Paper
DiffDock GitHub
FlowChart_ForToolsUse.pdf

Prerequisites

• Obtain LigPlot License

• Install Miniconda

• Open Conda Terminal

• Create folder C:\tmp (Windows) or /tmp (Mac/Linux)
  Linux/Mac Users: Run mkdir /tmp
  Windows Users: Run mkdir C:\tmp

• Download Chimera

• Download LigPlot

• Open LigPlot (ensure the tmp folder path is set correctly)

• Download 2hyy_single.pdb from this reposistory

Optional

• Run git clone https://github.com/tkella47/BioChem_DiffDock.git to get the pdb and python files you need to download.
  [Requires Git]

Troubleshooting:

Is LigPlot not opening?

• You may need to install Java
  Warning: Mac Users
  There is an x64 Java and ARM64 Java. This relates to the chip on your computer.
  If you have an M1 or M2 chip, ARM is for you. Otherwise, download x64 Java.
  Don't know how to check? Visit
• Mac Users: the installed above may give you OS error code 1. If so, please visit this issue
• For Mac users, See: Protocols_ .pdf

Obtaining Protein-Ligand Docked Poses Using DiffDock

1. Navigate to DiffDock HuggingFace Space
2. Upload the 2hyy_single.pdb and paste in the ligand SMILES sequence
3. Reduce number of samples to 10.
4. Download the result.
5. The results will be in zipped file. Extract them.

Merging the Samples in Chimera(X)

In Chimera(X), open the 2hyy_single.pdb

Then open rank#.sdf

(Sidenote, you can see all the test ligand positions by opening rank#_reversediffusion)
Take a moment to investigate the beauty that is biochemistry

Under File, Click save. Alternatively Ctrl(Windows/Linux)+S or Command(Mac)+S

Make sure that both the sdf model and pdb model are selected.

Give your file a name and save it.

Processing the PDB File

Next download process_ligand_protein.py

Navigate to the same directory as process_ligand_protein.py. Ensure that process_ligand_protein.py and the pdb from the previous step are in the same folder.

Then run

python process_ligand_protein.py PATH_TO_YOUR_SAVE_FILE.pdb

This will save a file diffdocked.pdb in the same place as the python file.

If the above code does not work, run the code below. Additionally, if you wish to provide a different name or path for your file simply run.

python process_ligand_protein.py PATH_TO_YOUR_SAVE_FILE.pdb --output_path DESIRED_NAME_OR_PATH/diffdocked.pdb

Viewing in LigPlot

Open the DiffDocked PDB file first

Make sure LigPlot is open
Click File: Open :
Browse : Open diffdocked.pdb

Compare to Crystal Structure

Download 2HYY.pdb
Click File: Open:
Browse : Open 2hyy.pdb

In the bottom right hand corner, you can split the screen to see the difference between the docked ligands