ianandersonlol / dockingpipeline

Protein Modeling and Docking Workflow

Table of Contents

1. Downloading FASTA Files
2. Folding Proteins
3. Organizing PDB Files
4. Cleaning PDB Files
5. Organizing Cleaned PDB Files
6. Relaxing Structures
7. Selecting Lowest Energy Models
8. Preparing for Docking
9. Docking Process
10. Final Steps

File Locations

main_folder/
├── getfastas.py
├── run_esmfold.sh
└── pdbs/
    ├── bulkrelax.py
    ├── organizecleanedpdbs.py
    ├── organizeenergies.py
    ├── organizepdbs.py
    └── lowest_energies/
        ├── makeresultsfolder.sh
        └── getscripts.py

Downloading FASTA Files

First, you need to download the FASTA files for all of the proteins you are using. Since we are getting them for this project from a list from PDB, we can use the script get_fastas.py documents for which can be found here. Run the script using:

python get_fastas.py

If you are using this workflow for another project, procure the FASTA files yourself.

Folding Proteins

Once you have all of the FASTA files, you can fold them using the script run_esmfold.sh documents for which can be found here. Run the script using:

bash run_esmfold.sh $fastapath $outputdirectory

Organizing PDB Files

Your PDB files might be disorganized, so you can use the organizepdbs.py script to organize everything. The script can be found here. Run the script using:

python organizepdbs.py

Cleaning PDB Files

Next, everything needs to be cleaned. Use the script cleanpdbs.sh documents for which can be found here. Run the script using:

bash cleanpdbs.sh

Organizing Cleaned PDB Files

Once everything is cleaned, we need to get them in the appropriate folders. This can be done with organizecleanedpdbs.py documents for which can be found here. Run the script using:

python organizecleanedpdbs.py

This will create new folders with the _cleaned tag for our later relaxing.

Relaxing Structures

Next, we need to relax everything. We can use the script bulkrelax.py documents for which can be found here. Run the script using:

python bulkrelax.py

This will submit SLURM jobs for each of the PDB files to relax them. IMPORTANT: Only do this step once, as it looks for PDB files within the cleaned folder. If you have new PDB files from your first run, it will submit a job for those too, potentially flooding CACAO. If you need to redo this step, delete everything in the cleaned folders and start from the cleaning step again.

Selecting Lowest Energy Models

Now we need to get the lowest energy model from the relaxed structures. This can be done using the script organizeenergies.py documents for which can be found here. Run the script using:

python organizeenergies.py

This script moves the lowest energy model to its own folder for docking.

Preparing for Docking

Next, we need to get the docking scripts from the other files. This can be done using the getscripts.py script documents for which can be found here in the lowest_energies folder. If you are copying this pipeline for your own proteins, you can view them in the docking_scripts folder and view their documentation here.

Docking Process

Remember, for docking, we need to have a results folder ready. You can use the makeresultsfolder.sh script documents for which can be found here. Run the script using:

bash makeresultsfolder.sh

If you prefer, you can create the results folder yourself before starting the docking.

Next, you need to merge your new protein with the ligand. There is no script for this step. We recommend using PyMOL:

1. Open the new protein file first (due to limitations with Rosetta).
2. Open the existing protein/ligand complex.
3. Align them.
4. Delete the existing protein.
5. Export the newly modeled protein complexed with the original ligand as a molecule under a PDB file.
6. Upload that back to the cluster and edit the run.sh file to include the name of the new PDB.
7. Add #SBATCH --array=1-10 at the bottom of the list of SLURM commands so that it runs in an array.

Final Steps

Finally, add the header to your PDB file by typing the following command:

cat headerfile inputpdb > outputfilename

If you are doing this with your own proteins, you must do this yourself.

To complete the process, run the following command:

sbatch run.sh

Wait for everything to finish, and you're done!