GROMACS-on-Colab

This repository provides four notebooks for running molecular dynamics simulations on Google Colab.

• Build_to_Google_Drive.ipynb
  Installs GROMACS to your Google Drive, where it can be loaded by the other notebooks.
  

• GROMACS_for_CHARMM-GUI.ipynb
  Equilibrates a system prepared with CHARMM-GUI. This notebook unpacks a CHARMM-GUI.tgz archive from "Solution Builder." Optionally, it can merge in an archive from "Ligand Reader," allowing for piecewise preparation of protein-ligand systems.
  

• GROMACS_for_production.ipynb
  Starts or resumes a production simulation.
  

• Trajectory_analysis_tools.ipynb
  Calculates data that can be derived from a production simulation trajectory, such as centroid structures, RMSDs, and interaction energies.
  

All inputs and outputs are stored on your Google Drive.

---

Tutorial

The following is a quick-start guide describing how to simulate a docked protein-ligand complex.

1. Find or make a protein structure.
   It should be complete (no gaps in the chain), though it's fine if a small number of residues at the termini are missing. You might use the Protein Data Bank or the AlphaFold Protein Structure Database.
   ↪ protein.pdb

2. Dock a ligand to the protein.
   You might use AutoDock Vina. If you need to convert the inputs to .mol2 or .pdbqt, or the output to .sdf, you might use Open Babel.
   ↪ docked_ligand.sdf

3. Process the protein in CHARMM-GUI Solution Builder or, if it's a membrane protein, Membrane Builder.
   The protein structure uploaded must be the one docked against (so its atomic coordinates line up with the ligand's). There is typically no need to change default settings or suggested values, except that you must enable GROMACS compatible outputs when you reach the last step.
   ↪ protein_CHARMM.tgz

4. Process the docked ligand in CHARMM-GUI Ligand Reader.
   The ligand structure uploaded must be the output from docking (so its atomic coordinates line up with the protein's). You must keep the default residue name: LIG for a custom ligand, or the canonical name for a ligand already parameterised in the CHARMM36 forcefield.
   ↪ docked_ligand_CHARMM.tgz

5. Upload the .tgz archives to Google Drive.
   ↪ {GoogleDrive}/protein_CHARMM.tgz and {GoogleDrive}/docked_ligand_CHARMM.tgz

6. Install GROMACS-on-Colab to your Google Drive using Build_to_Google_Drive.ipynb.
   ↪ {GoogleDrive}/gromacs-on-colab/

7. Merge and equilibrate the protein and ligand using GROMACS_for_CHARMM-GUI.ipynb.
   ↪ {GoogleDrive}/protein-ligand_project_folder/, containing conf.gro and other inputs for GROMACS

8. Run the production simulation using GROMACS_for_production.ipynb.
   ↪ {GoogleDrive}/protein-ligand_project_folder/sim_reference.xtc

9. Visualise the trajectory.
   This is typically achieved by loading conf.gro and sim_reference.xtc into a molecular visualisation software. You might use VMD.

10. Optionally, analyse the trajectory using Trajectory_analysis_tools.ipynb.
    These utilities allow you to compute the representative centroid frame for a timespan; to measure and plot the evolution of an observable, such as an interatomic distance or the RMSD of a group; or to summarise the interaction energy terms at a given time.