A middleware python tool for computational drug discovery on HPC architectures. It automates the setup of hamiltonian replica exchange HREM and non equilibrium alchemical transformations for protein-ligand systems. Its primary use is to calculate the binding free energy between a protein and a ligand with the virtual double system single box method vDSSB.

At the moment it only works for Gromacs (with Plumed) but it can easily be adapted for any MD program that is supported by Parmed.

The latest version of the software can be found on GitHub here, and the documentation can be found here

HPC_Drug offers some out of the box scripts to set up a vDSSB by starting from a protein-ligand PDB file and a ligand SDF file (for example downloaded from the PDB databank). Of course you can also create some custom ones.

To use them activate the conda environment you created

conda activate HPC_Drug

copy the script(s) that you want in the root directory

cp scripts/<SCRIPT NAME>.py .

At this point, for every script, you can read the documentation with the -h or --help option.

python <SCRIPT NAME>.py -h

Briefly what you can do:

• Once you have a PDB of the protein-ligand system and an SDF file of the ligand (for example downloaded from the PDB databank) you can add missing atoms and residues, protonate, rename the residues, and solvate the system with prepare_pdb.py

• Parametrize the system with any forcefield supported by Openmm/Openmmforcefields with parametrize.py

• Set up an hamiltonian replica exchange (solute tempering) for the protein-ligand system and for the ligand alone with setup_hrem.py

• Starting from the HREM create the input for the non equilibrium alchemical transformations, both for the bound (protein-ligand) and the unbound (only ligand) systems, with fsdam_input.py

• Postprocess the results of the non equilibrium alchemical transformations in order to get a binding free energy value and the various volume and charge corrections with fsdam_postprocessing.py

Of course this scripts will never be able to cover every use case, therefore advanced users might need to develop custom scripts and use HPC_Drug, but more so the other python packages that HPC_Drug depends on and that you can find on my github as libraries (for experienced python developers it should be straightforward in most of the cases).

The easy way is to create a conda environment by using the environment.yml file. Because of the fact that this program has a lot of dependecies it might happen that the environment created in this way will not always be up to date.

The command to give on the terminal is

conda env create -f environment.yml

This command will create an environment called HPC_Drug. While creating the environment because of the large number of dependencies conda might complain about conflicts or take ages to create it, so you might have to play aroud a bit.

At this point you only have to copy one of the scripts from the scripts directory in the root directory of HPC_Drug (or you can create a custom one) and run it!

The main difference is that version 2 dropped the support for the Orac MD program and doesn't use Primadorac to parametrize the ligands anymore; instead, through the Openmmforcefields interface, it allows to parametrize the ligands with the Gaff forcefields through antechamber or with the Openff or Smirnoff99Frost ones through the openff-toolkit. Another difference is that the version one was more automathic but also less flexible than this version.

If for any reason you want to use the version 1 (that is the version referenced in the papers DOI: 10.1021/acs.jctc.0c00634, and DOI: 10.1021/acs.jcim.1c00909) switch to the branch version v1.0

If you find a bug please open an issue on GitHub

If you would like to contribute please first open an issue in order to discuss about it and then feel free to make a pull request, but please follow good coding practices and write some tests for the code you wrote.

GNU Affero General Public License (agpl v3)

See the LICENSE file

Please cite these papers:

1. Addressing Suboptimal Poses in Nonequilibrium Alchemical Calculations Maurice Karrenbrock, Valerio Rizzi, Piero Procacci, Francesco Luigi Gervasio The Journal of Physical Chemistry B 2024 DOI: 10.1021/acs.jpcb.3c06516

2. Virtual Double-System Single-Box: A Nonequilibrium Alchemical Technique for Absolute Binding Free Energy Calculations: Application to Ligands of the SARS-CoV-2 Main Protease Marina Macchiagodena, Marco Pagliai, Maurice Karrenbrock, Guido Guarnieri, Francesco Iannone, and Piero Procacci Journal of Chemical Theory and Computation 2020 16 (11), 7160-7172 DOI: 10.1021/acs.jctc.0c00634

3. Virtual Double-System Single-Box for Absolute Dissociation Free Energy Calculations in GROMACS Marina Macchiagodena, Maurice Karrenbrock, Marco Pagliai, and Piero Procacci Journal of Chemical Information and Modeling 2021 61 (11), 5320-5326 DOI: 10.1021/acs.jcim.1c00909