Streamlined version of VirtualFlow combining both VFVS and VFLP
Please clone the repository using:
git clone [email protected]:VirtualFlow/VF-Unity.git
Please ensure that the following packages are installed:
run_vf_unity.py: Main file that initiates docking calculations.initiate_calc.py: File that initiates scoring, docking calculations.lig_process.py: Process provided ligand into 3D format compatible with docking program (used by )pose_prediction.py: File for running pose prediction on processed ligands (used by pose_prediction).scoring_functions.py: File for running scoring on already docked ligands (used by initiate_calc.py).config.txt: Config file concisting of user definable parameters for running calculation./ligands/: Directory created by VF-Unity which will contain all processed ligands in a ready-to-dock format./outputs/: Directory created by VF-Unity which will contain docked ligand files.
We will be running QuickVina on a processed protein located in the config directory (5wiu_test.pdbqt). We provide a config.txt file which contains all parmeters for this simple run.
Please edit this file based on your preferance:
# The choice of the docking method
# Possible choices:
# AutoDock-Koto, AutodockVina_1.2, AutodockZN, EquiBind, FRED
# FitDock, GalaxyDock3, LigandFit, LightDock, M-Dock
# MCDock, MM-GBSA, PLANTS, PSOVina, RLDock
# SEED, adfr, autodock_cpu, autodock_gpu, AutodockVina_1.1.2
# dock6, flexx, gnina, gnina-scoring
# gold, gwovina, iGemDock, idock, ledock
# molegro, nnscore2, qvina, qvina-w, rDock
# rf-score, rosetta-ligand, smina, smina-scoring, vina
# vina_carb, vina_xb, GlideSP, GlideXP, GlideHTVS
# qvina_gpu, qvina_w_gpu, vina_gpu, vina_gpu_2.0
# The choice of the scoring function
# Possible choices:
# nnscore2, rf-score, smina-scoring, ad4_scoring, vinandro_scoring
# vina_scoring, gnina_scoring, chemplp_scoring, PLP_scoring, PLP95_scoring
# contact_scoring, continuous_scoring, grid_scoring,
# asp, chemscore, goldscore, plp, mm_gbsa_scoring, Hawkins_gbsa
# Please note: different pose prediction/docking methods can be combined with scoring functions.
# For example: ’qvina+nnscore2’.
# For supported choices/combinations please see the VirtualFlow homepage.
program_choice=qvina+nnscore2
# The x,y&z coordinates of the center of the docking space. The binding space describes the location where a molecule
# is allowed to bind.
center_x=-17.820
center_y=16.140
center_z=-18.643
# The size (in Angstroms) of the docking space in the x,y&z directions.
size_x=20
size_y=20
size_z=20
# How many poses to search for, providing a limit to the maximum number of iterations that a docking program performs in
# search for good poses
# Large exhaustiveness settings lead to increased computational costs.
exhaustiveness=10
# Molecule (either a string in smiles, selfies or amino-acid sequence) to be used for docking
# If the is_selfies=True, or is_peptide=True, a conversion from selfies->smiles
# and aa-sequence->smiles is performed.
smi=C1=CC(=CC=C1CSCC2C(C(C(O2)N3C=NC4=C(N=CN=C43)N)O)O)Cl
is_selfies=False
is_peptide=False
# Location to the prepared receptor file
# The receptor needs to be in the correct format, supported by the user's selected docking program.
# Additionally, the file needs to be present in the config directory.
receptor=./config/5wiu_test.pdbqt
To execute the program, please run:
python3 run_vf_unity.py
We note:
- The processed ligands will be located within the newly created ligands directory.
- The docked output from running QuickVina will be located in the newly created outputs directory.
- The default behaviour is for the program (QuickVina) is to make use of all available CPUs.
- A summary csv file
docking_output.csvis created (for running QuickVina):Ligand File,Docking Values,Docking Pose 7.pdbqt,"-9.9,-9.8,-9.1,-9.0,-9.0,-8.8,-8.8,-8.7,-8.6",./outputs/pose_7.pdbqt 14.pdbqt,"-9.6,-9.1,-9.0,-8.8,-8.7,-8.6,-8.6,-8.4,-8.4",./outputs/pose_14.pdbqt 6.pdbqt,"-9.8,-9.7,-9.2,-8.9,-8.8,-8.8,-8.7,-8.7,-8.5",./outputs/pose_6.pdbqt 9.pdbqt,"-9.3,-9.3,-9.3,-9.1,-9.1,-9.0,-9.0,-8.9,-8.8",./outputs/pose_9.pdbqt 12.pdbqt,"-8.9,-8.9,-8.8,-8.5,-8.5,-8.5,-8.5,-8.1,-8.0",./outputs/pose_12.pdbqt 4.pdbqt,"-9.6,-9.3,-9.2,-9.2,-9.0,-9.0,-8.8,-8.7,-8.7",./outputs/pose_4.pdbqt 3.pdbqt,"-9.9,-9.2,-9.0,-8.9,-8.8,-8.7,-8.6,-8.5,-8.4",./outputs/pose_3.pdbqt 1.pdbqt,"-9.4,-9.4,-9.1,-9.0,-8.9,-8.8,-8.7,-8.7,-8.7",./outputs/pose_1.pdbqt 5.pdbqt,"-9.8,-9.5,-9.4,-9.3,-9.3,-9.2,-9.1,-9.1,-9.0",./outputs/pose_5.pdbqt 10.pdbqt,"-9.6,-9.5,-9.4,-9.3,-9.2,-9.2,-9.1,-8.9,-8.8",./outputs/pose_10.pdbqt 2.pdbqt,"-9.6,-9.3,-8.9,-8.9,-8.8,-8.8,-8.8,-8.6,-8.5",./outputs/pose_2.pdbqt 15.pdbqt,"-9.8,-9.3,-9.2,-9.2,-9.1,-9.0,-9.0,-8.8,-8.8",./outputs/pose_15.pdbqt 13.pdbqt,"-9.1,-9.1,-9.0,-8.8,-8.7,-8.7,-8.7,-8.7,-8.7",./outputs/pose_13.pdbqt 8.pdbqt,"-9.1,-9.1,-9.1,-8.8,-8.7,-8.7,-8.6,-8.6,-8.5",./outputs/pose_8.pdbqt 11.pdbqt,"-9.7,-9.5,-9.3,-9.2,-9.2,-9.1,-9.0,-8.9,-8.8",./outputs/pose_11.pdbqt 0.pdbqt,"-9.6,-9.3,-9.3,-9.0,-8.9,-8.9,-8.7,-8.7,-8.5",./outputs/pose_0.pdbqt - Sepperately, a summary csv file is created (
rescoring_output.csv) is created when running rescoring (nnscore2):Docked Ligand,Re-scored Value ./outputs/pose_11.pdbqt,Kd = 114.94 fM;Kd = 2.29 pM;Kd = 3.22 fM;Kd = 10.3 fM;Kd = 11.92 fM;Kd = 0.87 fM;Kd = 307.09 fM;Kd = 2.18 pM;Kd = 22.69 pM ./outputs/pose_10.pdbqt,Kd = 14.22 fM;Kd = 13.13 fM;Kd = 10.96 pM;Kd = 29.48 fM;Kd = 73.28 fM;Kd = 2.35 fM;Kd = 108.96 fM;Kd = 122.82 fM;Kd = 0.35 fM ./outputs/pose_9.pdbqt,Kd = 216.4 fM;Kd = 685.98 fM;Kd = 155.95 fM;Kd = 116.24 fM;Kd = 72.66 fM;Kd = 189.04 pM;Kd = 4.74 pM;Kd = 7.01 pM;Kd = 1.04 nM ./outputs/pose_5.pdbqt,Kd = 0.14 fM;Kd = 3.58 pM;Kd = 2.75 fM;Kd = 23.24 fM;Kd = 7.79 pM;Kd = 152.06 nM;Kd = 0.12 fM;Kd = 121.77 fM;Kd = 18.27 pM ./outputs/pose_12.pdbqt,Kd = 0.16 fM;Kd = 11.43 fM;Kd = 1.76 pM;Kd = 4.74 pM;Kd = 2.15 pM;Kd = 4.79 pM;Kd = 364.24 fM;Kd = 55.44 fM;Kd = 20.77 fM ./outputs/pose_7.pdbqt,Kd = 101.4 fM;Kd = 0.12 fM;Kd = 1.14 fM;Kd = 134.77 fM;Kd = 0.55 fM;Kd = 10.53 nM;Kd = 0.0 fM;Kd = 129.84 fM;Kd = 8.86 nM ./outputs/pose_13.pdbqt,Kd = 0.34 fM;Kd = 0.13 fM;Kd = 1.74 pM;Kd = 73.35 fM;Kd = 203.96 fM;Kd = 11.28 pM;Kd = 3.75 pM;Kd = 85.11 fM;Kd = 5.17 pM ./outputs/pose_0.pdbqt,Kd = 19.81 fM;Kd = 0.01 fM;Kd = 0.06 fM;Kd = 14.96 fM;Kd = 0.1 fM;Kd = 578.15 fM;Kd = 120.14 fM;Kd = 24.31 fM;Kd = 3.01 pM ./outputs/pose_4.pdbqt,Kd = 12.63 fM;Kd = 2.81 pM;Kd = 0.0 fM;Kd = 52.42 fM;Kd = 0.96 fM;Kd = 310.52 fM;Kd = 36.57 pM;Kd = 17.94 fM;Kd = 0.13 fM ./outputs/pose_6.pdbqt,Kd = 0.91 fM;Kd = 8.34 pM;Kd = 0.14 fM;Kd = 0.19 fM;Kd = 0.02 fM;Kd = 0.0 fM;Kd = 16.4 fM;Kd = 60.84 fM;Kd = 1.34 pM ./outputs/pose_8.pdbqt,Kd = 94.53 fM;Kd = 23.83 pM;Kd = 6.22 pM;Kd = 1.34 fM;Kd = 276.47 fM;Kd = 0.63 fM;Kd = 71.78 fM;Kd = 5.64 fM;Kd = 18.55 pM ./outputs/pose_1.pdbqt,Kd = 0.13 fM;Kd = 0.09 fM;Kd = 11.47 pM;Kd = 11.69 fM;Kd = 5.36 pM;Kd = 8.35 pM;Kd = 2.44 pM;Kd = 4.19 pM;Kd = 23.34 pM ./outputs/pose_3.pdbqt,Kd = 0.0 fM;Kd = 1.65 fM;Kd = 14.99 fM;Kd = 0.24 fM;Kd = 213.73 pM;Kd = 6.34 pM;Kd = 472.1 fM;Kd = 329.75 pM;Kd = 261.08 fM ./outputs/pose_15.pdbqt,Kd = 1.06 pM;Kd = 136.55 fM;Kd = 1.82 pM;Kd = 106.86 fM;Kd = 4.26 pM;Kd = 1.55 pM;Kd = 6.23 pM;Kd = 61.89 fM;Kd = 242.37 pM ./outputs/pose_2.pdbqt,Kd = 11.5 fM;Kd = 18.0 fM;Kd = 2.34 pM;Kd = 783.94 fM;Kd = 202.75 fM;Kd = 79.9 fM;Kd = 210.65 fM;Kd = 13.13 fM;Kd = 511.8 fM ./outputs/pose_14.pdbqt,Kd = 162.12 fM;Kd = 726.22 fM;Kd = 10.38 fM;Kd = 432.15 fM;Kd = 0.06 fM;Kd = 9.24 pM;Kd = 3.07 pM;Kd = 23.07 nM;Kd = 7.97 pM
from run_vf_unity import main
program_choice = 'qvina'
scoring_function = 'nnscore2'
center_x = -17.820
center_y = 16.140
center_z = -18.643
size_x = 20
size_y = 20
size_z = 20
exhaustiveness = 10
smi = 'C1=CC(=CC=C1CSCC2C(C(C(O2)N3C=NC4=C(N=CN=C43)N)O)O)Cl'
is_selfies = False
is_peptide = False
receptor = './config/5wiu_test.pdbqt'
pose_pred_out, re_scored_values = main(program_choice, scoring_function, center_x, center_y, center_z, size_x, size_y, size_z, exhaustiveness, smi, is_selfies, is_peptide, receptor)
We note:
- The processed ligands will be located within the newly created ligands directory.
- The docked output from running QuickVina will be located in the newly created outputs directory.
- The default behaviour is for the program (QuickVina) is to make use of all available CPUs.
- The output from running th QuickVina calculation will be stored in the dictionary pose_pred_out.
- The output from running th QuickVina NNScore2.0 will be stored in the dictionary re_scored_values.
- No output csv files are created in this case.
Please prepare a file concisting of a list of smiles. For example, molecules.txt:
Index,Smiles
0,CCCCCCCCCC
1,C1=CC(=CC=C1CSCC2C(C(C(O2)N3C=NC4=C(N=CN=C43)N)O)O)Cl
The molecules can be run simply using out python function call:
import os
from run_vf_unity import main
program_choice = 'qvina'
scoring_function = ''
center_x = -17.820
center_y = 16.140
center_z = -18.643
size_x = 20
size_y = 20
size_z = 20
exhaustiveness = 10
is_selfies = False
is_peptide = False
receptor = './config/5wiu_test.pdbqt'
with open('./molecules.smi', 'r') as f:
lines = f.readlines()
lines = lines[1: ]
for item in lines:
idx,smi = item.split(',')
pose_pred_out, re_scored_values = main(program_choice, scoring_function, center_x, center_y, center_z, size_x, size_y, size_z, exhaustiveness, smi, is_selfies, is_peptide, receptor)
os.system('rm -rf ligands')
os.system('cp -a outputs outputs_{}'.format(idx))
The corresponding index (column 1) of a molecule in the molecules.txt file will be used for storing the results in outputs_*index*
Please compile the code using instructions from: https://github.com/ccsb-scripps/AutoDock-GPU.
After successfull compilation, within the bin directory, an executable will be made (example name: autodock_gpu_1wi). Then, the code is ready to run.
We provide an example inside ./executables/vf_gpu_example.zip. Inside the directory, a prepared protein-ligand pair is provided and the code can be run
using: ./autodock_gpu_1wi --ffile 1stp_protein.maps.fld --lfile ./1stp_ligand.pdbqt
Please download the code using instructions from: https://github.com/HannesStark/EquiBind. Please create a conda enviroment per the instructions of the EquiBind repository. Copy paste all files inside the working directory of VF-Unity.
Please compile the code using the instruction provided in: https://rdock.sourceforge.net/installation/.
We installed rDock using anacond (with conda install -c bioconda rdock)
Please install AmberTools: https://ambermd.org/GetAmber.php#ambertools. We managed performed the download using conda.
Additionally, please note: the variable chimera_path should be updated to location of the Chimera on your system. Chimera can be downloaded: https://www.cgl.ucsf.edu/chimera/download.html.
Requirments
Please install the ADFR software suite: https://ccsb.scripps.edu/adfr/downloads/.
Please install meeko with: pip install meeko.
Instructions for running AutoDockZN can be found on their official website: https://autodock-vina.readthedocs.io/en/latest/docking_zinc.html.
Requirments Please install the OpenEye software suite: https://docs.eyesopen.com/toolkits/python/quickstart-python/install.html.
Requirments Please install AmberTools: https://ambermd.org/GetAmber.php#ambertools.
Please paste the executable from https://github.com/oddt/rfscorevs_binary in the executables directory.
A qvina_gpu executable (of name 'qvina_gpu') should be placed in directory: /executables Instructions for compilation are provided in https://github.com/DeltaGroupNJUPT/QuickVina2-GPU
A qvina_gpu executable (of name 'qvina_w_gpu') should be placed in directory: /executables Instructions for compilation are provided in https://github.com/DeltaGroupNJUPT/QVina-W-GPU
A qvina_gpu executable (of name 'vina_gpu') should be placed in directory: /executables Instructions for compilation are provided in https://github.com/DeltaGroupNJUPT/Vina-GPU
A qvina_gpu executable (of name 'vina_gpu_2.0') should be placed in directory: /executables Instructions for compilation are provided in https://github.com/DeltaGroupNJUPT/Vina-GPU-2.0
An adfr executable (of name adfr) should be placed in directory: /executables directory. An executable needs to be compiled based on a user’s system using instructions described in https://ccsb.scripps.edu/adfr/downloads/.
An HDock executable (of name hdock) should be placed in directory: /executables directory. An createpl executable (of name createpl) should be placed in directory: /executables directory.
Requirments Please install OpenEye: https://docs.eyesopen.com/toolkits/python/quickstart-python/install.html. Additionally, a valid OpenEye licence is required. Namely, a file named oe_license.txt needs to be placed in the working directory.
Please paste the executable from https://galaxy.seoklab.org/files/by2hsnvxjf/softwares/galaxydock.html in the executables directory. Note a data directory (name 'data') is required for successful runs.
We suggest downloading LightDock from: https://github.com/lightdock/lightdock. For this, within the current working directory (VF-Unity), please run:
git clone https://github.com/lightdock/lightdock.git
virtualenv venv
source venv/bin/activate
cd lightdock
pip install -e .
Please add the executable (of name MpSDock) in executables directory.
Please install AmberTools: https://ambermd.org/GetAmber.php#ambertools.
We managed performed the download using conda.
Please note: the variable chimera_path should be updated to location of the Chimera on your system. Chimera can be downloaded: https://www.cgl.ucsf.edu/chimera/download.html.
A working dock6 download (with a valid licence) is required. The variable dock6_path should be updated to location of the Chimera on your system.
A box.in input file is required for the program. Please specify the path in variable box_in_file.
A grid.in input file is required for the program. Please specify the path in variable grid_in_file.
A dock.in input file is required for the program. Please specify the path in variable dock_in_file.
A valid Schrödinger license is required to run CovDock.
A valid Schrödinger license is required to run GlideSP/XP/HTVS.
A valid GOLD license is required to re-score docking results with GOLD fitness functions.
An executable named gold_auto should be placed in the /executables directory.
When specifying a GOLD fitness function with which to re-score docking results, a dynamically loadable
shared object library must be available in the top-level directory of this repository and named correspondingly
with the parameter defined in the VFU configuration, i.e., one of plp, asp, chemscore, goldscore.
If you are interested in contributing to VirtualFlow, whether it is to report a bug or to extend VirtualFlow with your own code, please see the file CONTRIBUTING.md and the file CODE_OF_CONDUCT.md.
The project ist distributed under the GNU GPL v2.0. Please see the file LICENSE for more details.
Gorgulla, Christoph, et al. "VirtualFlow 2.0-The Next Generation Drug Discovery Platform Enabling Adaptive Screens of 69 Billion Molecules." bioRxiv (2023): 2023-04.