Multiple Ligand Docking

Jupyter Notebook is available at : Multiple Docking¶

In [ ]:

from docksuitex.batch_docking import BatchProtein, BatchLigand, BatchPocketFinder, BatchVinaDocking, BatchAD4Docking
from docksuitex.utils import fetch_pdb, fetch_sdf, view_molecule, parse_vina_log_to_csv, parse_ad4_dlg_to_csv

from docksuitex.batch_docking import BatchProtein, BatchLigand, BatchPocketFinder, BatchVinaDocking, BatchAD4Docking from docksuitex.utils import fetch_pdb, fetch_sdf, view_molecule, parse_vina_log_to_csv, parse_ad4_dlg_to_csv

Fetch and prepare Protein¶

In [ ]:

# Fetch multiple protein structures (1HVR, 2TRX) in parallel
fetch_pdb(pdbid = ["1HVR", "2TRX"], save_to="proteins", parallel=2)

# Fetch multiple protein structures (1HVR, 2TRX) in parallel fetch_pdb(pdbid = ["1HVR", "2TRX"], save_to="proteins", parallel=2)

In [ ]:

# Initialize BatchProtein to process all proteins in the 'proteins' directory
# This cleans the PDB, removes water/heterogens, and adds hydrogens/charges

batch = BatchProtein(
    inputs="proteins",

    #Default parameters
    fix_pdb=True,
    remove_heterogens=True,
    remove_water=True,
    add_hydrogens=True,
    add_charges=True,
    preserve_charge_types=None, #eg: ["Zn", "Fe"]
)


# Run preparation for all proteins using 8 CPUs
batch.prepare_all(save_to="prepared_proteins", cpu = 8)

# Initialize BatchProtein to process all proteins in the 'proteins' directory # This cleans the PDB, removes water/heterogens, and adds hydrogens/charges batch = BatchProtein( inputs="proteins", #Default parameters fix_pdb=True, remove_heterogens=True, remove_water=True, add_hydrogens=True, add_charges=True, preserve_charge_types=None, #eg: ["Zn", "Fe"] ) # Run preparation for all proteins using 8 CPUs batch.prepare_all(save_to="prepared_proteins", cpu = 8)

Find Binding Pockets¶

In [ ]:

# Identify binding pockets for all prepared proteins using P2Rank
batch = BatchPocketFinder(inputs="prepared_proteins",  max_centres=3)

# Run pocket detection and return receptors with their pocket centers
receptors_with_centers = batch.run_all(save_to="pockets_output", cpu=8)

# Identify binding pockets for all prepared proteins using P2Rank batch = BatchPocketFinder(inputs="prepared_proteins", max_centres=3) # Run pocket detection and return receptors with their pocket centers receptors_with_centers = batch.run_all(save_to="pockets_output", cpu=8)

Fetch and prepare Ligand¶

In [ ]:

# Fetch multiple ligand structures (CIDs 228, 338, 339, 340) in parallel
fetch_sdf(cid =  ["228", "338", "339", "340"], save_to="ligands", parallel=2)

# Fetch multiple ligand structures (CIDs 228, 338, 339, 340) in parallel fetch_sdf(cid = ["228", "338", "339", "340"], save_to="ligands", parallel=2)

In [ ]:

# Initialize BatchLigand to process all ligands in the 'ligands' directory
# This minimizes the structure, removes water and adds hydrogens/charges

batch = BatchLigand(
    inputs="ligands",

    #Default parameters
    minimize=None,  #Options: "mmff94", "mmff94s", "uff", "gaff"
    remove_water=True,
    add_hydrogens=True,
    add_charges=True,
    preserve_charge_types=None,
)
# Run preparation for all ligands using 8 CPUs
batch.prepare_all(save_to="prepared_ligands", cpu=8)

# Initialize BatchLigand to process all ligands in the 'ligands' directory # This minimizes the structure, removes water and adds hydrogens/charges batch = BatchLigand( inputs="ligands", #Default parameters minimize=None, #Options: "mmff94", "mmff94s", "uff", "gaff" remove_water=True, add_hydrogens=True, add_charges=True, preserve_charge_types=None, ) # Run preparation for all ligands using 8 CPUs batch.prepare_all(save_to="prepared_ligands", cpu=8)

Dock each ligand against a each receptor at all the predicted protein pockets¶

Using Vina¶

In [ ]:

# Initialize BatchVinaDocking with prepared receptors (and their pockets) and ligands
batch = BatchVinaDocking(
    receptors_with_centers=receptors_with_centers,
    ligands="prepared_ligands",

    # Default Parameters
    grid_size=(20,20,20),
    exhaustiveness=8,
    num_modes=9, 
    seed=42
)


# Run Vina docking for all receptor-ligand pairs using 8 CPUs
results_vina = batch.run_all(save_to="vina_batch_results", cpu=8)


# Parse all Vina output logs into a single CSV summary
parse_vina_log_to_csv(results_dir="vina_batch_results", output_csv="vina_batch_results/vina_summary.csv")

# Initialize BatchVinaDocking with prepared receptors (and their pockets) and ligands batch = BatchVinaDocking( receptors_with_centers=receptors_with_centers, ligands="prepared_ligands", # Default Parameters grid_size=(20,20,20), exhaustiveness=8, num_modes=9, seed=42 ) # Run Vina docking for all receptor-ligand pairs using 8 CPUs results_vina = batch.run_all(save_to="vina_batch_results", cpu=8) # Parse all Vina output logs into a single CSV summary parse_vina_log_to_csv(results_dir="vina_batch_results", output_csv="vina_batch_results/vina_summary.csv")

Using AutoDock4¶

In [ ]:

# Initialize BatchAD4Docking with prepared receptors (and their pockets) and ligands
batch_ad4 = BatchAD4Docking(
    receptors_with_centers=receptors_with_centers,
    ligands="prepared_ligands",

    #Default Parameters
    grid_size=(60,60,60),
    spacing=0.375,     
    dielectric=-0.1465, 
    smooth=0.5,          
    ga_pop_size=150,
    ga_num_evals=2500000,
    ga_num_generations=27000,
    ga_elitism=1,
    ga_mutation_rate=0.02,
    ga_crossover_rate=0.8,
    ga_run=10,
    rmstol=2.0,
    seed=("pid", "time")
)


# Run AutoDock4 docking for all receptor-ligand pairs using 8 CPUs
results_ad4 = batch_ad4.run_all(save_to="ad4_batch_results", cpu=8)


# Parse all AD4 output logs into a single CSV summary
parse_ad4_dlg_to_csv(results_dir="ad4_batch_results", output_csv="ad4_batch_results/ad4_summary.csv")

# Initialize BatchAD4Docking with prepared receptors (and their pockets) and ligands batch_ad4 = BatchAD4Docking( receptors_with_centers=receptors_with_centers, ligands="prepared_ligands", #Default Parameters grid_size=(60,60,60), spacing=0.375, dielectric=-0.1465, smooth=0.5, ga_pop_size=150, ga_num_evals=2500000, ga_num_generations=27000, ga_elitism=1, ga_mutation_rate=0.02, ga_crossover_rate=0.8, ga_run=10, rmstol=2.0, seed=("pid", "time") ) # Run AutoDock4 docking for all receptor-ligand pairs using 8 CPUs results_ad4 = batch_ad4.run_all(save_to="ad4_batch_results", cpu=8) # Parse all AD4 output logs into a single CSV summary parse_ad4_dlg_to_csv(results_dir="ad4_batch_results", output_csv="ad4_batch_results/ad4_summary.csv")