Last active
March 9, 2020 23:16
-
-
Save fabian-paul/74a3797b4ce1b728ea507bb01bbd56b3 to your computer and use it in GitHub Desktop.
delete all water molecules from MD trajectory, except molecules close to ligand
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
#!/usr/bin/env python | |
import numpy as np | |
import mdtraj | |
def tl(filename, chunk=100): | |
with mdtraj.open(filename, mode='r') as fh: | |
try: | |
return (len(fh) - 1) // chunk + 1 | |
except: | |
return None | |
def strip_most_water(traj, n_water=200, ligand_selection='resname IMAT'): | |
# TODO: take additional measures to preserve the identity of water molecules if possible | |
top = traj.top | |
# get atom indices for the different selections | |
protein_idx = top.select('protein') | |
ligand_idx = top.select(ligand_selection) | |
water_idx = np.array([[a.index for a in r.atoms] for r in top.residues if r.is_water]) | |
assert water_idx.shape[1] == 3 | |
ions_idx = top.select('resname CLA or resname POT') | |
first_water_idx = water_idx[0:n_water].reshape(-1) | |
# prepare data structure for distance computation | |
atom_pairs = [] | |
water_table_idx = [] | |
for ri_ligand, i_ligand in enumerate(ligand_idx): | |
for ri_wat, wat_mol in enumerate(water_idx): | |
for ri_wat_atom, i_wat in enumerate(wat_mol): | |
atom_pairs.append([i_ligand, i_wat]) | |
water_table_idx.append([ri_wat, ri_ligand*3 + ri_wat_atom]) | |
atom_pairs = np.array(atom_pairs) | |
water_table_idx = np.array(water_table_idx) | |
# restrict trajectory data to protein, ligand, ions and closest waters | |
new_xyz = [] | |
all_dist_traj = mdtraj.compute_distances(traj, atom_pairs=atom_pairs) | |
assert len(traj) == len(all_dist_traj) | |
for all_dist_frame, frame in zip(all_dist_traj, traj.xyz): | |
# rearrange result into (water_index x rest) matrix | |
D = np.zeros(shape=(len(water_idx), len(ligand_idx)*3)) | |
D[water_table_idx[:, 0], water_table_idx[:, 1]] = all_dist_frame | |
# compute minimal distance | |
mindist = np.min(D, axis=1) | |
# get atom coordiantes | |
i_closest_water_atom_idx = water_idx[np.argsort(mindist)[0:n_water], :].reshape(-1) | |
new_frame = frame[np.concatenate((protein_idx, ligand_idx, i_closest_water_atom_idx, ions_idx)), :] | |
new_xyz.append(new_frame) | |
# store result in mdtraj.Trajectory object | |
reduced_frame = traj[0].atom_slice(np.concatenate((protein_idx, ligand_idx, first_water_idx, ions_idx))) | |
reduced_top = reduced_frame.top | |
new_traj = mdtraj.Trajectory(new_xyz, reduced_top, time=traj.time, unitcell_lengths=traj.unitcell_lengths, unitcell_angles=traj.unitcell_angles) | |
return new_traj | |
if __name__=='__main__': | |
import argparse | |
import os | |
from tqdm import tqdm | |
parser = argparse.ArgumentParser(description='wrap molecules, remove far-away water molecules', | |
formatter_class=argparse.ArgumentDefaultsHelpFormatter) | |
parser.add_argument('--top', metavar='filename', required=True, | |
help='(in) file name of the full topology file') | |
parser.add_argument('--nwat', metavar='number', default=200, type=int, | |
help='(in) number of water molecules to keep') | |
parser.add_argument('--output', metavar='trajname', default='(auto)', | |
help='(out) file name for output trajectory. Choosing xtc format enables full online processing.') | |
parser.add_argument('--input', metavar='trajname', required=True, nargs='+', | |
help='(in) file name of input trajectory') | |
parser.add_argument('--ligand', metavar='selection', default='resname IMAT or resname IMA', | |
help='(in) MDTraj selection string for the ligand') | |
args = parser.parse_args() | |
if args.output == '(auto)': | |
base = os.path.splitext(args.input[0])[0] | |
out_name = base + '.small.xtc' | |
else: | |
out_name = args.output | |
top = mdtraj.load_topology(args.top) | |
if out_name.endswith('.xtc'): | |
with mdtraj.formats.XTCTrajectoryFile(out_name, 'w') as f: # full online version | |
for fname_input in args.input: | |
print('processing trajectory piece', fname_input, 'frame count:', tl(fname_input, chunk=1)) | |
traj = mdtraj.iterload(fname_input, top=top) | |
for chunk in tqdm(traj, total=tl(fname_input)): | |
chunk = strip_most_water(chunk, n_water=args.nwat, ligand_selection=args.ligand) | |
chunk.image_molecules(inplace=True) | |
f.write(chunk.xyz, time=chunk.time, box=chunk.unitcell_vectors) | |
else: | |
out = None | |
for fname_input in args.input: | |
print('processing trajectory piece', fname_input, 'frame count:', tl(fname_input, chunk=1)) | |
traj = mdtraj.iterload(fname_input, top=top) | |
for chunk in tqdm(traj, total=tl(fname_input)): | |
chunk = strip_most_water(chunk, n_water=args.nwat, ligand_selection=args.ligand) | |
chunk.image_molecules(inplace=True) | |
if out is None: | |
out = chunk | |
else: | |
out = out.join(chunk) | |
out.save(out_name) |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment