Repository navigation
Expand file tree
/
Copy pathgen_md_input_structures.py
More file actions
327 lines (275 loc) · 16.9 KB
/
Copy pathgen_md_input_structures.py
File metadata and controls
327 lines (275 loc) · 16.9 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
import argparse
import numpy as np
from Bio.PDB import PDBParser, PDBIO
import glob
import pandas as pd
import os
import pickle
from pathlib import Path
from sklearn.cluster import AgglomerativeClustering
import collections
from collections import defaultdict
import sys
import shutil
from scipy.spatial.distance import pdist
from pre_md_refine import refine_multiple_rounds
import json
from pymol import cmd
from custom_openfold_utils.pdb_utils import get_pdb_path_seq, get_af_disordered_domains, get_af_disordered_residues, get_af_nterm_disorder_tail_idx, get_af_cterm_disorder_tail_idx, get_af_disordered_tail_residues, get_ca_pairwise_dist, convert_pdb_to_mmcif, get_bfactor
asterisk_line = '******************************************************************************'
def remove_files(file_list):
for f in file_list:
print('removing old file: %s' % f)
os.remove(f)
def create_clustering_dist_matrix(ca_pdist_all):
#method derived from
#https://www.biorxiv.org/content/10.1101/2023.07.13.545008v2.supplementary-material
clustering_dist_matrix = np.zeros((ca_pdist_all.shape[0],ca_pdist_all.shape[0]))
for i in range(0,ca_pdist_all.shape[0]):
for j in range(i+1,ca_pdist_all.shape[0]):
pairwise_dist_diff = np.abs(ca_pdist_all[j,:] - ca_pdist_all[i,:])
pairwise_dist_diff[pairwise_dist_diff <= 1] = 0.
pairwise_dist_diff = np.sum(pairwise_dist_diff)
clustering_dist_matrix[i,j] = pairwise_dist_diff
return clustering_dist_matrix
def gen_md_input(conformation_info_dir, initial_pred_path, num_clusters, num_md_structures, plddt_threshold, disordered_threshold, remove_disordered_tails, max_candidate_conformations, overwrite, af_disordered_tail_residues_idx_custom=None):
if plddt_threshold is None and disordered_threshold is None:
md_structures_dir = '%s/md_starting_structures/num_clusters=%d/plddt_threshold=None/disordered_threshold=None' % (conformation_info_dir, num_clusters)
elif plddt_threshold is not None and disordered_threshold is None:
md_structures_dir = '%s/md_starting_structures/num_clusters=%d/plddt_threshold=%s/disordered_threshold=None' % (conformation_info_dir, num_clusters, str(plddt_threshold))
elif plddt_threshold is None and disordered_threshold is not None:
md_structures_dir = '%s/md_starting_structures/num_clusters=%d/plddt_threshold=None/disordered_threshold=%s' % (conformation_info_dir, num_clusters, str(disordered_threshold))
else:
md_structures_dir = '%s/md_starting_structures/num_clusters=%d/plddt_threshold=%s/disordered_threshold=%s' % (conformation_info_dir, num_clusters, str(plddt_threshold), str(disordered_threshold))
print("CREATING DIR: %s" % md_structures_dir)
os.makedirs(md_structures_dir, exist_ok=True)
if af_disordered_tail_residues_idx_custom is None:
try:
af_seq = get_pdb_path_seq(initial_pred_path, None)
af_disordered_domains_idx, _ = get_af_disordered_domains(initial_pred_path)
nterm_disorder_start_idx, nterm_disorder_end_idx = get_af_nterm_disorder_tail_idx(af_disordered_domains_idx)
cterm_disorder_start_idx, cterm_disorder_end_idx = get_af_cterm_disorder_tail_idx(af_disordered_domains_idx, len(af_seq))
print('nterminal disordered tail idx: %s-%s' % (str(nterm_disorder_start_idx),str(nterm_disorder_end_idx)))
print('cterminal disordered tail idx: %s-%s' % (str(cterm_disorder_start_idx),str(cterm_disorder_end_idx)))
af_disordered_tail_residues_idx = get_af_disordered_tail_residues(af_disordered_domains_idx, len(af_seq))
print('disordered tail residues idx:')
print(af_disordered_tail_residues_idx)
af_disordered_tail_residues_idx_complement = sorted(list(set(range(len(af_seq))) - set(af_disordered_tail_residues_idx)))
except ValueError as e:
print('TROUBLE PARSING AF PREDICTION %s' % initial_pred_path)
af_disordered_tail_residues_idx_complement = None
else:
af_seq = get_pdb_path_seq(initial_pred_path, None)
af_disordered_tail_residues_idx = af_disordered_tail_residues_idx_custom
af_disordered_tail_residues_idx_complement = sorted(list(set(range(len(af_seq))) - set(af_disordered_tail_residues_idx_custom)))
md_starting_structure_info_fname = '%s/md_starting_structure_info.pkl' % md_structures_dir
pdb_files = glob.glob('%s/openmm_refined_structures/*.pdb' % md_structures_dir)
json_files = glob.glob('%s/openmm_refined_structures/*.json' % md_structures_dir)
pdb_files.extend(glob.glob('%s/unrefined_structures/*.pdb' % md_structures_dir))
if os.path.exists(md_starting_structure_info_fname):
if not(overwrite):
print("SKIPPING CLUSTERING PROCEDURE, %s ALREADY EXISTS:" % md_starting_structure_info_fname)
return
else:
remove_files(pdb_files)
remove_files(json_files)
else:
remove_files(pdb_files)
remove_files(json_files)
pattern = "%s/conformation_info.pkl" % conformation_info_dir
files = glob.glob(pattern, recursive=True)
if len(files) == 0:
pattern = "%s/**/conformation_info.pkl" % conformation_info_dir
files = sorted(glob.glob(pattern, recursive=True))
if len(files) == 0:
print('conformation_info.pkl file(s) do not exist in dir: %s' % conformation_info_dir)
sys.exit()
else:
print('%d conformation_info.pkl files found under dir: %s' % ((len(files), conformation_info_dir)))
else:
print('conformation_info file(s) found:')
print(files)
#combine into single dictionary
for i in range(0,len(files)):
with open(files[i], 'rb') as f:
curr_conformation_info = pickle.load(f)
if i == 0:
conformation_info = curr_conformation_info.copy()
else:
curr_key = list(curr_conformation_info.keys())[0]
if curr_key not in conformation_info:
conformation_info.update(curr_conformation_info)
else:
conformation_info[curr_key].extend(curr_conformation_info[curr_key])
add_conformations = True
conformation_info_all = []
ca_pdist_all = []
for key in conformation_info:
for i,val in enumerate(conformation_info[key]):
pdb_path = val[0]
rmsd = val[1]
mean_plddt = val[2]
disordered_percentage = val[3]
if max_candidate_conformations is not None:
if len(conformation_info_all) >= max_candidate_conformations:
add_conformations = False
if add_conformations:
if plddt_threshold is None and disordered_threshold is None:
conformation_info_all.append([pdb_path,rmsd,mean_plddt])
ca_pdist = get_ca_pairwise_dist(pdb_path, residues_include_idx=af_disordered_tail_residues_idx_complement)
ca_pdist_all.append(ca_pdist)
else:
if plddt_threshold is not None and disordered_threshold is None:
if mean_plddt >= plddt_threshold:
conformation_info_all.append([pdb_path,rmsd,mean_plddt])
ca_pdist = get_ca_pairwise_dist(pdb_path, residues_include_idx=af_disordered_tail_residues_idx_complement)
ca_pdist_all.append(ca_pdist)
elif plddt_threshold is None and disordered_threshold is not None:
if disordered_percentage <= disordered_threshold:
conformation_info_all.append([pdb_path,rmsd,mean_plddt])
ca_pdist = get_ca_pairwise_dist(pdb_path, residues_include_idx=af_disordered_tail_residues_idx_complement)
ca_pdist_all.append(ca_pdist)
elif plddt_threshold is not None and disordered_threshold is not None:
if mean_plddt >= plddt_threshold and disordered_percentage <= disordered_threshold:
conformation_info_all.append([pdb_path,rmsd,mean_plddt])
ca_pdist = get_ca_pairwise_dist(pdb_path, residues_include_idx=af_disordered_tail_residues_idx_complement)
ca_pdist_all.append(ca_pdist)
total_candidate_conformations = len(conformation_info_all)
print("%d TOTAL CANDIDATE CONFORMATIONS" % total_candidate_conformations)
print("Building C-alpha pairwise distance matrix")
ca_pdist_all = np.array(ca_pdist_all)
clustering_dist_matrix = create_clustering_dist_matrix(ca_pdist_all)
print("Performing Agglomerative Clustering")
clustering = AgglomerativeClustering(n_clusters=num_clusters, metric="precomputed", linkage="average").fit(clustering_dist_matrix)
print(asterisk_line)
print('CLUSTER FREQUENCIES:')
counter = collections.Counter(clustering.labels_)
print(counter.most_common())
print(asterisk_line)
clustering_info_dict = {}
for idx,cluster_num in enumerate(clustering.labels_):
if cluster_num in clustering_info_dict:
clustering_info_dict[cluster_num].append(conformation_info_all[idx])
else:
clustering_info_dict[cluster_num] = [conformation_info_all[idx]]
for cluster_num in clustering_info_dict:
clustering_info_dict[cluster_num] = sorted(clustering_info_dict[cluster_num], key=lambda x: x[2], reverse=True) #sort by plddt
cluster_idx_dict = {}
for cluster_num,freq in counter.most_common():
cluster_idx_dict[cluster_num] = 0
md_starting_structure_info_dict = defaultdict(list)
curr_md_structures_count = 0
while curr_md_structures_count < min(total_candidate_conformations,num_md_structures):
for cluster_num,freq in counter.most_common():
curr_cluster_idx = cluster_idx_dict[cluster_num]
if len(clustering_info_dict[cluster_num]) > curr_cluster_idx and curr_md_structures_count < min(total_candidate_conformations,num_md_structures):
md_starting_structure_info_dict[cluster_num].append(clustering_info_dict[cluster_num][curr_cluster_idx])
cluster_idx_dict[cluster_num] += 1
curr_md_structures_count += 1
print("Adding structure from cluster %d, index %d" % (cluster_num, cluster_idx_dict[cluster_num]-1))
#add initial pred to md_starting_structure_info_dict
cif_path = convert_pdb_to_mmcif(initial_pred_path, './cif_temp')
_, mean_plddt = get_bfactor(cif_path)
mean_plddt = round(mean_plddt,2)
os.remove(cif_path)
md_starting_structure_info_dict['initial'] = [[initial_pred_path, 0, mean_plddt]]
print(asterisk_line)
print('MD STARTING STRUCTURE INFO:')
for key, value in md_starting_structure_info_dict.items():
print(f"Cluster number: {key}, Number of elements: {len(value)}")
print(asterisk_line)
structural_issues_info_summary = {}
md_starting_structure_info_dict_keys = list(md_starting_structure_info_dict.keys())
md_starting_structure_info_dict_keys.remove('initial')
md_starting_structure_info_dict_keys.insert(0,'initial')
unrefined_output_dir = '%s/unrefined_structures' % md_structures_dir
refined_output_dir = '%s/openmm_refined_structures' % md_structures_dir
os.makedirs(unrefined_output_dir, exist_ok=True)
os.makedirs(refined_output_dir, exist_ok=True)
for cluster_num in md_starting_structure_info_dict_keys:
for i in range(0,len(md_starting_structure_info_dict[cluster_num])):
if cluster_num != 'initial':
pdb_source_path = md_starting_structure_info_dict[cluster_num][i][0]
plddt = md_starting_structure_info_dict[cluster_num][i][2]
output_fname = 'cluster_%d_idx_%d_plddt_%s' % (cluster_num, i, str(round(plddt)))
else:
pdb_source_path = md_starting_structure_info_dict[cluster_num][i][0]
output_fname = 'initial_pred'
pdb_target_path = '%s/%s.pdb' % (unrefined_output_dir, output_fname)
shutil.copyfile(pdb_source_path, pdb_target_path)
if remove_disordered_tails:
cmd.reinitialize()
cmd.load(pdb_target_path, 'protein')
if len(af_disordered_tail_residues_idx) > 0:
disordered_tail_residues_str = [str(res+1) for res in af_disordered_tail_residues_idx]
residue_selection = '+'.join(disordered_tail_residues_str)
cmd.remove(f"resi {residue_selection}")
print(f"removed residues: {disordered_tail_residues_str}")
cmd.save(pdb_target_path, 'protein')
cmd.delete('all')
md_unrefined_structure_path = os.path.abspath(pdb_target_path)
md_refined_structure_path = md_unrefined_structure_path.replace('.pdb', '_openmm_refinement.pdb')
md_refined_structure_path = md_refined_structure_path.replace('unrefined_structures','openmm_refined_structures')
print('****************************************')
print('****************************************')
print("ATTEMPTING TO RESOLVE ANY STRUCTURAL ISSUES WITH %s" % md_unrefined_structure_path)
print('****************************************')
print('****************************************')
if cluster_num == 'initial':
structural_issues_info = refine_multiple_rounds(md_unrefined_structure_path, md_refined_structure_path, add_disulfide_bonds=True, disulfide_bond_residues_idx=None)
disulfide_bond_residues_idx_derived_from_initial_pred = structural_issues_info['disulfide_bond_ids']
else:
if len(disulfide_bond_residues_idx_derived_from_initial_pred) > 0:
structural_issues_info = refine_multiple_rounds(md_unrefined_structure_path, md_refined_structure_path, add_disulfide_bonds=True, disulfide_bond_residues_idx=disulfide_bond_residues_idx_derived_from_initial_pred)
else:
structural_issues_info = refine_multiple_rounds(md_unrefined_structure_path, md_refined_structure_path, add_disulfide_bonds=False)
md_starting_structure_info_dict[cluster_num][i].append(md_refined_structure_path)
structural_issues_info_summary[md_refined_structure_path] = structural_issues_info
print(structural_issues_info_summary)
output_file = '%s/structural_issues_info_summary.json' % md_structures_dir
print('saving %s' % output_file)
with open(output_file, 'w') as json_file:
json.dump(structural_issues_info_summary, json_file, sort_keys=True, indent=4)
md_starting_structure_info_fname = '%s/md_starting_structure_info.pkl' % md_structures_dir
with open(md_starting_structure_info_fname, 'wb') as f:
pickle.dump(md_starting_structure_info_dict, f)
if __name__ == "__main__":
parser = argparse.ArgumentParser()
parser.add_argument(
"--conformation_info_dir", type=str, default=None,
help="Either points directly to the parent directory containing conformation_info.pkl or an nth-parent directory if multiple conformation_info.pkl are present",
)
parser.add_argument(
"--initial_pred_path", type=str, default=None,
help="Points directly to the vanila AF prediction",
)
parser.add_argument(
"--num_clusters", type=int, default=10,
help="",
)
parser.add_argument(
"--num_md_structures", type=int, default=50,
help="number of structures to use as starting points for MD"
)
parser.add_argument(
"--plddt_threshold", type=float, default=70,
help="only clusters conformations whose plddt score is greater than this threshold (which is a value between 0-100)",
)
parser.add_argument(
"--disordered_threshold", type=float, default=70,
help="only clusters conformations whose disordered percentage is less than this threshold (which is a value between 0-100)",
)
parser.add_argument(
"--remove_disordered_tails", action="store_true", default=False
)
parser.add_argument(
"--overwrite", action="store_true", default=False
)
parser.add_argument(
"--max_candidate_conformations", type=int, default=None,
help="maximum number of conformations to use as candidate pool for selecting MD starting structures"
)
parser.add_argument('--af_disordered_tail_residues_idx_custom', nargs='+', default=None,
help='List of prespecified disordered tail residues indices to remove')
args = parser.parse_args()
gen_md_input(args.conformation_info_dir, args.initial_pred_path, args.num_clusters, args.num_md_structures, args.plddt_threshold, args.disordered_threshold, args.remove_disordered_tails, args.max_candidate_conformations, args.overwrite, args.af_disordered_tail_residues_idx_custom)