Mercurial > repos > bgruening > chembl_structure_pipeline

comparison chembl.py @ 0:2f59c6239f25 draft default tip

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
"planemo upload for repository https://github.com/chembl/chembl_webresource_client commit 78f2261af4e00c830ea311337d0aed9b297aad8e-dirty"
author bgruening
date Sat, 10 Oct 2020 09:43:40 +0000
parents
children
comparison
equal deleted inserted replaced
-1:000000000000 0:2f59c6239f25
1 import argparse
2
3 from chembl_webresource_client.new_client import new_client
4 from chembl_webresource_client.settings import Settings
5
6 Settings.Instance().CACHING = False
7
8
9 def open_file(filename):
10 with open(filename) as f:
11 return f.readline().split()[0]
12
13
14 def get_smiles(res):
15 """
16 Get a list of SMILES from function results
17 """
18 smiles = set()
19 for smi in res:
20 try:
21 smiles.add('{}\t{}'.format(smi['molecule_structures']['canonical_smiles'], smi['molecule_chembl_id']))
22 except TypeError:
23 continue
24 return smiles
25
26
27 def sim_search(smiles, tanimoto):
28 """
29 Return compounds which are within a Tanimoto range of the SMILES input
30 """
31 similarity = new_client.similarity
32 return similarity.filter(smiles=smiles, similarity=tanimoto).only(['molecule_structures', 'molecule_chembl_id'])
33
34
35 def substr_search(smiles):
36 """
37 Return compounds which contain the SMILES substructure input
38 """
39 substructure = new_client.substructure
40 return substructure.filter(smiles=smiles).only(['molecule_structures', 'molecule_chembl_id'])
41
42
43 def filter_drugs(mols):
44 """
45 Return only compounds which are approved drugs
46 """
47 return mols.filter(max_phase=4)
48
49
50 def filter_biotherapeutic(mols):
51 """
52 Return only biotherapeutic molecules
53 """
54 return mols.filter(biotherapeutic__isnull=False)
55
56
57 def filter_nat_prod(mols):
58 """
59 Return only natural products
60 """
61 return mols.filter(natural_product=1)
62
63
64 def filter_ro5(mols):
65 """
66 Return only compounds with no RO5 violations
67 """
68 return mols.filter(molecule_properties__num_ro5_violations=0)
69
70
71 def main():
72 parser = argparse.ArgumentParser(description='Search ChEMBL database for compounds')
73 parser.add_argument('-i', '--input', help='SMILES input')
74 parser.add_argument('-f', '--file', help='SMILES input as file')
75 parser.add_argument('-o', '--output', help="SMILES output")
76 parser.add_argument('-t', '--tanimoto', type=int, help='Tanimoto similarity score')
77 parser.add_argument('-s', '--substructure', action='store_true', help='Substructure search using the SMILES input.')
78 parser.add_argument('-d', '--drugs', action='store_true', help='Filter approved drugs')
79 parser.add_argument('-b', '--biotherapeutic', action='store_true', help='Filter biotherapeutic molecules')
80 parser.add_argument('-n', '--nat-prod', action='store_true', help='Filter natural products')
81 parser.add_argument('-r', '--ro5', action='store_true', help='Filter compounds that pass Lipinski RO5')
82
83 args = parser.parse_args()
84
85 if args.file: # get SMILES from file rather than -i option
86 args.input = open_file(args.file)
87
88 if len(args.input) < 5:
89 raise IOError('SMILES must be at least 5 characters long.')
90
91 if args.substructure: # specify search type: substructure or similarity
92 mols = substr_search(args.input)
93 else:
94 mols = sim_search(args.input, args.tanimoto)
95
96 # filter options:
97 if args.drugs:
98 mols = filter_drugs(mols)
99
100 if args.biotherapeutic:
101 mols = filter_biotherapeutic(mols)
102
103 if args.nat_prod:
104 mols = filter_nat_prod(mols)
105
106 if args.ro5:
107 mols = filter_ro5(mols)
108
109 # get SMILES from search output
110 mols = get_smiles(mols)
111
112 # write to file
113 with open(args.output, 'w') as f:
114 f.write('\n'.join(mols))
115
116
117 if __name__ == "__main__":
118 main()