|
1
|
|
|
from inspect import cleandoc as doc |
|
|
|
|
|
|
2
|
|
|
from pathlib import Path |
|
3
|
|
|
from typing import Mapping, Optional |
|
4
|
|
|
|
|
5
|
|
|
import typer |
|
|
|
|
|
|
6
|
|
|
|
|
7
|
|
|
from mandos.model.chembl_support import DataValidityComment |
|
8
|
|
|
from mandos.model.chembl_support.chembl_targets import TargetType, ConfidenceLevel |
|
9
|
|
|
from mandos.search.chembl.target_traversal import TargetTraversalStrategies |
|
10
|
|
|
|
|
11
|
|
|
|
|
12
|
|
|
def _stringify(keys: Mapping[str, str]): |
|
13
|
|
|
return ", ".join((k if v is None else f"{k} ({v.lower()})" for k, v in keys.items())) |
|
|
|
|
|
|
14
|
|
|
|
|
15
|
|
|
|
|
16
|
|
|
class EntryArgs: |
|
|
|
|
|
|
17
|
|
|
|
|
18
|
|
|
path = typer.Argument( |
|
19
|
|
|
..., |
|
20
|
|
|
exists=True, |
|
21
|
|
|
dir_okay=False, |
|
22
|
|
|
readable=True, |
|
23
|
|
|
help=doc( |
|
24
|
|
|
""" |
|
25
|
|
|
The path to the input file. |
|
26
|
|
|
One of: |
|
27
|
|
|
|
|
28
|
|
|
(A) *.txt, *.lines, or *.list (optionally with .gz/.zip/.xz/.bz2)), with one InChI Key per line; |
|
|
|
|
|
|
29
|
|
|
|
|
30
|
|
|
(B) A *.csv, *.tsv, *.tab file (or .gz/.zip/.xz/.bz2 variant) with a column called 'inchikey'; OR |
|
|
|
|
|
|
31
|
|
|
|
|
32
|
|
|
(C) An Arrow *.feather file or Parquet *.snappy file with a column called 'inchikey' |
|
33
|
|
|
""" |
|
34
|
|
|
), |
|
35
|
|
|
) |
|
36
|
|
|
|
|
37
|
|
|
to = typer.Option( |
|
38
|
|
|
None, |
|
39
|
|
|
show_default=False, |
|
40
|
|
|
help=doc( |
|
41
|
|
|
""" |
|
42
|
|
|
The path to the output file. |
|
43
|
|
|
If not set, chooses <input-path>-<search>.csv.gz |
|
44
|
|
|
The filename extension should be one of: .csv, .tsv, .tab, .json (with optional .gz/.bz2/.zip/.xz); |
|
|
|
|
|
|
45
|
|
|
.feather; .snappy (or .parquet); or .h5. |
|
46
|
|
|
Feather (.feather), Parquet (.snappy), and tab-delimited (.tsv.gz) are recommended. |
|
47
|
|
|
If H5, will add a new dataset named <key> to any existing archive. |
|
48
|
|
|
Will fail if the file exists unless the `--overwrite` flag is set. |
|
49
|
|
|
|
|
50
|
|
|
If only the filename extension is provided (e.g. --to '.feather'), will only change the output format |
|
|
|
|
|
|
51
|
|
|
(and filename extension). |
|
52
|
|
|
""" |
|
53
|
|
|
), |
|
54
|
|
|
) |
|
55
|
|
|
|
|
56
|
|
|
replace: bool = typer.Option(False, help="Replace output file if they exist. See also: --skip") |
|
57
|
|
|
|
|
58
|
|
|
skip: bool = typer.Option( |
|
59
|
|
|
False, help="Skip any search if the output file exists (only warns). See also: --replace" |
|
60
|
|
|
) |
|
61
|
|
|
|
|
62
|
|
|
in_cache: bool = typer.Option( |
|
63
|
|
|
False, |
|
64
|
|
|
help="Do not download any data. Fails if the needed data is not cached.", |
|
65
|
|
|
hidden=True, |
|
66
|
|
|
) |
|
67
|
|
|
|
|
68
|
|
|
verbose: bool = typer.Option( |
|
69
|
|
|
False, |
|
70
|
|
|
"--verbose", |
|
71
|
|
|
help="Configure logger to output INFO (use ``--quiet`` for less info)", |
|
72
|
|
|
) |
|
73
|
|
|
|
|
74
|
|
|
quiet: bool = typer.Option( |
|
75
|
|
|
False, |
|
76
|
|
|
"--quiet", |
|
77
|
|
|
help="Configure logger to output only ERROR (use ``--verbose`` for more info)", |
|
78
|
|
|
) |
|
79
|
|
|
|
|
80
|
|
|
log_path: Optional[Path] = typer.Option( |
|
81
|
|
|
None, |
|
82
|
|
|
"--log", |
|
83
|
|
|
help=""" |
|
84
|
|
|
Also log to a file. |
|
85
|
|
|
The suffix can be .log, .log.gz, .log.zip, or .json, .json.gz, or .json.gz. |
|
86
|
|
|
You can prefix the path with :LEVEL: to control the level. For example, :INFO:out.log |
|
87
|
|
|
""", |
|
88
|
|
|
) |
|
89
|
|
|
|
|
90
|
|
|
no_setup: bool = typer.Option( |
|
91
|
|
|
False, |
|
92
|
|
|
"--no-setup", |
|
93
|
|
|
hidden=True, |
|
94
|
|
|
help="Skip setup, such as configuring logging.", |
|
95
|
|
|
) |
|
96
|
|
|
|
|
97
|
|
|
@staticmethod |
|
98
|
|
|
def key(name: str) -> typer.Option: |
|
|
|
|
|
|
99
|
|
|
return typer.Option( |
|
100
|
|
|
name, |
|
101
|
|
|
min=1, |
|
102
|
|
|
max=120, |
|
103
|
|
|
help=""" |
|
104
|
|
|
A free-text unique key for the search. |
|
105
|
|
|
Should be a short, <60-character name that describes the search and any parameters. |
|
106
|
|
|
The output file will be named according to a 'sanitized' variant of this value. |
|
107
|
|
|
""", |
|
108
|
|
|
) |
|
109
|
|
|
|
|
110
|
|
|
test = typer.Option( |
|
111
|
|
|
False, |
|
112
|
|
|
"--check", |
|
113
|
|
|
help="Do not run searches; just check that the parameters are ok.", |
|
114
|
|
|
) |
|
115
|
|
|
|
|
116
|
|
|
taxa = typer.Option( |
|
117
|
|
|
"7742", |
|
118
|
|
|
show_default=False, |
|
119
|
|
|
help=doc( |
|
120
|
|
|
""" |
|
121
|
|
|
The IDs or names of UniProt taxa, comma-separated. |
|
122
|
|
|
Taxon names and common names can be used for vertebrate species (where available). |
|
123
|
|
|
|
|
124
|
|
|
This can have a significant effect on searches. See the docs for more info. |
|
125
|
|
|
|
|
126
|
|
|
[default: 7742] (Euteleostomi) |
|
127
|
|
|
""" |
|
128
|
|
|
), |
|
129
|
|
|
) |
|
130
|
|
|
|
|
131
|
|
|
atc_level = typer.Option( |
|
132
|
|
|
"1,2,3,4", min=1, max=4, help="""List of ATC levels, comma-separated.""" |
|
133
|
|
|
) |
|
134
|
|
|
|
|
135
|
|
|
min_cooccurrence_score = typer.Option( |
|
136
|
|
|
0.0, |
|
137
|
|
|
help="Minimum enrichment score, inclusive. See docs for more info.", |
|
138
|
|
|
min=0.0, |
|
139
|
|
|
) |
|
140
|
|
|
|
|
141
|
|
|
min_cooccurring_articles = typer.Option( |
|
142
|
|
|
0, |
|
143
|
|
|
help="Minimum number of articles for both the compound and object, inclusive.", |
|
144
|
|
|
min=0, |
|
145
|
|
|
) |
|
146
|
|
|
|
|
147
|
|
|
name_must_match = typer.Option( |
|
148
|
|
|
False, |
|
149
|
|
|
help=doc( |
|
150
|
|
|
""" |
|
151
|
|
|
Require that the name of the compound(s) exactly matches the compound name on PubChem (case-insensitive) |
|
|
|
|
|
|
152
|
|
|
""" |
|
153
|
|
|
), |
|
154
|
|
|
) |
|
155
|
|
|
|
|
156
|
|
|
acute_effect_level = typer.Option( |
|
157
|
|
|
2, |
|
158
|
|
|
min=1, |
|
159
|
|
|
max=2, |
|
160
|
|
|
help=""" |
|
161
|
|
|
The level in the ChemIDPlus hierarchy of effect names. |
|
162
|
|
|
Level 1: e.g. 'behavioral' |
|
163
|
|
|
Level 2: 'behavioral: excitement' |
|
164
|
|
|
""", |
|
165
|
|
|
) |
|
166
|
|
|
|
|
167
|
|
|
traversal_strategy = typer.Option( |
|
168
|
|
|
"@null", |
|
169
|
|
|
"--traversal", |
|
170
|
|
|
show_default=False, |
|
171
|
|
|
help=doc( |
|
172
|
|
|
""" |
|
173
|
|
|
Target traversal strategy name, file, or class. |
|
174
|
|
|
Dictates the way the network of ChEMBL targets is traversed (from the annotated target as a source). |
|
|
|
|
|
|
175
|
|
|
Specifies the network links that are followed and which targets are 'accepted' for final annotations. |
|
|
|
|
|
|
176
|
|
|
This option has a dramatic effect on the search. See the docs for more info. |
|
177
|
|
|
|
|
178
|
|
|
Can be one of: |
|
179
|
|
|
(A) A standard strategy name, starting with @; |
|
180
|
|
|
(B) The path to a ``*.strat`` file; OR |
|
181
|
|
|
(C) The fully qualified name of a ``TargetTraversal`` |
|
182
|
|
|
|
|
183
|
|
|
The standard traversal strategies are: {} |
|
184
|
|
|
|
|
185
|
|
|
[default: @null] (No traversal; targets as-is) |
|
186
|
|
|
""".format( |
|
187
|
|
|
"; ".join(TargetTraversalStrategies.standard_strategies()) |
|
|
|
|
|
|
188
|
|
|
) |
|
189
|
|
|
), |
|
190
|
|
|
) |
|
191
|
|
|
|
|
192
|
|
|
target_types = typer.Option( |
|
193
|
|
|
"@molecular", |
|
194
|
|
|
"--targets", |
|
195
|
|
|
show_default=False, |
|
196
|
|
|
help=doc( |
|
197
|
|
|
""" |
|
198
|
|
|
The accepted target types, comma-separated. |
|
199
|
|
|
|
|
200
|
|
|
NOTE: This affects only the types are are accepted after traversal, |
|
201
|
|
|
and the types must be included in the traversal. |
|
202
|
|
|
This means that this must be AT LEAST as restrictive as the traversal strategy. |
|
203
|
|
|
|
|
204
|
|
|
The ChEMBL-defined types are: |
|
205
|
|
|
|
|
206
|
|
|
{} |
|
207
|
|
|
|
|
208
|
|
|
These special names are also accepted: |
|
209
|
|
|
|
|
210
|
|
|
- {} |
|
211
|
|
|
|
|
212
|
|
|
[default: @molecular] |
|
213
|
|
|
""".format( |
|
214
|
|
|
"; ".join([s.name for s in TargetType.all_types()]), |
|
|
|
|
|
|
215
|
|
|
"\n\n - ".join( |
|
|
|
|
|
|
216
|
|
|
[f"{k} ({v})" for k, v in TargetType.special_type_names().items()] |
|
217
|
|
|
), |
|
218
|
|
|
) |
|
219
|
|
|
), |
|
220
|
|
|
) |
|
221
|
|
|
|
|
222
|
|
|
min_confidence = typer.Option( |
|
223
|
|
|
3, |
|
224
|
|
|
"--confidence", |
|
225
|
|
|
min=0, |
|
226
|
|
|
max=9, |
|
227
|
|
|
show_default=False, |
|
228
|
|
|
help=doc( |
|
229
|
|
|
""" |
|
230
|
|
|
Minimum target confidence score, inclusive. |
|
231
|
|
|
This is useful to modify in only some cases. More important options are min_pchembl and taxa. |
|
|
|
|
|
|
232
|
|
|
|
|
233
|
|
|
Values are: {} |
|
234
|
|
|
|
|
235
|
|
|
[default: 3] |
|
236
|
|
|
""".format( |
|
237
|
|
|
"; ".join([f"{s.value} ({s.name})" for s in ConfidenceLevel]) |
|
|
|
|
|
|
238
|
|
|
) |
|
239
|
|
|
), |
|
240
|
|
|
) |
|
241
|
|
|
|
|
242
|
|
|
relations = typer.Option( |
|
243
|
|
|
"<,<=,=", |
|
244
|
|
|
"--relations", |
|
245
|
|
|
show_default=False, |
|
246
|
|
|
help=doc( |
|
247
|
|
|
""" |
|
248
|
|
|
Assay activity relations allowed, comma-separated. |
|
249
|
|
|
If post-processing yourself, consider including all. |
|
250
|
|
|
Values are: <, <=, =, >, >=, ~. |
|
251
|
|
|
[default: <,<=,=] |
|
252
|
|
|
""" |
|
253
|
|
|
), |
|
254
|
|
|
) |
|
255
|
|
|
|
|
256
|
|
|
min_pchembl = typer.Option( |
|
257
|
|
|
6.0, |
|
258
|
|
|
"--pchembl", |
|
259
|
|
|
min=0.0, |
|
260
|
|
|
show_default=False, |
|
261
|
|
|
help=doc( |
|
262
|
|
|
""" |
|
263
|
|
|
Minimum pCHEMBL value, inclusive. |
|
264
|
|
|
If post-processing yourself, consider setting to 0.0. |
|
265
|
|
|
[default: 6.0] |
|
266
|
|
|
""" |
|
267
|
|
|
), |
|
268
|
|
|
) |
|
269
|
|
|
|
|
270
|
|
|
banned_flags = typer.Option( |
|
271
|
|
|
"@negative", |
|
272
|
|
|
show_default=False, |
|
273
|
|
|
help=doc( |
|
274
|
|
|
""" |
|
275
|
|
|
Exclude activity annotations with data validity flags, comma-separated. |
|
276
|
|
|
It is rare to need to change this. |
|
277
|
|
|
|
|
278
|
|
|
Values are: {}. |
|
279
|
|
|
|
|
280
|
|
|
Special sets are: |
|
281
|
|
|
|
|
282
|
|
|
- @all (all flags are banned) |
|
283
|
|
|
|
|
284
|
|
|
- @negative ({}) |
|
285
|
|
|
|
|
286
|
|
|
- @positive ({}) |
|
287
|
|
|
|
|
288
|
|
|
[default: @negative] |
|
289
|
|
|
""".format( |
|
290
|
|
|
"; ".join([s.name for s in DataValidityComment]), |
|
|
|
|
|
|
291
|
|
|
", ".join([s.name for s in DataValidityComment.negative_comments()]), |
|
|
|
|
|
|
292
|
|
|
", ".join([s.name for s in DataValidityComment.positive_comments()]), |
|
|
|
|
|
|
293
|
|
|
), |
|
294
|
|
|
), |
|
295
|
|
|
) |
|
296
|
|
|
|
|
297
|
|
|
binding_search_name = typer.Option( |
|
298
|
|
|
None, |
|
299
|
|
|
help=""" |
|
300
|
|
|
The fully qualified name of a class inheriting ``BindingSearch``. |
|
301
|
|
|
If specified, all parameters above are passed to its constructor. |
|
302
|
|
|
""", |
|
303
|
|
|
) |
|
304
|
|
|
|
|
305
|
|
|
chembl_trial = typer.Option( |
|
306
|
|
|
0, |
|
307
|
|
|
"--phase", |
|
308
|
|
|
show_default=False, |
|
309
|
|
|
help=doc( |
|
310
|
|
|
""" |
|
311
|
|
|
Minimum phase of a clinical trial, inclusive. |
|
312
|
|
|
Values are: 0, 1, 2, 3. |
|
313
|
|
|
[default: 0] |
|
314
|
|
|
""" |
|
315
|
|
|
), |
|
316
|
|
|
min=0, |
|
317
|
|
|
max=3, |
|
318
|
|
|
) |
|
319
|
|
|
|
|
320
|
|
|
KNOWN_USEFUL_KEYS: Mapping[str, str] = { |
|
321
|
|
|
"weight": "Molecular Weight", |
|
322
|
|
|
"xlogp3": None, |
|
323
|
|
|
"hydrogen-bond-donors": "Hydrogen Bond Donor Count", |
|
324
|
|
|
"hydrogen-bond-acceptors": "Hydrogen Bond Acceptor Count", |
|
325
|
|
|
"rotatable-bonds": "Rotatable Bond Count", |
|
326
|
|
|
"exact-mass": None, |
|
327
|
|
|
"monoisotopic-mass": None, |
|
328
|
|
|
"tpsa": "Topological Polar Surface Area", |
|
329
|
|
|
"heavy-atoms": "Heavy Atom Count", |
|
330
|
|
|
"charge": "Formal Charge", |
|
331
|
|
|
"complexity": None, |
|
332
|
|
|
} |
|
333
|
|
|
KNOWN_USELESS_KEYS: Mapping[str, str] = { |
|
334
|
|
|
"components": "Covalently-Bonded Unit Count", |
|
335
|
|
|
"isotope-atoms": "Isotope Atom Count", |
|
336
|
|
|
"defined-atom-stereocenter-count": None, |
|
337
|
|
|
"undefined-atom-stereocenter-count": None, |
|
338
|
|
|
"defined-bond-stereocenter-count": None, |
|
339
|
|
|
"undefined-bond-stereocenter-count": None, |
|
340
|
|
|
"compound-is-canonicalized": None, |
|
341
|
|
|
} |
|
342
|
|
|
|
|
343
|
|
|
pubchem_computed_keys = typer.Option( |
|
344
|
|
|
"weight,xlogp3,tpsa,complexity,exact-mass,heavy-atom-count,charge", |
|
345
|
|
|
help=""" |
|
346
|
|
|
The keys of the computed properties, comma-separated. |
|
347
|
|
|
Key names are case-insensitive and ignore punctuation like underscores and hyphens. |
|
348
|
|
|
|
|
349
|
|
|
Known keys are: {} |
|
350
|
|
|
|
|
351
|
|
|
Known, less-useful (metadata-like) keys are: {} |
|
352
|
|
|
""".format( |
|
353
|
|
|
_stringify(KNOWN_USEFUL_KEYS), _stringify(KNOWN_USELESS_KEYS) |
|
354
|
|
|
), |
|
355
|
|
|
) |
|
356
|
|
|
|
|
357
|
|
|
|
|
358
|
|
|
__all__ = ["EntryArgs"] |
|
359
|
|
|
|
dmyersturnbull /
mandos
Loading history...