|
1
|
|
|
#! /usr/bin/env python |
|
2
|
|
|
# |
|
3
|
|
|
# Copyright (C) 2016 Rich Lewis <[email protected]> |
|
4
|
|
|
# License: 3-clause BSD |
|
5
|
|
|
|
|
6
|
|
|
""" |
|
7
|
|
|
## skchem.standardizers.chemaxon |
|
8
|
|
|
|
|
9
|
|
|
Module wrapping ChemAxon Standardizer. Must have standardizer installed and |
|
10
|
|
|
license activated. |
|
11
|
|
|
""" |
|
12
|
|
|
|
|
13
|
|
|
import os |
|
14
|
|
|
from tempfile import NamedTemporaryFile |
|
15
|
|
|
import subprocess |
|
16
|
|
|
import logging |
|
17
|
|
|
|
|
18
|
|
|
logger = logging.getLogger(__file__) |
|
|
|
|
|
|
19
|
|
|
|
|
20
|
|
|
import pandas as pd |
|
|
|
|
|
|
21
|
|
|
|
|
22
|
|
|
from .. import core |
|
23
|
|
|
from .. import io |
|
24
|
|
|
|
|
25
|
|
|
# ideally we will programatically build this file, but for now just use it. |
|
26
|
|
|
DEFAULT_CONFIG = os.path.join(os.path.dirname(__file__), 'default_config.xml') |
|
27
|
|
|
|
|
28
|
|
|
class ChemAxonStandardizer(object): |
|
29
|
|
|
|
|
30
|
|
|
""" Object wrapping the ChemAxon Standardizer, for standardizing molecules. |
|
31
|
|
|
|
|
32
|
|
|
Args: |
|
33
|
|
|
config_path (str): |
|
34
|
|
|
The path of the config_file. If None, use the default one. |
|
35
|
|
|
""" |
|
36
|
|
|
def __init__(self, config_path=None, keep_failed=True): |
|
37
|
|
|
|
|
38
|
|
|
if not config_path: |
|
39
|
|
|
config_path = DEFAULT_CONFIG |
|
40
|
|
|
self.config_path = config_path |
|
41
|
|
|
self.keep_failed = keep_failed |
|
42
|
|
|
|
|
43
|
|
|
def transform(self, obj): |
|
|
|
|
|
|
44
|
|
|
if isinstance(obj, core.Mol): |
|
45
|
|
|
return self._transform_mol(obj) |
|
46
|
|
|
elif isinstance(obj, pd.Series): |
|
47
|
|
|
return self._transform_ser(obj) |
|
48
|
|
|
elif isinstance(obj, pd.DataFrame): |
|
49
|
|
|
obj = obj.copy() |
|
50
|
|
|
res = self._transform_ser(obj.structure) |
|
51
|
|
|
obj.iloc['structure'] = res |
|
52
|
|
|
return obj |
|
53
|
|
|
|
|
54
|
|
|
else: |
|
55
|
|
|
raise NotImplementedError |
|
56
|
|
|
|
|
57
|
|
|
def _transform_mol(self, mol): |
|
58
|
|
|
mol = pd.DataFrame([mol], index=[mol.name], columns=['structure']) |
|
59
|
|
|
return self.transform(mol).structure.iloc[0] |
|
60
|
|
|
|
|
61
|
|
|
|
|
62
|
|
|
def _transform_by(self, X, by='sdf'): |
|
|
|
|
|
|
63
|
|
|
|
|
64
|
|
|
with NamedTemporaryFile() as f_in, NamedTemporaryFile() as f_out: |
|
65
|
|
|
getattr(io, 'write_' + by)(X, f_in.name) |
|
66
|
|
|
args = ['standardize', f_in.name, |
|
67
|
|
|
'-c', self.config_path, |
|
|
|
|
|
|
68
|
|
|
'-f', 'sdf', |
|
|
|
|
|
|
69
|
|
|
'-o', f_out.name, |
|
|
|
|
|
|
70
|
|
|
'--ignore-error'] |
|
|
|
|
|
|
71
|
|
|
logger.debug('Running %s', ' '.join(args)) |
|
72
|
|
|
subprocess.call(args) |
|
73
|
|
|
out = getattr(io, 'read_' + by)(f_out.name).structure |
|
74
|
|
|
return out |
|
75
|
|
|
|
|
76
|
|
|
|
|
77
|
|
|
def _transform_ser(self, X, y=None): |
|
|
|
|
|
|
78
|
|
|
# TODO: should check for errors, and schedule multiple runs with |
|
|
|
|
|
|
79
|
|
|
# different serialization methods here. |
|
80
|
|
|
|
|
81
|
|
|
out = self._transform_by(X) |
|
82
|
|
|
return out |
|
83
|
|
|
|
|
84
|
|
|
# failed = X.ix[failed] |
|
85
|
|
|
# now_succ = self._transform_by(failed, 'smiles') |
|
86
|
|
|
# still_failed = set(now_succ.index).difference(set(failed.index)) |
|
87
|
|
|
# |
|
88
|
|
|
# for fail in still_failed: |
|
89
|
|
|
# logger.warn('%s failed to standardize.', fail) |
|
90
|
|
|
# |
|
91
|
|
|
# out = out.append(now_succ) |
|
92
|
|
|
# if self.keep_failed: |
|
93
|
|
|
# out = out.append(X.ix[still_failed]) |
|
94
|
|
|
# return out.structure |
|
95
|
|
|
|
richlewis42 /
scikit-chem
Loading history...
This check looks for invalid names for a range of different identifiers.
You can set regular expressions to which the identifiers must conform if the defaults do not match your requirements.
If your project includes a Pylint configuration file, the settings contained in that file take precedence.
To find out more about Pylint, please refer to their site.