def _pipe_from_textfile(self, finp):
while True:
s = finp.readline()
if s == '':
return
s = s.rstrip('\n').lstrip()
if '##' in s:
s = s[:s.index('##')]
if '=' in s:
s = s.split('=', 1)
if s[0].lower().startswith('formula'):
self.filter.append(
encode_formula_to_array(parse_formula(s[1])))
elif s.lower().startswith('dict_form'):
self.vector_form = False
elif s.lower().startswith('end'):
if not self.vector_form:
self.required_fields = ['Formula']
for i in range(len(self.filter)):
self.filter[i] = decode_formula_from_array(
self.filter[i])
self.rejected = self.__rejected_formula
return
def _pipe_from_textfile(self, finp):
while True:
s = finp.readline()
if s == '':
return
s = s.rstrip('\n').lstrip()
if '##' in s:
s = s[:s.index('##')]
if '=' in s:
s = s.split('=', 1)
if s[0].lower().startswith('formula'):
s = s[1].split(',')
self.formulas.append(
encode_formula_to_array(parse_formula(s[0])))
self.scores.append(float(s[1]))
elif s[0].lower().startswith('unknown_score'):
self.unknown_score = float(s[1])
elif s.lower().startswith('dict_form'):
self.vector_form = False
elif s.lower().startswith('end'):
if not self.vector_form:
self.required_fields = ['Formula']
for i in range(len(self.formulas)):
self.formulas[i] = decode_formula_from_array(
self.formulas[i])
self.process_molecular_candidate_record = self.__process_molecular_candidate_record_formula
return
def __repr__(self):
flas = []
for filtervalue in self.filter:
if self.vector_form:
flas.append(
'"%s"' %
decode_formula_from_array(filtervalue).formula_to_string())
else:
flas.append('"%s"' % filtervalue.formula_to_string())
s = ', '.join(flas)
return "FormulasFilter(%s)" % s
def __repr__(self):
l = []
for i in range(len(self.formulas)):
if self.vector_form:
print(self.scores[i])
l.append('"%s"=%s' % (decode_formula_from_array(
self.formulas[i]).formula_to_string(), self.scores[i]))
else:
l.append(
'"%s"=%s' %
(self.formulas[i].formula_to_string(), self.scores[i]))
s = ', '.join(l)
return 'FormulaScorer(%s)' % s
def _pipe_to_textfile(self, fout, indent=''):
#subindent='%s\t'%indent;
if self.vector_form:
fout.write('%s%s\n' % (indent, 'vector_form'))
else:
fout.write('%s%s\n' % (indent, 'dict_form'))
for filtervalue in self.filter:
if self.vector_form:
fout.write(
'%s%s=%s\n' %
(indent, 'formula', decode_formula_from_array(
filtervalue).formula_to_string()))
else:
fout.write(
'%s%s=%s\n' %
(indent, 'formula', filtervalue.formula_to_string()))
def __init__(self, formulas=None, use_vector_form=True):
self.filter = []
self.vector_form = use_vector_form
#self.supported_adducts=set();
if self.vector_form:
self.required_fields = ['FormulaVector']
else:
self.required_fields = ['Formula']
if formulas is None:
return
if isinstance(formulas, str):
formulas = formulas.split(',')
for formula in formulas:
self.filter.append(
encode_formula_to_array(parse_formula(formula)))
elif isinstance(formulas, list):
for formula in formulas:
if isinstance(formula, dict):
self.filter.append(encode_formula_to_array(formula))
elif isinstance(formula, np.ndarray):
self.filter.append(formula)
elif isinstance(formula, str):
fs = formula.split(',')
for f in fs:
self.filter.append(
encode_formula_to_array(parse_formula(f)))
else:
raise TypeError(
'Wrong type argument for FormulasFilter initialization! str, dict, list of (dict or str) supported only!'
)
elif isinstance(formulas, dict):
self.filter.append(encode_formula_to_array(formulas))
elif isinstance(formulas, np.ndarray):
self.filter.append(formulas)
else:
raise TypeError(
'Wrong type argument for FormulasFilter initialization! str, dict, list of (dict or str) supported only!'
)
if not self.vector_form:
for i in range(len(self.filter)):
self.filter[i] = decode_formula_from_array(self.filter[i])
def _pipe_to_textfile(self, fout, indent=''):
if self.vector_form:
fout.write('%svector_form\n' % (indent))
else:
fout.write('%sdict_form\n' % (indent))
for i in range(len(self.formulas)):
if self.vector_form:
fout.write('%s%s=%s,%s\n' %
(indent, 'formula',
decode_formula_from_array(
self.formulas[i]).formula_to_string(),
self.scores[i]))
else:
fout.write(
'%s%s=%s,%s\n' %
(indent, 'formula', self.formulas[i].formula_to_string()),
self.scores[i])
fout.write('%s%s=%s\n' % (indent, 'unknown_score', self.unknown_score))
def _get_next_hdf5_record(self):
self.mzcurrent+=1;
if self.mzcurrent>self.mzmax_index:
raise StopIteration();
record=MolecularRecord();
if self.charged:
record['MZ']=self.mz_mass_charge_dataset[self.mzcurrent,0];
record['Mass']=self.mz_mass_charge_dataset[self.mzcurrent,1];
record['Charge']=int(self.mz_mass_charge_dataset[self.mzcurrent,2]);
else:
record['MZ']=self.mz_dataset[self.mzcurrent];
record['Mass']=record['MZ'];
record['Charge']=0;
if 'FPT' in self.required_fields:
record['FPT']=self.fpt_dataset[self.mzcurrent,:];
if 'Frag' in self.required_fields:
i0=self.fragprint_index_dataset[self.mzcurrent,0];
i1=self.fragprint_index_dataset[self.mzcurrent,1];
record['Frag']=list(self.fragprint_values_dataset[i0:i1]);
if 'CFM' in self.required_fields:
i0=self.CFM_negative_index_dataset[self.mzcurrent,0,0];
j0=self.CFM_negative_index_dataset[self.mzcurrent,0,1];
i1=self.CFM_negative_index_dataset[self.mzcurrent,1,0];
j1=self.CFM_negative_index_dataset[self.mzcurrent,1,1];
i2=self.CFM_negative_index_dataset[self.mzcurrent,2,0];
j2=self.CFM_negative_index_dataset[self.mzcurrent,2,1];
if i0>=0 and i1>=0 and i2>=0:
n=[self.CFM_values_dataset[i0:j0,:], self.CFM_values_dataset[i1:j1,:], self.CFM_values_dataset[i2:j2,:]]
else:
n=[];
i0=self.CFM_positive_index_dataset[self.mzcurrent,0,0];
j0=self.CFM_positive_index_dataset[self.mzcurrent,0,1];
i1=self.CFM_positive_index_dataset[self.mzcurrent,1,0];
j1=self.CFM_positive_index_dataset[self.mzcurrent,1,1];
i2=self.CFM_positive_index_dataset[self.mzcurrent,2,0];
j2=self.CFM_positive_index_dataset[self.mzcurrent,2,1];
if i0>=0 and i1>=0 and i2>=0:
p=[self.CFM_values_dataset[i0:j0,:], self.CFM_values_dataset[i1:j1,:], self.CFM_values_dataset[i2:j2,:]]
else:
p=[];
record['CFM']=[n, p];
if 'CFM2' in self.required_fields:
i0=self.CFM2_negative_index_dataset[self.mzcurrent,0,0];
j0=self.CFM2_negative_index_dataset[self.mzcurrent,0,1];
i1=self.CFM2_negative_index_dataset[self.mzcurrent,1,0];
j1=self.CFM2_negative_index_dataset[self.mzcurrent,1,1];
i2=self.CFM2_negative_index_dataset[self.mzcurrent,2,0];
j2=self.CFM2_negative_index_dataset[self.mzcurrent,2,1];
if i0>=0 and i1>=0 and i2>=0:
n=[self.CFM2_values_dataset[i0:j0,:], self.CFM2_values_dataset[i1:j1,:], self.CFM2_values_dataset[i2:j2,:]]
else:
n=[];
i0=self.CFM2_positive_index_dataset[self.mzcurrent,0,0];
j0=self.CFM2_positive_index_dataset[self.mzcurrent,0,1];
i1=self.CFM2_positive_index_dataset[self.mzcurrent,1,0];
j1=self.CFM2_positive_index_dataset[self.mzcurrent,1,1];
i2=self.CFM2_positive_index_dataset[self.mzcurrent,2,0];
j2=self.CFM2_positive_index_dataset[self.mzcurrent,2,1];
if i0>=0 and i1>=0 and i2>=0:
p=[self.CFM2_values_dataset[i0:j0,:], self.CFM2_values_dataset[i1:j1,:], self.CFM2_values_dataset[i2:j2,:]]
else:
p=[];
record['CFM2']=[n, p];
if 'SMILES' in self.required_fields:
i0=self.smiles_dataset[self.mzcurrent,0];
i1=self.smiles_dataset[self.mzcurrent,1];
arr=self.ascii_dataset[i0:i1];
if python_version==2:
record['SMILES']=str(bytearray(arr));
else:
record['SMILES']=bytes(list(arr)).decode('ascii');
if 'IDs' in self.required_fields:
i0=self.ids_dataset[self.mzcurrent,0];
i1=self.ids_dataset[self.mzcurrent,1];
arr=self.ascii_dataset[i0:i1];
if python_version==2:
record['IDs']=str(bytearray(arr));
else:
record['IDs']=bytes(list(arr)).decode('ascii');
if ('InChIKeyValues' in self.required_fields) or ('InChIKey' in self.required_fields):
kv=self.inchikey_values_dataset[self.mzcurrent,:];
if 'InChIKeyValues' in self.required_fields:
record['InChIKeyValues']=kv;
if 'InChIKey' in self.required_fields:
record['InChIKey']=inchikey_from_inchikeyvalues(kv);
if ('ShortInChI' in self.required_fields) or ('InChI' in self.required_fields):# or ('Formula' in self.required_fields):
i=self.inchi_dataset[self.mzcurrent,0,0];
j=self.inchi_dataset[self.mzcurrent,0,1];
arr=self.ascii_dataset[i:j];
if python_version==2:
i0=str(bytearray(arr));
else:
i0=bytes(list(arr)).decode('ascii');
i=self.inchi_dataset[self.mzcurrent,1,0];
j=self.inchi_dataset[self.mzcurrent,1,1];
arr=self.ascii_dataset[i:j];
if python_version==2:
i1=str(bytearray(arr));
else:
i1=bytes(list(arr)).decode('ascii');
i=self.inchi_dataset[self.mzcurrent,2,0];
j=self.inchi_dataset[self.mzcurrent,2,1];
arr=self.ascii_dataset[i:j];
if python_version==2:
i2=str(bytearray(arr));
else:
i2=bytes(list(arr)).decode('ascii');
i=self.inchi_dataset[self.mzcurrent,3,0];
j=self.inchi_dataset[self.mzcurrent,3,1];
arr=self.ascii_dataset[i:j];
if python_version==2:
i3=str(bytearray(arr));
else:
i3=bytes(list(arr)).decode('ascii');
#if 'Formula' in self.required_fields: #check which is faster
# record['Formula']=parse_formula(i0);
if 'ShortInChI' in self.required_fields:
record['ShortInChI']='%s/%s'%(i0,i1);
if 'InChI' in self.required_fields:
record['InChI']='%s/%s%s%s'%(i0,i1,i2,i3);
if ('FormulaVector' in self.required_fields) or ('Formula' in self.required_fields):
fv=self.formulavector_dataset[self.mzcurrent,:];
if ('FormulaVector' in self.required_fields):
record['FormulaVector']=fv;
if ('Formula' in self.required_fields):
record['Formula']=decode_formula_from_array(fv);
if 'ElementVector' in self.required_fields:
record['ElementVector']=self.elementvector_dataset[self.mzcurrent,:];
#Likeness
if ('MetaLike' in self.required_fields):
record['MetaLike']=self.metalikeness_dataset[self.mzcurrent];
#End of likeness
return record;
def setup_scorer(self, formulas, scores, unknown_score=0.0):
self.formulas = []
self.scores = []
self.unknown_score = unknown_score
if isinstance(formulas, str):
formulas = formulas.split(',')
for formula in formulas:
self.formulas.append(
encode_formula_to_array(parse_formula(formula)))
elif isinstance(formulas, list):
for formula in formulas:
if isinstance(formula, dict):
self.formulas.append(encode_formula_to_array(formula))
elif isinstance(formula, np.ndarray):
self.formulas.append(formula)
elif isinstance(formula, str):
fs = formula.split(',')
for f in fs:
self.formulas.append(
encode_formula_to_array(parse_formula(f)))
else:
raise TypeError(
'Wrong type argument (formulas) for FormulaVectors initialization! str, formula, list of (formula, formulavector or str) supported only!'
)
elif isinstance(formulas, dict):
self.formulas.append(encode_formula_to_array(formulas))
elif isinstance(formulas, np.ndarray):
self.formulas.append(formulas)
else:
raise TypeError(
'Wrong type argument (formulas) for FormulaVectors initialization! str, formula, list of (formula, formulavector or str) supported only!'
)
if isinstance(scores, str):
scores = scores.split(',')
for score in scores:
self.scores.append(float(score))
elif isinstance(scores, float):
self.scores.append(scores)
elif isinstance(scores, int):
self.scores.append(float(scores))
elif isinstance(scores, list):
for score in scores:
if isinstance(score, float):
self.scores.append(score)
elif isinstance(score, dict):
self.scores.append(score)
elif isinstance(score, int):
self.scores.append(float(score))
elif isinstance(score, str):
score = score.split(',')
for s in score:
self.scores.append(float(s))
else:
raise TypeError(
'Wrong type argument (scores) for FormulaScorer initialization! str, float, int, list of (float, int or str) or dictionary supported only!'
)
else:
raise TypeError(
'Wrong type argument (scores) for FormulaScorer initialization! str, float, int, list of (float, int or str) supported only!'
)
if len(self.scores) != len(self.formulas):
raise TypeError(
'Number of formulas and number of scores supplied do not match!'
)
if not self.vector_form:
for i in range(len(self.formulas)):
self.formulas[i] = decode_formula_from_array(self.formulas[i])
def generate_candidate_html(mainHTML,
candidate,
outpath,
index,
subpath,
merged=False):
if 'Annotation' in candidate:
if candidate['Annotation'] == 'Correct':
mainHTML.write('<tr><td bgcolor="#ccffcc">\n')
else:
mainHTML.write('<tr><td bgcolor="#e0e0e0">\n')
else:
mainHTML.write('<tr><td>\n')
if 'SMILES' in candidate:
smiles = candidate['SMILES']
has_image = False
if smiles != '':
fname = os.path.join(outpath, 'candidate_%s.png' % index)
if generate_candidate_image(fname, smiles):
has_image = True
if has_image:
mainHTML.write(
'<table><tr><td><img src="%s/candidate_%s.png"></td><td>\n' %
(subpath, index))
mainHTML.write('<b>Total Score:</b> %s\n' % candidate['TotalScore'])
mainHTML.write('<br><b>Individual Scores:</b>\n')
for score in candidate['Scores']:
mainHTML.write('<br><b>%s:</b> %s\n' %
(score, candidate['Scores'][score]))
mainHTML.write('<br><b>Mass:</b> %s, <b>Charge:</b> %s' %
(candidate['Mass'], candidate['Charge']))
smiles = ''
inchi = ''
formula = ''
if 'Formula' in candidate:
formula = candidate['Formula'].formula_to_string()
elif 'FormulaVector' in candidate:
formula = decode_formula_from_array(
candidate['FormulaVector']).formula_to_string()
if formula != '':
mainHTML.write(', <b>Formula:</b> %s' % formula)
if merged:
adduct = ''
isotope = ''
if 'Adduct' in candidate:
adduct = candidate['Adduct']
if 'Isotope' in candidate:
isotope = candidate['Isotope']
mainHTML.write('<br><b>Adduct:</b> %s, <b>Isotope:</b> %s\n' %
(adduct, isotope))
if 'SMILES' in candidate:
mainHTML.write('<br><b>SMILES:</b> %s\n' % candidate['SMILES'])
if 'InChI' in candidate:
mainHTML.write('<br><b>InChI</b>=1S/%s\n' % candidate['InChI'])
inchi = 'InChI=1S/%s' % candidate['InChI']
inchikey = inchikey_from_inchi(inchi)
mainHTML.write('<br><b>InChIKey:</b> %s\n' % inchikey)
mainHTML.write(
'<br><button onclick="CallPubChem(this.id)" id="%s">PubChem</button><br>\n'
% (inchikey))
if has_image:
mainHTML.write('</td></tr></table>\n')
mainHTML.write('</td></tr>\n')