def _writeMuts2Tsv(self, muts, path):
"""
Given a mutation generator, this methods writes a tab separated file for all mutations in the mutation
generator. In addition, this method computes the appropriate sample name in scenarios where the mutation is
missing sample name annotation. It also computes a list of all chromosomes and sample names contained within
the generator.
:param path: temporary filename
:param muts: generator object with mutations
"""
sampleNames = set()
chroms = set()
writer = None
# create a temporary file to write tab-separated file
tempTsvFile = tempfile.NamedTemporaryFile(dir=path, delete=False)
self.logger.debug("Creating intermediate tsv file at %s" % tempTsvFile.name)
mutAttributeNames = []
sampleNameSelector = SampleNameSelector(self.mutation,
configFile=self.configTable.getConfigFilename(),
section="OTHER")
with open(tempTsvFile.name, 'w') as fptr:
ctr = 0
sampleNameAnnotationName = sampleNameSelector.getOutputAnnotationName()
sampleNameSource = sampleNameSelector.getAnnotationSource()
for mut in muts:
if len(mutAttributeNames) == 0:
mutAttributeNames = mut.getAttributeNames()
sampleName = sampleNameSelector.getSampleName(mut)
if sampleName is not None:
if mut.get(sampleNameAnnotationName, None) is None:
mut.createAnnotation(sampleNameAnnotationName, sampleName, sampleNameSource)
sampleNames.add(sampleName)
# Parse chromosome
chroms.add(mut.chr)
updated_start, updated_ref_allele, updated_alt_allele = MutUtils.retrieveMutCoordinatesForRendering(mut)
mut.ref_allele = updated_ref_allele
mut.alt_allele = updated_alt_allele
mut.start = updated_start
if ctr == 0:
fieldnames2Render = MutUtils.getAllAttributeNames(mut)
if sampleNameAnnotationName is not None:
fieldnames2Render += [sampleNameAnnotationName]
for fieldname in fieldnames2Render: # fieldnames that start "_" aren't rendered
if fieldname.startswith("_"):
fieldnames2Render.remove(fieldname)
writer = csv.DictWriter(fptr, fieldnames2Render, extrasaction='ignore', delimiter=self.delimiter,
lineterminator=self.lineterminator)
writer.writeheader()
writer.writerow(mut)
ctr += 1
if (ctr % 1000) == 0:
self.logger.info("Wrote " + str(ctr) + " mutations to tsv.")
sampleNames = list(sampleNames)
sampleNames.sort()
chroms = list(chroms)
return chroms, sampleNames, tempTsvFile.name
def _writeMuts2Tsv(self, muts, path):
"""
Given a mutation generator, this methods writes a tab separated file for all mutations in the mutation
generator. In addition, this method computes the appropriate sample name in scenarios where the mutation is
missing sample name annotation. It also computes a list of all chromosomes and sample names contained within
the generator.
:param filename: temporary filename
:param muts: generator object with mutations
"""
sampleNames = set()
chroms = set()
writer = None
# create a temporary file to write tab-separated file
tempTsvFile = tempfile.NamedTemporaryFile(dir=path, delete=False)
self.logger.info("Creating intermediate tsv file...")
sampleNameAnnotationNames = self.getAnnotationNames("SAMPLE_NAME")
tumorSampleNameAnnotationNames = self.getAnnotationNames("SAMPLE_TUMOR_NAME")
normalSampleNameAnnotationNames = self.getAnnotationNames("SAMPLE_NORMAL_NAME")
mutAttributeNames = []
with open(tempTsvFile.name, 'w') as fptr:
ctr = 0
for mut in muts:
sampleName = None
sampleNameAnnotationName = None
if len(mutAttributeNames) == 0:
mutAttributeNames = mut.getAttributeNames()
# Sample name annotation is present
if len(sampleNameAnnotationNames) != 0:
sampleNameAnnotationName = sampleNameAnnotationNames[0]
sampleName = mut[sampleNameAnnotationName]
# Both, tumor and normal sample name annotations are present
elif len(tumorSampleNameAnnotationNames) != 0 and len(normalSampleNameAnnotationNames) != 0:
tumorSampleNameAnnotationName = tumorSampleNameAnnotationNames[0]
normalSampleNameAnnotationName = normalSampleNameAnnotationNames[0]
sampleName = string.join([mut[normalSampleNameAnnotationName],
mut[tumorSampleNameAnnotationName]], sep="-")
sampleNameAnnotationName = MutUtils.SAMPLE_NAME_ANNOTATION_NAME
mut.createAnnotation(sampleNameAnnotationName, sampleName, "OUTPUT")
if ctr == 0:
self.logger.info("Sample name is the concatenation of %s and %s columns."
% (normalSampleNameAnnotationName, tumorSampleNameAnnotationName))
# Only tumor sample name is present
elif len(tumorSampleNameAnnotationNames) != 0:
tumorSampleNameAnnotationName = tumorSampleNameAnnotationNames[0]
sampleName = mut[tumorSampleNameAnnotationName]
sampleNameAnnotationName = MutUtils.SAMPLE_NAME_ANNOTATION_NAME
mut.createAnnotation(sampleNameAnnotationName, sampleName, "INPUT")
if ctr == 0:
self.logger.info("Sample name is %s column." % tumorSampleNameAnnotationName)
# Only normal sample name is present
elif len(normalSampleNameAnnotationNames) != 0:
normalSampleNameAnnotationName = normalSampleNameAnnotationNames[0]
sampleName = mut[normalSampleNameAnnotationName]
sampleNameAnnotationName = MutUtils.SAMPLE_NAME_ANNOTATION_NAME
mut.createAnnotation(sampleNameAnnotationName, sampleName, "INPUT")
if ctr == 0:
self.logger.info("Sample name is %s column." % normalSampleNameAnnotationName)
if sampleName is not None:
sampleNames.add(sampleName)
# Parse chromosome
chroms.add(mut.chr)
updated_start, updated_ref_allele, updated_alt_allele = MutUtils.retrieveMutCoordinatesForRendering(mut)
mut.ref_allele = updated_ref_allele
mut.alt_allele = updated_alt_allele
mut.start = updated_start
if ctr == 0:
fieldnames2Render = MutUtils.getAllAttributeNames(mut)
if sampleNameAnnotationName is not None:
fieldnames2Render += [sampleNameAnnotationName]
for fieldname in fieldnames2Render: # fieldnames that start "_" aren't rendered
if fieldname.startswith("_"):
fieldnames2Render.remove(fieldname)
writer = csv.DictWriter(fptr, fieldnames2Render, extrasaction='ignore', delimiter=self.delimiter,
lineterminator=self.lineterminator)
writer.writeheader()
writer.writerow(mut)
ctr += 1
if (ctr % 1000) == 0:
self.logger.info("Wrote " + str(ctr) + " mutations to tsv.")
sampleNames = list(sampleNames)
sampleNames.sort()
chroms = list(chroms)
return chroms, sampleNames, tempTsvFile.name
