def _renderSortedTsv(self, templateFilename, vcfFilename, tsvFilename, sampleNames, dataManager, inferGenotypes):
"""
Turn a sorted tsv into a VCF
:param templateFilename: basic VCF to model output VCF.
:param vcfFilename: output VCF filename
:param tsvFilename: input sorted tsv
:param sampleNames: sample names that should be used in output
:param dataManager: dataManager instance used in creating pyvcf records.
:param inferGenotypes: whether we should try to infer the genotypes, since we may not have add GT explicitly
on input
"""
tempVcfReader = vcf.Reader(filename=templateFilename, strict_whitespace=True)
pointer = file(vcfFilename, "w")
tsvReader = GenericTsvReader(tsvFilename, delimiter=self.delimiter)
index = 0
nrecords = 1000
chrom = None
pos = None
refAllele = None
recordBuilder = None
vcfWriter = vcf.Writer(pointer, tempVcfReader, self.lineterminator)
for ctr, m in enumerate(tsvReader):
isNewRecord = self._isNewVcfRecordNeeded(chrom, m["chr"], pos, m["start"], refAllele, m["ref_allele"])
if isNewRecord:
if recordBuilder is not None:
record = recordBuilder.createRecord()
vcfWriter.write_record(record)
index += 1
if index % nrecords == 0:
self.logger.info("Rendered " + str(index) + " vcf records.")
vcfWriter.flush()
chrom = m["chr"]
pos = m["start"]
refAllele = m["ref_allele"]
recordBuilder = RecordBuilder(chrom, int(pos), refAllele, sampleNames)
recordBuilder = self._parseRecordBuilder(m, recordBuilder, dataManager, inferGenotypes)
if recordBuilder is not None:
record = recordBuilder.createRecord()
vcfWriter.write_record(record)
vcfWriter.close()
tsvReader.close()
self.logger.info("Rendered all " + str(index) + " vcf records.")
def index(destDir, inputFilename, fileColumnNumList=None, preset=None):
"""
Create a tabix index file for genomic position datasource tsv files.
Prerequisites (for genomic position indexed):
Input file has three columns that can be mapped to chromosome, start position, and end position without any modification.
For example, ['hg19.oreganno.chrom', 'hg19.oreganno.chromStart', 'hg19.oreganno.chromEnd'] in oreganno.hg19.txt
This will overwrite an existing index (since the force parameter is set to True in pysam.tabix_index() call).
Also, in cases where the inputFilename doesn't end with a ".gz", the a compressed file will be created and indexed.
If the gz and tbi files already exist, this will simply copy the files to the specified destination.
:param destDir: destination directory
:param fileColumnNumList: ordered list. This list contains the corresponding entries (column numbers)
in the tsv file. Typically, this would be [chr,start,end] or [gene, startAA, endAA]
:param inputFilename: tsv file input
:param preset: if preset is provided, the column coordinates are taken from a preset. Valid values for preset
are "gff", "bed", "sam", "vcf", "psltbl", and "pileup". "tsv" is also recognized, but this will use the tabix
generic indexing (after commenting out the header line)
"""
fileColumnNumList = [] if fileColumnNumList is None else fileColumnNumList
inputFilename = os.path.abspath(inputFilename)
fileDir = os.path.dirname(inputFilename)
fileName, fileExtension = os.path.splitext(os.path.basename(inputFilename))
if fileExtension in (".gz",):
# Ensure .gz.tbi file is there as well
inputIndexFilename = os.path.join(fileDir, string.join([inputFilename, "tbi"], "."))
if not os.path.exists(inputIndexFilename):
msg = "Missing tabix index file %s." % inputIndexFilename
raise TabixIndexerFileMissingError(msg)
outputFilename = os.path.join(destDir, string.join([fileName, "gz"], "."))
shutil.copyfile(inputFilename, outputFilename)
outputIndexFilename = os.path.join(destDir, string.join([fileName, "gz", "tbi"], "."))
shutil.copyfile(inputIndexFilename, outputIndexFilename)
return outputFilename
outputFilename = os.path.join(destDir, string.join([fileName, ".tabix_indexed", fileExtension], ""))
# Load the file into a tsvReader.
if preset in ("gff", "bed", "sam", "vcf", "psltbl", "pileup"):
# Copy the input file to output file.
shutil.copyfile(inputFilename, outputFilename)
tabix_index = pysam.tabix_index(filename=outputFilename, force=True, preset=preset)
else:
# Need to comment out the header line with a "#", so we cannot simply copy the file.
input_reader = GenericTsvReader(inputFilename)
with file(outputFilename, 'w') as output_writer:
output_writer.writelines(input_reader.getCommentsAsList())
# Add "#" for the header line.
output_writer.write("#")
field_names = input_reader.getFieldNames()
output_writer.write("\t".join(field_names))
output_writer.write("\n")
output_writer.flush()
# Write the rest of the file
# This might be too slow, since a raw reader would be pretty fast.
for line_dict in input_reader:
line_list = [line_dict[k] for k in field_names]
line_rendered = "\t".join(line_list) + "\n"
output_writer.write(line_rendered)
input_reader.close()
tabix_index = pysam.tabix_index(filename=outputFilename, force=True, seq_col=fileColumnNumList[0],
start_col=fileColumnNumList[1], end_col=fileColumnNumList[2])
if tabix_index is None:
raise OncotatorException("Could not create a tabix index from this input file: " + outputFilename)
return tabix_index
def index(destDir, inputFilename, fileColumnNumList=None, preset=None):
"""
Create a tabix index file for genomic position datasource tsv files.
Prerequisites (for genomic position indexed):
Input file has three columns that can be mapped to chromosome, start position, and end position without any modification.
For example, ['hg19.oreganno.chrom', 'hg19.oreganno.chromStart', 'hg19.oreganno.chromEnd'] in oreganno.hg19.txt
This will overwrite an existing index (since the force parameter is set to True in pysam.tabix_index() call).
Also, in cases where the inputFilename doesn't end with a ".gz", the a compressed file will be created and indexed.
If the gz and tbi files already exist, this will simply copy the files to the specified destination.
:param destDir: destination directory
:param fileColumnNumList: ordered list. This list contains the corresponding entries (column numbers)
in the tsv file. Typically, this would be [chr,start,end] or [gene, startAA, endAA]
:param inputFilename: tsv file input
:param preset: if preset is provided, the column coordinates are taken from a preset. Valid values for preset
are "gff", "bed", "sam", "vcf", "psltbl", and "pileup". "tsv" is also recognized, but this will use the tabix
generic indexing (after commenting out the header line)
"""
fileColumnNumList = [] if fileColumnNumList is None else fileColumnNumList
inputFilename = os.path.abspath(inputFilename)
fileDir = os.path.dirname(inputFilename)
fileName, fileExtension = os.path.splitext(
os.path.basename(inputFilename))
if fileExtension in (".gz", ):
# Ensure .gz.tbi file is there as well
inputIndexFilename = os.path.join(
fileDir, string.join([inputFilename, "tbi"], "."))
if not os.path.exists(inputIndexFilename):
msg = "Missing tabix index file %s." % inputIndexFilename
raise TabixIndexerFileMissingError(msg)
outputFilename = os.path.join(destDir,
string.join([fileName, "gz"], "."))
shutil.copyfile(inputFilename, outputFilename)
outputIndexFilename = os.path.join(
destDir, string.join([fileName, "gz", "tbi"], "."))
shutil.copyfile(inputIndexFilename, outputIndexFilename)
return outputFilename
outputFilename = os.path.join(
destDir,
string.join([fileName, ".tabix_indexed", fileExtension], ""))
# Load the file into a tsvReader.
if preset in ("gff", "bed", "sam", "vcf", "psltbl", "pileup"):
# Copy the input file to output file.
shutil.copyfile(inputFilename, outputFilename)
tabix_index = pysam.tabix_index(filename=outputFilename,
force=True,
preset=preset)
else:
# Need to comment out the header line with a "#", so we cannot simply copy the file.
input_reader = GenericTsvReader(inputFilename)
with file(outputFilename, 'w') as output_writer:
output_writer.writelines(input_reader.getCommentsAsList())
# Add "#" for the header line.
output_writer.write("#")
field_names = input_reader.getFieldNames()
output_writer.write("\t".join(field_names))
output_writer.write("\n")
output_writer.flush()
# Write the rest of the file
# This might be too slow, since a raw reader would be pretty fast.
for line_dict in input_reader:
line_list = [line_dict[k] for k in field_names]
line_rendered = "\t".join(line_list) + "\n"
output_writer.write(line_rendered)
input_reader.close()
tabix_index = pysam.tabix_index(filename=outputFilename,
force=True,
seq_col=fileColumnNumList[0],
start_col=fileColumnNumList[1],
end_col=fileColumnNumList[2])
if tabix_index is None:
raise OncotatorException(
"Could not create a tabix index from this input file: " +
outputFilename)
return tabix_index
