def addtags(bam,
tagfile,
output,
sam=False,
trim_suffix=True,
mode="tag",
nproc=1):
"""Add tags to reads from individual cells
Copies BAM entries to a new file, adding a read tag to cells matching an input table
Parameters
----------
bam : str
Path to BAM file.
tagfile : str
Tab-delimited file containing cell barcode, read tag to be added, tag information
output : str
Name for output BAM file.
sam : bool, optional
Output SAM format. Default is BAM format.
trim_suffix: bool, optional
Remove trailing 2 characters from cell barcode in bam file (sometimes needed to match 10x barcodes).
nproc : int, optional
Number of processors to use. Default is 1.
mode : str
Either tag (default) or readname. Some BAM file store the cell barcode in the readname rather than under
a read tag.
Raises
------
Exception
If samtools merge of temporary BAM files fails
"""
nproc = int(nproc)
tags = _readtags(tagfile)
inputBam = pysam.AlignmentFile(bam, "rb")
intervals = utils.chunk_bam(inputBam, nproc)
inputBam.close()
p = Pool(nproc)
tempfiles = p.map_async(
functools.partial(
_add_read_tags,
bam=bam,
sam=sam,
output=output,
cb=tags,
trim_suffix=trim_suffix,
mode=mode,
),
intervals.values(),
).get(9999999)
mergestring = ("samtools merge -@ " + str(nproc) + " " + output + " " +
" ".join(tempfiles))
call(mergestring, shell=True)
if os.path.exists(output):
[os.remove(i) for i in tempfiles]
else:
raise Exception("samtools merge failed, temp files not deleted")
def filterbarcodes(cells,
bam,
output,
sam=False,
trim_suffix=True,
nproc=1,
mode="tag"):
"""Filter reads based on input list of cell barcodes
Copy BAM entries matching a list of cell barcodes to a new BAM file.
Parameters
----------
cells : str
Path to file containing cell barcodes, or comma-separated list of cell barcodes. File can be gzip compressed.
bam : str
Path to BAM file.
output : str
Name for output BAM file.
sam : bool, optional
Output SAM format. Default is BAM format.
trim_suffix: bool, optional
Remove trailing 2 characters from cell barcode in bam file (sometimes needed to match 10x barcodes).
nproc : int, optional
Number of processors to use. Default is 1.
mode : str
Either tag (default) or readname. Some BAM file store the cell barcode in the readname rather than under
a read tag.
Raises
------
Exception
If samtools merge of temporary BAM files fails
"""
nproc = int(nproc)
cb = utils.read_cells(cells)
inputBam = pysam.AlignmentFile(bam, "rb")
intervals = utils.chunk_bam(inputBam, nproc)
inputBam.close()
p = Pool(nproc)
tempfiles = p.map_async(
functools.partial(
_iterate_reads,
bam=bam,
sam=sam,
output=output,
cb=cb,
trim_suffix=trim_suffix,
mode=mode,
),
intervals.values(),
).get(9999999)
mergestring = ("samtools merge -@ " + str(nproc) + " " + output + " " +
" ".join(tempfiles))
call(mergestring, shell=True)
if os.path.exists(output):
[os.remove(i) for i in tempfiles]
else:
raise Exception("samtools merge failed, temp files not deleted")
def filterbarcodes(
cells, bam, readname_barcode, cellbarcode, sam=False, trim_suffix=True, nproc=1
):
"""Filter reads based on input list of cell barcodes
Copy BAM entries matching a list of cell barcodes to a new BAM file.
Output BAM files will be named according to the group name in the
file provided.
Parameters
----------
cells : str
Path to file containing cell barcodes and the group associated with each barcode.
File can be gzip compressed. A separate BAM file will be created for each
group of cells.
bam : str
Path to BAM file.
trim_suffix: bool, optional
Remove trailing 2 characters from cell barcode in bam file (sometimes needed to match 10x barcodes).
nproc : int, optional
Number of processors to use. Default is 1.
cellbarcode : str
Tag used for cell barcode. Default is CB (used by cellranger)
readname_barcode : regex
A regular expression for matching cell barcode in read name. If None (default),
use the read tags.
Raises
------
Exception
If samtools merge of temporary BAM files fails
"""
nproc = int(nproc)
cb = utils.read_cell_barcode_file(cells)
unique_classes = list(set(chain.from_iterable(cb.values())))
inputBam = pysam.AlignmentFile(bam, "rb")
intervals = utils.chunk_bam(inputBam, nproc)
inputBam.close()
if readname_barcode is not None:
readname_barcode = re.compile(readname_barcode)
p = Pool(nproc)
idents = p.map_async(
functools.partial(
_iterate_reads,
bam=bam,
cb=cb,
classes=unique_classes,
trim_suffix=trim_suffix,
cellbarcode=cellbarcode,
readname_barcode=readname_barcode
),
intervals.values(),
).get(9999999)
mergeAll(idents=idents, classes=unique_classes, nproc=nproc, remove=True)