def peFragmentSize(bamFile, bamFileIndex=None,
return_lengths=False,
numberOfProcessors=None, verbose=False):
bamHandle = bamHandler.openBam(bamFile, bamFileIndex)
chromSizes = zip(bamHandle.references, bamHandle.lengths)
chunkSize = int(
float(sum(bamHandle.lengths)) * 0.3 / max(numberOfProcessors,
len(bamHandle.lengths)))
imap_res = mapReduce.mapReduce((bamHandle.filename, ),
getFragmentLength_wrapper,
chromSizes,
genomeChunkLength=chunkSize,
numberOfProcessors=numberOfProcessors,
verbose=verbose)
fl = np.concatenate(imap_res)
if len(fl):
fragLength = {'sample_size': len(fl),
'min': fl.min(),
'qtile25': np.percentile(fl, 25),
'mean': np.mean(fl),
'median': np.median(fl),
'qtile75': np.percentile(fl, 75),
'max': fl.max(),
'std': np.std(fl)}
if return_lengths:
fragLength['lengths'] = fl
else:
fragLength = None
return fragLength
def getScorePerBin(bigwigFilesList, binLength,
numberOfProcessors=1, skipZeros=True,
verbose=False, region=None,
bedFile=None,
chrsToSkip=[]):
"""
This function returns a matrix containing scores (median) for the coverage
of fragments within a region. Each row corresponds to a sampled region.
Likewise, each column corresponds to a bigwig file.
Test dataset with two samples covering 200 bp.
>>> test = Tester()
>>> np.transpose(getScorePerBin([test.bwFile1, test.bwFile2], 50, 5,))
array([[ 1., 1., 2., 2.],
[ 1., 1., 1., 3.]])
"""
# Try to determine an optimal fraction of the genome (chunkSize)
# that is sent to workers for analysis. If too short, too much time
# is spend loading the files
# if too long, some processors end up free.
# the following values are empirical
# get list of common chromosome names and sizes
chromSizes = getChromSizes(bigwigFilesList)
# skip chromosome in the list. This is usually for the
# X chromosome which may have either one copy in a male sample
# or a mixture of male/female and is unreliable.
# Also the skip may contain heterochromatic regions and
# mitochondrial DNA
if len(chrsToSkip): chromSizes = [ x for x in chromSizes if x[0] not in chrsToSkip ]
chrNames, chrLengths = zip(*chromSizes)
genomeSize = sum(chrLengths)
max_mapped = max( map(lambda x: getNumberOfFragmentsPerRegionFromBigWig(x, chromSizes), bigwigFilesList) )
reads_per_bp = float(max_mapped) / genomeSize
stepSize = binLength #for consecutive bins
chunkSize = int(stepSize * 1e3 / ( reads_per_bp * len(bigwigFilesList)) )
if verbose:
print "step size is {}".format(stepSize)
if region:
# in case a region is used, append the tilesize
region += ":{}".format(binLength)
# mapReduce( (staticArgs), func, chromSize, etc. )
imap_res = mapReduce.mapReduce((bigwigFilesList, stepSize, binLength, skipZeros),
countReadsInRegions_wrapper,
chromSizes,
genomeChunkLength=chunkSize,
bedFile=bedFile,
region=region,
numberOfProcessors=numberOfProcessors)
score_per_bin = np.concatenate(imap_res, axis=0)
return score_per_bin
def writeBedGraph(
bamOrBwFileList, outputFileName, fragmentLength,
func, funcArgs, tileSize=25, region=None, numberOfProcessors=None,
format="bedgraph", extendPairedEnds=True, missingDataAsZero=False,
smoothLength=0, fixed_step=False):
r"""
Given a list of bamfiles, a function and a function arguments,
this method writes a bedgraph file (or bigwig) file
for a partition of the genome into tiles of given size
and a value for each tile that corresponds to the given function
and that is related to the coverage underlying the tile.
"""
bamHandlers = [bamHandler.openBam(indexedFile) for
indexedFile,
fileFormat in bamOrBwFileList if fileFormat == 'bam']
if len(bamHandlers):
genomeChunkLength = getGenomeChunkLength(bamHandlers, tileSize)
# check if both bam files correspond to the same species
# by comparing the chromosome names:
chromNamesAndSize, __ = getCommonChrNames(bamHandlers, verbose=False)
else:
genomeChunkLength = int(10e6)
bigwigs = [fileName for fileName,
fileFormat in bamOrBwFileList if fileFormat == 'bigwig']
cCommon = []
chromNamesAndSize = {}
for bw in bigwigs:
bwh = pyBigWig.open(bw)
for chromName, size in bwh.chroms().items():
if chromName in chromNamesAndSize:
cCommon.append(chromName)
if chromNamesAndSize[chromName] != size:
print "\nWARNING\n" \
"Chromosome {} length reported in the " \
"bigwig files differ.\n{} for {}\n" \
"{} for {}.\n\nThe smallest " \
"length will be used".format(
chromName, chromNamesAndSize[chromName],
bigwigs[0], size, bw)
chromNamesAndSize[chromName] = min(
chromNamesAndSize[chromName], size)
else:
chromNamesAndSize[chromName] = size
bwh.close()
# get the list of common chromosome names and sizes
chromNamesAndSize = [(k, v) for k, v in chromNamesAndSize.iteritems()
if k in cCommon]
if region:
# in case a region is used, append the tilesize
region += ":{}".format(tileSize)
res = mapReduce.mapReduce((tileSize, fragmentLength, bamOrBwFileList,
func, funcArgs, extendPairedEnds, smoothLength,
missingDataAsZero, fixed_step),
writeBedGraph_wrapper,
chromNamesAndSize,
genomeChunkLength=genomeChunkLength,
region=region,
numberOfProcessors=numberOfProcessors)
# concatenate intermediary bedgraph files
outFile = open(outputFileName + ".bg", 'wb')
for tempFileName in res:
if tempFileName:
# concatenate all intermediate tempfiles into one
# bedgraph file
shutil.copyfileobj(open(tempFileName, 'rb'), outFile)
os.remove(tempFileName)
bedGraphFile = outFile.name
outFile.close()
if format == 'bedgraph':
os.rename(bedGraphFile, outputFileName)
if debug:
print "output file: %s" % (outputFileName)
else:
bedGraphToBigWig(
chromNamesAndSize, bedGraphFile, outputFileName, True)
if debug:
print "output file: %s" % (outputFileName)
os.remove(bedGraphFile)
def get_read_and_fragment_length(bamFile,
bamFileIndex=None,
return_lengths=False,
numberOfProcessors=None,
verbose=False):
"""
Estimates the fragment length and read length through sampling
:param bamFile: bamfile name
:param bamFileIndex: bamfile index name
:param return_lengths: bool,
:param numberOfProcessors:
:param verbose:
:return: tuple of two dictionaries, one for the fragment length and the other
for the read length. The dictionaries summarise the mean, median etc. values
"""
bam_handle = bamHandler.openBam(bamFile, bamFileIndex)
chrom_sizes = zip(bam_handle.references, bam_handle.lengths)
chunk_size = int(
float(sum(bam_handle.lengths)) * 0.3 /
max(numberOfProcessors, len(bam_handle.lengths)))
# avoid small chunk sizes to split the computation
chunk_size = max(chunk_size, 100000)
imap_res = mapReduce.mapReduce((bam_handle.filename, ),
getFragmentLength_wrapper,
chrom_sizes,
genomeChunkLength=chunk_size,
numberOfProcessors=numberOfProcessors,
verbose=verbose)
fl = np.concatenate(imap_res)
if len(fl):
fragment_length = fl[:, 0]
read_length = fl[:, 1]
if fragment_length.mean() > 0:
fragment_len_dict = {
'sample_size': len(fragment_length),
'min': fragment_length.min(),
'qtile25': np.percentile(fragment_length, 25),
'mean': np.mean(fragment_length),
'median': np.median(fragment_length),
'qtile75': np.percentile(fragment_length, 75),
'max': fragment_length.max(),
'std': np.std(fragment_length)
}
else:
fragment_len_dict = None
if return_lengths:
fragment_len_dict['lengths'] = fragment_length
read_len_dict = {
'sample_size': len(read_length),
'min': read_length.min(),
'qtile25': np.percentile(read_length, 25),
'mean': np.mean(read_length),
'median': np.median(read_length),
'qtile75': np.percentile(read_length, 75),
'max': read_length.max(),
'std': np.std(read_length)
}
if return_lengths:
read_len_dict['lengths'] = read_length
else:
fragment_len_dict = None
read_len_dict = None
return fragment_len_dict, read_len_dict
def run(self):
# Try to determine an optimal fraction of the genome (chunkSize) that is sent to
# workers for analysis. If too short, too much time is spend loading the files
# if too long, some processors end up free.
# the following values are empirical
bamFilesHandlers = [bamHandler.openBam(x) for x in self.bamFilesList]
chromSizes, non_common = deeptools.utilities.getCommonChrNames(bamFilesHandlers, verbose=self.verbose)
# skip chromosome in the list. This is usually for the
# X chromosome which may have either one copy in a male sample
# or a mixture of male/female and is unreliable.
# Also the skip may contain heterochromatic regions and
# mitochondrial DNA
if len(self.chrsToSkip):
chromSizes = [x for x in chromSizes if x[0] not in self.chrsToSkip]
chrNames, chrLengths = zip(*chromSizes)
genomeSize = sum(chrLengths)
if self.stepSize is None:
if self.region is None:
self.stepSize = max(int(float(genomeSize) / self.numberOfSamples), 1)
else:
# compute the step size, based on the number of samples
# and the length of the region studied
(chrom, start, end) = mapReduce.getUserRegion(chromSizes, self.region)[:3]
self.stepSize = max(int(float(end - start) / self.numberOfSamples), 1)
# number of samples is better if large
if np.mean(chrLengths) < self.stepSize:
min_num_of_samples = int(genomeSize / np.mean(chrLengths))
raise ValueError("numberOfSamples has to be bigger than {} ".format(min_num_of_samples))
max_mapped = max([x.mapped for x in bamFilesHandlers])
reads_per_bp = float(max_mapped) / genomeSize
# chunkSize = int(100 / ( reads_per_bp * len(bamFilesList)) )
chunkSize = int(self.stepSize * 1e3 / (reads_per_bp * len(bamFilesHandlers)))
[bam_h.close() for bam_h in bamFilesHandlers]
if self.verbose:
print "step size is {}".format(self.stepSize)
if self.region:
# in case a region is used, append the tilesize
self.region += ":{}".format(self.binLength)
# use map reduce to call countReadsInRegions_wrapper
imap_res = mapReduce.mapReduce([],
countReadsInRegions_wrapper,
chromSizes,
self_=self,
genomeChunkLength=chunkSize,
bedFile=self.bedFile,
region=self.region,
numberOfProcessors=self.numberOfProcessors)
if self.out_file_for_raw_data:
if len(non_common):
sys.stderr.write("*Warning*\nThe resulting bed file does not contain information for "
"the chromosomes that were not common between the bigwig files\n")
# concatenate intermediary bedgraph files
for _values, tempFileName in imap_res:
if tempFileName:
# concatenate all intermediate tempfiles into one
shutil.copyfileobj(open(tempFileName, 'r'), self.out_file_for_raw_data)
os.remove(tempFileName)
# self.out_file_for_raw_data.close()
try:
num_reads_per_bin = np.concatenate([x[0] for x in imap_res], axis=0)
return num_reads_per_bin
except ValueError:
if self.bedFile:
sys.exit('\nNo coverage values could be computed.\n\n'
'Please check that the chromosome names in the BED file are found on the bam files.\n\n'
'The valid chromosome names are:\n{}'.format(chrNames))
else:
sys.exit('\nNo coverage values could be computed.\n\nCheck that all bam files are valid and '
'contain mapped reads.')
def run(self):
# Try to determine an optimal fraction of the genome (chunkSize) that is sent to
# workers for analysis. If too short, too much time is spend loading the files
# if too long, some processors end up free.
# the following values are empirical
bamFilesHandlers = [bamHandler.openBam(x) for x in self.bamFilesList]
chromSizes, non_common = deeptools.utilities.getCommonChrNames(bamFilesHandlers, verbose=self.verbose)
# skip chromosome in the list. This is usually for the
# X chromosome which may have either one copy in a male sample
# or a mixture of male/female and is unreliable.
# Also the skip may contain heterochromatic regions and
# mitochondrial DNA
if len(self.chrsToSkip):
chromSizes = [x for x in chromSizes if x[0] not in self.chrsToSkip]
chrNames, chrLengths = zip(*chromSizes)
genomeSize = sum(chrLengths)
if self.stepSize is None:
if self.region is None:
self.stepSize = max(int(float(genomeSize) / self.numberOfSamples), 1)
else:
# compute the step size, based on the number of samples
# and the length of the region studied
(chrom, start, end) = mapReduce.getUserRegion(chromSizes, self.region)[:3]
self.stepSize = max(int(float(end - start) / self.numberOfSamples), 1)
# number of samples is better if large
if np.mean(chrLengths) < self.stepSize:
min_num_of_samples = int(genomeSize / np.mean(chrLengths))
raise ValueError("numberOfSamples has to be bigger than {} ".format(min_num_of_samples))
max_mapped = max([x.mapped for x in bamFilesHandlers])
reads_per_bp = float(max_mapped) / genomeSize
# chunkSize = int(100 / ( reads_per_bp * len(bamFilesList)) )
chunkSize = int(self.stepSize * 1e3 / (reads_per_bp * len(bamFilesHandlers)))
[bam_h.close() for bam_h in bamFilesHandlers]
if self.verbose:
print "step size is {}".format(self.stepSize)
if self.region:
# in case a region is used, append the tilesize
self.region += ":{}".format(self.binLength)
# use map reduce to call countReadsInRegions_wrapper
imap_res = mapReduce.mapReduce([],
countReadsInRegions_wrapper,
chromSizes,
self_=self,
genomeChunkLength=chunkSize,
bedFile=self.bedFile,
blackListFileName=self.blackListFileName,
region=self.region,
numberOfProcessors=self.numberOfProcessors)
if self.out_file_for_raw_data:
if len(non_common):
sys.stderr.write("*Warning*\nThe resulting bed file does not contain information for "
"the chromosomes that were not common between the bigwig files\n")
# concatenate intermediary bedgraph files
for _values, tempFileName in imap_res:
if tempFileName:
# concatenate all intermediate tempfiles into one
shutil.copyfileobj(open(tempFileName, 'r'), self.out_file_for_raw_data)
os.remove(tempFileName)
self.out_file_for_raw_data.close()
try:
num_reads_per_bin = np.concatenate([x[0] for x in imap_res], axis=0)
return num_reads_per_bin
except ValueError:
if self.bedFile:
sys.exit('\nNo coverage values could be computed.\n\n'
'Please check that the chromosome names in the BED file are found on the bam files.\n\n'
'The valid chromosome names are:\n{}'.format(chrNames))
else:
sys.exit('\nNo coverage values could be computed.\n\nCheck that all bam files are valid and '
'contain mapped reads.')
def get_read_and_fragment_length(bamFile, return_lengths=False, blackListFileName=None,
binSize=50000, distanceBetweenBins=1000000,
numberOfProcessors=None, verbose=False):
"""
Estimates the fragment length and read length through sampling
Parameters
----------
bamFile : str
BAM file name
return_lengths : bool
numberOfProcessors : int
verbose : bool
binSize : int
distanceBetweenBins : int
Returns
-------
d : dict
tuple of two dictionaries, one for the fragment length and the other
for the read length. The dictionaries summarise the mean, median etc. values
"""
bam_handle = bamHandler.openBam(bamFile)
chrom_sizes = zip(bam_handle.references, bam_handle.lengths)
distanceBetweenBins *= 2
fl = []
while len(fl) < 1000 and distanceBetweenBins > 1:
distanceBetweenBins /= 2
stepsize = binSize + distanceBetweenBins
imap_res = mapReduce.mapReduce((bam_handle.filename, distanceBetweenBins),
getFragmentLength_wrapper,
chrom_sizes,
genomeChunkLength=stepsize,
blackListFileName=blackListFileName,
numberOfProcessors=numberOfProcessors,
verbose=verbose)
fl = np.concatenate(imap_res)
if len(fl):
fragment_length = fl[:, 0]
read_length = fl[:, 1]
if fragment_length.mean() > 0:
fragment_len_dict = {'sample_size': len(fragment_length),
'min': fragment_length.min(),
'qtile25': np.percentile(fragment_length, 25),
'mean': np.mean(fragment_length),
'median': np.median(fragment_length),
'qtile75': np.percentile(fragment_length, 75),
'max': fragment_length.max(),
'std': np.std(fragment_length)}
else:
fragment_len_dict = None
if return_lengths and fragment_len_dict is not None:
fragment_len_dict['lengths'] = fragment_length
read_len_dict = {'sample_size': len(read_length),
'min': read_length.min(),
'qtile25': np.percentile(read_length, 25),
'mean': np.mean(read_length),
'median': np.median(read_length),
'qtile75': np.percentile(read_length, 75),
'max': read_length.max(),
'std': np.std(read_length)}
if return_lengths:
read_len_dict['lengths'] = read_length
else:
fragment_len_dict = None
read_len_dict = None
return fragment_len_dict, read_len_dict
def getNumReadsPerBin(bamFilesList, binLength, numberOfSamples,
defaultFragmentLength, numberOfProcessors=1,
skipZeros=True, verbose=False, region=None,
bedFile=None, extendPairedEnds=True,
minMappingQuality=None,
ignoreDuplicates=False,
chrsToSkip=[],
stepSize=None,
samFlag=None):
r"""
This function visits a number of sites and returs a matrix containing read
counts. Each row to one sampled site and each column correspond to each of
the bamFiles.
If the chrsToSkip is given, then counts are filter out from this
chromosome which, unless a female sample is used, the counts are less
compared to autosomes. For most applications this is irrelevant but for
other cases, like when stimating the best scaling factor, this is
important.
The test data contains reads for 200 bp
>>> test = Tester()
The transpose function is used to get a nicer looking output.
The first line corresponds to the number of reads per bin in bam file 1
>>> np.transpose(getNumReadsPerBin([test.bamFile1, test.bamFile2],
... 50, 4, 0, skipZeros=True))
array([[ 0., 1., 1.],
[ 1., 1., 2.]])
>>> aa = np.transpose(getNumReadsPerBin([test.bamFile1, test.bamFile2],
... 50, 4, 0, skipZeros=True))
>>> np.savez('/tmp/aa', aa)
"""
# Try to determine an optimal fraction of the genome (chunkSize) that is sent to
# workers for analysis. If too short, too much time is spend loading the files
# if too long, some processors end up free.
# the following values are empirical
bamFilesHandlers = [ bamHandler.openBam(x) for x in bamFilesList ]
chromSizes = getCommonChrNames(bamFilesHandlers, verbose=verbose)
# skip chromosome in the list. This is usually for the
# X chromosome which may have either one copy in a male sample
# or a mixture of male/female and is unreliable.
# Also the skip may contain heterochromatic regions and
# mitochondrial DNA
if len(chrsToSkip): chromSizes = [ x for x in chromSizes if x[0] not in chrsToSkip ]
chrNames, chrLengths = zip(*chromSizes)
genomeSize = sum(chrLengths)
max_mapped = max( [ x.mapped for x in bamFilesHandlers ] )
reads_per_bp = float(max_mapped) / genomeSize
# chunkSize = int(100 / ( reads_per_bp * len(bamFilesList)) )
if stepSize is None:
stepSize = max(int( float(genomeSize) / numberOfSamples ), 1 )
chunkSize = int (stepSize * 1e3 / ( reads_per_bp * len(bamFilesHandlers)) )
[ bam_h.close() for bam_h in bamFilesHandlers]
if verbose:
print "step size is {}".format(stepSize)
if region:
# in case a region is used, append the tilesize
region += ":{}".format(binLength)
imap_res = mapReduce.mapReduce( (bamFilesList, stepSize, binLength,
defaultFragmentLength, skipZeros,
extendPairedEnds, minMappingQuality,
ignoreDuplicates, samFlag),
countReadsInRegions_wrapper,
chromSizes,
genomeChunkLength=chunkSize,
bedFile=bedFile,
region=region,
numberOfProcessors=numberOfProcessors)
try:
num_reads_per_bin = np.concatenate(imap_res, axis=0)
except ValueError:
if bedFile:
exit('\nNo coverage values could be computed.\n\n'
'Please check that the chromosome names in the BED file are found on the bam files.\n\n'
'The valid chromosome names are:\n{}'.format(chrNames))
else:
exit('\nNo coverage values could be computed.\n\nCheck that all bam files are valid and '
'contain mapped reads.')
return num_reads_per_bin
def writeBedGraph(
bamOrBwFileList,
outputFileName,
fragmentLength,
func,
funcArgs,
tileSize=25,
region=None,
numberOfProcessors=None,
format="bedgraph",
extendPairedEnds=True,
missingDataAsZero=False,
smoothLength=0,
fixed_step=False,
):
r"""
Given a list of bamfiles, a function and a function arguments,
this method writes a bedgraph file (or bigwig) file
for a partition of the genome into tiles of given size
and a value for each tile that corresponds to the given function
and that is related to the coverage underlying the tile.
"""
bamHandlers = [
bamHandler.openBam(indexedFile) for indexedFile, fileFormat in bamOrBwFileList if fileFormat == "bam"
]
if len(bamHandlers):
genomeChunkLength = getGenomeChunkLength(bamHandlers, tileSize)
# check if both bam files correspond to the same species
# by comparing the chromosome names:
chromNamesAndSize, __ = getCommonChrNames(bamHandlers, verbose=False)
else:
genomeChunkLength = int(10e6)
bigwigs = [fileName for fileName, fileFormat in bamOrBwFileList if fileFormat == "bigwig"]
cCommon = []
chromNamesAndSize = {}
for bw in bigwigs:
bwh = pyBigWig.open(bw)
for chromName, size in bwh.chroms().items():
if chromName in chromNamesAndSize:
cCommon.append(chromName)
if chromNamesAndSize[chromName] != size:
print "\nWARNING\n" "Chromosome {} length reported in the " "bigwig files differ.\n{} for {}\n" "{} for {}.\n\nThe smallest " "length will be used".format(
chromName, chromNamesAndSize[chromName], bigwigs[0], size, bw
)
chromNamesAndSize[chromName] = min(chromNamesAndSize[chromName], size)
else:
chromNamesAndSize[chromName] = size
bwh.close()
# get the list of common chromosome names and sizes
chromNamesAndSize = [(k, v) for k, v in chromNamesAndSize.iteritems() if k in cCommon]
if region:
# in case a region is used, append the tilesize
region += ":{}".format(tileSize)
res = mapReduce.mapReduce(
(
tileSize,
fragmentLength,
bamOrBwFileList,
func,
funcArgs,
extendPairedEnds,
smoothLength,
missingDataAsZero,
fixed_step,
),
writeBedGraph_wrapper,
chromNamesAndSize,
genomeChunkLength=genomeChunkLength,
region=region,
numberOfProcessors=numberOfProcessors,
)
# concatenate intermediary bedgraph files
outFile = open(outputFileName + ".bg", "wb")
for tempFileName in res:
if tempFileName:
# concatenate all intermediate tempfiles into one
# bedgraph file
shutil.copyfileobj(open(tempFileName, "rb"), outFile)
os.remove(tempFileName)
bedGraphFile = outFile.name
outFile.close()
if format == "bedgraph":
os.rename(bedGraphFile, outputFileName)
if debug:
print "output file: %s" % (outputFileName)
else:
bedGraphToBigWig(chromNamesAndSize, bedGraphFile, outputFileName, True)
if debug:
print "output file: %s" % (outputFileName)
os.remove(bedGraphFile)
def getNumReadsPerBin(bamFilesList, binLength, numberOfSamples,
defaultFragmentLength, numberOfProcessors=1,
skipZeros=True, verbose=False, region=None,
bedFile=None, extendPairedEnds=True,
minMappingQuality=None,
ignoreDuplicates=False,
chrsToSkip=[],
stepSize=None):
r"""
This function visits a number of sites and returs a matrix containing read
counts. Each row to one sampled site and each column correspond to each of
the bamFiles.
If the chrsToSkip is given, then counts are filter out from this
chromosome which, unless a female sample is used, the counts are less
compared to autosomes. For most applications this is irrelevant but for
other cases, like when stimating the best scaling factor, this is
important.
The test data contains reads for 200 bp
>>> test = Tester()
The transpose function is used to get a nicer looking output.
The first line corresponds to the number of reads per bin in bam file 1
>>> np.transpose(getNumReadsPerBin([test.bamFile1, test.bamFile2],
... 50, 4, 0, skipZeros=True))
array([[ 0., 1., 1.],
[ 1., 1., 2.]])
>>> aa = np.transpose(getNumReadsPerBin([test.bamFile1, test.bamFile2],
... 50, 4, 0, skipZeros=True))
>>> np.savez('/tmp/aa', aa)
"""
# Try to determine an optimal fraction of the genome (chunkSize) that is sent to
# workers for analysis. If too short, too much time is spend loading the files
# if too long, some processors end up free.
# the following values are empirical
bamFilesHandlers = [ bamHandler.openBam(x) for x in bamFilesList ]
chromSizes = getCommonChrNames(bamFilesHandlers, verbose=verbose)
# skip chromosome in the list. This is usually for the
# X chromosome which may have either one copy in a male sample
# or a mixture of male/female and is unreliable.
# Also the skip may contain heterochromatic regions and
# mitochondrial DNA
if len(chrsToSkip): chromSizes = [ x for x in chromSizes if x[0] not in chrsToSkip ]
chrNames, chrLengths = zip(*chromSizes)
genomeSize = sum(chrLengths)
max_mapped = max( [ x.mapped for x in bamFilesHandlers ] )
reads_per_bp = float(max_mapped) / genomeSize
# chunkSize = int(100 / ( reads_per_bp * len(bamFilesList)) )
if stepSize is None:
stepSize = max(int( float(genomeSize) / numberOfSamples ), 1 )
chunkSize = int (stepSize * 1e3 / ( reads_per_bp * len(bamFilesHandlers)) )
[ bam_h.close() for bam_h in bamFilesHandlers]
if verbose:
print "step size is {}".format(stepSize)
if region:
# in case a region is used, append the tilesize
region += ":{}".format(binLength)
imap_res = mapReduce.mapReduce( (bamFilesList, stepSize, binLength,
defaultFragmentLength, skipZeros,
extendPairedEnds, minMappingQuality,
ignoreDuplicates),
countReadsInRegions_wrapper,
chromSizes,
genomeChunkLength=chunkSize,
bedFile=bedFile,
region=region,
numberOfProcessors=numberOfProcessors)
num_reads_per_bin = np.concatenate(imap_res, axis=0)
return num_reads_per_bin
def writeBedGraph(
bamOrBwFileList, outputFileName, fragmentLength,
func, funcArgs, tileSize=25, region=None, numberOfProcessors=None,
format="bedgraph", extendPairedEnds=True, zerosToNans=True,
smoothLength=0, fixed_step=False):
r"""
Given a list of bamfiles, a function and a function arguments,
this method writes a bedgraph file (or bigwig) file
for a partition of the genome into tiles of given size
and a value for each tile that corresponds to the given function
and that is related to the coverage underlying the tile.
>>> test = Tester()
>>> outFile = tempfile.NamedTemporaryFile()
>>> funcArgs = {'scaleFactor': 1.0}
>>> writeBedGraph([(test.bamFile1, 'bam')], outFile.name,
... 0, scaleCoverage, funcArgs, region='3R:0:200')
>>> open(outFile.name, 'r').readlines()
['3R\t100\t200\t1.0\n']
>>> outFile.close()
"""
bigwig_info = cfg.config.get('external_tools', 'bigwig_info')
bamHandlers = [openBam(indexedFile) for
indexedFile,
fileFormat in bamOrBwFileList if fileFormat == 'bam']
if len(bamHandlers):
genomeChunkLength = getGenomeChunkLength(bamHandlers, tileSize)
# check if both bam files correspond to the same species
# by comparing the chromosome names:
chromNamesAndSize = getCommonChrNames(bamHandlers, verbose=False)
else:
genomeChunkLength = int(10e6)
bigwigs = [fileName for fileName,
fileFormat in bamOrBwFileList if fileFormat == 'bigwig']
cCommon = []
chromNamesAndSize = {}
for bw in bigwigs:
inBlock = False
for line in os.popen(
"{} -chroms {}".format(bigwig_info, bw)).readlines():
if line[0:10] == "chromCount":
inBlock = True
continue
if line[0:5] == "bases":
break
if inBlock:
chromName, id, size = line.strip().split(" ")
size = int(size)
if chromName in chromNamesAndSize:
cCommon.append(chromName)
if chromNamesAndSize[chromName] != size:
print "\nWARNING\n" \
"Chromosome {} length reported in the " \
"bigwig files differ.\n{} for {}\n" \
"{} for {}.\n\nThe smallest " \
"length will be used".format(
chromName, chromNamesAndSize[chromName],
bigwigs[0], size, bigwigs[1])
chromNamesAndSize[chromName] = min(
chromNamesAndSize[chromName], size)
else:
chromNamesAndSize[chromName] = size
# get the list of common chromosome names and sizes
chromNamesAndSize = [(k, v) for k, v in chromNamesAndSize.iteritems()
if k in cCommon]
if region:
# in case a region is used, append the tilesize
region += ":{}".format(tileSize)
res = mapReduce.mapReduce((tileSize, fragmentLength, bamOrBwFileList,
func, funcArgs, extendPairedEnds, smoothLength,
zerosToNans, fixed_step),
writeBedGraph_wrapper,
chromNamesAndSize,
genomeChunkLength=genomeChunkLength,
region=region,
numberOfProcessors=numberOfProcessors)
# concatenate intermediary bedgraph files
outFile = open(outputFileName + ".bg", 'wb')
for tempFileName in res:
if tempFileName:
# concatenate all intermediate tempfiles into one
# bedgraph file
shutil.copyfileobj(open(tempFileName, 'rb'), outFile)
os.remove(tempFileName)
bedGraphFile = outFile.name
outFile.close()
if format == 'bedgraph':
os.rename(bedGraphFile, outputFileName)
if debug:
print "output file: %s" % (outputFileName)
else:
bedGraphToBigWig(
chromNamesAndSize, bedGraphFile, outputFileName, False)
if debug:
print "output file: %s" % (outputFileName)
os.remove(bedGraphFile)
def writeBedGraph(bamFilesList, outputFileName, fragmentLength,
func, funcArgs, tileSize=25, region=None,
numberOfProcessors=None, format="bedgraph",
extendPairedEnds=True, zerosToNans=True, smoothLength=0,
minMappingQuality=None, ignoreDuplicates=False,
fragmentFromRead_func=None):
r"""
Given a list of bamfiles, a function and a function arguments,
this method writes a bedgraph file (or bigwig) file
for a partition of the genome into tiles of given size
and a value for each tile that corresponds to the given function
and that is related to the coverage underlying the tile.
>>> test = Tester()
>>> outFile = tempfile.NamedTemporaryFile()
>>> funcArgs = {'scaleFactor': 1.0}
>>> writeBedGraph( [test.bamFile1], outFile.name,
... 0, scaleCoverage, funcArgs, region='3R:0:200')
>>> open(outFile.name, 'r').readlines()
['3R\t100\t200\t1.0\n']
>>> outFile.close()
"""
bamHandlers = [openBam(x) for x in bamFilesList]
genomeChunkLength = getGenomeChunkLength(bamHandlers, tileSize)
# check if both bam files correspond to the same species
# by comparing the chromosome names:
chromNamesAndSize = getCommonChrNames(bamHandlers, verbose=False)
if region:
# in case a region is used, append the tilesize
region += ":{}".format(tileSize)
res = mapReduce.mapReduce((tileSize, fragmentLength, bamFilesList,
func, funcArgs, extendPairedEnds, smoothLength,
zerosToNans, minMappingQuality,
ignoreDuplicates,
fragmentFromRead_func),
writeBedGraph_wrapper,
chromNamesAndSize,
genomeChunkLength=genomeChunkLength,
region=region,
numberOfProcessors=numberOfProcessors)
# concatenate intermediary bedgraph files
outFile = open(outputFileName + ".bg", 'wb')
for tempFileName in res:
if tempFileName:
# concatenate all intermediate tempfiles into one
# bedgraph file
shutil.copyfileobj(open(tempFileName, 'rb'), outFile)
os.remove(tempFileName)
bedGraphFile = outFile.name
outFile.close()
if format == 'bedgraph':
os.rename(bedGraphFile, outputFileName)
if debug:
print "output file: %s" % (outputFileName)
else:
bedGraphToBigWig(
chromNamesAndSize, bedGraphFile, outputFileName, False)
if debug:
print "output file: %s" % (outputFileName)
os.remove(bedGraphFile)
def get_read_and_fragment_length(bamFile,
return_lengths=False,
blackListFileName=None,
binSize=50000,
distanceBetweenBins=1000000,
numberOfProcessors=None,
verbose=False):
"""
Estimates the fragment length and read length through sampling
Parameters
----------
bamFile : str
BAM file name
return_lengths : bool
numberOfProcessors : int
verbose : bool
binSize : int
distanceBetweenBins : int
Returns
-------
d : dict
tuple of two dictionaries, one for the fragment length and the other
for the read length. The dictionaries summarise the mean, median etc. values
"""
bam_handle = bamHandler.openBam(bamFile)
chrom_sizes = zip(bam_handle.references, bam_handle.lengths)
distanceBetweenBins *= 2
fl = []
while len(fl) < 1000 and distanceBetweenBins > 1:
distanceBetweenBins /= 2
stepsize = binSize + distanceBetweenBins
imap_res = mapReduce.mapReduce(
(bam_handle.filename, distanceBetweenBins),
getFragmentLength_wrapper,
chrom_sizes,
genomeChunkLength=stepsize,
blackListFileName=blackListFileName,
numberOfProcessors=numberOfProcessors,
verbose=verbose)
fl = np.concatenate(imap_res)
if len(fl):
fragment_length = fl[:, 0]
read_length = fl[:, 1]
if fragment_length.mean() > 0:
fragment_len_dict = {
'sample_size': len(fragment_length),
'min': fragment_length.min(),
'qtile25': np.percentile(fragment_length, 25),
'mean': np.mean(fragment_length),
'median': np.median(fragment_length),
'qtile75': np.percentile(fragment_length, 75),
'max': fragment_length.max(),
'std': np.std(fragment_length)
}
else:
fragment_len_dict = None
if return_lengths and fragment_len_dict is not None:
fragment_len_dict['lengths'] = fragment_length
read_len_dict = {
'sample_size': len(read_length),
'min': read_length.min(),
'qtile25': np.percentile(read_length, 25),
'mean': np.mean(read_length),
'median': np.median(read_length),
'qtile75': np.percentile(read_length, 75),
'max': read_length.max(),
'std': np.std(read_length)
}
if return_lengths:
read_len_dict['lengths'] = read_length
else:
fragment_len_dict = None
read_len_dict = None
return fragment_len_dict, read_len_dict
def getNumReadsPerBin(bamFilesList, binLength, numberOfSamples,
defaultFragmentLength, numberOfProcessors=1,
skipZeros=True, verbose=False, region=None,
bedFile=None, extendPairedEnds=True,
minMappingQuality=None,
ignoreDuplicates=False,
chrsToSkip=[],
stepSize=None,
samFlag=None):
r"""
This function collects read counts (coverage) from several bam files and returns
an numpy array with the results. This function does not explicitly do the
coverage computation, instead divides the work into smaller chunks that are
sent to individual processors.
Parameters
----------
bamFilesList : list
List containing the names of indexed bam files. E.g. ['file1.bam', 'file2.bam']
binLength : int
Length of the window/bin. This value is overruled by ``bedFile`` if present.
numberOfSamples : int
Total number of samples. The genome is divided into ``numberOfSamples``, each
with a window/bin length equal to ``binLength``. This value is overruled
by ``stepSize`` in case such value is present and by ``bedFile`` in which
case the number of samples and bins are defined in the bed file
defaultFragmentLength : int
fragment length to extend reads that are not paired. Paired reads are extended to
the fragment length defined by the mate distance. For Illumina reads, usual values
are around 300. This value can be determined using the peak caller MACS2 or can be
approximated by the fragment lengths computed when preparing the library for sequencing.
numberOfProcessors : int
Number of processors to use. Default is 4
skipZeros : bool
Default is True. This option decides if regions having zero coverage in all bam files
should be skipped or kept.
verbose : bool
Output messages. Default: False
region : str
Region to limit the computation in the form chrom:start:end.
bedFile : str
Name of a bed file containing the regions for wich to compute the coverage. This option
overrules ``binLength``, ``numberOfSamples`` and ``stepSize``.
extendPairedEnds : bool
Whether coverage should be computed for the extended read length (i.e. the region covered
by the two mates or the regions expected to be covered by single-reads). Default: true
minMappingQuality : int
Reads of a mapping quality less than the give value are not considered. Default: None
ignoreDuplicates : bool
Whether read duplicates (same start, end position. If paired-end, same start-end for mates) are
to be excluded. Default: false
chrToSkip: list
List with names of chromosomes that do not want to be included in the coverage computation.
This is useful to remove unwanted chromosomes (e.g. 'random' or 'Het').
stepSize : int
the positions for which the coverage is computed are defined as follows:
``range(start, end, stepSize)``. Thus, a stepSize of 1, will compute
the coverage at each base pair. If the stepSize is equal to the
binLength then the coverage is computed for consecutive bins. If seepSize is
smaller than the binLength, then teh bins will overlap.
samFlag : int
If given, only reads having such flag are considered. For example, to get only
reads that are the first mates a samFlag of 64 could be used. Similarly, the
samFlag can be used to select only reads mapping on the forward (or reverse) strand
or to get only properly paired reads.
Returns
-------
numpy array
Each row correspond to each bin/bed region and each column correspond to each of
the bamFiles. If ``skipZeros`` is used, then the result may have less rows
than expected
Examples
--------
The test data contains reads for 200 bp.
>>> test = Tester()
The transpose function is used to get a nicer looking output.
The first line corresponds to the number of reads per bin in bam file 1
>>> np.transpose(getNumReadsPerBin([test.bamFile1, test.bamFile2],
... 50, 4, 0, skipZeros=True))
array([[ 0., 1., 1.],
[ 1., 1., 2.]])
"""
# Try to determine an optimal fraction of the genome (chunkSize) that is sent to
# workers for analysis. If too short, too much time is spend loading the files
# if too long, some processors end up free.
# the following values are empirical
bamFilesHandlers = [ bamHandler.openBam(x) for x in bamFilesList ]
chromSizes = getCommonChrNames(bamFilesHandlers, verbose=verbose)
# skip chromosome in the list. This is usually for the
# X chromosome which may have either one copy in a male sample
# or a mixture of male/female and is unreliable.
# Also the skip may contain heterochromatic regions and
# mitochondrial DNA
if len(chrsToSkip): chromSizes = [ x for x in chromSizes if x[0] not in chrsToSkip ]
chrNames, chrLengths = zip(*chromSizes)
genomeSize = sum(chrLengths)
max_mapped = max( [ x.mapped for x in bamFilesHandlers ] )
reads_per_bp = float(max_mapped) / genomeSize
# chunkSize = int(100 / ( reads_per_bp * len(bamFilesList)) )
if stepSize is None:
stepSize = max(int( float(genomeSize) / numberOfSamples ), 1 )
chunkSize = int (stepSize * 1e3 / ( reads_per_bp * len(bamFilesHandlers)) )
[ bam_h.close() for bam_h in bamFilesHandlers]
if verbose:
print "step size is {}".format(stepSize)
if region:
# in case a region is used, append the tilesize
region += ":{}".format(binLength)
imap_res = mapReduce.mapReduce( (bamFilesList, stepSize, binLength,
defaultFragmentLength, skipZeros,
extendPairedEnds, minMappingQuality,
ignoreDuplicates, samFlag),
countReadsInRegions_wrapper,
chromSizes,
genomeChunkLength=chunkSize,
bedFile=bedFile,
region=region,
numberOfProcessors=numberOfProcessors)
try:
num_reads_per_bin = np.concatenate(imap_res, axis=0)
except ValueError:
if bedFile:
exit('\nNo coverage values could be computed.\n\n'
'Please check that the chromosome names in the BED file are found on the bam files.\n\n'
'The valid chromosome names are:\n{}'.format(chrNames))
else:
exit('\nNo coverage values could be computed.\n\nCheck that all bam files are valid and '
'contain mapped reads.')
return num_reads_per_bin
def writeBedGraph(
bamOrBwFileList, outputFileName, fragmentLength,
func, funcArgs, tileSize=25, region=None, numberOfProcessors=None,
format="bedgraph", extendPairedEnds=True, missingDataAsZero=False,
smoothLength=0, fixed_step=False):
r"""
Given a list of bamfiles, a function and a function arguments,
this method writes a bedgraph file (or bigwig) file
for a partition of the genome into tiles of given size
and a value for each tile that corresponds to the given function
and that is related to the coverage underlying the tile.
>>> test = Tester()
>>> outFile = tempfile.NamedTemporaryFile()
>>> funcArgs = {'scaleFactor': 1.0}
>>> writeBedGraph([(test.bamFile1, 'bam')], outFile.name,
... 0, scaleCoverage, funcArgs, region='3R:0:200')
>>> open(outFile.name, 'r').readlines()
['3R\t100\t200\t1.0\n']
>>> outFile.close()
"""
bigwig_info = cfg.config.get('external_tools', 'bigwig_info')
bamHandlers = [openBam(indexedFile) for
indexedFile,
fileFormat in bamOrBwFileList if fileFormat == 'bam']
if len(bamHandlers):
genomeChunkLength = getGenomeChunkLength(bamHandlers, tileSize)
# check if both bam files correspond to the same species
# by comparing the chromosome names:
chromNamesAndSize = getCommonChrNames(bamHandlers, verbose=False)
else:
genomeChunkLength = int(10e6)
bigwigs = [fileName for fileName,
fileFormat in bamOrBwFileList if fileFormat == 'bigwig']
cCommon = []
chromNamesAndSize = {}
for bw in bigwigs:
inBlock = False
for line in os.popen(
"{} -chroms {}".format(bigwig_info, bw)).readlines():
if line[0:10] == "chromCount":
inBlock = True
continue
if line[0:5] == "bases":
break
if inBlock:
chromName, id, size = line.strip().split(" ")
size = int(size)
if chromName in chromNamesAndSize:
cCommon.append(chromName)
if chromNamesAndSize[chromName] != size:
print "\nWARNING\n" \
"Chromosome {} length reported in the " \
"bigwig files differ.\n{} for {}\n" \
"{} for {}.\n\nThe smallest " \
"length will be used".format(
chromName, chromNamesAndSize[chromName],
bigwigs[0], size, bigwigs[1])
chromNamesAndSize[chromName] = min(
chromNamesAndSize[chromName], size)
else:
chromNamesAndSize[chromName] = size
# get the list of common chromosome names and sizes
chromNamesAndSize = [(k, v) for k, v in chromNamesAndSize.iteritems()
if k in cCommon]
if region:
# in case a region is used, append the tilesize
region += ":{}".format(tileSize)
res = mapReduce.mapReduce((tileSize, fragmentLength, bamOrBwFileList,
func, funcArgs, extendPairedEnds, smoothLength,
missingDataAsZero, fixed_step),
writeBedGraph_wrapper,
chromNamesAndSize,
genomeChunkLength=genomeChunkLength,
region=region,
numberOfProcessors=numberOfProcessors)
# concatenate intermediary bedgraph files
outFile = open(outputFileName + ".bg", 'wb')
for tempFileName in res:
if tempFileName:
# concatenate all intermediate tempfiles into one
# bedgraph file
shutil.copyfileobj(open(tempFileName, 'rb'), outFile)
os.remove(tempFileName)
bedGraphFile = outFile.name
outFile.close()
if format == 'bedgraph':
os.rename(bedGraphFile, outputFileName)
if debug:
print "output file: %s" % (outputFileName)
else:
bedGraphToBigWig(
chromNamesAndSize, bedGraphFile, outputFileName, False)
if debug:
print "output file: %s" % (outputFileName)
os.remove(bedGraphFile)
def get_read_and_fragment_length(bamFile, return_lengths=False,
numberOfProcessors=None, verbose=False):
"""
Estimates the fragment length and read length through sampling
Parameters
----------
bamFile : str
BAM file name
return_lengths : bool
numberOfProcessors : int
verbose : bool
Returns
-------
d : dict
tuple of two dictionaries, one for the fragment length and the other
for the read length. The dictionaries summarise the mean, median etc. values
"""
bam_handle = bamHandler.openBam(bamFile)
chrom_sizes = zip(bam_handle.references, bam_handle.lengths)
chunk_size = int(float(sum(bam_handle.lengths)) * 0.3 / max(numberOfProcessors, len(bam_handle.lengths)))
# avoid small chunk sizes to split the computation
chunk_size = max(chunk_size, 100000)
imap_res = mapReduce.mapReduce((bam_handle.filename, ),
getFragmentLength_wrapper,
chrom_sizes,
genomeChunkLength=chunk_size,
numberOfProcessors=numberOfProcessors,
verbose=verbose)
fl = np.concatenate(imap_res)
if len(fl):
fragment_length = fl[:, 0]
read_length = fl[:, 1]
if fragment_length.mean() > 0:
fragment_len_dict = {'sample_size': len(fragment_length),
'min': fragment_length.min(),
'qtile25': np.percentile(fragment_length, 25),
'mean': np.mean(fragment_length),
'median': np.median(fragment_length),
'qtile75': np.percentile(fragment_length, 75),
'max': fragment_length.max(),
'std': np.std(fragment_length)}
else:
fragment_len_dict = None
if return_lengths:
fragment_len_dict['lengths'] = fragment_length
read_len_dict = {'sample_size': len(read_length),
'min': read_length.min(),
'qtile25': np.percentile(read_length, 25),
'mean': np.mean(read_length),
'median': np.median(read_length),
'qtile75': np.percentile(read_length, 75),
'max': read_length.max(),
'std': np.std(read_length)}
if return_lengths:
read_len_dict['lengths'] = read_length
else:
fragment_len_dict = None
read_len_dict = None
return fragment_len_dict, read_len_dict