def extractContigs(self,
timer,
bids=[],
fasta=[],
prefix='',
cutoff=0):
"""Extract contigs and write to file"""
if prefix is None or prefix == '':
prefix=os.path.basename(self.dbFileName) \
.replace(".gm", "") \
.replace(".sm", "")
profile = self.loadProfile(timer, bids, cutoff)
bm = BinManager(profile)
# load all the contigs which have been assigned to bins
cp = ContigParser()
# contigs looks like cid->seq
contigs = {}
import mimetypes
try:
for file_name in fasta:
gm_open = open
try:
# handle gzipped files
mime = mimetypes.guess_type(file_name)
if mime[1] == 'gzip':
import gzip
gm_open = gzip.open
except:
print "Error when guessing contig file mimetype"
raise
with gm_open(file_name, "r") as f:
cp.getWantedSeqs(f, profile.contigNames, out_dict=contigs)
except:
print "Could not parse contig file:",fasta[0],sys.exc_info()[0]
raise
# now print out the sequences
print "Writing files"
for bid in bm.getBids():
file_name = os.path.join(self._outDir, "%s_bin_%d.fna" % (prefix, bid))
try:
with open(file_name, 'w') as f:
for cid in bm.profile.contigNames[bm.getBinIndices(bid)]:
if(cid in contigs):
f.write(">%s\n%s\n" % (cid, contigs[cid]))
else:
print "These are not the contigs you're looking for. ( %s )" % (cid)
except:
print "Could not open file for writing:",file_name,sys.exc_info()[0]
raise
def extractReads(self,
timer,
bids=[],
bams=[],
prefix="",
mixBams=False,
mixGroups=False,
mixReads=False,
interleaved=False,
bigFile=False,
headersOnly=False,
minMapQual=0,
maxMisMatches=1000,
useSuppAlignments=False,
useSecondaryAlignments=False,
threads=1,
verbose=False):
"""Extract reads from bam files and write to file
All logic is handled by BamM <- soon to be wrapped by StoreM"""
# load data
profile = self.loadProfile(timer, bids)
bm = BinManager(profile) # bins
print "Extracting reads"
# work out a set of targets to pass to the parser
targets = []
group_names = []
for bid in bm.getBids():
group_names.append("BIN_%d" % bid)
row_indices = bm.getBinIndices(bid)
targets.append(list(bm.profile.contigNames[row_indices]))
# get something to parse the bams with
bam_parser = BMBE(targets,
bams,
groupNames=group_names,
prefix=prefix,
outFolder=self._outDir,
mixBams=mixBams,
mixGroups=mixGroups,
mixReads=mixReads,
interleaved=interleaved,
bigFile=bigFile,
headersOnly=headersOnly,
minMapQual=minMapQual,
maxMisMatches=maxMisMatches,
useSuppAlignments=useSuppAlignments,
useSecondaryAlignments=useSecondaryAlignments)
bam_parser.extract(threads=threads,
verbose=verbose)
def extractReads(self,
timer,
bids=[],
bams=[],
prefix="",
mixBams=False,
mixGroups=False,
mixReads=False,
interleaved=False,
bigFile=False,
headersOnly=False,
minMapQual=0,
maxMisMatches=1000,
useSuppAlignments=False,
useSecondaryAlignments=False,
threads=1,
verbose=False):
"""Extract reads from bam files and write to file
All logic is handled by BamM <- soon to be wrapped by StoreM"""
# load data
profile = self.loadProfile(timer, bids)
bm = BinManager(profile) # bins
print "Extracting reads"
# work out a set of targets to pass to the parser
targets = []
group_names = []
for bid in bm.getBids():
group_names.append("BIN_%d" % bid)
row_indices = bm.getBinIndices(bid)
targets.append(list(bm.profile.contigNames[row_indices]))
# get something to parse the bams with
bam_parser = BMBE(targets,
bams,
groupNames=group_names,
prefix=prefix,
outFolder=self._outDir,
mixBams=mixBams,
mixGroups=mixGroups,
mixReads=mixReads,
interleaved=interleaved,
bigFile=bigFile,
headersOnly=headersOnly,
minMapQual=minMapQual,
maxMisMatches=maxMisMatches,
useSuppAlignments=useSuppAlignments,
useSecondaryAlignments=useSecondaryAlignments)
bam_parser.extract(threads=threads, verbose=verbose)
def extractContigs(self, timer, bids=[], fasta=[], prefix='', cutoff=0):
"""Extract contigs and write to file"""
if prefix is None or prefix == '':
prefix=os.path.basename(self.dbFileName) \
.replace(".gm", "") \
.replace(".sm", "")
profile = self.loadProfile(timer, bids, cutoff)
bm = BinManager(profile)
# load all the contigs which have been assigned to bins
cp = ContigParser()
# contigs looks like cid->seq
contigs = {}
import mimetypes
try:
for file_name in fasta:
gm_open = open
try:
# handle gzipped files
mime = mimetypes.guess_type(file_name)
if mime[1] == 'gzip':
import gzip
gm_open = gzip.open
except:
print "Error when guessing contig file mimetype"
raise
with gm_open(file_name, "r") as f:
cp.getWantedSeqs(f, profile.contigNames, out_dict=contigs)
except:
print "Could not parse contig file:", fasta[0], sys.exc_info()[0]
raise
# now print out the sequences
print "Writing files"
for bid in bm.getBids():
file_name = os.path.join(self._outDir,
"%s_bin_%d.fna" % (prefix, bid))
try:
with open(file_name, 'w') as f:
for cid in bm.profile.contigNames[bm.getBinIndices(bid)]:
if (cid in contigs):
f.write(">%s\n%s\n" % (cid, contigs[cid]))
else:
print "These are not the contigs you're looking for. ( %s )" % (
cid)
except:
print "Could not open file for writing:", file_name, sys.exc_info(
)[0]
raise
def extractMappingInfo(self,
timer,
bids=[],
prefix='',
separator='\t',
cutoff=0):
"""Extract markers from bins and write to file"""
if prefix is None or prefix == '':
prefix=os.path.basename(self.dbFileName) \
.replace(".gm", "") \
.replace(".sm", "")
profile = self.loadProfile(timer, bids, cutoff)
bm = BinManager(profile)
mt = MarkerCheckTreePrinter(profile)
# now print out the marker info
print "Writing files"
for bid in bm.getBids():
file_name = os.path.join(self._outDir,
"%s_bin_%d.txt" % (prefix, bid))
bin_indices = bm.getBinIndices([bid])
idx = np.flatnonzero(
np.in1d(profile.mapping.rowIndices, bin_indices))
labels = profile.mapping.markerNames[idx]
cnames = profile.contigNames[profile.mapping.rowIndices[idx]]
taxstrings = profile.mapping.taxstrings[idx]
try:
with open(file_name, 'w') as f:
#labels and lineages
f.write(
'#info table\n%s\n' %
separator.join(['label', 'taxonomy', 'contig_name']))
for (label, taxstring,
cname) in zip(labels, taxstrings, cnames):
f.write('%s\n' % separator.join(
[label, '\'%s\'' % taxstring, cname]))
#marker tree
f.write('\n#marker tree\n')
f.write(
mt.printTree(profile.mapping.rowIndices[idx],
leaves_list=bin_indices))
except:
print "Could not open file for writing:", file_name, sys.exc_info(
)[0]
raise
def extractMappingInfo(self,
timer,
bids=[],
prefix='',
separator='\t',
cutoff=0
):
"""Extract markers from bins and write to file"""
if prefix is None or prefix == '':
prefix=os.path.basename(self.dbFileName) \
.replace(".gm", "") \
.replace(".sm", "")
profile = self.loadProfile(timer, bids, cutoff)
bm = BinManager(profile)
mt = MarkerCheckTreePrinter(profile)
# now print out the marker info
print "Writing files"
for bid in bm.getBids():
file_name = os.path.join(self._outDir, "%s_bin_%d.txt" % (prefix, bid))
bin_indices = bm.getBinIndices([bid])
idx = np.flatnonzero(np.in1d(profile.mapping.rowIndices, bin_indices))
labels = profile.mapping.markerNames[idx]
cnames = profile.contigNames[profile.mapping.rowIndices[idx]]
taxstrings = profile.mapping.taxstrings[idx]
try:
with open(file_name, 'w') as f:
#labels and lineages
f.write('#info table\n%s\n' % separator.join(['label', 'taxonomy', 'contig_name']))
for (label, taxstring, cname) in zip(labels, taxstrings, cnames):
f.write('%s\n' % separator.join([label, '\'%s\'' % taxstring, cname]))
#marker tree
f.write('\n#marker tree\n')
f.write(mt.printTree(profile.mapping.rowIndices[idx], leaves_list=bin_indices))
except:
print "Could not open file for writing:",file_name,sys.exc_info()[0]
raise
