def test_compressor_dsrc():
"""Test dsrc codecs to gz and bz2"""
# Create a temporary directory and chdir in it:
tempdir = tempfile.TemporaryDirectory()
filename = sequana_data("test.fastq.gz")
shutil.copy(filename, tempdir.name)
cwd = os.path.abspath(os.curdir)
os.chdir(tempdir.name)
# We concert gz -> dsrc -> bz2 -> dsrc -> gz and must get the exact same
# However, since the compression is not deterministic, we should compare the
# content of the uncompressed file (input and output)
try:
compressor.main([
prog, "--source", "fastq.gz", "--target", "fastq.dsrc", "--quiet"
])
compressor.main([
prog, "--source", "fastq.dsrc", "--target", "fastq.bz2", "--quiet"
])
compressor.main([
prog, "--source", "fastq.bz2", "--target", "fastq.dsrc", "--quiet"
])
compressor.main([
prog, "--source", "fastq.dsrc", "--target", "fastq.gz", "--quiet"
])
except Exception as err:
raise Exception(err)
finally:
os.chdir(cwd)
f1 = FastQ(filename)
f2 = FastQ(tempdir.name + os.sep + os.path.basename(filename))
assert f1 == f2
def is_synchronised(self):
from sequana import FastQ
N = 0
for a, b in zip(FastQ(self.fq1), FastQ(self.fq2)):
a = a['identifier'].decode()
b = b['identifier'].decode()
a = a.split()[0]
b = b.split()[0]
if a.endswith("/1"):
id1 = a.rsplit("/1")[0]
elif a.endswith("/2"):
id1 = a.rsplit("/2")[0]
else:
id1 = a
if b.endswith("/1"):
id2 = b.rsplit("/1")[0]
elif b.endswith("/2"):
id2 = b.rsplit("/2")[0]
else:
id2 = b
if id1 != id2:
print("%s differs from %s" % (id1, id2))
print(a)
print(b)
return False
N += 1
print(N)
return True
def __init__(self, directory=".", prefix=""):
self.prefix = prefix
self.directory = directory
self.sample_name = "undefined"
# low quality isoforms
filename = "all.polished_lq.fastq"
self.lq_isoforms = self.get_file(filename)
if self.lq_isoforms:
logger.info("Reading {}".format(filename))
self.lq_sequence = FastQ(self.lq_isoforms)
# high quality isoforms
filename = "all.polished_hq.fastq"
self.hq_isoforms = self.get_file(filename)
if self.hq_isoforms:
logger.info("Reading {}".format(filename))
self.hq_sequence = FastQ(self.hq_isoforms)
# General info
filename = "file.csv"
self.csv = self.get_file(filename)
if self.csv:
logger.info("Reading {}".format(filename))
self.data = pd.read_csv(self.csv)
# CCS fasta sequence
#self.ccs = self.get_file("-ccs.tar.gz")
filename = "ccs.fasta"
self.ccs = self.get_file(filename, noprefix=True)
if self.ccs:
logger.info("Reading {}".format(filename))
self.ccs = FastA(self.ccs)
def __init__(self, directory=".", prefix=""):
self.prefix = prefix
self.directory = directory
self.sample_name = "undefined"
# low quality isoforms
filename = "all.polished_lq.fastq"
self.lq_isoforms = self.get_file(filename)
if self.lq_isoforms:
logger.info("Reading {}".format(filename))
self.lq_sequence = FastQ(self.lq_isoforms)
# high quality isoforms
filename = "all.polished_hq.fastq"
self.hq_isoforms = self.get_file(filename)
if self.hq_isoforms:
logger.info("Reading {}".format(filename))
self.hq_sequence = FastQ(self.hq_isoforms)
# General info
filename = "file.csv"
self.csv = self.get_file(filename)
if self.csv:
logger.info("Reading {}".format(filename))
self.data = pd.read_csv(self.csv)
# CCS fasta sequence
#self.ccs = self.get_file("-ccs.tar.gz")
filename = "ccs.fasta"
self.ccs = self.get_file(filename, noprefix=True)
if self.ccs:
logger.info("Reading {}".format(filename))
self.ccs = FastA(self.ccs)
def main(args=None):
if args is None:
args = sys.argv[:]
user_options = Options(prog="sequana")
# If --help or no options provided, show the help
if len(args) == 1:
user_options.parse_args(["prog", "--help"])
else:
options = user_options.parse_args(args[1:])
reference = options.reference
if options.file1 and options.file2:
fastq = "%s %s" % (options.file1, options.file2)
elif options.file1 and not options.file2:
fastq = "%s" % (options.file1)
elif options.file1 is None:
raise ValueError("--file1 must be used")
from sequana import FastQ
from sequana import FastA
S = 0
for this in FastQ(options.file1):
S += len(this['sequence'])
if options.file2:
for this in FastQ(options.file2):
S += len(this['sequence'])
ref = FastA(options.reference)
coverage = float(S) / len(ref.sequences[0])
print('Theoretical Depth of Coverage : %s' % coverage)
params = {"reference": reference, "fastq": fastq, "thread": options.thread}
# indexing
shellcmd("bwa index %(reference)s " % params)
cmd = "samtools faidx %(reference)s " % params
# mapping
cmd = "bwa mem -M " # mark shorter split read as secondary; -M is not compulsary but recommended
if options.pacbio:
cmd += "-x pacbio "
cmd += r" -t %(thread)s -R @RG\\tID:1\\tSM:1\\tPL:illumina -T 30 %(reference)s %(fastq)s "
# Samtools options:
# S:ignore input format
# h:include header
# b:bam output
if options.sambamba is False:
cmd += "| samtools view -Sbh | "
# sorting BAM
cmd += "samtools sort -@ %(thread)s -o %(reference)s.sorted.bam -"
shellcmd(cmd % params)
else:
# FIXME use sambamba for the view as well
cmd += "| samtools view -Sbu - | sambamba sort /dev/stdin -o %(reference)s.sorted.bam -t %(thread)s --tmpdir=./tmp " % params
shellcmd(cmd % params)
def test_fastq_unzipped():
for thisdata in [data, datagz]:
# isntanciation
f = fastq.FastQ(thisdata)
assert f.data_format == "Illumina_1.8+"
# count lines
# rune it twice because we want to make sure re-running count_lines
# (decompression with zlib) works when run again.
assert f.count_lines() == 1000
assert f.count_lines() == 1000
assert f.count_reads() == 250
assert f.count_reads() == 250
# extract head of the file into an unzipped file
ft = TempFile()
f.extract_head(100, ft.name)
fcheck = fastq.FastQ(ft.name)
assert fcheck.count_lines() == 100
ft.delete()
# extract head of the file and zip output
ft = TempFile(suffix=".gz")
f.extract_head(100, ft.name)
fcheck = fastq.FastQ(ft.name)
assert fcheck.count_lines() == 100
ft.delete()
with FastQ(thisdata) as ff:
assert len(ff) == 250
with TempFile() as fh:
selection = f.select_random_reads(10, fh.name)
f.select_random_reads(selection, fh.name)
def test_filter():
f = fastq.FastQ(data)
# keeps all
with TempFile() as fh:
f.filter(min_bp=80,
max_bp=120,
output_filename=fh.name,
progressbar=False)
assert len(f) == 250
ff = FastQ(fh.name)
assert len(ff) == 250
# keeps nothing
with TempFile() as fh:
f.filter(min_bp=80, max_bp=90, output_filename=fh.name)
assert len(f) == 250
ff = FastQ(fh.name)
assert len(ff) == 0
def test_compressor_running():
# Here we test gz -> bz2 -> gz -> dsrc -> gz using recursive or not
# get a fastq.gz in a temp file and process it
tempdir = tempfile.TemporaryDirectory()
filename = sequana_data("test.fastq.gz")
shutil.copy(filename, tempdir.name)
cwd = os.path.abspath(os.curdir)
os.chdir(tempdir.name)
# We concert gz -> bz2 -> gz and must get the exact same
# However, since the compression is not deterministic, we should compare the
# content of the uncompressed file (input and output)
try:
# seems to fail on travis with a subprocess issue
# https://travis-ci.org/sequana/sequana/builds/162466158
compressor.main(
[prog, "--source", "fastq.gz", "--target", "fastq.bz2", "--quiet"])
compressor.main([
prog, "--source", "fastq.bz2", "--target", "fastq.gz",
"--recursive", "--quiet"
])
compressor.main([
prog, "--source", "fastq.gz", "--target", "fastq.dsrc",
"--recursive", "--quiet"
])
compressor.main([
prog, "--source", "fastq.dsrc", "--target", "fastq.gz", "--quiet"
])
except Exception as err:
raise Exception(err)
finally:
os.chdir(cwd)
f1 = FastQ(filename)
f2 = FastQ(tempdir.name + os.sep + os.path.basename(filename))
assert f1 == f2
def __init__(self, directory=".", prefix="job-*"):
self.prefix = prefix
self.directory = directory
# low quality isoforms
self.lq_isoforms = self.get_file("lq_isoforms.fastq")
if self.lq_isoforms:
self.lq_sequence = FastQ(self.lq_isoforms)
# high quality isoforms
self.hq_isoforms = self.get_file("hq_isoforms.fastq")
if self.hq_isoforms:
self.hq_sequence = FastQ(self.hq_isoforms)
# General info
self.csv = self.get_file("-file.csv")
if self.csv:
self.data = pd.read_csv(self.csv)
# CCS fasta sequence
#self.ccs = self.get_file("-ccs.tar.gz")
self.ccs = self.get_file("ccs.fasta", noprefix=True)
if self.ccs:
self.ccs = FastA(self.ccs)
def test_bam(tmpdir):
datatest = sequana_data("test.bam", "testing")
s = BAM(datatest)
assert len(s) == 1000
assert s.is_sorted is True
df = s.get_df_concordance()
assert s.is_paired is True
assert int(df.length.sum()) == 67938
assert int(df.M.sum()) == 67788
df = s.get_df()
# call this here before other computations on purpose
with TempFile(suffix=".json") as fh:
s.bam_analysis_to_json(fh.name)
assert s.get_read_names()
s.get_mapped_read_length()
s.get_stats()
s.get_stats_full()
s.get_samtools_stats_as_df()
with TempFile() as fh:
s.to_fastq(fh.name)
from sequana import FastQ
ff = FastQ(fh.name)
len(ff) == len(s)
# plotting
with TempFile(suffix='.png') as fh:
s.plot_bar_flags(filename=fh.name, logy=True)
s.plot_bar_flags(filename=fh.name)
with TempFile(suffix='.png') as fh:
s.plot_bar_mapq(filename=fh.name)
s.get_gc_content()
s.get_length_count()
s.plot_gc_content()
s.boxplot_qualities()
s.boxplot_qualities(max_sample=50)
try:
s.plot_gc_content(bins=[1, 2, 10])
assert False
except:
assert True
def test_bam(tmpdir):
s = BAM(datatest)
assert len(s) == 1000
assert s.is_sorted is True
assert len(list(s.iter_unmapped_reads())) == 2
s.reset()
assert len(list(s.iter_mapped_reads())) == 998
s.reset()
# call this here before other computations on purpose
with TempFile(suffix=".json") as fh:
s.bam_analysis_to_json(fh.name)
assert s.get_read_names()
s.get_mapped_read_length()
s.get_stats()
s.get_full_stats_as_df()
with TempFile(suffix='.png') as fh:
s.plot_bar_flags(filename=fh.name, logy=True)
s.plot_bar_flags(filename=fh.name)
with TempFile(suffix='.png') as fh:
s.plot_bar_mapq(filename=fh.name)
with TempFile() as fh:
s.to_fastq(fh.name)
from sequana import FastQ
ff = FastQ(fh.name)
len(ff) == len(s)
s.get_gc_content()
s.get_length_count()
s.plot_gc_content()
try:
s.plot_gc_content(bins=[1, 2, 10])
assert False
except:
assert True
def __init__(self, infile, references, outdir, mapper, threads=4):
self.infile = infile
self.references = references
self.outdir = outdir
self.threads = threads
if os.path.exists(outdir):
logger.info("using {} for output".format(outdir))
else:
os.mkdir(outdir)
# this may be used later on for other mapper or methodology
if mapper == "minimap2":
self.mapper_cmd = "minimap2 -x map-pb -t {} {} {} -a > {}"
elif mapper == "bwa":
self.mapper_cmd = "bwa mem -M -t {} {} {} > {}"
f = FastQ(self.infile)
self.L = len(f)
logger.info("Found {} reads in input FastQ file\n\n".format(self.L))
class IsoSeqQC(object):
"""
Use get_isoseq_files on smrtlink to get the proper files
iso = IsoSeqQC()
iso.hist_read_length_consensus_isoform() # histo CCS
iso.stats() # "CCS" key is equivalent to summary metrics in CCS report
todo: get CCS passes histogram . Where to get the info of passes ?
"""
def __init__(self, directory=".", prefix=""):
self.prefix = prefix
self.directory = directory
self.sample_name = "undefined"
# low quality isoforms
filename = "all.polished_lq.fastq"
self.lq_isoforms = self.get_file(filename)
if self.lq_isoforms:
logger.info("Reading {}".format(filename))
self.lq_sequence = FastQ(self.lq_isoforms)
# high quality isoforms
filename = "all.polished_hq.fastq"
self.hq_isoforms = self.get_file(filename)
if self.hq_isoforms:
logger.info("Reading {}".format(filename))
self.hq_sequence = FastQ(self.hq_isoforms)
# General info
filename = "file.csv"
self.csv = self.get_file(filename)
if self.csv:
logger.info("Reading {}".format(filename))
self.data = pd.read_csv(self.csv)
# CCS fasta sequence
#self.ccs = self.get_file("-ccs.tar.gz")
filename = "ccs.fasta"
self.ccs = self.get_file(filename, noprefix=True)
if self.ccs:
logger.info("Reading {}".format(filename))
self.ccs = FastA(self.ccs)
def get_file(self, tag, noprefix=False):
if noprefix:
filenames = glob.glob(self.directory + os.sep + tag)
else:
filenames = glob.glob(self.directory + os.sep + self.prefix + tag)
if len(filenames) == 1:
return filenames[0]
elif len(filenames) > 1:
print("Found several files ending in %s" % tag)
else:
print("No files matching %s" % tag)
return None
def stats(self):
results = {}
if self.data is not None:
logger.info("Reading strand")
results['strand'] = {
"+": sum(self.data.strand == "+"),
"-": sum(self.data.strand == "-"),
"?": sum(self.data.strand.isnull())
}
results['classification'] = {
"total_ccs_reads" : len(self.data),
"five_prime_reads" : int(self.data.fiveseen.sum()),
"three_prime_reads" : int(self.data.threeseen.sum()),
"chimera" : int(self.data.chimera.sum()),
"polyA_reads" : int(self.data.polyAseen.sum()),
}
if self.lq_isoforms:
logger.info("Reading LQ isoforms")
results['lq_isoform'] = self.lq_sequence.stats() # number of
if self.hq_isoforms:
logger.info("Reading HQ isoforms")
results['hq_isoform'] = self.hq_sequence.stats() # number of polished HQ isoform
if self.ccs:
seq = [ len(read.sequence) for read in self.ccs]
results["CCS"] = {
"mean_length" : pylab.mean(seq),
"number_ccs_bases" : sum(seq),
"number_ccs_reads" : len(seq)
}
self.idents_v = []
self.full_v = []
self.non_full_v = []
self.isoform_lengths = []
for read in self.lq_sequence:
ident, full, non_full, length = read['identifier'].decode().split(";")
self.idents_v.append(ident)
self.full_v.append(int(full.split("=")[1]))
self.non_full_v.append(int(non_full.split("=")[1]))
self.isoform_lengths.append(int(length.split("=")[1]))
return results
def to_summary(self, filename="sequana_summary_isoseq.json", data=None):
"""Save statistics into a JSON file
:param filename:
:param data: dictionary to save. If not provided, use :meth:`stats`
"""
from sequana.summary import Summary
if data is None:
data = self.stats()
Summary("isoseq",self.sample_name, data=data).to_json(filename)
def hist_read_length_consensus_isoform(self, mode="all", bins=80, rwidth=0.8,
align="left", fontsize=16, edgecolor="k", **kwargs):
"""
mode can be all, lq, hq
"""
pylab.clf()
L1 = [len(read['sequence']) for read in self.lq_sequence]
L2 = [len(read['sequence']) for read in self.hq_sequence]
if mode == "all":
L = L1 + L2
elif mode == "lq":
L = L1
else:
L = L2
Y, X, _ = pylab.hist(L, bins=bins, rwidth=rwidth, align=align,
ec=edgecolor, **kwargs)
pylab.gca().set_ylim(bottom=0)
pylab.gca().set_xlim(left=0)
pylab.xlabel("Read length", fontsize=fontsize)
pylab.ylabel("Number of reads", fontsize=fontsize)
pylab.grid()
ax_twin = pylab.gca().twinx()
shift = (X[1] - X[0]) / 2
ax_twin.plot(X[0:-1]- shift, len(L) - pylab.cumsum(Y), "k")
ax_twin.set_ylim(bottom=0)
pylab.ylabel("CDF", fontsize=fontsize)
pylab.title("Read length of Consensus isoforms reads")
def hist_average_quality(self, fontsize=16, bins=None):
"""
bins is from 0 to 94
"""
hq_qv = [pylab.mean([ord(X)-33 for X in read['quality'].decode()])
for read in self.hq_sequence]
lq_qv = [pylab.mean([ord(X) -33 for X in read['quality'].decode()])
for read in self.lq_sequence]
if bins is None:
bins = range(0,94)
Y1, X = np.histogram(hq_qv, bins=bins)
Y2, X = np.histogram(lq_qv, bins=bins)
pylab.bar(X[1:], Y1, width=1, label="HQ")
pylab.bar(X[1:], Y2, bottom=Y1, width=1, label="LQ")
pylab.xlim([0.5, 93.5])
pylab.xlabel("Isoform average QV")
pylab.ylabel("# Isoform")
pylab.legend(fontsize=fontsize)
ax = pylab.twinx()
N = np.sum(Y1+Y2)
ax.plot(X, [N] + list(N-np.cumsum(Y1+Y2)), "k")
class IsoSeqQC(object):
"""
Use get_isoseq_files on smrtlink to get the proper files
iso = IsoSeqQC()
iso.hist_read_length_consensus_isoform() # histo CCS
iso.stats() # "CCS" key is equivalent to summary metrics in CCS report
todo: get CCS passes histogram . Where to get the info of passes ?
"""
def __init__(self, directory=".", prefix=""):
self.prefix = prefix
self.directory = directory
self.sample_name = "undefined"
# low quality isoforms
filename = "all.polished_lq.fastq"
self.lq_isoforms = self.get_file(filename)
if self.lq_isoforms:
logger.info("Reading {}".format(filename))
self.lq_sequence = FastQ(self.lq_isoforms)
# high quality isoforms
filename = "all.polished_hq.fastq"
self.hq_isoforms = self.get_file(filename)
if self.hq_isoforms:
logger.info("Reading {}".format(filename))
self.hq_sequence = FastQ(self.hq_isoforms)
# General info
filename = "file.csv"
self.csv = self.get_file(filename)
if self.csv:
logger.info("Reading {}".format(filename))
self.data = pd.read_csv(self.csv)
# CCS fasta sequence
#self.ccs = self.get_file("-ccs.tar.gz")
filename = "ccs.fasta"
self.ccs = self.get_file(filename, noprefix=True)
if self.ccs:
logger.info("Reading {}".format(filename))
self.ccs = FastA(self.ccs)
def get_file(self, tag, noprefix=False):
if noprefix:
filenames = glob.glob(self.directory + os.sep + tag)
else:
filenames = glob.glob(self.directory + os.sep + self.prefix + tag)
if len(filenames) == 1:
return filenames[0]
elif len(filenames) > 1:
print("Found several files ending in %s" % tag)
else:
print("No files matching %s" % tag)
return None
def stats(self):
results = {}
if self.data is not None:
logger.info("Reading strand")
results['strand'] = {
"+": sum(self.data.strand == "+"),
"-": sum(self.data.strand == "-"),
"?": sum(self.data.strand.isnull())
}
results['classification'] = {
"total_ccs_reads" : len(self.data),
"five_prime_reads" : int(self.data.fiveseen.sum()),
"three_prime_reads" : int(self.data.threeseen.sum()),
"chimera" : int(self.data.chimera.sum()),
"polyA_reads" : int(self.data.polyAseen.sum()),
}
if self.lq_isoforms:
logger.info("Reading LQ isoforms")
results['lq_isoform'] = self.lq_sequence.stats() # number of
if self.hq_isoforms:
logger.info("Reading HQ isoforms")
results['hq_isoform'] = self.hq_sequence.stats() # number of polished HQ isoform
if self.ccs:
seq = [ len(read.sequence) for read in self.ccs]
results["CCS"] = {
"mean_length" : pylab.mean(seq),
"number_ccs_bases" : sum(seq),
"number_ccs_reads" : len(seq)
}
self.idents_v = []
self.full_v = []
self.non_full_v = []
self.isoform_lengths = []
for read in self.lq_sequence:
ident, full, non_full, length = read['identifier'].decode().split(";")
self.idents_v.append(ident)
self.full_v.append(int(full.split("=")[1]))
self.non_full_v.append(int(non_full.split("=")[1]))
self.isoform_lengths.append(int(length.split("=")[1]))
return results
def to_summary(self, filename="sequana_summary_isoseq.json", data=None):
"""Save statistics into a JSON file
:param filename:
:param data: dictionary to save. If not provided, use :meth:`stats`
"""
from sequana.summary import Summary
if data is None:
data = self.stats()
Summary("isoseq",self.sample_name, data=data).to_json(filename)
def hist_read_length_consensus_isoform(self, mode="all", bins=80, rwidth=0.8,
align="left", fontsize=16, edgecolor="k", **kwargs):
"""
mode can be all, lq, hq
"""
pylab.clf()
L1 = [len(read['sequence']) for read in self.lq_sequence]
L2 = [len(read['sequence']) for read in self.hq_sequence]
if mode == "all":
L = L1 + L2
elif mode == "lq":
L = L1
else:
L = L2
Y, X, _ = pylab.hist(L, bins=bins, rwidth=rwidth, align=align,
ec=edgecolor, **kwargs)
pylab.gca().set_ylim(bottom=0)
pylab.gca().set_xlim(left=0)
pylab.xlabel("Read length", fontsize=fontsize)
pylab.ylabel("Number of reads", fontsize=fontsize)
pylab.grid()
ax_twin = pylab.gca().twinx()
shift = (X[1] - X[0]) / 2
ax_twin.plot(X[0:-1]- shift, len(L) - pylab.cumsum(Y), "k")
ax_twin.set_ylim(bottom=0)
pylab.ylabel("CDF", fontsize=fontsize)
pylab.title("Read length of Consensus isoforms reads")
def hist_average_quality(self, fontsize=16, bins=None):
"""
bins is from 0 to 94
"""
hq_qv = [pylab.mean([ord(X)-33 for X in read['quality'].decode()])
for read in self.hq_sequence]
lq_qv = [pylab.mean([ord(X) -33 for X in read['quality'].decode()])
for read in self.lq_sequence]
if bins is None:
bins = range(0,94)
Y1, X = np.histogram(hq_qv, bins=bins)
Y2, X = np.histogram(lq_qv, bins=bins)
pylab.bar(X[1:], Y1, width=1, label="HQ")
pylab.bar(X[1:], Y2, bottom=Y1, width=1, label="LQ")
pylab.xlim([0.5, 93.5])
pylab.xlabel("Isoform average QV")
pylab.ylabel("# Isoform")
pylab.legend(fontsize=fontsize)
ax = pylab.twinx()
N = np.sum(Y1+Y2)
ax.plot(X, [N] + list(N-np.cumsum(Y1+Y2)), "k")
class IsoSeq(object):
"""
ftp://ftp-trace.ncbi.nih.gov/1000genomes/ftp/technical/reference/
"""
def __init__(self, directory=".", prefix="job-*"):
self.prefix = prefix
self.directory = directory
# low quality isoforms
self.lq_isoforms = self.get_file("lq_isoforms.fastq")
if self.lq_isoforms:
self.lq_sequence = FastQ(self.lq_isoforms)
# high quality isoforms
self.hq_isoforms = self.get_file("hq_isoforms.fastq")
if self.hq_isoforms:
self.hq_sequence = FastQ(self.hq_isoforms)
# General info
self.csv = self.get_file("-file.csv")
if self.csv:
self.data = pd.read_csv(self.csv)
# CCS fasta sequence
#self.ccs = self.get_file("-ccs.tar.gz")
self.ccs = self.get_file("ccs.fasta", noprefix=True)
if self.ccs:
self.ccs = FastA(self.ccs)
def get_file(self, tag, noprefix=False):
if noprefix:
filenames = glob.glob(self.directory + os.sep + tag)
else:
filenames = glob.glob(self.directory + os.sep + self.prefix + tag)
if len(filenames) == 1:
return filenames[0]
elif len(filenames) > 1:
print("Found several files ending in %s" % tag)
else:
print("No files matching %s" % tag)
return None
def stats(self):
results = {}
if self.data is not None:
print("Reading strand")
results['strand'] = {
"+": sum(self.data.strand == "+"),
"-": sum(self.data.strand == "-"),
"?": sum(self.data.strand.isnull())
}
results['classification'] = {
"total_ccs_reads": len(self.data),
"five_prime_reads": sum(self.data.fiveseen),
"three_prime_reads": sum(self.data.threeseen),
"polyA_reads": sum(self.data.polyAseen),
}
if self.lq_isoforms:
print("Reading LQ isoform")
results['lq_isoform'] = self.lq_sequence.stats() # number of
if self.hq_isoforms:
print("Reading HQ isoform")
results['hq_isoform'] = self.hq_sequence.stats(
) # number of polished HQ isoform
if self.ccs:
seq = [read.sequence for read in self.ccs]
results["CCS"] = {
"mean_length": pylab.mean(seq),
"number_ccs_bases": sum(seq),
"number_ccs_reads": len(seq)
}
self.idents_v = []
self.full_v = []
self.non_full_v = []
self.isoform_lengths = []
for read in self.lq_sequence:
ident, full, non_full, length = read['identifier'].decode().split(
";")
self.idents_v.append(ident)
self.full_v.append(int(full.split("=")[1]))
self.non_full_v.append(int(non_full.split("=")[1]))
self.isoform_lengths.append(int(length.split("=")[1]))
return results
def hist_read_length_consensus_isoform(self,
mode="all",
bins=80,
rwidth=0.8,
align="left",
fontsize=16,
edgecolor="k",
**kwargs):
"""
mode can be all, lq, hq
"""
pylab.clf()
L1 = [len(read['sequence']) for read in self.lq_sequence]
L2 = [len(read['sequence']) for read in self.hq_sequence]
if mode == "all":
L = L1 + L2
elif mode == "lq":
L = L1
else:
L = L2
Y, X, _ = pylab.hist(L,
bins=bins,
rwidth=rwidth,
align=align,
ec=edgecolor,
**kwargs)
pylab.gca().set_ylim(bottom=0)
pylab.gca().set_xlim(left=0)
pylab.xlabel("Read length", fontsize=fontsize)
pylab.ylabel("Number of reads", fontsize=fontsize)
pylab.grid()
ax_twin = pylab.gca().twinx()
shift = (X[1] - X[0]) / 2
ax_twin.plot(X[1:] - shift, len(L) - pylab.cumsum(Y), "k")
ax_twin.set_ylim(bottom=0)
pylab.ylabel("CDF", fontsize=fontsize)
pylab.title("Read length of Consensus isoforms reads")
def hist_average_quality(self, fontsize=16):
hq_qv = [
mean([phred.ascii_to_quality(X) for X in read['quality'].decode()])
for read in iso.hq_sequence
]
lq_qv = [
mean([phred.ascii_to_quality(X) for X in read['quality'].decode()])
for read in iso.lq_sequence
]
Y1, X = numpy.histogram(hq_qv, bins=range(0, 94))
Y2, X = numpy.histogram(lq_qv, bins=range(0, 94))
pylab.bar(X[1:], Y1, width=1, label="HQ")
pylab.bar(X[1:], Y2, bottom=Y1, width=1, label="LQ")
pylab.xlabel("Isoform average QV")
pylab.ylabel("# Isoform")
pylab.legend(fontsize=fontsize)
