def pairs(args):
"""
%prog pairs folder reference.fasta
Estimate insert size distribution. Compatible with a variety of aligners,
including BOWTIE and BWA.
"""
p = OptionParser(pairs.__doc__)
p.set_firstN()
p.set_mates()
p.set_aligner()
opts, args = p.parse_args(args)
if len(args) != 2:
sys.exit(not p.print_help())
cwd = os.getcwd()
aligner = opts.aligner
work = "-".join(("pairs", aligner))
mkdir(work)
from jcvi.formats.sam import pairs as ps
if aligner == "bowtie":
from jcvi.apps.bowtie import align
elif aligner == "bwa":
from jcvi.apps.bwa import align
folder, ref = args
ref = get_abs_path(ref)
messages = []
for p, prefix in iter_project(folder):
samplefq = []
for i in range(2):
samplefq.append(
op.join(work, prefix + "_{0}.first.fastq".format(i + 1)))
first([str(opts.firstN)] + [p[i]] + ["-o", samplefq[i]])
os.chdir(work)
align_args = [ref] + [op.basename(fq) for fq in samplefq]
outfile, logfile = align(align_args)
bedfile, stats = ps([outfile, "--rclip={0}".format(opts.rclip)])
os.chdir(cwd)
median = stats.median
tag = "MP" if median > 1000 else "PE"
median = str(median)
pf, sf = median[:2], median[2:]
if sf and int(sf) != 0:
pf = str(int(pf) + 1) # Get the first two effective digits
lib = "{0}-{1}".format(tag, pf + "0" * len(sf))
for i, xp in enumerate(p):
suffix = "fastq.gz" if xp.endswith(".gz") else "fastq"
link = "{0}-{1}.{2}.{3}".format(lib, prefix.replace("-", ""),
i + 1, suffix)
m = "\t".join(str(x) for x in (xp, link))
messages.append(m)
messages = "\n".join(messages)
write_file("f.meta", messages, tee=True)
def pairs(args):
"""
%prog pairs folder reference.fasta
Estimate insert size distribution. Compatible with a variety of aligners,
including CLC, BOWTIE and BWA.
"""
p = OptionParser(pairs.__doc__)
p.set_firstN()
p.set_mates()
p.set_aligner()
opts, args = p.parse_args(args)
if len(args) != 2:
sys.exit(not p.print_help())
cwd = os.getcwd()
aligner = opts.aligner
work = "-".join(("pairs", aligner))
mkdir(work)
if aligner == "clc":
from jcvi.apps.clc import align
from jcvi.formats.cas import pairs as ps
else:
from jcvi.formats.sam import pairs as ps
if aligner == "bowtie":
from jcvi.apps.bowtie import align
elif aligner == "bwa":
from jcvi.apps.bwa import align
folder, ref = args
ref = get_abs_path(ref)
messages = []
for p, prefix in iter_project(folder, 2):
samplefq = op.join(work, prefix + ".first.fastq")
first([str(opts.firstN)] + p + ["-o", samplefq])
os.chdir(work)
align_args = [ref, op.basename(samplefq)]
outfile, logfile = align(align_args)
bedfile, stats = ps([outfile, "--rclip={0}".format(opts.rclip)])
os.chdir(cwd)
median = stats.median
tag = "MP" if median > 1000 else "PE"
median = str(median)
pf, sf = median[:2], median[2:]
if sf and int(sf) != 0:
pf = str(int(pf) + 1) # Get the first two effective digits
lib = "{0}-{1}".format(tag, pf + "0" * len(sf))
for i, xp in enumerate(p):
suffix = "fastq.gz" if xp.endswith(".gz") else "fastq"
link = "{0}-{1}.{2}.{3}".format(lib, prefix.replace("-", ""), i + 1, suffix)
m = "\t".join(str(x) for x in (xp, link))
messages.append(m)
messages = "\n".join(messages)
write_file("f.meta", messages, tee=True)
def mates(args):
"""
%prog mates bedfile
Generate the mates file by inferring from the names.
"""
p = OptionParser(mates.__doc__)
p.add_option(
"--lib",
default=False,
action="store_true",
help="Output library information along with pairs [default: %default]")
p.add_option(
"--nointra",
default=False,
action="store_true",
help="Remove mates that are intra-scaffold [default: %default]")
p.add_option(
"--prefix",
default=False,
action="store_true",
help="Only keep links between IDs with same prefix [default: %default]"
)
p.set_mates()
opts, args = p.parse_args(args)
if len(args) != 1:
sys.exit(not p.print_help())
bedfile, = args
rclip = opts.rclip
key = (lambda x: x.accn[:-rclip]) if rclip else (lambda x: x.accn)
bed = Bed(bedfile, key=key)
pf = bedfile.rsplit(".", 1)[0]
matesfile = pf + ".mates"
lib = pf if opts.lib else None
fw = open(matesfile, "w")
if lib:
bedfile, stats = pairs([bedfile, \
"--rclip={0}".format(rclip),
"--cutoff={0}".format(opts.cutoff)])
sv = int(2 * stats.sd)
mindist = max(stats.mean - sv, 1)
maxdist = stats.mean + sv
print >> fw, "\t".join(str(x) for x in \
("library", pf, mindist, maxdist))
num_fragments = num_pairs = 0
matesbedfile = matesfile + ".bed"
fwm = open(matesbedfile, "w")
for pe, lines in groupby(bed, key=key):
lines = list(lines)
if len(lines) != 2:
num_fragments += len(lines)
continue
a, b = lines
if opts.nointra and a.seqid == b.seqid:
continue
# Use --prefix to limit the links between seqids with the same prefix
# For example, contigs of the same BAC, mth2-23j10_001, mth-23j10_002
if opts.prefix:
aprefix = a.seqid.split("_")[0]
bprefix = b.seqid.split("_")[0]
if aprefix != bprefix:
continue
num_pairs += 1
pair = [a.accn, b.accn]
if lib:
pair.append(lib)
print >> fw, "\t".join(pair)
print >> fwm, a
print >> fwm, b
logging.debug("Discard {0} frags and write {1} pairs to `{2}` and `{3}`.".\
format(num_fragments, num_pairs, matesfile, matesbedfile))
fw.close()
fwm.close()
return matesfile, matesbedfile
def mates(args):
"""
%prog mates bedfile
Generate the mates file by inferring from the names.
"""
p = OptionParser(mates.__doc__)
p.add_option("--lib", default=False, action="store_true",
help="Output library information along with pairs [default: %default]")
p.add_option("--nointra", default=False, action="store_true",
help="Remove mates that are intra-scaffold [default: %default]")
p.add_option("--prefix", default=False, action="store_true",
help="Only keep links between IDs with same prefix [default: %default]")
p.set_mates()
opts, args = p.parse_args(args)
if len(args) != 1:
sys.exit(not p.print_help())
bedfile, = args
rclip = opts.rclip
key = (lambda x: x.accn[:-rclip]) if rclip else (lambda x: x.accn)
bed = Bed(bedfile, key=key)
pf = bedfile.rsplit(".", 1)[0]
matesfile = pf + ".mates"
lib = pf if opts.lib else None
fw = open(matesfile, "w")
if lib:
bedfile, stats = pairs([bedfile, \
"--rclip={0}".format(rclip),
"--cutoff={0}".format(opts.cutoff)])
sv = int(2 * stats.sd)
mindist = max(stats.mean - sv, 1)
maxdist = stats.mean + sv
print >> fw, "\t".join(str(x) for x in \
("library", pf, mindist, maxdist))
num_fragments = num_pairs = 0
matesbedfile = matesfile + ".bed"
fwm = open(matesbedfile, "w")
for pe, lines in groupby(bed, key=key):
lines = list(lines)
if len(lines) != 2:
num_fragments += len(lines)
continue
a, b = lines
if opts.nointra and a.seqid == b.seqid:
continue
# Use --prefix to limit the links between seqids with the same prefix
# For example, contigs of the same BAC, mth2-23j10_001, mth-23j10_002
if opts.prefix:
aprefix = a.seqid.split("_")[0]
bprefix = b.seqid.split("_")[0]
if aprefix != bprefix:
continue
num_pairs += 1
pair = [a.accn, b.accn]
if lib:
pair.append(lib)
print >> fw, "\t".join(pair)
print >> fwm, a
print >> fwm, b
logging.debug("Discard {0} frags and write {1} pairs to `{2}` and `{3}`.".\
format(num_fragments, num_pairs, matesfile, matesbedfile))
fw.close()
fwm.close()
return matesfile, matesbedfile
def link(args):
"""
%prog link bedfile fastafile
Construct contig links based on bed file. Use --prefix to limit the links
between contigs that start with the same prefix_xxx.
"""
p = OptionParser(link.__doc__)
p.set_mates(rclip=1, mateorientation="+-")
p.add_option("--insert", type="int", default=0, help="Mean insert size [default: estimate from data]")
p.add_option(
"--prefix",
default=False,
action="store_true",
help="Only keep links between IDs with same prefix [default: %default]",
)
p.add_option(
"--debug",
dest="debug",
default=False,
action="store_true",
help="Print verbose info when checking mates [default: %default]",
)
opts, args = p.parse_args(args)
if len(args) != 2:
sys.exit(not p.print_help())
bedfile, fastafile = args
debug = opts.debug
cutoff = opts.cutoff
sizes = Sizes(fastafile)
cutoffopt = "--cutoff={0}".format(cutoff)
mateorientationopt = "--mateorientation={0}".format(opts.mateorientation)
bedfile, stats = pairs([bedfile, cutoffopt, mateorientationopt, "--rclip={0}".format(opts.rclip)])
maxcutoff = cutoff or stats.p2
insert = opts.insert or stats.median
logging.debug("Mate hangs must be <= {0}, --cutoff to override".format(maxcutoff))
rs = lambda x: x.accn[:-1]
fp = open(bedfile)
linksfile = bedfile.rsplit(".", 1)[0] + ".links"
fw = open(linksfile, "w")
for a, b in pairwise(fp):
"""
Criteria for valid contig edge
1. for/rev do not mapping to the same scaffold (useful for linking)
2. assuming innie (outie must be flipped first), order the contig pair
3. calculate sequence hangs, valid hangs are smaller than insert size
"""
a, b = BedLine(a), BedLine(b)
if rs(a) != rs(b):
continue
pe = rs(a)
# Intra-contig links
if a.seqid == b.seqid:
continue
# Use --prefix to limit the links between seqids with the same prefix
# For example, contigs of the same BAC, mth2-23j10_001, mth-23j10_002
if opts.prefix:
aprefix = a.seqid.split("_")[0]
bprefix = b.seqid.split("_")[0]
if aprefix != bprefix:
continue
cl = ContigLink(a, b, insert=insert, cutoff=maxcutoff)
if cl.flip_innie(sizes, debug=debug):
print >> fw, "\t".join((pe, str(cl)))
