def phase(accession):
gbdir = "gb"
gbfile = op.join(gbdir, accession + ".gb")
if not op.exists(gbfile):
entrez([accession, "--skipcheck", "--outdir=" + gbdir, "--format=gb"])
rec = SeqIO.parse(gbfile, "gb").next()
ph, keywords = get_phase(rec)
return ph, len(rec)
def phase(accession):
gbdir = "gb"
gbfile = op.join(gbdir, accession + ".gb")
if not op.exists(gbfile):
entrez([accession, "--skipcheck", "--outdir=" + gbdir, "--format=gb"])
rec = next(SeqIO.parse(gbfile, "gb"))
ph, keywords = get_phase(rec)
return ph, len(rec)
def _get_records(self):
gbdir = "gb"
dirmade = mkdir(gbdir)
if not dirmade:
sh("rm -rf {0}_old; mv -f {0} {0}_old".format(gbdir,))
assert mkdir(gbdir)
entrez([self.idfile, "--format=gb", "--database=nuccore", "--outdir={0}"\
.format(gbdir)])
logging.debug('GenBank records written to {0}.'.format(gbdir))
return gbdir
def bes(args):
"""
%prog bes bacfasta clonename
Use the clone name to download BES gss sequences from Genbank, map and then
visualize.
"""
from jcvi.apps.align import run_blat
p = OptionParser(bes.__doc__)
p.set_cpus()
opts, args = p.parse_args(args)
if len(args) != 2:
sys.exit(not p.print_help())
bacfasta, clonename = args
entrez([clonename, "--database=nucgss", "--skipcheck"])
besfasta = clonename + ".fasta"
blatfile = clonename + ".bes.blat"
run_blat(
infile=besfasta,
outfile=blatfile,
db=bacfasta,
pctid=95,
hitlen=100,
cpus=opts.cpus,
)
aid, asize = next(Fasta(bacfasta).itersizes())
width = 50
msg = "=" * width
msg += " " + aid
print(msg, file=sys.stderr)
ratio = width * 1.0 / asize
_ = lambda x: int(round(x * ratio, 0))
blasts = [BlastLine(x) for x in open(blatfile)]
for b in blasts:
if b.orientation == "+":
msg = " " * _(b.sstart) + "->"
else:
msg = " " * (_(b.sstop) - 2) + "<-"
msg += " " * (width - len(msg) + 2)
msg += b.query
if b.orientation == "+":
msg += " (hang={0})".format(b.sstart - 1)
else:
msg += " (hang={0})".format(asize - b.sstop)
print(msg, file=sys.stderr)
def _get_records(self):
gbdir = "gb"
dirmade = mkdir(gbdir)
if not dirmade:
sh("rm -rf {0}_old; mv -f {0} {0}_old".format(gbdir,))
assert mkdir(gbdir)
entrez([self.idfile, "--format=gb", "--database=nuccore", "--outdir={0}"\
.format(gbdir)])
logging.debug('GenBank records written to {0}.'.format(gbdir))
return gbdir
def blast(args):
"""
%prog blast allfasta clonename
Insert a component into agpfile by aligning to the best hit in pool and see
if they have good overlaps.
"""
from jcvi.apps.align import run_megablast
p = OptionParser(blast.__doc__)
p.add_option("-n",
type="int",
default=2,
help="Take best N hits [default: %default]")
opts, args = p.parse_args(args)
if len(args) != 2:
sys.exit(not p.print_help())
allfasta, clonename = args
fastadir = "fasta"
infile = op.join(fastadir, clonename + ".fasta")
if not op.exists(infile):
entrez([clonename, "--skipcheck", "--outdir=" + fastadir])
outfile = "{0}.{1}.blast".format(clonename, allfasta.split(".")[0])
run_megablast(infile=infile, outfile=outfile, db=allfasta, \
pctid=GoodPct, hitlen=GoodOverlap)
blasts = [BlastLine(x) for x in open(outfile)]
besthits = []
for b in blasts:
if b.query.count("|") >= 3:
b.query = b.query.split("|")[3]
if b.subject.count("|") >= 3:
b.subject = b.subject.split("|")[3]
b.query = b.query.rsplit(".", 1)[0]
b.subject = b.subject.rsplit(".", 1)[0]
if b.query == b.subject:
continue
if b.subject not in besthits:
besthits.append(b.subject)
if len(besthits) == opts.n:
break
for b in besthits:
overlap([clonename, b, "--dir=" + fastadir])
def blast(args):
"""
%prog blast allfasta clonename
Insert a component into agpfile by aligning to the best hit in pool and see
if they have good overlaps.
"""
from jcvi.apps.align import run_megablast
p = OptionParser(blast.__doc__)
p.add_option("-n", type="int", default=2,
help="Take best N hits [default: %default]")
opts, args = p.parse_args(args)
if len(args) != 2:
sys.exit(not p.print_help())
allfasta, clonename = args
fastadir = "fasta"
infile = op.join(fastadir, clonename + ".fasta")
if not op.exists(infile):
entrez([clonename, "--skipcheck", "--outdir=" + fastadir])
outfile = "{0}.{1}.blast".format(clonename, allfasta.split(".")[0])
run_megablast(infile=infile, outfile=outfile, db=allfasta, \
pctid=GoodPct, hitlen=GoodOverlap)
blasts = [BlastLine(x) for x in open(outfile)]
besthits = []
for b in blasts:
if b.query.count("|") >= 3:
b.query = b.query.split("|")[3]
if b.subject.count("|") >= 3:
b.subject = b.subject.split("|")[3]
b.query = b.query.rsplit(".", 1)[0]
b.subject = b.subject.rsplit(".", 1)[0]
if b.query == b.subject:
continue
if b.subject not in besthits:
besthits.append(b.subject)
if len(besthits) == opts.n:
break
for b in besthits:
overlap([clonename, b, "--dir=" + fastadir])
def bes(args):
"""
%prog bes bacfasta clonename
Use the clone name to download BES gss sequences from Genbank, map and then
visualize.
"""
from jcvi.apps.align import run_blat
p = OptionParser(bes.__doc__)
p.set_cpus()
opts, args = p.parse_args(args)
if len(args) != 2:
sys.exit(not p.print_help())
bacfasta, clonename = args
entrez([clonename, "--database=nucgss", "--skipcheck"])
besfasta = clonename + ".fasta"
blatfile = clonename + ".bes.blat"
run_blat(infile=besfasta, outfile=blatfile, db=bacfasta, \
pctid=95, hitlen=100, cpus=opts.cpus)
aid, asize = Fasta(bacfasta).itersizes().next()
width = 50
msg = "=" * width
msg += " " + aid
print >> sys.stderr, msg
ratio = width * 1. / asize
_ = lambda x: int(round(x * ratio, 0))
blasts = [BlastLine(x) for x in open(blatfile)]
for b in blasts:
if b.orientation == '+':
msg = " " * _(b.sstart) + "->"
else:
msg = " " * (_(b.sstop) - 2) + "<-"
msg += " " * (width - len(msg) + 2)
msg += b.query
if b.orientation == '+':
msg += " (hang={0})".format(b.sstart - 1)
else:
msg += " (hang={0})".format(asize - b.sstop)
print >> sys.stderr, msg
def certificate(args):
"""
%prog certificate tpffile certificatefile
Generate certificate file for all overlaps in tpffile. tpffile can be
generated by jcvi.formats.agp.tpf().
North chr1 2 0 AC229737.8 telomere 58443
South chr1 2 1 AC229737.8 AC202463.29 58443 37835 58443 + Non-terminal
Each line describes a relationship between the current BAC and the
north/south BAC. First, "North/South" tag, then the chromosome, phases of
the two BACs, ids of the two BACs, the size and the overlap start-stop of
the CURRENT BAC, and orientation. Each BAC will have two lines in the
certificate file.
"""
p = OptionParser(certificate.__doc__)
opts, args = p.parse_args(args)
if len(args) != 2:
sys.exit(not p.print_help())
tpffile, certificatefile = args
fastadir = "fasta"
tpf = TPF(tpffile)
data = check_certificate(certificatefile)
fw = must_open(certificatefile, "w")
for i, a in enumerate(tpf):
if a.is_gap:
continue
aid = a.component_id
af = op.join(fastadir, aid + ".fasta")
if not op.exists(af): # Check to avoid redownload
entrez([aid, "--skipcheck", "--outdir=" + fastadir])
north, south = tpf.getNorthSouthClone(i)
aphase, asize = phase(aid)
for tag, p in (("North", north), ("South", south)):
if not p: # end of the chromosome
ov = "telomere\t{0}".format(asize)
elif p.isCloneGap:
bphase = "0"
ov = "{0}\t{1}".format(p.gap_type, asize)
else:
bid = p.component_id
bphase, bsize = phase(bid)
key = (tag, aid, bid)
if key in data:
print >> fw, data[key]
continue
ar = [aid, bid, "--dir=" + fastadir]
o = overlap(ar)
ov = o.certificateline if o \
else "{0}\t{1}\tNone".format(bid, asize)
print >> fw, "\t".join(str(x) for x in \
(tag, a.object, aphase, bphase, aid, ov))
fw.flush()
def overlap(args):
"""
%prog overlap <a|a.fasta> <b|b.fasta>
Check overlaps between two fasta records. The arguments can be genBank IDs
instead of FASTA files. In case of IDs, the sequences will be downloaded
first.
"""
from jcvi.formats.blast import chain_HSPs
p = OptionParser(overlap.__doc__)
p.add_option("--dir", default=os.getcwd(),
help="Download sequences to dir [default: %default]")
p.add_option("--suffix", default="fasta",
help="Suffix of the sequence file in dir [default: %default]")
p.add_option("--qreverse", default=False, action="store_true",
help="Reverse seq a [default: %default]")
p.add_option("--nochain", default=False, action="store_true",
help="Do not chain adjacent HSPs [default: chain HSPs]")
p.set_align(pctid=GoodPct, hitlen=GoodOverlap, evalue=.01)
p.set_outfile(outfile=None)
opts, args = p.parse_args(args)
if len(args) != 2:
sys.exit(not p.print_help())
afasta, bfasta = args
dir = opts.dir
chain = not opts.nochain
suffix = opts.suffix
evalue = opts.evalue
pctid = opts.pctid
hitlen = opts.hitlen
cutoff = Cutoff(pctid, hitlen)
# Check first whether it is file or accession name
if not op.exists(afasta):
af = op.join(dir, ".".join((afasta, suffix)))
if not op.exists(af): # Check to avoid redownload
entrez([afasta, "--skipcheck", "--outdir=" + dir])
afasta = af
if not op.exists(bfasta):
bf = op.join(dir, ".".join((bfasta, suffix)))
if not op.exists(bf):
entrez([bfasta, "--skipcheck", "--outdir=" + dir])
bfasta = bf
assert op.exists(afasta) and op.exists(bfasta)
cmd = "blastn -dust no"
cmd += " -query {0} -subject {1}".format(afasta, bfasta)
cmd += " -evalue {0} -outfmt 6 -perc_identity {1}".format(evalue, pctid)
fp = popen(cmd)
hsps = fp.readlines()
hsps = [BlastLine(x) for x in hsps]
hsps = [x for x in hsps if x.hitlen >= hitlen]
if chain:
logging.debug("Chain HSPs in the Blast output.")
dist = 2 * hitlen # Distance to chain the HSPs
hsps = chain_HSPs(hsps, xdist=dist, ydist=dist)
if len(hsps) == 0:
print >> sys.stderr, "No match found."
return None
besthsp = hsps[0]
aid, asize = Fasta(afasta).itersizes().next()
bid, bsize = Fasta(bfasta).itersizes().next()
o = Overlap(besthsp, asize, bsize, cutoff, qreverse=opts.qreverse)
o.print_graphic()
if opts.outfile:
fw = must_open(opts.outfile, "w")
print >> fw, str(o)
fw.close()
return o
def htg(args):
"""
%prog htg fastafile template.sbt
Prepare sqnfiles for Genbank HTG submission to update existing records.
`fastafile` contains the records to update, multiple records are allowed
(with each one generating separate sqn file in the sqn/ folder). The record
defline has the accession ID. For example,
>AC148290.3
Internally, this generates two additional files (phasefile and namesfile)
and download records from Genbank. Below is implementation details:
`phasefile` contains, for each accession, phase information. For example:
AC148290.3 3 HTG 2 mth2-45h12
which means this is a Phase-3 BAC. Record with only a single contig will be
labeled as Phase-3 regardless of the info in the `phasefile`. Template file
is the Genbank sbt template. See jcvi.formats.sbt for generation of such
files.
Another problem is that Genbank requires the name of the sequence to stay
the same when updating and will kick back with a table of name conflicts.
For example:
We are unable to process the updates for these entries
for the following reason:
Seqname has changed
Accession Old seq_name New seq_name
--------- ------------ ------------
AC239792 mtg2_29457 AC239792.1
To prepare a submission, this script downloads genbank and asn.1 format,
and generate the phase file and the names file (use formats.agp.phase() and
apps.gbsubmit.asn(), respectively). These get automatically run.
However, use --phases if the genbank files contain outdated information.
For example, the clone name changes or phase upgrades. In this case, run
formats.agp.phase() manually, modify the phasefile and use --phases to override.
"""
from jcvi.formats.fasta import sequin, ids
from jcvi.formats.agp import phase
from jcvi.apps.fetch import entrez
p = OptionParser(htg.__doc__)
p.add_option("--phases", default=None,
help="Use another phasefile to override [default: %default]")
p.add_option("--comment", default="",
help="Comments for this update [default: %default]")
opts, args = p.parse_args(args)
if len(args) != 2:
sys.exit(not p.print_help())
fastafile, sbtfile = args
pf = fastafile.rsplit(".", 1)[0]
idsfile = pf + ".ids"
phasefile = pf + ".phases"
namesfile = pf + ".names"
ids([fastafile, "--outfile={0}".format(idsfile)])
asndir = "asn.1"
mkdir(asndir)
entrez([idsfile, "--format=asn.1", "--outdir={0}".format(asndir)])
asn(glob("{0}/*".format(asndir)) + \
["--outfile={0}".format(namesfile)])
if opts.phases is None:
gbdir = "gb"
mkdir(gbdir)
entrez([idsfile, "--format=gb", "--outdir={0}".format(gbdir)])
phase(glob("{0}/*".format(gbdir)) + \
["--outfile={0}".format(phasefile)])
else:
phasefile = opts.phases
assert op.exists(namesfile) and op.exists(phasefile)
newphasefile = phasefile + ".new"
newphasefw = open(newphasefile, "w")
comment = opts.comment
fastadir = "fasta"
sqndir = "sqn"
mkdir(fastadir)
mkdir(sqndir)
from jcvi.graphics.histogram import stem_leaf_plot
names = DictFile(namesfile)
assert len(set(names.keys())) == len(set(names.values()))
phases = DictFile(phasefile)
ph = [int(x) for x in phases.values()]
# vmin 1, vmax 4, bins 3
stem_leaf_plot(ph, 1, 4, 3, title="Counts of phases before updates")
logging.debug("Information loaded for {0} records.".format(len(phases)))
assert len(names) == len(phases)
newph = []
cmd = "faSplit byname {0} {1}/".format(fastafile, fastadir)
sh(cmd, outfile="/dev/null", errfile="/dev/null")
acmd = 'tbl2asn -a z -p fasta -r {sqndir}'
acmd += ' -i {splitfile} -t {sbtfile} -C tigr'
acmd += ' -j "{qualifiers}"'
acmd += ' -A {accession_nv} -o {sqndir}/{accession_nv}.sqn -V Vbr'
acmd += ' -y "{comment}" -W T -T T'
qq = "[tech=htgs {phase}] [organism=Medicago truncatula] [strain=A17]"
nupdated = 0
for row in open(phasefile):
atoms = row.rstrip().split("\t")
# see formats.agp.phase() for column contents
accession, phase, clone = atoms[0], atoms[1], atoms[-1]
fafile = op.join(fastadir, accession + ".fa")
accession_nv = accession.split(".", 1)[0]
newid = names[accession_nv]
newidopt = "--newid={0}".format(newid)
cloneopt = "--clone={0}".format(clone)
splitfile, gaps = sequin([fafile, newidopt, cloneopt])
splitfile = op.basename(splitfile)
phase = int(phase)
assert phase in (1, 2, 3)
oldphase = phase
if gaps == 0 and phase != 3:
phase = 3
if gaps != 0 and phase == 3:
phase = 2
print("{0}\t{1}\t{2}".\
format(accession_nv, oldphase, phase), file=newphasefw)
newph.append(phase)
qualifiers = qq.format(phase=phase)
if ";" in clone:
qualifiers += " [keyword=HTGS_POOLED_MULTICLONE]"
cmd = acmd.format(accession=accession, accession_nv=accession_nv,
sqndir=sqndir, sbtfile=sbtfile, splitfile=splitfile,
qualifiers=qualifiers, comment=comment)
sh(cmd)
verify_sqn(sqndir, accession)
nupdated += 1
stem_leaf_plot(newph, 1, 4, 3, title="Counts of phases after updates")
print("A total of {0} records updated.".format(nupdated), file=sys.stderr)
def overlap(args):
"""
%prog overlap <a|a.fasta> <b|b.fasta>
Check overlaps between two fasta records. The arguments can be genBank IDs
instead of FASTA files. In case of IDs, the sequences will be downloaded
first.
"""
from jcvi.formats.blast import chain_HSPs
p = OptionParser(overlap.__doc__)
p.add_option("--dir",
default=os.getcwd(),
help="Download sequences to dir [default: %default]")
p.add_option("--suffix",
default="fasta",
help="Suffix of the sequence file in dir [default: %default]")
p.add_option("--qreverse",
default=False,
action="store_true",
help="Reverse seq a [default: %default]")
p.add_option("--nochain",
default=False,
action="store_true",
help="Do not chain adjacent HSPs [default: chain HSPs]")
p.set_align(pctid=GoodPct, hitlen=GoodOverlap, evalue=.01)
p.set_outfile(outfile=None)
opts, args = p.parse_args(args)
if len(args) != 2:
sys.exit(not p.print_help())
afasta, bfasta = args
dir = opts.dir
chain = not opts.nochain
suffix = opts.suffix
evalue = opts.evalue
pctid = opts.pctid
hitlen = opts.hitlen
cutoff = Cutoff(pctid, hitlen)
# Check first whether it is file or accession name
if not op.exists(afasta):
af = op.join(dir, ".".join((afasta, suffix)))
if not op.exists(af): # Check to avoid redownload
entrez([afasta, "--skipcheck", "--outdir=" + dir])
afasta = af
if not op.exists(bfasta):
bf = op.join(dir, ".".join((bfasta, suffix)))
if not op.exists(bf):
entrez([bfasta, "--skipcheck", "--outdir=" + dir])
bfasta = bf
assert op.exists(afasta) and op.exists(bfasta)
cmd = "blastn -dust no"
cmd += " -query {0} -subject {1}".format(afasta, bfasta)
cmd += " -evalue {0} -outfmt 6 -perc_identity {1}".format(evalue, pctid)
fp = popen(cmd)
hsps = fp.readlines()
hsps = [BlastLine(x) for x in hsps]
hsps = [x for x in hsps if x.hitlen >= hitlen]
if chain:
logging.debug("Chain HSPs in the Blast output.")
dist = 2 * hitlen # Distance to chain the HSPs
hsps = chain_HSPs(hsps, xdist=dist, ydist=dist)
if len(hsps) == 0:
print >> sys.stderr, "No match found."
return None
besthsp = hsps[0]
aid, asize = Fasta(afasta).itersizes().next()
bid, bsize = Fasta(bfasta).itersizes().next()
o = Overlap(besthsp, asize, bsize, cutoff, qreverse=opts.qreverse)
o.print_graphic()
if opts.outfile:
fw = must_open(opts.outfile, "w")
print >> fw, str(o)
fw.close()
return o
def certificate(args):
"""
%prog certificate tpffile certificatefile
Generate certificate file for all overlaps in tpffile. tpffile can be
generated by jcvi.formats.agp.tpf().
North chr1 2 0 AC229737.8 telomere 58443
South chr1 2 1 AC229737.8 AC202463.29 58443 37835 58443 + Non-terminal
Each line describes a relationship between the current BAC and the
north/south BAC. First, "North/South" tag, then the chromosome, phases of
the two BACs, ids of the two BACs, the size and the overlap start-stop of
the CURRENT BAC, and orientation. Each BAC will have two lines in the
certificate file.
"""
p = OptionParser(certificate.__doc__)
opts, args = p.parse_args(args)
if len(args) != 2:
sys.exit(not p.print_help())
tpffile, certificatefile = args
fastadir = "fasta"
tpf = TPF(tpffile)
data = check_certificate(certificatefile)
fw = must_open(certificatefile, "w")
for i, a in enumerate(tpf):
if a.is_gap:
continue
aid = a.component_id
af = op.join(fastadir, aid + ".fasta")
if not op.exists(af): # Check to avoid redownload
entrez([aid, "--skipcheck", "--outdir=" + fastadir])
north, south = tpf.getNorthSouthClone(i)
aphase, asize = phase(aid)
for tag, p in (("North", north), ("South", south)):
if not p: # end of the chromosome
ov = "telomere\t{0}".format(asize)
elif p.isCloneGap:
bphase = "0"
ov = "{0}\t{1}".format(p.gap_type, asize)
else:
bid = p.component_id
bphase, bsize = phase(bid)
key = (tag, aid, bid)
if key in data:
print >> fw, data[key]
continue
ar = [aid, bid, "--dir=" + fastadir]
o = overlap(ar)
ov = o.certificateline if o \
else "{0}\t{1}\tNone".format(bid, asize)
print >> fw, "\t".join(str(x) for x in \
(tag, a.object, aphase, bphase, aid, ov))
fw.flush()
def htg(args):
"""
%prog htg fastafile template.sbt
Prepare sqnfiles for Genbank HTG submission to update existing records.
`fastafile` contains the records to update, multiple records are allowed
(with each one generating separate sqn file in the sqn/ folder). The record
defline has the accession ID. For example,
>AC148290.3
Internally, this generates two additional files (phasefile and namesfile)
and download records from Genbank. Below is implementation details:
`phasefile` contains, for each accession, phase information. For example:
AC148290.3 3 HTG 2 mth2-45h12
which means this is a Phase-3 BAC. Record with only a single contig will be
labeled as Phase-3 regardless of the info in the `phasefile`. Template file
is the Genbank sbt template. See jcvi.formats.sbt for generation of such
files.
Another problem is that Genbank requires the name of the sequence to stay
the same when updating and will kick back with a table of name conflicts.
For example:
We are unable to process the updates for these entries
for the following reason:
Seqname has changed
Accession Old seq_name New seq_name
--------- ------------ ------------
AC239792 mtg2_29457 AC239792.1
To prepare a submission, this script downloads genbank and asn.1 format,
and generate the phase file and the names file (use formats.agp.phase() and
apps.gbsubmit.asn(), respectively). These get automatically run.
However, use --phases if the genbank files contain outdated information.
For example, the clone name changes or phase upgrades. In this case, run
formats.agp.phase() manually, modify the phasefile and use --phases to override.
"""
from jcvi.formats.fasta import sequin, ids
from jcvi.formats.agp import phase
from jcvi.apps.fetch import entrez
p = OptionParser(htg.__doc__)
p.add_option(
"--phases",
default=None,
help="Use another phasefile to override",
)
p.add_option("--comment", default="", help="Comments for this update")
opts, args = p.parse_args(args)
if len(args) != 2:
sys.exit(not p.print_help())
fastafile, sbtfile = args
pf = fastafile.rsplit(".", 1)[0]
idsfile = pf + ".ids"
phasefile = pf + ".phases"
namesfile = pf + ".names"
ids([fastafile, "--outfile={0}".format(idsfile)])
asndir = "asn.1"
mkdir(asndir)
entrez([idsfile, "--format=asn.1", "--outdir={0}".format(asndir)])
asn(glob("{0}/*".format(asndir)) + ["--outfile={0}".format(namesfile)])
if opts.phases is None:
gbdir = "gb"
mkdir(gbdir)
entrez([idsfile, "--format=gb", "--outdir={0}".format(gbdir)])
phase(
glob("{0}/*".format(gbdir)) + ["--outfile={0}".format(phasefile)])
else:
phasefile = opts.phases
assert op.exists(namesfile) and op.exists(phasefile)
newphasefile = phasefile + ".new"
newphasefw = open(newphasefile, "w")
comment = opts.comment
fastadir = "fasta"
sqndir = "sqn"
mkdir(fastadir)
mkdir(sqndir)
from jcvi.graphics.histogram import stem_leaf_plot
names = DictFile(namesfile)
assert len(set(names.keys())) == len(set(names.values()))
phases = DictFile(phasefile)
ph = [int(x) for x in phases.values()]
# vmin 1, vmax 4, bins 3
stem_leaf_plot(ph, 1, 4, 3, title="Counts of phases before updates")
logging.debug("Information loaded for {0} records.".format(len(phases)))
assert len(names) == len(phases)
newph = []
cmd = "faSplit byname {0} {1}/".format(fastafile, fastadir)
sh(cmd, outfile="/dev/null", errfile="/dev/null")
acmd = "tbl2asn -a z -p fasta -r {sqndir}"
acmd += " -i {splitfile} -t {sbtfile} -C tigr"
acmd += ' -j "{qualifiers}"'
acmd += " -A {accession_nv} -o {sqndir}/{accession_nv}.sqn -V Vbr"
acmd += ' -y "{comment}" -W T -T T'
qq = "[tech=htgs {phase}] [organism=Medicago truncatula] [strain=A17]"
nupdated = 0
for row in open(phasefile):
atoms = row.rstrip().split("\t")
# see formats.agp.phase() for column contents
accession, phase, clone = atoms[0], atoms[1], atoms[-1]
fafile = op.join(fastadir, accession + ".fa")
accession_nv = accession.split(".", 1)[0]
newid = names[accession_nv]
newidopt = "--newid={0}".format(newid)
cloneopt = "--clone={0}".format(clone)
splitfile, gaps = sequin([fafile, newidopt, cloneopt])
splitfile = op.basename(splitfile)
phase = int(phase)
assert phase in (1, 2, 3)
oldphase = phase
if gaps == 0 and phase != 3:
phase = 3
if gaps != 0 and phase == 3:
phase = 2
print("{0}\t{1}\t{2}".format(accession_nv, oldphase, phase),
file=newphasefw)
newph.append(phase)
qualifiers = qq.format(phase=phase)
if ";" in clone:
qualifiers += " [keyword=HTGS_POOLED_MULTICLONE]"
cmd = acmd.format(
accession=accession,
accession_nv=accession_nv,
sqndir=sqndir,
sbtfile=sbtfile,
splitfile=splitfile,
qualifiers=qualifiers,
comment=comment,
)
sh(cmd)
verify_sqn(sqndir, accession)
nupdated += 1
stem_leaf_plot(newph, 1, 4, 3, title="Counts of phases after updates")
print("A total of {0} records updated.".format(nupdated), file=sys.stderr)
