def write_csv(header, contents, sep=",", filename="stdout", thousands=False, tee=False):
"""
Write csv that are aligned with the column headers.
>>> header = ["x_value", "y_value"]
>>> contents = [(1, 100), (2, 200)]
>>> write_csv(header, contents)
x_value, y_value
1, 100
2, 200
"""
from jcvi.formats.base import must_open, is_number
from jcvi.utils.cbook import thousands as th
fw = must_open(filename, "w")
allcontents = [header] + contents if header else contents
cols = len(contents[0])
for content in allcontents:
assert len(content) == cols
# Stringify the contents
for i, content in enumerate(allcontents):
if thousands:
content = [int(x) if is_number(x, cast=int) else x for x in content]
content = [th(x) if (is_number(x, cast=int) and x >= 1000) else x for x in content]
allcontents[i] = [str(x) for x in content]
colwidths = [max(len(x[i]) for x in allcontents) for i in xrange(cols)]
sep += " "
for content in allcontents:
rjusted = [x.rjust(cw) for x, cw in zip(content, colwidths)]
formatted = sep.join(rjusted)
print >> fw, formatted
if tee and filename != "stdout":
print formatted
def load_csv(header, contents, sep=",", thousands=False, align=True):
from jcvi.formats.base import is_number
from jcvi.utils.cbook import thousands as th
allcontents = [header] + contents if header else contents
cols = len(contents[0])
for content in allcontents:
assert len(content) == cols
# Stringify the contents
for i, content in enumerate(allcontents):
if thousands:
content = [int(x) if is_number(x, cast=int) else x \
for x in content]
content = [th(x) if (is_number(x, cast=int) and x >= 1000) else x \
for x in content]
allcontents[i] = [str(x) for x in content]
colwidths = [max(len(x[i]) for x in allcontents) for i in xrange(cols)]
sep += " "
formatted_contents = []
for content in allcontents:
rjusted = [x.rjust(cw) for x, cw in zip(content, colwidths)] \
if align else content
formatted = sep.join(rjusted)
formatted_contents.append(formatted)
return formatted_contents
def load_csv(header, contents, sep=",", thousands=False, align=True):
from jcvi.formats.base import is_number
from jcvi.utils.cbook import thousands as th
allcontents = [header] + contents if header else contents
cols = len(contents[0])
for content in allcontents:
assert len(content) == cols
# Stringify the contents
for i, content in enumerate(allcontents):
if thousands:
content = [int(x) if is_number(x, cast=int) else x \
for x in content]
content = [th(x) if (is_number(x, cast=int) and x >= 1000) else x \
for x in content]
allcontents[i] = [str(x) for x in content]
colwidths = [max(len(x[i]) for x in allcontents) for i in xrange(cols)]
sep += " "
formatted_contents = []
for content in allcontents:
rjusted = [x.rjust(cw) for x, cw in zip(content, colwidths)] \
if align else content
formatted = sep.join(rjusted)
formatted_contents.append(formatted)
return formatted_contents
def random(args):
"""
%prog random bedfile number_of_features
Extract a random subset of features. Number of features can be an integer
number, or a fractional number in which case a random fraction (for example
0.1 = 10% of all features) will be extracted.
"""
from random import sample
from jcvi.formats.base import flexible_cast
p = OptionParser(random.__doc__)
opts, args = p.parse_args(args)
if len(args) != 2:
sys.exit(not p.print_help())
bedfile, N = args
assert is_number(N)
b = Bed(bedfile)
NN = flexible_cast(N)
if NN < 1:
NN = int(round(NN * len(b)))
beds = sample(b, NN)
new_bed = Bed()
new_bed.extend(beds)
outfile = bedfile.rsplit(".", 1)[0] + ".{0}.bed".format(N)
new_bed.print_to_file(outfile)
logging.debug("Write {0} features to `{1}`".format(NN, outfile))
def guess_method(tag):
from jcvi.formats.base import is_number
jobids = tag.split(",")
for jobid in jobids:
if not is_number(jobid):
return "pattern"
return "jobid"
def gffline(self, type='match', source='default'):
score = "." if not self.score or \
(self.score and not is_number(self.score)) \
else self.score
strand = "." if not self.strand else self.strand
row = "\t".join((self.seqid, source, type, str(self.start + 1),
str(self.end), score, strand, '.', 'ID=' + self.accn))
return row
def guess_method(tag):
from jcvi.formats.base import is_number
jobids = tag.split(",")
for jobid in jobids:
if not is_number(jobid):
return "pattern"
return "jobid"
def gffline(self, type='match', source='default'):
score = "." if not self.score or \
(self.score and not is_number(self.score)) \
else self.score
strand = "." if not self.strand else self.strand
row = "\t".join((self.seqid, source, type,
str(self.start + 1), str(self.end), score,
strand, '.', 'ID=' + self.accn))
return row
def blat(args):
"""
%prog blat map1.txt ref.fasta
Make ALLMAPS input csv based on sequences. The tab-delimited txt file
include: name, LG, position, sequence.
"""
from jcvi.formats.base import is_number
from jcvi.formats.blast import best as blast_best, bed as blast_bed
from jcvi.apps.align import blat as blat_align
p = OptionParser(blat.__doc__)
opts, args = p.parse_args(args)
if len(args) != 2:
sys.exit(not p.print_help())
maptxt, ref = args
pf = maptxt.rsplit(".", 1)[0]
register = {}
fastafile = pf + ".fasta"
fp = open(maptxt)
fw = open(fastafile, "w")
for row in fp:
name, lg, pos, seq = row.split()
if not is_number(pos):
continue
register[name] = (pf + '-' + lg, pos)
print(">{0}\n{1}\n".format(name, seq), file=fw)
fw.close()
blatfile = blat_align([ref, fastafile])
bestfile = blast_best([blatfile])
bedfile = blast_bed([bestfile])
b = Bed(bedfile).order
pf = ".".join((op.basename(maptxt).split(".")[0],
op.basename(ref).split(".")[0]))
csvfile = pf + ".csv"
fp = open(maptxt)
fw = open(csvfile, "w")
for row in fp:
name, lg, pos, seq = row.split()
if name not in b:
continue
bbi, bb = b[name]
scaffold, scaffold_pos = bb.seqid, bb.start
print(",".join(str(x) for x in \
(scaffold, scaffold_pos, lg, pos)), file=fw)
fw.close()
def blat(args):
"""
%prog blat map1.txt ref.fasta
Make ALLMAPS input csv based on sequences. The tab-delimited txt file
include: name, LG, position, sequence.
"""
from jcvi.formats.base import is_number
from jcvi.formats.blast import best as blast_best, bed as blast_bed
from jcvi.apps.align import blat as blat_align
p = OptionParser(blat.__doc__)
opts, args = p.parse_args(args)
if len(args) != 2:
sys.exit(not p.print_help())
maptxt, ref = args
pf = maptxt.rsplit(".", 1)[0]
register = {}
fastafile = pf + ".fasta"
fp = open(maptxt)
fw = open(fastafile, "w")
for row in fp:
name, lg, pos, seq = row.split()
if not is_number(pos):
continue
register[name] = (pf + '-' + lg, pos)
print(">{0}\n{1}\n".format(name, seq), file=fw)
fw.close()
blatfile = blat_align([ref, fastafile])
bestfile = blast_best([blatfile])
bedfile = blast_bed([bestfile])
b = Bed(bedfile).order
pf = ".".join(
(op.basename(maptxt).split(".")[0], op.basename(ref).split(".")[0]))
csvfile = pf + ".csv"
fp = open(maptxt)
fw = open(csvfile, "w")
for row in fp:
name, lg, pos, seq = row.split()
if name not in b:
continue
bbi, bb = b[name]
scaffold, scaffold_pos = bb.seqid, bb.start
print(",".join(str(x) for x in \
(scaffold, scaffold_pos, lg, pos)), file=fw)
fw.close()
def read_meme(fi):
mtfs = []
fhi = open(fi, 'r')
for head, content in read_block(fhi, 'MOTIF'):
ps = head.split(' ')
pre, mid = ps[:2]
score = ''
if len(ps) >= 3:
score = ps[2]
#mtf = mid.split("-")[1]
if is_number(score):
score = float(score)
width = len(content) - 2
mtfs.append([mid, width, score])
#print(mid,'\t',width)
return mtfs
def rename_seqid(seqid):
seqid = seqid.split("_")[-1]
seqid = seqid.replace("supercont", "s")
seqid = seqid.replace("contig", "c").replace("scaffold", "s")
return "c{}".format(int(seqid)) if is_number(seqid, int) else seqid
def frommaf(args):
"""
%prog frommaf maffile
Convert to four-column tabular format from MAF.
"""
p = OptionParser(frommaf.__doc__)
p.add_option("--validate", help="Validate coordinates against FASTA")
opts, args = p.parse_args(args)
if len(args) != 1:
sys.exit(not p.print_help())
(maf,) = args
snpfile = maf.rsplit(".", 1)[0] + ".vcf"
fp = open(maf)
fw = open(snpfile, "w")
total = 0
id = "."
qual = 20
filter = "PASS"
info = "DP=20"
print("##fileformat=VCFv4.0", file=fw)
print("#CHROM POS ID REF ALT QUAL FILTER INFO".replace(" ", "\t"), file=fw)
for row in fp:
atoms = row.split()
c, pos, ref, alt = atoms[:4]
if is_number(c, int):
c = int(c)
else:
continue
c = "chr{0:02d}".format(c)
pos = int(pos)
print(
"\t".join(str(x) for x in (c, pos, id, ref, alt, qual, filter, info)),
file=fw,
)
total += 1
fw.close()
validate = opts.validate
if not validate:
return
from jcvi.utils.cbook import percentage
f = Fasta(validate)
fp = open(snpfile)
nsnps = 0
for row in fp:
if row[0] == "#":
continue
c, pos, id, ref, alt, qual, filter, info = row.split("\t")
pos = int(pos)
feat = dict(chr=c, start=pos, stop=pos)
s = f.sequence(feat)
s = str(s)
assert s == ref, "Validation error: {0} is {1} (expect: {2})".format(
feat, s, ref
)
nsnps += 1
if nsnps % 50000 == 0:
logging.debug("SNPs parsed: {0}".format(percentage(nsnps, total)))
logging.debug(
"A total of {0} SNPs validated and written to `{1}`.".format(nsnps, snpfile)
)
