def summary(args):
"""
%prog summary gffile fastafile
Print summary stats, including:
- Gene/Exon/Intron
- Number
- Average size (bp)
- Median size (bp)
- Total length (Mb)
- % of genome
- % GC
"""
p = OptionParser(summary.__doc__)
opts, args = p.parse_args(args)
if len(args) != 2:
sys.exit(not p.print_help())
gff_file, ref = args
s = Fasta(ref)
g = make_index(gff_file)
geneseqs, exonseqs, intronseqs = [], [], [] # Calc % GC
for f in g.features_of_type("gene"):
fid = f.id
fseq = s.sequence({'chr': f.chrom, 'start': f.start, 'stop': f.stop})
geneseqs.append(fseq)
exons = set((c.chrom, c.start, c.stop) for c in g.children(fid, 2) \
if c.featuretype == "exon")
exons = list(exons)
for chrom, start, stop in exons:
fseq = s.sequence({'chr': chrom, 'start': start, 'stop': stop})
exonseqs.append(fseq)
introns = range_interleave(exons)
for chrom, start, stop in introns:
fseq = s.sequence({'chr': chrom, 'start': start, 'stop': stop})
intronseqs.append(fseq)
r = {} # Report
for t, tseqs in zip(("Gene", "Exon", "Intron"),
(geneseqs, exonseqs, intronseqs)):
tsizes = [len(x) for x in tseqs]
tsummary = SummaryStats(tsizes, dtype="int")
r[t, "Number"] = tsummary.size
r[t, "Average size (bp)"] = tsummary.mean
r[t, "Median size (bp)"] = tsummary.median
r[t, "Total length (Mb)"] = human_size(tsummary.sum,
precision=0,
target="Mb")
r[t, "% of genome"] = percentage(tsummary.sum,
s.totalsize,
precision=0,
mode=-1)
r[t, "% GC"] = gc(tseqs)
print >> sys.stderr, tabulate(r)
def summary(args):
"""
%prog summary gffile fastafile
Print summary stats, including:
- Gene/Exon/Intron
- Number
- Average size (bp)
- Median size (bp)
- Total length (Mb)
- % of genome
- % GC
"""
p = OptionParser(summary.__doc__)
opts, args = p.parse_args(args)
if len(args) != 2:
sys.exit(not p.print_help())
gff_file, ref = args
s = Fasta(ref)
g = make_index(gff_file)
geneseqs, exonseqs, intronseqs = [], [], [] # Calc % GC
for f in g.features_of_type("gene"):
fid = f.id
fseq = s.sequence({'chr': f.chrom, 'start': f.start, 'stop': f.stop})
geneseqs.append(fseq)
exons = set((c.chrom, c.start, c.stop) for c in g.children(fid, 2) \
if c.featuretype == "exon")
exons = list(exons)
for chrom, start, stop in exons:
fseq = s.sequence({'chr': chrom, 'start': start, 'stop': stop})
exonseqs.append(fseq)
introns = range_interleave(exons)
for chrom, start, stop in introns:
fseq = s.sequence({'chr': chrom, 'start': start, 'stop': stop})
intronseqs.append(fseq)
r = {} # Report
for t, tseqs in zip(("Gene", "Exon", "Intron"), (geneseqs, exonseqs, intronseqs)):
tsizes = [len(x) for x in tseqs]
tsummary = SummaryStats(tsizes, dtype="int")
r[t, "Number"] = tsummary.size
r[t, "Average size (bp)"] = tsummary.mean
r[t, "Median size (bp)"] = tsummary.median
r[t, "Total length (Mb)"] = human_size(tsummary.sum, precision=0, target="Mb")
r[t, "% of genome"] = percentage(tsummary.sum, s.totalsize, precision=0, mode=-1)
r[t, "% GC"] = gc(tseqs)
print(tabulate(r), file=sys.stderr)
opts, args = p.parse_args(args)
try:
agpfile, componentfasta, targetfasta = args
except Exception, e:
sys.exit(p.print_help())
agp = AGP(agpfile)
build = Fasta(targetfasta)
bacs = Fasta(componentfasta, index=False)
# go through this line by line
for aline in agp:
try:
build_seq = build.sequence(dict(chr=aline.object,
start=aline.object_beg, stop=aline.object_end))
if aline.is_gap:
assert build_seq.upper() == aline.gap_length * 'N', \
"gap mismatch: %s" % aline
else:
bac_seq = bacs.sequence(dict(chr=aline.component_id,
start=aline.component_beg, stop=aline.component_end,
strand=aline.orientation))
assert build_seq.upper() == bac_seq.upper(), \
"sequence mismatch: %s" % aline
logging.debug("%s:%d-%d verified" % (aline.object,
aline.object_beg, aline.object_end))
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 [default: %default]")
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 >> fw, "##fileformat=VCFv4.0"
print >> fw, "#CHROM POS ID REF ALT QUAL FILTER INFO".replace(" ", "\t")
for row in fp:
atoms = row.split()
c, pos, ref, alt = atoms[:4]
try:
c = int(c)
except:
continue
c = "chr{0:02d}".format(c)
pos = int(pos)
print >> fw, "\t".join(str(x) for x in \
(c, pos, id, ref, alt, qual, filter, info))
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))
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 [default: %default]")
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 >> fw, "##fileformat=VCFv4.0"
print >> fw, "#CHROM POS ID REF ALT QUAL FILTER INFO".replace(" ", "\t")
for row in fp:
atoms = row.split()
c, pos, ref, alt = atoms[:4]
try:
c = int(c)
except:
continue
c = "chr{0:02d}".format(c)
pos = int(pos)
print >> fw, "\t".join(str(x) for x in \
(c, pos, id, ref, alt, qual, filter, info))
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))
def load(args):
'''
%prog load gff_file fasta_file [--options]
Parses the selected features out of GFF, with subfeatures concatenated.
For example, to get the CDS sequences, do this::
$ %prog load athaliana.gff athaliana.fa --parents mRNA --children CDS
'''
from jcvi.formats.fasta import Seq, SeqRecord
p = OptionParser(load.__doc__)
p.add_option("--parents", dest="parents", default="mRNA",
help="list of features to extract, use comma to separate (e.g."
"'gene,mRNA') [default: %default]")
p.add_option("--children", dest="children", default="CDS",
help="list of features to extract, use comma to separate (e.g."
"'five_prime_UTR,CDS,three_prime_UTR') [default: %default]")
p.add_option("--attribute",
help="The attribute field to extract [default: %default]")
set_outfile(p)
opts, args = p.parse_args(args)
if len(args) != 2:
sys.exit(p.print_help())
gff_file, fasta_file = args
g = make_index(gff_file)
f = Fasta(fasta_file, index=False)
fw = must_open(opts.outfile, "w")
parents = set(opts.parents.split(','))
children_list = set(opts.children.split(','))
attr = opts.attribute
for feat in get_parents(gff_file, parents):
children = []
for c in g.children(feat.id, 1):
if c.featuretype not in children_list:
continue
child = f.sequence(dict(chr=c.chrom, start=c.start, stop=c.stop,
strand=c.strand))
children.append((child, c))
if not children:
print >>sys.stderr, "[warning] %s has no children with type %s" \
% (feat.id, ','.join(children_list))
continue
# sort children in incremental position
children.sort(key=lambda x: x[1].start)
# reverse children if negative strand
if feat.strand == '-':
children.reverse()
feat_seq = ''.join(x[0] for x in children)
description = ",".join(feat.attributes[attr]) \
if attr and attr in feat.attributes else ""
description = description.replace("\"", "")
rec = SeqRecord(Seq(feat_seq), id=feat.id, description=description)
SeqIO.write([rec], fw, "fasta")
fw.flush()
opts, args = p.parse_args(args)
try:
agpfile, componentfasta, targetfasta = args
except Exception, e:
sys.exit(p.print_help())
agp = AGP(agpfile)
build = Fasta(targetfasta)
bacs = Fasta(componentfasta, index=False)
# go through this line by line
for aline in agp:
try:
build_seq = build.sequence(
dict(chr=aline.object,
start=aline.object_beg,
stop=aline.object_end))
if aline.is_gap:
assert build_seq.upper() == aline.gap_length * 'N', \
"gap mismatch: %s" % aline
else:
bac_seq = bacs.sequence(
dict(chr=aline.component_id,
start=aline.component_beg,
stop=aline.component_end,
strand=aline.orientation))
assert build_seq.upper() == bac_seq.upper(), \
"sequence mismatch: %s" % aline
def load(args):
'''
%prog load gff_file fasta_file [--options]
Parses the selected features out of GFF, with subfeatures concatenated.
For example, to get the CDS sequences, do this::
$ %prog load athaliana.gff athaliana.fa --parents mRNA --children CDS
'''
from jcvi.formats.fasta import Seq, SeqRecord
p = OptionParser(load.__doc__)
p.add_option(
"--parents",
dest="parents",
default="mRNA",
help="list of features to extract, use comma to separate (e.g."
"'gene,mRNA') [default: %default]")
p.add_option(
"--children",
dest="children",
default="CDS",
help="list of features to extract, use comma to separate (e.g."
"'five_prime_UTR,CDS,three_prime_UTR') [default: %default]")
p.add_option("--attribute",
help="The attribute field to extract [default: %default]")
set_outfile(p)
opts, args = p.parse_args(args)
if len(args) != 2:
sys.exit(p.print_help())
gff_file, fasta_file = args
g = make_index(gff_file)
f = Fasta(fasta_file, index=False)
fw = must_open(opts.outfile, "w")
parents = set(opts.parents.split(','))
children_list = set(opts.children.split(','))
attr = opts.attribute
for feat in get_parents(gff_file, parents):
children = []
for c in g.children(feat.id, 1):
if c.featuretype not in children_list:
continue
child = f.sequence(
dict(chr=c.chrom, start=c.start, stop=c.stop, strand=c.strand))
children.append((child, c))
if not children:
print >>sys.stderr, "[warning] %s has no children with type %s" \
% (feat.id, ','.join(children_list))
continue
# sort children in incremental position
children.sort(key=lambda x: x[1].start)
# reverse children if negative strand
if feat.strand == '-':
children.reverse()
feat_seq = ''.join(x[0] for x in children)
description = ",".join(feat.attributes[attr]) \
if attr and attr in feat.attributes else ""
description = description.replace("\"", "")
rec = SeqRecord(Seq(feat_seq), id=feat.id, description=description)
SeqIO.write([rec], fw, "fasta")
fw.flush()
