def A50(args):
"""
%prog A50 contigs_A.fasta contigs_B.fasta ...
Plots A50 graphics, see blog post (http://blog.malde.org/index.php/a50/)
"""
p = OptionParser(A50.__doc__)
p.add_option("--overwrite", default=False, action="store_true",
help="overwrite .rplot file if exists [default: %default]")
p.add_option("--cutoff", default=0, type="int", dest="cutoff",
help="use contigs above certain size [default: %default]")
p.add_option("--stepsize", default=10, type="int", dest="stepsize",
help="stepsize for the distribution [default: %default]")
opts, args = p.parse_args(args)
if not args:
sys.exit(p.print_help())
import numpy as np
from jcvi.utils.table import loadtable
stepsize = opts.stepsize # use stepsize to speed up drawing
rplot = "A50.rplot"
if not op.exists(rplot) or opts.overwrite:
fw = open(rplot, "w")
header = "\t".join(("index", "cumsize", "fasta"))
statsheader = ("Fasta", "L50", "N50", "Min", "Max", "Average", "Sum",
"Counts")
statsrows = []
print >>fw, header
for fastafile in args:
f = Fasta(fastafile, index=False)
ctgsizes = [length for k, length in f.itersizes()]
ctgsizes = np.array(ctgsizes)
a50, l50, n50 = calculate_A50(ctgsizes, cutoff=opts.cutoff)
cmin, cmax, cmean = min(ctgsizes), max(ctgsizes), np.mean(ctgsizes)
csum, counts = np.sum(ctgsizes), len(ctgsizes)
cmean = int(round(cmean))
statsrows.append((fastafile, l50, n50, cmin, cmax, cmean, csum,
counts))
logging.debug("`{0}` ctgsizes: {1}".format(fastafile, ctgsizes))
tag = "{0} (L50={1})".format(\
op.basename(fastafile).rsplit(".", 1)[0], l50)
logging.debug(tag)
for i, s in zip(xrange(0, len(a50), stepsize), a50[::stepsize]):
print >> fw, "\t".join((str(i), str(s / 1000000.), tag))
fw.close()
table = loadtable(statsheader, statsrows)
print >> sys.stderr, table
generate_plot(rplot)
def A50(args):
"""
%prog A50 contigs_A.fasta contigs_B.fasta ...
Plots A50 graphics, see blog post (http://blog.malde.org/index.php/a50/)
"""
p = OptionParser(A50.__doc__)
p.add_option("--overwrite", default=False, action="store_true",
help="overwrite .rplot file if exists [default: %default]")
p.add_option("--cutoff", default=0, type="int", dest="cutoff",
help="use contigs above certain size [default: %default]")
p.add_option("--stepsize", default=10, type="int", dest="stepsize",
help="stepsize for the distribution [default: %default]")
opts, args = p.parse_args(args)
if not args:
sys.exit(p.print_help())
import numpy as np
from jcvi.utils.table import loadtable
stepsize = opts.stepsize # use stepsize to speed up drawing
rplot = "A50.rplot"
if not op.exists(rplot) or opts.overwrite:
fw = open(rplot, "w")
header = "\t".join(("index", "cumsize", "fasta"))
statsheader = ("Fasta", "L50", "N50", "Min", "Max", "Average", "Sum",
"Counts")
statsrows = []
print >>fw, header
for fastafile in args:
f = Fasta(fastafile, index=False)
ctgsizes = [length for k, length in f.itersizes()]
ctgsizes = np.array(ctgsizes)
a50, l50, n50 = calculate_A50(ctgsizes, cutoff=opts.cutoff)
cmin, cmax, cmean = min(ctgsizes), max(ctgsizes), np.mean(ctgsizes)
csum, counts = np.sum(ctgsizes), len(ctgsizes)
cmean = int(round(cmean))
statsrows.append((fastafile, l50, n50, cmin, cmax, cmean, csum,
counts))
logging.debug("`{0}` ctgsizes: {1}".format(fastafile, ctgsizes))
tag = "{0} (L50={1})".format(\
op.basename(fastafile).rsplit(".", 1)[0], l50)
logging.debug(tag)
for i, s in zip(xrange(0, len(a50), stepsize), a50[::stepsize]):
print >> fw, "\t".join((str(i), str(s / 1000000.), tag))
fw.close()
table = loadtable(statsheader, statsrows)
print >> sys.stderr, table
generate_plot(rplot)
def rstats(self, object, bacs, components, scaffold_sizes, length):
from jcvi.utils.cbook import human_size
nbacs = len(bacs)
nscaffolds = len(scaffold_sizes)
a50, l50, n50 = calculate_A50(scaffold_sizes)
l50 = human_size(l50)
length = human_size(length)
return (object, nbacs, components, nscaffolds, n50, l50, length)
def rstats(self, object, bacs, components, scaffold_sizes, length):
from jcvi.utils.cbook import human_size
nbacs = len(bacs)
nscaffolds = len(scaffold_sizes)
a50, l50, n50 = calculate_A50(scaffold_sizes)
l50 = human_size(l50)
length = human_size(length)
return (object, nbacs, components, nscaffolds, n50, l50, length)
def get_stats(blastfile, strict=False):
from jcvi.utils.range import range_union, range_span
from .pyblast import BlastLine
logging.debug("Report stats on `%s`" % blastfile)
fp = open(blastfile)
ref_ivs = []
qry_ivs = []
identicals = 0
ngaps = 0
alignlens = []
for row in fp:
c = BlastLine(row)
qstart, qstop = c.qstart, c.qstop
if qstart > qstop:
qstart, qstop = qstop, qstart
qry_ivs.append((c.query, qstart, qstop))
sstart, sstop = c.sstart, c.sstop
if sstart > sstop:
sstart, sstop = sstop, sstart
ref_ivs.append((c.subject, sstart, sstop))
alen = c.hitlen
ngaps += c.ngaps
identicals += c.hitlen - c.nmismatch - c.ngaps
alignlens.append(alen)
qrycovered = range_union(qry_ivs)
refcovered = range_union(ref_ivs)
if strict:
# We discount gaps in counting covered bases, since we
# did not track individually gaps in qry and ref, we assume
# the gaps are opened evenly in the two sequences
qrycovered -= ngaps / 2
refcovered -= ngaps / 2
qryspan = range_span(qry_ivs)
refspan = range_span(ref_ivs)
_, AL50, _ = calculate_A50(alignlens)
filename = op.basename(blastfile)
alignstats = AlignStats(
filename, qrycovered, refcovered, qryspan, refspan, identicals, AL50
)
return alignstats
def get_stats(blastfile, strict=False):
from jcvi.utils.range import range_union, range_span
from .pyblast import BlastLine
logging.debug("Report stats on `%s`" % blastfile)
fp = open(blastfile)
ref_ivs = []
qry_ivs = []
identicals = 0
ngaps = 0
alignlens = []
for row in fp:
c = BlastLine(row)
qstart, qstop = c.qstart, c.qstop
if qstart > qstop:
qstart, qstop = qstop, qstart
qry_ivs.append((c.query, qstart, qstop))
sstart, sstop = c.sstart, c.sstop
if sstart > sstop:
sstart, sstop = sstop, sstart
ref_ivs.append((c.subject, sstart, sstop))
alen = c.hitlen
ngaps += c.ngaps
identicals += c.hitlen - c.nmismatch - c.ngaps
alignlens.append(alen)
qrycovered = range_union(qry_ivs)
refcovered = range_union(ref_ivs)
if strict:
# We discount gaps in counting covered bases, since we
# did not track individually gaps in qry and ref, we assume
# the gaps are opened evenly in the two sequences
qrycovered -= ngaps / 2
refcovered -= ngaps / 2
qryspan = range_span(qry_ivs)
refspan = range_span(ref_ivs)
_, AL50, _ = calculate_A50(alignlens)
filename = op.basename(blastfile)
alignstats = AlignStats(filename, qrycovered, refcovered,
qryspan, refspan, identicals, AL50)
return alignstats
def get_stats(coordsfile):
from jcvi.utils.range import range_union
logging.debug("Report stats on `%s`" % coordsfile)
coords = Coords(coordsfile)
ref_ivs = []
qry_ivs = []
identicals = 0
alignlen = 0
alignlens = []
for c in coords:
qstart, qstop = c.start2, c.end2
if qstart > qstop:
qstart, qstop = qstop, qstart
qry_ivs.append((c.query, qstart, qstop))
sstart, sstop = c.start1, c.end1
if sstart > sstop:
sstart, sstop = sstop, sstart
ref_ivs.append((c.ref, sstart, sstop))
alen = sstop - sstart
alignlen += alen
identicals += c.identity / 100. * alen
alignlens.append(alen)
qrycovered = range_union(qry_ivs)
refcovered = range_union(ref_ivs)
_, AL50, _ = calculate_A50(alignlens)
filename = op.basename(coordsfile)
alignstats = AlignStats(filename, qrycovered, refcovered,
None, None, identicals)
return alignstats
def get_stats(coordsfile):
from jcvi.utils.range import range_union
logging.debug("Report stats on `%s`" % coordsfile)
coords = Coords(coordsfile)
ref_ivs = []
qry_ivs = []
identicals = 0
alignlen = 0
alignlens = []
for c in coords:
qstart, qstop = c.start2, c.end2
if qstart > qstop:
qstart, qstop = qstop, qstart
qry_ivs.append((c.query, qstart, qstop))
sstart, sstop = c.start1, c.end1
if sstart > sstop:
sstart, sstop = sstop, sstart
ref_ivs.append((c.ref, sstart, sstop))
alen = sstop - sstart
alignlen += alen
identicals += c.identity / 100. * alen
alignlens.append(alen)
qrycovered = range_union(qry_ivs)
refcovered = range_union(ref_ivs)
_, AL50, _ = calculate_A50(alignlens)
filename = op.basename(coordsfile)
alignstats = AlignStats(filename, qrycovered, refcovered, None, None,
identicals)
return alignstats
def summary(self):
from jcvi.assembly.base import calculate_A50
ctgsizes = self.sizes
a50, l50, n50 = calculate_A50(ctgsizes)
return sum(ctgsizes), l50, n50
def summary(self):
from jcvi.assembly.base import calculate_A50
ctgsizes = self.sizes
a50, l50, n50 = calculate_A50(ctgsizes)
return sum(ctgsizes), l50, n50
