def print_gffline(self, fw, f, seqid, parent=None):
score = phase = "."
type = f.type
if type == "source":
type = "contig"
attr = "ID=tmp"
source = self.source
start = get_number(f.location.start) + 1
end = get_number(f.location.end)
strand = '-' if f.strand < 0 else '+'
g = "\t".join(str(x) for x in \
(seqid, source, type, start, end, score, strand, phase, attr))
g = GffLine(g)
qual = f.qualifiers
id = "tmp"
if MT in qual:
id = seqid
elif LT in qual:
id, = qual[LT]
else:
qual[LT] = [self.current_id]
id, = qual[LT]
id = id.split()[0]
if parent:
id, = parent.qualifiers[LT]
id = id.split()[0]
if type == 'CDS':
parent_id = id
self.counter[id] += 1
suffix = ".cds.{0}".format(self.counter[id])
id = parent_id + suffix
g.attributes["Parent"] = [parent_id]
assert id != "tmp", f
g.attributes["ID"] = [id]
if type == "mRNA":
g.attributes["Name"] = g.attributes["ID"]
if "product" in qual:
note, = qual["product"]
g.attributes["Note"] = [note]
if "pseudo" in qual:
note = "Pseudogene"
g.attributes["Note"] = [note]
g.update_attributes()
print >> fw, g
self.current_id = id
def get_target(achr, bchr):
if "chr" not in achr and "chr" not in bchr:
return None
achr, bchr = get_number(achr), get_number(bchr)
if achr > bchr:
achr, bchr = bchr, achr
if (achr, bchr) in target_pairs:
return achr, bchr
return None
def print_gffline(self, fw, f, seqid, parent=None):
score = phase = "."
type = f.type
if type == "source":
type = "contig"
attr = "ID=tmp"
source = self.source
start = get_number(f.location.start) + 1
end = get_number(f.location.end)
strand = '-' if f.strand < 0 else '+'
g = "\t".join(str(x) for x in \
(seqid, source, type, start, end, score, strand, phase, attr))
g = GffLine(g)
qual = f.qualifiers
id = "tmp"
if MT in qual:
id = seqid
elif LT in qual:
id, = qual[LT]
else:
qual[LT] = [self.current_id]
id, = qual[LT]
id = id.split()[0]
if parent:
id, = parent.qualifiers[LT]
id = id.split()[0]
assert id != "tmp", f
oid = id
self.counter[(oid, type)].append((start, end))
count = len(self.counter[(oid, type)])
if type in ("mRNA", "gene"):
if type == "gene" and count > 1:
return
self.start = min(a for a, b in self.counter[(id, type)])
self.end = max(a for a, b in self.counter[(id, type)])
self.set_attribute("gene", "Alias", qual, g)
self.set_attribute("product", "Note", qual, g)
else:
suffix = ".{0}.{1}".format(type.lower(), count)
id = id + suffix
g.attributes["Parent"] = [oid]
self.set_attribute("product", "Note", qual, g)
g.attributes["ID"] = [id]
g.update_attributes()
print(g, file=fw)
self.current_id = oid
def print_gffline(self, fw, f, seqid, parent=None):
score = phase = "."
type = f.type
if type == "source":
type = "contig"
attr = "ID=tmp"
source = self.source
start = get_number(f.location.start) + 1
end = get_number(f.location.end)
strand = '-' if f.strand < 0 else '+'
g = "\t".join(str(x) for x in \
(seqid, source, type, start, end, score, strand, phase, attr))
g = GffLine(g)
qual = f.qualifiers
id = "tmp"
if MT in qual:
id = seqid
elif LT in qual:
id, = qual[LT]
else:
qual[LT] = [self.current_id]
id, = qual[LT]
id = id.split()[0]
if parent:
id, = parent.qualifiers[LT]
id = id.split()[0]
assert id != "tmp", f
oid = id
self.counter[(oid, type)].append((start, end))
count = len(self.counter[(oid, type)])
if type in ("mRNA", "gene"):
if type == "gene" and count > 1:
return
self.start = min(a for a, b in self.counter[(id, type)])
self.end = max(a for a, b in self.counter[(id, type)])
self.set_attribute("gene", "Alias", qual, g)
self.set_attribute("product", "Note", qual, g)
else:
suffix = ".{0}.{1}".format(type.lower(), count)
id = id + suffix
g.attributes["Parent"] = [oid]
self.set_attribute("product", "Note", qual, g)
g.attributes["ID"] = [id]
g.update_attributes()
print >> fw, g
self.current_id = oid
def atg_name(name, retval="chr,rank", trimpad0=True):
atg_name_pat = re.compile(
r"""
^(?P<locus>
(?:(?P<prefix>\D+[\D\d\D])\.?)(?P<chr>[\d|C|M]+)(?P<sep>[A-z]+)(?P<rank>\d+)
)
\.?(?P<iso>\d+)?
""", re.VERBOSE)
seps = ["g", "te", "trna", "s", "u"]
pad0s = ["rank"]
if name is not None:
m = re.match(atg_name_pat, name)
if m is not None and m.group('sep').lower() in seps:
retvals = []
for grp in retval.split(","):
if grp == 'chr':
val = chr_number(m.group(grp))
else:
val = get_number(m.group(grp)) \
if trimpad0 and grp in pad0s \
else m.group(grp)
retvals.append(val)
return (x for x in retvals) if len(retvals) > 1 \
else retvals[0]
return (None for x in retval.split(","))
def coge(args):
"""
%prog coge *.gff
Prepare coge datasets.
"""
p = OptionParser(coge.__doc__)
opts, args = p.parse_args(args)
if len(args) < 1:
sys.exit(not p.print_help())
gffs = args
for gff in gffs:
atoms = op.basename(gff).split(".")
gid = atoms[-2]
assert gid.startswith("gid")
gid = get_number(gid)
genomefasta = "genome_{0}.faa.fasta".format(gid)
species = "_".join(atoms[0].split("_")[:2])
cdsfasta = species + ".cds.fasta"
load([
gff,
genomefasta,
"--id_attribute=Parent",
"--outfile={0}".format(cdsfasta),
])
def bed(args):
"""
%prog bed anchorsfile
Convert ANCHORS file to BED format.
"""
from collections import defaultdict
from jcvi.compara.synteny import AnchorFile, check_beds
from jcvi.formats.bed import Bed
from jcvi.formats.base import get_number
p = OptionParser(bed.__doc__)
p.add_option("--switch", default=False, action="store_true",
help="Switch reference and aligned map elements")
p.add_option("--scale", type="float",
help="Scale the aligned map distance by factor")
p.set_beds()
p.set_outfile()
opts, args = p.parse_args(args)
if len(args) != 1:
sys.exit(not p.print_help())
anchorsfile, = args
switch = opts.switch
scale = opts.scale
ac = AnchorFile(anchorsfile)
pairs = defaultdict(list)
for a, b, block_id in ac.iter_pairs():
pairs[a].append(b)
qbed, sbed, qorder, sorder, is_self = check_beds(anchorsfile, p, opts)
bd = Bed()
for q in qbed:
qseqid, qstart, qend, qaccn = q.seqid, q.start, q.end, q.accn
if qaccn not in pairs:
continue
for s in pairs[qaccn]:
si, s = sorder[s]
sseqid, sstart, send, saccn = s.seqid, s.start, s.end, s.accn
if switch:
qseqid, sseqid = sseqid, qseqid
qstart, sstart = sstart, qstart
qend, send = send, qend
qaccn, saccn = saccn, qaccn
if scale:
sstart /= scale
try:
newsseqid = get_number(sseqid)
except ValueError:
raise ValueError, "`{0}` is on `{1}` with no number to extract".\
format(saccn, sseqid)
bedline = "\t".join(str(x) for x in (qseqid, qstart - 1, qend,
"{0}:{1}".format(newsseqid, sstart)))
bd.add(bedline)
bd.print_to_file(filename=opts.outfile, sorted=True)
def bed(args):
"""
%prog bed anchorsfile
Convert ANCHORS file to BED format.
"""
from collections import defaultdict
from jcvi.compara.synteny import AnchorFile, check_beds
from jcvi.formats.bed import Bed, BedLine
from jcvi.formats.base import get_number
p = OptionParser(bed.__doc__)
p.add_option("--switch",
default=False,
action="store_true",
help="Switch reference and aligned map elements")
p.add_option("--scale",
type="float",
help="Scale the aligned map distance by factor")
p.set_beds()
p.set_outfile()
opts, args = p.parse_args(args)
if len(args) != 1:
sys.exit(not p.print_help())
anchorsfile, = args
switch = opts.switch
scale = opts.scale
ac = AnchorFile(anchorsfile)
pairs = defaultdict(list)
for a, b, block_id in ac.iter_pairs():
pairs[a].append(b)
qbed, sbed, qorder, sorder, is_self = check_beds(anchorsfile, p, opts)
bd = Bed()
for q in qbed:
qseqid, qstart, qend, qaccn = q.seqid, q.start, q.end, q.accn
if qaccn not in pairs:
continue
for s in pairs[qaccn]:
si, s = sorder[s]
sseqid, sstart, send, saccn = s.seqid, s.start, s.end, s.accn
if switch:
qseqid, sseqid = sseqid, qseqid
qstart, sstart = sstart, qstart
qend, send = send, qend
qaccn, saccn = saccn, qaccn
if scale:
sstart /= scale
bedline = "\t".join(
str(x) for x in (qseqid, qstart - 1, qend,
"{0}:{1}".format(get_number(sseqid), sstart)))
bd.append(BedLine(bedline))
bd.print_to_file(filename=opts.outfile, sorted=True)
def remove_isoforms(ids):
"""
This is more or less a hack to remove the GMAP multiple mappings. Multiple
GMAP mappings can be seen given the names .mrna1, .mrna2, etc.
"""
key = lambda x: x.rsplit(".", 1)[0]
iso_number = lambda x: get_number(x.split(".")[-1])
ids = sorted(ids, key=key)
newids = []
for k, ii in groupby(ids, key=key):
min_i = min(list(ii), key=iso_number)
newids.append(min_i)
return newids
def remove_isoforms(ids):
"""
This is more or less a hack to remove the GMAP multiple mappings. Multiple
GMAP mappings can be seen given the names .mrna1, .mrna2, etc.
"""
key = lambda x: x.rsplit(".", 1)[0]
iso_number = lambda x: get_number(x.split(".")[-1])
ids = sorted(ids, key=key)
newids = []
for k, ii in groupby(ids, key=key):
min_i = min(list(ii), key=iso_number)
newids.append(min_i)
return newids
def atg_name(name, retval="chr,rank", trimpad0=True):
seps = ["g", "te", "trna", "s", "u", "nc"]
pad0s = ["rank"]
if name is not None:
m = re.match(atg_name_pat, name)
if m is not None and m.group("sep").lower() in seps:
retvals = []
for grp in retval.split(","):
if grp == "chr":
val = chr_number(m.group(grp))
else:
val = (get_number(m.group(grp))
if trimpad0 and grp in pad0s else m.group(grp))
retvals.append(val)
return (x for x in retvals) if len(retvals) > 1 else retvals[0]
return (None for _ in retval.split(","))
def annotate_chr(chr, chrbed, g, scores, nbedline, abedline, opts, splits):
current_chr = get_number(chr)
for line in chrbed:
accn = line.accn
if accn not in g or (opts.atg_name and "chr" not in chr):
abedline[accn] = line
continue
gaccns = g[accn]
new = [a for a in gaccns if re.search(new_id_pat, a)]
newgrp = ";".join(sorted(new))
if accn in scores:
scores[accn] = sorted(scores[accn], key=lambda x: x[1])
scores[accn] = sorted(scores[accn], key=lambda x: float(x[3]), reverse=True)
accns = []
print >> sys.stderr, accn
for elem in scores[accn]:
print >> sys.stderr, "\t" + ", ".join([str(x)\
for x in elem[1:]])
if opts.atg_name:
achr, arank = atg_name(elem[1])
if not achr or achr != current_chr:
continue
accns.append(elem[1])
if len(new) > 1:
if newgrp not in scores: scores[newgrp] = []
scores[newgrp].append(elem)
else:
accns[0:0] = [accn]
line.accn = ";".join([str(x) for x in accns])
if len(scores[accn]) > 1: break
if len(new) > 1:
splits.add(newgrp)
else:
abedline[line.accn] = line
return abedline, splits
def atg_name(name, retval="chr,rank", trimpad0=True):
seps = ["g", "te", "trna", "s", "u", "nc"]
pad0s = ["rank"]
if name is not None:
m = re.match(atg_name_pat, name)
if m is not None and m.group('sep').lower() in seps:
retvals = []
for grp in retval.split(","):
if grp == 'chr':
val = chr_number(m.group(grp))
else:
val = get_number(m.group(grp)) \
if trimpad0 and grp in pad0s \
else m.group(grp)
retvals.append(val)
return (x for x in retvals) if len(retvals) > 1 \
else retvals[0]
return (None for x in retval.split(","))
def collinear(args):
"""
%prog collinear a.b.anchors
Reduce synteny blocks to strictly collinear, use dynamic programming in a
procedure similar to DAGchainer.
"""
p = OptionParser(collinear.__doc__)
p.set_beds()
opts, args = p.parse_args(args)
if len(args) != 1:
sys.exit(not p.print_help())
(anchorfile, ) = args
qbed, sbed, qorder, sorder, is_self = check_beds(anchorfile, p, opts)
af = AnchorFile(anchorfile)
newanchorfile = anchorfile.rsplit(".", 1)[0] + ".collinear.anchors"
fw = open(newanchorfile, "w")
blocks = af.blocks
for block in blocks:
print("#" * 3, file=fw)
iblock = []
for q, s, score in block:
qi, q = qorder[q]
si, s = sorder[s]
score = get_number(score)
iblock.append([qi, si, score])
block = get_collinear(iblock)
for q, s, score in block:
q = qbed[q].accn
s = sbed[s].accn
print("\t".join((q, s, str(score))), file=fw)
fw.close()
def coge(args):
"""
%prog coge *.gff
Prepare coge datasets.
"""
p = OptionParser(coge.__doc__)
opts, args = p.parse_args(args)
if len(args) < 1:
sys.exit(not p.print_help())
gffs = args
for gff in gffs:
atoms = op.basename(gff).split(".")
gid = atoms[-2]
assert gid.startswith("gid")
gid = get_number(gid)
genomefasta = "genome_{0}.faa.fasta".format(gid)
species = "_".join(atoms[0].split("_")[:2])
cdsfasta = species + ".cds.fasta"
load([gff, genomefasta, "--id_attribute=Parent",
"--outfile={0}".format(cdsfasta)])
def atg_name(name, retval="chr,rank", trimpad0=True):
atg_name_pat = re.compile(r"""
^(?P<locus>
(?:(?P<prefix>\D+[\D\d\D])\.?)(?P<chr>[\d|C|M]+)?(?P<sep>[A-z]+)(?P<rank>\d+)
)
\.?(?P<iso>\d+)?
""", re.VERBOSE)
seps = ["g", "te", "trna", "s", "u"]
pad0s = ["chr", "rank"]
if name is not None:
m = re.match(atg_name_pat, name)
if m is not None and m.group('sep').lower() in seps:
retvals = []
for grp in retval.split(","):
val = get_number(m.group(grp)) \
if trimpad0 and grp in pad0s \
else m.group(grp)
retvals.append(val)
return (x for x in retvals)
else:
return (None for x in retval.split(","))
def multihistogram(args):
"""
%prog multihistogram *.histogram species
Plot the histogram based on a set of K-mer hisotograms. The method is based
on Star et al.'s method (Atlantic Cod genome paper).
"""
p = OptionParser(multihistogram.__doc__)
p.add_option("--kmin",
default=15,
type="int",
help="Minimum K-mer size, inclusive")
p.add_option("--kmax",
default=30,
type="int",
help="Maximum K-mer size, inclusive")
p.add_option("--vmin",
default=2,
type="int",
help="Minimum value, inclusive")
p.add_option("--vmax",
default=100,
type="int",
help="Maximum value, inclusive")
opts, args, iopts = p.set_image_options(args, figsize="10x5", dpi=300)
if len(args) < 1:
sys.exit(not p.print_help())
histfiles = args[:-1]
species = args[-1]
fig = plt.figure(1, (iopts.w, iopts.h))
root = fig.add_axes([0, 0, 1, 1])
A = fig.add_axes([0.08, 0.12, 0.38, 0.76])
B = fig.add_axes([0.58, 0.12, 0.38, 0.76])
lines = []
legends = []
genomesizes = []
for histfile in histfiles:
ks = KmerSpectrum(histfile)
x, y = ks.get_xy(opts.vmin, opts.vmax)
K = get_number(op.basename(histfile).split(".")[0].split("-")[-1])
if not opts.kmin <= K <= opts.kmax:
continue
(line, ) = A.plot(x, y, "-", lw=1)
lines.append(line)
legends.append("K = {0}".format(K))
ks.analyze(K=K, method="allpaths")
genomesizes.append((K, ks.genomesize / 1e6))
leg = A.legend(lines, legends, shadow=True, fancybox=True)
leg.get_frame().set_alpha(0.5)
title = "{0} genome K-mer histogram".format(species)
A.set_title(markup(title))
xlabel, ylabel = "Coverage (X)", "Counts"
A.set_xlabel(xlabel)
A.set_ylabel(ylabel)
set_human_axis(A)
title = "{0} genome size estimate".format(species)
B.set_title(markup(title))
x, y = zip(*genomesizes)
B.plot(x, y, "ko", mfc="w")
t = np.linspace(opts.kmin - 0.5, opts.kmax + 0.5, 100)
p = np.poly1d(np.polyfit(x, y, 2))
B.plot(t, p(t), "r:")
xlabel, ylabel = "K-mer size", "Estimated genome size (Mb)"
B.set_xlabel(xlabel)
B.set_ylabel(ylabel)
set_ticklabels_helvetica(B)
labels = ((0.04, 0.96, "A"), (0.54, 0.96, "B"))
panel_labels(root, labels)
normalize_axes(root)
imagename = species + ".multiK.pdf"
savefig(imagename, dpi=iopts.dpi, iopts=iopts)
def ancestral(args):
"""
%prog ancestral ancestral.txt assembly.fasta
Karyotype evolution of pineapple. The figure is inspired by Amphioxus paper
Figure 3 and Tetradon paper Figure 9.
"""
p = OptionParser(ancestral.__doc__)
opts, args, iopts = p.set_image_options(args, figsize="8x7")
if len(args) != 2:
sys.exit(not p.print_help())
regionsfile, sizesfile = args
regions = RegionsFile(regionsfile)
sizes = Sizes(sizesfile).mapping
sizes = dict((k, v) for (k, v) in sizes.iteritems() if k[:2] == "LG")
maxsize = max(sizes.values())
ratio = .5 / maxsize
fig = plt.figure(1, (iopts.w, iopts.h))
root = fig.add_axes((0, 0, 1, 1))
from jcvi.graphics.base import set2
a, b, c, d, e, f, g = set2[:7]
set2 = (c, g, b, e, d, a, f)
# Upper panel is the evolution of segments
# All segments belong to one of seven karyotypes 1 to 7
karyotypes = regions.karyotypes
xgap = 1. / (1 + len(karyotypes))
ygap = .05
mgap = xgap / 4.5
gwidth = mgap * .75
tip = .02
coords = {}
for i, k in enumerate(regions.karyotypes):
x = (i + 1) * xgap
y = .9
root.text(x, y + tip, "Anc" + k, ha="center")
root.plot((x, x), (y, y - ygap), "k-", lw=2)
y -= 2 * ygap
coords['a'] = (x - 1.5 * mgap , y)
coords['b'] = (x - .5 * mgap , y)
coords['c'] = (x + .5 * mgap , y)
coords['d'] = (x + 1.5 * mgap , y)
coords['ab'] = join_nodes_vertical(root, coords, 'a', 'b', y + ygap / 2)
coords['cd'] = join_nodes_vertical(root, coords, 'c', 'd', y + ygap / 2)
coords['abcd'] = join_nodes_vertical(root, coords, 'ab', 'cd', y + ygap)
for n in 'abcd':
nx, ny = coords[n]
root.text(nx, ny - tip, n, ha="center")
coords[n] = (nx, ny - ygap / 2)
kdata = regions.get_karyotype(k)
for kd in kdata:
g = kd.group
gx, gy = coords[g]
gsize = ratio * kd.span
gy -= gsize
p = Rectangle((gx - gwidth / 2, gy),
gwidth, gsize, lw=0, color=set2[i])
root.add_patch(p)
root.text(gx, gy + gsize / 2, kd.chromosome,
ha="center", va="center", color='w')
coords[g] = (gx, gy - tip)
# Bottom panel shows the location of segments on chromosomes
# TODO: redundant code, similar to graphics.chromosome
ystart = .54
chr_number = len(sizes)
xstart, xend = xgap - 2 * mgap, 1 - xgap + 2 * mgap
xinterval = (xend - xstart - gwidth) / (chr_number - 1)
chrpos = {}
for a, (chr, clen) in enumerate(sorted(sizes.items())):
chr = get_number(chr)
xx = xstart + a * xinterval + gwidth / 2
chrpos[chr] = xx
root.text(xx, ystart + .01, chr, ha="center")
Chromosome(root, xx, ystart, ystart - clen * ratio, width=gwidth)
# Start painting
for r in regions:
xx = chrpos[r.chromosome]
yystart = ystart - r.start * ratio
yyend = ystart - r.end * ratio
p = Rectangle((xx - gwidth / 2, yystart), gwidth, yyend - yystart,
color=set2[int(r.karyotype) - 1], lw=0)
root.add_patch(p)
root.set_xlim(0, 1)
root.set_ylim(0, 1)
root.set_axis_off()
pf = "pineapple-karyotype"
image_name = pf + "." + iopts.format
savefig(image_name, dpi=iopts.dpi, iopts=iopts)
def multihistogram(args):
"""
%prog multihistogram *.histogram species
Plot the histogram based on a set of K-mer hisotograms. The method is based
on Star et al.'s method (Atlantic Cod genome paper).
"""
p = OptionParser(multihistogram.__doc__)
p.add_option("--kmin", default=15, type="int", help="Minimum K-mer size, inclusive")
p.add_option("--kmax", default=30, type="int", help="Maximum K-mer size, inclusive")
p.add_option("--vmin", default=2, type="int", help="Minimum value, inclusive")
p.add_option("--vmax", default=100, type="int", help="Maximum value, inclusive")
opts, args, iopts = p.set_image_options(args, figsize="10x5", dpi=300)
histfiles = args[:-1]
species = args[-1]
fig = plt.figure(1, (iopts.w, iopts.h))
root = fig.add_axes([0, 0, 1, 1])
A = fig.add_axes([0.08, 0.12, 0.38, 0.76])
B = fig.add_axes([0.58, 0.12, 0.38, 0.76])
lines = []
legends = []
genomesizes = []
for histfile in histfiles:
ks = KmerSpectrum(histfile)
x, y = ks.get_xy(opts.vmin, opts.vmax)
K = get_number(op.basename(histfile).split(".")[0].split("-")[-1])
if not opts.kmin <= K <= opts.kmax:
continue
line, = A.plot(x, y, "-", lw=1)
lines.append(line)
legends.append("K = {0}".format(K))
ks.analyze(K=K)
genomesizes.append((K, ks.genomesize / 1e6))
leg = A.legend(lines, legends, shadow=True, fancybox=True)
leg.get_frame().set_alpha(0.5)
title = "{0} genome K-mer histogram".format(species)
A.set_title(markup(title))
xlabel, ylabel = "Coverage (X)", "Counts"
A.set_xlabel(xlabel)
A.set_ylabel(ylabel)
set_human_axis(A)
title = "{0} genome size estimate".format(species)
B.set_title(markup(title))
x, y = zip(*genomesizes)
B.plot(x, y, "ko", mfc="w")
t = np.linspace(opts.kmin - 0.5, opts.kmax + 0.5, 100)
p = np.poly1d(np.polyfit(x, y, 2))
B.plot(t, p(t), "r:")
xlabel, ylabel = "K-mer size", "Estimated genome size (Mb)"
B.set_xlabel(xlabel)
B.set_ylabel(ylabel)
set_ticklabels_helvetica(B)
labels = ((0.04, 0.96, "A"), (0.54, 0.96, "B"))
panel_labels(root, labels)
normalize_axes(root)
imagename = species + ".multiK.pdf"
savefig(imagename, dpi=iopts.dpi, iopts=iopts)
def renumber(args):
"""
%prog renumber Mt35.consolidated.bed > tagged.bed
Renumber genes for annotation updates.
"""
from jcvi.algorithms.lis import longest_increasing_subsequence
from jcvi.utils.grouper import Grouper
p = OptionParser(renumber.__doc__)
p.add_option("--pad0", default=6, type="int",
help="Pad gene identifiers with 0 [default: %default]")
p.add_option("--prefix", default="Medtr",
help="Genome prefix [default: %default]")
opts, args = p.parse_args(args)
if len(args) != 1:
sys.exit(not p.print_help())
bedfile, = args
pf = bedfile.rsplit(".", 1)[0]
abedfile = pf + ".a.bed"
bbedfile = pf + ".b.bed"
if need_update(bedfile, (abedfile, bbedfile)):
prepare(bedfile)
mbed = Bed(bbedfile)
g = Grouper()
for s in mbed:
accn = s.accn
g.join(*accn.split(";"))
bed = Bed(abedfile)
for chr, sbed in bed.sub_beds():
if "chr" not in chr:
continue
current_chr = get_number(chr)
ranks = []
gg = set()
for s in sbed:
accn = s.accn
achr, arank = atg_name(accn)
if achr != current_chr:
continue
ranks.append(arank)
gg.add(accn)
lranks = longest_increasing_subsequence(ranks)
print >> sys.stderr, current_chr, len(sbed), "==>", len(ranks), \
"==>", len(lranks)
granks = set(gene_name(current_chr, x) for x in lranks) | \
set(gene_name(current_chr, x, sep="te") for x in lranks)
tagstore = {}
for s in sbed:
achr, arank = atg_name(s.accn)
accn = s.accn
if accn in granks:
tag = (accn, FRAME)
elif accn in gg:
tag = (accn, RETAIN)
else:
tag = (".", NEW)
tagstore[accn] = tag
# Find cases where genes overlap
for s in sbed:
accn = s.accn
gaccn = g[accn]
tags = [((tagstore[x][-1] if x in tagstore else NEW), x) for x in gaccn]
group = [(PRIORITY.index(tag), x) for tag, x in tags]
best = min(group)[-1]
if accn != best:
tag = (best, OVERLAP)
else:
tag = tagstore[accn]
print "\t".join((str(s), "|".join(tag)))
def ancestral(args):
"""
%prog ancestral ancestral.txt assembly.fasta
Karyotype evolution of pineapple. The figure is inspired by Amphioxus paper
Figure 3 and Tetradon paper Figure 9.
"""
p = OptionParser(ancestral.__doc__)
opts, args, iopts = p.set_image_options(args, figsize="8x7")
if len(args) != 2:
sys.exit(not p.print_help())
regionsfile, sizesfile = args
regions = RegionsFile(regionsfile)
sizes = Sizes(sizesfile).mapping
sizes = dict((k, v) for (k, v) in sizes.iteritems() if k[:2] == "LG")
maxsize = max(sizes.values())
ratio = .5 / maxsize
fig = plt.figure(1, (iopts.w, iopts.h))
root = fig.add_axes((0, 0, 1, 1))
from jcvi.graphics.base import set2
a, b, c, d, e, f, g = set2[:7]
set2 = (c, g, b, e, d, a, f)
# Upper panel is the evolution of segments
# All segments belong to one of seven karyotypes 1 to 7
karyotypes = regions.karyotypes
xgap = 1. / (1 + len(karyotypes))
ygap = .05
mgap = xgap / 4.5
gwidth = mgap * .75
tip = .02
coords = {}
for i, k in enumerate(regions.karyotypes):
x = (i + 1) * xgap
y = .9
root.text(x, y + tip, "Anc" + k, ha="center")
root.plot((x, x), (y, y - ygap), "k-", lw=2)
y -= 2 * ygap
coords['a'] = (x - 1.5 * mgap , y)
coords['b'] = (x - .5 * mgap , y)
coords['c'] = (x + .5 * mgap , y)
coords['d'] = (x + 1.5 * mgap , y)
coords['ab'] = join_nodes_vertical(root, coords, 'a', 'b', y + ygap / 2)
coords['cd'] = join_nodes_vertical(root, coords, 'c', 'd', y + ygap / 2)
coords['abcd'] = join_nodes_vertical(root, coords, 'ab', 'cd', y + ygap)
for n in 'abcd':
nx, ny = coords[n]
root.text(nx, ny - tip, n, ha="center")
coords[n] = (nx, ny - ygap / 2)
kdata = regions.get_karyotype(k)
for kd in kdata:
g = kd.group
gx, gy = coords[g]
gsize = ratio * kd.span
gy -= gsize
p = Rectangle((gx - gwidth / 2, gy),
gwidth, gsize, lw=0, color=set2[i])
root.add_patch(p)
root.text(gx, gy + gsize / 2, kd.chromosome,
ha="center", va="center", color='w')
coords[g] = (gx, gy - tip)
# Bottom panel shows the location of segments on chromosomes
# TODO: redundant code, similar to graphics.chromosome
ystart = .54
chr_number = len(sizes)
xstart, xend = xgap - 2 * mgap, 1 - xgap + 2 * mgap
xinterval = (xend - xstart - gwidth) / (chr_number - 1)
chrpos = {}
for a, (chr, clen) in enumerate(sorted(sizes.items())):
chr = get_number(chr)
xx = xstart + a * xinterval + gwidth / 2
chrpos[chr] = xx
root.text(xx, ystart + .01, chr, ha="center")
Chromosome(root, xx, ystart, ystart - clen * ratio, width=gwidth)
# Start painting
for r in regions:
xx = chrpos[r.chromosome]
yystart = ystart - r.start * ratio
yyend = ystart - r.end * ratio
p = Rectangle((xx - gwidth / 2, yystart), gwidth, yyend - yystart,
color=set2[int(r.karyotype) - 1], lw=0)
root.add_patch(p)
root.set_xlim(0, 1)
root.set_ylim(0, 1)
root.set_axis_off()
pf = "pineapple-karyotype"
image_name = pf + "." + iopts.format
savefig(image_name, dpi=iopts.dpi, iopts=iopts)
def allocate(self, info, chr, start_id, end_id, id_table):
start_bp = info[0].start
end_bp = info[-1].end
current_chr = get_number(chr)
needed = info
assert end_id > start_id, \
"end ({0}) > start ({1})".format(end_id, start_id)
spots = end_id - start_id - 1
available = [x for x in xrange(start_id + 1, end_id) if
(current_chr, x) not in self.black]
message = "chr{0} need {1} ids, has {2} spots ({3} available)".\
format(current_chr, len(needed), spots, len(available))
start_gene = gene_name(current_chr, start_id)
end_gene = gene_name(current_chr, end_id)
message += " between {0} - {1}\n".format(start_gene, end_gene)
assert end_bp > start_bp
b = "\t".join(str(x) for x in (chr, start_bp - 1, end_bp))
cmd = "echo '{0}' |".format(b)
cmd += " intersectBed -a {0} -b stdin".format(self.gapfile)
gaps = list(BedLine(x) for x in popen(cmd, debug=False))
ngaps = len(gaps)
gapsexpanded = []
GeneDensity = 10000. # assume 10Kb per gene
for gap in gaps:
gap_bp = int(gap.score)
gap_ids = int(round(gap_bp / GeneDensity))
gapsexpanded += [gap] * gap_ids
lines = sorted(info + gapsexpanded, key=lambda x: x.start)
message += "between bp: {0} - {1}, there are {2} gaps (total {3} ids)".\
format(start_bp, end_bp, ngaps, len(lines))
needed = lines
stride = Stride(needed, available)
conf = stride.conf
message += " stride: {0}".format(conf)
print >> sys.stderr, message
nneeded = len(needed)
if conf is None: # prefix rule - prepend version number for spills
magic = 400000 # version 4
firstdigit = 100000
step = 10 # stride for the prefixed ids
rank = start_id + magic
if rank > magic + firstdigit:
rank -= firstdigit
available = []
while len(available) != nneeded:
rank += step
if (current_chr, rank) in self.black: # avoid blacklisted ids
continue
available.append(rank)
else: # follow the best stride
available = stride.available
if start_id == 0: # follow right flank at start of chr
available = available[- nneeded:]
else: # follow left flank otherwise
available = available[:nneeded]
# Finally assign the ids
assert len(needed) == len(available)
for b, rank in zip(needed, available):
name = gene_name(current_chr, rank)
print >> sys.stderr, "\t".join((str(b), name))
id_table[b.accn] = name
self.black.add((current_chr, rank))
print >> sys.stderr
def draw_chromosomes(
root,
bedfile,
sizes,
iopts,
mergedist,
winsize,
imagemap,
mappingfile=None,
gauge=False,
legend=True,
empty=False,
title=None,
):
bed = Bed(bedfile)
prefix = bedfile.rsplit(".", 1)[0]
if imagemap:
imgmapfile = prefix + ".map"
mapfh = open(imgmapfile, "w")
print('<map id="' + prefix + '">', file=mapfh)
if mappingfile:
mappings = DictFile(mappingfile, delimiter="\t")
classes = sorted(set(mappings.values()))
preset_colors = (DictFile(
mappingfile, keypos=1, valuepos=2, delimiter="\t")
if DictFile.num_columns(mappingfile) >= 3 else {})
else:
classes = sorted(set(x.accn for x in bed))
mappings = dict((x, x) for x in classes)
preset_colors = {}
logging.debug("A total of {} classes found: {}".format(
len(classes), ",".join(classes)))
# Assign colors to classes
ncolors = max(3, min(len(classes), 12))
palette = set1_n if ncolors <= 8 else set3_n
colorset = palette(number=ncolors)
colorset = sample_N(colorset, len(classes))
class_colors = dict(zip(classes, colorset))
class_colors.update(preset_colors)
logging.debug("Assigned colors: {}".format(class_colors))
chr_lens = {}
centromeres = {}
if sizes:
chr_lens = Sizes(sizes).sizes_mapping
else:
for b, blines in groupby(bed, key=(lambda x: x.seqid)):
blines = list(blines)
maxlen = max(x.end for x in blines)
chr_lens[b] = maxlen
for b in bed:
accn = b.accn
if accn == "centromere":
centromeres[b.seqid] = b.start
if accn in mappings:
b.accn = mappings[accn]
else:
b.accn = "-"
chr_number = len(chr_lens)
if centromeres:
assert chr_number == len(
centromeres), "chr_number = {}, centromeres = {}".format(
chr_number, centromeres)
r = 0.7 # width and height of the whole chromosome set
xstart, ystart = 0.15, 0.85
xinterval = r / chr_number
xwidth = xinterval * 0.5 # chromosome width
max_chr_len = max(chr_lens.values())
ratio = r / max_chr_len # canvas / base
# first the chromosomes
for a, (chr, clen) in enumerate(sorted(chr_lens.items())):
xx = xstart + a * xinterval + 0.5 * xwidth
root.text(xx, ystart + 0.01, str(get_number(chr)), ha="center")
if centromeres:
yy = ystart - centromeres[chr] * ratio
ChromosomeWithCentromere(root,
xx,
ystart,
yy,
ystart - clen * ratio,
width=xwidth)
else:
Chromosome(root, xx, ystart, ystart - clen * ratio, width=xwidth)
chr_idxs = dict((a, i) for i, a in enumerate(sorted(chr_lens.keys())))
alpha = 1
# color the regions
for chr in sorted(chr_lens.keys()):
segment_size, excess = 0, 0
bac_list = []
prev_end, prev_klass = 0, None
for b in bed.sub_bed(chr):
clen = chr_lens[chr]
idx = chr_idxs[chr]
klass = b.accn
if klass == "centromere":
continue
start = b.start
end = b.end
if start < prev_end + mergedist and klass == prev_klass:
start = prev_end
xx = xstart + idx * xinterval
yystart = ystart - end * ratio
yyend = ystart - start * ratio
root.add_patch(
Rectangle(
(xx, yystart),
xwidth,
yyend - yystart,
fc=class_colors.get(klass, "lightslategray"),
lw=0,
alpha=alpha,
))
prev_end, prev_klass = b.end, klass
if imagemap:
"""
`segment` : size of current BAC being investigated + `excess`
`excess` : left-over bases from the previous BAC, as a result of
iterating over `winsize` regions of `segment`
"""
if excess == 0:
segment_start = start
segment = (end - start + 1) + excess
while True:
if segment < winsize:
bac_list.append(b.accn)
excess = segment
break
segment_end = segment_start + winsize - 1
tlx, tly, brx, bry = (
xx,
(1 - ystart) + segment_start * ratio,
xx + xwidth,
(1 - ystart) + segment_end * ratio,
)
print(
"\t" + write_ImageMapLine(
tlx,
tly,
brx,
bry,
iopts.w,
iopts.h,
iopts.dpi,
chr + ":" + ",".join(bac_list),
segment_start,
segment_end,
),
file=mapfh,
)
segment_start += winsize
segment -= winsize
bac_list = []
if imagemap and excess > 0:
bac_list.append(b.accn)
segment_end = end
tlx, tly, brx, bry = (
xx,
(1 - ystart) + segment_start * ratio,
xx + xwidth,
(1 - ystart) + segment_end * ratio,
)
print(
"\t" + write_ImageMapLine(
tlx,
tly,
brx,
bry,
iopts.w,
iopts.h,
iopts.dpi,
chr + ":" + ",".join(bac_list),
segment_start,
segment_end,
),
file=mapfh,
)
if imagemap:
print("</map>", file=mapfh)
mapfh.close()
logging.debug("Image map written to `{0}`".format(mapfh.name))
if gauge:
xstart, ystart = 0.9, 0.85
Gauge(root, xstart, ystart - r, ystart, max_chr_len)
if "centromere" in class_colors:
del class_colors["centromere"]
# class legends, four in a row
if legend:
xstart = 0.1
xinterval = 0.8 / len(class_colors)
xwidth = 0.04
yy = 0.08
for klass, cc in sorted(class_colors.items()):
if klass == "-":
continue
root.add_patch(
Rectangle((xstart, yy),
xwidth,
xwidth,
fc=cc,
lw=0,
alpha=alpha))
root.text(xstart + xwidth + 0.01, yy, latex(klass), fontsize=10)
xstart += xinterval
if empty:
root.add_patch(
Rectangle((xstart, yy), xwidth, xwidth, fill=False, lw=1))
root.text(xstart + xwidth + 0.01, yy, empty, fontsize=10)
if title:
root.text(0.5, 0.95, markup(title), ha="center", va="center")