def wgd(args):
"""
%prog wgd vplanifoliaA_blocks.bed vplanifoliaA.sizes
Create a wgd figure.
"""
from jcvi.graphics.chromosome import draw_chromosomes
p = OptionParser(synteny.__doc__)
opts, args, iopts = p.set_image_options(args, figsize="8x5")
(bedfile, sizesfile) = args
fig = plt.figure(1, (iopts.w, iopts.h))
ax1 = fig.add_axes([0, 0, 1, 1])
title = r"Genome duplication $\alpha^{O}$ event in $\textit{Vanilla}$"
draw_chromosomes(
ax1,
bedfile,
sizes=sizesfile,
iopts=iopts,
mergedist=200000,
winsize=50000,
imagemap=False,
gauge=True,
legend=False,
title=title,
)
normalize_axes([ax1])
image_name = "wgd.pdf"
savefig(image_name, dpi=iopts.dpi, iopts=iopts)
def main(args):
"""
%prog table.csv
Render a table on canvas. Input is a CSV file.
"""
p = OptionParser(main.__doc__)
opts, args, iopts = p.set_image_options(args, figsize="7x7")
if len(args) != 1:
sys.exit(not p.print_help())
(csvfile, ) = args
pf = csvfile.rsplit(".", 1)[0]
fig = plt.figure(1, (iopts.w, iopts.h))
root = fig.add_axes([0, 0, 1, 1])
csv_table = CsvTable(csvfile)
draw_table(root, csv_table)
normalize_axes(root)
image_name = pf + "." + iopts.format
savefig(image_name, dpi=iopts.dpi, iopts=iopts)
def draw(self):
self.om1.draw()
self.om2.draw()
self.sax.set_xlim(0, self.om1.amax)
self.sax.set_ylim(-8, 8)
normalize_axes(self.ax)
self.sax.set_axis_off()
def draw(self, width=0.03):
HorizontalChromosome(
self.ax,
self.xpad,
1 - self.xpad,
self.ypad - 0.05,
height=width * 1.5,
patch=self.apatch,
lw=2,
)
Chromosome(
self.ax,
self.xpad - 0.05,
self.ypad,
1 - self.ypad,
width=width,
patch=self.bpatch,
patchcolor=self.bpatchcolor,
lw=2,
)
for a, b in zip(self.a, self.b):
self.sax.plot(a, b, "-", color="darkslategrey", lw=2)
self.sax.set_xticklabels([])
self.sax.set_yticklabels([])
self.sax.set_xlim((1, self.amax))
self.sax.set_ylim((1, self.bmax))
normalize_axes(self.ax)
def draw(self):
self.om1.draw()
self.om2.draw()
self.sax.set_xlim(0, self.om1.amax)
self.sax.set_ylim(-8, 8)
normalize_axes(self.ax)
self.sax.set_axis_off()
def allelefreq(args):
"""
%prog allelefreq HD,DM1,SCA1,SCA17
Plot the allele frequencies of some STRs.
"""
p = OptionParser(allelefreq.__doc__)
opts, args, iopts = p.set_image_options(args, figsize="10x10")
if len(args) != 1:
sys.exit(not p.print_help())
loci, = args
fig, ((ax1, ax2), (ax3, ax4)) = plt.subplots(ncols=2,
nrows=2,
figsize=(iopts.w, iopts.h))
plt.tight_layout(pad=4)
treds, df = read_treds()
df = df.set_index(["abbreviation"])
for ax, locus in zip((ax1, ax2, ax3, ax4), loci.split(",")):
plot_allelefreq(ax, df, locus)
root = fig.add_axes([0, 0, 1, 1])
pad = .03
panel_labels(root, ((pad / 2, 1 - pad, "A"), (1 / 2., 1 - pad, "B"),
(pad / 2, 1 / 2., "C"), (1 / 2., 1 / 2., "D")))
normalize_axes(root)
image_name = "allelefreq." + iopts.format
savefig(image_name, dpi=iopts.dpi, iopts=iopts)
def resample(args):
"""
%prog resample yellow-catfish-resample.txt medicago-resample.txt
Plot ALLMAPS performance across resampled real data.
"""
p = OptionParser(resample.__doc__)
opts, args, iopts = p.set_image_options(args, figsize="8x4", dpi=300)
if len(args) != 2:
sys.exit(not p.print_help())
dataA, dataB = args
fig = plt.figure(1, (iopts.w, iopts.h))
root = fig.add_axes([0, 0, 1, 1])
A = fig.add_axes([0.1, 0.18, 0.32, 0.64])
B = fig.add_axes([0.6, 0.18, 0.32, 0.64])
dataA = import_data(dataA)
dataB = import_data(dataB)
xlabel = "Fraction of markers"
ylabels = ("Anchor rate", "Runtime (m)")
legend = ("anchor rate", "runtime")
subplot_twinx(A, dataA, xlabel, ylabels, title="Yellow catfish", legend=legend)
subplot_twinx(B, dataB, xlabel, ylabels, title="Medicago", legend=legend)
labels = ((0.04, 0.92, "A"), (0.54, 0.92, "B"))
panel_labels(root, labels)
normalize_axes(root)
image_name = "resample." + iopts.format
savefig(image_name, dpi=iopts.dpi, iopts=iopts)
def likelihood2(args):
"""
%prog likelihood2 100_20.json
Plot the likelihood surface and marginal distributions.
"""
from matplotlib import gridspec
p = OptionParser(likelihood2.__doc__)
opts, args, iopts = p.set_image_options(args,
figsize="10x5",
style="white",
cmap="coolwarm")
if len(args) != 1:
sys.exit(not p.print_help())
jsonfile, = args
fig = plt.figure(figsize=(iopts.w, iopts.h))
gs = gridspec.GridSpec(2, 2)
ax1 = fig.add_subplot(gs[:, 0])
ax2 = fig.add_subplot(gs[0, 1])
ax3 = fig.add_subplot(gs[1, 1])
plt.tight_layout(pad=3)
pf = plot_panel(jsonfile, ax1, ax2, ax3, opts.cmap)
root = fig.add_axes([0, 0, 1, 1])
normalize_axes(root)
image_name = "likelihood2.{}.".format(pf) + iopts.format
savefig(image_name, dpi=iopts.dpi, iopts=iopts)
def resample(args):
"""
%prog resample yellow-catfish-resample.txt medicago-resample.txt
Plot ALLMAPS performance across resampled real data.
"""
p = OptionParser(resample.__doc__)
opts, args, iopts = p.set_image_options(args, figsize="8x4", dpi=300)
if len(args) != 2:
sys.exit(not p.print_help())
dataA, dataB = args
fig = plt.figure(1, (iopts.w, iopts.h))
root = fig.add_axes([0, 0, 1, 1])
A = fig.add_axes([.1, .18, .32, .64])
B = fig.add_axes([.6, .18, .32, .64])
dataA = import_data(dataA)
dataB = import_data(dataB)
xlabel = "Fraction of markers"
ylabels = ("Anchor rate", "Runtime (m)")
legend = ("anchor rate", "runtime")
subplot_twinx(A, dataA, xlabel, ylabels,
title="Yellow catfish", legend=legend)
subplot_twinx(B, dataB, xlabel, ylabels,
title="Medicago", legend=legend)
labels = ((.04, .92, "A"), (.54, .92, "B"))
panel_labels(root, labels)
normalize_axes(root)
image_name = "resample." + iopts.format
savefig(image_name, dpi=iopts.dpi, iopts=iopts)
def compare2(args):
"""
%prog compare2
Compare performances of various variant callers on simulated STR datasets.
"""
p = OptionParser(compare2.__doc__)
p.add_option('--maxinsert',
default=300,
type="int",
help="Maximum number of repeats")
add_simulate_options(p)
opts, args, iopts = p.set_image_options(args, figsize="10x5")
if len(args) != 0:
sys.exit(not p.print_help())
depth = opts.depth
readlen = opts.readlen
distance = opts.distance
max_insert = opts.maxinsert
fig, (ax1, ax2) = plt.subplots(ncols=2,
nrows=1,
figsize=(iopts.w, iopts.h))
plt.tight_layout(pad=2)
# ax1: lobSTR vs TREDPARSE with haploid model
lobstr_results = parse_results("lobstr_results_homo.txt")
tredparse_results = parse_results("tredparse_results_homo.txt")
title = SIMULATED_HAPLOID + \
r" ($D=%s\times, L=%dbp, V=%dbp$)" % (depth, readlen, distance)
plot_compare(ax1,
title,
tredparse_results,
lobstr_results,
max_insert=max_insert)
# ax2: lobSTR vs TREDPARSE with diploid model
lobstr_results = parse_results("lobstr_results_het.txt", exclude=20)
tredparse_results = parse_results("tredparse_results_het.txt", exclude=20)
title = SIMULATED_DIPLOID + \
r" ($D=%s\times, L=%dbp, V=%dbp$)" % (depth, readlen, distance)
plot_compare(ax2,
title,
tredparse_results,
lobstr_results,
max_insert=max_insert)
for ax in (ax1, ax2):
ax.set_xlim(0, max_insert)
ax.set_ylim(0, max_insert)
root = fig.add_axes([0, 0, 1, 1])
pad = .03
panel_labels(root, ((pad / 2, 1 - pad, "A"), (1 / 2., 1 - pad, "B")))
normalize_axes(root)
image_name = "tredparse." + iopts.format
savefig(image_name, dpi=iopts.dpi, iopts=iopts)
def compare(args):
"""
%prog compare Evaluation.csv
Compare performances of various variant callers on simulated STR datasets.
"""
p = OptionParser(compare.__doc__)
opts, args, iopts = p.set_image_options(args, figsize="10x10")
if len(args) != 1:
sys.exit(not p.print_help())
datafile, = args
pf = datafile.rsplit(".", 1)[0]
fig, ((ax1, ax2), (ax3, ax4)) = plt.subplots(ncols=2,
nrows=2,
figsize=(iopts.w, iopts.h))
plt.tight_layout(pad=3)
bbox = {'facecolor': 'tomato', 'alpha': .2, 'ec': 'w'}
pad = 2
# Read benchmark data
df = pd.read_csv("Evaluation.csv")
truth = df["Truth"]
axes = (ax1, ax2, ax3, ax4)
progs = ("Manta", "Isaac", "GATK", "lobSTR")
markers = ("bx-", "yo-", "md-", "c+-")
for ax, prog, marker in zip(axes, progs, markers):
ax.plot(truth, df[prog], marker)
ax.plot(truth, truth, 'k--') # to show diagonal
ax.axhline(infected_thr, color='tomato')
ax.text(max(truth) - pad,
infected_thr + pad,
'Risk threshold',
bbox=bbox,
ha="right")
ax.axhline(ref_thr, color='tomato')
ax.text(max(truth) - pad,
ref_thr - pad,
'Reference repeat count',
bbox=bbox,
ha="right",
va="top")
ax.set_title(SIMULATED_HAPLOID)
ax.set_xlabel(r'Num of CAG repeats inserted ($\mathit{h}$)')
ax.set_ylabel('Num of CAG repeats called')
ax.legend([prog, 'Truth'], loc='best')
root = fig.add_axes([0, 0, 1, 1])
pad = .03
panel_labels(root, ((pad / 2, 1 - pad, "A"), (1 / 2., 1 - pad, "B"),
(pad / 2, 1 / 2., "C"), (1 / 2., 1 / 2., "D")))
normalize_axes(root)
image_name = pf + "." + iopts.format
savefig(image_name, dpi=iopts.dpi, iopts=iopts)
def snpplot(args):
"""
%prog counts.cdt
Illustrate the histogram per SNP site.
"""
p = OptionParser(snpplot.__doc__)
opts, args, iopts = p.set_image_options(args, format="png")
if len(args) != 1:
sys.exit(not p.print_help())
(datafile,) = args
# Read in CDT file
fp = open(datafile)
next(fp)
next(fp)
data = []
for row in fp:
atoms = row.split()[4:]
nval = len(atoms)
values = [float(x) for x in atoms]
# normalize
values = [x * 1.0 / sum(values) for x in values]
data.append(values)
pf = datafile.rsplit(".", 1)[0]
fig = plt.figure(1, (iopts.w, iopts.h))
root = fig.add_axes([0, 0, 1, 1])
xmin, xmax = 0.1, 0.9
ymin, ymax = 0.1, 0.9
yinterval = (ymax - ymin) / len(data)
colors = "rbg" if nval == 3 else ["lightgray"] + list("rbg")
ystart = ymax
for d in data:
xstart = xmin
for dd, c in zip(d, colors):
xend = xstart + (xmax - xmin) * dd
root.plot((xstart, xend), (ystart, ystart), "-", color=c)
xstart = xend
ystart -= yinterval
root.text(
0.05,
0.5,
"{0} LMD50 SNPs".format(len(data)),
ha="center",
va="center",
rotation=90,
color="lightslategray",
)
for x, t, c in zip((0.3, 0.5, 0.7), ("REF", "ALT", "HET"), "rbg"):
root.text(x, 0.95, t, color=c, ha="center", va="center")
normalize_axes(root)
image_name = pf + "." + iopts.format
savefig(image_name, dpi=iopts.dpi, iopts=iopts)
def draw(self, width=.03):
HorizontalChromosome(self.ax, self.xpad, 1 - self.xpad,
self.ypad - width / 2, height=width * 1.5,
patch=self.apatch, lw=2)
for r in self.reads:
r.draw(self.sax)
self.sax.set_xlim((1, self.amax))
self.sax.set_ylim((-1, self.ymax))
normalize_axes(self.ax)
self.sax.set_axis_off()
def venn(args):
"""
%prog venn *.benchmark
Display benchmark results as Venn diagram.
"""
from matplotlib_venn import venn2
p = OptionParser(venn.__doc__)
opts, args, iopts = p.set_image_options(args, figsize="9x9")
if len(args) < 1:
sys.exit(not p.print_help())
bcs = args
fig = plt.figure(1, (iopts.w, iopts.h))
root = fig.add_axes([0, 0, 1, 1])
pad = .02
ystart = 1
ywidth = 1. / len(bcs)
tags = ("Bowers", "YGOB", "Schnable")
for bc, tag in zip(bcs, tags):
fp = open(bc)
data = []
for row in fp:
prog, pcounts, tcounts, shared = row.split()
pcounts = int(pcounts)
tcounts = int(tcounts)
shared = int(shared)
data.append((prog, pcounts, tcounts, shared))
xstart = 0
xwidth = 1. / len(data)
for prog, pcounts, tcounts, shared in data:
a, b, c = pcounts - shared, tcounts - shared, shared
ax = fig.add_axes([xstart + pad, ystart - ywidth + pad,
xwidth - 2 * pad, ywidth - 2 * pad])
venn2(subsets=(a, b, c), set_labels=(prog, tag), ax=ax)
message = "Sn={0} Pu={1}".\
format(percentage(shared, tcounts, precision=0, mode=-1),
percentage(shared, pcounts, precision=0, mode=-1))
print >> sys.stderr, message
ax.text(.5, .92, latex(message), ha="center", va="center",
transform=ax.transAxes, color='b')
ax.set_axis_off()
xstart += xwidth
ystart -= ywidth
panel_labels(root, ((.04, .96, "A"), (.04, .96 - ywidth, "B"),
(.04, .96 - 2 * ywidth, "C")))
panel_labels(root, ((.5, .98, "A. thaliana duplicates"),
(.5, .98 - ywidth, "14 Yeast genomes"),
(.5, .98 - 2 * ywidth, "4 Grass genomes")))
normalize_axes(root)
savefig("venn.pdf", dpi=opts.dpi)
def venn(args):
"""
%prog venn *.benchmark
Display benchmark results as Venn diagram.
"""
from matplotlib_venn import venn2
p = OptionParser(venn.__doc__)
opts, args, iopts = p.set_image_options(args, figsize="9x9")
if len(args) < 1:
sys.exit(not p.print_help())
bcs = args
fig = plt.figure(1, (iopts.w, iopts.h))
root = fig.add_axes([0, 0, 1, 1])
pad = .02
ystart = 1
ywidth = 1. / len(bcs)
tags = ("Bowers", "YGOB", "Schnable")
for bc, tag in zip(bcs, tags):
fp = open(bc)
data = []
for row in fp:
prog, pcounts, tcounts, shared = row.split()
pcounts = int(pcounts)
tcounts = int(tcounts)
shared = int(shared)
data.append((prog, pcounts, tcounts, shared))
xstart = 0
xwidth = 1. / len(data)
for prog, pcounts, tcounts, shared in data:
a, b, c = pcounts - shared, tcounts - shared, shared
ax = fig.add_axes([xstart + pad, ystart - ywidth + pad,
xwidth - 2 * pad, ywidth - 2 * pad])
venn2(subsets=(a, b, c), set_labels=(prog, tag), ax=ax)
message = "Sn={0} Pu={1}".\
format(percentage(shared, tcounts, precision=0, mode=-1),
percentage(shared, pcounts, precision=0, mode=-1))
print(message, file=sys.stderr)
ax.text(.5, .92, latex(message), ha="center", va="center",
transform=ax.transAxes, color='b')
ax.set_axis_off()
xstart += xwidth
ystart -= ywidth
panel_labels(root, ((.04, .96, "A"), (.04, .96 - ywidth, "B"),
(.04, .96 - 2 * ywidth, "C")))
panel_labels(root, ((.5, .98, "A. thaliana duplicates"),
(.5, .98 - ywidth, "14 Yeast genomes"),
(.5, .98 - 2 * ywidth, "4 Grass genomes")))
normalize_axes(root)
savefig("venn.pdf", dpi=opts.dpi)
def snpplot(args):
"""
%prog counts.cdt
Illustrate the histogram per SNP site.
"""
p = OptionParser(snpplot.__doc__)
opts, args, iopts = p.set_image_options(args, format="png")
if len(args) != 1:
sys.exit(not p.print_help())
datafile, = args
# Read in CDT file
fp = open(datafile)
next(fp)
next(fp)
data = []
for row in fp:
atoms = row.split()[4:]
nval = len(atoms)
values = [float(x) for x in atoms]
# normalize
values = [x * 1. / sum(values) for x in values]
data.append(values)
pf = datafile.rsplit(".", 1)[0]
fig = plt.figure(1, (iopts.w, iopts.h))
root = fig.add_axes([0, 0, 1, 1])
xmin, xmax = .1, .9
ymin, ymax = .1, .9
yinterval = (ymax - ymin) / len(data)
colors = "rbg" if nval == 3 else ["lightgray"] + list("rbg")
ystart = ymax
for d in data:
xstart = xmin
for dd, c in zip(d, colors):
xend = xstart + (xmax - xmin) * dd
root.plot((xstart, xend), (ystart, ystart), "-", color=c)
xstart = xend
ystart -= yinterval
root.text(.05, .5, "{0} LMD50 SNPs".format(len(data)),
ha="center", va="center", rotation=90, color="lightslategray")
for x, t, c in zip((.3, .5, .7), ("REF", "ALT", "HET"), "rbg"):
root.text(x, .95, t, color=c, ha="center", va="center")
normalize_axes(root)
image_name = pf + "." + iopts.format
savefig(image_name, dpi=iopts.dpi, iopts=iopts)
def draw(self, width=.03):
HorizontalChromosome(self.ax,
self.xpad,
1 - self.xpad,
self.ypad - width / 2,
height=width * 1.5,
patch=self.apatch,
lw=2)
for r in self.reads:
r.draw(self.sax)
self.sax.set_xlim((1, self.amax))
self.sax.set_ylim((-1, self.ymax))
normalize_axes(self.ax)
self.sax.set_axis_off()
def draw(self, width=.03):
HorizontalChromosome(self.ax, self.xpad, 1 - self.xpad,
self.ypad - .05, height=width * 1.5,
patch=self.apatch, lw=2)
Chromosome(self.ax, self.xpad - .05, self.ypad, 1 - self.ypad,
width=width, patch=self.bpatch,
patchcolor=self.bpatchcolor, lw=2)
for a, b in zip(self.a, self.b):
self.sax.plot(a, b, "-", color="darkslategrey", lw=2)
self.sax.set_xticklabels([])
self.sax.set_yticklabels([])
self.sax.set_xlim((1, self.amax))
self.sax.set_ylim((1, self.bmax))
normalize_axes(self.ax)
def movieframe(args):
"""
%prog movieframe tour test.clm contigs.ref.anchors
Draw heatmap and synteny in the same plot.
"""
p = OptionParser(movieframe.__doc__)
p.add_option("--label", help="Figure title")
p.set_beds()
p.set_outfile(outfile=None)
opts, args, iopts = p.set_image_options(args,
figsize="16x8",
style="white",
cmap="coolwarm",
format="png",
dpi=120)
if len(args) != 3:
sys.exit(not p.print_help())
tour, clmfile, anchorsfile = args
tour = tour.split(",")
image_name = opts.outfile or ("movieframe." + iopts.format)
label = opts.label or op.basename(image_name).rsplit(".", 1)[0]
clm = CLMFile(clmfile)
totalbins, bins, breaks = make_bins(tour, clm.tig_to_size)
M = read_clm(clm, totalbins, bins)
fig = plt.figure(1, (iopts.w, iopts.h))
root = fig.add_axes([0, 0, 1, 1]) # whole canvas
ax1 = fig.add_axes([.05, .1, .4, .8]) # heatmap
ax2 = fig.add_axes([.55, .1, .4, .8]) # dot plot
ax2_root = fig.add_axes([.5, 0, .5, 1]) # dot plot canvas
# Left axis: heatmap
plot_heatmap(ax1, M, breaks, iopts)
# Right axis: synteny
qbed, sbed, qorder, sorder, is_self = check_beds(anchorsfile,
p,
opts,
sorted=False)
dotplot(anchorsfile, qbed, sbed, fig, ax2_root, ax2, sep=False, title="")
root.text(.5, .98, clm.name, color="g", ha="center", va="center")
root.text(.5, .95, label, color="darkslategray", ha="center", va="center")
normalize_axes(root)
savefig(image_name, dpi=iopts.dpi, iopts=iopts)
def simulation(args):
"""
%prog simulation inversion.txt translocation.txt maps.txt multimaps.txt
Plot ALLMAPS accuracy across a range of simulated datasets.
"""
p = OptionParser(simulation.__doc__)
opts, args, iopts = p.set_image_options(args, dpi=300)
if len(args) != 4:
sys.exit(not p.print_help())
dataA, dataB, dataC, dataD = args
fig = plt.figure(1, (iopts.w, iopts.h))
root = fig.add_axes([0, 0, 1, 1])
A = fig.add_axes([0.12, 0.62, 0.35, 0.35])
B = fig.add_axes([0.62, 0.62, 0.35, 0.35])
C = fig.add_axes([0.12, 0.12, 0.35, 0.35])
D = fig.add_axes([0.62, 0.12, 0.35, 0.35])
dataA = import_data(dataA)
dataB = import_data(dataB)
dataC = import_data(dataC)
dataD = import_data(dataD)
subplot(A, dataA, "Inversion error rate", "Accuracy", xlim=0.5)
subplot(
B,
dataB,
"Translocation error rate",
"Accuracy",
xlim=0.5,
legend=("intra-chromosomal", "inter-chromosomal", "75\% intra + 25\% inter"),
)
subplot(C, dataC, "Number of input maps", "Accuracy", xcast=int)
subplot(D, dataD, "Number of input maps", "Accuracy", xcast=int)
labels = (
(0.03, 0.97, "A"),
(0.53, 0.97, "B"),
(0.03, 0.47, "C"),
(0.53, 0.47, "D"),
)
panel_labels(root, labels)
normalize_axes(root)
image_name = "simulation." + iopts.format
savefig(image_name, dpi=iopts.dpi, iopts=iopts)
def movieframe(args):
"""
%prog movieframe tour test.clm contigs.ref.anchors
Draw heatmap and synteny in the same plot.
"""
p = OptionParser(movieframe.__doc__)
p.add_option("--label", help="Figure title")
p.set_beds()
p.set_outfile(outfile=None)
opts, args, iopts = p.set_image_options(args, figsize="16x8",
style="white", cmap="coolwarm",
format="png", dpi=120)
if len(args) != 3:
sys.exit(not p.print_help())
tour, clmfile, anchorsfile = args
tour = tour.split(",")
image_name = opts.outfile or ("movieframe." + iopts.format)
label = opts.label or op.basename(image_name).rsplit(".", 1)[0]
clm = CLMFile(clmfile)
totalbins, bins, breaks = make_bins(tour, clm.tig_to_size)
M = read_clm(clm, totalbins, bins)
fig = plt.figure(1, (iopts.w, iopts.h))
root = fig.add_axes([0, 0, 1, 1]) # whole canvas
ax1 = fig.add_axes([.05, .1, .4, .8]) # heatmap
ax2 = fig.add_axes([.55, .1, .4, .8]) # dot plot
ax2_root = fig.add_axes([.5, 0, .5, 1]) # dot plot canvas
# Left axis: heatmap
plot_heatmap(ax1, M, breaks, iopts)
# Right axis: synteny
qbed, sbed, qorder, sorder, is_self = check_beds(anchorsfile, p, opts,
sorted=False)
dotplot(anchorsfile, qbed, sbed, fig, ax2_root, ax2, sep=False, title="")
root.text(.5, .98, clm.name, color="g", ha="center", va="center")
root.text(.5, .95, label, color="darkslategray", ha="center", va="center")
normalize_axes(root)
savefig(image_name, dpi=iopts.dpi, iopts=iopts)
def tree(args):
"""
%prog tree treefile
Create a tree figure.
"""
from jcvi.graphics.tree import parse_tree, LeafInfoFile, WGDInfoFile, draw_tree
p = OptionParser(tree.__doc__)
opts, args, iopts = p.set_image_options(args, figsize="10x8")
(datafile, ) = args
logging.debug("Load tree file `{0}`".format(datafile))
t, hpd = parse_tree(datafile)
fig = plt.figure(1, (iopts.w, iopts.h))
ax1 = fig.add_axes([0, 0, 1, 1])
supportcolor = "k"
margin, rmargin = 0.1, 0.2 # Left and right margin
leafinfo = LeafInfoFile("leafinfo.csv").cache
wgdinfo = WGDInfoFile("wgdinfo.csv").cache
outgroup = "ginkgo"
# Panel A
draw_tree(
ax1,
t,
hpd=hpd,
margin=margin,
rmargin=rmargin,
supportcolor=None,
internal=False,
outgroup=outgroup,
reroot=False,
leafinfo=leafinfo,
wgdinfo=wgdinfo,
geoscale=True,
)
normalize_axes([ax1])
image_name = "tree.pdf"
savefig(image_name, dpi=iopts.dpi, iopts=iopts)
def synteny(args):
"""
%prog synteny vplanifoliaA_blocks.bed vplanifoliaA.sizes \
b1.blocks all.bed b1.layout
Create a composite figure with (A) wgd and (B) microsynteny.
"""
from jcvi.graphics.chromosome import draw_chromosomes
p = OptionParser(synteny.__doc__)
opts, args, iopts = p.set_image_options(args, figsize="12x12")
(bedfile, sizesfile, blocksfile, allbedfile, blockslayout) = args
fig = plt.figure(1, (iopts.w, iopts.h))
root = fig.add_axes([0, 0, 1, 1])
ax1 = fig.add_axes([0, 0.5, 1, 0.5])
ax2 = fig.add_axes([0.02, 0, 0.98, 0.5])
# Panel A
title = r"Genome duplication $\alpha^{O}$ event in $\textit{Vanilla}$"
draw_chromosomes(
ax1,
bedfile,
sizes=sizesfile,
iopts=iopts,
mergedist=200000,
winsize=50000,
imagemap=False,
gauge=True,
legend=False,
title=title,
)
# Panel B
draw_ploidy(fig, ax2, blocksfile, allbedfile, blockslayout)
normalize_axes([root, ax1, ax2])
labels = ((0.05, 0.95, "A"), (0.05, 0.5, "B"))
panel_labels(root, labels)
image_name = "synteny.pdf"
savefig(image_name, dpi=iopts.dpi, iopts=iopts)
def microsynteny(args):
"""
%prog microsynteny b1.blocks all.bed b1.layout
Create a microsynteny figure.
"""
p = OptionParser(synteny.__doc__)
opts, args, iopts = p.set_image_options(args, figsize="12x6")
(blocksfile, allbedfile, blockslayout) = args
fig = plt.figure(1, (iopts.w, iopts.h))
ax2 = fig.add_axes([0, 0, 1, 1])
draw_ploidy(fig, ax2, blocksfile, allbedfile, blockslayout)
normalize_axes([ax2])
image_name = "microsynteny.pdf"
savefig(image_name, dpi=iopts.dpi, iopts=iopts)
def simulation(args):
"""
%prog simulation inversion.txt translocation.txt maps.txt multimaps.txt
Plot ALLMAPS accuracy across a range of simulated datasets.
"""
p = OptionParser(simulation.__doc__)
opts, args, iopts = p.set_image_options(args, dpi=300)
if len(args) != 4:
sys.exit(not p.print_help())
dataA, dataB, dataC, dataD = args
fig = plt.figure(1, (iopts.w, iopts.h))
root = fig.add_axes([0, 0, 1, 1])
A = fig.add_axes([.12, .62, .35, .35])
B = fig.add_axes([.62, .62, .35, .35])
C = fig.add_axes([.12, .12, .35, .35])
D = fig.add_axes([.62, .12, .35, .35])
dataA = import_data(dataA)
dataB = import_data(dataB)
dataC = import_data(dataC)
dataD = import_data(dataD)
subplot(A, dataA, "Inversion error rate", "Accuracy", xlim=.5)
subplot(B, dataB, "Translocation error rate", "Accuracy", xlim=.5,
legend=("intra-chromosomal", "inter-chromosomal",
"75\% intra + 25\% inter"))
subplot(C, dataC, "Number of input maps", "Accuracy", xcast=int)
subplot(D, dataD, "Number of input maps", "Accuracy", xcast=int)
labels = ((.03, .97, "A"), (.53, .97, "B"),
(.03, .47, "C"), (.53, .47, "D"))
panel_labels(root, labels)
normalize_axes(root)
image_name = "simulation." + iopts.format
savefig(image_name, dpi=iopts.dpi, iopts=iopts)
def likelihood3(args):
"""
%prog likelihood2 200_20.json 200_100.json
Plot the likelihood surface and marginal distributions for two settings.
"""
from matplotlib import gridspec
p = OptionParser(likelihood3.__doc__)
opts, args, iopts = p.set_image_options(args,
figsize="10x10",
style="white",
cmap="coolwarm")
if len(args) != 2:
sys.exit(not p.print_help())
jsonfile1, jsonfile2 = args
fig = plt.figure(figsize=(iopts.w, iopts.h))
gs = gridspec.GridSpec(9, 2)
ax1 = fig.add_subplot(gs[:4, 0])
ax2 = fig.add_subplot(gs[:2, 1])
ax3 = fig.add_subplot(gs[2:4, 1])
ax4 = fig.add_subplot(gs[5:, 0])
ax5 = fig.add_subplot(gs[5:7, 1])
ax6 = fig.add_subplot(gs[7:, 1])
plt.tight_layout(pad=2)
plot_panel(jsonfile1, ax1, ax2, ax3, opts.cmap)
plot_panel(jsonfile2, ax4, ax5, ax6, opts.cmap)
root = fig.add_axes([0, 0, 1, 1])
pad = .02
panel_labels(root, ((pad, 1 - pad, "A"), (pad, 4. / 9, "B")))
normalize_axes(root)
image_name = "likelihood3." + iopts.format
savefig(image_name, dpi=iopts.dpi, iopts=iopts)
def lms(args):
"""
%prog lms
ALLMAPS cartoon to illustrate LMS metric.
"""
from random import randint
from jcvi.graphics.chromosome import HorizontalChromosome
p = OptionParser(lms.__doc__)
opts, args, iopts = p.set_image_options(args, figsize="6x6", dpi=300)
fig = plt.figure(1, (iopts.w, iopts.h))
root = fig.add_axes([0, 0, 1, 1])
# Panel A
w, h = .7, .35
ax = fig.add_axes([.15, .6, w, h])
xdata = [x + randint(-3, 3) for x in range(10, 110, 10)]
ydata = [x + randint(-3, 3) for x in range(10, 110, 10)]
ydata[3:7] = ydata[3:7][::-1]
xydata = zip(xdata, ydata)
lis = xydata[:3] + [xydata[4]] + xydata[7:]
lds = xydata[3:7]
xlis, ylis = zip(*lis)
xlds, ylds = zip(*lds)
ax.plot(xlis, ylis, "r-", lw=12, alpha=.3,
solid_capstyle="round", solid_joinstyle="round")
ax.plot(xlds, ylds, "g-", lw=12, alpha=.3,
solid_capstyle="round", solid_joinstyle="round")
ax.plot(xdata, ydata, "k.", mec="k", mfc="w", mew=3, ms=12)
HorizontalChromosome(root, .15, .15 + w, .57, height=.02, lw=2)
root.text(.15 + w / 2, .55, "Chromosome location (bp)", ha="center", va="top")
ax.text(80, 30, "LIS = 7", color="r", ha="center", va="center")
ax.text(80, 20, "LDS = 4", color="g", ha="center", va="center")
ax.text(80, 10, "LMS = $max$(LIS, LDS) = 7", ha="center", va="center")
normalize_lms_axis(ax)
# Panel B
w = .37
p = (0, 45, 75, 110)
ax = fig.add_axes([.1, .12, w, h])
xdata = [x for x in range(10, 110, 10)]
ydata = ydata_orig = [x for x in range(10, 110, 10)]
ydata = ydata[:4] + ydata[7:] + ydata[4:7][::-1]
xydata = zip(xdata, ydata)
lis = xydata[:7]
xlis, ylis = zip(*lis)
ax.plot(xlis, ylis, "r-", lw=12, alpha=.3,
solid_capstyle="round", solid_joinstyle="round")
ax.plot(xdata, ydata, "k.", mec="k", mfc="w", mew=3, ms=12)
ax.vlines(p, 0, 110, colors="beige", lw=3)
normalize_lms_axis(ax)
patch = [.1 + w * x / 110. for x in p]
HorizontalChromosome(root, .1, .1 + w, .09, patch=patch,
height=.02, lw=2)
scaffolds = ("a", "b", "c")
for i, s in enumerate(scaffolds):
xx = (patch[i] + patch[i + 1]) / 2
root.text(xx, .09, s, va="center", ha="center")
root.text(.1 + w / 2, .04, "LMS($a||b||c$) = 7", ha="center")
# Panel C
ax = fig.add_axes([.6, .12, w, h])
patch = [.6 + w * x / 110. for x in p]
ydata = ydata_orig
ax.plot(xdata, ydata, "r-", lw=12, alpha=.3,
solid_capstyle="round", solid_joinstyle="round")
ax.plot(xdata, ydata, "k.", mec="k", mfc="w", mew=3, ms=12)
ax.vlines(p, [0], [110], colors="beige", lw=3)
normalize_lms_axis(ax)
HorizontalChromosome(root, .6, .6 + w, .09, patch=patch,
height=.02, lw=2)
scaffolds = ("a", "-c", "b")
for i, s in enumerate(scaffolds):
xx = (patch[i] + patch[i + 1]) / 2
root.text(xx, .09, s, va="center", ha="center")
root.text(.6 + w / 2, .04, "LMS($a||-c||b$) = 10", ha="center")
labels = ((.05, .95, 'A'), (.05, .48, 'B'), (.55, .48, 'C'))
panel_labels(root, labels)
normalize_axes(root)
pf = "lms"
image_name = pf + "." + iopts.format
savefig(image_name, dpi=iopts.dpi, iopts=iopts)
def wheel(args):
"""
%prog wheel datafile.csv groups.csv
Wheel plot that shows continous data in radial axes.
"""
p = OptionParser(wheel.__doc__)
p.add_option("--column", default="score", choices=("score", "percentile"),
help="Which column to extract from `datafile.csv`")
opts, args, iopts = p.set_image_options(args, figsize="5x5", format="png")
if len(args) != 2:
sys.exit(not p.print_help())
datafile, groupsfile = args
column = opts.column
linecolor = "#d6d6d6"
df = parse_data(datafile, score_column=opts.column)
groups = parse_groups(groupsfile)
labels = [g for g in groups if g in df]
print labels
df = [df[g] for g in labels]
print df
groups = [groups[g] for g in labels]
print groups
pf = datafile.rsplit(".", 1)[0]
fig = plt.figure(1, (iopts.w, iopts.h))
root = fig.add_axes([0, 0, 1, 1])
categories = len(df)
#ax = plt.subplot(111, projection='polar')
ax = fig.add_axes([0.1, 0.1, 0.8, 0.8], polar=True)
brewer = [ "#FF3B30", "#DD43A0", "#5856D6",
"#007AFE", "#56BDEC", "#4CD8BA",
"#4CD864", "#B0F457", "#FEF221",
"#FFCC01", "#FF9500", "#FF3B30",
]
# Baseline
theta = np.linspace(1.5 * np.pi, 3.5 * np.pi, endpoint=False, num=categories)
_theta = np.linspace(1.5 * np.pi, 3.5 * np.pi)
R = max(max(df), 10)
xlim = (-R, R) if column == "score" else (-100, 100)
plim = (-R / 2, R) if column == "score" else (0, 100)
ci = (-.5, 2) if column == "score" else (10, 90)
# Grid
if column == "score":
for t in theta:
ax.plot([t, t], plim, color=linecolor)
ax.axis('off')
# Contours
for t in plim:
ax.plot(_theta, [t] * len(_theta), color=linecolor)
# Sectors (groupings)
collapsed_groups = []
gg = []
for group, c in groupby(enumerate(groups), lambda x: x[1]):
c = [x[0] for x in list(c)]
collapsed_groups.append(group)
gg.append(c)
sector = False
if sector:
theta_interval = 2 * np.pi / categories
theta_pad = theta_interval / 2 * .9
for color, group in zip(brewer, gg):
tmin, tmax = min(group), max(group)
sector(ax, theta[tmin], theta[tmax], theta_pad, R * .95,
"-", color=color, lw=2)
# Data
r = df
closed_plot(ax, theta, r, color="lightslategray", alpha=.25)
all_data = []
for color, group in zip(brewer, gg):
hidden_data = [(theta[x], r[x]) for x in group if \
(ci[0] <= r[x] <= ci[1])]
shown_data = [(theta[x], r[x]) for x in group if (r[x] < ci[0] \
or r[x] > ci[1])]
all_data.append((theta[x], labels[x], r[x]))
for alpha, data in zip((1, 1), (hidden_data, shown_data)):
if not data:
continue
color_theta, color_r = zip(*data)
ax.plot(color_theta, color_r, "o", color=color, alpha=alpha)
# Print out data
diseaseNames, risks = labels, df
print "var theta = [{}]".format(",".join("{:.1f}".format(degrees(x)) for x in theta))
print "var risks = [{}]".format(",".join(str(x) for x in risks))
print "var diseaseNames = [{}]".format(",".join(\
['"{}"'.format(x) for x in diseaseNames]))
# Labels
from math import cos, sin
r = .5
for i, label in enumerate(labels):
tl = theta[i]
x, y = .5 + r * cos(tl), .5 + r * sin(tl)
d = degrees(tl)
if 90 < d % 360 < 270: # On the left quardrants
d -= 180
root.text(x, y, label, size=4, rotation=d,
ha="center", va="center", color=linecolor)
print x, y, label
# Add baseline
baseline = 0 if column == "score" else 50
_r = len(_theta) * [baseline]
closed_plot(ax, _theta, _r, "k:", lw=1, ms=4)
# Add confidence interval
if column == "percentile":
barcolor = "#eeeeee"
ax.bar([0], [ci[1] - ci[0]], width=2 * np.pi, bottom=ci[0], fc=barcolor)
ax.set_rmin(xlim[0])
ax.set_rmax(xlim[1])
normalize_axes(root)
image_name = pf + "-" + column + "." + iopts.format
savefig(image_name, dpi=iopts.dpi, iopts=iopts)
def seeds(args):
"""
%prog seeds [pngfile|jpgfile]
Extract seed metrics from [pngfile|jpgfile]. Use --rows and --cols to crop image.
"""
p = OptionParser(seeds.__doc__)
p.set_outfile()
opts, args, iopts = add_seeds_options(p, args)
if len(args) != 1:
sys.exit(not p.print_help())
pngfile, = args
pf = opts.prefix or op.basename(pngfile).rsplit(".", 1)[0]
sigma, kernel = opts.sigma, opts.kernel
rows, cols = opts.rows, opts.cols
labelrows, labelcols = opts.labelrows, opts.labelcols
ff = opts.filter
calib = opts.calibrate
outdir = opts.outdir
if outdir != '.':
mkdir(outdir)
if calib:
calib = json.load(must_open(calib))
pixel_cm_ratio, tr = calib["PixelCMratio"], calib["RGBtransform"]
tr = np.array(tr)
resizefile, mainfile, labelfile, exif = \
convert_image(pngfile, pf, outdir=outdir,
rotate=opts.rotate,
rows=rows, cols=cols,
labelrows=labelrows, labelcols=labelcols)
oimg = load_image(resizefile)
img = load_image(mainfile)
fig, (ax1, ax2, ax3, ax4) = plt.subplots(ncols=4, nrows=1,
figsize=(iopts.w, iopts.h))
# Edge detection
img_gray = rgb2gray(img)
logging.debug("Running {0} edge detection ...".format(ff))
if ff == "canny":
edges = canny(img_gray, sigma=opts.sigma)
elif ff == "roberts":
edges = roberts(img_gray)
elif ff == "sobel":
edges = sobel(img_gray)
edges = clear_border(edges, buffer_size=opts.border)
selem = disk(kernel)
closed = closing(edges, selem) if kernel else edges
filled = binary_fill_holes(closed)
# Watershed algorithm
if opts.watershed:
distance = distance_transform_edt(filled)
local_maxi = peak_local_max(distance, threshold_rel=.05, indices=False)
coordinates = peak_local_max(distance, threshold_rel=.05)
markers, nmarkers = label(local_maxi, return_num=True)
logging.debug("Identified {0} watershed markers".format(nmarkers))
labels = watershed(closed, markers, mask=filled)
else:
labels = label(filled)
# Object size filtering
w, h = img_gray.shape
canvas_size = w * h
min_size = int(round(canvas_size * opts.minsize / 100))
max_size = int(round(canvas_size * opts.maxsize / 100))
logging.debug("Find objects with pixels between {0} ({1}%) and {2} ({3}%)"\
.format(min_size, opts.minsize, max_size, opts.maxsize))
# Plotting
ax1.set_title('Original picture')
ax1.imshow(oimg)
params = "{0}, $\sigma$={1}, $k$={2}".format(ff, sigma, kernel)
if opts.watershed:
params += ", watershed"
ax2.set_title('Edge detection\n({0})'.format(params))
closed = gray2rgb(closed)
ax2_img = labels
if opts.edges:
ax2_img = closed
elif opts.watershed:
ax2.plot(coordinates[:, 1], coordinates[:, 0], 'g.')
ax2.imshow(ax2_img, cmap=iopts.cmap)
ax3.set_title('Object detection')
ax3.imshow(img)
filename = op.basename(pngfile)
if labelfile:
accession = extract_label(labelfile)
else:
accession = pf
# Calculate region properties
rp = regionprops(labels)
rp = [x for x in rp if min_size <= x.area <= max_size]
nb_labels = len(rp)
logging.debug("A total of {0} objects identified.".format(nb_labels))
objects = []
for i, props in enumerate(rp):
i += 1
if i > opts.count:
break
y0, x0 = props.centroid
orientation = props.orientation
major, minor = props.major_axis_length, props.minor_axis_length
major_dx = cos(orientation) * major / 2
major_dy = sin(orientation) * major / 2
minor_dx = sin(orientation) * minor / 2
minor_dy = cos(orientation) * minor / 2
ax2.plot((x0 - major_dx, x0 + major_dx),
(y0 + major_dy, y0 - major_dy), 'r-')
ax2.plot((x0 - minor_dx, x0 + minor_dx),
(y0 - minor_dy, y0 + minor_dy), 'r-')
npixels = int(props.area)
# Sample the center of the blob for color
d = min(int(round(minor / 2 * .35)) + 1, 50)
x0d, y0d = int(round(x0)), int(round(y0))
square = img[(y0d - d):(y0d + d), (x0d - d):(x0d + d)]
pixels = []
for row in square:
pixels.extend(row)
logging.debug("Seed #{0}: {1} pixels ({2} sampled) - {3:.2f}%".\
format(i, npixels, len(pixels), 100. * npixels / canvas_size))
rgb = pixel_stats(pixels)
objects.append(Seed(filename, accession, i, rgb, props, exif))
minr, minc, maxr, maxc = props.bbox
rect = Rectangle((minc, minr), maxc - minc, maxr - minr,
fill=False, ec='w', lw=1)
ax3.add_patch(rect)
mc, mr = (minc + maxc) / 2, (minr + maxr) / 2
ax3.text(mc, mr, "{0}".format(i), color='w',
ha="center", va="center", size=6)
for ax in (ax2, ax3):
ax.set_xlim(0, h)
ax.set_ylim(w, 0)
# Output identified seed stats
ax4.text(.1, .92, "File: {0}".format(latex(filename)), color='g')
ax4.text(.1, .86, "Label: {0}".format(latex(accession)), color='m')
yy = .8
fw = must_open(opts.outfile, "w")
if not opts.noheader:
print(Seed.header(calibrate=calib), file=fw)
for o in objects:
if calib:
o.calibrate(pixel_cm_ratio, tr)
print(o, file=fw)
i = o.seedno
if i > 7:
continue
ax4.text(.01, yy, str(i), va="center", bbox=dict(fc='none', ec='k'))
ax4.text(.1, yy, o.pixeltag, va="center")
yy -= .04
ax4.add_patch(Rectangle((.1, yy - .025), .12, .05, lw=0,
fc=rgb_to_hex(o.rgb)))
ax4.text(.27, yy, o.hashtag, va="center")
yy -= .06
ax4.text(.1 , yy, "(A total of {0} objects displayed)".format(nb_labels),
color="darkslategrey")
normalize_axes(ax4)
for ax in (ax1, ax2, ax3):
xticklabels = [int(x) for x in ax.get_xticks()]
yticklabels = [int(x) for x in ax.get_yticks()]
ax.set_xticklabels(xticklabels, family='Helvetica', size=8)
ax.set_yticklabels(yticklabels, family='Helvetica', size=8)
image_name = op.join(outdir, pf + "." + iopts.format)
savefig(image_name, dpi=iopts.dpi, iopts=iopts)
return objects
def compare(args):
"""
%prog compare Evaluation.csv
Compare performances of various variant callers on simulated STR datasets.
"""
p = OptionParser(__doc__)
opts, args, iopts = p.set_image_options(args, figsize="15x5")
if len(args) != 1:
sys.exit(not p.print_help())
datafile, = args
pf = datafile.rsplit(".", 1)[0]
fig, (ax1, ax2, ax3) = plt.subplots(ncols=3,
nrows=1,
figsize=(iopts.w, iopts.h))
plt.tight_layout(pad=2)
# Huntington risk allele
infected_thr = 40
ref_thr = 19
# ax1: Multiple callers at lower range
df = pd.read_csv("Evaluation.csv")
truth = df["Truth"]
ax1.plot(truth, df["Manta"], 'bx-')
ax1.plot(truth, df["Isaac"], 'yo-')
ax1.plot(truth, df["GATK"], 'md-')
ax1.plot(truth, df["lobSTR"], 'c+-')
ax1.plot(truth, truth, 'k--') # to show diagonal
bbox = {'facecolor': 'tomato', 'alpha': .2, 'ec': 'w'}
pad = 2
ax1.axhline(infected_thr, color='tomato')
ax1.text(max(truth) - pad,
infected_thr + pad,
'Risk threshold',
bbox=bbox,
ha="right")
ax1.axhline(ref_thr, color='tomato')
ax1.text(max(truth) - pad,
ref_thr - pad,
'Reference repeat count',
bbox=bbox,
ha="right",
va="top")
ax1.set_xlabel(r'Num of CAG repeats inserted ($\mathit{h}$)')
ax1.set_ylabel('Num of CAG repeats called')
ax1.set_title(r'Simulated haploid $\mathit{h}$')
ax1.legend(['Manta', 'Isaac', 'GATK', 'lobSTR', 'Truth'], loc='best')
max_insert = 120
# ax2: lobSTR vs TREDPARSE with haploid model
lobstr_results = parse_results("lobstr_results_homo.txt")
tredparse_results = parse_results("tredparse_results_homo.txt")
truth = range(10, max_insert + 1)
lx, ly = zip(*lobstr_results)
tx, ty = zip(*tredparse_results)
ax2.plot(lx, ly, 'c+-')
ax2.plot(tx, ty, 'gx-')
ax2.plot(truth, truth, 'k--')
ax2.set_xlabel(r'Num of CAG repeats inserted ($\mathit{h}$)')
ax2.set_ylabel('Num of CAG repeats called')
ax2.set_title(r'Simulated haploid $\mathit{h}$')
ax2.legend(['lobSTR', 'TREDPARSE', 'Truth'], loc='best')
pad *= 2
ax2.axhline(infected_thr, color='tomato')
ax2.text(max(truth) - pad,
infected_thr + pad,
'Risk threshold',
bbox=bbox,
ha="right")
ax2.set_xlim(10, max_insert)
# ax3: lobSTR vs TREDPARSE with haploid model
lobstr_results = parse_results("lobstr_results_het.txt", exclude=20)
tredparse_results = parse_results("tredparse_results_het.txt", exclude=20)
truth = range(10, max_insert + 1)
lx, ly = zip(*lobstr_results)
tx, ty = zip(*tredparse_results)
ax3.plot(lx, ly, 'c+-')
ax3.plot(tx, ty, 'gx-')
ax3.plot(truth, truth, 'k--')
ax3.set_xlabel(r'Num of CAG repeats inserted ($\mathit{h}$)')
ax3.set_ylabel('Num of CAG repeats called')
ax3.set_title(r'Simulated diploid $\mathit{20/h}$')
ax3.legend(['lobSTR', 'TREDPARSE', 'Truth'], loc='best')
ax3.axhline(infected_thr, color='tomato')
ax3.text(max(truth) - pad,
infected_thr + pad,
'Risk threshold',
bbox=bbox,
ha="right")
ax3.set_xlim(10, max_insert)
root = fig.add_axes([0, 0, 1, 1])
pad = .03
panel_labels(root, ((pad / 2, 1 - pad, "A"), (1 / 3., 1 - pad, "B"),
(2 / 3., 1 - pad, "C")))
normalize_axes(root)
image_name = pf + "." + iopts.format
savefig(image_name, dpi=iopts.dpi, iopts=iopts)
def lms(args):
"""
%prog lms
ALLMAPS cartoon to illustrate LMS metric.
"""
from random import randint
from jcvi.graphics.chromosome import HorizontalChromosome
p = OptionParser(lms.__doc__)
opts, args, iopts = p.set_image_options(args, figsize="6x6", dpi=300)
fig = plt.figure(1, (iopts.w, iopts.h))
root = fig.add_axes([0, 0, 1, 1])
# Panel A
w, h = 0.7, 0.35
ax = fig.add_axes([0.15, 0.6, w, h])
xdata = [x + randint(-3, 3) for x in range(10, 110, 10)]
ydata = [x + randint(-3, 3) for x in range(10, 110, 10)]
ydata[3:7] = ydata[3:7][::-1]
xydata = zip(xdata, ydata)
lis = xydata[:3] + [xydata[4]] + xydata[7:]
lds = xydata[3:7]
xlis, ylis = zip(*lis)
xlds, ylds = zip(*lds)
ax.plot(
xlis,
ylis,
"r-",
lw=12,
alpha=0.3,
solid_capstyle="round",
solid_joinstyle="round",
)
ax.plot(
xlds,
ylds,
"g-",
lw=12,
alpha=0.3,
solid_capstyle="round",
solid_joinstyle="round",
)
ax.plot(xdata, ydata, "k.", mec="k", mfc="w", mew=3, ms=12)
HorizontalChromosome(root, 0.15, 0.15 + w, 0.57, height=0.02, lw=2)
root.text(0.15 + w / 2, 0.55, "Chromosome location (bp)", ha="center", va="top")
ax.text(80, 30, "LIS = 7", color="r", ha="center", va="center")
ax.text(80, 20, "LDS = 4", color="g", ha="center", va="center")
ax.text(80, 10, "LMS = $max$(LIS, LDS) = 7", ha="center", va="center")
normalize_lms_axis(ax, xlim=110, ylim=110)
# Panel B
w = 0.37
p = (0, 45, 75, 110)
ax = fig.add_axes([0.1, 0.12, w, h])
xdata = [x for x in range(10, 110, 10)]
ydata = ydata_orig = [x for x in range(10, 110, 10)]
ydata = ydata[:4] + ydata[7:] + ydata[4:7][::-1]
xydata = zip(xdata, ydata)
lis = xydata[:7]
xlis, ylis = zip(*lis)
ax.plot(
xlis,
ylis,
"r-",
lw=12,
alpha=0.3,
solid_capstyle="round",
solid_joinstyle="round",
)
ax.plot(xdata, ydata, "k.", mec="k", mfc="w", mew=3, ms=12)
ax.vlines(p, 0, 110, colors="beige", lw=3)
normalize_lms_axis(ax, xlim=110, ylim=110)
patch = [0.1 + w * x / 110.0 for x in p]
HorizontalChromosome(root, 0.1, 0.1 + w, 0.09, patch=patch, height=0.02, lw=2)
scaffolds = ("a", "b", "c")
for i, s in enumerate(scaffolds):
xx = (patch[i] + patch[i + 1]) / 2
root.text(xx, 0.09, s, va="center", ha="center")
root.text(0.1 + w / 2, 0.04, "LMS($a||b||c$) = 7", ha="center")
# Panel C
ax = fig.add_axes([0.6, 0.12, w, h])
patch = [0.6 + w * x / 110.0 for x in p]
ydata = ydata_orig
ax.plot(
xdata,
ydata,
"r-",
lw=12,
alpha=0.3,
solid_capstyle="round",
solid_joinstyle="round",
)
ax.plot(xdata, ydata, "k.", mec="k", mfc="w", mew=3, ms=12)
ax.vlines(p, [0], [110], colors="beige", lw=3)
normalize_lms_axis(ax, xlim=110, ylim=110)
HorizontalChromosome(root, 0.6, 0.6 + w, 0.09, patch=patch, height=0.02, lw=2)
scaffolds = ("a", "-c", "b")
for i, s in enumerate(scaffolds):
xx = (patch[i] + patch[i + 1]) / 2
root.text(xx, 0.09, s, va="center", ha="center")
root.text(0.6 + w / 2, 0.04, "LMS($a||-c||b$) = 10", ha="center")
labels = ((0.05, 0.95, "A"), (0.05, 0.48, "B"), (0.55, 0.48, "C"))
panel_labels(root, labels)
normalize_axes(root)
pf = "lms"
image_name = pf + "." + iopts.format
savefig(image_name, dpi=iopts.dpi, iopts=iopts)
def plot(args):
"""
%prog plot input.bed seqid
Plot the matchings between the reconstructed pseudomolecules and the maps.
Two types of visualizations are available in one canvas:
1. Parallel axes, and matching markers are shown in connecting lines;
2. Scatter plot.
"""
from jcvi.graphics.base import plt, savefig, normalize_axes, \
set2, panel_labels
from jcvi.graphics.chromosome import Chromosome, GeneticMap, \
HorizontalChromosome
p = OptionParser(plot.__doc__)
add_allmaps_plot_options(p)
opts, args, iopts = p.set_image_options(args, figsize="10x6")
if len(args) != 2:
sys.exit(not p.print_help())
inputbed, seqid = args
pf = inputbed.rsplit(".", 1)[0]
bedfile = pf + ".lifted.bed"
agpfile = pf + ".agp"
weightsfile = opts.weightsfile
links = opts.links
function = get_function(opts.distance)
cc = Map(bedfile, function)
allseqids = cc.seqids
mapnames = cc.mapnames
weights = Weights(weightsfile, mapnames)
assert seqid in allseqids, "{0} not in {1}".format(seqid, allseqids)
s = Scaffold(seqid, cc)
mlgs = [k for k, v in s.mlg_counts.items() if v >= links]
while not mlgs:
links /= 2
logging.error("No markers to plot, --links reset to {0}".format(links))
mlgs = [k for k, v in s.mlg_counts.items() if v >= links]
mlgsizes = {}
for mlg in mlgs:
mm = cc.extract_mlg(mlg)
mlgsize = max(function(x) for x in mm)
mlgsizes[mlg] = mlgsize
fig = plt.figure(1, (iopts.w, iopts.h))
root = fig.add_axes([0, 0, 1, 1])
ax1 = fig.add_axes([0, 0, .5, 1])
ax2 = fig.add_axes([.5, 0, .5, 1])
# Find the layout first
ystart, ystop = .9, .1
L = Layout(mlgsizes)
coords = L.coords
tip = .02
marker_pos = {}
# Palette
colors = dict((mapname, set2[i]) for i, mapname in enumerate(mapnames))
colors = dict((mlg, colors[mlg.split("-")[0]]) for mlg in mlgs)
rhos = {}
# Parallel coordinates
for mlg, (x, y1, y2) in coords.items():
mm = cc.extract_mlg(mlg)
markers = [(m.accn, function(m)) for m in mm] # exhaustive marker list
xy = [(m.pos, function(m)) for m in mm if m.seqid == seqid]
mx, my = zip(*xy)
rho = spearmanr(mx, my)
rhos[mlg] = rho
flip = rho < 0
g = GeneticMap(ax1, x, y1, y2, markers, tip=tip, flip=flip)
extra = -3 * tip if x < .5 else 3 * tip
ha = "right" if x < .5 else "left"
mapname = mlg.split("-")[0]
tlg = mlg.replace("_", ".") # Latex does not like underscore char
label = "{0} (w={1})".format(tlg, weights[mapname])
ax1.text(x + extra, (y1 + y2) / 2, label, color=colors[mlg],
ha=ha, va="center", rotation=90)
marker_pos.update(g.marker_pos)
agp = AGP(agpfile)
agp = [x for x in agp if x.object == seqid]
chrsize = max(x.object_end for x in agp)
# Pseudomolecules in the center
r = ystart - ystop
ratio = r / chrsize
f = lambda x: (ystart - ratio * x)
patchstart = [f(x.object_beg) for x in agp if not x.is_gap]
Chromosome(ax1, .5, ystart, ystop, width=2 * tip, patch=patchstart, lw=2)
label = "{0} ({1})".format(seqid, human_size(chrsize, precision=0))
ax1.text(.5, ystart + tip, label, ha="center")
scatter_data = defaultdict(list)
# Connecting lines
for b in s.markers:
marker_name = b.accn
if marker_name not in marker_pos:
continue
cx = .5
cy = f(b.pos)
mx = coords[b.mlg][0]
my = marker_pos[marker_name]
extra = -tip if mx < cx else tip
extra *= 1.25 # leave boundaries for aesthetic reasons
cx += extra
mx -= extra
ax1.plot((cx, mx), (cy, my), "-", color=colors[b.mlg])
scatter_data[b.mlg].append((b.pos, function(b)))
# Scatter plot, same data as parallel coordinates
xstart, xstop = sorted((ystart, ystop))
f = lambda x: (xstart + ratio * x)
pp = [x.object_beg for x in agp if not x.is_gap]
patchstart = [f(x) for x in pp]
HorizontalChromosome(ax2, xstart, xstop, ystop,
height=2 * tip, patch=patchstart, lw=2)
gap = .03
ratio = (r - gap * len(mlgs) - tip) / sum(mlgsizes.values())
tlgs = []
for mlg, mlgsize in sorted(mlgsizes.items()):
height = ratio * mlgsize
ystart -= height
xx = .5 + xstart / 2
width = r / 2
color = colors[mlg]
ax = fig.add_axes([xx, ystart, width, height])
ypos = ystart + height / 2
ystart -= gap
sd = scatter_data[mlg]
xx, yy = zip(*sd)
ax.vlines(pp, 0, mlgsize, colors="beige")
ax.plot(xx, yy, ".", color=color)
rho = rhos[mlg]
ax.text(.5, 1 - .4 * gap / height, r"$\rho$={0:.3f}".format(rho),
ha="center", va="top", transform=ax.transAxes, color="gray")
tlg = mlg.replace("_", ".")
tlgs.append((tlg, ypos, color))
ax.set_xlim(0, chrsize)
ax.set_ylim(0, mlgsize)
ax.set_xticks([])
while height / len(ax.get_yticks()) < .03 and len(ax.get_yticks()) >= 2:
ax.set_yticks(ax.get_yticks()[::2]) # Sparsify the ticks
yticklabels = [int(x) for x in ax.get_yticks()]
ax.set_yticklabels(yticklabels, family='Helvetica')
if rho < 0:
ax.invert_yaxis()
for i, (tlg, ypos, color) in enumerate(tlgs):
ha = "center"
if len(tlgs) > 4:
ha = "right" if i % 2 else "left"
root.text(.5, ypos, tlg, color=color, rotation=90,
ha=ha, va="center")
if opts.panels:
labels = ((.04, .96, 'A'), (.48, .96, 'B'))
panel_labels(root, labels)
normalize_axes((ax1, ax2, root))
image_name = seqid + "." + iopts.format
savefig(image_name, dpi=iopts.dpi, iopts=iopts)
plt.close(fig)
def compare3(args):
"""
%prog compare3
Compare performances of various variant callers on simulated STR datasets.
This compares the power of various evidence types.
"""
p = OptionParser(compare3.__doc__)
p.add_option('--maxinsert',
default=300,
type="int",
help="Maximum number of repeats")
add_simulate_options(p)
opts, args, iopts = p.set_image_options(args, figsize="10x10")
if len(args) != 0:
sys.exit(not p.print_help())
max_insert = opts.maxinsert
fig, ((ax1, ax2), (ax3, ax4)) = plt.subplots(ncols=2,
nrows=2,
figsize=(iopts.w, iopts.h))
plt.tight_layout(pad=3)
color = "lightslategray"
# ax1: Spanning
tredparse_results = parse_results("tredparse_results_het-spanning.txt")
title = SIMULATED_DIPLOID + "( Sub-model 1: Spanning reads)"
plot_compare(ax1,
title,
tredparse_results,
None,
color=color,
max_insert=max_insert,
risk=False)
# ax2: Partial
tredparse_results = parse_results("tredparse_results_het-partial.txt",
exclude=20)
title = SIMULATED_DIPLOID + " (Sub-model 2: Partial reads)"
plot_compare(ax2,
title,
tredparse_results,
None,
color=color,
max_insert=max_insert,
risk=False)
# ax3: Repeat
tredparse_results = parse_results("tredparse_results_het-repeat.txt",
exclude=20)
# HACK (repeat reads won't work under 50)
tredparse_results = [x for x in tredparse_results if x[0] > 50]
title = SIMULATED_DIPLOID + " (Sub-model 3: Repeat-only reads)"
plot_compare(ax3,
title,
tredparse_results,
None,
color=color,
max_insert=max_insert,
risk=False)
# ax4: Pair
tredparse_results = parse_results("tredparse_results_het-pair.txt",
exclude=20)
title = SIMULATED_DIPLOID + " (Sub-model 4: Paired-end reads)"
plot_compare(ax4,
title,
tredparse_results,
None,
color=color,
max_insert=max_insert,
risk=False)
for ax in (ax1, ax2, ax3, ax4):
ax.set_xlim(0, max_insert)
ax.set_ylim(0, max_insert)
root = fig.add_axes([0, 0, 1, 1])
pad = .03
panel_labels(root, ((pad / 2, 1 - pad, "A"), (1 / 2., 1 - pad, "B"),
(pad / 2, 1 / 2., "C"), (1 / 2., 1 / 2., "D")))
normalize_axes(root)
image_name = "tredparse." + iopts.format
savefig(image_name, dpi=iopts.dpi, iopts=iopts)
def dotplot(anchorfile, qbed, sbed, fig, root, ax, vmin=0, vmax=1,
is_self=False, synteny=False, cmap_text=None, cmap="copper",
genomenames=None, sample_number=10000, minfont=5, palette=None,
chrlw=.1, title=None, sep=True, sepcolor="g", stdpf=True):
fp = open(anchorfile)
# add genome names
if genomenames:
gx, gy = genomenames.split("_")
else:
to_ax_label = lambda fname: op.basename(fname).split(".")[0]
gx, gy = [to_ax_label(x.filename) for x in (qbed, sbed)]
gx, gy = markup(gx), markup(gy)
qorder = qbed.order
sorder = sbed.order
data = []
if cmap_text:
logging.debug("Capping values within [{0:.1f}, {1:.1f}]"\
.format(vmin, vmax))
block_id = 0
for row in fp:
atoms = row.split()
block_color = None
if row[0] == "#":
block_id += 1
if palette:
block_color = palette.get(block_id, "k")
continue
# first two columns are query and subject, and an optional third column
if len(atoms) < 2:
continue
query, subject = atoms[:2]
value = atoms[-1]
if cmap_text:
try:
value = float(value)
except ValueError:
value = vmax
if value < vmin:
continue
if value > vmax:
continue
else:
value = 0
if query not in qorder:
continue
if subject not in sorder:
continue
qi, q = qorder[query]
si, s = sorder[subject]
nv = value if block_color is None else block_color
data.append((qi, si, nv))
if is_self: # Mirror image
data.append((si, qi, nv))
npairs = downsample(data, sample_number=sample_number)
x, y, c = zip(*data)
if palette:
ax.scatter(x, y, c=c, edgecolors="none", s=2, lw=0)
else:
ax.scatter(x, y, c=c, edgecolors="none", s=2, lw=0, cmap=cmap,
vmin=vmin, vmax=vmax)
if synteny:
clusters = batch_scan(data, qbed, sbed)
draw_box(clusters, ax)
if cmap_text:
draw_cmap(root, cmap_text, vmin, vmax, cmap=cmap)
xsize, ysize = len(qbed), len(sbed)
logging.debug("xsize=%d ysize=%d" % (xsize, ysize))
qbreaks = qbed.get_breaks()
sbreaks = sbed.get_breaks()
xlim, ylim = plot_breaks_and_labels(fig, root, ax, gx, gy, xsize, ysize,
qbreaks, sbreaks, sep=sep, chrlw=chrlw,
sepcolor=sepcolor, minfont=minfont, stdpf=stdpf)
# create a diagonal to separate mirror image for self comparison
if is_self:
ax.plot(xlim, (0, ysize), 'm-', alpha=.5, lw=2)
if palette: # bottom-left has the palette, if available
colors = palette.colors
xstart, ystart = .1, .05
for category, c in sorted(colors.items()):
root.add_patch(Rectangle((xstart, ystart), .03, .02, lw=0, fc=c))
root.text(xstart + .04, ystart, category, color=c)
xstart += .1
if title is None:
title = "Inter-genomic comparison: {0} vs {1}".format(gx, gy)
if is_self:
title = "Intra-genomic comparison within {0}".format(gx)
npairs /= 2
title += " ({0} gene pairs)".format(thousands(npairs))
root.set_title(title, x=.5, y=.96, color="k")
if title:
logging.debug("Dot plot title: {}".format(title))
normalize_axes(root)
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 estimategaps(args):
"""
%prog estimategaps JM-4 chr1 JMMale-1
Illustrate ALLMAPS gap estimation algorithm.
"""
p = OptionParser(estimategaps.__doc__)
opts, args, iopts = p.set_image_options(args, figsize="6x6", dpi=300)
if len(args) != 3:
sys.exit(not p.print_help())
pf, seqid, mlg = args
bedfile = pf + ".lifted.bed"
agpfile = pf + ".agp"
function = lambda x: x.cm
cc = Map(bedfile, scaffold_info=True, function=function)
agp = AGP(agpfile)
g = GapEstimator(cc, agp, seqid, mlg, function=function)
pp, chrsize, mlgsize = g.pp, g.chrsize, g.mlgsize
spl, spld = g.spl, g.spld
g.compute_all_gaps(verbose=False)
fig = plt.figure(1, (iopts.w, iopts.h))
root = fig.add_axes([0, 0, 1, 1])
# Panel A
xstart, ystart = .15, .65
w, h = .7, .3
t = np.linspace(0, chrsize, 1000)
ax = fig.add_axes([xstart, ystart, w, h])
mx, my = zip(*g.scatter_data)
rho = spearmanr(mx, my)
dsg = "g"
ax.vlines(pp, 0, mlgsize, colors="beige")
ax.plot(mx, my, ".", color=set2[3])
ax.plot(t, spl(t), "-", color=dsg)
ax.text(.05, .95, mlg, va="top", transform=ax.transAxes)
normalize_lms_axis(ax, xlim=chrsize, ylim=mlgsize,
ylabel="Genetic distance (cM)")
if rho < 0:
ax.invert_yaxis()
# Panel B
ystart -= .28
h = .25
ax = fig.add_axes([xstart, ystart, w, h])
ax.vlines(pp, 0, mlgsize, colors="beige")
ax.plot(t, spld(t), "-", lw=2, color=dsg)
ax.plot(pp, spld(pp), "o", mfc="w", mec=dsg, ms=5)
normalize_lms_axis(ax, xlim=chrsize, ylim=25 * 1e-6,
xfactor=1e-6, xlabel="Physical position (Mb)",
yfactor=1000000, ylabel="Recomb. rate\n(cM / Mb)")
# Panel C (specific to JMMale-1)
a, b = "scaffold_1076", "scaffold_861"
sizes = dict((x.component_id, (x.object_beg, x.object_end,
x.component_span, x.orientation)) \
for x in g.agp if not x.is_gap)
a_beg, a_end, asize, ao = sizes[a]
b_beg, b_end, bsize, bo = sizes[b]
gapsize = g.get_gapsize(a)
total_size = asize + gapsize + bsize
ratio = .6 / total_size
y = .16
pad = .03
pb_ratio = w / chrsize
# Zoom
lsg = "lightslategray"
root.plot((.15 + pb_ratio * a_beg, .2),
(ystart, ystart - .14), ":", color=lsg)
root.plot((.15 + pb_ratio * b_end, .3),
(ystart, ystart - .08), ":", color=lsg)
ends = []
for tag, size, marker, beg in zip((a, b), (asize, bsize), (49213, 81277),
(.2, .2 + (asize + gapsize) * ratio)):
end = beg + size * ratio
marker = beg + marker * ratio
ends.append((beg, end, marker))
root.plot((marker,), (y,), "o", color=lsg)
root.text((beg + end) / 2, y + pad, latex(tag),
ha="center", va="center")
HorizontalChromosome(root, beg, end, y, height=.025, fc='gainsboro')
begs, ends, markers = zip(*ends)
fontprop = dict(color=lsg, ha="center", va="center")
ypos = y + pad * 2
root.plot(markers, (ypos, ypos), "-", lw=2, color=lsg)
root.text(sum(markers) / 2, ypos + pad,
"Distance: 1.29cM $\Leftrightarrow$ 211,824bp (6.1 cM/Mb)", **fontprop)
ypos = y - pad
xx = markers[0], ends[0]
root.plot(xx, (ypos, ypos), "-", lw=2, color=lsg)
root.text(sum(xx) / 2, ypos - pad, "34,115bp", **fontprop)
xx = markers[1], begs[1]
root.plot(xx, (ypos, ypos), "-", lw=2, color=lsg)
root.text(sum(xx) / 2, ypos - pad, "81,276bp", **fontprop)
root.plot((ends[0], begs[1]), (y, y), ":", lw=2, color=lsg)
root.text(sum(markers) / 2, ypos - 3 * pad, r"$\textit{Estimated gap size: 96,433bp}$",
color="r", ha="center", va="center")
labels = ((.05, .95, 'A'), (.05, .6, 'B'), (.05, .27, 'C'))
panel_labels(root, labels)
normalize_axes(root)
pf = "estimategaps"
image_name = pf + "." + iopts.format
savefig(image_name, dpi=iopts.dpi, iopts=iopts)
def heatmap(args):
"""
%prog heatmap input.npy genome.json
Plot heatmap based on .npy data file. The .npy stores a square matrix with
bins of genome, and cells inside the matrix represent number of links
between bin i and bin j. The `genome.json` contains the offsets of each
contig/chr so that we know where to draw boundary lines, or extract per
contig/chromosome heatmap.
"""
p = OptionParser(heatmap.__doc__)
p.add_option("--resolution", default=500000, type="int",
help="Resolution when counting the links")
p.add_option("--chr", help="Plot this contig/chr only")
opts, args, iopts = p.set_image_options(args, figsize="10x10",
style="white", cmap="coolwarm",
format="png", dpi=120)
if len(args) != 2:
sys.exit(not p.print_help())
npyfile, jsonfile = args
contig = opts.chr
# Load contig/chromosome starts and sizes
header = json.loads(open(jsonfile).read())
# Load the matrix
A = np.load(npyfile)
# Select specific submatrix
if contig:
contig_start = header["starts"][contig]
contig_size = header["sizes"][contig]
contig_end = contig_start + contig_size
A = A[contig_start: contig_end, contig_start: contig_end]
# Several concerns in practice:
# The diagonal counts may be too strong, this can either be resolved by
# masking them. Or perform a log transform on the entire heatmap.
B = A.astype("float64")
B += 1.0
B = np.log(B)
vmin, vmax = 1, 7
B[B < vmin] = vmin
B[B > vmax] = vmax
print B
logging.debug("Matrix log-transformation and thresholding ({}-{}) done"
.format(vmin, vmax))
# Canvas
fig = plt.figure(1, (iopts.w, iopts.h))
root = fig.add_axes([0, 0, 1, 1]) # whole canvas
ax = fig.add_axes([.05, .05, .9, .9]) # just the heatmap
breaks = header["starts"].values()
breaks += [header["total_bins"]] # This is actually discarded
breaks = sorted(breaks)[1:]
if contig:
breaks = []
plot_heatmap(ax, B, breaks, iopts, binsize=opts.resolution)
# Title
pf = npyfile.rsplit(".", 1)[0]
title = pf
if contig:
title += "-{}".format(contig)
root.text(.5, .98, title, color="darkslategray", size=18,
ha="center", va="center")
normalize_axes(root)
image_name = title + "." + iopts.format
savefig(image_name, dpi=iopts.dpi, iopts=iopts)
def draw_tree(ax, tx, rmargin=.3, leafcolor="k", supportcolor="k",
outgroup=None, reroot=True, gffdir=None, sizes=None,
trunc_name=None, SH=None, scutoff=0, barcodefile=None,
leafcolorfile=None, leaffont=12):
"""
main function for drawing phylogenetic tree
"""
t = Tree(tx)
if reroot:
if outgroup:
R = t.get_common_ancestor(*outgroup)
else:
# Calculate the midpoint node
R = t.get_midpoint_outgroup()
if R != t:
t.set_outgroup(R)
farthest, max_dist = t.get_farthest_leaf()
margin = .05
xstart = margin
ystart = 1 - margin
canvas = 1 - rmargin - 2 * margin
tip = .005
# scale the tree
scale = canvas / max_dist
num_leaves = len(t.get_leaf_names())
yinterval = canvas / (num_leaves + 1)
# get exons structures, if any
structures = {}
if gffdir:
gffiles = glob("{0}/*.gff*".format(gffdir))
setups, ratio = get_setups(gffiles, canvas=rmargin / 2, noUTR=True)
structures = dict((a, (b, c)) for a, b, c in setups)
if sizes:
sizes = Sizes(sizes).mapping
if barcodefile:
barcodemap = DictFile(barcodefile, delimiter="\t")
if leafcolorfile:
leafcolors = DictFile(leafcolorfile, delimiter="\t")
coords = {}
i = 0
for n in t.traverse("postorder"):
dist = n.get_distance(t)
xx = xstart + scale * dist
if n.is_leaf():
yy = ystart - i * yinterval
i += 1
if trunc_name:
name = truncate_name(n.name, rule=trunc_name)
else:
name = n.name
if barcodefile:
name = decode_name(name, barcodemap)
sname = name.replace("_", "-")
try:
lc = leafcolors[n.name]
except Exception:
lc = leafcolor
else:
# if color is given as "R,G,B"
if "," in lc:
lc = map(float, lc.split(","))
ax.text(xx + tip, yy, sname, va="center",
fontstyle="italic", size=leaffont, color=lc)
gname = n.name.split("_")[0]
if gname in structures:
mrnabed, cdsbeds = structures[gname]
ExonGlyph(ax, 1 - rmargin / 2, yy, mrnabed, cdsbeds,
align="right", ratio=ratio)
if sizes and gname in sizes:
size = sizes[gname]
size = size / 3 - 1 # base pair converted to amino acid
size = "{0}aa".format(size)
ax.text(1 - rmargin / 2 + tip, yy, size, size=leaffont)
else:
children = [coords[x] for x in n.get_children()]
children_x, children_y = zip(*children)
min_y, max_y = min(children_y), max(children_y)
# plot the vertical bar
ax.plot((xx, xx), (min_y, max_y), "k-")
# plot the horizontal bar
for cx, cy in children:
ax.plot((xx, cx), (cy, cy), "k-")
yy = sum(children_y) * 1. / len(children_y)
support = n.support
if support > 1:
support = support / 100.
if not n.is_root():
if support > scutoff / 100.:
ax.text(xx, yy+.005, "{0:d}".format(int(abs(support * 100))),
ha="right", size=leaffont, color=supportcolor)
coords[n] = (xx, yy)
# scale bar
br = .1
x1 = xstart + .1
x2 = x1 + br * scale
yy = ystart - i * yinterval
ax.plot([x1, x1], [yy - tip, yy + tip], "k-")
ax.plot([x2, x2], [yy - tip, yy + tip], "k-")
ax.plot([x1, x2], [yy, yy], "k-")
ax.text((x1 + x2) / 2, yy - tip, "{0:g}".format(br),
va="top", ha="center", size=leaffont)
if SH is not None:
xs = x1
ys = (margin + yy) / 2.
ax.text(xs, ys, "SH test against ref tree: {0}"\
.format(SH), ha="left", size=leaffont, color="g")
normalize_axes(ax)
def wheel(args):
"""
%prog wheel datafile.csv groups.csv
Wheel plot that shows continous data in radial axes.
"""
p = OptionParser(wheel.__doc__)
p.add_option("--column",
default="score",
choices=("score", "percentile"),
help="Which column to extract from `datafile.csv`")
opts, args, iopts = p.set_image_options(args, figsize="5x5", format="png")
if len(args) != 2:
sys.exit(not p.print_help())
datafile, groupsfile = args
column = opts.column
linecolor = "#d6d6d6"
df = parse_data(datafile, score_column=opts.column)
groups = parse_groups(groupsfile)
labels = [g for g in groups if g in df]
print(labels)
df = [df[g] for g in labels]
print(df)
groups = [groups[g] for g in labels]
print(groups)
pf = datafile.rsplit(".", 1)[0]
fig = plt.figure(1, (iopts.w, iopts.h))
root = fig.add_axes([0, 0, 1, 1])
categories = len(df)
#ax = plt.subplot(111, projection='polar')
ax = fig.add_axes([0.1, 0.1, 0.8, 0.8], polar=True)
brewer = [
"#FF3B30",
"#DD43A0",
"#5856D6",
"#007AFE",
"#56BDEC",
"#4CD8BA",
"#4CD864",
"#B0F457",
"#FEF221",
"#FFCC01",
"#FF9500",
"#FF3B30",
]
# Baseline
theta = np.linspace(1.5 * np.pi,
3.5 * np.pi,
endpoint=False,
num=categories)
_theta = np.linspace(1.5 * np.pi, 3.5 * np.pi)
R = max(max(df), 10)
xlim = (-R, R) if column == "score" else (-100, 100)
plim = (-R / 2, R) if column == "score" else (0, 100)
ci = (-.5, 2) if column == "score" else (10, 90)
# Grid
if column == "score":
for t in theta:
ax.plot([t, t], plim, color=linecolor)
ax.axis('off')
# Contours
for t in plim:
ax.plot(_theta, [t] * len(_theta), color=linecolor)
# Sectors (groupings)
collapsed_groups = []
gg = []
for group, c in groupby(enumerate(groups), lambda x: x[1]):
c = [x[0] for x in list(c)]
collapsed_groups.append(group)
gg.append(c)
sector = False
if sector:
theta_interval = 2 * np.pi / categories
theta_pad = theta_interval / 2 * .9
for color, group in zip(brewer, gg):
tmin, tmax = min(group), max(group)
sector(ax,
theta[tmin],
theta[tmax],
theta_pad,
R * .95,
"-",
color=color,
lw=2)
# Data
r = df
closed_plot(ax, theta, r, color="lightslategray", alpha=.25)
all_data = []
for color, group in zip(brewer, gg):
hidden_data = [(theta[x], r[x]) for x in group if \
(ci[0] <= r[x] <= ci[1])]
shown_data = [(theta[x], r[x]) for x in group if (r[x] < ci[0] \
or r[x] > ci[1])]
all_data.append((theta[x], labels[x], r[x]))
for alpha, data in zip((1, 1), (hidden_data, shown_data)):
if not data:
continue
color_theta, color_r = zip(*data)
ax.plot(color_theta, color_r, "o", color=color, alpha=alpha)
# Print out data
diseaseNames, risks = labels, df
print("var theta = [{}]".format(",".join("{:.1f}".format(degrees(x))
for x in theta)))
print("var risks = [{}]".format(",".join(str(x) for x in risks)))
print("var diseaseNames = [{}]".format(",".join(\
['"{}"'.format(x) for x in diseaseNames])))
# Labels
from math import cos, sin
r = .5
for i, label in enumerate(labels):
tl = theta[i]
x, y = .5 + r * cos(tl), .5 + r * sin(tl)
d = degrees(tl)
if 90 < d % 360 < 270: # On the left quardrants
d -= 180
root.text(x,
y,
label,
size=4,
rotation=d,
ha="center",
va="center",
color=linecolor)
print(x, y, label)
# Add baseline
baseline = 0 if column == "score" else 50
_r = len(_theta) * [baseline]
closed_plot(ax, _theta, _r, "k:", lw=1, ms=4)
# Add confidence interval
if column == "percentile":
barcolor = "#eeeeee"
ax.bar([0], [ci[1] - ci[0]],
width=2 * np.pi,
bottom=ci[0],
fc=barcolor)
ax.set_rmin(xlim[0])
ax.set_rmax(xlim[1])
normalize_axes(root)
image_name = pf + "-" + column + "." + iopts.format
savefig(image_name, dpi=iopts.dpi, iopts=iopts)
def main():
p = OptionParser(__doc__)
opts, args, iopts = p.set_image_options(figsize="9x7")
if len(args) != 1:
sys.exit(not p.print_help())
mode, = args
assert mode == "demo"
a, b = 30, 70
pad = .08
w = .31
fig = plt.figure(1, (iopts.w, iopts.h))
root = fig.add_axes([0, 0, 1, 1])
# Row separators
yy = 1 - pad
for i in xrange(3):
root.plot((0, 1), (yy, yy), "-", lw=2, color="lightgray")
yy -= w
# Row headers
xx = pad * .6
yy = 1 - pad - .5 * w
for title in ("Inversion", "Indel", "Duplication"):
root.text(xx, yy, title, ha="center", va="center")
yy -= w
# Column headers
xx = pad + .5 * w
yy = 1 - pad / 2
for title in ("Assembly alignment", "Read alignment",
"Optical map alignment"):
root.text(xx, yy, title, ha="center", va="center")
xx += w
p = PairwiseAlign(fig, [pad, 2 * w, w, w])
p.invert(a, b)
p.draw()
p = PairwiseAlign(fig, [pad, w, w, w])
p.delete(a, b)
p.draw()
p = PairwiseAlign(fig, [pad, 0, w, w])
p.duplicate(a, b, gap=5)
p.draw()
p = ReadAlign(fig, [pad + w, 2 * w, w, w])
p.invert(a, b)
p.draw()
p = ReadAlign(fig, [pad + w, w, w, w])
p.delete(a, b)
p.draw()
p = ReadAlign(fig, [pad + w, 0, w, w])
p.duplicate(a, b)
p.draw()
p = OpticalMapAlign(fig, [pad + 2 * w, 2 * w, w, w])
p.invert(a, b)
p.draw()
p = OpticalMapAlign(fig, [pad + 2 * w, w, w, w])
p.delete(a, b)
p.draw()
p = OpticalMapAlign(fig, [pad + 2 * w, 0, w, w])
p.duplicate(a, b)
p.draw()
normalize_axes(root)
image_name = mode + "." + iopts.format
savefig(image_name, dpi=iopts.dpi, iopts=iopts)
def compare4(args):
"""
%prog compare4
Compare performances of various variant callers on simulated STR datasets.
Adds coverage comparisons as panel C and D.
"""
p = OptionParser(compare4.__doc__)
p.add_option('--maxinsert',
default=300,
type="int",
help="Maximum number of repeats")
add_simulate_options(p)
opts, args, iopts = p.set_image_options(args, figsize="10x10")
if len(args) != 0:
sys.exit(not p.print_help())
depth = opts.depth
max_insert = opts.maxinsert
fig, ((ax1, ax2), (ax3, ax4)) = plt.subplots(ncols=2,
nrows=2,
figsize=(iopts.w, iopts.h))
plt.tight_layout(pad=3)
# ax1: lobSTR vs TREDPARSE with haploid model
lobstr_results = parse_results("lobstr_results_homo-20x-150bp-500bp.txt")
tredparse_results = parse_results(
"tredparse_results_homo-20x-150bp-500bp.txt")
title = SIMULATED_HAPLOID + r" ($Depth=%s\times)" % depth
plot_compare(ax1,
title,
tredparse_results,
lobstr_results,
max_insert=max_insert)
# ax2: lobSTR vs TREDPARSE with diploid model (depth=20x)
lobstr_results = parse_results("lobstr_results_het-20x-150bp-500bp.txt",
exclude=20)
tredparse_results = parse_results(
"tredparse_results_het-20x-150bp-500bp.txt", exclude=20)
title = SIMULATED_DIPLOID + r" ($Depth=%s\times$)" % depth
plot_compare(ax2,
title,
tredparse_results,
lobstr_results,
max_insert=max_insert)
# ax3: lobSTR vs TREDPARSE with diploid model (depth=5x)
lobstr_results = parse_results("lobstr_results_het-5x-150bp-500bp.txt",
exclude=20)
tredparse_results = parse_results(
"tredparse_results_het-5x-150bp-500bp.txt", exclude=20)
title = SIMULATED_DIPLOID + r" ($Depth=%s\times$)" % 5
plot_compare(ax3,
title,
tredparse_results,
lobstr_results,
max_insert=max_insert)
# ax4: lobSTR vs TREDPARSE with diploid model (depth=80x)
lobstr_results = parse_results("lobstr_results_het-80x-150bp-500bp.txt",
exclude=20)
tredparse_results = parse_results(
"tredparse_results_het-80x-150bp-500bp.txt", exclude=20)
title = SIMULATED_DIPLOID + r" ($Depth=%s\times$)" % 80
plot_compare(ax4,
title,
tredparse_results,
lobstr_results,
max_insert=max_insert)
for ax in (ax1, ax2, ax3, ax4):
ax.set_xlim(0, max_insert)
ax.set_ylim(0, max_insert)
root = fig.add_axes([0, 0, 1, 1])
pad = .03
panel_labels(root, ((pad / 2, 1 - pad, "A"), (1 / 2., 1 - pad, "B"),
(pad / 2, 1 / 2., "C"), (1 / 2., 1 / 2., "D")))
normalize_axes(root)
image_name = "tredparse." + iopts.format
savefig(image_name, dpi=iopts.dpi, iopts=iopts)
def fig3(args):
"""
%prog fig3 chrA02,A02,C2,chrC02 chr.sizes all.bed data
Napus Figure 3 displays alignments between quartet chromosomes, inset
with read histograms.
"""
from jcvi.formats.bed import Bed
p = OptionParser(fig3.__doc__)
p.add_option("--gauge_step",
default=10000000,
type="int",
help="Step size for the base scale")
opts, args, iopts = p.set_image_options(args, figsize="12x9")
if len(args) != 4:
sys.exit(not p.print_help())
chrs, sizes, bedfile, datadir = args
gauge_step = opts.gauge_step
diverge = iopts.diverge
rr, gg = diverge
chrs = [[x] for x in chrs.split(",")]
sizes = Sizes(sizes).mapping
fig = plt.figure(1, (iopts.w, iopts.h))
root = fig.add_axes([0, 0, 1, 1])
chr_sizes, chr_sum_sizes, ratio = calc_ratio(chrs, sizes)
# Synteny panel
seqidsfile = make_seqids(chrs)
klayout = make_layout(chrs, chr_sum_sizes, ratio, template_f3a, shift=.05)
height = .07
r = height / 4
K = Karyotype(fig,
root,
seqidsfile,
klayout,
gap=gap,
height=height,
lw=2,
generank=False,
sizes=sizes,
heightpad=r,
roundrect=True,
plot_label=False)
# Chromosome labels
for kl in K.layout:
if kl.empty:
continue
lx, ly = kl.xstart, kl.y
if lx < .11:
lx += .1
ly += .06
label = kl.label
root.text(lx - .015, ly, label, fontsize=15, ha="right", va="center")
# Inset with datafiles
datafiles = ("chrA02.bzh.forxmgr", "parent.A02.per10kb.forxmgr",
"parent.C2.per10kb.forxmgr", "chrC02.bzh.forxmgr")
datafiles = [op.join(datadir, x) for x in datafiles]
tracks = K.tracks
hlfile = op.join(datadir, "bzh.regions.forhaibao")
xy_axes = []
for t, datafile in zip(tracks, datafiles):
ax = make_affix_axis(fig, t, -r, height=2 * r)
xy_axes.append(ax)
chr = t.seqids[0]
xy = XYtrack(ax, datafile, color="lightslategray")
start, end = 0, t.total
xy.interpolate(end)
xy.cap(ymax=40)
xy.import_hlfile(hlfile, chr, diverge=diverge)
xy.draw()
ax.set_xlim(start, end)
gauge_ax = make_affix_axis(fig, t, -r)
adjust_spines(gauge_ax, ["bottom"])
setup_gauge_ax(gauge_ax, start, end, gauge_step)
# Converted gene tracks
ax_Ar = make_affix_axis(fig, tracks[1], r, height=r / 2)
ax_Co = make_affix_axis(fig, tracks[2], r, height=r / 2)
order = Bed(bedfile).order
for asterisk in (False, True):
conversion_track(order,
"data/Genes.Converted.seuil.0.6.AtoC.txt",
0,
"A02",
ax_Ar,
rr,
asterisk=asterisk)
conversion_track(order,
"data/Genes.Converted.seuil.0.6.AtoC.txt",
1,
"C2",
ax_Co,
gg,
asterisk=asterisk)
conversion_track(order,
"data/Genes.Converted.seuil.0.6.CtoA.txt",
0,
"A02",
ax_Ar,
gg,
ypos=1,
asterisk=asterisk)
conversion_track(order,
"data/Genes.Converted.seuil.0.6.CtoA.txt",
1,
"C2",
ax_Co,
rr,
ypos=1,
asterisk=asterisk)
Ar, Co = xy_axes[1:3]
annotations = ((Ar, "Bra028920 Bra028897", "center",
"1DAn2+"), (Ar, "Bra020081 Bra020171", "right", "2DAn2+"),
(Ar, "Bra020218 Bra020286", "left",
"3DAn2+"), (Ar, "Bra008143 Bra008167", "left", "4DAn2-"),
(Ar, "Bra029317 Bra029251", "right",
"5DAn2+ (GSL)"), (Co, "Bo2g001000 Bo2g001300", "left",
"1DCn2-"), (Co, "Bo2g018560 Bo2g023700",
"right", "2DCn2-"),
(Co, "Bo2g024450 Bo2g025390", "left",
"3DCn2-"), (Co, "Bo2g081060 Bo2g082340", "left", "4DCn2+"),
(Co, "Bo2g161510 Bo2g164260", "right", "5DCn2-"))
for ax, genes, ha, label in annotations:
g1, g2 = genes.split()
x1, x2 = order[g1][1].start, order[g2][1].start
if ha == "center":
x = (x1 + x2) / 2 * .8
elif ha == "left":
x = x2
else:
x = x1
label = r"\textit{{{0}}}".format(label)
color = rr if "+" in label else gg
ax.text(x, 30, label, color=color, fontsize=9, ha=ha, va="center")
ax_Ar.set_xlim(0, tracks[1].total)
ax_Ar.set_ylim(-1, 1)
ax_Co.set_xlim(0, tracks[2].total)
ax_Co.set_ylim(-1, 1)
# Plot coverage in resequencing lines
gstep = 5000000
order = "swede,kale,h165,yudal,aviso,abu,bristol".split(",")
labels_dict = {"h165": "Resynthesized (H165)", "abu": "Aburamasari"}
hlsuffix = "regions.forhaibao"
chr1, chr2 = "chrA02", "chrC02"
t1, t2 = tracks[0], tracks[-1]
s1, s2 = sizes[chr1], sizes[chr2]
canvas1 = (t1.xstart, .75, t1.xend - t1.xstart, .2)
c = Coverage(fig,
root,
canvas1,
chr1, (0, s1),
datadir,
order=order,
gauge=None,
plot_chr_label=False,
gauge_step=gstep,
palette="gray",
cap=40,
hlsuffix=hlsuffix,
labels_dict=labels_dict,
diverge=diverge)
yys = c.yys
x1, x2 = .37, .72
tip = .02
annotations = ((x1, yys[2] + .3 * tip, tip, tip / 2,
"FLC"), (x1, yys[3] + .6 * tip, tip, tip / 2, "FLC"),
(x1, yys[5] + .6 * tip, tip, tip / 2,
"FLC"), (x2, yys[0] + .9 * tip, -1.2 * tip, 0, "GSL"),
(x2, yys[4] + .9 * tip, -1.2 * tip, 0,
"GSL"), (x2, yys[6] + .9 * tip, -1.2 * tip, 0, "GSL"))
arrowprops = dict(facecolor='black',
shrink=.05,
frac=.5,
width=1,
headwidth=4)
for x, y, dx, dy, label in annotations:
label = r"\textit{{{0}}}".format(label)
root.annotate(label,
xy=(x, y),
xytext=(x + dx, y + dy),
arrowprops=arrowprops,
color=rr,
fontsize=9,
ha="center",
va="center")
canvas2 = (t2.xstart, .05, t2.xend - t2.xstart, .2)
Coverage(fig,
root,
canvas2,
chr2, (0, s2),
datadir,
order=order,
gauge=None,
plot_chr_label=False,
gauge_step=gstep,
palette="gray",
cap=40,
hlsuffix=hlsuffix,
labels_dict=labels_dict,
diverge=diverge)
pad = .03
labels = ((.1, .67, "A"), (t1.xstart - 3 * pad, .95 + pad, "B"),
(t2.xstart - 3 * pad, .25 + pad, "C"))
panel_labels(root, labels)
normalize_axes(root)
image_name = "napus-fig3." + iopts.format
savefig(image_name, dpi=iopts.dpi, iopts=iopts)
def seeds(args):
"""
%prog seeds [pngfile|jpgfile]
Extract seed metrics from [pngfile|jpgfile]. Use --rows and --cols to crop image.
"""
p = OptionParser(seeds.__doc__)
p.set_outfile()
opts, args, iopts = add_seeds_options(p, args)
if len(args) != 1:
sys.exit(not p.print_help())
(pngfile, ) = args
pf = opts.prefix or op.basename(pngfile).rsplit(".", 1)[0]
sigma, kernel = opts.sigma, opts.kernel
rows, cols = opts.rows, opts.cols
labelrows, labelcols = opts.labelrows, opts.labelcols
ff = opts.filter
calib = opts.calibrate
outdir = opts.outdir
if outdir != ".":
mkdir(outdir)
if calib:
calib = json.load(must_open(calib))
pixel_cm_ratio, tr = calib["PixelCMratio"], calib["RGBtransform"]
tr = np.array(tr)
nbcolor = opts.changeBackground
pngfile = convert_background(pngfile, nbcolor)
resizefile, mainfile, labelfile, exif = convert_image(
pngfile,
pf,
outdir=outdir,
rotate=opts.rotate,
rows=rows,
cols=cols,
labelrows=labelrows,
labelcols=labelcols,
)
oimg = load_image(resizefile)
img = load_image(mainfile)
fig, (ax1, ax2, ax3, ax4) = plt.subplots(ncols=4,
nrows=1,
figsize=(iopts.w, iopts.h))
# Edge detection
img_gray = rgb2gray(img)
logging.debug("Running {0} edge detection ...".format(ff))
if ff == "canny":
edges = canny(img_gray, sigma=opts.sigma)
elif ff == "roberts":
edges = roberts(img_gray)
elif ff == "sobel":
edges = sobel(img_gray)
edges = clear_border(edges, buffer_size=opts.border)
selem = disk(kernel)
closed = closing(edges, selem) if kernel else edges
filled = binary_fill_holes(closed)
# Watershed algorithm
if opts.watershed:
distance = distance_transform_edt(filled)
local_maxi = peak_local_max(distance,
threshold_rel=0.05,
indices=False)
coordinates = peak_local_max(distance, threshold_rel=0.05)
markers, nmarkers = label(local_maxi, return_num=True)
logging.debug("Identified {0} watershed markers".format(nmarkers))
labels = watershed(closed, markers, mask=filled)
else:
labels = label(filled)
# Object size filtering
w, h = img_gray.shape
canvas_size = w * h
min_size = int(round(canvas_size * opts.minsize / 100))
max_size = int(round(canvas_size * opts.maxsize / 100))
logging.debug(
"Find objects with pixels between {0} ({1}%) and {2} ({3}%)".format(
min_size, opts.minsize, max_size, opts.maxsize))
# Plotting
ax1.set_title("Original picture")
ax1.imshow(oimg)
params = "{0}, $\sigma$={1}, $k$={2}".format(ff, sigma, kernel)
if opts.watershed:
params += ", watershed"
ax2.set_title("Edge detection\n({0})".format(params))
closed = gray2rgb(closed)
ax2_img = labels
if opts.edges:
ax2_img = closed
elif opts.watershed:
ax2.plot(coordinates[:, 1], coordinates[:, 0], "g.")
ax2.imshow(ax2_img, cmap=iopts.cmap)
ax3.set_title("Object detection")
ax3.imshow(img)
filename = op.basename(pngfile)
if labelfile:
accession = extract_label(labelfile)
else:
accession = pf
# Calculate region properties
rp = regionprops(labels)
rp = [x for x in rp if min_size <= x.area <= max_size]
nb_labels = len(rp)
logging.debug("A total of {0} objects identified.".format(nb_labels))
objects = []
for i, props in enumerate(rp):
i += 1
if i > opts.count:
break
y0, x0 = props.centroid
orientation = props.orientation
major, minor = props.major_axis_length, props.minor_axis_length
major_dx = cos(orientation) * major / 2
major_dy = sin(orientation) * major / 2
minor_dx = sin(orientation) * minor / 2
minor_dy = cos(orientation) * minor / 2
ax2.plot((x0 - major_dx, x0 + major_dx),
(y0 + major_dy, y0 - major_dy), "r-")
ax2.plot((x0 - minor_dx, x0 + minor_dx),
(y0 - minor_dy, y0 + minor_dy), "r-")
npixels = int(props.area)
# Sample the center of the blob for color
d = min(int(round(minor / 2 * 0.35)) + 1, 50)
x0d, y0d = int(round(x0)), int(round(y0))
square = img[(y0d - d):(y0d + d), (x0d - d):(x0d + d)]
pixels = []
for row in square:
pixels.extend(row)
logging.debug("Seed #{0}: {1} pixels ({2} sampled) - {3:.2f}%".format(
i, npixels, len(pixels), 100.0 * npixels / canvas_size))
rgb = pixel_stats(pixels)
objects.append(Seed(filename, accession, i, rgb, props, exif))
minr, minc, maxr, maxc = props.bbox
rect = Rectangle((minc, minr),
maxc - minc,
maxr - minr,
fill=False,
ec="w",
lw=1)
ax3.add_patch(rect)
mc, mr = (minc + maxc) / 2, (minr + maxr) / 2
ax3.text(mc,
mr,
"{0}".format(i),
color="w",
ha="center",
va="center",
size=6)
for ax in (ax2, ax3):
ax.set_xlim(0, h)
ax.set_ylim(w, 0)
# Output identified seed stats
ax4.text(0.1, 0.92, "File: {0}".format(latex(filename)), color="g")
ax4.text(0.1, 0.86, "Label: {0}".format(latex(accession)), color="m")
yy = 0.8
fw = must_open(opts.outfile, "w")
if not opts.noheader:
print(Seed.header(calibrate=calib), file=fw)
for o in objects:
if calib:
o.calibrate(pixel_cm_ratio, tr)
print(o, file=fw)
i = o.seedno
if i > 7:
continue
ax4.text(0.01, yy, str(i), va="center", bbox=dict(fc="none", ec="k"))
ax4.text(0.1, yy, o.pixeltag, va="center")
yy -= 0.04
ax4.add_patch(
Rectangle((0.1, yy - 0.025),
0.12,
0.05,
lw=0,
fc=rgb_to_hex(o.rgb)))
ax4.text(0.27, yy, o.hashtag, va="center")
yy -= 0.06
ax4.text(
0.1,
yy,
"(A total of {0} objects displayed)".format(nb_labels),
color="darkslategray",
)
normalize_axes(ax4)
for ax in (ax1, ax2, ax3):
xticklabels = [int(x) for x in ax.get_xticks()]
yticklabels = [int(x) for x in ax.get_yticks()]
ax.set_xticklabels(xticklabels, family="Helvetica", size=8)
ax.set_yticklabels(yticklabels, family="Helvetica", size=8)
image_name = op.join(outdir, pf + "." + iopts.format)
savefig(image_name, dpi=iopts.dpi, iopts=iopts)
return objects
def estimategaps(args):
"""
%prog estimategaps JM-4 chr1 JMMale-1
Illustrate ALLMAPS gap estimation algorithm.
"""
p = OptionParser(estimategaps.__doc__)
opts, args, iopts = p.set_image_options(args, figsize="6x6", dpi=300)
if len(args) != 3:
sys.exit(not p.print_help())
pf, seqid, mlg = args
bedfile = pf + ".lifted.bed"
agpfile = pf + ".agp"
function = lambda x: x.cm
cc = Map(bedfile, scaffold_info=True, function=function)
agp = AGP(agpfile)
g = GapEstimator(cc, agp, seqid, mlg, function=function)
pp, chrsize, mlgsize = g.pp, g.chrsize, g.mlgsize
spl, spld = g.spl, g.spld
g.compute_all_gaps(verbose=False)
fig = plt.figure(1, (iopts.w, iopts.h))
root = fig.add_axes([0, 0, 1, 1])
# Panel A
xstart, ystart = 0.15, 0.65
w, h = 0.7, 0.3
t = np.linspace(0, chrsize, 1000)
ax = fig.add_axes([xstart, ystart, w, h])
mx, my = zip(*g.scatter_data)
rho = spearmanr(mx, my)
dsg = "g"
ax.vlines(pp, 0, mlgsize, colors="beige")
ax.plot(mx, my, ".", color=set2[3])
ax.plot(t, spl(t), "-", color=dsg)
ax.text(0.05, 0.95, mlg, va="top", transform=ax.transAxes)
normalize_lms_axis(ax, xlim=chrsize, ylim=mlgsize, ylabel="Genetic distance (cM)")
if rho < 0:
ax.invert_yaxis()
# Panel B
ystart -= 0.28
h = 0.25
ax = fig.add_axes([xstart, ystart, w, h])
ax.vlines(pp, 0, mlgsize, colors="beige")
ax.plot(t, spld(t), "-", lw=2, color=dsg)
ax.plot(pp, spld(pp), "o", mfc="w", mec=dsg, ms=5)
normalize_lms_axis(
ax,
xlim=chrsize,
ylim=25 * 1e-6,
xfactor=1e-6,
xlabel="Physical position (Mb)",
yfactor=1000000,
ylabel="Recomb. rate\n(cM / Mb)",
)
ax.xaxis.grid(False)
# Panel C (specific to JMMale-1)
a, b = "scaffold_1076", "scaffold_861"
sizes = dict(
(x.component_id, (x.object_beg, x.object_end, x.component_span, x.orientation))
for x in g.agp
if not x.is_gap
)
a_beg, a_end, asize, ao = sizes[a]
b_beg, b_end, bsize, bo = sizes[b]
gapsize = g.get_gapsize(a)
total_size = asize + gapsize + bsize
ratio = 0.6 / total_size
y = 0.16
pad = 0.03
pb_ratio = w / chrsize
# Zoom
lsg = "lightslategray"
root.plot((0.15 + pb_ratio * a_beg, 0.2), (ystart, ystart - 0.14), ":", color=lsg)
root.plot((0.15 + pb_ratio * b_end, 0.3), (ystart, ystart - 0.08), ":", color=lsg)
ends = []
for tag, size, marker, beg in zip(
(a, b), (asize, bsize), (49213, 81277), (0.2, 0.2 + (asize + gapsize) * ratio)
):
end = beg + size * ratio
marker = beg + marker * ratio
ends.append((beg, end, marker))
root.plot((marker,), (y,), "o", color=lsg)
root.text((beg + end) / 2, y + pad, latex(tag), ha="center", va="center")
HorizontalChromosome(root, beg, end, y, height=0.025, fc="gainsboro")
begs, ends, markers = zip(*ends)
fontprop = dict(color=lsg, ha="center", va="center")
ypos = y + pad * 2
root.plot(markers, (ypos, ypos), "-", lw=2, color=lsg)
root.text(
sum(markers) / 2,
ypos + pad,
"Distance: 1.29cM $\Leftrightarrow$ 211,824bp (6.1 cM/Mb)",
**fontprop
)
ypos = y - pad
xx = markers[0], ends[0]
root.plot(xx, (ypos, ypos), "-", lw=2, color=lsg)
root.text(sum(xx) / 2, ypos - pad, "34,115bp", **fontprop)
xx = markers[1], begs[1]
root.plot(xx, (ypos, ypos), "-", lw=2, color=lsg)
root.text(sum(xx) / 2, ypos - pad, "81,276bp", **fontprop)
root.plot((ends[0], begs[1]), (y, y), ":", lw=2, color=lsg)
root.text(
sum(markers) / 2,
ypos - 3 * pad,
r"$\textit{Estimated gap size: 96,433bp}$",
color="r",
ha="center",
va="center",
)
labels = ((0.05, 0.95, "A"), (0.05, 0.6, "B"), (0.05, 0.27, "C"))
panel_labels(root, labels)
normalize_axes(root)
pf = "estimategaps"
image_name = pf + "." + iopts.format
savefig(image_name, dpi=iopts.dpi, iopts=iopts)
def fig3(args):
"""
%prog fig3 chrA02,A02,C2,chrC02 chr.sizes all.bed data
Napus Figure 3 displays alignments between quartet chromosomes, inset
with read histograms.
"""
from jcvi.formats.bed import Bed
p = OptionParser(fig3.__doc__)
p.add_option("--gauge_step", default=10000000, type="int",
help="Step size for the base scale")
opts, args, iopts = p.set_image_options(args, figsize="12x9")
if len(args) != 4:
sys.exit(not p.print_help())
chrs, sizes, bedfile, datadir = args
gauge_step = opts.gauge_step
diverge = iopts.diverge
rr, gg = diverge
chrs = [[x] for x in chrs.split(",")]
sizes = Sizes(sizes).mapping
fig = plt.figure(1, (iopts.w, iopts.h))
root = fig.add_axes([0, 0, 1, 1])
chr_sizes, chr_sum_sizes, ratio = calc_ratio(chrs, sizes)
# Synteny panel
seqidsfile = make_seqids(chrs)
klayout = make_layout(chrs, chr_sum_sizes, ratio, template_f3a, shift=.05)
height = .07
r = height / 4
K = Karyotype(fig, root, seqidsfile, klayout, gap=gap,
height=height, lw=2, generank=False, sizes=sizes,
heightpad=r, roundrect=True, plot_label=False)
# Chromosome labels
for kl in K.layout:
if kl.empty:
continue
lx, ly = kl.xstart, kl.y
if lx < .11:
lx += .1
ly += .06
label = kl.label
root.text(lx - .015, ly, label, fontsize=15,
ha="right", va="center")
# Inset with datafiles
datafiles = ("chrA02.bzh.forxmgr", "parent.A02.per10kb.forxmgr",
"parent.C2.per10kb.forxmgr", "chrC02.bzh.forxmgr")
datafiles = [op.join(datadir, x) for x in datafiles]
tracks = K.tracks
hlfile = op.join(datadir, "bzh.regions.forhaibao")
xy_axes = []
for t, datafile in zip(tracks, datafiles):
ax = make_affix_axis(fig, t, -r, height=2 * r)
xy_axes.append(ax)
chr = t.seqids[0]
xy = XYtrack(ax, datafile, color="lightslategray")
start, end = 0, t.total
xy.interpolate(end)
xy.cap(ymax=40)
xy.import_hlfile(hlfile, chr, diverge=diverge)
xy.draw()
ax.set_xlim(start, end)
gauge_ax = make_affix_axis(fig, t, -r)
adjust_spines(gauge_ax, ["bottom"])
setup_gauge_ax(gauge_ax, start, end, gauge_step)
# Converted gene tracks
ax_Ar = make_affix_axis(fig, tracks[1], r, height=r/2)
ax_Co = make_affix_axis(fig, tracks[2], r, height=r/2)
order = Bed(bedfile).order
for asterisk in (False, True):
conversion_track(order, "data/Genes.Converted.seuil.0.6.AtoC.txt",
0, "A02", ax_Ar, rr, asterisk=asterisk)
conversion_track(order, "data/Genes.Converted.seuil.0.6.AtoC.txt",
1, "C2", ax_Co, gg, asterisk=asterisk)
conversion_track(order, "data/Genes.Converted.seuil.0.6.CtoA.txt",
0, "A02", ax_Ar, gg, ypos=1, asterisk=asterisk)
conversion_track(order, "data/Genes.Converted.seuil.0.6.CtoA.txt",
1, "C2", ax_Co, rr, ypos=1, asterisk=asterisk)
Ar, Co = xy_axes[1:3]
annotations = ((Ar, "Bra028920 Bra028897", "center", "1DAn2+"),
(Ar, "Bra020081 Bra020171", "right", "2DAn2+"),
(Ar, "Bra020218 Bra020286", "left", "3DAn2+"),
(Ar, "Bra008143 Bra008167", "left", "4DAn2-"),
(Ar, "Bra029317 Bra029251", "right", "5DAn2+ (GSL)"),
(Co, "Bo2g001000 Bo2g001300", "left", "1DCn2-"),
(Co, "Bo2g018560 Bo2g023700", "right", "2DCn2-"),
(Co, "Bo2g024450 Bo2g025390", "left", "3DCn2-"),
(Co, "Bo2g081060 Bo2g082340", "left", "4DCn2+"),
(Co, "Bo2g161510 Bo2g164260", "right", "5DCn2-"))
for ax, genes, ha, label in annotations:
g1, g2 = genes.split()
x1, x2 = order[g1][1].start, order[g2][1].start
if ha == "center":
x = (x1 + x2) / 2 * .8
elif ha == "left":
x = x2
else:
x = x1
label = r"\textit{{{0}}}".format(label)
color = rr if "+" in label else gg
ax.text(x, 30, label, color=color, fontsize=9, ha=ha, va="center")
ax_Ar.set_xlim(0, tracks[1].total)
ax_Ar.set_ylim(-1, 1)
ax_Co.set_xlim(0, tracks[2].total)
ax_Co.set_ylim(-1, 1)
# Plot coverage in resequencing lines
gstep = 5000000
order = "swede,kale,h165,yudal,aviso,abu,bristol".split(",")
labels_dict = {"h165": "Resynthesized (H165)", "abu": "Aburamasari"}
hlsuffix = "regions.forhaibao"
chr1, chr2 = "chrA02", "chrC02"
t1, t2 = tracks[0], tracks[-1]
s1, s2 = sizes[chr1], sizes[chr2]
canvas1 = (t1.xstart, .75, t1.xend - t1.xstart, .2)
c = Coverage(fig, root, canvas1, chr1, (0, s1), datadir,
order=order, gauge=None, plot_chr_label=False,
gauge_step=gstep, palette="gray",
cap=40, hlsuffix=hlsuffix, labels_dict=labels_dict,
diverge=diverge)
yys = c.yys
x1, x2 = .37, .72
tip = .02
annotations = ((x1, yys[2] + .3 * tip, tip, tip / 2, "FLC"),
(x1, yys[3] + .6 * tip, tip, tip / 2, "FLC"),
(x1, yys[5] + .6 * tip, tip, tip / 2, "FLC"),
(x2, yys[0] + .9 * tip, -1.2 * tip, 0, "GSL"),
(x2, yys[4] + .9 * tip, -1.2 * tip, 0, "GSL"),
(x2, yys[6] + .9 * tip, -1.2 * tip, 0, "GSL"))
arrowprops=dict(facecolor='black', shrink=.05, frac=.5,
width=1, headwidth=4)
for x, y, dx, dy, label in annotations:
label = r"\textit{{{0}}}".format(label)
root.annotate(label, xy=(x, y), xytext=(x + dx, y + dy),
arrowprops=arrowprops, color=rr, fontsize=9,
ha="center", va="center")
canvas2 = (t2.xstart, .05, t2.xend - t2.xstart, .2)
Coverage(fig, root, canvas2, chr2, (0, s2), datadir,
order=order, gauge=None, plot_chr_label=False,
gauge_step=gstep, palette="gray",
cap=40, hlsuffix=hlsuffix, labels_dict=labels_dict,
diverge=diverge)
pad = .03
labels = ((.1, .67, "A"), (t1.xstart - 3 * pad, .95 + pad, "B"),
(t2.xstart - 3 * pad, .25 + pad, "C"))
panel_labels(root, labels)
normalize_axes(root)
image_name = "napus-fig3." + iopts.format
savefig(image_name, dpi=iopts.dpi, iopts=iopts)
def diagram(args):
"""
%prog diagram
Plot the predictive power of various evidences.
"""
p = OptionParser(diagram.__doc__)
opts, args, iopts = p.set_image_options(args, figsize="8x4")
if len(args) != 0:
sys.exit(not p.print_help())
fig = plt.figure(1, (iopts.w, iopts.h))
root = fig.add_axes([0, 0, 1, 1])
# Gauge on top, this is log-scale
lsg = "lightslategray"
yy = .7
yinterval = .1
height = .05
yp = yy - yinterval - height
canvas = .95
xstart = .025
convert = lambda x: xstart + x * canvas / 1200
# Symbols
root.text(.5,
.9,
r"$L$: Read length, $F$: Flank size, $V$: Pair distance",
ha="center")
root.text(.5, .85, r"ex. $L=150bp, F=9bp, V=500bp$", ha="center")
root.text(xstart + canvas,
yy - height,
"STR repeat length",
ha="center",
color=lsg,
size=10)
# Mark the key events
pad = .02
arrowlen = canvas * 1.05
arrowprops = dict(length_includes_head=True,
width=.01,
fc=lsg,
lw=0,
head_length=arrowlen * .12,
head_width=.04)
p = FancyArrow(xstart, yy, arrowlen, 0, shape="right", **arrowprops)
root.add_patch(p)
ppad = 30
keyevents = (
(0, 0, -1, r"$0$"),
(150 - 18, 150 - 18 - ppad, 0, r"$L - 2F$"),
(150 - 9, 150 - 9, 1, r"$L - F$"),
(150, 150 + ppad, 2, r"$L$"),
(500 - 9, 500 - 9, 3, r"$V - F$"),
(500 * 2 - 18, 500 * 2 - 18, 2, r"$2(V - F)$"),
)
for event, pos, i, label in keyevents:
_event = convert(event)
_pos = convert(pos)
root.plot((_event, _event), (yy - height / 4, yy + height / 4),
'-',
color='k')
root.text(_pos, yy + pad, label, rotation=45, va="bottom", size=8)
if i < 0:
continue
ystart = yp - i * yinterval
root.plot((_event, _event), (ystart, yy - height / 4), ':', color=lsg)
# Range on bottom. These are simple 4 rectangles, with the range indicating
# the predictive range.
CLOSED, OPEN = range(2)
ranges = (
(0, 150 - 18, CLOSED, "Spanning reads"),
(9, 150 - 9, OPEN, "Partial reads"),
(150, 500 * 2 - 18, CLOSED, "Repeat reads"),
(0, 500 - 9, CLOSED, "Paired-end reads"),
)
for start, end, starttag, label in ranges:
_start = convert(start)
_end = convert(end)
data = [[0., 1.], [0., 1.]] if starttag == OPEN else \
[[1., 0.], [1., 0.]]
root.imshow(data,
interpolation='bicubic',
cmap=plt.cm.Greens,
extent=[_start, _end, yp, yp + height])
root.text(_end + pad, yp + height / 2, label, va="center")
yp -= yinterval
normalize_axes(root)
image_name = "diagram." + iopts.format
savefig(image_name, dpi=iopts.dpi, iopts=iopts)
def main():
p = OptionParser(__doc__)
opts, args, iopts = p.set_image_options(figsize="9x7")
if len(args) != 1:
sys.exit(not p.print_help())
mode, = args
assert mode == "demo"
a, b = 30, 70
pad = .08
w = .31
fig = plt.figure(1, (iopts.w, iopts.h))
root = fig.add_axes([0, 0, 1, 1])
# Row separators
yy = 1 - pad
for i in xrange(3):
root.plot((0, 1), (yy, yy), "-", lw=2, color="lightgray")
yy -= w
# Row headers
xx = pad * .6
yy = 1 - pad - .5 * w
for title in ("Inversion", "Indel", "Duplication"):
root.text(xx, yy, title, ha="center", va="center")
yy -= w
# Column headers
xx = pad + .5 * w
yy = 1 - pad / 2
for title in ("Assembly alignment", "Read alignment", "Optical map alignment"):
root.text(xx, yy, title, ha="center", va="center")
xx += w
p = PairwiseAlign(fig, [pad, 2 * w, w, w])
p.invert(a, b)
p.draw()
p = PairwiseAlign(fig, [pad, w, w, w])
p.delete(a, b)
p.draw()
p = PairwiseAlign(fig, [pad, 0, w, w])
p.duplicate(a, b, gap=5)
p.draw()
p = ReadAlign(fig, [pad + w, 2 * w, w, w])
p.invert(a, b)
p.draw()
p = ReadAlign(fig, [pad + w, w, w, w])
p.delete(a, b)
p.draw()
p = ReadAlign(fig, [pad + w, 0, w, w])
p.duplicate(a, b)
p.draw()
p = OpticalMapAlign(fig, [pad + 2 * w, 2 * w, w, w])
p.invert(a, b)
p.draw()
p = OpticalMapAlign(fig, [pad + 2 * w, w, w, w])
p.delete(a, b)
p.draw()
p = OpticalMapAlign(fig, [pad + 2 * w, 0, w, w])
p.duplicate(a, b)
p.draw()
normalize_axes(root)
image_name = mode + "." + 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)
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 depth(args):
"""
%prog depth anchorfile --qbed qbedfile --sbed sbedfile
Calculate the depths in the two genomes in comparison, given in --qbed and
--sbed. The synteny blocks will be layered on the genomes, and the
multiplicity will be summarized to stderr.
"""
from jcvi.utils.range import range_depth
p = OptionParser(depth.__doc__)
p.add_option("--depthfile",
help="Generate file with gene and depth [default: %default]")
p.add_option("--histogram", default=False, action="store_true",
help="Plot histograms in PDF")
p.add_option("--xmax", type="int", help="x-axis maximum to display in plot")
p.add_option("--title", default=None, help="Title to display in plot")
p.add_option("--quota", help="Force to use this quota, e.g. 1:1, 1:2 ...")
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)
depthfile = opts.depthfile
ac = AnchorFile(anchorfile)
qranges = []
sranges = []
blocks = ac.blocks
for ib in blocks:
q, s, t = zip(*ib)
q = [qorder[x] for x in q]
s = [sorder[x] for x in s]
qrange = (min(q)[0], max(q)[0])
srange = (min(s)[0], max(s)[0])
qranges.append(qrange)
sranges.append(srange)
if is_self:
qranges.append(srange)
qgenome = op.basename(qbed.filename).split(".")[0]
sgenome = op.basename(sbed.filename).split(".")[0]
qtag = "Genome {0} depths".format(qgenome)
print("{}:".format(qtag), file=sys.stderr)
dsq, details = range_depth(qranges, len(qbed))
if depthfile:
fw = open(depthfile, "w")
write_details(fw, details, qbed)
if is_self:
return
stag = "Genome {0} depths".format(sgenome)
print("{}:".format(stag), file=sys.stderr)
dss, details = range_depth(sranges, len(sbed))
if depthfile:
write_details(fw, details, sbed)
fw.close()
logging.debug("Depth written to `{0}`.".format(depthfile))
if not opts.histogram:
return
from jcvi.graphics.base import plt, quickplot_ax, savefig, normalize_axes
# Plot two histograms one for query genome, one for subject genome
plt.figure(1, (6, 3))
f, (ax1, ax2) = plt.subplots(1, 2, sharey=True)
xmax = opts.xmax or max(4, max(dsq.keys() + dss.keys()))
if opts.quota:
speak, qpeak = opts.quota.split(":")
qpeak, speak = int(qpeak), int(speak)
else:
qpeak = find_peak(dsq)
speak = find_peak(dss)
qtag = "# of {} blocks per {} gene".format(sgenome, qgenome)
stag = "# of {} blocks per {} gene".format(qgenome, sgenome)
quickplot_ax(ax1, dss, 0, xmax, stag, ylabel="Percentage of genome",
highlight=range(1, speak + 1))
quickplot_ax(ax2, dsq, 0, xmax, qtag, ylabel=None,
highlight=range(1, qpeak + 1))
title = opts.title or "{} vs {} syntenic depths\n{}:{} pattern"\
.format(qgenome, sgenome, speak, qpeak)
root = f.add_axes([0, 0, 1, 1])
vs, pattern = title.split('\n')
root.text(.5, .97, vs, ha="center", va="center", color="darkslategray")
root.text(.5, .925, pattern, ha="center", va="center",
color="tomato", size=16)
print(title, file=sys.stderr)
normalize_axes(root)
pf = anchorfile.rsplit(".", 1)[0] + ".depth"
image_name = pf + ".pdf"
savefig(image_name)
