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 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 multilineplot(args):
"""
%prog multilineplot fastafile chr1
Combine multiple line plots in one vertical stack
Inputs must be BED-formatted.
--lines: traditional line plots, useful for plotting feature freq
"""
p = OptionParser(multilineplot.__doc__)
p.add_option("--lines",
help="Features to plot in lineplot [default: %default]")
p.add_option("--colors",
help="List of colors matching number of input bed files")
p.add_option("--mode", default="span", choices=("span", "count", "score"),
help="Accumulate feature based on [default: %default]")
p.add_option("--binned", default=False, action="store_true",
help="Specify whether the input is already binned; " +
"if True, input files are considered to be binfiles")
add_window_options(p)
opts, args, iopts = p.set_image_options(args, figsize="8x5")
if len(args) != 2:
sys.exit(not p.print_help())
fastafile, chr = args
window, shift, subtract = check_window_options(opts)
linebeds = []
colors = opts.colors
if opts.lines:
lines = opts.lines.split(",")
assert len(colors) == len(lines), "Number of chosen colors must match" + \
" number of input bed files"
linebeds = get_beds(lines, binned=opts.binned)
linebins = get_binfiles(linebeds, fastafile, shift, mode=opts.mode, binned=opts.binned)
clen = Sizes(fastafile).mapping[chr]
nbins = get_nbins(clen, shift)
plt.rcParams["xtick.major.size"] = 0
plt.rcParams["ytick.major.size"] = 0
plt.rcParams["figure.figsize"] = iopts.w, iopts.h
fig, axarr = plt.subplots(nrows=len(lines))
if len(linebeds) == 1:
axarr = (axarr, )
fig.suptitle(chr, color="darkslategray")
for i, ax in enumerate(axarr):
lineplot(ax, [linebins[i]], nbins, chr, window, shift, \
color="{0}{1}".format(colors[i], 'r'))
plt.subplots_adjust(hspace=0.5)
image_name = chr + "." + 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(__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 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 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 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 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 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)
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 >> sys.stderr, "{}:".format(qtag)
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 >> sys.stderr, "{}:".format(stag)
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 >> sys.stderr, title
normalize_axes(root)
pf = anchorfile.rsplit(".", 1)[0] + ".depth"
image_name = pf + ".pdf"
savefig(image_name)
def likelihood(args):
"""
%prog likelihood
Plot likelihood surface. Look for two files in the current folder:
- 100_100.log, haploid model
- 100_20.log, diploid model
"""
p = OptionParser(likelihood.__doc__)
opts, args, iopts = p.set_image_options(args,
figsize="10x5",
style="white",
cmap="coolwarm")
if len(args) != 0:
sys.exit(not p.print_help())
fig, (ax1, ax2) = plt.subplots(ncols=2,
nrows=1,
figsize=(iopts.w, iopts.h))
plt.tight_layout(pad=4)
# Haploid model
LL, CI_h1, CI_h2, MLE = parse_log("100_100.log")
data = []
for k, v in LL.items():
data.append((k[0], v))
data.sort()
x, y = zip(*data)
x = np.array(x)
curve, = ax1.plot(x, y, "-", color=lsg, lw=2)
ax1.set_title("Simulated haploid ($h^{truth}=100$)")
h_hat, max_LL = max(data, key=lambda x: x[-1])
_, min_LL = min(data, key=lambda x: x[-1])
ymin, ymax = ax1.get_ylim()
ax1.set_ylim([ymin, ymax + 30])
LL_label = "log(Likelihood)"
ax1.plot([h_hat, h_hat], [ymin, max_LL], ":", color=lsg, lw=2)
ax1.text(h_hat, max_LL + 10, r"$\hat{h}=93$", color=lsg)
ax1.set_xlabel(r"$h$")
ax1.set_ylabel(LL_label)
a, b = CI_h1
ci = ax1.fill_between(x, [ymin] * len(x),
y,
where=(x >= a) & (x <= b),
color=lsg,
alpha=.5)
ax1.legend([curve, ci], ["Likelihood curve", r'95$\%$ CI'], loc='best')
# Diploid model
LL, CI_h1, CI_h2, MLE = parse_log("100_20.log")
h_hat, max_LL = max(data, key=lambda x: x[-1])
_, min_LL = min(data, key=lambda x: x[-1])
data = np.ones((301, 301)) * min_LL
for k, v in LL.items():
a, b = k
data[a, b] = v
data[b, a] = v
data = mask_upper_triangle(data)
ax_imshow(ax2, data, opts.cmap, LL_label, 20, 104)
root = fig.add_axes([0, 0, 1, 1])
pad = .04
panel_labels(root, ((pad / 2, 1 - pad, "A"), (1 / 2., 1 - pad, "B")))
normalize_axes(root)
image_name = "likelihood." + iopts.format
savefig(image_name, dpi=iopts.dpi, iopts=iopts)