def drawPca(rownames, transformedM, outfile, options):
#Draw V usage distribution of each sample onto the same pdf
options.out = outfile
fig, pdf = immunoseqLib.initImage( 12.5, 8.0, options )
axes = immunoseqLib.setAxes(fig)
drawPcaData( axes, rownames, transformedM, options )
immunoseqLib.writeImage( fig, pdf, options )
def drawOverlapReads(exps, options):
sample = options.sample
options.out = os.path.join(options.outdir, "overlapPlot-%s" % sample)
fig, pdf = iseqlib.initImage( 10.0, 12.0, options )
axes = iseqlib.setAxes(fig)
drawOverlapReadsData(axes, exps, sample, options.sampleOrder)
iseqlib.writeImage(fig, pdf, options)
def drawClonesizeDist(outdir, options, sample2data):
options.out = os.path.join( outdir, "cloneVsRead" )
fig, pdf = iseqlib.initImage( 10.0, 8.0, options )
axes = fig.add_axes( [0.12, 0.15, 0.85, 0.75] )
lines = []
sampleNames = sorted(sample2data.keys())
#draw
colors = iseqlib.getColors6()
lightColors = iseqlib.getColors6light()
markers = ['o', '*', 's', 'd', '^', 'p', 'v']
c = 0
xdata = []
xticklabels = []
for sample in sampleNames:
#xticklabels, ydata = binData( sample2freqs[sample] )
(xticklabels, ydata) = sample2data[sample]
print sample
print xdata
print ydata
xdata = range(0, len(xticklabels), 1)
markersize = 10.0
m = markers[c]
if m == '*':
markersize = 12.0
elif m == 's':
markersize = 8.0
l = axes.plot(xdata, ydata, color=colors[c], marker=m, markeredgecolor=colors[c], markersize=markersize, linestyle='none')
axes.plot(xdata, ydata, color=lightColors[c], linestyle='-', linewidth=0.2)
lines.append(l)
c += 1
iseqlib.editSpine(axes)
axes.set_title('Clone Size Distribution', size='xx-large')
axes.set_xlabel('Clone size as percentage of total reads', size='large')
axes.set_ylabel('Percentage of total clones', size='large')
fontP = FontProperties()
fontP.set_size('medium')
legend = axes.legend( lines, sampleNames, numpoints=1, loc='best', ncol=1, prop=fontP)
legend._drawFrame = False
axes.xaxis.set_ticklabels( xticklabels )
axes.xaxis.set_ticks( [x - 0.5 for x in xdata] )
axes.set_xlim(-0.5, len(xdata) - 0.5)
axes.set_ylim(-0.5, 50)
for label in axes.get_xticklabels():
label.set_fontsize('medium')
label.set_rotation(45)
for label in axes.get_yticklabels():
label.set_fontsize('medium')
axes.yaxis.grid(b=True, color="#BDBDBD", linestyle='-', linewidth=0.005)
axes.xaxis.grid(b=True, color="#BDBDBD", linestyle='-', linewidth=0.005)
iseqlib.writeImage(fig, pdf, options)
def drawUsageDist( samples, stds, options, genetype, intersectGenes ):
#Draw V usage distribution of each sample onto the same pdf
options.out = os.path.join( options.outdir, "%sUsage" %genetype)
fig, pdf = immunoseqLib.initImage( 10.0, 12.0, options )
axesList = setUsageDistAxes( fig, len(samples), options.samplesPerPlot )
drawUsageData( axesList, samples, stds, options, genetype, intersectGenes )
immunoseqLib.writeImage( fig, pdf, options )
def drawVJ( sample, vgenes, jgenes, options, minvj, maxvj ):
options.out = os.path.join( options.outdir, "%s-vjUsage" %sample.name )
fig, pdf = immunoseqLib.initImage( 8.0, 10.0, options )
axes = immunoseqLib.setAxes(fig)
#axesList = setCompareAxes( fig )
drawVJdata( fig, axes, sample, vgenes, jgenes, options, minvj, maxvj )
immunoseqLib.writeImage( fig, pdf, options )
def drawCloneVsRead( samples, options, isAbs ):
options.out = os.path.join( options.outdir, "cloneVsRead" )
if not isAbs:
options.out = os.path.join( options.outdir, "cloneVsRead-Rel" )
#fig, pdf = libplot.initImage( 10.0, 8.0, options )
fig, pdf = iseqlib.initImage( 10.0, 8.0, options )
axesList = setUsageDistAxes( fig, len(samples), options.samplesPerPlot )
drawCloneVsReadData( axesList, samples, options.samplesPerPlot, options, isAbs )
#libplot.writeImage( fig, pdf, options )
iseqlib.writeImage( fig, pdf, options )
def drawCloneSizeDist( samples, options, isAbs, yaxisPcReads, yaxisPcClones, cumulative ):
options.out = os.path.join( options.outdir, "cloneSizeDist" )
if cumulative:
options.out += "-cumulative"
if not isAbs:
options.out = os.path.join( options.outdir, "cloneSizeDist-Rel" )
if yaxisPcReads:
options.out += "-pcReads"
if yaxisPcClones:
options.out += "-pcClones"
#fig, pdf = libplot.initImage( 10.0, 8.0, options )
fig, pdf = iseqlib.initImage( 10.0, 8.0, options )
axesList = setUsageDistAxes( fig, len(samples), options.samplesPerPlot )
drawCloneSizeData( axesList, samples, options.samplesPerPlot, options, isAbs, yaxisPcReads, yaxisPcClones, cumulative )
#libplot.writeImage( fig, pdf, options )
iseqlib.writeImage( fig, pdf, options )
def drawCombine( samples, options ):
options.out = os.path.join(options.outdir, "combine")
fig, pdf = iseqlib.initImage( 8.0, 12.0, options )
#fig, (ax1, ax2, ax3) = plt.subplots(nrows=3, ncols=1)
fig.subplots_adjust(hspace=.5)
#drawCloneSizeDist( samples, options, isAbs, yaxisPcReads, yaxisPcClones, cumulative )
ax1 = fig.add_subplot(311)
drawCloneSizeData( [ax1], samples, options.samplesPerPlot, options, False, False, True, True )
ax2 = fig.add_subplot(312)
drawCloneSizeData( [ax2], samples, options.samplesPerPlot, options, False, True, False, True )
ax3 = fig.add_subplot(313)
#drawCloneVsRead( samples, options, True )
drawCloneVsReadData( [ax3], samples, options.samplesPerPlot, options, True )
#Write to output file
#import matplotlib.backends.backend_pdf as pltBack
#pdf = None
#if options.outFormat == 'pdf' or options.outFormat == 'all':
# pdf = pltBack.PdfPages( options.out + '.pdf' )
iseqlib.writeImage( fig, pdf, options )
def drawDist(sam2nt2aa, options):
options.out = os.path.join(options.outdir, 'nt2aa')
fig, pdf = iseqlib.initImage(10.0, 10.0, options)
axes = iseqlib.setAxes(fig)
drawDistData(axes, sam2nt2aa)
iseqlib.writeImage(fig, pdf, options)
def drawDist(samples, options):
options.out = os.path.join(options.outdir, "sharedSeqsDist")
fig, pdf = iseqlib.initImage(10.0, 10.0, options)
axes = iseqlib.setAxes(fig)
drawDistData(axes, samples)
iseqlib.writeImage(fig, pdf, options)
def drawAll(options, outdir, rowname2cells, index2colname):
if options.infile == '-':
outname = 'all'
else:
outname = os.path.basename(options.infile).split('.')[0]
options.out = os.path.join(outdir, outname)
fig, pdf = iseqlib.initImage(10.0, 8.0, options)
axes = fig.add_axes( [0.12, 0.15, 0.85, 0.75] )
lines = []
#rownames = sorted( rowname2cells.keys() )
rownames = ['as11D', 'as16D', 'as1D', 'asBD', 'as20D', 'as15D', 'as8D']
name2color = sample2color(rownames)
#xdata = sorted( [ int(x) for x in colname2index.keys() ] )
xmax = 0.0
xmin = float('inf')
#ymax = 0
markersize = 12.0
xindices = []
#for rowname in rownames:
r = 0
while r < len(rownames):
rowname = rownames[r]
row = rowname2cells[rowname]
means = []
stds = []
xdata = []
for i, m in enumerate( row ):
if i % 2 == 0 and m != 'NA' and m != '' and m != '-':
colname = index2colname[i]
try:
colname = int(colname)
xdata.append( colname )
except:
xdata.append(i/2)
if i not in xindices:
xindices.append(i)
means.append( float(m) )
elif i%2 == 1 and m != 'NA' and m != '' and m != '-':
stds.append( float(m) )
xmax = max([xmax, max(xdata)])
xmin = min([xmin, min(xdata)])
#HACK
#if rowname == 'uniqClones':
#if rowname == 'mountford' or rowname == 'horn':
#if rowname != 'horn':
#if rowname == 'manhattan' or rowname == 'euclidean' or rowname == 'binomial':
exceptions = ['manhattan', 'euclidean', 'binomial', 'kulczynski', 'canberra', 'jaccard']
if rowname in exceptions:
rownames.remove(rowname)
continue
#means = [m/min(means) for m in means]
#stds = [0.0 for s in stds]
#END HACK
color = name2color[rowname]
axes.errorbar(xdata, means, yerr=stds, color=color, markeredgecolor=color, markersize=markersize, fmt='.')
line = axes.plot(xdata, means, color=color, linestyle='-', linewidth=4.0)
lines.append(line)
r += 1
#axes.set_xscale('log')
#axes.set_yscale('log')
iseqlib.editSpine(axes)
#axes.set_title("%s index across different sampling sizes" %outname, size='xx-large')
#axes.set_xlabel("Sampling size (number of reads)", size='large' )
#axes.set_ylabel("%s index" %outname, size='large')
#HACK
axes.set_title("Sequencing Saturation", size='xx-large', weight='bold')
axes.set_xlabel("Sampling size (number of sequences, in millions) ", size='x-large', weight='bold' )
#axes.set_xlabel("Sampling size (number of sequences, in thousands) ", size='x-large', weight='bold' )
axes.set_ylabel("Number of clones (in thousands)", size='x-large', weight='bold')
fontP = FontProperties()
fontP.set_size('medium')
#rownames = ["A", "B"]
rownames = [iseqlib.properName(n) for n in rownames]
legend = axes.legend( lines, rownames, numpoints = 1, loc='best', ncol = 1, prop=fontP)
legend._drawFrame = False
if len(xindices) > 0:
xticklabels = []
for i in sorted(xindices):
xticklabels.append( index2colname[i] )
axes.xaxis.set_ticks( xrange(len(xticklabels)) )
axes.xaxis.set_ticklabels( xticklabels )
#HACK:
#xticks = [ 10000, 50000, 100000, 200000, 300000, 400000, 500000]
#xticklabels = [ str(x/1000) for x in xticks]
xticks = [0, 1000000,2000000,3000000,4000000,5000000,6000000,7000000,8000000,9000000]
xticklabels = [ str(x) for x in xrange(0, 10) ]
axes.xaxis.set_ticks(xticks)
axes.xaxis.set_ticklabels( xticklabels )
#yticks = xrange(0, 121000, 20000)
#yticklabels = [ str(y) for y in xrange(0, 121, 20) ]
yticks = xrange(0, 250000, 50000)
yticklabels = [ str(y) for y in xrange(0, 250, 50) ]
axes.yaxis.set_ticks(yticks)
axes.yaxis.set_ticklabels(yticklabels)
for label in axes.get_xticklabels():
label.set_fontsize('large')
label.set_fontweight ('bold')
#label.set_rotation(45)
for label in axes.get_yticklabels():
label.set_fontsize('large')
label.set_fontweight ('bold')
axes.xaxis.grid(b=True, color='#3F3F3F', linestyle='-', linewidth=0.05)
axes.yaxis.grid(b=True, color='#3F3F3F', linestyle='-', linewidth=0.05)
#HACK
#axes.set_ylim(0.996, 1)
xspan = xmax - xmin
#axes.set_xlim(xmin - xspan*0.01, xmax + xspan*0.01)
#axes.set_ylim(10000, 121000)
#axes.set_xlim(-10, 110000)
#axes.set_ylim(-10, 60000)
axes.set_xlim(-10, 5100000)
axes.set_ylim(-10, 205000)
iseqlib.writeImage(fig, pdf, options)
