def reorder_pov_by_cluster(pov, mutIDs, config):
# reorder rows and columns of pov matrix
# to put mutations assigned to common clusters next to
# each other
clusterPath = os.path.join(config.working_dir, \
config.mutation_to_cluster_assignment)
if not os.path.exists(clusterPath):
# if no cluster assignment available
# return the matrix and ids as is
return pov, mutIDs
cluster2mut = read_cluster_assignments(clusterPath)
pairs = []
for cl in cluster2mut:
pairs.extend(zip(cluster2mut[cl], [cl] * len(cluster2mut[cl])))
print pairs
pairs = sorted(pairs, key=lambda x: x[1])
print pairs
sortedmutIDs = zip(*pairs)[0]
print sortedmutIDs
sortedPov = np.zeros(pov.shape)
for pnID in sortedmutIDs:
for cnID in sortedmutIDs:
sortedPov[sortedmutIDs.index(pnID), sortedmutIDs.index(cnID)] = pov[mutIDs.index(pnID), \
mutIDs.index(cnID)]
return sortedPov, sortedmutIDs
def reorder_pov_by_cluster(pov, mutIDs, config):
# reorder rows and columns of pov matrix
# to put mutations assigned to common clusters next to
# each other
clusterPath = os.path.join(config.working_dir, \
config.mutation_to_cluster_assignment)
if not os.path.exists(clusterPath):
# if no cluster assignment available
# return the matrix and ids as is
return pov, mutIDs
cluster2mut = read_cluster_assignments(clusterPath)
pairs = []
for cl in cluster2mut:
pairs.extend(zip(cluster2mut[cl], [cl] * len(cluster2mut[cl])))
print pairs
pairs = sorted(pairs, key = lambda x: x[1])
print pairs
sortedmutIDs = zip(*pairs)[0]
print sortedmutIDs
sortedPov = np.zeros(pov.shape)
for pnID in sortedmutIDs:
for cnID in sortedmutIDs:
sortedPov[sortedmutIDs.index(pnID), sortedmutIDs.index(cnID)] = pov[mutIDs.index(pnID), \
mutIDs.index(cnID)]
return sortedPov, sortedmutIDs
def plot_mut_clust_cellularity(args):
# plot estimated cellularity for mutations and clusters
# across samples
try:
import matplotlib
matplotlib.use('Agg')
import matplotlib.pyplot as plt
except:
print >>sys.stderr, \
'Plotting mutation/cluster cellularities requires matplotlib \n' +\
'module. Failed to import matplotlib. Please install \n'+\
'matplotlib package and try again'
return
config = Config(args.config_file)
#--------------------------------------
# figure out path to input data files
if config.cellularity_estimation == 'schism':
mutCellularityPath = os.path.join(config.working_dir,\
config.output_prefix + '.mutation.cellularity')
else:
mutCellularityPath = os.path.join(config.working_dir,\
config.mutation_cellularity_input)
clustCellularityPath = os.path.join(config.working_dir,\
config.output_prefix + '.cluster.cellularity')
clusterPath = os.path.join(config.working_dir, \
config.mutation_to_cluster_assignment)
cluster2mut = read_cluster_assignments(clusterPath)
#--------------------------------------
# read in cluster cellularity data
cl2index, samples = read_input_samples(clustCellularityPath)
clBounds = {}
samples = sorted(samples, key=lambda x: x.name)
x = range(len(samples))
# generate serial information each cluster
for cl in cl2index.keys():
lowerBound = ['NA'] * len(samples)
upperBound = ['NA'] * len(samples)
for id, sample in enumerate(samples):
if sample.mutCellularity[cl2index[cl]] != -1:
lowerBound[id] = sample.mutCellularity[cl2index[cl]] - \
sample.mutSigma[cl2index[cl]]
upperBound[id] = sample.mutCellularity[cl2index[cl]] + \
sample.mutSigma[cl2index[cl]]
else:
pass
clBounds[cl] = [x, lowerBound, upperBound]
# visualize cluster cellularity estimates and standard error
# as ribbon plot
#colorPalette = ["#E62E41", "#0A893D", "#455593", "#F19131", \
# "#F375A0", "#874B2C", "#C894F1", "#C3401E", "#91430F", "#F65348"]
# modified color palette
colorPalette = ['#323f7b', '#cb3245', '#638e4d', '#9a336d', '#e2a86a',\
'#246c8f', '#7d303d', '#734d85', '#077783', '#9c7688',\
'#b48b73', '#7da1bf', '#4b6b6c', '#7b7282', '#263246']
if len(cl2index) > len(colorPalette):
print >> sys.stderr, 'Cellularity plot not supported for more than 15 clusters.'
sys.exit()
plt.grid(True)
fig, ax = plt.subplots(1, 1)
for cl, index in cl2index.items():
trace = np.mean(np.array([clBounds[cl][1], clBounds[cl][2]]), 0)
trace = map(bound_fix, trace)
clBounds[cl][1] = map(bound_fix, clBounds[cl][1])
clBounds[cl][2] = map(bound_fix, clBounds[cl][2])
ax.fill_between(clBounds[cl][0], \
clBounds[cl][1], \
clBounds[cl][2], color = colorPalette[index], \
alpha = 0.5)
ax.plot(clBounds[cl][0], list(trace), \
color = colorPalette[index], label = cl, \
linestyle="dotted", marker="o", linewidth = 1)
#-----------------------------------------#
# visualize cluster cellularity for individual mutations as
# scatter plot overlay
mut2index, samples = read_input_samples(mutCellularityPath)
samples = sorted(samples, key=lambda x: x.name)
for cl in cluster2mut:
coords = []
for mut in cluster2mut[cl]:
xm = range(len(samples))
ym = [sample.mutCellularity[mut2index[mut]] \
for sample in samples]
coords.extend(zip(xm, ym))
coords = filter(lambda x: x[1] != -1, coords)
ax.scatter(zip(*coords)[0], zip(*coords)[1], color = colorPalette[cl2index[cl]],\
marker = '+', s = 20.0)
#-----------------------------------------#
# adjust visual properties of the plot
handles, labels = ax.get_legend_handles_labels()
handles, labels = zip(*sorted(zip(handles, labels), key=lambda x: x[1]))
ax.legend(handles, labels)
plt.xticks(range(len(samples)))
plt.yticks([x * 0.1 for x in range(11)])
plt.xlim((-0.5, 2.5))
plt.ylim((-0.1, 1.1))
xtext = [sample.name for sample in samples]
ax.set_xticklabels(xtext)
plt.gca().yaxis.grid(True)
outputPath = os.path.join(config.working_dir, \
config.output_prefix +'.cellularity.png')
plt.savefig(outputPath)
return
def plot_mut_clust_cellularity(args):
# plot estimated cellularity for mutations and clusters
# across samples
try:
import matplotlib
matplotlib.use('Agg')
import matplotlib.pyplot as plt
except:
print >>sys.stderr, \
'Plotting mutation/cluster cellularities requires matplotlib \n' +\
'module. Failed to import matplotlib. Please install \n'+\
'matplotlib package and try again'
return
config = Config(args.config_file)
#--------------------------------------
# figure out path to input data files
if config.cellularity_estimation == 'schism':
mutCellularityPath = os.path.join(config.working_dir,\
config.output_prefix + '.mutation.cellularity')
else:
mutCellularityPath = os.path.join(config.working_dir,\
config.mutation_cellularity_input)
clustCellularityPath = os.path.join(config.working_dir,\
config.output_prefix + '.cluster.cellularity')
clusterPath = os.path.join(config.working_dir, \
config.mutation_to_cluster_assignment)
cluster2mut = read_cluster_assignments(clusterPath)
#--------------------------------------
# read in cluster cellularity data
cl2index, samples = read_input_samples(clustCellularityPath)
clBounds = {}
samples = sorted(samples, key = lambda x: x.name)
x = range(len(samples))
# generate serial information each cluster
for cl in cl2index.keys():
lowerBound = ['NA'] * len(samples)
upperBound = ['NA'] * len(samples)
for id, sample in enumerate(samples):
if sample.mutCellularity[cl2index[cl]] != -1:
lowerBound[id] = sample.mutCellularity[cl2index[cl]] - \
sample.mutSigma[cl2index[cl]]
upperBound[id] = sample.mutCellularity[cl2index[cl]] + \
sample.mutSigma[cl2index[cl]]
else:
pass
clBounds[cl] = [x, lowerBound, upperBound]
# visualize cluster cellularity estimates and standard error
# as ribbon plot
colorPalette = ["#E62E41", "#0A893D", "#455593", "#F19131", \
"#F375A0", "#874B2C", "#C894F1", "#C3401E", "#91430F", "#F65348"]
if len(cl2index) > len(colorPalette):
print >>sys.stderr, 'Cellularity plot not supported for more than 8 clusters.'
sys.exit()
plt.grid(True)
fig, ax = plt.subplots(1,1)
for cl, index in cl2index.items():
trace = np.mean(np.array([clBounds[cl][1], clBounds[cl][2]]), 0)
trace = map(bound_fix, trace)
clBounds[cl][1] = map(bound_fix, clBounds[cl][1])
clBounds[cl][2] = map(bound_fix, clBounds[cl][2])
ax.fill_between(clBounds[cl][0], \
clBounds[cl][1], \
clBounds[cl][2], color = colorPalette[index], \
alpha = 0.5)
ax.plot(clBounds[cl][0], list(trace), \
color = colorPalette[index], label = cl, \
linestyle="dotted", marker="o", linewidth = 1)
#-----------------------------------------#
# visualize cluster cellularity for individual mutations as
# scatter plot overlay
mut2index, samples = read_input_samples(mutCellularityPath)
samples = sorted(samples, key = lambda x: x.name)
for cl in cluster2mut:
coords = []
for mut in cluster2mut[cl]:
xm = range(len(samples))
ym = [sample.mutCellularity[mut2index[mut]] \
for sample in samples]
coords.extend(zip(xm,ym))
coords = filter(lambda x: x[1] != -1, coords)
ax.scatter(zip(*coords)[0], zip(*coords)[1], color = colorPalette[cl2index[cl]],\
marker = '+', s = 20.0)
#-----------------------------------------#
# adjust visual properties of the plot
handles, labels = ax.get_legend_handles_labels()
handles, labels = zip(*sorted(zip(handles, labels), key = lambda x: x[1]))
ax.legend(handles, labels)
plt.xticks(range(len(samples)))
plt.yticks([x * 0.1 for x in range(11)])
plt.xlim((-0.5,2.5))
plt.ylim((-0.1,1.1))
xtext = [sample.name for sample in samples]
ax.set_xticklabels(xtext)
plt.gca().yaxis.grid(True)
outputPath = os.path.join(config.working_dir, \
config.output_prefix +'.cellularity.png')
plt.savefig(outputPath)
return