def run_strip_plot(panc_df, gtex_df, panc_labels, gtex_labels):
assert panc_df.columns.equals(gtex_df.columns)
psi_df = pd.concat((panc_df, gtex_df), axis=0)
assert psi_df.shape[0] == panc_df.shape[0] + gtex_df.shape[0]
assert psi_df.columns.unique().size == psi_df.shape[1]
assert panc_df.index.equals(panc_labels.index)
assert gtex_df.index.equals(gtex_labels.index)
event_list = psi_df.columns.tolist()
psi_df_aug = psi_df.copy()
psi_df_aug['cnc'] = None
psi_df_aug.loc[panc_df.index, ['cnc']] = panc_labels.loc[panc_df.index]
psi_df_aug.loc[gtex_df.index, ['cnc']] = gtex_labels.loc[gtex_df.index]
unq_panc_labels = sorted(panc_labels.unique().tolist())
unq_gtex_labels = sorted(gtex_labels.unique().tolist())
assert np.intersect1d(unq_panc_labels, unq_gtex_labels).size == 0
plt.close()
label_list = unq_panc_labels + unq_gtex_labels
color_lut = _get_stripplot_color_lut(unq_panc_labels, unq_gtex_labels)
for event in event_list:
outpath = os.path.join(PLOT_DIR, 'stripplots', '%s_stripplot.png'%event)
if not os.path.exists(os.path.dirname(outpath)): os.makedirs(os.path.dirname(outpath))
fig, ax = plt.subplots(figsize=(20,3))
sns.stripplot(x='cnc', y=event, data=psi_df_aug,
palette=color_lut, s=3,
order=label_list,
jitter=True, ax=ax)
ax.axvline(len(unq_panc_labels) - .5, color='black', ls='--')
ax.set_xticklabels(ax.get_xticklabels(), rotation = 90)
ax.set_ylim(-.05, 1.05)
ax.title.set_text('Gene: %s Event type: %s Event ID: %d'%(_decode_event_name(event)))
ax.set_ylabel('psi')
ax.set_xlabel('')
axs.clean_axis(ax)
print "Writing: %s" %outpath
plt.savefig(outpath, bbox_inches='tight')
pdf_outpath = re.sub('.png$', '.pdf', outpath)
print "Writing: %s" %pdf_outpath
plt.savefig(pdf_outpath, bbox_inches='tight')
plt.close()
return
def plot_full(ax, vals, ylabel, title):
### detected / detected conf / detected high-conf
cax = ax.bar([2, 5, 8],
vals[1],
color=['b', 'g', 'm'],
alpha=0.5,
width=0.8,
linewidth=0)
axs.label_bars(ax, cax, vals[1], rotation=90)
### annotated fraction of the counts above
cax = ax.bar([3, 6, 9],
vals[2],
color=['b', 'g', 'm'],
alpha=0.9,
width=0.8,
linewidth=0)
axs.label_bars(ax, cax, vals[2], rotation=90)
### high-conf novel events
cax = ax.bar([11], vals[3], color=['c'], alpha=0.9, width=0.8, linewidth=0)
axs.label_bars(ax, cax, vals[3], rotation=90)
### annotated
cax = ax.bar([0], vals[0], color='r', alpha=0.5, width=0.8, linewidth=0)
axs.label_bars(ax, cax, vals[0], rotation=90)
ax.set_ylabel(ylabel)
ax.set_xlabel('')
ax.set_xticks(sp.array([0, 2, 3, 5, 6, 8, 9, 11], dtype='float') + 0.4)
ax.set_xticklabels([
'annotated', 'detected', 'detected\n(anno)', 'detected\n(conf)',
'detected\n(conf+anno)', 'detected\n(h_conf)',
'detected\n(h_conf+anno)', 'novel\n(h_conf)'
],
rotation=90)
ax.set_xlim([-0.2, ax.get_xlim()[1]])
ax.set_title(title)
axs.set_ticks_outer(ax)
axs.clean_axis(ax)
def main():
figs = dict()
gss = dict()
gss['stats'] = gridspec.GridSpec(3, 2) #, wspace=0.0, hspace=0.0)
gss['stats_log'] = gridspec.GridSpec(3, 2) #, wspace=0.0, hspace=0.0)
for e, event_type in enumerate(event_types):
print >> sys.stderr, 'Handling %s' % event_type
### load annotation index
is_anno_gtex = cPickle.load(
open(
os.path.join(
paths.basedir_as_gtex,
'merge_graphs_%s_C%i.anno_only.pickle' %
(event_type, CONF)), 'r'))
is_anno_icgc = cPickle.load(
open(
os.path.join(
paths.basedir_as, 'merge_graphs_%s_C%i.anno_only.pickle' %
(event_type, CONF)), 'r'))
### load confident events
IN = h5py.File(
os.path.join(
paths.basedir_as,
'merge_graphs_%s_C%i.counts.hdf5' % (event_type, CONF)), 'r')
idx_conf_tcga = sp.zeros((is_anno_icgc.shape[0], ), dtype='bool')
idx_conf_tcga[IN['conf_idx'][:]] = 1
IN.close()
### for mutex events load cluster idx
if event_type == 'mutex_exons':
mutex_clusters = cPickle.load(
open(
os.path.join(
paths.basedir_as,
'merge_graphs_%s_C%i.cluster_idx.pickle' %
(event_type, CONF)), 'r'))
### load psi filtered events
idx_psi_icgc = cPickle.load(
open(
os.path.join(
paths.basedir_as,
'merge_graphs_%s_C%i.counts.psi_filt_per_ct.pickle' %
(event_type, CONF)), 'r'))[1]
### get all cancer types
ctypes = sp.unique([x[1] for x in idx_psi_icgc.keys()])
colors = sp.array(tc.get_color_scheme('study', labels=ctypes))
for dp in [0.0, 0.1, 0.3, 0.5]:
for mr in [5, 20, 50]:
### get counts
if event_type == 'mutex_exons':
counts_anno = []
counts_new = []
for ct in ctypes:
if not (mr, ct, dp) in idx_psi_icgc:
counts_anno.append(0)
counts_new.append(0)
continue
tmp_idx = sp.zeros((is_anno_icgc.shape[0], ),
dtype='bool')
tmp_idx[idx_psi_icgc[(mr, ct, dp)]] = 1
counts_anno.append(
sp.unique(
mutex_clusters[idx_conf_tcga & is_anno_icgc
& tmp_idx]).shape[0])
counts_new.append(
sp.unique(
mutex_clusters[idx_conf_tcga & ~is_anno_icgc
& tmp_idx]).shape[0])
counts_anno = sp.array(counts_anno)
counts_new = sp.array(counts_new)
else:
counts_anno = sp.array([
sp.sum((idx_conf_tcga
& is_anno_icgc)[idx_psi_icgc[(mr, ct, dp)]]) if
(mr, ct, dp) in idx_psi_icgc else 0 for ct in ctypes
])
counts_new = sp.array([
sp.sum(idx_conf_tcga[idx_psi_icgc[(mr, ct, dp)]]) -
sp.sum((idx_conf_tcga
& is_anno_icgc)[idx_psi_icgc[(mr, ct, dp)]]) if
(mr, ct, dp) in idx_psi_icgc else 0 for ct in ctypes
])
### plot stats for events by cancer type
if e == 0:
figs['stats_mr%i_dp%.1f' %
(mr, dp)] = plt.figure(figsize=(15, 9))
ax = figs['stats_mr%i_dp%.1f' % (mr, dp)].add_subplot(
gss['stats'][e / 2, e % 2])
ax.bar(sp.arange(ctypes.shape[0]) + 0.2,
counts_anno,
0.6,
color=colors,
linewidth=0.5)
ax.bar(sp.arange(ctypes.shape[0]) + 0.2,
counts_new,
0.6,
bottom=counts_anno,
color=colors,
linewidth=0.5,
alpha=0.5)
axs.set_ticks_outer(ax)
axs.clean_axis(ax)
ax.set_xticks(sp.arange(ctypes.shape[0]) + 0.15)
ax.set_xlim([-1, ctypes.shape[0]])
ax.set_xticklabels(ctypes, rotation=90, fontsize=10)
ax.set_title(event_dict[event_type])
ax.yaxis.grid(True)
continue ### skip log for now
### plot stats for events by histotype (log)
if e == 0:
figs['stats_log_mr%i_dp%.1f' %
(mr, dp)] = plt.figure(figsize=(15, 9))
ax = figs['stats_log_mr%i_dp%.1f' % (mr, dp)].add_subplot(
gss['stats'][e / 2, e % 2])
ax.bar(sp.arange(ctypes.shape[0]) + 0.1,
sp.log10(counts_anno + 1),
0.8,
color=colors,
linewidth=0.5)
ax.bar(sp.arange(ctypes.shape[0]) + 0.1,
sp.log10(counts_new + 1),
0.8,
bottom=sp.log10(counts_anno + 1),
color=colors,
linewidth=0.5,
alpha=0.3)
axs.set_ticks_outer(ax)
axs.clean_axis(ax)
ax.set_xticks(range(ctypes.shape[0]))
ax.set_xlim([-1, ctypes.shape[0]])
ax.set_xticklabels(ctypes, rotation=90, fontsize=10)
ax.set_title(event_dict[event_type])
ax.yaxis.grid(True)
for p in figs:
print 'plotting %s' % p
figs[p].tight_layout()
figs[p].savefig(os.path.join(
PLOTDIR, 'event_overview_per_ct_C%i_%s.pdf' % (CONF, p)),
format='pdf',
bbox_inches='tight')
figs[p].savefig(os.path.join(
PLOTDIR, 'event_overview_per_ct_C%i_%s.png' % (CONF, p)),
format='png',
bbox_inches='tight')
plt.close(figs[p])
def main():
figs = dict()
figs['stats'] = plt.figure(figsize=(15, 8))
figs['stats_rel'] = plt.figure(figsize=(15, 8))
figs['stats_rel_sampsize'] = plt.figure(figsize=(15, 8))
gss = dict()
gss['stats'] = gridspec.GridSpec(2, 3) #, wspace=0.0, hspace=0.0)
gss['stats_rel'] = gridspec.GridSpec(2, 3) #, wspace=0.0, hspace=0.0)
gss['stats_rel_sampsize'] = gridspec.GridSpec(
2, 3) #, wspace=0.0, hspace=0.0)
for e, event_type in enumerate(event_types):
print('Handling %s' % event_type, file=sys.stderr)
### is exonization
is_exonization = pickle.load(
open(
os.path.join(
BASEDIR_ICGC,
'merge_graphs_%s_C%i.exonize_candidates_step2.pickle' %
(event_type, CONF)), 'r'))
### load confident events
IN = h5py.File(
os.path.join(
BASEDIR_ICGC,
'merge_graphs_%s_C%i.counts.hdf5' % (event_type, CONF)), 'r')
idx_conf_icgc = IN['conf_idx'][:]
[tumor_dict, histo_dict] = tm.translate([('analysis_id', 'is_tumour'),
('analysis_id', 'histotype')])
htypes_all = sp.array([
histo_dict[x.split('.')[0]]
if x.split('.')[0] in histo_dict else 'NA'
for x in IN['strains'][:]
])
htypes_all_u, htypes_all_cnt = sp.unique(htypes_all,
return_counts=True)
htypes_all_med = sp.median(htypes_all_cnt).astype('float')
htypes_sf = htypes_all_med / htypes_all_cnt
IN.close()
htypes_sf = dict([(htypes_all_u[_], htypes_sf[_])
for _ in range(htypes_all_u.shape[0])])
### load psi filtered events
idx_psi_icgc = pickle.load(
open(
os.path.join(
BASEDIR_ICGC,
'merge_graphs_%s_C%i.counts.hdf5.psi_filt_per_ht_normalized.pickle'
% (event_type, CONF)), 'r'))[1]
### get all histotypes
htypes = sp.unique([x[0] for x in list(idx_psi_icgc.keys())])
#colors = dict(zip(htypes, ic.get_color_scheme('tumor_subtype', labels=htypes)))
colors = ic.get_color_scheme('tumor_subtype', labels=htypes)
### get counts
counts = []
dp = 0.3
counts_anno = sp.array([
sp.sum(is_exonization[idx_psi_icgc[(ht, dp)]]) if
(ht, dp) in idx_psi_icgc else 0 for ht in htypes
])
counts_all = sp.array([
idx_psi_icgc[(ht, dp)].shape[0] if (ht, dp) in idx_psi_icgc else 0
for ht in htypes
])
counts_sf = sp.array([htypes_sf[ht] for ht in htypes], dtype='float')
### plot stats for events by histotype
ax = figs['stats'].add_subplot(gss['stats'][e / 3, e % 3])
ax.bar(sp.arange(htypes.shape[0]) + 0.2,
counts_anno,
0.6,
color=colors,
linewidth=0.5)
#li, = ax.plot(sp.arange(data1_icgc.shape[0]), data1_icgc, '-r', label='ICGC')
axs.set_ticks_outer(ax)
axs.clean_axis(ax)
#if e == len(event_types) - 1:
# ax.legend(handles=[lg, lga, li, lia], loc='upper right', frameon=False, fontsize=10)
ax.set_xticks(sp.arange(htypes.shape[0]) + 0.5)
ax.set_xlim([-0.2, htypes.shape[0]])
if e < len(event_types) - 3:
ax.set_xticklabels([])
else:
ax.set_xticklabels(htypes, rotation=90, fontsize=10)
ax.set_title(event_dict[event_type])
ax.yaxis.grid(True)
### plot stats for events by histotype - relative to events detected
ax = figs['stats_rel'].add_subplot(gss['stats_rel'][e / 3, e % 3])
ax.bar(sp.arange(htypes.shape[0]) + 0.2,
counts_anno / counts_all.astype('float'),
0.6,
color=colors,
linewidth=0.5)
axs.set_ticks_outer(ax)
axs.clean_axis(ax)
ax.set_xticks(sp.arange(htypes.shape[0]) + 0.5)
ax.set_xlim([-0.2, htypes.shape[0]])
if e < len(event_types) - 3:
ax.set_xticklabels([])
else:
ax.set_xticklabels(htypes, rotation=90, fontsize=10)
ax.set_title(event_dict[event_type])
ax.yaxis.grid(True)
### plot stats for events by histotype - relative to sample size
ax = figs['stats_rel_sampsize'].add_subplot(
gss['stats_rel_sampsize'][e / 3, e % 3])
ax.bar(sp.arange(htypes.shape[0]) + 0.2,
counts_anno * counts_sf,
0.6,
color=colors,
linewidth=0.5)
axs.set_ticks_outer(ax)
axs.clean_axis(ax)
ax.set_xticks(sp.arange(htypes.shape[0]) + 0.5)
ax.set_xlim([-0.2, htypes.shape[0]])
if e < len(event_types) - 3:
ax.set_xticklabels([])
else:
ax.set_xticklabels(htypes, rotation=90, fontsize=10)
ax.set_title(event_dict[event_type])
ax.yaxis.grid(True)
for p in figs:
figs[p].tight_layout()
figs[p].savefig(os.path.join(
PLOTDIR,
'event_overview_per_ht_exonize_level2_C%i_%s.pdf' % (CONF, p)),
format='pdf',
bbox_inches='tight')
figs[p].savefig(os.path.join(
PLOTDIR,
'event_overview_per_ht_exonize_level2_C%i_%s.png' % (CONF, p)),
format='png',
bbox_inches='tight')
plt.close(figs[p])
def main():
figs = dict()
figs['stats'] = plt.figure(figsize=(15, 8))
figs['stats_log'] = plt.figure(figsize=(15, 8))
gss = dict()
gss['stats'] = gridspec.GridSpec(2, 3) #, wspace=0.0, hspace=0.0)
gss['stats_log'] = gridspec.GridSpec(2, 3) #, wspace=0.0, hspace=0.0)
for e, event_type in enumerate(event_types):
print('Handling %s' % event_type, file=sys.stderr)
### load events detected in annotation only
#anno = cPickle.load(open(os.path.join(BASEDIR_ANNO, 'merge_graphs_%s_C%i.pickle' % (event_type, CONF)), 'r'))
#if isinstance(anno, tuple):
# anno = anno[0]
### load annotation index
is_anno_gtex = pickle.load(
open(
os.path.join(
BASEDIR_GTEX, 'merge_graphs_%s_C%i.anno_only.pickle' %
(event_type, CONF)), 'r'))
is_anno_icgc = pickle.load(
open(
os.path.join(
BASEDIR_ICGC, 'merge_graphs_%s_C%i.anno_only.pickle' %
(event_type, CONF)), 'r'))
### load confident events
IN = h5py.File(
os.path.join(
BASEDIR_GTEX,
'merge_graphs_%s_C%i.counts.hdf5' % (event_type, CONF)), 'r')
idx_conf_gtex = IN['conf_idx'][:]
IN.close()
IN = h5py.File(
os.path.join(
BASEDIR_ICGC,
'merge_graphs_%s_C%i.counts.hdf5' % (event_type, CONF)), 'r')
idx_conf_icgc = IN['conf_idx'][:]
IN.close()
### load psi filtered events
#idx_psi_gtex = cPickle.load(open(os.path.join(BASEDIR_GTEX, 'merge_graphs_%s_C%i.counts.hdf5.psi_filt_per_ht.pickle' % (event_type, CONF)), 'r'))[1]
idx_psi_icgc = pickle.load(
open(
os.path.join(
BASEDIR_ICGC,
'merge_graphs_%s_C%i.counts.hdf5.psi_filt_per_ht.pickle' %
(event_type, CONF)), 'r'))[1]
### get all histotypes
htypes = sp.unique([x[0] for x in list(idx_psi_icgc.keys())])
#colors = dict(zip(htypes, ic.get_color_scheme('tumor_subtype', labels=htypes)))
colors = ic.get_color_scheme('tumor_subtype', labels=htypes)
### get counts
counts = []
dp = 0.3
counts_anno = sp.array([
sp.sum(is_anno_icgc[idx_psi_icgc[(ht, dp)]]) if
(ht, dp) in idx_psi_icgc else 0 for ht in htypes
])
counts_new = sp.array([
idx_psi_icgc[(ht, dp)].shape[0] -
sp.sum(is_anno_icgc[idx_psi_icgc[(ht, dp)]]) if
(ht, dp) in idx_psi_icgc else 0 for ht in htypes
])
### plot stats for events by histotype
ax = figs['stats'].add_subplot(gss['stats'][e / 3, e % 3])
ax.bar(sp.arange(htypes.shape[0]) + 0.2,
counts_anno,
0.6,
color=colors,
linewidth=0.5)
ax.bar(sp.arange(htypes.shape[0]) + 0.2,
counts_new,
0.6,
bottom=counts_anno,
color=colors,
linewidth=0.5,
alpha=0.5)
#li, = ax.plot(sp.arange(data1_icgc.shape[0]), data1_icgc, '-r', label='ICGC')
axs.set_ticks_outer(ax)
axs.clean_axis(ax)
#if e == len(event_types) - 1:
# ax.legend(handles=[lg, lga, li, lia], loc='upper right', frameon=False, fontsize=10)
ax.set_xticks(sp.arange(htypes.shape[0]) + 0.5)
ax.set_xlim([-0.2, htypes.shape[0]])
if e < len(event_types) - 3:
ax.set_xticklabels([])
else:
ax.set_xticklabels(htypes, rotation=90, fontsize=10)
ax.set_title(event_dict[event_type])
ax.yaxis.grid(True)
### plot stats for events by histotype (log)
ax = figs['stats_log'].add_subplot(gss['stats'][e / 3, e % 3])
ax.bar(sp.arange(htypes.shape[0]) + 0.1,
sp.log10(counts_anno + 1),
0.8,
color=colors,
linewidth=0.5)
ax.bar(sp.arange(htypes.shape[0]) + 0.1,
sp.log10(counts_new + 1),
0.8,
bottom=sp.log10(counts_anno + 1),
color=colors,
linewidth=0.5,
alpha=0.3)
axs.set_ticks_outer(ax)
axs.clean_axis(ax)
ax.set_xticks(list(range(htypes.shape[0])))
ax.set_xlim([-0.2, htypes.shape[0]])
if e < len(event_types) - 3:
ax.set_xticklabels([])
else:
ax.set_xticklabels(htypes, rotation=90, fontsize=10)
ax.set_title(event_dict[event_type])
ax.yaxis.grid(True)
for p in figs:
figs[p].tight_layout()
figs[p].savefig(os.path.join(
PLOTDIR, 'event_overview_per_ht_C%i_%s.pdf' % (CONF, p)),
format='pdf',
bbox_inches='tight')
figs[p].savefig(os.path.join(
PLOTDIR, 'event_overview_per_ht_C%i_%s.png' % (CONF, p)),
format='png',
bbox_inches='tight')
plt.close(figs[p])
intron_filter=intron_filter,
return_legend_handle=True,
label='Ref: %s' % ref_donors[d]))
labels.append('Ref: %s' % ref_donors[d])
plt.legend(cax, labels, fontsize=10)
#for vp in var_pos:
# ax.plot(vp, 0, 'bo', markersize=1)
#ylim = ax.get_ylim()
#xlim = ax.get_xlim()
#xspan = xlim[1] - xlim[0]
#yspan = ylim[1] - ylim[0]
#for vp in var_pos:
# ax.arrow(vp, yspan * 0.2, 0, yspan * -0.15, head_width=0.01*xspan, head_length=0.01*yspan, fc='b')
x_range = ax.get_xlim()
ax.set_ylim([0, ax.get_ylim()[1]])
axs.clean_axis(ax)
axs.set_ticks_outer(ax)
### plot event layout
ax = fig.add_subplot(gs[rows - 1, 0])
gl.multiple([exons1, exons2], x_range=x_range, ax=ax)
axs.clean_axis(ax, left=True, bottom=True)
plt.tight_layout()
outbase = os.path.join(
PLOTDIR, 'result_exonization_exon_skip_%s_%s_%s' %
(results[i, 1], results[i, 0], un.get_ID(results[i, 9],
lookup=lookup)))
print('saving to %s' % outbase)
plt.savefig('%s%s.pdf' % (outbase, logtag),
def main():
figs = dict()
figs['stats'] = plt.figure(figsize=(12, 8))
figs['stats_log'] = plt.figure(figsize=(12, 8))
figs['stats_full'] = plt.figure(figsize=(12, 8))
figs['stats_full_log'] = plt.figure(figsize=(12, 8))
gss = dict()
gss['stats'] = gridspec.GridSpec(2, 3) #, wspace=0.0, hspace=0.0)
gss['stats_log'] = gridspec.GridSpec(2, 3) #, wspace=0.0, hspace=0.0)
gss['stats_full'] = gridspec.GridSpec(2, 3) #, wspace=0.0, hspace=0.0)
gss['stats_full_log'] = gridspec.GridSpec(2, 3) #, wspace=0.0, hspace=0.0)
for e, event_type in enumerate(event_types):
print('Handling %s' % event_type, file=sys.stderr)
### load events detected in annotation only
anno = pickle.load(open(os.path.join(BASEDIR_ANNO, 'merge_graphs_%s_C%i.pickle' % (event_type, CONF)), 'r'))
if isinstance(anno, tuple):
anno = anno[0]
### load annotation index
is_anno_gtex = pickle.load(open(os.path.join(BASEDIR_GTEX, 'merge_graphs_%s_C%i.anno_only.pickle' % (event_type, CONF)), 'r'))
is_anno_icgc_t = pickle.load(open(os.path.join(BASEDIR_ICGC_T, 'merge_graphs_%s_C%i.anno_only.pickle' % (event_type, CONF)), 'r'))
is_anno_icgc_n = pickle.load(open(os.path.join(BASEDIR_ICGC_N, 'merge_graphs_%s_C%i.anno_only.pickle' % (event_type, CONF)), 'r'))
### load confident events
IN = h5py.File(os.path.join(BASEDIR_GTEX, 'merge_graphs_%s_C%i.counts.hdf5' % (event_type, CONF)), 'r')
idx_conf_gtex = IN['conf_idx'][:]
IN.close()
IN = h5py.File(os.path.join(BASEDIR_ICGC_T, 'merge_graphs_%s_C%i.counts.hdf5' % (event_type, CONF)), 'r')
idx_conf_icgc_t = IN['conf_idx'][:]
IN.close()
IN = h5py.File(os.path.join(BASEDIR_ICGC_N, 'merge_graphs_%s_C%i.counts.hdf5' % (event_type, CONF)), 'r')
idx_conf_icgc_n = IN['conf_idx'][:]
IN.close()
### load filtered events
#IN = h5py.File(os.path.join(BASEDIR_GTEX, 'merge_graphs_%s_C%i.counts.r10_s50_V10.hdf5' % (event_type, CONF)), 'r')
#idx_filt_gtex = IN['filter_idx'][:]
#IN.close()
#IN = h5py.File(os.path.join(BASEDIR_ICGC, 'merge_graphs_%s_C%i.counts.r10_s50_V10.hdf5' % (event_type, CONF)), 'r')
#idx_filt_icgc = IN['filter_idx'][:]
#IN.close()
### load psi filtered events
idx_psi_gtex = pickle.load(open(os.path.join(BASEDIR_GTEX, 'merge_graphs_%s_C%i.counts.hdf5.psi_filt.pickle' % (event_type, CONF)), 'r'))[1]
idx_psi_icgc_t = pickle.load(open(os.path.join(BASEDIR_ICGC_T, 'merge_graphs_%s_C%i.counts.hdf5.psi_filt.pickle' % (event_type, CONF)), 'r'))[1]
idx_psi_icgc_n = pickle.load(open(os.path.join(BASEDIR_ICGC_N, 'merge_graphs_%s_C%i.counts.hdf5.psi_filt.pickle' % (event_type, CONF)), 'r'))[1]
### plot stats for normal counts (FULL)
ax = figs['stats_full'].add_subplot(gss['stats_full'][e / 3, e % 3])
xlabels_full = ['detected', 'confident']
xlabels_part = ['confident']
xlabels_full.extend(['dpsi > %.1f' % _ for _ in sorted(idx_psi_gtex.keys())])
xlabels_part.extend(['dpsi > %.1f' % _ for _ in sorted(idx_psi_gtex.keys())])
# all confirmed events, further filtered by PSI - GTEX
data1_gtex = [is_anno_gtex.shape[0], idx_conf_gtex.shape[0]]
data1_gtex.extend([sp.intersect1d(idx_conf_gtex, idx_psi_gtex[_]).shape[0] for _ in sorted(idx_psi_gtex.keys())])
data1_gtex = sp.array(data1_gtex)
lg, = ax.plot(sp.arange(data1_gtex.shape[0]), data1_gtex, '-b', label='GTEx')
# all annotated confirmed events, further filtered by PSI - GTEX
data2_gtex = [sp.sum(is_anno_gtex), sp.sum(is_anno_gtex[idx_conf_gtex])]
data2_gtex.extend([sp.sum(is_anno_gtex[sp.intersect1d(idx_conf_gtex, idx_psi_gtex[_])]) for _ in sorted(idx_psi_gtex.keys())])
data2_gtex = sp.array(data2_gtex)
lga, = ax.plot(sp.arange(data2_gtex.shape[0]), data2_gtex, '--b', label='GTEx (anno)')
# all confirmed events, further filtered by PSI - ICGC_T
data1_icgc_t = [is_anno_icgc_t.shape[0], idx_conf_icgc_t.shape[0]]
data1_icgc_t.extend([sp.intersect1d(idx_conf_icgc_t, idx_psi_icgc_t[_]).shape[0] for _ in sorted(idx_psi_icgc_t.keys())])
data1_icgc_t = sp.array(data1_icgc_t)
lit, = ax.plot(sp.arange(data1_icgc_t.shape[0]), data1_icgc_t, '-r', label='ICGC Tumor')
# all annotated confirmed events, further filtered by PSI - ICGC
data2_icgc_t = [sp.sum(is_anno_icgc_t), sp.sum(is_anno_icgc_t[idx_conf_icgc_t])]
data2_icgc_t.extend([sp.sum(is_anno_icgc_t[sp.intersect1d(idx_conf_icgc_t, idx_psi_icgc_t[_])]) for _ in sorted(idx_psi_icgc_t.keys())])
data2_icgc_t = sp.array(data2_icgc_t)
lita, = ax.plot(sp.arange(data2_icgc_t.shape[0]), data2_icgc_t, '--r', label='ICGC Tumor (anno)')
# all confirmed events, further filtered by PSI - ICGC_T
data1_icgc_n = [is_anno_icgc_n.shape[0], idx_conf_icgc_n.shape[0]]
data1_icgc_n.extend([sp.intersect1d(idx_conf_icgc_n, idx_psi_icgc_n[_]).shape[0] for _ in sorted(idx_psi_icgc_n.keys())])
data1_icgc_n = sp.array(data1_icgc_n)
lin, = ax.plot(sp.arange(data1_icgc_n.shape[0]), data1_icgc_n, '-g', label='ICGC Normal')
# all annotated confirmed events, further filtered by PSI - ICGC
data2_icgc_n = [sp.sum(is_anno_icgc_n), sp.sum(is_anno_icgc_n[idx_conf_icgc_n])]
data2_icgc_n.extend([sp.sum(is_anno_icgc_n[sp.intersect1d(idx_conf_icgc_n, idx_psi_icgc_n[_])]) for _ in sorted(idx_psi_icgc_n.keys())])
data2_icgc_n = sp.array(data2_icgc_n)
lina, = ax.plot(sp.arange(data2_icgc_n.shape[0]), data2_icgc_n, '--g', label='ICGC Normal (anno)')
axs.set_ticks_outer(ax)
axs.clean_axis(ax)
if e == len(event_types) - 1:
ax.legend(handles=[lit, lita, lin, lina, lg, lga], loc='upper right', frameon=False, fontsize=10)
ax.set_xticks(list(range(len(xlabels_full))))
if e < len(event_types) - 3:
ax.set_xticklabels([])
else:
ax.set_xticklabels(xlabels_full, rotation=90, fontsize=10)
ax.set_title(event_dict[event_type])
ax.xaxis.grid(True)
### plots stats for log10 counts (FULL)
ax = figs['stats_full_log'].add_subplot(gss['stats_full_log'][e / 3, e % 3])
lg, = ax.plot(sp.arange(data1_gtex.shape[0]), sp.log10(data1_gtex + 1), '-b', label='GTEx')
lga, = ax.plot(sp.arange(data2_gtex.shape[0]), sp.log10(data2_gtex + 1), '--b', label='GTEx (anno)')
lit, = ax.plot(sp.arange(data1_icgc_t.shape[0]), sp.log10(data1_icgc_t + 1), '-r', label='ICGC Tumor')
lita, = ax.plot(sp.arange(data2_icgc_t.shape[0]), sp.log10(data2_icgc_t + 1), '--r', label='ICGC Tumor (anno)')
lin, = ax.plot(sp.arange(data1_icgc_n.shape[0]), sp.log10(data1_icgc_n + 1), '-g', label='ICGC Normal')
lina, = ax.plot(sp.arange(data2_icgc_n.shape[0]), sp.log10(data2_icgc_n + 1), '--g', label='ICGC Normal (anno)')
axs.set_ticks_outer(ax)
axs.clean_axis(ax)
if e == len(event_types) - 1:
ax.legend(handles=[lit, lita, lin, lina, lg, lga], loc='lower left', frameon=False, fontsize=10)
ax.set_xticks(list(range(len(xlabels_full))))
if e < len(event_types) - 3:
ax.set_xticklabels([])
else:
ax.set_xticklabels(xlabels_full, rotation=90, fontsize=10)
ax.set_title(event_dict[event_type])
ax.xaxis.grid(True)
### plot stats for normal counts (only conf)
ax = figs['stats'].add_subplot(gss['stats'][e / 3, e % 3])
lg, = ax.plot(sp.arange(data1_gtex.shape[0] - 1), data1_gtex[1:], '-b', label='GTEx')
lga, = ax.plot(sp.arange(data2_gtex.shape[0] - 1), data2_gtex[1:], '--b', label='GTEx (anno)')
lit, = ax.plot(sp.arange(data1_icgc_t.shape[0] - 1), data1_icgc_t[1:], '-r', label='ICGC Tumor')
lita, = ax.plot(sp.arange(data2_icgc_t.shape[0] - 1), data2_icgc_t[1:], '--r', label='ICGC Tumor (anno)')
lin, = ax.plot(sp.arange(data1_icgc_n.shape[0] - 1), data1_icgc_n[1:], '-g', label='ICGC Normal')
lina, = ax.plot(sp.arange(data2_icgc_n.shape[0] - 1), data2_icgc_n[1:], '--g', label='ICGC Normal (anno)')
axs.set_ticks_outer(ax)
axs.clean_axis(ax)
if e == len(event_types) - 1:
ax.legend(handles=[lit, lita, lin, lina, lg, lga], loc='upper right', frameon=False, fontsize=10)
ax.set_xticks(list(range(len(xlabels_part))))
if e < len(event_types) - 3:
ax.set_xticklabels([])
else:
ax.set_xticklabels(xlabels_part, rotation=90, fontsize=10)
ax.set_title(event_dict[event_type])
ax.xaxis.grid(True)
### plots stats for log10 counts (only cony)
ax = figs['stats_log'].add_subplot(gss['stats_log'][e / 3, e % 3])
lg, = ax.plot(sp.arange(data1_gtex.shape[0] - 1), sp.log10(data1_gtex[1:] + 1), '-b', label='GTEx')
lga, = ax.plot(sp.arange(data2_gtex.shape[0] - 1), sp.log10(data2_gtex[1:] + 1), '--b', label='GTEx (anno)')
lit, = ax.plot(sp.arange(data1_icgc_t.shape[0] - 1), sp.log10(data1_icgc_t[1:] + 1), '-r', label='ICGC Tumor')
lita, = ax.plot(sp.arange(data2_icgc_t.shape[0] - 1), sp.log10(data2_icgc_t[1:] + 1), '--r', label='ICGC Tumor (anno)')
lin, = ax.plot(sp.arange(data1_icgc_n.shape[0] - 1), sp.log10(data1_icgc_n[1:] + 1), '-g', label='ICGC Normal')
lina, = ax.plot(sp.arange(data2_icgc_n.shape[0] - 1), sp.log10(data2_icgc_n[1:] + 1), '--g', label='ICGC Normal (anno)')
axs.set_ticks_outer(ax)
axs.clean_axis(ax)
if e == len(event_types) - 1:
ax.legend(handles=[lit, lita, lin, lina, lg, lga], loc='lower left', frameon=False, fontsize=10)
ax.set_xticks(list(range(len(xlabels_part))))
if e < len(event_types) - 3:
ax.set_xticklabels([])
else:
ax.set_xticklabels(xlabels_part, rotation=90, fontsize=10)
ax.set_title(event_dict[event_type])
ax.xaxis.grid(True)
for p in figs:
figs[p].tight_layout()
figs[p].savefig(os.path.join(PLOTDIR, 'event_overview_cumm_Liver_C%i_%s.pdf' % (CONF, p)), format='pdf', bbox_inches='tight')
figs[p].savefig(os.path.join(PLOTDIR, 'event_overview_cumm_Liver_C%i_%s.png' % (CONF, p)), format='png', bbox_inches='tight')
plt.close(figs[p])
