-
Notifications
You must be signed in to change notification settings - Fork 10
/
plot_contingency_figure.py
1060 lines (770 loc) · 41.4 KB
/
plot_contingency_figure.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
import pylab
import matplotlib as mpl
import matplotlib.colors as colors
import matplotlib.cm as cmx
import matplotlib.pyplot as plt
import matplotlib.gridspec as gridspec
import numpy
import parse_file
from numpy.random import shuffle
from math import ceil, exp,fabs
from scipy.special import digamma
from scipy.interpolate import interp1d
import sys
from stats_utils import *
import mutation_spectrum_utils
import matplotlib.colors as colors
import matplotlib.cm as cmx
import CBcm
import figure_utils
import stats_utils
debug=False
if len(sys.argv) > 1:
level = sys.argv[1]
else:
level = 'gene'
populations = parse_file.complete_nonmutator_lines
FDR=0.05
##############################################################################
#
# Set up figures
#
##############################################################################
mpl.rcParams['font.size'] = 5.0
mpl.rcParams['lines.linewidth'] = 1.0
mpl.rcParams['legend.frameon'] = False
mpl.rcParams['legend.fontsize'] = 'small'
mpl.rcParams['axes.labelpad'] = 2
mpl.rcParams['pdf.fonttype'] = 42
mpl.rcParams['font.sans-serif'] = 'Arial'
mpl.rcParams['font.serif'] = 'Times New Roman'
mpl.rcParams['mathtext.rm'] = 'serif'
mpl.rcParams['mathtext.it'] = 'serif:italic'
mpl.rcParams['mathtext.bf'] = 'serif:bold'
mpl.rcParams['mathtext.fontset'] = 'custom'
####################
#
# Fig 6: signatures of historical contingency
#
####################
#fig6 = plt.figure(figsize=(7.2, 4.5))
#fig6 = plt.figure(figsize=(7.08, 4.2))
fig6 = plt.figure(figsize=(8.5,2.7))
outer_grid = gridspec.GridSpec(2, 1, height_ratios=[1.4,1], hspace=0.1)
matrix_grid = gridspec.GridSpecFromSubplotSpec(1, 6,
width_ratios=[0.02,1,1,1,0.05,0.05],
wspace=0.25,
subplot_spec=outer_grid[1])
#####
#
# Gene time distribution panel
#
#####
time_axis = plt.Subplot(fig6, outer_grid[0])
fig6.add_subplot(time_axis)
time_axis.set_ylabel('Appearance time',labelpad=3)
time_axis.get_xaxis().tick_top()
time_axis.get_yaxis().tick_left()
time_axis.get_yaxis().set_tick_params(direction='out',length=3,pad=1)
time_axis.get_xaxis().set_tick_params(direction='out',length=3,pad=1)
time_axis.set_ylim([-2000,62000])
time_axis.set_yticks(figure_utils.time_xticks)
time_axis.set_yticklabels(figure_utils.time_xticklabels)
time_vmin = 0
time_vmax = 1
time_cmap = CBcm.CB2cm['redblue']
time_cNorm = colors.Normalize(vmin=time_vmin, vmax=time_vmax)
time_scalarMap = cmx.ScalarMappable(norm=time_cNorm, cmap=time_cmap)
######
#
# Dispersion matrix panels
#
######
all_matrix_axis = plt.Subplot(fig6, matrix_grid[1])
fig6.add_subplot(all_matrix_axis)
all_matrix_axis.set_yticks([0.5,1.5,2.5,3.5,4.5])
all_matrix_axis.set_yticklabels(['6+','5','4','3','2'])
all_matrix_axis.get_yaxis().set_tick_params(pad=4)
all_matrix_axis.set_xticks([0.5,1.5,2.5,3.5,4.5,5.5])
all_matrix_axis.set_xticklabels(['1','2','3','4','5','6'])
all_matrix_axis.set_xlabel('# populations with mutation',labelpad=3)
all_matrix_axis.set_ylabel('# mutations in gene',labelpad=3)
all_matrix_axis.spines['top'].set_visible(False)
all_matrix_axis.spines['right'].set_visible(False)
all_matrix_axis.get_xaxis().tick_bottom()
all_matrix_axis.get_yaxis().tick_left()
all_matrix_axis.tick_params(axis='both', which='both',length=0)
all_matrix_axis.set_xlim([0,6.05])
all_matrix_axis.set_ylim([0,5.05])
early_matrix_axis = plt.Subplot(fig6, matrix_grid[2])
fig6.add_subplot(early_matrix_axis)
early_matrix_axis.set_yticks([0.5,1.5,2.5,3.5,4.5])
early_matrix_axis.set_yticklabels([])
early_matrix_axis.set_xticks([0.5,1.5,2.5,3.5,4.5,5.5])
early_matrix_axis.set_xticklabels(['1','2','3','4','5','6'])
early_matrix_axis.set_xlabel('# populations with mutation')
early_matrix_axis.spines['top'].set_visible(False)
early_matrix_axis.spines['right'].set_visible(False)
early_matrix_axis.get_xaxis().tick_bottom()
early_matrix_axis.get_yaxis().tick_left()
early_matrix_axis.tick_params(axis='both', which='both',length=0)
early_matrix_axis.set_xlim([0,6.05])
early_matrix_axis.set_ylim([0,5.05])
late_matrix_axis = plt.Subplot(fig6, matrix_grid[3])
fig6.add_subplot(late_matrix_axis)
late_matrix_axis.set_yticks([0.5,1.5,2.5,3.5,4.5])
late_matrix_axis.set_yticklabels([])
late_matrix_axis.set_xticks([0.5,1.5,2.5,3.5,4.5,5.5])
late_matrix_axis.set_xticklabels(['1','2','3','4','5','6'])
late_matrix_axis.set_xlabel('# populations with mutation')
late_matrix_axis.spines['top'].set_visible(False)
late_matrix_axis.spines['right'].set_visible(False)
late_matrix_axis.get_xaxis().tick_bottom()
late_matrix_axis.get_yaxis().tick_left()
late_matrix_axis.tick_params(axis='both', which='both',length=0)
late_matrix_axis.set_xlim([0,6.05])
late_matrix_axis.set_ylim([0,5.05])
cax = plt.Subplot(fig6, matrix_grid[4])
fig6.add_subplot(cax)
matrix_vmin = -20
matrix_vmax = 20
#cmap = CBcm.CB2cm['redblue']
matrix_cmap = pylab.get_cmap('RdBu_r')
matrix_cNorm = colors.Normalize(vmin=matrix_vmin, vmax=matrix_vmax)
matrix_scalarMap = cmx.ScalarMappable(norm=matrix_cNorm, cmap=matrix_cmap)
####################################
#
# Supplemental Fig: Pooled distribution of appearance times
#
####################################
pooled_fig = plt.figure(figsize=(3, 1.7))
pooled_grid = gridspec.GridSpec(1, 1)
pooled_time_axis = plt.Subplot(pooled_fig, pooled_grid[0])
pooled_fig.add_subplot(pooled_time_axis)
pooled_time_axis.spines['top'].set_visible(False)
pooled_time_axis.spines['right'].set_visible(False)
pooled_time_axis.get_xaxis().tick_bottom()
pooled_time_axis.get_yaxis().tick_left()
pooled_time_axis.set_ylabel('Fraction mutations $\geq t$',fontsize=6)
pooled_time_axis.set_xlabel('Appearance time, $t$',fontsize=6)
pooled_time_axis.set_xticks(figure_utils.time_xticks)
pooled_time_axis.set_xticklabels(figure_utils.time_xticklabels)
pooled_time_axis.set_xlim([0,60000])
####################################
#
# Supplemental Fig: Temporal LRT as function of time
#
####################################
#####
#
# Mutation spectrum panel
#
#####
LRT_fig = plt.figure(figsize=(8.5, 0.9))
LRT_grid = gridspec.GridSpec(1, 2, width_ratios=[5.6,1.4], wspace=0.175)
probability_axis = plt.Subplot(LRT_fig, LRT_grid[0])
LRT_fig.add_subplot(probability_axis)
probability_axis.set_ylabel('Relative weight (%)',labelpad=3)
probability_axis.spines['top'].set_visible(False)
probability_axis.spines['right'].set_visible(False)
probability_axis.get_xaxis().tick_bottom()
probability_axis.get_yaxis().tick_left()
probability_axis.get_yaxis().set_tick_params(direction='out',length=2,pad=1)
probability_axis.get_xaxis().set_tick_params(direction='out',length=2,pad=1)
LRT_axis = plt.Subplot(LRT_fig, LRT_grid[1])
LRT_fig.add_subplot(LRT_axis)
LRT_axis.spines['top'].set_visible(False)
LRT_axis.spines['right'].set_visible(False)
LRT_axis.get_xaxis().tick_bottom()
LRT_axis.get_yaxis().tick_left()
LRT_axis.set_ylabel('LRT statistic, $\\Delta \ell$',fontsize=6)
LRT_axis.set_xlabel('Partition time, $t^*$',fontsize=6)
LRT_axis.set_xticks(figure_utils.time_xticks)
LRT_axis.set_xticklabels(figure_utils.time_xticklabels)
LRT_axis.set_xlim([0,55000])
LRT_axis.get_yaxis().set_tick_params(direction='out',length=2,pad=1)
LRT_axis.get_xaxis().set_tick_params(direction='out',length=2,pad=1)
####################################
#
# Supplemental Fig: 2-hit stuff
#
####################################
twohit_fig = plt.figure(figsize=(3, 1.7))
twohit_grid = gridspec.GridSpec(1, 1)
twohit_axis = plt.Subplot(twohit_fig, twohit_grid[0])
twohit_fig.add_subplot(twohit_axis)
twohit_axis.spines['top'].set_visible(False)
twohit_axis.spines['right'].set_visible(False)
twohit_axis.get_xaxis().tick_bottom()
twohit_axis.get_yaxis().tick_left()
twohit_axis.set_ylabel('Fraction of genes $\geq \Delta t$',fontsize=6)
twohit_axis.set_xlabel('Time difference, $\Delta t$', fontsize=6)
twohit_axis.set_xticks(figure_utils.time_xticks)
twohit_axis.set_xticklabels(figure_utils.time_xticklabels)
twohit_axis.set_xlim([0,60000])
twohit_axis.set_ylim([0,1])
####################################
#
# Supplemental Fig: Net missed opportunities as function of time
#
####################################
missed_opportunity_fig = plt.figure(figsize=(3, 1.7))
missed_opportunity_grid = gridspec.GridSpec(1, 1)
early_color = '#a50f15'
late_color = '#045a8d'
missed_opportunity_axis = plt.Subplot(missed_opportunity_fig, missed_opportunity_grid[0])
missed_opportunity_fig.add_subplot(missed_opportunity_axis)
missed_opportunity_axis.spines['top'].set_visible(False)
missed_opportunity_axis.spines['right'].set_visible(False)
missed_opportunity_axis.get_xaxis().tick_bottom()
missed_opportunity_axis.get_yaxis().tick_left()
missed_opportunity_axis.set_ylabel('Net missed opportunities',fontsize=6)
missed_opportunity_axis.set_xlabel('Partition time, $t^*$',fontsize=6)
missed_opportunity_axis.set_xticks(figure_utils.time_xticks)
missed_opportunity_axis.set_xticklabels(figure_utils.time_xticklabels)
missed_opportunity_axis.set_xlim([0,55000])
#missed_opportunity_axis.set_ylim([-20,35])
#######################################
#
# Now do calculations and plot figures
#
#######################################
tstars = numpy.arange(0,111)*500
# Load convergence matrix
convergence_matrix = parse_file.parse_convergence_matrix(parse_file.data_directory+('%s_convergence_matrix.txt' % level))
# Load significant genes
parallel_genes = parse_file.parse_parallel_genes(parse_file.data_directory+('parallel_%ss.txt' % level))
# Calculate gene parallelism statistics
gene_parallelism_statistics = mutation_spectrum_utils.calculate_parallelism_statistics(convergence_matrix,populations)
# Calculate gene name, pop, and time vectors
# All genes
all_gene_names = []
all_pops = []
all_times = []
for gene_name in convergence_matrix.keys():
for population in populations:
for t,L,Lclade,f in convergence_matrix[gene_name]['mutations'][population]:
all_gene_names.append(gene_name)
all_pops.append(population)
all_times.append(t)
# Multi-hit genes (ni>=2)
multihit_gene_names = []
multihit_pops = []
multihit_times = []
for gene_name in convergence_matrix.keys():
if gene_parallelism_statistics[gene_name]['observed']>1.5:
for population in populations:
for t,L,Lclade,f in convergence_matrix[gene_name]['mutations'][population]:
multihit_gene_names.append(gene_name)
multihit_pops.append(population)
multihit_times.append(t)
# One-hit genes (ni=1)
onehit_gene_names = []
onehit_pops = []
onehit_times = []
for gene_name in convergence_matrix.keys():
if fabs(gene_parallelism_statistics[gene_name]['observed']-1) < 0.5:
for population in populations:
for t,L,Lclade,f in convergence_matrix[gene_name]['mutations'][population]:
onehit_gene_names.append(gene_name)
onehit_pops.append(population)
onehit_times.append(t)
# Two-hit genes (ni=2)
twohit_gene_names = []
twohit_pops = []
twohit_times = []
for gene_name in convergence_matrix.keys():
if fabs(gene_parallelism_statistics[gene_name]['observed']-2) < 0.5:
for population in populations:
for t,L,Lclade,f in convergence_matrix[gene_name]['mutations'][population]:
twohit_gene_names.append(gene_name)
twohit_pops.append(population)
twohit_times.append(t)
# Significantly parallel genes
parallel_gene_names = []
parallel_pops = []
parallel_times = []
# Calculate gene name, pop, and time vectors
for gene_name in parallel_genes:
for population in populations:
for t,L,Lclade,f in convergence_matrix[gene_name]['mutations'][population]:
parallel_gene_names.append(gene_name)
parallel_pops.append(population)
parallel_times.append(t)
parallel_gene_names = numpy.array(parallel_gene_names)
parallel_pops = numpy.array(parallel_pops)
parallel_times = numpy.array(parallel_times)
# Calculate total number of mutations and total number of significant mutations
ntot = len(all_gene_names)
nsig = len(parallel_gene_names)
# Helper function for converting vectors into time distribution
def calculate_time_distribution(gene_names, times):
time_distribution = {}
for gene_name, t in zip(gene_names,times):
if gene_name not in time_distribution:
time_distribution[gene_name] = []
time_distribution[gene_name].append(t)
for gene_name in time_distribution.keys():
time_distribution[gene_name] = numpy.array(time_distribution[gene_name])
time_distribution[gene_name].sort()
return time_distribution
# Helper function for converting vectors into population distribution
def calculate_population_distribution(gene_names, pops):
pop_distribution = {}
for gene_name, pop in zip(gene_names,pops):
if gene_name not in pop_distribution:
pop_distribution[gene_name] = {population: 0 for population in populations}
pop_distribution[gene_name][pop] += 1
for gene_name in pop_distribution.keys():
pop_distribution[gene_name] = numpy.array([pop_distribution[gene_name][pop] for pop in populations])
return pop_distribution
# Helper function for converting vectors into distributions
def calculate_early_late_time_distributions(gene_names, times, tstar=-1):
time_distribution = {}
for gene_name, t in zip(gene_names,times):
if gene_name not in time_distribution:
time_distribution[gene_name] = {'early': [], 'late' : []}
if t<=tstar:
time_distribution[gene_name]['early'].append(t)
else:
time_distribution[gene_name]['late'].append(t)
return time_distribution
# Calculate pooled time distribution for all genes
pooled_all_time_distribution = numpy.array(all_times, copy=True)
pooled_all_time_distribution.sort()
# Calculate pooled time distribution for multihit genes
pooled_multihit_time_distribution = numpy.array(multihit_times, copy=True)
pooled_multihit_time_distribution.sort()
# Calculate pooled time distribution for 1-hit genes
pooled_onehit_time_distribution = numpy.array(onehit_times, copy=True)
pooled_onehit_time_distribution.sort()
# Calculate pooled time distribution for 2-hit genes
pooled_twohit_time_distribution = numpy.array(twohit_times, copy=True)
pooled_twohit_time_distribution.sort()
# Calculate gene-specific time distribution for 2-hit genes
twohit_time_distribution = calculate_time_distribution(twohit_gene_names, twohit_times)
# Calculate pooled time distribution for significantly parallel genes
pooled_parallel_time_distribution = numpy.array(parallel_times, copy=True)
pooled_parallel_time_distribution.sort()
# Calculate gene-specific time distribution for significantly parallel genes
parallel_time_distribution = calculate_time_distribution(parallel_gene_names, parallel_times)
# Calculate KS distance between gene-specific distributions and pooled subset
parallel_kss = numpy.array([stats_utils.calculate_ks_distance(parallel_time_distribution[gene_name], pooled_parallel_time_distribution) for gene_name in parallel_genes])
# Re-sort parallel genes by descending (num_hits, ks_statistic)
parallel_genes, parallel_kss = (numpy.array(x) for x in zip(*sorted(zip(parallel_genes, parallel_kss), key=lambda pair: (gene_parallelism_statistics[pair[0]]['observed'],pair[1]), reverse=True)))
# Calculate median times of gene-specific time distributions
parallel_median_times = numpy.array([numpy.median(parallel_time_distribution[gene_name]) for gene_name in parallel_genes])
######################
#
# Plot pooled time distributions
#
######################
# use same color map as for distribution of appearance times by var-type
colors = [figure_utils.get_var_type_color(var_type) for var_type in parse_file.var_types]
all_ts, all_survivals = calculate_unnormalized_survival_from_vector(pooled_all_time_distribution, min_x=0, max_x=60000, min_p=1e-10)
pooled_time_axis.step(all_ts, all_survivals*1.0/all_survivals[0],'-',color='k',label='All genes')
onehit_ts, onehit_survivals = calculate_unnormalized_survival_from_vector(pooled_onehit_time_distribution, min_x=0, max_x=60000, min_p=1e-10)
pooled_time_axis.step(onehit_ts, onehit_survivals*1.0/onehit_survivals[0],'-',color=colors[0],label='1-hit genes', alpha=0.5)
twohit_ts, twohit_survivals = calculate_unnormalized_survival_from_vector(pooled_twohit_time_distribution, min_x=0, max_x=60000, min_p=1e-10)
pooled_time_axis.step(twohit_ts, twohit_survivals*1.0/twohit_survivals[0],'-',color=colors[1],label='2-hit genes', alpha=0.5)
multihit_ts, multihit_survivals = calculate_unnormalized_survival_from_vector(pooled_multihit_time_distribution, min_x=0, max_x=60000, min_p=1e-10)
pooled_time_axis.step(multihit_ts, multihit_survivals*1.0/multihit_survivals[0],'-',color=colors[2],label='Multi-hit genes', alpha=0.5)
parallel_ts, parallel_survivals = calculate_unnormalized_survival_from_vector(pooled_parallel_time_distribution, min_x=0, max_x=60000, min_p=1e-10)
interpolated_time_CDF = interp1d(parallel_ts, 1.0-parallel_survivals/parallel_survivals[0],kind='linear',bounds_error=True)
pooled_time_axis.step(parallel_ts, parallel_survivals*1.0/parallel_survivals[0],'-',color=colors[3],label='Significant genes',alpha=0.5)
pooled_time_axis.legend(loc='upper right', frameon=False)
###############
#
# Two-hit gene calculation
#
###############
# Calculate time difference between earliest and latest genes in 2-hit genes
observed_twohit_differences = numpy.array([twohit_time_distribution[gene_name][-1]-twohit_time_distribution[gene_name][0] for gene_name in twohit_gene_names])
observed_twohit_differences.sort()
sys.stderr.write("Bootstrap resampling 2-hit genes...\t")
if debug==True:
num_bootstraps=10
else:
num_bootstraps=10000
bootstrapped_twohit_differences = []
for bootstrap_idx in xrange(0,num_bootstraps):
bootstrapped_twohit_gene_names = numpy.array([gene_name for gene_name in twohit_gene_names])
shuffle(bootstrapped_twohit_gene_names)
bootstrapped_twohit_time_distribution = calculate_time_distribution(bootstrapped_twohit_gene_names, twohit_times)
bootstrapped_twohit_differences.append( numpy.array([bootstrapped_twohit_time_distribution[gene_name][-1]-bootstrapped_twohit_time_distribution[gene_name][0] for gene_name in bootstrapped_twohit_gene_names]) )
bootstrapped_twohit_differences = numpy.array(bootstrapped_twohit_differences)
null_twohit_differences = bootstrapped_twohit_differences.flatten()
# Calculate p-value for mean of distribution
observed_mean_twohit_difference = observed_twohit_differences.mean()
bootstrapped_mean_twohit_differences = bootstrapped_twohit_differences.mean(axis=1)
twohit_difference_pvalue = stats_utils.calculate_empirical_pvalue(-observed_mean_twohit_difference, -bootstrapped_mean_twohit_differences)
sys.stderr.write("Done!\n")
sys.stdout.write("Observed two-hit dt = %g, expected = %g, pvalue = %g (%d bootstraps)\n" % (observed_mean_twohit_difference, bootstrapped_mean_twohit_differences.mean(), twohit_difference_pvalue, num_bootstraps))
###############
#
# Plot two-hit gene calculation
#
###############
null_dts, null_survivals = calculate_unnormalized_survival_from_vector(null_twohit_differences, min_x=0, max_x=60000, min_p=1e-10)
observed_dts, observed_survivals = calculate_unnormalized_survival_from_vector(observed_twohit_differences, min_x=0, max_x=60000, min_p=1e-10)
twohit_axis.step(observed_dts, observed_survivals*1.0/observed_survivals[0],'-',color=parse_file.nonmutator_group_color,label='Observed')
twohit_axis.step(null_dts, null_survivals*1.0/null_survivals[0],'-',color='0.7',linewidth=0.5,label='Expected')
twohit_axis.legend(loc='upper right',frameon=False)
################
#
# Parallel gene calculation
#
################
observed_LRTs = []
for tstar in tstars:
time_distribution = calculate_early_late_time_distributions(parallel_gene_names, parallel_times, tstar)
early_ns = numpy.array([len(time_distribution[gene_name]['early'])*1.0 for gene_name in parallel_genes])
late_ns = numpy.array([len(time_distribution[gene_name]['late'])*1.0 for gene_name in parallel_genes])
observed_LRTs.append( mutation_spectrum_utils.calculate_LRT_statistic(early_ns, late_ns) )
observed_LRTs = numpy.array(observed_LRTs)
sys.stderr.write("Bootstrap resampling parallel genes...\n")
if debug==True:
num_bootstraps=10
else:
num_bootstraps=10000
bootstrapped_LRTs = []
bootstrapped_kss = []
for bootstrap_idx in xrange(1,num_bootstraps+1):
if bootstrap_idx%1000==0:
sys.stderr.write("%dk\n" % (bootstrap_idx/1000))
bootstrapped_gene_names = numpy.array([gene_name for gene_name in parallel_gene_names])
shuffle(bootstrapped_gene_names)
bootstrapped_time_distribution = calculate_time_distribution(bootstrapped_gene_names, parallel_times)
bootstrapped_kss.append( [stats_utils.calculate_ks_distance(bootstrapped_time_distribution[gene_name], pooled_parallel_time_distribution) for gene_name in parallel_genes] )
LRTs = []
for tstar in tstars:
bootstrapped_time_distribution = calculate_early_late_time_distributions(bootstrapped_gene_names, parallel_times, tstar)
early_ns = numpy.array([len(bootstrapped_time_distribution[gene_name]['early'])*1.0 for gene_name in parallel_genes])
late_ns = numpy.array([len(bootstrapped_time_distribution[gene_name]['late'])*1.0 for gene_name in parallel_genes])
LRTs.append( mutation_spectrum_utils.calculate_LRT_statistic(early_ns, late_ns) )
bootstrapped_LRTs.append(LRTs)
bootstrapped_kss = numpy.array(bootstrapped_kss) # bootstrap x genes matrix
bootstrapped_LRTs = numpy.array(bootstrapped_LRTs) # bootstrap x tstars matrix
sys.stderr.write("Done!\n")
# Calculate ks pvalues
parallel_ks_pvalues = numpy.array([stats_utils.calculate_empirical_pvalue(parallel_kss[i],bootstrapped_kss[:,i]) for i in xrange(0,len(parallel_genes))])
# Calculate ks qvalues
parallel_ks_qvalues = stats_utils.calculate_qvalues(parallel_ks_pvalues)
# Calculate set of individually significant genes
sys.stdout.write("Temporal nonuniformity of significantly parallel genes (%d bootstraps):\n" % num_bootstraps)
for gene_name, qvalue in zip(parallel_genes, parallel_ks_qvalues):
sys.stdout.write("%s: q=%g\n" % (gene_name, qvalue))
individually_significant_genes = set(parallel_genes[parallel_ks_qvalues<FDR])
# And remaining genes
nonsignificant_genes = parallel_genes[parallel_ks_qvalues>=FDR]
# Recalculate time distribution for nonsignificant genes
nonsignificant_time_distribution = {gene_name: parallel_time_distribution[gene_name] for gene_name in nonsignificant_genes}
nonsignificant_gene_names = []
nonsignificant_times = []
for gene_name in nonsignificant_genes:
nonsignificant_gene_names.extend( [gene_name]*len(nonsignificant_time_distribution[gene_name]) )
nonsignificant_times.extend( nonsignificant_time_distribution[gene_name] )
pooled_nonsignificant_time_distribution = numpy.array(nonsignificant_times, copy=True)
pooled_nonsignificant_time_distribution.sort()
nonsignificant_kss = numpy.array([stats_utils.calculate_ks_distance( nonsignificant_time_distribution[gene_name], pooled_nonsignificant_time_distribution) for gene_name in nonsignificant_genes])
sys.stderr.write("Bootstrap resampling non-individually significant genes...\n")
bootstrapped_nonsignificant_kss = []
for bootstrap_idx in xrange(1,num_bootstraps+1):
if bootstrap_idx%1000==0:
sys.stderr.write("%dk\n" % (bootstrap_idx/1000))
bootstrapped_gene_names = numpy.array([gene_name for gene_name in nonsignificant_gene_names])
shuffle(bootstrapped_gene_names)
bootstrapped_time_distribution = calculate_time_distribution(bootstrapped_gene_names, nonsignificant_times)
bootstrapped_nonsignificant_kss.append( [stats_utils.calculate_ks_distance(bootstrapped_time_distribution[gene_name], pooled_nonsignificant_time_distribution) for gene_name in nonsignificant_genes] )
bootstrapped_nonsignificant_kss = numpy.array(bootstrapped_nonsignificant_kss) # bootstrap x genes matrix
sys.stderr.write("Done!\n")
# Calculate pvalue for global ks sum
observed_total_ks = nonsignificant_kss.sum()
bootstrapped_total_kss = bootstrapped_nonsignificant_kss.sum(axis=1)
total_ks_pvalue = stats_utils.calculate_empirical_pvalue(observed_total_ks, bootstrapped_total_kss)
sys.stdout.write("Individually significant_genes: %s\n" % (", ".join(individually_significant_genes)))
sys.stdout.write("Remaining total KS distance = %g, expected = %g (+/- %g), pvalue = %g\n" % (observed_total_ks, bootstrapped_total_kss.mean(), bootstrapped_total_kss.std(), total_ks_pvalue))
######################
#
# Plot early-late LRT as function of time
#
######################
probability_tstar = numpy.median(parallel_times)
upper_null_LRTs = []
for i in xrange(0,len(tstars)):
LRTs = numpy.array(bootstrapped_LRTs[:,i],copy=True)
LRTs.sort()
upper_null_LRTs.append( LRTs[long(len(LRTs)*0.95)] )
upper_null_LRTs = numpy.array(upper_null_LRTs)
LRT_axis.fill_between(tstars, numpy.zeros_like(tstars), upper_null_LRTs,color='0.7')
LRT_axis.plot(tstars,upper_null_LRTs,'-',linewidth=0.25, color='0.6')
line, = LRT_axis.plot([probability_tstar, probability_tstar],[0,observed_LRTs.max()*1.1],'k:',linewidth=0.25)
line.set_dashes((1,1))
LRT_axis.plot(tstars,observed_LRTs,'-',color=parse_file.nonmutator_group_color)
LRT_axis.set_ylim([0,observed_LRTs.max()*1.1])
observed_max_LRT = observed_LRTs.max()
bootstrapped_max_LRTs = bootstrapped_LRTs.max(axis=1)
sys.stdout.write("Max LRT at %d: Observed = %g, Expected = %g +/- %g, p=%g\n" % (tstars[observed_LRTs.argmax()], observed_max_LRT, bootstrapped_max_LRTs.mean(), bootstrapped_max_LRTs.std(), stats_utils.calculate_empirical_pvalue(observed_max_LRT, bootstrapped_max_LRTs)))
######################
#
# Plot gene-specific time distribution
#
######################
positions = []
current_position = 0
previous_block_position = 0
current_num_hits = gene_parallelism_statistics[parallel_genes[0]]['observed']
grey = True
for i in xrange(0,len(parallel_genes)):
if gene_parallelism_statistics[parallel_genes[i]]['observed'] != current_num_hits:
# reached the end of a block
current_position+=1.25
if grey:
time_axis.fill_between( [previous_block_position-0.5,current_position-0.5],[-2000,-2000],[62000,62000],facecolor='0.85',linewidth=0.0)
grey= not grey
previous_block_position = current_position
current_num_hits = gene_parallelism_statistics[parallel_genes[i]]['observed']
current_position-=0.75
current_position += 1
positions.append(current_position)
gene_labels = []
for i in xrange(0,len(parallel_genes)):
colorVal = time_scalarMap.to_rgba(interpolated_time_CDF(parallel_median_times[i]))
time_axis.plot([positions[i]]*len(parallel_time_distribution[parallel_genes[i]]), parallel_time_distribution[parallel_genes[i]], 'o-',color=colorVal,markersize=1.5,linewidth=0.25,markeredgewidth=0)
time_axis.plot([positions[i]],[parallel_median_times[i]],'k_',alpha=0.5,markersize=2)
if parallel_genes[i] in individually_significant_genes:
significance_string = "*"
else:
significance_string = ""
gene_labels.append('%s%s (%d)' % (significance_string, parallel_genes[i], len(parallel_time_distribution[parallel_genes[i]])))
time_axis.set_xticks(positions)
time_axis.set_xticklabels(gene_labels, rotation='vertical',fontsize=5)
time_axis.set_xlim([positions[0]-1,positions[-1]+1])
probability_axis.set_xticks(positions)
probability_axis.set_xticklabels(gene_labels, rotation='vertical',fontsize=4)
probability_axis.set_xlim([positions[0]-1,positions[-1]+1])
#######
#
# Plot mutation spectrum for significant genes
#
#######
tstar = numpy.median(parallel_times)
probability_tstar = tstar
early_late_time_distribution = calculate_early_late_time_distributions(parallel_gene_names, parallel_times, tstar)
all_ns = numpy.array([len(parallel_time_distribution[gene_name]) for gene_name in parallel_genes])
early_ns = numpy.array([len(early_late_time_distribution[gene_name]['early'])*1.0 for gene_name in parallel_genes])
late_ns = numpy.array([len(early_late_time_distribution[gene_name]['late'])*1.0 for gene_name in parallel_genes])
all_ps = all_ns*1.0/all_ns.sum()*nsig*1.0/ntot
early_ps = early_ns*1.0/early_ns.sum()*nsig*1.0/ntot
late_ps = late_ns*1.0/late_ns.sum()*nsig*1.0/ntot
probability_axis.plot(positions, all_ps*100, 'k.-',label='All',linewidth=0.5,markersize=3.0)
probability_axis.plot(positions, early_ps*100, '.-', color=time_scalarMap.to_rgba(interpolated_time_CDF(10000)), label='Early',alpha=0.5,linewidth=0.5,markersize=3.0)
probability_axis.plot(positions, late_ps*100, '.-', color=time_scalarMap.to_rgba(interpolated_time_CDF(50000)),label='Late',alpha=0.5,linewidth=0.5,markersize=3.0)
sys.stdout.write("Median time of significant genes: %g\n" % tstar)
probability_axis.legend(loc='upper right', frameon=False, ncol=3,fontsize=5)
#######
#
# Missed opportunity calculation
#
#######
observed_time_distribution = calculate_time_distribution(all_gene_names, all_times)
observed_population_distribution = calculate_population_distribution(all_gene_names, all_pops)
desired_genes = sorted(observed_time_distribution.keys())
observed_population_matrix = numpy.array([observed_population_distribution[gene_name] for gene_name in desired_genes])
observed_median_times = numpy.array([numpy.median(observed_time_distribution[gene_name]) for gene_name in desired_genes])
observed_missed_opportunities = mutation_spectrum_utils.calculate_scaled_missed_opportunities_from_matrix(observed_population_matrix)
sys.stdout.write("Missed opportunities for >= 4-hit genes:")
for i in xrange(0,len(desired_genes)):
if observed_population_matrix[i,:].sum() > 3.5:
sys.stdout.write("%s n=%g m=%g\n" % (desired_genes[i], observed_population_matrix[i,:].sum(), observed_missed_opportunities[i]))
sys.stderr.write("Calculating excess missed opportunities as function of tstar...\t")
tstars = numpy.arange(0,111)*500
early_dms = []
late_dms = []
bootstrapped_early_dms = []
bootstrapped_late_dms = []
if debug==True:
num_bootstraps=10
else:
num_bootstraps=10000
for tstar in tstars:
# First look at genes with median time <= tstar
early_population_matrix = observed_population_matrix[observed_median_times<=tstar,:]
if early_population_matrix.shape[0]>0:
early_total_m = mutation_spectrum_utils.calculate_scaled_missed_opportunities_from_matrix(early_population_matrix).sum()
bootstrapped_early_total_ms = []
for bootstrap_idx in xrange(1,num_bootstraps+1):
bootstrapped_early_population_matrix = mutation_spectrum_utils.resample_population_matrix(early_population_matrix)
bootstrapped_early_total_ms.append( mutation_spectrum_utils.calculate_scaled_missed_opportunities_from_matrix(bootstrapped_early_population_matrix).sum() )
bootstrapped_early_total_ms = numpy.array(bootstrapped_early_total_ms)
expected_early_total_m = bootstrapped_early_total_ms.mean()
early_dms.append( early_total_m - expected_early_total_m )
bootstrapped_early_dms.append( bootstrapped_early_total_ms - expected_early_total_m )
else:
early_dms.append( 0 )
bootstrapped_early_dms.append( numpy.zeros(num_bootstraps)*1.0 )
# Now look at genes with median time > tstar
late_population_matrix = observed_population_matrix[observed_median_times>tstar,:]
if late_population_matrix.shape[0]>0:
late_total_m = mutation_spectrum_utils.calculate_scaled_missed_opportunities_from_matrix(late_population_matrix).sum()
bootstrapped_late_total_ms = []
for bootstrap_idx in xrange(1,num_bootstraps+1):
bootstrapped_late_population_matrix = mutation_spectrum_utils.resample_population_matrix(late_population_matrix)
bootstrapped_late_total_ms.append( mutation_spectrum_utils.calculate_scaled_missed_opportunities_from_matrix(bootstrapped_late_population_matrix).sum() )
bootstrapped_late_total_ms = numpy.array(bootstrapped_late_total_ms)
expected_late_total_m = bootstrapped_late_total_ms.mean()
late_dms.append( late_total_m - expected_late_total_m )
bootstrapped_late_dms.append( bootstrapped_late_total_ms - expected_late_total_m )
else:
late_dms.append(0)
bootstrapped_late_dms.append( numpy.zeros(num_bootstraps)*1.0 )
early_dms = numpy.array(early_dms)
late_dms = numpy.array(late_dms)
bootstrapped_early_dms = numpy.transpose(numpy.array(bootstrapped_early_dms))
bootstrapped_late_dms = numpy.transpose(numpy.array(bootstrapped_late_dms))
sys.stderr.write("Done!\n")
# Calculate lower 95% CI on the (negative) early excess multiplicity
lower_early_dms = []
for i in xrange(0,bootstrapped_early_dms.shape[1]):
null_dms = numpy.array(bootstrapped_early_dms[:,i],copy=True)
null_dms.sort()
lower_early_dms.append( null_dms[len(null_dms)*0.05] )
lower_early_dms = numpy.array(lower_early_dms)
# Calculate upper 95% CI on the (positive) late excess multiplicity
upper_late_dms = []
for i in xrange(0, bootstrapped_late_dms.shape[1]):
null_dms = numpy.array(bootstrapped_late_dms[:,i],copy=True)
null_dms.sort()
upper_late_dms.append( null_dms[long(len(null_dms)*0.95)] )
upper_late_dms = numpy.array(upper_late_dms)
lower_dm = numpy.fmin(early_dms, lower_early_dms).min()*1.1
upper_dm = numpy.fmax(late_dms, upper_late_dms).max()*1.1
ddms = late_dms-early_dms
tstar_idx = ddms.argmax()
tstar = tstars[tstar_idx]
max_ddm = ddms[tstar_idx]
total_ddm = ddms[0]
bootstrapped_ddms = bootstrapped_late_dms-bootstrapped_early_dms
bootstrapped_max_ddms = bootstrapped_ddms.max(axis=1)
max_ddm_pvalue = stats_utils.calculate_empirical_pvalue(max_ddm, bootstrapped_max_ddms)
total_ddm_pvalue = stats_utils.calculate_empirical_pvalue(total_ddm, bootstrapped_ddms[:,0])
missed_opportunity_axis.plot(tstars,numpy.zeros_like(tstars),'k-',linewidth=0.25)
missed_opportunity_axis.fill_between(tstars, lower_early_dms, numpy.zeros_like(lower_early_dms),color=early_color,alpha=0.25)
#missed_opportunity_axis.plot(tstars, lower_early_dms, '-',color=early_color,linewidth=0.25)
missed_opportunity_axis.plot(tstars, early_dms, '-', color=early_color,label='$\leq t^*$')
missed_opportunity_axis.fill_between(tstars, numpy.zeros_like(upper_late_dms), upper_late_dms,color=late_color,alpha=0.25)
#missed_opportunity_axis.plot(tstars, upper_late_dms, '-',color=late_color,linewidth=0.25)
missed_opportunity_axis.plot(tstars, late_dms, '-', color=late_color,label='$> t^*$')
#missed_opportunity_axis.plot([tstar,tstar],[lower_dm,upper_dm],'k-',linewidth=0.25)
#missed_opportunity_axis.set_ylim([lower_dm,upper_dm])
missed_opportunity_axis.set_ylim([-15,20])
missed_opportunity_axis.legend(loc='upper right', frameon=False)
early_ymax = early_dms[tstar_idx]
ymax = late_dms[tstar_idx]
y0 = late_dms[0]
#print tstar, late_dms[tstar_idx], early_dms[tstar_idx], max_ddm, bootstrapped_max_ddms.mean(), bootstrapped_max_ddms.std(), max_ddm_pvalue
sys.stdout.write("Total ddM = %g (p=%g, %d bootstraps)\n" % (total_ddm, total_ddm_pvalue, num_bootstraps))
sys.stdout.write("t^* for largest excess opportunities = %g (p=%g, %d bootstraps)\n" % (tstar, max_ddm_pvalue, num_bootstraps))
sys.stdout.write("Early=%g, Late=%g\n" % (early_dms[tstar_idx], late_dms[tstar_idx]))
########
#
# Now plot dispersion matrices for (0,60000), (0,tstar), and (tstar, 60000)
#
########
for tmin, tmax, matrix_axis in zip([-1000, -1000, tstar],[65000, tstar, 65000], [all_matrix_axis, early_matrix_axis, late_matrix_axis]):
sys.stderr.write("Calculating dispersion matrix for %d<=t<=%d\t" % (tmin,tmax))
population_matrix = observed_population_matrix[(observed_median_times>tmin)*(observed_median_times<=tmax),:]
observed_num_pops = (population_matrix>0.5).sum(axis=1)
observed_num_muts = (population_matrix).sum(axis=1)
bootstrapped_num_pops = []
# num hits is conserved in bootstraping
for bootstrap_idx in xrange(1,num_bootstraps+1):
bootstrapped_population_matrix = mutation_spectrum_utils.resample_population_matrix(population_matrix)
bootstrapped_num_pops.append( (bootstrapped_population_matrix>0.5).sum(axis=1) )
bootstrapped_num_pops = numpy.array(bootstrapped_num_pops)
raw_count_matrix = [] # the number of counts in that tile
excess_probability_matrix = [] # the deviation in expected probability of that tile
missed_opportunity_matrix = [] # the number of missed opportunities in that tile
for hlower, h, hupper in [(1.5,2,2.5),(2.5,3,3.5),(3.5,4,4.5),(4.5,5,5.5),(5.5,6,100.5)]:
# The genes that belong to this row...
desired_idxs = numpy.nonzero((observed_num_muts>hlower)*(observed_num_muts<hupper))[0]
observed_counts = numpy.zeros(6)*1.0
null_counts = numpy.zeros(6)*1.0
for idx in desired_idxs:
observed_counts[observed_num_pops[idx]-1]+=1
bootstrapped_counts = numpy.array([(bootstrapped_num_pops[:,idx]==k).sum() for k in xrange(1,6+1)])
null_counts += bootstrapped_counts*1.0/(bootstrapped_counts.sum())
missed_opportunities = h-numpy.arange(1,h+1)
raw_count_matrix.append(observed_counts)
excess_probability_matrix.append((observed_counts-null_counts)/(null_counts.sum())*100)
missed_opportunity_matrix.append(missed_opportunities)
raw_count_matrix = numpy.array([row for row in reversed(raw_count_matrix)])
excess_probability_matrix = numpy.array([row for row in reversed(excess_probability_matrix)])
missed_opportunity_matrix = numpy.array([row for row in reversed(missed_opportunity_matrix)])
# plot matrix
m = matrix_axis.pcolor(excess_probability_matrix, cmap=matrix_cmap, vmin=matrix_vmin, vmax=matrix_vmax)
# plot border
matrix_axis.plot([0,2,2,3,3,4,4,5,5,6,6], [5,5,4,4,3,3,2,2,1,1,0], 'k-')
# plot grid
matrix_axis.plot([1,1],[0,5],'k-',linewidth=0.25)
matrix_axis.plot([2,2],[0,4],'k-',linewidth=0.25)
matrix_axis.plot([3,3],[0,3],'k-',linewidth=0.25)
matrix_axis.plot([4,4],[0,2],'k-',linewidth=0.25)
matrix_axis.plot([5,5],[0,1],'k-',linewidth=0.25)
matrix_axis.plot([0,2],[4,4],'k-',linewidth=0.25)
matrix_axis.plot([0,3],[3,3],'k-',linewidth=0.25)
matrix_axis.plot([0,4],[2,2],'k-',linewidth=0.25)
matrix_axis.plot([0,5],[1,1],'k-',linewidth=0.25)
# plot missed opportunity # guides
for row_idx in xrange(0,len(missed_opportunity_matrix)):
for col_idx in xrange(0,len(missed_opportunity_matrix[row_idx])):
missed_opportunities = missed_opportunity_matrix[row_idx][col_idx]
raw_count = raw_count_matrix[row_idx][col_idx]
x = col_idx+0.25
y = row_idx+0.35