-
Notifications
You must be signed in to change notification settings - Fork 27
/
layer_optimal_models.R
executable file
·1005 lines (922 loc) · 33.6 KB
/
layer_optimal_models.R
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
library('SingleCellExperiment')
library('here')
library('jaffelab')
library('scater')
library('scran')
library('pheatmap')
library('readxl')
library('Polychrome')
library('cluster')
library('limma')
library('sessioninfo')
library('ggplot2')
library('viridisLite')
load("rda/sce_layer.Rdata", verbose = TRUE)
source("layer_specificity_functions.R", echo = TRUE)
## Use genes_km_raw and genes_bm_raw from layer_specificity_functions.R
## Match the layer names, combine 6 and 6b into a single layer for thje KM list
## Combine 5a and 6b for the BM list
genes_km_mat <- genes_km_raw[, c(1:6, 'WM')]
colnames(genes_km_mat)[-7] <-
paste0('Layer', colnames(genes_km_mat)[-7])
genes_km_mat[['Layer6']] <-
pmin(genes_km_raw[['6']] + genes_km_raw[['6b']], 1)
genes_bm_mat <- genes_bm_raw[, as.character(c(1:4))]
genes_bm_mat[['5']] <-
pmin(genes_bm_raw[['5a']] + genes_bm_raw[['5b']], 1)
genes_bm_mat[['6']] <- genes_bm_raw[['6']]
colnames(genes_bm_mat) <- paste0('Layer', colnames(genes_bm_mat))
genes_bm_mat[['WM']] <- 0
## Genes to query
genes_query <- data.frame(
symbol = c(genes_km_raw$Gene, genes_bm_raw$Gene),
list = rep(c('KM', 'BM'), c(nrow(genes_km_raw), nrow(genes_bm_raw))),
species = c(genes_km_raw$Species, rep('MH', nrow(genes_bm_raw))),
stringsAsFactors = FALSE
)
genes_query$m <-
match(tolower(genes_query$symbol), tolower(rowData(sce_layer)$gene_name))
genes_query$gene_name <- rowData(sce_layer)$gene_name[genes_query$m]
genes_query$ensembl <- rownames(sce_layer)[genes_query$m]
## Combine the matrices into one
genes_query_mat <- rbind(genes_km_mat, genes_bm_mat)
dim(genes_query_mat)
# [1] 146 7
## Note this includes 4 genes that are fully duplicated in the lists
## as determined further below
addmargins(table(
'Expressed in our data' = !is.na(genes_query$m),
'Has useful model info' = rowSums(genes_query_mat) > 0
))
# Has useful model info
# Expressed in our data FALSE TRUE Sum
# FALSE 0 14 14
# TRUE 2 130 132
# Sum 2 144 146
## Keep only those that are present in the data and have useful matrix info
## otherwise the model fails later
marker_genes <- list(gene_info = genes_query[!is.na(genes_query$m) &
rowSums(genes_query_mat) > 0,],
layer_mat = genes_query_mat[!is.na(genes_query$m) &
rowSums(genes_query_mat) > 0,])
## How many of these are unique models?
marker_genes$gene_info$models <-
apply(marker_genes$layer_mat > 0, 1, function(x) {
paste(colnames(marker_genes$layer_mat[x]), collapse = '+')
})
length(unique(marker_genes$gene_info$models))
# [1] 29
## Check if any genes are duplicated and if the layer assignment agree
dups_ensembl <-
unique(marker_genes$gene_info$ensembl[duplicated(marker_genes$gene_info$ensembl)])
dups_models <- lapply(dups_ensembl, function(ens) {
i <- which(marker_genes$gene_info$ensembl == ens)
marker_genes$gene_info$models[i]
})
names(dups_models) <- dups_ensembl
## At most a gene shows up twice: once per list
stopifnot(all(lengths(dups_models) == 2))
## Find which ones are truly duplicated: the gene and the model are the same
dups_ensembl_remove <-
dups_ensembl[lengths(sapply(dups_models, unique)) == 1]
## Update the list section to label them as fully duplicated in KM and BM
marker_genes$gene_info$list[marker_genes$gene_info$ensembl %in% dups_ensembl_remove &
marker_genes$gene_info$list == 'KM'] <- 'KM+BM'
## Drop the redundant entries
dups_drop <-
which(
marker_genes$gene_info$ensembl %in% dups_ensembl_remove &
marker_genes$gene_info$list == 'BM'
)
marker_genes$gene_info[dups_drop,]
# symbol list species m gene_name ensembl models
# 84 RELN BM MH 8638 RELN ENSG00000189056 Layer1
# 93 RorB BM MH 10825 RORB ENSG00000198963 Layer4
# 101 cux2 BM MH 14444 CUX2 ENSG00000111249 Layer2+Layer3+Layer4
# 137 Foxp2 BM MH 8689 FOXP2 ENSG00000128573 Layer6
marker_genes <- lapply(marker_genes, function(x) {
x[-dups_drop,]
})
##### Compute the duplicate correlation globally
## Extract the data
mat <- assays(sce_layer)$logcounts
## Build a group model
mod <- with(colData(sce_layer), model.matrix(~ layer_guess))
colnames(mod) <- gsub('layer_guess', '', colnames(mod))
## Takes like 2 min to run
corfit <-
duplicateCorrelation(mat, mod, block = sce_layer$subject_position)
#######
## What is the estimated block correlation?
corfit$consensus.correlation
# [1] 0.07771005 ## note that ~ 0 + layer_guess vs ~ layer_guess has the same cor
## Process each marker gene (use the full gene matrix to get global stats)
## takes about 3 min to run
layer_idx <- splitit(sce_layer$layer_guess)
eb_markers_list <-
lapply(seq_len(nrow(marker_genes$gene_info)), function(i) {
res <- rep(0, ncol(sce_layer))
layers <-
colnames(marker_genes$layer_mat)[marker_genes$layer_mat[i,] > 0]
res[unlist(layer_idx[layers])] <- 1
m <- model.matrix(~ res)
eBayes(
lmFit(
mat,
design = m,
block = sce_layer$subject_position,
correlation = corfit$consensus.correlation
)
)
})
## Extract p-values, FDR, t-stat, logFC
pvals_markers <- sapply(eb_markers_list, function(x) {
x$p.value[, 2, drop = FALSE]
})
fdr_markers <- apply(pvals_markers, 2, p.adjust, 'fdr')
tstats_markers <- sapply(eb_markers_list, function(x) {
x$t[, 2, drop = FALSE]
})
logFC_markers <- sapply(eb_markers_list, function(x) {
x$coefficients[, 2, drop = FALSE]
})
## Create a summary table
marker_genes_summary <- cbind(marker_genes[[1]], marker_genes[[2]])
## Function to extract the relevant info
marker_extract <- function(m, j = marker_genes_summary$m) {
sapply(seq_along(j), function(i) {
m[j[i], i]
})
}
marker_genes_summary$p_value <- marker_extract(pvals_markers)
marker_genes_summary$fdr <- marker_extract(fdr_markers)
marker_genes_summary$t_stat <- marker_extract(tstats_markers)
marker_genes_summary$log_fc <- marker_extract(logFC_markers)
## Get the ranks in decreasing order, so it's:
# x <- c(-3, 4, 5, 1)
# rank(-x)
## Thus -log10(pval) * -1 = log10(pval)
marker_genes_summary$rank <-
marker_extract(apply(log10(pvals_markers), 2, rank))
## Or I could have done apply(pvals_markers, 2, rank) and gotten the same result :P
stopifnot(identical(marker_genes_summary$rank,
marker_extract(apply(
pvals_markers, 2, rank
))))
## And it's the same as the abs t-stats ranks
stopifnot(identical(marker_genes_summary$rank,
marker_extract(apply(
-abs(tstats_markers), 2, rank
))))
## t-stats should ideally be positive, not all are
addmargins(table(
sign(marker_genes_summary$t_stat),
marker_genes_summary$species
))
# H MH Sum
# -1 5 16 21
# 1 17 88 105
# Sum 22 104 126
## The ranks are much better for some results when the t-stat is positive
tapply(marker_genes_summary$rank,
paste(
sign(marker_genes_summary$t_stat),
marker_genes_summary$species
),
summary)
# $`-1 H`
# Min. 1st Qu. Median Mean 3rd Qu. Max.
# 1792 1977 4155 6862 8980 17405
#
# $`-1 MH`
# Min. 1st Qu. Median Mean 3rd Qu. Max.
# 1027 5618 9434 9786 12660 21827
#
# $`1 H`
# Min. 1st Qu. Median Mean 3rd Qu. Max.
# 6 29 738 2039 1883 12092
#
# $`1 MH`
# Min. 1st Qu. Median Mean 3rd Qu. Max.
# 1.00 46.25 314.50 3487.25 3401.00 21847.00
## 33 have p-value < 1e-8 for the positive t-stats
addmargins(do.call(
rbind,
tapply(marker_genes_summary$p_value,
paste(
sign(marker_genes_summary$t_stat),
marker_genes_summary$species
), function(x)
table(factor(x < 1e-8, levels = c('FALSE', 'TRUE'))))
))
# FALSE TRUE Sum
# -1 H 5 0 5
# -1 MH 16 0 16
# 1 H 13 4 17
# 1 MH 59 29 88
# Sum 93 33 126
## 47 have p-value < 1e-6 for the positive t-stats
addmargins(do.call(
rbind,
tapply(marker_genes_summary$p_value,
paste(
sign(marker_genes_summary$t_stat),
marker_genes_summary$species
), function(x)
table(factor(x < 1e-6, levels = c('FALSE', 'TRUE'))))
))
# FALSE TRUE Sum
# -1 H 5 0 5
# -1 MH 16 0 16
# 1 H 9 8 17
# 1 MH 49 39 88
# Sum 79 47 126
## 75 have FDR < 0.05 for the positive t-stats
addmargins(do.call(
rbind,
tapply(marker_genes_summary$fdr,
paste(
sign(marker_genes_summary$t_stat),
marker_genes_summary$species
), function(x)
table(factor(x < 0.05, levels = c('FALSE', 'TRUE'))))
))
# FALSE TRUE Sum
# -1 H 5 0 5
# -1 MH 16 0 16
# 1 H 5 12 17
# 1 MH 25 63 88
# Sum 51 75 126
## Add mean expr
marker_genes_summary$mean_expr <-
rowMeans(mat[marker_genes_summary$m,])
## Find the best gene for each model
marker_genes_summary$best_gene_m <-
apply(log10(pvals_markers), 2, function(x) {
which(rank(x) == 1)
})
marker_genes_summary$best_gene_name <-
rowData(sce_layer)$gene_name[marker_genes_summary$best_gene_m]
marker_genes_summary$best_gene_ensembl <-
rownames(sce_layer)[marker_genes_summary$best_gene_m]
marker_genes_summary$best_gene_p_value <-
marker_extract(pvals_markers, marker_genes_summary$best_gene_m)
marker_genes_summary$best_gene_fdr <-
marker_extract(fdr_markers, marker_genes_summary$best_gene_m)
marker_genes_summary$best_gene_t_stat <-
marker_extract(tstats_markers, marker_genes_summary$best_gene_m)
marker_genes_summary$best_gene_log_fc <-
marker_extract(logFC_markers, marker_genes_summary$best_gene_m)
marker_genes_summary$best_gene_mean_expr <-
rowMeans(mat[marker_genes_summary$best_gene_m,])
## Re-order by p-value
marker_genes <- lapply(marker_genes, function(x) {
x[order(marker_genes_summary$p_value),]
})
marker_genes_summary <-
marker_genes_summary[order(marker_genes_summary$p_value),]
## Save for later
save(eb_markers_list, file = 'rda/eb_markers_list.Rdata')
save(marker_genes, marker_genes_summary, file = 'rda/marker_genes_summary.Rdata')
write.csv(
marker_genes_summary,
file = 'marker_genes_summary.csv',
quote = FALSE,
row.names = FALSE
)
## Make boxplots
layer_guess_reordered <-
factor(sce_layer$layer_guess, levels = c(paste0('Layer', 1:6), 'WM'))
pdf('pdf/markers_gene_optimal_boxplots.pdf', useDingbats = FALSE)
set.seed(20200206)
for (i in seq_len(nrow(marker_genes_summary))) {
# i <- 1
message(
paste(
Sys.time(),
'making the plot for',
i,
'gene',
marker_genes_summary$gene_name[i]
)
)
curr_layers <-
strsplit(marker_genes_summary$models[i], '\\+')[[1]]
boxplot(
mat[marker_genes_summary$m[i], ] ~ layer_guess_reordered,
xlab = 'Layer',
ylab = 'logcounts',
main = paste(
marker_genes_summary$gene_name[i],
marker_genes_summary$ensembl[i],
paste(
ifelse(marker_genes_summary$species[i] == 'H', 'human', 'mouse'),
'marker'
),
paste0('(', marker_genes_summary$list[i], ')'),
'\n',
paste(marker_genes_summary$models[i], 'vs others'),
'\n',
'tstat',
formatC(
marker_genes_summary$t_stat[i],
format = "e",
digits = 2
),
'p',
formatC(
marker_genes_summary$p_val[i],
format = "e",
digits = 2
),
'rank',
marker_genes_summary$rank[i]
),
outline = FALSE,
cex = 1.5,
col = ifelse(
levels(layer_guess_reordered) %in% curr_layers,
'skyblue',
'violet'
)
)
points(
mat[marker_genes_summary$m[i], ] ~ jitter(as.integer(layer_guess_reordered)),
pch = 21,
bg = ifelse(
layer_guess_reordered %in% curr_layers,
'dodgerblue4',
'darkviolet'
),
cex = 1.5
)
# legend(
# "top",
# levels(sce_layer$layer_guess),
# col = Polychrome::palette36.colors(7),
# lwd = 2
# )
}
dev.off()
layer_guess_reordered_short <- layer_guess_reordered
levels(layer_guess_reordered_short) <-
gsub('ayer', '', levels(layer_guess_reordered_short))
pdf('pdf/markers_gene_optimal_boxplots_short.pdf', useDingbats = FALSE)
set.seed(20200206)
for (i in seq_len(nrow(marker_genes_summary))) {
# i <- 1
par(mar = c(2, 3, 2, 1) + 0.1)
message(
paste(
Sys.time(),
'making the plot for',
i,
'gene',
marker_genes_summary$gene_name[i]
)
)
curr_layers <-
strsplit(marker_genes_summary$models[i], '\\+')[[1]]
boxplot(
mat[marker_genes_summary$m[i], ] ~ layer_guess_reordered_short,
xlab = '',
ylab = '',
main = paste(
marker_genes_summary$gene_name[i],
paste(ifelse(
marker_genes_summary$species[i] == 'H', 'HM', 'MM'
)),
paste0(
'p=',
formatC(
marker_genes_summary$p_val[i],
format = "e",
digits = 2
)
),
'rank',
marker_genes_summary$rank[i]
),
outline = FALSE,
cex = 2,
cex.axis = 2,
cex.lab = 2,
cex.main = 2,
col = ifelse(
levels(layer_guess_reordered) %in% curr_layers,
'skyblue',
'violet'
)
)
points(
mat[marker_genes_summary$m[i], ] ~ jitter(as.integer(layer_guess_reordered_short)),
pch = 21,
bg = ifelse(
layer_guess_reordered %in% curr_layers,
'dodgerblue4',
'darkviolet'
),
cex = 2
)
}
dev.off()
marker_genes_summary$species <-
ifelse(marker_genes_summary$species == 'MH', 'mouse', 'human')
pdf('pdf/markers_gene_optimal_scatter.pdf',
width = 12,
useDingbats = FALSE)
ggplot(marker_genes_summary,
aes(
x = -log10(best_gene_p_value),
y = -log10(p_value),
color = -log10(rank / max(rank))
)) + geom_point(size = 4) + theme_bw(base_size = 25) +
ylab('-log10 p-value') + xlab('-log10 p-value for the best gene') +
geom_abline(
slope = 0,
intercept = -log10(1e-05),
color = 'red',
linetype = 2
) +
geom_abline(
slope = 1,
intercept = 0,
color = 'grey50',
linetype = 3
) +
facet_grid(~ species) +
scale_color_gradientn(name = '-log10\nrank\npercentile', colors = viridis(11))
ggplot(marker_genes_summary,
aes(
x = best_gene_t_stat,
y = t_stat,
color = -log10(rank / max(rank))
)) + geom_point(size = 4) + coord_fixed() + theme_bw(base_size = 25) +
ylab('t-statistic') + xlab('t-statistic for the best gene') +
geom_abline(
slope = 1,
intercept = 0,
color = 'grey50',
linetype = 3
) +
facet_grid(~ species) +
scale_color_gradientn(name = '-log10\nrank\npercentile', colors = viridis(11))
ggplot(marker_genes_summary,
aes(
x = best_gene_log_fc,
y = log_fc,
color = -log10(rank / max(rank))
)) + geom_point(size = 4) + coord_fixed() + theme_bw(base_size = 25) +
ylab('log fold change') + xlab('log fold change for the best gene') +
geom_abline(
slope = 1,
intercept = 0,
color = 'grey50',
linetype = 3
) +
facet_grid(~ species) +
scale_color_gradientn(name = '-log10\nrank\npercentile', colors = viridis(11))
dev.off()
with(marker_genes_summary, cor(-log10(best_gene_p_value), -log10(p_value)))
# [1] 0.4224215
with(marker_genes_summary, cor(-log10(best_gene_fdr), -log10(fdr)))
# [1] 0.4719611
with(marker_genes_summary, cor(best_gene_t_stat, t_stat))
# [1] 0.144047
with(marker_genes_summary, cor(best_gene_log_fc, log_fc))
# [1] 0.1767757
### Numbers for:
## First, we tested for enrichment of these genes - as a set - among the DEGs
## described above, and found strong enrichment (p = XX). However, only
## XX of the genes were significant DEGs in our human data.
## Load summary t-stats
load('rda/modeling_results.Rdata', verbose = TRUE)
compute_GST <-
function(type = 'f_stat_full',
ranks.only = TRUE,
orgs = c('mouse', 'human')) {
set.seed(20200218)
df <- subset(marker_genes_summary, species %in% orgs)
geneSetTest(
index = match(unique(df$ensembl), results_anova$ensembl),
statistics = results_anova[[type]],
ranks.only = ranks.only
)
}
apply_GST <- function(type) {
data.frame(
'all' = compute_GST(type),
'all_sim' = compute_GST(type, ranks.only = FALSE),
'mouse' = compute_GST(type, orgs = 'mouse'),
'mouse_sim' = compute_GST(type, ranks.only = FALSE, orgs = 'mouse'),
'human' = compute_GST(type, orgs = 'human'),
'human_sim' = compute_GST(type, ranks.only = FALSE, orgs = 'human'),
'type' = type,
stringsAsFactors = FALSE
)
}
results_GST <- rbind(apply_GST('f_stat_full'),
apply_GST('f_stat_noWM'))
options(width = 120)
results_GST
# all all_sim mouse mouse_sim human human_sim type
# 1 1.225341e-41 1e-04 8.072116e-35 1e-04 1.308392e-08 1e-04 f_stat_full
# 2 1.216983e-44 1e-04 6.685479e-36 1e-04 1.380023e-10 1e-04 f_stat_noWM
addmargins(with(marker_genes_summary, table('FDR <5%' = fdr < 0.05, species)))
# species
# FDR <5% human mouse Sum
# FALSE 10 41 51
# TRUE 12 63 75
# Sum 22 104 126
## Six genes appear twice in the table
marker_genes_summary$gene_name[duplicated(marker_genes_summary$ensembl)]
# [1] "ETV1" "PCP4" "TLE4" "CARTPT" "KCNIP2" "CRYM"
dups <-
marker_genes_summary$ensembl %in% marker_genes_summary$ensembl[duplicated(marker_genes_summary$ensembl)]
table(dups)
# FALSE TRUE
# 114 12
addmargins(with(
marker_genes_summary,
table(
'FDR <5%' = fdr < 0.05,
species,
'Has more than 1 model' = dups
)
))
# , , Has more than 1 model = FALSE
#
# species
# FDR <5% human mouse Sum
# FALSE 10 39 49
# TRUE 12 53 65
# Sum 22 92 114
#
# , , Has more than 1 model = TRUE
#
# species
# FDR <5% human mouse Sum
# FALSE 0 2 2
# TRUE 0 10 10
# Sum 0 12 12
#
# , , Has more than 1 model = Sum
#
# species
# FDR <5% human mouse Sum
# FALSE 10 41 51
# TRUE 12 63 75
# Sum 22 104 126
75 / 126 * 100
# [1] 59.52381
65 / 114 * 100
# [1] 57.01754
## Lets look at those genes
with(marker_genes_summary[dups, ],
table('FDR <5%' = fdr < 0.05, species, 'Gene' = gene_name))
# , , Gene = CARTPT
#
# species
# FDR <5% mouse
# FALSE 0
# TRUE 2
#
# , , Gene = CRYM
#
# species
# FDR <5% mouse
# FALSE 1
# TRUE 1
#
# , , Gene = ETV1
#
# species
# FDR <5% mouse
# FALSE 0
# TRUE 2
#
# , , Gene = KCNIP2
#
# species
# FDR <5% mouse
# FALSE 1
# TRUE 1
#
# , , Gene = PCP4
#
# species
# FDR <5% mouse
# FALSE 0
# TRUE 2
#
# , , Gene = TLE4
#
# species
# FDR <5% mouse
# FALSE 0
# TRUE 2
options(width = 200)
with(marker_genes_summary[dups, ],
table('FDR <5%' = fdr < 0.05, models, 'Gene' = gene_name))
# , , Gene = CARTPT
#
# models
# FDR <5% Layer2 Layer2+Layer3 Layer2+Layer3+Layer5+Layer6 Layer3 Layer3+Layer4+Layer5+Layer6 Layer4 Layer5 Layer5+Layer6 Layer6
# FALSE 0 0 0 0 0 0 0 0 0
# TRUE 1 0 0 1 0 0 0 0 0
#
# , , Gene = CRYM
#
# models
# FDR <5% Layer2 Layer2+Layer3 Layer2+Layer3+Layer5+Layer6 Layer3 Layer3+Layer4+Layer5+Layer6 Layer4 Layer5 Layer5+Layer6 Layer6
# FALSE 0 0 0 0 0 0 0 1 0
# TRUE 0 0 1 0 0 0 0 0 0
#
# , , Gene = ETV1
#
# models
# FDR <5% Layer2 Layer2+Layer3 Layer2+Layer3+Layer5+Layer6 Layer3 Layer3+Layer4+Layer5+Layer6 Layer4 Layer5 Layer5+Layer6 Layer6
# FALSE 0 0 0 0 0 0 0 0 0
# TRUE 0 0 0 0 0 0 1 1 0
#
# , , Gene = KCNIP2
#
# models
# FDR <5% Layer2 Layer2+Layer3 Layer2+Layer3+Layer5+Layer6 Layer3 Layer3+Layer4+Layer5+Layer6 Layer4 Layer5 Layer5+Layer6 Layer6
# FALSE 0 0 0 0 0 1 0 0 0
# TRUE 0 1 0 0 0 0 0 0 0
#
# , , Gene = PCP4
#
# models
# FDR <5% Layer2 Layer2+Layer3 Layer2+Layer3+Layer5+Layer6 Layer3 Layer3+Layer4+Layer5+Layer6 Layer4 Layer5 Layer5+Layer6 Layer6
# FALSE 0 0 0 0 0 0 0 0 0
# TRUE 0 0 0 0 1 0 1 0 0
#
# , , Gene = TLE4
#
# models
# FDR <5% Layer2 Layer2+Layer3 Layer2+Layer3+Layer5+Layer6 Layer3 Layer3+Layer4+Layer5+Layer6 Layer4 Layer5 Layer5+Layer6 Layer6
# FALSE 0 0 0 0 0 0 0 0 0
# TRUE 0 0 0 0 0 0 0 1 1
## Answer the following question:
## The biggest factor contributing to replication of markers in our data was XXX
addmargins(with(
marker_genes_summary,
table(
'pval < 1e-5' = p_value < 1e-5,
species,
'Has more than 1 model' = dups
)
))
# , , Has more than 1 model = FALSE
#
# species
# pval < 1e-5 human mouse Sum
# FALSE 14 54 68
# TRUE 8 38 46
# Sum 22 92 114
#
# , , Has more than 1 model = TRUE
#
# species
# pval < 1e-5 human mouse Sum
# FALSE 0 7 7
# TRUE 0 5 5
# Sum 0 12 12
#
# , , Has more than 1 model = Sum
#
# species
# pval < 1e-5 human mouse Sum
# FALSE 14 61 75
# TRUE 8 43 51
# Sum 22 104 126
51 / 126 * 100
# [1] 40.47619
46 / 114 * 100
# [1] 40.35088
marker_genes_logistic <- data.frame(
replicated = marker_genes_summary$p_value < 1e-05,
replicated_fdr = marker_genes_summary$fdr < 0.05,
n_layers = stringr::str_count(marker_genes_summary$models, '\\+') + 1,
species = marker_genes_summary$species,
list_km = grepl('KM', marker_genes_summary$list),
list_bm = grepl('BM', marker_genes_summary$list),
is_dup = dups,
stringsAsFactors = FALSE
)
fit_log <-
glm(
replicated ~ n_layers + species + list_km + list_bm + is_dup,
family = binomial,
data = marker_genes_logistic
)
summary(fit_log)
# Call:
# glm(formula = replicated ~ n_layers + species + list_km + list_bm +
# is_dup, family = binomial, data = marker_genes_logistic)
#
# Deviance Residuals:
# Min 1Q Median 3Q Max
# -1.6148 -1.0078 -0.7411 1.2159 1.6081
#
# Coefficients:
# Estimate Std. Error z value Pr(>|z|)
# (Intercept) -3.0965 1.3727 -2.256 0.0241 *
# n_layers 0.2463 0.1976 1.247 0.2125
# speciesmouse 0.3747 0.5514 0.680 0.4968
# list_kmTRUE 2.1390 1.2139 1.762 0.0780 .
# list_bmTRUE 1.3233 1.2096 1.094 0.2739
# is_dupTRUE 0.1798 0.6547 0.275 0.7836
# ---
# Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
#
# (Dispersion parameter for binomial family taken to be 1)
#
# Null deviance: 170.07 on 125 degrees of freedom
# Residual deviance: 162.16 on 120 degrees of freedom
# AIC: 174.16
#
# Number of Fisher Scoring iterations: 4
anova(fit_log, test = 'Chisq')
# Analysis of Deviance Table
#
# Model: binomial, link: logit
#
# Response: replicated
#
# Terms added sequentially (first to last)
#
#
# Df Deviance Resid. Df Resid. Dev Pr(>Chi)
# NULL 125 170.07
# n_layers 1 1.5324 124 168.54 0.21575
# species 1 0.0292 123 168.51 0.86443
# list_km 1 4.9448 122 163.57 0.02617 *
# list_bm 1 1.3312 121 162.24 0.24859
# is_dup 1 0.0750 120 162.16 0.78420
# ---
# Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
fit_log_fdr <-
glm(
replicated_fdr ~ n_layers + species + list_km + list_bm + is_dup,
family = binomial,
data = marker_genes_logistic
)
summary(fit_log_fdr)
# Call:
# glm(formula = replicated_fdr ~ n_layers + species + list_km +
# list_bm + is_dup, family = binomial, data = marker_genes_logistic)
#
# Deviance Residuals:
# Min 1Q Median 3Q Max
# -2.0582 -0.9220 0.6484 0.8207 1.6986
#
# Coefficients:
# Estimate Std. Error z value Pr(>|z|)
# (Intercept) -18.1615 1168.7051 -0.016 0.9876
# n_layers 0.5375 0.2246 2.394 0.0167 *
# speciesmouse 0.4949 0.5743 0.862 0.3888
# list_kmTRUE 17.5067 1168.7049 0.015 0.9880
# list_bmTRUE 15.9561 1168.7049 0.014 0.9891
# is_dupTRUE 1.9547 0.8833 2.213 0.0269 *
# ---
# Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
#
# (Dispersion parameter for binomial family taken to be 1)
#
# Null deviance: 170.07 on 125 degrees of freedom
# Residual deviance: 143.13 on 120 degrees of freedom
# AIC: 155.13
#
# Number of Fisher Scoring iterations: 15
anova(fit_log_fdr, test = 'Chisq')
# Analysis of Deviance Table
#
# Model: binomial, link: logit
#
# Response: replicated_fdr
#
# Terms added sequentially (first to last)
#
#
# Df Deviance Resid. Df Resid. Dev Pr(>Chi)
# NULL 125 170.07
# n_layers 1 4.3650 124 165.71 0.0366844 *
# species 1 0.0039 123 165.71 0.9505145
# list_km 1 13.7117 122 151.99 0.0002131 ***
# list_bm 1 2.7375 121 149.26 0.0980165 .
# is_dup 1 6.1288 120 143.13 0.0133000 *
# ---
# Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
## Reproducibility information
print('Reproducibility information:')
Sys.time()
proc.time()
options(width = 120)
session_info()
# ─ Session info ───────────────────────────────────────────────────────────────────────────────────────────────────────
# setting value
# version R version 3.6.1 Patched (2019-10-31 r77350)
# os CentOS Linux 7 (Core)
# system x86_64, linux-gnu
# ui X11
# language (EN)
# collate en_US.UTF-8
# ctype en_US.UTF-8
# tz US/Eastern
# date 2020-02-21
#
# ─ Packages ───────────────────────────────────────────────────────────────────────────────────────────────────────────
# package * version date lib source
# assertthat 0.2.1 2019-03-21 [2] CRAN (R 3.6.1)
# backports 1.1.5 2019-10-02 [1] CRAN (R 3.6.1)
# beeswarm 0.2.3 2016-04-25 [2] CRAN (R 3.6.1)
# Biobase * 2.46.0 2019-10-29 [2] Bioconductor
# BiocGenerics * 0.32.0 2019-10-29 [1] Bioconductor
# BiocNeighbors 1.4.1 2019-11-01 [2] Bioconductor
# BiocParallel * 1.20.1 2019-12-21 [1] Bioconductor
# BiocSingular 1.2.2 2020-02-14 [2] Bioconductor
# bitops 1.0-6 2013-08-17 [2] CRAN (R 3.6.1)
# cellranger 1.1.0 2016-07-27 [1] CRAN (R 3.6.1)
# cli 2.0.0 2019-12-09 [1] CRAN (R 3.6.1)
# cluster * 2.1.0 2019-06-19 [3] CRAN (R 3.6.1)
# colorout * 1.2-2 2019-10-31 [1] Github (jalvesaq/colorout@641ed38)
# colorspace 1.4-1 2019-03-18 [2] CRAN (R 3.6.1)
# crayon 1.3.4 2017-09-16 [1] CRAN (R 3.6.1)
# DelayedArray * 0.12.2 2020-01-06 [2] Bioconductor
# DelayedMatrixStats 1.8.0 2019-10-29 [2] Bioconductor
# digest 0.6.23 2019-11-23 [1] CRAN (R 3.6.1)
# dqrng 0.2.1 2019-05-17 [1] CRAN (R 3.6.1)
# edgeR 3.28.0 2019-10-29 [1] Bioconductor
# fansi 0.4.0 2018-10-05 [1] CRAN (R 3.6.1)
# GenomeInfoDb * 1.22.0 2019-10-29 [1] Bioconductor
# GenomeInfoDbData 1.2.2 2019-10-28 [2] Bioconductor
# GenomicRanges * 1.38.0 2019-10-29 [1] Bioconductor
# ggbeeswarm 0.6.0 2017-08-07 [2] CRAN (R 3.6.1)
# ggplot2 * 3.2.1 2019-08-10 [1] CRAN (R 3.6.1)
# glue 1.3.1 2019-03-12 [1] CRAN (R 3.6.1)
# googledrive 1.0.0 2019-08-19 [1] CRAN (R 3.6.1)
# gridExtra 2.3 2017-09-09 [2] CRAN (R 3.6.1)
# gtable 0.3.0 2019-03-25 [2] CRAN (R 3.6.1)
# here * 0.1 2017-05-28 [1] CRAN (R 3.6.1)
# htmltools 0.4.0 2019-10-04 [1] CRAN (R 3.6.1)
# htmlwidgets 1.5.1 2019-10-08 [1] CRAN (R 3.6.1)
# httpuv 1.5.2 2019-09-11 [1] CRAN (R 3.6.1)
# igraph 1.2.4.2 2019-11-27 [2] CRAN (R 3.6.1)
# IRanges * 2.20.1 2019-11-20 [1] Bioconductor
# irlba 2.3.3 2019-02-05 [1] CRAN (R 3.6.1)
# jaffelab * 0.99.29 2019-11-04 [1] Github (LieberInstitute/jaffelab@a7d87cb)
# jsonlite 1.6.1 2020-02-02 [2] CRAN (R 3.6.1)
# later 1.0.0 2019-10-04 [1] CRAN (R 3.6.1)
# lattice 0.20-38 2018-11-04 [3] CRAN (R 3.6.1)
# lazyeval 0.2.2 2019-03-15 [2] CRAN (R 3.6.1)
# lifecycle 0.1.0 2019-08-01 [1] CRAN (R 3.6.1)
# limma * 3.42.0 2019-10-29 [1] Bioconductor
# locfit 1.5-9.1 2013-04-20 [2] CRAN (R 3.6.1)
# magrittr 1.5 2014-11-22 [1] CRAN (R 3.6.1)
# Matrix 1.2-17 2019-03-22 [3] CRAN (R 3.6.1)
# matrixStats * 0.55.0 2019-09-07 [1] CRAN (R 3.6.1)
# munsell 0.5.0 2018-06-12 [2] CRAN (R 3.6.1)
# pheatmap * 1.0.12 2019-01-04 [2] CRAN (R 3.6.1)
# pillar 1.4.3 2019-12-20 [1] CRAN (R 3.6.1)
# pkgconfig 2.0.3 2019-09-22 [1] CRAN (R 3.6.1)
# png 0.1-7 2013-12-03 [2] CRAN (R 3.6.1)
# Polychrome * 1.2.3 2019-08-01 [1] CRAN (R 3.6.1)
# promises 1.1.0 2019-10-04 [1] CRAN (R 3.6.1)
# R6 2.4.1 2019-11-12 [2] CRAN (R 3.6.1)
# rafalib * 1.0.0 2015-08-09 [1] CRAN (R 3.6.1)
# RColorBrewer 1.1-2 2014-12-07 [2] CRAN (R 3.6.1)
# Rcpp 1.0.3 2019-11-08 [1] CRAN (R 3.6.1)
# RCurl 1.98-1.1 2020-01-19 [2] CRAN (R 3.6.1)
# readxl * 1.3.1 2019-03-13 [2] CRAN (R 3.6.1)
# rlang 0.4.2 2019-11-23 [1] CRAN (R 3.6.1)
# rmote * 0.3.4 2019-10-31 [1] Github (cloudyr/rmote@fbce611)
# rprojroot 1.3-2 2018-01-03 [2] CRAN (R 3.6.1)
# rsvd 1.0.2 2019-07-29 [1] CRAN (R 3.6.1)
# S4Vectors * 0.24.1 2019-12-01 [1] Bioconductor
# scales 1.1.0 2019-11-18 [2] CRAN (R 3.6.1)
# scater * 1.14.6 2019-12-16 [2] Bioconductor
# scatterplot3d 0.3-41 2018-03-14 [1] CRAN (R 3.6.1)
# scran * 1.14.5 2019-11-19 [1] Bioconductor
# segmented 1.1-0 2019-12-10 [2] CRAN (R 3.6.1)
# servr 0.15 2019-08-07 [1] CRAN (R 3.6.1)
# sessioninfo * 1.1.1 2018-11-05 [1] CRAN (R 3.6.1)
# SingleCellExperiment * 1.8.0 2019-10-29 [2] Bioconductor
# statmod 1.4.34 2020-02-17 [2] CRAN (R 3.6.1)
# stringi 1.4.6 2020-02-17 [2] CRAN (R 3.6.1)
# stringr 1.4.0 2019-02-10 [1] CRAN (R 3.6.1)
# SummarizedExperiment * 1.16.1 2019-12-19 [1] Bioconductor
# tibble 2.1.3 2019-06-06 [1] CRAN (R 3.6.1)
# vipor 0.4.5 2017-03-22 [2] CRAN (R 3.6.1)
# viridis 0.5.1 2018-03-29 [2] CRAN (R 3.6.1)
# viridisLite * 0.3.0 2018-02-01 [2] CRAN (R 3.6.1)
# withr 2.1.2 2018-03-15 [2] CRAN (R 3.6.1)
# xfun 0.11 2019-11-12 [1] CRAN (R 3.6.1)
# XVector 0.26.0 2019-10-29 [1] Bioconductor