-
Notifications
You must be signed in to change notification settings - Fork 16
/
global.R
1958 lines (1601 loc) · 64.9 KB
/
global.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
################################################################################################################
## Filename: global.R
## Created: October 09, 2015
## Author(s): Karsten Krug, Ozan Aygun
##
## Purpose: Shiny-app to perform differential expression analysis, primarily on proteomics data, to perform
## simple data QC, to interactively browse through the results and to download high-quality result
## figures.
##
## This file defines global parameters, loads all required R-packages and defines the actual functions to perform data filtering,
## data normalization, the moderated test statistics, and visualization. The code for moderated t-tests,
## two-component normalization and the reproducibility filter has been written by D. R. Mani and
## adopted by me for integration into a Shiny-Server environment.
##
##
## Last updated April 20, 2022 by Natalie Clark (nclark@broadinstitute.org) - v 1.0
##################################################################################################################
#options( stringsAsFactors = F )
## R package managing tool
if (!require("pacman")) install.packages ("pacman")
require('pacman')
if (!require("BiocManager")) install.packages ("BiocManager")
require(BiocManager)
options(repos = BiocManager::repositories())
#################################################################
## global parameters
#################################################################
## use PACMAN R package manager?
## set to FALSE if deployed to RStudio Connect
PACMAN <- FALSE
## version number
VER <- "1.1.7"
## maximal file size for upload
MAXSIZEMB <<- 1024
## list of strings indicating missing data
NASTRINGS <<- c("NA", "<NA>", "#N/A", "#NUM!", "#DIV/0!", "#NA", "#NAME?", "na", "#VALUE!","Na","nA","NaN")
## separator tested in the uploaded file
SEPARATOR <<- c('\t', ',', ';')
## Colors used throughout the app to color the defined groups
GRPCOLORS <<- c(RColorBrewer::brewer.pal(8, "Set1"), RColorBrewer::brewer.pal(8, "Dark2"), RColorBrewer::brewer.pal(8, "Set2"), terrain.colors(20), cm.colors(20), topo.colors(20))
## number of characters to display in plots/tables for column names
STRLENGTH <<- 25
## operating system
OS <<- Sys.info()['sysname']
## temp directory to write the Excel file
TMPDIR <<- ifelse(OS=='Windows', "./", "/tmp/")
## app name
APPNAME <<- 'Protigy'
## app folder
APPDIR <<- getwd()
## directory to store data files
DATADIR <<- ifelse(OS=='Linux', "/shiny-data/", tempdir())
## User database for shared sessions
USERDB <<- file.path(APPDIR, 'conf/user-roles.txt')
## R version
RVERSION <- as.numeric(paste(R.version$major, sub('\\..*','',R.version$minor), sep='.'))
## email for trouble shooting
MAIL <<- 'nclark@broadinstitute.org'
## URL to configuration app (SSP/RSC only)
CONFAPP <<- 'https://rstudio-connect.broadapps.org/protigy-sessions/'
## PIWIK location
PIWIKURL <<- ''
#################################################################
## load required packages
#################################################################
if(PACMAN){
p_load(RColorBrewer)
p_load(shiny)
p_load(shinydashboard)
p_load(shinyjs)
p_load(shinyjqui)
p_load(shinyalert)
p_load(markdown)
p_load(tippy)
p_load(magrittr)
p_load(tibble)
p_load(dplyr)
p_load(stringr)
## heatmap
p_load(pheatmap)
p_load(heatmaply)
## clustering
p_load(ape)
p_load(dendextend)
p_load(scales)
p_load(gtable)
p_load(fastcluster)
## rmarkdown
p_load(rmarkdown)
p_load(knitr)
## moderated tests
p_load(limma)
p_load(statmod)
## multiscatter
p_load(hexbin)
p_load(Hmisc)
p_load(grid)
p_load(scatterplot3d)
p_load(plotly)
## Excel
p_load(readxl)
p_load(WriteXLS)
## reproducibility filter
p_load(reshape)
p_load(nlme)
p_load(BlandAltmanLeh)
## normalization
p_load(preprocessCore)
p_load(vsn)
## normalization 2-component
p_load (mice)
p_load (mixtools)
p_load (mclust)
## table preview
p_load(DT)
## label placements without overlap
p_load(maptools)
p_load(ggrepel)
p_load(UpSetR)
p_load(gridExtra)
p_load(vioplot)
## id mapping
p_load(RSQLite)
p_load(org.Hs.eg.db)
p_load(org.Mm.eg.db)
p_load(org.Rn.eg.db)
p_load(org.Dr.eg.db)
## enrichment analysis
p_load(gprofiler2)
#other needed packages
p_load(seriation)
p_load(preprocessCore)
} else { ## RStudio Connect
library(RColorBrewer)
library(shiny)
library(shinydashboard)
library(shinyjs)
library(shinyjqui)
library(shinyalert)
library(tippy)
library(magrittr)
library(tibble)
library(dplyr)
library(stringr)
## heatmap
library(pheatmap)
library(heatmaply)
## clustering
library(ape)
library(dendextend)
library(scales)
library(gtable)
library(fastcluster)
## rmarkdown
library(rmarkdown)
library(markdown)
library(knitr)
## moderated tests
library(limma)
library(statmod)
## multiscatter
library(hexbin)
library(Hmisc)
library(grid)
library(scatterplot3d)
library(plotly)
## Excel
library(WriteXLS)
library(readxl)
## reproducibility filter
library(reshape)
library(nlme)
library(BlandAltmanLeh)
## normalization
library(preprocessCore)
library(vsn)
## normalization 2-component
library (mice)
library (mixtools)
library (mclust)
## table preview
library(DT)
## label placements without overlap
library(maptools)
library(ggrepel)
library(UpSetR)
library(gridExtra)
library(vioplot)
## id mapping
library(RSQLite)
library(org.Hs.eg.db)
library(org.Mm.eg.db)
library(org.Rn.eg.db)
library(org.Dr.eg.db)
## enrichment analysis
library(gprofiler2)
}
## required to install limma in R >= 3.5
if(RVERSION >= 3.5){
if(PACMAN)
p_load(BiocManager)
else
library(BiocManager)
options(repos = BiocManager::repositories())
}
## pca
# ChemometricsWithR - removed from CRAN
# use GH version instead
if(!require("ChemometricsWithR")){
install.packages("remotes")
library(remotes)
install_github("rwehrens/ChemometricsWithR")
library(ChemometricsWithR)
}
## morpheus
#p_load_gh('morpheus')
#if(!require(morpheus))
# devtools::install_github('cmap/morpheus.R')
#p_load(morpheus)
source('src/modT.R')
source('src/helptext.R')
source('src/my_io.R')
source('src/gct-io.R')
source('src/plots.R')
source('src/manage_sessions.R')
#source('src/linear_model.R')
##################################################
## import header descriptions
headerDesc <- read_excel('docs/description-column-headers.xlsx')
## #####################################
## CSS for loading animation
appCSS <- "
#loading-content {
position: absolute;
background: #000000;
opacity: 0.9;
z-index: 100;
left: 0;
right: 0;
height: 100%;
text-align: center;
color: #FFFFFF;
}
"
## #################################################################
##
## color scheme for PPI databases
##
## #################################################################
ppi.db.col <- c('InWeb'='deepskyblue',
'BioGRID'='yellow',
'Reactome'='magenta',
'Shared'='orange'
)
## ##################################################################
##
## import PPi databases
##
## ##################################################################
import.ppi.db <- function(){
if( file.exists('ppi/ppi.RData')){
load('ppi/ppi.RData')
} else {
ppi <- list()
## ###################################
## InWeb
ppi$iw <- readRDS('ppi/core.psimitab.rds')
## uniprot
ppi$iw$V1 <- toupper(sub('uniprotkb\\:', '', ppi$iw$V1))
ppi$iw$V2 <- toupper(sub('uniprotkb\\:', '', ppi$iw$V2))
## gene names
ppi$iw$V5 <- toupper(sub('^uniprotkb\\:(.*?)\\(gene name\\).*', '\\1', ppi$iw$V5))
ppi$iw$V6 <- toupper(sub('^uniprotkb\\:(.*?)\\(gene name\\).*', '\\1', ppi$iw$V6))
## ###################################
## biogrid
ppi$bg <- readRDS('ppi/BIOGRID-HUMAN-3.4.147.mitab.rds')
ppi$bg$Alt.IDs.Interactor.A <- sapply(strsplit(ppi$bg$Alt.IDs.Interactor.A, '\\|'), function(x) sub('.*\\:','',x[2]) )
ppi$bg$Alt.IDs.Interactor.B <- sapply(strsplit(ppi$bg$Alt.IDs.Interactor.B, '\\|'), function(x) sub('.*\\:','',x[2]) )
## ###################################
## Reactome
ppi$react <- readRDS('ppi/homo_sapiens.mitab.interactions.rds')
ppi$react$alternative.id.A <- sub('^.*_(.*?)\\(.*$','\\1', ppi$react$alternative.id.A)
ppi$react$alternative.id.B <- sub('^.*_(.*?)\\(.*$','\\1', ppi$react$alternative.id.B)
# export
save(ppi, file='ppi/ppi.RData')
}
return(ppi)
}
# run the function to import
ppi <- import.ppi.db()
## ###############################################
##
## map uniprot/refseq to gene names
## n.try = number of ids taken from 'ids' to try to
## determine organism
## ###############################################
mapIDs <- function(ids,rdesc=NULL,
n.try=100
){
withProgress(message='Mapping gene names...', {
## ###################################
## id type
## ###################################
keytype <- 'UNKNOWN'
## Uniprot or RefSeq or Ensembl?
if(length(grep('^(Q|P|O|A|E|H|F)', ids)) > 0)
keytype='UNIPROT'
if(length(grep('^(NP_|XP_|YP_)', ids)) > 0)
keytype='REFSEQ'
if(length(grep('ENSP', ids)) > 0)
keytype='ENSEMBLPROT'
## ###################################
## extract query strings
## ###################################
if(keytype == 'UNIPROT' ){
id.query <- sub('(-|;|\\.|_|\\|).*', '', ids) ## first id
} else if(keytype == 'REFSEQ') {
id.query <- sub('(\\.|;).*', '', ids) ## first id
} else if(keytype=='ENSEMBLPROT') {
id.query <- sub('(\\.|;).*', '', ids) ## first id
}else {
id.query <- ids
}
names(id.query) <- ids
## ###################################
## determine organism
## ###################################
orgtype <- 'UNKNOWN'
if(keytype != 'UNKNOWN'){
# try human
if(orgtype == 'UNKNOWN'){
cat('trying human...\n')
id.map.tmp <- try( mapIds(org.Hs.eg.db, keys=id.query[ sample(1:length(id.query), n.try)] , column=c('SYMBOL'), keytype=keytype, multiVals='first') )
if(class(id.map.tmp) != 'try-error'){
orgtype='HSA'
}
}
# try mouse
if(orgtype == 'UNKNOWN'){
cat('trying mouse...\n')
id.map.tmp <- try( mapIds(org.Mm.eg.db, keys=id.query[ sample(1:length(id.query), n.try)] , column=c('SYMBOL'), keytype=keytype, multiVals='first') )
if(class(id.map.tmp) != 'try-error'){
orgtype='MMU'
}
}
# try rat
if(orgtype == 'UNKNOWN'){
cat('trying rat...\n')
id.map.tmp <- try( mapIds(org.Rn.eg.db, keys=id.query[ sample(1:length(id.query), n.try)] , column=c('SYMBOL'), keytype=keytype, multiVals='first') )
if(class(id.map.tmp) != 'try-error'){
orgtype='RNO'
}
}
# try zebrafish
if(orgtype == 'UNKNOWN'){
cat('trying zebrafish...\n')
id.map.tmp <- try( mapIds(org.Dr.eg.db, keys=id.query[ sample(1:length(id.query), n.try)] , column=c('SYMBOL'), keytype=keytype, multiVals='first') )
if(class(id.map.tmp) != 'try-error'){
orgtype='DRE'
}
}
}
## ##################################
## map
#if geneSymbol column is included in rdesc, use that
if(!is.null(rdesc)&("geneSymbol"%in%colnames(rdesc))){
id.map.tmp <- sub('(-|;|\\.|_|\\|).*', '', rdesc$geneSymbol) #take first if there is a list
}else{
if(keytype != 'UNKNOWN' & orgtype != 'UNKNOWN'){
if(orgtype == 'HSA')
id.map.tmp <- try(mapIds(org.Hs.eg.db, keys=id.query , column=c('SYMBOL'), keytype=keytype, multiVals='first'))
if(orgtype == 'MMU')
id.map.tmp <- try(mapIds(org.Mm.eg.db, keys=id.query , column=c('SYMBOL'), keytype=keytype, multiVals='first'))
if(orgtype == 'RNO')
id.map.tmp <- try(mapIds(org.Rn.eg.db, keys=id.query , column=c('SYMBOL'), keytype=keytype, multiVals='first'))
if(orgtype == 'DRE')
id.map.tmp <- try(mapIds(org.Dr.eg.db, keys=id.query , column=c('SYMBOL'), keytype=keytype, multiVals='first'))
} else {
id.map.tmp <- c()
}
}
if(class(id.map.tmp) == 'try-error' | is.null( class(id.map.tmp) ) | class(id.map.tmp) == 'NULL' ){
id.map <- data.frame(id=names(id.query), id.query=id.query, id.mapped=names(id.query), id.concat=ids, stringsAsFactors=F)
#cat('test2\n')
keytype <- 'UNKNOWN'
} else {
## if successful
id.mapped <- id.map.tmp
id.mapped[which(is.na(id.mapped) | id.mapped=="")] <- ids[which(is.na(id.mapped) | id.mapped=="")]
id.map.tmp[which(is.na(id.map.tmp) | id.map.tmp=="")] <- 'NotFound'
id.map <- data.frame(id=names(id.query), id.query=id.query, id.mapped=as.character(id.mapped), id.concat=paste(ids, id.map.tmp, sep='_'), stringsAsFactors=F)
}
## results
res <- list()
res[[1]] <- keytype
res[[2]] <- id.map
res[[3]] <- orgtype
names(res) <- c('keytype', 'id.map', 'orgtype')
})
return(res)
}
############################################################
## given a vector of group names, return all distinct
## pairwise comparison
## mode = 'sort' : group names will be sorted alphanumerically
############################################################
get_pairwise_group_comparisons <- function(groups.unique,
mode=c('sort') ## sort: group names are ordered alphanumerically
## e.g. "Exp_B" and "Exp_A" -> "Exp_A.vs.Exp_B"
## which translates to Exp_B over Exp_A
){
mode <- match.arg(mode)
groups.comp <- c()
count=1
for(i in 1:(length(groups.unique)-1))
for(j in (i+1):length(groups.unique)){
## order alphabetically
if(mode == 'sort')
groups.tmp <- sort( groups.unique[c(i,j)] )
else
groups.tmp <- groups.unique[c(i,j)]
groups.comp[count] <- paste(groups.tmp[1], groups.tmp[2], sep='.vs.')
count <- count+1
}
return(groups.comp)
}
#################################################################################
##
## calculate average abundance per group groups
##
#################################################################################
calculate_fc <- function(tab, grp.vec, groups.comp, test,
mode='sort', ## for assigning pairwise comparisons
## sort: group names are ordered alphanumerically
## e.g. "Exp_B" and "Exp_A" -> "Exp_A.vs.Exp_B",
center=F ## if TRUE median centering will be applied
){
#browser()
groups <- unique(grp.vec)
## ##########################################
## average groups
group_avg <- lapply(groups, function(gg){
gg.idx = names(grp.vec)[ which(grp.vec == gg) ]
apply(tab[, gg.idx], 1, mean, na.rm=T)
})
group_avg <- data.frame(Reduce('cbind', group_avg))
colnames(group_avg) <- groups
## ##########################################
## one sample & mod F: just average the log values
if(test %in% c("One-sample mod T", "none", "mod F")){
group_fc <- group_avg
}
## ##########################################
## moderated F: average log values and subtract
## groups as specified by the user
#if(test == "mod F"){
# groups.comp <- get_pairwise_group_comparisons(groups, mode=mode)
#}
## ##########################################
## two sample: subtract averaged groups
if(test %in% c("Two-sample mod T")){
groups.comp.split <- strsplit(groups.comp, split = '\\.vs\\.')
names(groups.comp.split) <- groups.comp
## calculate log FCs
group_fc <- lapply(groups.comp.split, function(x){
g1 <- x[1]
g2 <- x[2]
## subtract logs
g2.over.g1 <- group_avg[, g2] - group_avg[, g1]
g2.over.g1
})
group_fc <- data.frame(Reduce('cbind', group_fc))
colnames(group_fc) <- groups.comp
}
## median center data
if(center)
group_fc <- apply(group_fc, 2, function(column) column - median(column, na.rm=T))
## for one-sample/mod F/none, it is average, not FC
if(test %in% c("One-sample mod T", "none", "mod F")){
colnames(group_fc) <- paste0('RawAveExpr.', colnames(group_fc))
}else{
colnames(group_fc) <- paste0('RawlogFC.', colnames(group_fc))
}
return(group_fc)
}
#################################################################################
##
## - deduplicate ids and show warning in a modal shiny window
##
#################################################################################
de_duplicate_ids <- function(ids, global.param=NULL, show_modal = TRUE){
if(is.null(global.param)) global.param$id.col.value <- 'id'
## dups
dup_idx <- which(duplicated(ids))
dup_n <- length(dup_idx)
dup_ids <- ids[ dup_idx ]
## make dups unique
ids <- make.unique(ids , sep='_')
if(show_modal){
require(shiny)
## dups after making unique
dup_ids_fix <- ids[dup_idx]
## show the first 3 dups in a table
html_tab <- paste0('<table border="1">
<th style="padding:10px">duplciated</th><th style="padding:10px">deduplicated</th>')
for(tr in 1:min(c(3, dup_n))){
html_tab <- paste0(html_tab,
'<tr><td style="padding:10px">', dup_ids[tr], '</td><td style="padding:10px">', dup_ids_fix[tr], '</td></tr>\n')
}
html_tab <- paste0(html_tab, '</table>')
#######################
## give a warning about duplicated ids
showModal(modalDialog(
size='m',
title = paste("Warning: ids are not unique!"),
HTML(paste0('Found <b>', dup_n, '</b> duplicated entries in column <b>',
global.param$id.col.value,
'</b> which will be made unique as shown below (first ', min(c(3, dup_n)), ' duplicated ids shown).<br><br>')),
HTML(html_tab),
HTML('<br><br>Click "Dismiss" to continue.')
))
} ## end if(show.modal)
return(ids)
}
## ##############################################################################
##
## generate links to external databases
##
## ##############################################################################
link.db <- function(id, # vetcor of ids
keytype=c('UNKNOWN', 'UNIPROT', 'REFSEQ','ENSEMBLPROT'),
db=c('GENECARDS', 'UNIPROT')){
keytype <- match.arg(keytype)
db <- match.arg(db)
if(keytype == 'UNIPROT'){
up.link <- paste("<a href='https://www.uniprot.org/uniprot/", sub('(_|,|;|\\.).*', '', id),"' target='_blank'>", id, "</a>", sep='')
}
if(keytype %in% c('REFSEQ', 'ENSEMBLPROT','UNKNOWN')){
up.link <- paste("<a href='http://www.genecards.org/Search/Keyword?queryString=", sub('^(NP_|NM_|NR_.*?)(_|,|;|\\.).*', '\\1', id),"' target='_blank'>", id, "</a>", sep='')
}
return(up.link)
}
#############################################################################################
normalize.data <- function(data, id.col,
method=c('Median',
'Median (non-zero)',
'Quantile',
'VSN',
'Median-MAD',
'Median-MAD (non-zero)',
'2-component',
'Upper-quartile'),
grp.vec=NULL ## if NULL apply global normalization strategy
## if vector of group assingments, normalization will be applied to each group separately
){
#################################
## global normalization
if(is.null(grp.vec)){
data.norm <- normalize.data.helper(data, id.col, method=match.arg(method))
################################
## group-specific normalization
} else {
## extract groups
groups <- unique(grp.vec)
############################################
## loop over replicate groups
for(gg in groups){
gg.idx = names(grp.vec)[ which(grp.vec == gg) ]
data.group <- data[, c(id.col, gg.idx)]
data.group.norm <- normalize.data.helper(data.group, id.col, method=match.arg(method), per_group = TRUE)
if(gg == groups[1]){
data.norm <- data.group.norm
} else {
data.norm <- cbind(data.norm, data.group.norm[, gg.idx])
}
} ## end for
} ## end else
return(data.norm)
}
#############################################################################################
##
## different normalization methods for expression data
##
## 20160235
#############################################################################################
normalize.data.helper <- function(data, id.col,
method=c('Median',
'Median (non-zero)',
'Quantile',
'VSN',
'Median-MAD',
'Median-MAD (non-zero)',
'2-component',
'Upper-quartile'),
per_group=FALSE ## for Median & Median-MAD
){
cat('\n\n-- normalize data --\n\n')
method = match.arg(method)
cat(' normalization method: ', method, '\n')
ids = data[, id.col]
data = data[ , -grep(paste('^', id.col, '$', sep=''), colnames(data))]
data <- data.matrix(data)
## quantile
if(method == 'Quantile'){
p_load("preprocessCore")
data.norm <- normalize.quantiles(data)
rownames(data.norm) <- rownames(data)
colnames(data.norm) <- paste( colnames(data))
## shift median to zero
## data.norm <- apply(data.norm, 2, function(x) x - median(x, na.rm=T))
}
## median only
if(method == 'Median'){
data.norm <- apply(data, 2, function(x) x - median(x, na.rm=T))
colnames(data.norm) <- paste( colnames(data), sep='.')
if(per_group){
all_medians <- apply(data, 2, median, na.rm=T)
data.norm <- data.norm + median( all_medians, na.rm=T)
}
}
## median plus shifting by medians of medians
if(method == 'Median (non-zero)'){
all_medians <- apply(data, 2, median, na.rm=T)
data.norm <- apply(data, 2, function(x) x - median(x, na.rm=T))
data.norm <- data.norm + median( all_medians, na.rm=T )
colnames(data.norm) <- paste( colnames(data), sep='.')
}
## median & MAD
if(method == 'Median-MAD'){
data.norm <- apply(data, 2, function(x) (x - median(x, na.rm=T))/mad(x, na.rm=T) )
colnames(data.norm) <- paste( colnames(data), sep='.')
if(per_group){
all_medians <- apply(data, 2, median, na.rm=T)
data.norm <- data.norm + median( all_medians, na.rm=T)
}
}
## median & MAD plus shifting by medians of medians
if(method == 'Median-MAD (non-zero)'){
all_medians <- apply(data, 2, median, na.rm=T)
data.norm <- apply(data, 2, function(x) (x - median(x, na.rm=T))/mad(x, na.rm=T) )
data.norm <- data.norm + median( all_medians, na.rm=T )
colnames(data.norm) <- paste( colnames(data), sep='.')
}
## 2-component normalization
if(method == '2-component'){
data.norm.list <- vector('list', ncol(data))
names(data.norm.list) <- colnames(data)
for(x in colnames(data)){
res <- try(two.comp.normalize(data[, x], type="unimodal"))
data.norm.list[[x]] <- res
if(class(res) == 'try-error') break;
}
## check if all runs were successful
## return the 'try-error' object to
## catch the error in server.R
for(i in 1:length(data.norm.list)){
if(class(data.norm.list[[i]]) == 'try-error'){
msg <- data.norm.list[[i]]
return(msg)
}
}
## if 2-comp was successful on all data column convert list to matrix
data.norm = matrix( unlist(lapply(data.norm.list, function(x)x$norm.sample)), ncol=length(data.norm.list), dimnames=list(rownames(data), names(data.norm.list)) )
}
## Upper quartile
if(method == 'Upper-quartile'){
data.norm <- apply(data, 2, function(x) x - quantile(x, c(0.75),na.rm=T))
colnames(data.norm) <- paste( colnames(data), sep='.')
}
## VSN - variance stabilizing normalization
if(method == 'VSN'){
p_load(vsn)
data.norm <- justvsn(data)
}
## add id column
data.norm <- data.frame(ids, data.norm)
colnames(data.norm)[1] <- id.col
cat('\n\n-- normalize data exit--\n\n')
return(data.norm)
}
##############################################################################
##
## - perform principle component analysis
## - calculate variances explained by components
## - plot the results
##
## changelog: 20131001 implementation
## 20131007 3D plot now plot pc1 vs. pc3 vs. pc2
## instead of pc1 vs. pc2 vs. pc3
## 20151208 legend
## 20161103 check number of rows (N), 2D plot only
##############################################################################
my.prcomp.static <- function(x, pca.x, pca.y, pca.z, col=NULL, cor=T, plot=T, rgl=F, scale=T, pch=20, cex.points=3, rgl.point.size=30, main="PCA", leg.vec=NULL, leg.col=NULL, ...){
cex.font = 1.8
## number of data columns, N=2 -> 2D plot only
N <- nrow(x)
## color
if( is.null(col) ) col="black"
## perform pca
pca <- prcomp(x, scale=scale)
## calculate variance
comp.var <- eigen(cov(pca$x))$values
## extract the principle components
pc1=pca$x[,1]
pc2=pca$x[,2]
if(N>2)
pc3=pca$x[,3]
##############
# rgl plot
##############
if(rgl & N > 2){
p_load(rgl)
plot3d(pc1, pc2, pc3, xlab=paste("PC 1 (", round(100*comp.var[1]/sum(comp.var),1),"%)", sep=""), ylab=paste("PC 2 (",round(100*comp.var[2]/sum(comp.var),1),"%)", sep=""), zlab=paste("PC 3 (", round(100*comp.var[3]/sum(comp.var),1),"%)", sep=""), type="s", col=col, expand=1.2, size=rgl.point.size)
}
########################################
# scatterplot 2D/3D
########################################
if( plot){
p_load(scatterplot3d)
if(N > 2)
par(mfrow=c(1,3), mar=c(7,7,3,1))
if(N <= 2)
par(mfrow=c(1,2), mar=c(7,7,3,1))
## PC 1-2
plot(pc1, pc2, xlab=paste("PC 1 (", round(100*comp.var[1]/sum(comp.var),1),"%)", sep=""), ylab=paste("PC 2 (",round(100*comp.var[2]/sum(comp.var),1),"%)", sep=""), pch=pch, main=main, col=col, sub=paste("Cumulative variance = ", round(100*sum(comp.var[1:2]/sum(comp.var)),1),"%", sep=""), cex=cex.points, ylim=c( min(pc2), max(pc2)+.15*max(pc2)), cex.axis=cex.font, cex.lab=cex.font, cex.sub=cex.font )
if(N > 2) {
## PC 1-3
scatterplot3d( pca$x[,1], pca$x[,3], pca$x[,2], xlab=paste("PC 1 (", round(100*comp.var[1]/sum(comp.var),1),"%)", sep=""), zlab=paste("PC 2 (",round(100*comp.var[2]/sum(comp.var),1),"%)", sep=""), ylab=paste("PC 3 (", round(100*comp.var[3]/sum(comp.var),1),"%)", sep=""), color=col, cex.symbols=cex.points, pch=pch, main=main, sub=paste("Cumulative variance = ", round(100*sum(comp.var[1:3]/sum(comp.var)),1),"%", sep=""), type="h" )
}
## legend
plot.new()
plot.window(xlim=c(0,1), ylim=c(0, 1))
if(!is.null(leg.vec) & !is.null(leg.col))
legend('topleft', legend=leg.vec, col=leg.col, pch=pch, pt.cex=max(1, cex.points-1.5), ncol=ifelse( length(leg.vec)> 10, 2, 1), bty='n', cex=2 )
par(mfrow=c(1,1))
}
return(pca)
}
#####################################################
##
## calculate PCs and their variance variance
##
## res - results after testing
## grp - class vector, names are column names of input matrix
#####################################################
my.prcomp2 <- function(res, grp){
## remove missing values
rm.idx <- apply(res, 1, function(x) sum(is.na(x)) + sum(is.infinite(x)))
rm.idx <- which(rm.idx > 0)
if(length(rm.idx)>0) res <- res[-rm.idx, ]
## extract expression data
res = res[, names(grp)]
## perform pca
pca <- PCA(scale(t(res)))
return(pca)
}
#
# #####################################################
# ##
# ##
# ##
# #####################################################
# plotPCAloadings <- function(pca, topn, pca.x, pca.y, pca.z){
#
# ##load = loadings(pca)[, c(pc1, pc2, pc3)]
#
# ## ###################
# ## extract loadings
# load.pca.x <- pca$loadings[, pca.x]
# load.pca.y <- pca$loadings[, pca.y]
# load.pca.z <- pca$loadings[, pca.z]
#
# n=length(load.pca.x)
#
# ## ###################
# ## choose top N
# x <- rev(sort(abs( load.pca.x ), decreasing=T )[1:min(topn, n)])
# y <- rev(sort(abs( load.pca.y ), decreasing=T )[1:min(topn, n)])
# z <- rev(sort(abs( load.pca.z ), decreasing=T )[1:min(topn, n)])
#
#
# ## ###################################################
# ## base plotting system
# par(mfrow=c(1,3))
# barplot(x, horiz=T, main=paste('PC', pca.x), las=2, border='blue', space=0, col='grey95', ylab='Features', xlab='Absolute coefficient', names.arg=rev(1:length(x)))
# text(rep(0, length(x)), 1:length(x)-.5, labels=names(x), pos=4)
#
# barplot(y, horiz=T, main=paste('PC', pca.y), las=2, border='blue', space=0, col='grey95', axisnames=T, xlab='Absolute coefficient',names.arg=rev(1:length(x)))
# text(rep(0, length(y)), 1:(length(y))-.5, labels=names(y), pos=4)
#
# barplot(z, horiz=T, main=paste('PC', pca.z), las=2, border='blue', space=0, col='grey95', axisnames=T, xlab='Absolute coefficient', names.arg=rev(1:length(x)))
# text(rep(0, length(z)), 1:(length(z))-.5, labels=names(z), pos=4)
#
# }
#
# #####################################################
# ##
# ## scatterPlotPCAloadings
# ##
# #####################################################
#
# scatterPlotPCAloadings <- function(pca, topn, pca.x, pca.y, pca.z){
#
# ## extract loadings
# load.pca.x <- pca$loadings[, pca.x]
# load.pca.y <- pca$loadings[, pca.y]
# load.pca.z <- pca$loadings[, pca.z]
#
# n=length(load.pca.x)
#
# ## ###################
# ## choose top N
# x <- rev(sort(abs( load.pca.x ), decreasing=T )[1:min(topn, n)])
# y <- rev(sort(abs( load.pca.y ), decreasing=T )[1:min(topn, n)])
# z <- rev(sort(abs( load.pca.z ), decreasing=T )[1:min(topn, n)])
#
#
# tmp.pcaloadings <- as.data.frame(pca$loadings[,c(pca.x,pca.y,pca.z)])
# names(tmp.pcaloadings) <- make.names(names(tmp.pcaloadings))
#
# PC1 <- names(tmp.pcaloadings)[pca.x]
# PC2 <- names(tmp.pcaloadings)[pca.y]
# PC3 <- names(tmp.pcaloadings)[pca.z]
#
# my.scatter <- function(datafr,xa,ya,topx,topy){
#
# #scatterplot dimensions pairwise (pairs passed to function as x,y)
# x.axis <- xa ; y.axis <- ya ;topxy <- c(names(topx),names(topy))
# # This is the data.frame of the TopN loadings to be marked
# mark.frame <- tmp.pcaloadings[which(row.names(tmp.pcaloadings) %in% topxy),]
#
# xmin = min(datafr[,x.axis]);xmax = max(datafr[,x.axis])
# ymin = min(datafr[,y.axis]);ymax = max(datafr[,y.axis])
#
# #make the scatterplot
# ggplot(data = datafr,aes_string(x = x.axis,y = y.axis))+
# geom_hline(yintercept = 0, linetype = "dashed",size =0.7)+