-
Notifications
You must be signed in to change notification settings - Fork 8
/
trajectoryAnalysis.R
651 lines (509 loc) · 30.2 KB
/
trajectoryAnalysis.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
# script to perform trajectory analysis
# setwd("~/KP/singleCellProjects/multiomeProject_MatkarS")
library(tidyverse)
library(monocle)
library(Seurat)
library(devtools)
library(roxygen2)
# install_github('hb-gitified/cellrangerRkit', user = 'kpatel427',
# auth_token = '' )
library(cellrangerRkit)
#BiocManager::install("slingshot")
library(slingshot)
library(RColorBrewer)
library(scales)
library(pals)
library(grDevices)
library(ggbeeswarm)
library(tradeSeq)
library(pheatmap)
library(reactome.db)
library(org.Hs.eg.db)
library(fgsea)
# reading in split_seurat after SCTransform ------------------
split_seurat <- readRDS("data/split_seurat.rds")
FelixLRX1_Multiome <- split_seurat$FelixLRX1_Multiome
FelixLRX2_Multiome <- split_seurat$FelixLRX2_Multiome
FelixParental1_Multiome <- split_seurat$FelixParental1_Multiome
FelixParental2_Multiome <- split_seurat$FelixParental2_Multiome
# merging seurat object post SCTransormation into one object
merged_seurat <- merge(x = FelixLRX1_Multiome,
y = c(FelixLRX2_Multiome, FelixParental1_Multiome, FelixParental2_Multiome),
add.cell.id = c("LRX1", "LRX2", "parent1", "parent2"))
saveRDS(merged_seurat, file = paste0('data/',Sys.Date(),'_beforeIntegration.rds'))
View([email protected])
# just a quick check!
# getting variable features from SCT assay
VariableFeatures(merged_seurat[["SCT"]]) <- rownames(merged_seurat[["SCT"]]@scale.data)
merged_seurat <- RunPCA(merged_seurat, verbose = FALSE)
merged_seurat <- RunUMAP(merged_seurat, dims = 1:30, verbose = FALSE)
merged_seurat <- FindNeighbors(merged_seurat, dims = 1:30, verbose = FALSE)
merged_seurat <- FindClusters(merged_seurat, verbose = FALSE)
merged_seurat$sample <- gsub('PDX','LRX', merged_seurat$sample)
a1 <- DimPlot(merged_seurat, label = TRUE)
a2 <- DimPlot(merged_seurat,
label = FALSE,
group.by = 'sample')
a3 <- DimPlot(merged_seurat,
label = FALSE,
group.by = 'seq_folder')
plot <- plot_grid(a1, a2, a3, labels=c("Cluster all cells", "Cluster by Source", "Cluster by Samples"), ncol = 2, nrow = 2)
ggsave(plot, filename = paste0('figures/',Sys.Date(),'_postSCTransform_preIntegration_clustering.pdf'), width = 15, height = 10)
saveRDS(merged_seurat, file = paste0('data/',Sys.Date(), '_postSCTransform_beforeIntegration.rds'))
# reading in seurat objects pre and post SCTransform ---------------
post.sct <- readRDS('~/KP/singleCellProjects/multiomeProject_MatkarS/data/2021-10-11_postSCTransform_beforeIntegration.rds')
pre.sct <- readRDS('~/KP/singleCellProjects/multiomeProject_MatkarS/data/2021-10-11_preSCTransform_beforeIntegration.rds')
# getting overlapping ADRN and MES genes
adrn.mes.sig <- read.delim('~/KP/singleCellProjects/PRAME_ARDN_MES/ADRN_MES_signature_genes_list.txt', header = F)
mesench <- data.frame('V1'=adrn.mes.sig[which(adrn.mes.sig[,2] == "MES"),1])
adrenal <- data.frame('V1'=adrn.mes.sig[which(adrn.mes.sig[,2] == "ADRN"),1])
mesench$group <- 'MES'
adrenal$group <- 'ADRN'
ADRN.overlap <- unique(intersect(adrenal$V1, VariableFeatures(pre.sct)))
MES.overlap <- unique(intersect(mesench$V1, VariableFeatures(pre.sct)))
# starting with pre-SCT data
class(pre.sct)
head(rownames(pre.sct))
head(colnames(pre.sct))
# visualizations
t <- ggplot([email protected], aes(nCount_RNA, nFeature_RNA, color = seq_folder)) +
geom_point(size = 0.5) +
scale_color_brewer(type = "qual", palette = "Set2", name = "cell type") +
scale_x_log10() +
scale_y_log10() +
theme_bw() +
# Make points larger in legend
guides(color = guide_legend(override.aes = list(size = 3))) +
labs(x = "Total UMI counts", y = "Number of genes detected")
ggsave(t, filename = 'figures/test.pdf', width = 15, height = 10)
# looking at PCs
e <- ElbowPlot(pre.sct, ndims = 70)
ggsave(e, filename = 'figures/elbowplot.pdf', width = 15, height = 10)
# looking at reduced dimensions and clustering
d <- DimPlot(pre.sct, reduction = "umap",
group.by = "seq_folder", pt.size = 0.5, label = TRUE, repel = TRUE) +
scale_color_brewer(type = "qual", palette = "Set2")
ggsave(d, filename = 'figures/dimplot.pdf', width = 15, height = 10)
# Slingshot -----------------------------------
# convert seurat to singleCellExperiment obj
pre.sct.sc <- as.SingleCellExperiment(pre.sct)
post.sct.sc <- as.SingleCellExperiment(post.sct)
# running slingshot ------------------------------
# ....using seurat object to calculate lineages -------------------
pre.sct.ss <- slingshot(Embeddings(pre.sct, "umap"), clusterLabels = pre.sct$seurat_clusters,
start.clus = 4, stretch = 0)
post.sct.ss <- slingshot(Embeddings(post.sct, "umap"), clusterLabels = post.sct$seurat_clusters,
start.clus = 10, stretch = 0, omega = TRUE)
# to test whether we have disjoint trajectories - no disjoint trajectories detected
# to give a different start cluster - 7: Still does not change the lineage structure
pre.sct.ss.test <- slingshot(Embeddings(pre.sct, "umap"), clusterLabels = pre.sct$seurat_clusters,
start.clus = 7, stretch = 0, omega = TRUE)
# using singleCellExperiment object
#sce <- slingshot(pre.sct.sc, clusterLabels = pre.sct.sc$seurat_clusters, reducedDim = 'PCA')
# assign colors to cells
cell_pal <- function(cell_vars, pal_fun,...) {
if (is.numeric(cell_vars)) {
pal <- pal_fun(100, ...)
return(pal[cut(cell_vars, breaks = 100)])
} else {
categories <- sort(unique(cell_vars))
pal <- setNames(pal_fun(length(categories), ...), categories)
return(pal[cell_vars])
}
}
# pre.SCT
cell_colors <- cell_pal(pre.sct$seq_folder, brewer_pal("qual", "Set2"))
cell_colors_clust <- cell_pal(pre.sct$seurat_clusters, hue_pal())
# post.SCT
cell_colors <- cell_pal(post.sct$seq_folder, brewer_pal("qual", "Set2"))
cell_colors_clust <- cell_pal(post.sct$seurat_clusters, hue_pal())
# ...plotting trajectory ------------------
# for seurat object
plot(Embeddings(pre.sct, "umap"), col = cell_colors, pch = 16, cex = 0.5)
pdf(file = 'figures/trajectoryAnalysis_slingshot_trajectories_clusters.pdf', width = 10, height = 10)
plot(Embeddings(pre.sct, "umap"), col = cell_colors_clust, pch = 16, cex = 0.5)
lines(SlingshotDataSet(pre.sct.ss), lwd=2, col='black', type = "lineages", show.constraints = TRUE)
dev.off()
# to get pseudotime for each cell across lineages
slingPseudotime(post.sct.ss, na = TRUE)
# ....using SCE object -------------------
#sce <- slingshot(pre.sct.sc, clusterLabels = pre.sct.sc$seurat_clusters, reducedDim = 'umap', approx_points = 200, start.clus = 4) # step was run on test.R and sbatched using qsub-test.R script
load('data/slingobj.RData')
pre.sce <- sce
load('data/slingobj_postSCT.RData')
post.sce <- sce
summary(sce$slingPseudotime_1)
str(sce)
# ....to visualize inferred lineage for dta with points colored by pseudotime ---------------
colors <- colorRampPalette(brewer.pal(11,'Spectral')[-6])(100)
plotcol <- colors[cut(sce$slingPseudotime_1, breaks=100)]
plot.new()
pdf('figures/trajectoryAnalysis_slingshot_SCEobj_UMAP.pdf', width = 10, height = 10)
plot(reducedDims(sce)$UMAP, col = plotcol, pch=16, asp = 1)
lines(SlingshotDataSet(sce), lwd=2, col='black', type = 'lineages', show.constraints = TRUE)
dev.off()
# ....plotting pseudotime variable (one per lineage) for cells in all clusters ---------------
time <- as.data.frame(as.matrix(sce@colData$slingPseudotime_3))
cluster.info <- as.data.frame(as.matrix(sce@colData$seurat_clusters))
samples <- as.data.frame(as.matrix(sce@colData$seq_folder))
slingshot_df <- cbind(slingPseudotime_1 = time,
seurat_clusters = cluster.info,
seq_folder = samples)
names(slingshot_df) <- c('slingPseudotime_3', 'seurat_clusters', 'seq_folder')
plot.new()
pdf('figures/slingshot_orderedby_pseudotime3_postSCT.pdf', width = 10, height = 10)
ggplot(slingshot_df, aes(x = slingPseudotime_3, y = seurat_clusters,
colour = seq_folder)) +
geom_quasirandom(groupOnX = FALSE) +
theme_classic() +
xlab("Third Slingshot pseudotime") +
ylab("cluster number") +
ggtitle("Cells ordered by Slingshot pseudotime")
dev.off()
# ....identifying temporally dynamic genes --------------------------------
# finding genes that change their expression over the course of development.
#library(tradeSeq)
# lin <- getLineages(pre.sct.sc, clusterLabels = pre.sct.sc$seurat_clusters, start.clus = 4, reducedDim = "UMAP")
# crv <- getCurves(lin)
# a2 <- plotGeneCount(curve = crv, clusters = pre.sct.sc$seurat_clusters)
# ggsave(a2, filename = 'figures/trajectoryCurves_tradeSeq_preSCT_UMAP.pdf', width = 10, height = 10)
#
# # determining number of knots
# icMat <- evaluateK(counts = counts(sce), k=3:10, nGenes = 200,
# pseudotime = slingPseudotime(crv, na = FALSE),
# cellWeights = slingCurveWeights(crv))
#
# ggsave(icMat, filename = 'figures/PseudotimeAnalysis_determineKnots_pre.SCT_UMAP.pdf', width = 10, height = 10)
#
# # 1. fit negative binomial GAM --------
# # since fitGAM runtimes are long, run it on most variable features from seurat object
# hvg <- VariableFeatures(pre.sct)
# length(rownames(sce)[rownames(sce) %in% hvg])
#
# # to be fed into genes parameter
# idx <- which(rownames(sce) %in% hvg)
#
# # following lines of code were run on a separate script on a scheduler (test2.R)
# # fitting GAM on top variable genes from seurat object
# # ordering rownames in sce as per variableFeatures in SCE
# sce <- sce[order(match(rownames(sce),hvg)),]
#
# sce <- fitGAM(sce, nknots = 6, genes = 1:10)
#
# # 2. test for dynamic expression -----------
# ATres <- associationTest(sce, lineages = TRUE)
# steps above are run in separate scripts - test2.R, test3.R and test4.R
# ____1. Pre.SCT per lineage ------------------
load('data/associateTestResult_perLineages_k6_preSCT_UMAP.RData') # pre.sct, per lineage
# We want to inspect two lineages i.e. lineage 1 and 2
# pre.sct.ss@metadata$lineages
# pre.sct.ss is a "PseudotimeOrdering" object i.e. it has lineage information
lineage1.DE.genes <- rownames(ATres)[which(p.adjust(ATres$pvalue_1, "fdr") <= 0.05)]
length(unique(intersect(lineage1.DE.genes, ADRN.overlap)))
length(unique(intersect(lineage1.DE.genes, MES.overlap)))
# only parental cluster
lineage6.DE.genes <- rownames(ATres)[which(p.adjust(ATres$pvalue_6, "fdr") <= 0.05)]
length(unique(intersect(lineage6.DE.genes, ADRN.overlap)))
length(unique(intersect(lineage6.DE.genes, MES.overlap)))
# ____2. Pre.SCT global ------------------
load('data/associateTestResult_global_k6_preSCT_UMAP.RData') # pre.sct global
global.DE.genes <- rownames(ATres)[which(p.adjust(ATres$pvalue, "fdr") <= 0.05)]
length(unique(intersect(global.DE.genes, ADRN.overlap)))
length(unique(intersect(global.DE.genes, MES.overlap)))
# ____3. Post.SCT per lineage ------------------
load('data/associateTestResult_perLineages_k6_postSCT_UMAP.RData') # post.sct, per lineage
# We want to inspect two lineages i.e. lineage 1 and 2
# pre.sct.ss@metadata$lineages
# pre.sct.ss is a "PseudotimeOrdering" object i.e. it has lineage information
lineage1.DE.genes <- rownames(ATres)[which(p.adjust(ATres$pvalue_1, "fdr") <= 0.05)]
length(unique(intersect(lineage1.DE.genes, ADRN.overlap)))
length(unique(intersect(lineage1.DE.genes, MES.overlap)))
# only parental cluster
lineage6.DE.genes <- rownames(ATres)[which(p.adjust(ATres$pvalue_6, "fdr") <= 0.05)]
length(unique(intersect(lineage6.DE.genes, ADRN.overlap)))
length(unique(intersect(lineage6.DE.genes, MES.overlap)))
# ____4. Post.SCT global ------------------
load('data/associateTestResult_global_k6_postSCT_UMAP.RData') # post.sct global
global.DE.genes <- rownames(ATres)[which(p.adjust(ATres$pvalue, "fdr") <= 0.05)]
length(unique(intersect(global.DE.genes, ADRN.overlap)))
length(unique(intersect(global.DE.genes, MES.overlap)))
# 3. plotting a heatmap to visualize expression of DE genes across lineages ------------
#topgenes <- rownames(ATres[order(ATres$pvalue_1), ])[1:200]
topgenes <- rownames(ATres[order(ATres$pvalue), ])[1:200]
pst.ord <- order(sce$slingPseudotime_1, na.last = NA)
heatdata <- assays(sce)$logcounts[topgenes, pst.ord]
#write.table(topgenes, file = 'data/topgenes_assoTest_preSCT.txt', col.names = F, row.names = F, quote = F)
col_anno <- colData(sce)[,c(1,12)]
col_anno.ordered <- col_anno[colnames(heatdata),]
any(rownames(col_anno.ordered) == colnames(heatdata))
col_anno.ordered <- as.data.frame(col_anno.ordered)
# setting annotation colors
n <- 18
qual_col_pals = brewer.pal.info[brewer.pal.info$category == 'qual',]
col_vector = unlist(mapply(brewer.pal, qual_col_pals$maxcolors, rownames(qual_col_pals)))
col=sample(col_vector, n)
cluster_cols <- setNames(col, unique(col_anno.ordered$seurat_clusters))
anno_colors <- list(
orig.ident = c('parent2' = '#6D9886', 'parent1' = 'royalblue', 'LRX1' = '#FF7777', 'LRX2' = '#90AACB'))
# plot heatmap
p <- pheatmap(heatdata,
annotation_col = col_anno.ordered,
cluster_cols = F,
show_colnames = F,
fontsize_row = 3,
annotation_colors = anno_colors)
#ggsave(p, filename = paste0('figures/',Sys.Date(),'_heatmap_pseudotime1_200genes_preSCT.pdf'), width = 10, height = 10)
ggsave(p, filename = paste0('figures/',Sys.Date(),'_heatmap_global_pseudotime_200genes_preSCT.pdf'), width = 10, height = 10)
rownames(heatdata[p$tree_row[["order"]],])
gene.clusters <- c("IQCJ-SCHIP1","FGF13","KCNIP4","PCDH7","AL163541.1","XYLT1","FMN1","CREB5","PTPRG","PTPRM","PTCHD1-AS","PRKN","ROBO2","MT-ND3","CCSER1","CLSTN2","AC073050.1","SLC8A1","SGCZ","ROBO1","MAST4","CHGB","ZNF804A","AC069277.1")
intersect(rownames(heatdata[p$tree_row[["order"]],]),mesench$V1)
intersect(rownames(heatdata[p$tree_row[["order"]],]),adrenal$V1)
wnt.signaling <- c("APC","RHOA","ATF4","AXIN1","BCL9","BIRC5","BTRC","CBY1","GCFC2","CAMK2A","CAMK2B","CAMK2D","CAMK2G","CCND1","CDH1","CDH10","CDH11","CDH12","CDH13","CDH15","CDH16","CDH17","CDH18","CDH19","CDH2","CDH20","CDH22","CDH23","CDH24","CDH26","CDH3","CDH4","CDH5","CDH6","CDH7","CDH8","CDH9","CEBPB","CER1","CHUK","CLDN1","CREB1","CREB3","CREB3L4","CREBBP","CSNK1A1","CSNK1D","CSNK1E","CSNK1G1","CSNK1G2","CSNK1G3","CTNNA1","CTNNB1","DAAM1","DAAM2","DKK1","DKK2","DKK4","DVL1","ENC1","FOSL1","FRZB","FZD1","FZD10","FZD2","FZD3","FZD4","FZD5","FZD6","FZD7","FZD8","FZD9","GAST","GNA11","GNA12","GNA13","GNA14","GNA15","GNAI1","GNAI2","GNAI3","GNAL","GNAO1","GNAS","GNAT1","GNAT2","GNAZ","GNB1","GNB2","GNB3","GNB4","GNB5","GNG10","GNG11","GNG12","GNG13","GNG2","GNG3","GNG4","GNG5","GNG7","GNG8","GNGT1","GNGT2","GSK3B","HNF4A","JUN","KREMEN1","KREMEN2","LEF1","LRP5","LRP6","MAP2K4","MAP2K7","MAP3K1","MAP3K10","MAP3K11","MAP3K12","MAP3K13","MAP3K14","MAP3K2","MAP3K3","MAP3K4","MAP3K5","MAP3K6","MAP3K7","MAP3K8","MAP3K9","MAPK10","MAPK8","MAPK9","MMP7","MYC","NFAT5","NFATC1","NFATC2","NFATC3","NFATC4","NFE2L1","NRCAM","PSMD6","DCHS1","PLAUR","PLCB1","PLCB2","PLCB3","PLCB4","PLCD1","PLCD3","PLCD4","PLCE1","PLCG1","PLCG2","PLCZ1","PPARD","PPP3CA","PPP3CB","PPP3CC","PPP3R1","PPP3R2","PRKCA","PRKCB","PRKCD","PRKCE","PRKCG","PRKCH","PRKCI","PRKD1","PRKD3","PRKCQ","PRKCZ","PSMA1","PSMA2","PSMA3","PSMA4","PSMA5","PSMA6","PSMA7","PSMB1","PSMB10","PSMB2","PSMB3","PSMB4","PSMB5","PSMB6","PSMB7","PSMB8","PSMB9","PSMC1","PSMC2","PSMC3","PSMC4","PSMC5","PSMC6","PSMD1","PSMD2","PSMD3","PSMD4","PSMD5","PSMD7","PTGS2","PYGO1","PYGO2","RAC1","REST","ROCK2","SMARCA4","HNF1A","TCF12","TCF15","TCF19","HNF1B","TCF20","TCF21","TCF23","TCF3","TCF4","TCF7","TCF7L2","ZEB1","TEAD1","TEAD4","TFAM","TP53","UBB","UBC","UBD","UBE2D2","VIM","WIF1","WNT1","WNT10A","WNT10B","WNT11","WNT9A","WNT9B","WNT16","WNT2","WNT2B","WNT3","WNT3A","WNT4","WNT5A","WNT5B","WNT6","WNT7A","WNT7B","WNT8A","WNT8B","ZNF354A")
wnt <- intersect(topgenes, wnt.signaling)
# canon.wnt.signaling <- c("CDH15","CDH13","CDH12","CDH11","CDH1","DKK4","DKK3","DKK2","DKK1","LEF1","TCF4","TCF25","TCF21","CDH16","CDH17","CDH18","CDH19","CDH2","CDH24","CDH26","CDH3","CDH4","CDH5","CDH6","CDH8","CTNNB1","CSNK1A1","GSK3B","ANAPC10","ANAPC11","ANAPC13","ANAPC15","ANAPC16","ANAPC2","ANAPC5","WNT1","WNT10A","WNT11","WNT16","WNT2","WNT3","WNT3A","WNT4","WNT5A","WNT5B","WNT6","WNT7A","WNT7B","WNT9A")
# unique(intersect(canon.wnt.signaling, global.topgenes))
#
#
# wnt.hedgehog.notch <- c("AFP","ACVR1","ALPL","RUNX3","RIPK4","APC","ARRB1","ARRB2","AXIN1","AXIN2","BAG6","BCL9","BCL9L","TGFBI","BMP4","BMP6","BMP7","BMPR2","BRD2","CACYBP","PROM1","CDC73","CDCP1","CDX2","CELSR2","CSNK1A1","CSNK1D","CSNK1E","CCN2","CTNNB1","CTNND1","CUX1","CYLD","DDX5","DIXDC1","DKK1","DKK2","DLK1","DLL1","DLL3","DLL4","DVL2","DVL3","EMD","ENG","FSCN1","FOXD3","FOXK2","FOXP1","FOXP2","FOXP3","FUT4","FZD5","FZD6","GDF15","GLI1","GLI2","GREM1","GSK3A","GSK3B","HES1","ID2","ID3","IGF2BP1","ITCH","JAG1","JAG2","WWC1","KIF3A","KLF4","LATS1","LEF1","LEFTY1","LFNG","LIN28A","LIN28B","LRP5","LRP6","MAML1","MAML2","MARK3","MESD","MIB1","AMHR2","MOB1A","MOB1B","STK4","MSX1","MSI1","MYB","NANOG","NDRG1","NDRG4","NEDD4L","CSPG4","NCSTN","NKD1","NKD2","NKX2-2","NKX2-5","NLK","NOTCH1","NOTCH2","NOTCH3","NOTCH4","NR6A1","NUMB","POU5F1","PARN","PAX2","PAX3","PAX5","PAX9","PPM1A","PSEN1","PSEN2","PSENEN","PTCH1","PTCH2","PTPN14","RBPJ","RING1","RUVBL1","ZFYVE9","SAV1","SCRIB","SERPINE1","SFRP1","SHH","SIX1","SKIL","SMAD1","SMAD2","SMAD3","SMAD4","SMAD5","SMAD9","SMURF1","SMURF2","SNAI1","SNAI2","SOX2","SP1","SPARC","SUFU","TAB1","TAB2","ADAM17","MAP3K7","WWTR1","TBK1","TCF12","TCF7L1","TCF7","TCF7L2","ZEB1","TDGF1","TEAD1","TEAD2","TEAD3","TEAD4","TGFB1","TGFB2","TGFB3","TGFBR1","TGFBR3","TRIM33","TLE1","TLE2","TLE3","TLE4","TRIB2","USP15","UTF1","VANGL1","WBP2","WIF1","WNT3A","WNT5A","WNT5B","AMER1","YAP1")
# unique(intersect(wnt.hedgehog.notch, lineage.topgenes))
#
#
# non.canon.wnt <- c("DAG1","TK1","PFN1","PFN2","PFN4","PXN","ROCK2","CDC42","PPP3CA","NR2C2","NFATC1","NFATC3","TCF21","TCF25","TCF4","LEF1","WNT1","WNT10A","WNT11","WNT16","WNT2","WNT3","WNT3A","WNT4","WNT5A","WNT5B","WNT6","WNT7A","WNT7B","WNT9A","FZD1","FZD10","FZD2","FZD4","FZD5","FZD9","MAP2K3","MAP2K6","PRKCE","PRKCA","PRKCB","PRKCD","PRKCG","PRKCH","PRKCI","PRKCZ","CSNK1A1","RHOA","RAC1","CAMK2A","CAMK2B","CAMK2G","MAPK14","MAPK1","MAPK6","MAPK7","CTNNB1","NLK","ATF2")
# unique(intersect(non.canon.wnt, lineage.topgenes))
#hippo <- c("PATJ","TPTEP2-CSNK1E","APC2","YAP1","YWHAQ","RASSF1","FZD10","FRMD6","WTIP","CSNK1D","CSNK1E","CCN2","CTNNA1","CTNNA2","CTNNB1","GDF7","AMOT","RASSF6","DLG1","AFP","DLG2","DLG3","DLG4","DVL1","DVL2","DVL3","FGF1","WWC1","FBXW11","CRB1","SCRIB","FZD2","WWTR1","LATS2","AMH","BBC3","GLI2","CRB2","CTNNA3","GSK3B","APC","BIRC2","BIRC3","BIRC5","ID1","ID2","BMP8A","ITGB2","AREG","GDF6","LLGL2","LLGL1","SMAD1","SMAD2","SMAD3","SMAD4","SMAD7","MYC","NF2","SERPINE1","PARD6A","LEF1","WNT16","WNT4","PPP1CA","PPP1CB","PPP1CC","PPP2CA","PPP2CB","PPP2R1A","PPP2R1B","PPP2R2A","PPP2R2B","PPP2R2C","MOB1A","PRKCI","PPP2R2D","PRKCZ","PARD3","CCND1","ACTB","SAV1","PALS1","BMP2","BMP4","BMP5","BMP6","BMP7","BMP8B","BMPR1A","BMPR1B","BMPR2","SNAI2","SOX2","STK3","TCF7","TCF7L2","TEAD1","TEAD4","TEAD3","TGFB1","TGFB2","TGFB3","TGFBR1","TGFBR2","ACTG1","TP53BP2","TP73","WNT1","WNT2","WNT3","WNT5A","WNT6","WNT7A","WNT7B","WNT8A","WNT8B","WNT10B","WNT11","WNT2B","WNT9A","WNT9B","YWHAB","YWHAE","YWHAG","YWHAH","YWHAZ","FZD5","FZD3","FRMD1","WNT10A","WNT5B","GDF5","AXIN1","AXIN2","FZD1","FZD4","FZD6","FZD7","FZD8","FZD9","TCF7L1","PARD6G","PARD6B","TEAD2","AJUBA","NKD1","NKD2","CCND2","BTRC","CCND3","WNT3A","LIMD1","LATS1","DLG5","MOB1B","CDH1")
# unique(intersect(hippo, global.topgenes))
robo.receptor.signaling <- c("ROBO1","SLIT2","GPC1","RAC1","ABL1","ABL2","CAP2","CAP1","CLASP2","CLASP1","ROBO2","ROBO3","CDC42","NCK1","PAK1","PAK2","PAK3","SOS1","SOS2","NCK2","SRGAP3","PAK4","SRGAP2","PAK6","PAK5","SRGAP1","ARHGAP39","BUB1B-PAK6","PFN1","PFN2","VASP","EVL","ENAH")
robo <- intersect(topgenes, robo.receptor.signaling)
signaling.by.slit <- c("ABL1","ABL2","ACTR2","ACTR3","ARHGEF1","ARHGEF10","ARHGEF11","ARHGEF12","ARHGEF15","ARHGEF16","ARHGEF2","ARHGEF3","ARHGEF4","ARHGEF5","ARHGEF6","ARHGEF7","ARHGEF9","CDC42","CLASP1","CLASP2","CXCL12","CXCR4","DCC","ENAH","GNA11","GNA12","GNA13","GNA14","GNA15","GNAS","GNB1","GNB2","GNB3","GNB4","GNB5","GNG2","GNG3","GNG4","MCF2L","NET1","ROBO1","SLIT1","SLIT2","SLIT3","SRGAP1","SRGAP2","SRGAP3","WASL")
slit <- intersect(topgenes, signaling.by.slit)
EMT <- c("ARHGEF18","CGN","F11R","FKBP1A","MIR6869","PARD3","PARD6A","PRKCZ","RHOA","RPS27A","SMURF1","TGFB1","TGFBR1","TGFBR2","UBA52","UBB","UBC")
intersect(topgenes, EMT)
# 4. extracting genes of interest ----------
# reshape data
# converting sparse matrix to data frame
summ <- summary(heatdata)
df.heatdata <- data.frame(gene = rownames(heatdata)[summ$i],
cells = colnames(heatdata)[summ$j],
logcounts = summ$x)
# merge df.heatdata with col_anno.ordered
df.heatdata <- merge(df.heatdata, col_anno.ordered, by.x = 'cells', by.y = 'row.names')
# comparing mean expression of ADRN/MES genes between Parental and LRX cells
unique(intersect(topgenes, ADRN.overlap))
unique(intersect(topgenes, MES.overlap))
#genes.of.interest <- rbind(adrenal, mesench)
genes.of.interest <- unique(c(wnt, robo, slit, 'SOX6', 'MCC', 'BCL2'))
genes.of.interest <- as.data.frame(genes.of.interest)
names(genes.of.interest)[1] <- 'V1'
df.heatdata <- merge(df.heatdata, genes.of.interest, by.x = 'gene', by.y = 'V1')
# plotting expression
df.heatdata$seurat_clusters <- factor(df.heatdata$seurat_clusters, levels=c(13,5,6,3,10,8,0,18,15,9,19,7,17,16,11,4,2,14,1,12))
a1 <- ggplot(df.heatdata, aes(seurat_clusters, logcounts, fill = group)) +
geom_boxplot() +
facet_wrap(. ~ gene)
#ggsave(a1, filename = 'figures/boxplot_pseudotime1_top200genes_preSCT.pdf', width = 15, height = 10)
#ggsave(a1, filename = 'figures/boxplot_global_pseudotime_top200genes_postSCT.pdf', width = 15, height = 10)
df.heatdata$group <- ifelse(grepl('LRX', df.heatdata$orig.ident), 'LRX', 'Parental')
a2 <- ggplot(df.heatdata, aes(seurat_clusters, logcounts)) +
geom_boxplot() +
facet_wrap(. ~ gene)
#ggsave(a2, filename = 'figures/boxplot_global_GOI_pseudotime_top200genes_preSCT.pdf', width = 15, height = 10)
#ggsave(a2, filename = 'figures/boxplot_GOI_pseudotime1_top200genes_preSCT.pdf', width = 15, height = 10)
ggsave(a2, filename = paste0('figures/',Sys.Date(),'_boxplot_test_pseudotime1_top200genes_preSCT.pdf'), width = 15, height = 10)
p1 <- ggplot(df.heatdata, aes(reorder(gene, logcounts), logcounts, fill = orig.ident)) +
geom_boxplot() +
facet_wrap(. ~ group) +
theme(axis.text.x = element_text(angle = 45, hjust = 1, vjust = 1))
ggsave(p1, filename = 'figures/boxplot_pseudotime2_top100genes_preSCT.pdf', width = 15, height = 10)
p2 <- ggplot(df.heatdata, aes(logcounts, fill = orig.ident)) +
geom_density(alpha = 0.4) +
facet_wrap(. ~ group) +
theme(axis.text.x = element_text(angle = 45, hjust = 1, vjust = 1))
ggsave(p2, filename = 'figures/densityplot_pseudotime1_top50genes_preSCT.pdf', width = 15, height = 10)
# 5. GSEA on DE genes for lineages -------------------------------------------
# top 100,200 and 500 genes give very generalized hits
# hs <- org.Hs.eg.db
#
# my.symbols <- c(as.character(topgenes))
# df <- AnnotationDbi::select(hs,
# keys = my.symbols,
# columns = c("ENTREZID", "SYMBOL"),
# keytype = "SYMBOL")
#
#
# # getting entrezsymbol for top genes
# top.genes.df <- as.data.frame(topgenes)
# top.genes.gsea <- merge(top.genes.df, df, by.x = 'topgenes', by.y = 'SYMBOL')
# top.genes.gsea <- na.omit(top.genes.gsea)
#
# # setting ranks using wald stat for lineage 1
# ranks <- ATres[rownames(ATres) %in% top.genes.gsea$topgenes, 'waldStat_1']
# names(ranks) <- top.genes.gsea$ENTREZID
# head(ranks)
#
# # load pathways
# my_pathways <- reactomePathways(names(ranks))
#
# eaRes <- fgsea(pathways = my_pathways, stats = ranks, nperm = 100000, minSize = 15)
# out <- eaRes
#' trying monocle! -------------
# # creating feature data
# gene_annotation <- as.data.frame(rownames(pre.sct@reductions[["pca"]]@feature.loadings),
# row.names = rownames(pre.sct@reductions[["pca"]]@feature.loadings))
# colnames(gene_annotation) <- "gene_short_name"
#
# # creating pheno data
# cell_metadata <- as.data.frame(pre.sct@assays[["RNA"]]@counts@Dimnames[[2]],
# row.names = pre.sct@assays[["RNA"]]@counts@Dimnames[[2]])
# colnames(cell_metadata) <- "barcode"
#
# # creating a CellDataSet for Monocle
# New_matrix <- pre.sct@assays[["RNA"]]@counts
# New_matrix <- New_matrix[rownames(pre.sct@reductions[["pca"]]@feature.loadings), ]
# expression_matrix <- New_matrix
#
#
# cds_from_seurat <- newCellDataSet(expression_matrix,
# phenoData = new("AnnotatedDataFrame", data = cell_metadata),
# featureData = new("AnnotatedDataFrame", data = gene_annotation),
# expressionFamily=negbinomial.size())
#
# # exploring slot names
# slotNames(cds_from_seurat)
#
#
#
# # Estimating size factors and dispersions (required)
# cds_from_seurat <- estimateSizeFactors(cds_from_seurat)
# cds_from_seurat <- estimateDispersions(cds_from_seurat)
#
#
# # tally number of cells expressing a gene and # of genes expressed among all cells (recommended step for all monocle analysis)
# cds_from_seurat <- detectGenes(cds_from_seurat, min_expr = 0.1)
# print(head(fData(cds_from_seurat), n = 20))
#
# summary(fData(cds_from_seurat)$num_cells_expressed)
#
# expressed_genes <- row.names(subset(fData(cds_from_seurat),
# num_cells_expressed >= 10))
#
#
#
# # to find number of genes expressed per cell
# head(pData(cds_from_seurat))
#
#
#
# summary(pData(cds_from_seurat)$num_genes_expressed)
#
#
# # standardise to Z-distribution
# x <- pData(cds_from_seurat)$num_genes_expressed
# x_1 <- (x - mean(x)) / sd(x)
# summary(x_1)
#
# library(ggplot2)
# # I like the default theme of cowplot
# library(cowplot)
#
# df <- data.frame(x = x_1)
# g <- ggplot(df, aes(x)) +
# geom_histogram(bins = 50) +
# geom_vline(xintercept = c(-2, 2), linetype = "dotted", color = 'red')
# ggsave(g, filename = 'figures/QC_monocle_num_of_genes.pdf', width = 10, height = 10)
#
#
# # clustering cells without marker genes
# disp_table <- dispersionTable(cds_from_seurat)
# head(disp_table)
#
# # select genes with mean expression >= 0.1 to use in clustering step
# table(disp_table$mean_expression>=0.1)
#
# unsup_clustering_genes <- subset(disp_table, mean_expression >= 0.1)
#
# cds_from_seurat <- setOrderingFilter(cds_from_seurat, unsup_clustering_genes$gene_id)
# p <- plot_ordering_genes(cds_from_seurat)
# # The plot_ordering_genes() function plots mean expression against the empirical dispersion and highlights the set of genes (as black dots) that will be used for clustering.
# ggsave(p, filename = 'figures/QC_monocle_custering.pdf', width = 10, height = 10)
#
#
# # Constructing Single Cell Trajectories
# # The trajectory analysis consists of three stages:
# # 1. Choose genes that define progress
# # 2. Reduce the dimensionality of the data
# # 3. Order cells in pseudotime
#
#
# # just using expressed_genes (genes_expressed >= 10 cells)
# cds_from_seurat_subset <- cds_from_seurat[expressed_genes,]
# cds_from_seurat_subset
#
# # ordering - based on genes that differ between clusters
# cds_from_seurat_subset <- detectGenes(cds_from_seurat, min_expr = 0.1)
# fData(cds_from_seurat_subset)$use_for_ordering <- fData(cds_from_seurat_subset)$num_cells_expressed > 0.05 * ncol(cds_from_seurat_subset)
# head(fData(cds_from_seurat_subset))
#
# # how many genes are used?
# table(fData(cds_from_seurat_subset)$use_for_ordering)
#
# # FALSE TRUE
# # 973 1027
#
#
# p1 <- plot_pc_variance_explained(cds_from_seurat_subset, return_all = FALSE)
# ggsave(p1, filename = 'figures/trajectoryAnalysis_monocle_PCA_preSCT.pdf', width = 10, height = 10)
#
#
# # reduce Dimensions
# cds_from_seurat_subset <- reduceDimension(cds_from_seurat_subset,
# max_components = 2,
# norm_method = 'log',
# num_dim = 20,
# reduction_method = 'tSNE',
# verbose = TRUE)
#
# cds_from_seurat_subset <- clusterCells(cds_from_seurat_subset, verbose = FALSE)
#
# p2 <- plot_rho_delta(cds_from_seurat_subset, rho_threshold = 2, delta_threshold = 10)
# ggsave(p2, filename = 'figures/trajectoryAnalysis_monocle_clustering_rhoDelta_preSCT.pdf', width = 10, height = 10)
#
# # use rho = 2 and delta = 10 to cluster the cells again.
# cds_from_seurat_subset <- clusterCells(cds_from_seurat_subset,
# rho_threshold = 2,
# delta_threshold = 10,
# skip_rho_sigma = T,
# verbose = FALSE)
#
# table(pData(cds_from_seurat_subset)$Cluster)
#
# p3 <- plot_cell_clusters(cds_from_seurat_subset)
# ggsave(p3, filename = 'figures/trajectoryAnalysis_monocle_clustering_preSCT.pdf', width = 10, height = 10)
#
#
# # performing the differential gene expression analysis across all cell clusters.
# # to use the top 1,000 most significantly differentially expressed genes as the set of ordering genes and perform the dimension reduction and the trajectory analysis (using the orderCells() function).
#
# clustering_DEG_genes <- differentialGeneTest(cds_from_seurat_subset,
# fullModelFormulaStr = '~Cluster',
# cores = 8)
#
# dim(clustering_DEG_genes)
#
# clustering_DEG_genes %>% arrange(qval) %>% head()
#
#
# my_ordering_genes <- row.names(clustering_DEG_genes)[order(clustering_DEG_genes$qval)][1:1000]
# cds_from_seurat_subset <- setOrderingFilter(cds_from_seurat_subset, ordering_genes = my_ordering_genes)
# cds_from_seurat_subset <- reduceDimension(cds_from_seurat_subset, method = 'DDRTree')
# cds_from_seurat_subset <- orderCells(cds_from_seurat_subset)
#
# p4 <- plot_cell_trajectory(cds_from_seurat_subset, color_by = "Cluster")
# ggsave(p4, filename = 'figures/trajectoryAnalysis_monocle_trajectory_preSCT.pdf', width = 10, height = 10)
#
#
# # Finding Genes that Change as a Function of Pseudotime
# # Once we have a trajectory, we can use differentialGeneTest() to find genes
# # that have an expression pattern that varies according to pseudotime.
#
# # pseudotime is now a column in the phenotypic data as well as the cell state
# head(pData(cds_from_seurat_subset))