-
Notifications
You must be signed in to change notification settings - Fork 0
/
2A_TB_DGE.R
344 lines (253 loc) · 14.7 KB
/
2A_TB_DGE.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
#############################################
## Differential Gene Expression, ORA, GSEA ##
#############################################
library(DESeq2)
library(tidyverse)
library(ggplot2)
library(EnhancedVolcano)
library(ggpubr)
library(car)
library(ggsignif)
library(enrichplot)
# DGE
#----------------------------------------------------------------------------------
# Load counts and metadata
setwd("C:/Users/manfa/OneDrive - Universitaet Bern/MMF working folder/Scripts")
data_annot <- read_csv(file = "../TB_epithelium/data_annot_tb_e.csv")
data_raw <- read_csv(file = "../TB_epithelium/data_raw_tb_e.csv", col_select = c(2:29))
data_annot$subgroup.a <- as.factor(data_annot$subgroup.a)
data_annot <- data_annot %>% column_to_rownames(var="SegmentDisplayName")
data_raw <- data_raw %>% column_to_rownames(var="TargetName")
data_annot[which(data_annot$T == TRUE), 12] <- "A"
data_annot[which(data_annot$T == FALSE), 12] <- "B"
colnames(data_annot)[12] <- "Slide"
# Exclude TMA X4-Y2 (mislabeled)
data_raw <- data_raw %>% dplyr::select(-"ROI 03_22 Bern A 15/03/23 | TMA-2203-A-X4-Y2-415-TB | T")
data_annot <- data_annot[-4,-33] # Drop CMS column too
# Make all count values integers
data_raw[] <- sapply(data_raw, as.integer)
# Create DESeq2 object
all(colnames(data_raw) %in% rownames(data_annot))
all(colnames(data_raw) == rownames(data_annot)) # check that row and colnames align
design1 <- ~ subgroup.a
design2 <- ~ Slide + subgroup.a # Batch effect
dds <- DESeqDataSetFromMatrix(countData = data_raw, colData = data_annot, design = design1)
# Run the DGE
dds <- DESeq(dds)
# Extract results for each pairwise comparison
contrast_SMSS <- c("subgroup.a", "TSS", "TSM") # TSM is base
contrast_SSSM <- c("subgroup.a", "TSM", "TSS") # For plotting GSEA
res_SMSS_unshrunken <- results(dds, contrast=contrast_SMSS, alpha = 0.05, independentFiltering=T, cooksCutoff = T) # filtering always on
res_SMSS <- lfcShrink(dds, contrast=contrast_SMSS, res=res_SMSS_unshrunken, type = "normal") # Shrink Log2 fold-change
res_SMSS_tbl <- res_SMSS %>% data.frame()
res_SSSM <- results(dds, contrast=contrast_SSSM, alpha = 0.05, independentFiltering=T, cooksCutoff = T) # filtering always on
res_SSSM <- lfcShrink(dds, contrast=contrast_SSSM, res=res_SSSM, type = "normal") # Shrink Log2 fold-change
res_SSSM_tbl <- res_SSSM %>% data.frame()
contrast_SMMP <- c("subgroup.a", "TMP", "TSM") # TSM is base
res_SMMP <- results(dds, contrast=contrast_SMMP, alpha = 0.05, independentFiltering=T, cooksCutoff = T)
res_SMMP <- lfcShrink(dds, contrast=contrast_SMMP, res=res_SMMP, type = "normal")
res_SMMP_tbl <- res_SMMP %>% data.frame()
contrast_SSMP <- c("subgroup.a", "TMP", "TSS") # TSS is base
contrast_MPSS <- c("subgroup.a", "TSS", "TMP") # For plotting GSEA
res_SSMP <- results(dds, contrast=contrast_SSMP, alpha = 0.05, independentFiltering=T, cooksCutoff = T)
res_SSMP <- lfcShrink(dds, contrast=contrast_SSMP, res=res_SSMP, type = "normal")
res_SSMP_tbl <- res_SSMP %>% data.frame()
res_MPSS <- results(dds, contrast=contrast_MPSS, alpha = 0.05, independentFiltering=T, cooksCutoff = T)
res_MPSS <- lfcShrink(dds, contrast=contrast_MPSS, res=res_MPSS, type = "normal")
res_MPSS_tbl <- res_MPSS %>% data.frame()
summary(res_SMSS)
summary(res_SMMP)
summary(res_SSMP)
# Extract sig DE genes in SMSS
sig_SMSS <- res_SMSS_tbl %>% filter(padj < 0.05)
# MA Plot, visualize shrinkage
plotMA(res_SMSS_unshrunken, ylim=c(-3,3), main ="TSS vs TSM, unshrunken", colSig="red2")
plotMA(res_SMSS, ylim=c(-3,3), main ="TSS vs TSM, shrunken", colSig="red2")
plotMA(res_SMMP, ylim=c(-3,3), main ="TMP vs TSM, shrunken", colSig="red2")
plotMA(res_SSMP, ylim=c(-3,3), main ="TMP vs TSS, shrunken", colSig="red2")
# Volcano plots
pdf("../TB_epithelium/Figures/volcano_SSSM.pdf", width = 8)
EnhancedVolcano(res_SSSM_tbl, lab = rownames(res_SSSM_tbl), x = 'log2FoldChange', y = 'pvalue',
title = "DGE in Tumor Buds: TSM vs TSS", subtitle = element_blank(),
pCutoff = 0.05, pCutoffCol = 'padj', FCcutoff = 0.58,
legendPosition = 'bottom', legendLabSize = 10, legendIconSize = 3.0,
drawConnectors = TRUE, widthConnectors = 0.5,
max.overlaps = 15, pointSize = 1, labSize = 4)
dev.off()
pdf("../TB_epithelium/Figures/volcano_SMMP.pdf", width = 8)
EnhancedVolcano(res_SMMP_tbl, lab = rownames(res_SMMP_tbl), x = 'log2FoldChange', y = 'pvalue',
title = "DGE in TB: TMP vs TSM (Base)", subtitle = element_blank(),
pCutoff = 0.05, pCutoffCol = 'padj', FCcutoff = 0.58,
legendPosition = 'bottom', legendLabSize = 10, legendIconSize = 3.0,
drawConnectors = TRUE, widthConnectors = 0.5,
max.overlaps = 15, pointSize = 1, labSize = 4)
dev.off()
pdf("../TB_epithelium/Figures/volcano_SSMP.pdf", width = 8)
EnhancedVolcano(res_SSMP_tbl, lab = rownames(res_SSMP_tbl), x = 'log2FoldChange', y = 'pvalue',
title = "DGE in TB: TMP vs TSS (Base)", subtitle = element_blank(),
pCutoff = 0.05, pCutoffCol = 'padj', FCcutoff = 0.58,
legendPosition = 'bottom', legendLabSize = 10, legendIconSize = 3.0,
drawConnectors = TRUE, widthConnectors = 0.5,
max.overlaps = 15, pointSize = 1, labSize = 4)
dev.off()
# Pathway enrichment analyses
#---------------------------------------------------------------------------------------------------
###### GSEA, hallmark gene sets ######
library(org.Hs.eg.db)
library(clusterProfiler)
library(enrichplot)
# Shift gene names to SYMBOL column
res_SMSS_gsea <- res_SMSS_tbl %>% rownames_to_column(var="SYMBOL")
res_SSSM_gsea <- res_SSSM_tbl %>% rownames_to_column(var="SYMBOL")
res_SMMP_gsea <- res_SMMP_tbl %>% rownames_to_column(var="SYMBOL")
res_SSMP_gsea <-res_SSMP_tbl %>% rownames_to_column(var="SYMBOL")
res_MPSS_gsea <-res_MPSS_tbl %>% rownames_to_column(var="SYMBOL")
# Map SYMBOL name to the Entrez IDs
symbol2enz <- AnnotationDbi::select(org.Hs.eg.db,
key=res_SMSS_gsea$SYMBOL,
columns=("ENTREZID"), # desired format
keytype="SYMBOL")
res_SMSS_gsea <- dplyr::inner_join(res_SMSS_gsea, symbol2enz, by = "SYMBOL")
res_SSSM_gsea <- dplyr::inner_join(res_SSSM_gsea, symbol2enz, by = "SYMBOL")
res_SMMP_gsea <- dplyr::inner_join(res_SMMP_gsea, symbol2enz, by = "SYMBOL")
res_SSMP_gsea <- dplyr::inner_join(res_SSMP_gsea, symbol2enz, by = "SYMBOL")
res_MPSS_gsea <- dplyr::inner_join(res_MPSS_gsea, symbol2enz, by = "SYMBOL")
# Remove the unmapped genes, and only keep log2FC and ID columns
res_SMSS_gsea <- res_SMSS_gsea[!is.na(res_SMSS_gsea$ENTREZID), ] %>% dplyr::select(ENTREZID, log2FoldChange)
res_SSSM_gsea <- res_SSSM_gsea[!is.na(res_SSSM_gsea$ENTREZID), ] %>% dplyr::select(ENTREZID, log2FoldChange)
res_SMMP_gsea <- res_SMMP_gsea[!is.na(res_SMMP_gsea$ENTREZID), ] %>% dplyr::select(ENTREZID, log2FoldChange)
res_SSMP_gsea <- res_SSMP_gsea[!is.na(res_SSMP_gsea$ENTREZID), ] %>% dplyr::select(ENTREZID, log2FoldChange)
res_MPSS_gsea <- res_MPSS_gsea[!is.na(res_MPSS_gsea$ENTREZID), ] %>% dplyr::select(ENTREZID, log2FoldChange)
# Check duplicates
sum(duplicated(res_SMSS_gsea$ENTREZID))
# Rank according to Log2FC
res_SMSS_gsea <- res_SMSS_gsea[order(-res_SMSS_gsea$log2FoldChange), ]
log2fc_SMSS <- res_SMSS_gsea$log2FoldChange
names(log2fc_SMSS) <- res_SMSS_gsea[, 1]
res_SSSM_gsea <- res_SSSM_gsea[order(-res_SSSM_gsea$log2FoldChange), ]
log2fc_SSSM <- res_SSSM_gsea$log2FoldChange
names(log2fc_SSSM) <- res_SSSM_gsea[, 1]
res_SMMP_gsea <- res_SMMP_gsea[order(-res_SMMP_gsea$log2FoldChange), ]
log2fc_SMMP <- res_SMMP_gsea$log2FoldChange
names(log2fc_SMMP) <- res_SMMP_gsea[, 1]
res_SSMP_gsea <- res_SSMP_gsea[order(-res_SSMP_gsea$log2FoldChange), ]
log2fc_SSMP <- res_SSMP_gsea$log2FoldChange
names(log2fc_SSMP) <- res_SSMP_gsea[, 1]
res_MPSS_gsea <- res_MPSS_gsea[order(-res_MPSS_gsea$log2FoldChange), ]
log2fc_MPSS <- res_MPSS_gsea$log2FoldChange
names(log2fc_MPSS) <- res_MPSS_gsea[, 1]
# Fetch Hallmark gene sets from msigdb
hallmark_gs <- msigdbr::msigdbr(
species = "Homo sapiens",
category = "H") %>% dplyr::select(gs_name, entrez_gene)
# Run GSEA
gsea_SMSS <- GSEA(log2fc_SMSS, TERM2GENE = hallmark_gs)
gsea_SMMP <- GSEA(log2fc_SMMP, TERM2GENE = hallmark_gs)
gsea_SSMP <- GSEA(log2fc_SSMP, TERM2GENE = hallmark_gs)
# Visualize GSEA results
dot_SMSS <- dotplot(gsea_SMSS, showCategory=15, split=".sign", font.size=9, size = NULL, label_format = 90) +
ggtitle("Hallmark, TSS vs TSM (Base)") +
facet_grid(~.sign) +
theme(plot.title.position = "plot")
# Note: GeneRatio = # of core (contributed to the enrichment score) DE genes in pathway / total count of genes in pathway
dot_SMMP <- dotplot(gsea_SMMP, showCategory=15, split=".sign", font.size=9, size = NULL, label_format = 90) +
ggtitle("Hallmark, TMP vs TSM (Base)") +
facet_grid(~.sign) +
theme(plot.title.position = "plot")
dot_SSMP <- dotplot(gsea_SSMP, showCategory=15, split=".sign", font.size=9, size = NULL, label_format = 90) +
ggtitle("Hallmark, TMP vs TSS (Base)") +
facet_grid(~.sign) +
theme(plot.title.position = "plot")
pdf("../TB_epithelium/Figures/gsea_SMSS_dot.pdf", width = 9)
dot_SMSS
dev.off()
pdf("../TB_epithelium/Figures/gsea_SMMP_dot.pdf", width = 9)
dot_SMMP
dev.off()
pdf("../TB_epithelium/Figures/gsea_SSMP_dot.pdf", width = 9)
dot_SSMP
dev.off()
####### Overrepresentation Analysis (ORA) #######
#Get ENTREZ IDs of sig. DE genes
sig_SMSS <- sig_SMSS %>% rownames_to_column(var="SYMBOL")
symbol2enz <- AnnotationDbi::select(org.Hs.eg.db,
key=sig_SMSS$SYMBOL,
columns=("ENTREZID"), # desired format
keytype="SYMBOL")
sig_SMSS <- dplyr::inner_join(sig_SMSS, symbol2enz, by = "SYMBOL")
gene <- sig_SMSS$ENTREZID
# Run ORA and make upset plot, hallmark
ora_SMSS <- enricher(gene, TERM2GENE=hallmark_gs)
pdf("../TB_epithelium/Figures/ora_SMSS.pdf", width = 10)
upsetplot(ora_SMSS) + labs(title = "Hallmark, TSS vs TSM") + theme(plot.title.position = "plot")
dev.off()
# Save ORA, GSEA and Sig. DE tables
write.table(ora_SMSS@result, file="../TB_epithelium/ora_SMSS.tsv", sep="\t", quote=F, col.names=NA)
write.table(gsea_SMSS@result, file="../TB_epithelium/gsea_SMSS.tsv", sep="\t", quote=F, col.names=NA)
write.table(gsea_SMMP@result, file="../TB_epithelium/gsea_SMMP.tsv", sep="\t", quote=F, col.names=NA)
write.table(gsea_SSMP@result, file="../TB_epithelium/gsea_SSMP.tsv", sep="\t", quote=F, col.names=NA)
write.table(sig_SMSS, file="../TB_epithelium/sigDE.tsv", sep="\t", quote=F, col.names=NA)
###### Reactome pathways #######
library(ReactomePA)
#Run GSEA, same params as above
gsea_SMSS2 <- gsePathway(log2fc_SMSS, organism = "human", pvalueCutoff = 0.05, pAdjustMethod = "BH")
gsea_SSSM2 <- gsePathway(log2fc_SSSM, organism = "human", pvalueCutoff = 0.05, pAdjustMethod = "BH")
gsea_SMMP2 <- gsePathway(log2fc_SMMP, organism = "human", pvalueCutoff = 0.05, pAdjustMethod = "BH")
gsea_SSMP2 <- gsePathway(log2fc_SSMP, organism = "human", pvalueCutoff = 0.05, pAdjustMethod = "BH")
gsea_MPSS2 <- gsePathway(log2fc_MPSS, organism = "human", pvalueCutoff = 0.05, pAdjustMethod = "BH")
write.table(gsea_SSMP2@result, file="../TB_epithelium/gsea_SSMP2.tsv", sep="\t", quote=F, col.names=NA)
write.table(gsea_SMMP2@result, file="../TB_epithelium/gsea_SMMP2.tsv", sep="\t", quote=F, col.names=NA)
write.table(gsea_SMSS2@result, file="../TB_epithelium/gsea_SMSS2.tsv", sep="\t", quote=F, col.names=NA)
# Visualize GSEA results
pdf("../TB_epithelium/Figures/gsea_SMSS2_dot.pdf", width = 10)
dotplot(gsea_SMSS2, showCategory=20, split=".sign", font.size=9, size = NULL, label_format = 90) +
ggtitle("Reactome, TSS vs TSM (Base)") +
facet_grid(~.sign) +
theme(plot.title.position = "plot")
dev.off()
pdf("../TB_epithelium/Figures/gsea_SMMP2_dot.pdf", width = 10)
dotplot(gsea_SMMP2, showCategory=20, split=".sign", font.size=9, size = NULL, label_format = 90) +
ggtitle("Reactome, TMP vs TSM (Base)") +
facet_grid(~.sign) +
theme(plot.title.position = "plot")
dev.off()
pdf("../TB_epithelium/Figures/gsea_SSMP2_dot.pdf", width = 10)
dotplot(gsea_SSMP2, showCategory=20, split=".sign", font.size=9, size = NULL, label_format = 90) +
ggtitle("Reactome, TMP vs TSS (Base)") +
facet_grid(~.sign) +
theme(plot.title.position = "plot")
dev.off()
#Run ORA and make upset plot
ora_SMSS2 <- enrichPathway(gene=gene, organism = "human",pvalueCutoff = 0.05, pAdjustMethod = "BH", readable=TRUE)
pdf("../TB_epithelium/Figures/ora_SMSS2.pdf", width = 10)
upsetplot(ora_SMSS2) + labs(title = "Reactome, TSS vs TSM") + theme(plot.title.position = "plot")
dev.off()
############ Presentation/Report additional figures ###########
# Format data
normalized_counts <- counts(dds, normalized=TRUE)
normalized_counts <- normalized_counts %>%
data.frame() %>%
rownames_to_column(var="gene") %>%
as_tibble()
group_df <- data_annot[, c(1, 5)] %>% rownames_to_column(var = 'Sample')
group_df <- group_df[, c(1, 3)]
normalized_counts <- normalized_counts %>%
gather(colnames(normalized_counts)[2:27], key = "Sample", value = "counts")
normalized_counts['subgroup.a'] <- rep(group_df$subgroup.a, each = 6920) # add metadata
normalized_counts$subgroup.a <- factor(normalized_counts$subgroup.a, levels = c("TSM", "TMP", "TSS"))
comparisons <- list(c('TSM, TMP'), c('TSM, TSS'), c('TSS, TMP'))
# Boxplot PD-L1
wilcox_result <- compare_means(counts ~ subgroup.a, data = normalized_counts[normalized_counts$gene == 'CD274',], method = 'wilcox.test', paired = F)
shapiro_result_tmp <- shapiro.test(normalized_counts$counts[normalized_counts$subgroup.a == 'TMP' & normalized_counts$gene == 'CD274'])
bp <- ggplot(normalized_counts[normalized_counts$gene %in% c('CD274'),], aes(x=gene, y=counts, fill = subgroup.a)) +
geom_boxplot() + geom_dotplot(binaxis='y', stackdir='center', position= position_dodge(width = 0.75), dotsize=0.5) +
labs(x = 'Gene', y='Normalized counts',fill = 'Region')
pdf("../TB_epithelium/Figures/boxplot_pdl1.pdf", width = 12, height = 7)
bp + scale_fill_manual(values = c('#00ba38','#F8766D','#83b0fc')) + theme_bw() #+ stat_compare_means(method = "kruskal.test", comparisons = comparisons)
dev.off()
# Additional GSEA plots
## Running score
pdf("../TB_epithelium/Figures/running_ecm.pdf", width = 12)
gseaplot2(gsea_SMMP2, geneSetID = c(5), title = paste0(gsea_SMMP2@result$Description[5], ", TMP vs TSM (Base)"))
dev.off()
gseaplot2(gsea_SMMP2, geneSetID = c(1, 5), title = "TMP vs TSM (Base)")