-
Notifications
You must be signed in to change notification settings - Fork 17
/
module4_vitro_vs_vivo_II.R
290 lines (234 loc) · 26.6 KB
/
module4_vitro_vs_vivo_II.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
# ----------------------------------------------------------------------------------------------------
# Module 4. Evaluating expression heterogeneity in HNSCC and melanoma by analyzing human tumors and cell lines as a single dataset
# ----------------------------------------------------------------------------------------------------
# **************************************************************************
# Basic setup
# load necessary R packages and functions
library(reshape2)
library(ggplot2)
library(scales)
# read scRNA-seq data from cell lines and tumors
expr_ccle <- readRDS("CCLE_heterogeneity_Rfiles/CCLE_scRNAseq_CPM.RDS") # CCLE cell lines
expr_tumor <- readRDS("CCLE_heterogeneity_Rfiles/tumors_scRNAseq_logTPM.RDS") # human tumors
# calculate average gene expression
rowmeans_ccle <- lapply(expr_ccle, rowMeans)
rowmeans_tumor <- lapply(expr_tumor, function(x) rowMeans(10*((2^x)-1)))
# process data
expr_ccle<- lapply(expr_ccle, function(x) {log2((x/10) + 1)})
expr_ccle <- lapply(expr_ccle, function(x) {x-rowMeans(x)})
expr_tumor <- lapply(expr_tumor, function(x) {x-rowMeans(x)})
# select relevant cell lines and tumors
nmf_metaprograms_programs_nc_ccle<- readRDS("Expected_results/module2/nmf_metaprograms_programs_nc_ccle.RDS") # programs in each cell line metaprogram
nmf_metaprograms_programs_nc_tumor<- readRDS("Expected_results/module2/nmf_metaprograms_programs_nc_tumor.RDS") # programs in each tumor metaprogram
episen_emt_ccle <- intersect(gsub(".{4}$", "",nmf_metaprograms_programs_nc_ccle$episen), gsub(".{4}$", "",nmf_metaprograms_programs_nc_ccle$emtII))
episen_emt_ccle <- unique(episen_emt_ccle[grep("UPPER", episen_emt_ccle)]) # HNCSS cell lines harboring both EpiSen and EMTII
episen_emt_tumor <- unique(intersect(gsub(".{4}$", "",nmf_metaprograms_programs_nc_tumor$episen), gsub(".{4}$", "",nmf_metaprograms_programs_nc_tumor$emt))) # HNCSS tumors harboring both EpiSen and EMTII
skinpig_emt_ccle <- intersect(gsub(".{4}$", "",nmf_metaprograms_programs_nc_ccle$skinpig), gsub(".{4}$", "",nmf_metaprograms_programs_nc_ccle$emtI)) # melanoma cell lines harboring both skinpig and EMTI
skinpig_emt_tumor <- names(expr_tumor)[grep("mel", names(expr_tumor))] # melanoma tumors harboring both skinpig and EMTI
# read metaprogram signatures from cell lines
meta_sig_ccle <- readRDS("Expected_results/module2/nmf_metaprograms_sig_nc_ccle.RDS")
meta_sig_ccle <- meta_sig_ccle[c("skinpig", "emtI", "episen", "emtII")]
names(meta_sig_ccle) <- paste0(names(meta_sig_ccle), "_vitro")
meta_sig_ccle <- lapply(meta_sig_ccle, function(x) names(x[1:100]))
# read metaprogram signatures from tumors
meta_sig_tumor <- unlist(list(read.table("CCLE_heterogeneity_Rfiles/metaprograms_tumors_literature.txt", sep = "\t", header = T,stringsAsFactors = F)), recursive=F)
meta_sig_tumor <- meta_sig_tumor[c("melanoma.MITF.program", "melanoma.AXL.program", "HNSCC.PEMT","HNSCC.Epidif.1" )]
meta_sig_tumor <- lapply(meta_sig_tumor, function(x) x[x!=""])
# get genes shared between programs in tumors and cell lines
meta_intersect <- lapply(meta_sig_ccle, function(x) lapply(meta_sig_tumor, function(y) intersect(x,y)))
# **************************************************************************
# Combine datasets and run PCA - HNSCC
# get expression data from selected cell lines
expr_emt_episen_vitro <- expr_ccle[episen_emt_ccle]
expr_emt_episen_vivo <- expr_tumor[episen_emt_tumor]
# select common genes in the cell line and tumor datasets
expr_emt_episen_vitro <- sapply(expr_emt_episen_vitro, function(x) x[Reduce(intersect, c(lapply( expr_emt_episen_vitro, rownames), lapply(expr_emt_episen_vivo, rownames))),])
expr_emt_episen_vivo <- sapply(expr_emt_episen_vivo, function(x) x[Reduce(intersect, c(lapply(expr_emt_episen_vitro, rownames), lapply(expr_emt_episen_vivo, rownames))),])
# unlist datasets
expr_emt_episen_vitro <- do.call(cbind,expr_emt_episen_vitro )
expr_emt_episen_vivo <- do.call(cbind, expr_emt_episen_vivo)
# select top expressed genes among cell lines and tumors
rowmeans_ccle_hnscc <-rowMeans(sapply(rowmeans_ccle[episen_emt_ccle], function(x) x[rownames(expr_emt_episen_vitro)]))
rowmeans_tumor_hnscc <- rowMeans(sapply(rowmeans_tumor[episen_emt_tumor], function(x) x[rownames(expr_emt_episen_vitro)]))
expr_emt_episen_vitro_filt <- expr_emt_episen_vitro[order(rowMeans(cbind(rowmeans_ccle_hnscc, rowmeans_tumor_hnscc)), decreasing=T)[1:4500] ,]
expr_emt_episen_vivo_filt <- expr_emt_episen_vivo[order(rowMeans(cbind(rowmeans_ccle_hnscc, rowmeans_tumor_hnscc)), decreasing=T)[1:4500] ,]
# combine the two datasets
expr_emt_episen_final <- cbind(expr_emt_episen_vitro_filt,expr_emt_episen_vivo_filt)
# run pca
expr_emt_episen_pca <- prcomp(t(expr_emt_episen_final))
# **************************************************************************
# Combine datasets and run PCA - melanoma
# get expression data from selected cell lines
expr_skinpig_emt_vitro <- expr_ccle[skinpig_emt_ccle]
expr_skinpig_emt_vivo <- expr_tumor[skinpig_emt_tumor]
# select common genes in the cell line and tumor datasets
expr_skinpig_emt_vitro <- sapply(expr_skinpig_emt_vitro, function(x) x[Reduce(intersect, c(lapply( expr_skinpig_emt_vitro, rownames), lapply(expr_skinpig_emt_vivo, rownames))),])
expr_skinpig_emt_vivo <- sapply(expr_skinpig_emt_vivo, function(x) x[Reduce(intersect, c(lapply(expr_skinpig_emt_vitro, rownames), lapply(expr_skinpig_emt_vivo, rownames))),])
# unlist datasets
expr_skinpig_emt_vitro <- do.call(cbind, expr_skinpig_emt_vitro)
expr_skinpig_emt_vivo <- do.call(cbind, expr_skinpig_emt_vivo)
# select top expressed genes among cell lines and tumors
rowmeans_ccle_melanoma <-rowMeans(sapply(rowmeans_ccle[skinpig_emt_ccle], function(x) x[rownames(expr_skinpig_emt_vitro)]))
rowmeans_tumor_melanoma<- rowMeans(sapply(rowmeans_tumor[skinpig_emt_tumor], function(x) x[rownames(expr_skinpig_emt_vitro)]))
expr_skinpig_emt_vitro_filt <- expr_skinpig_emt_vitro[order(rowMeans(cbind(rowmeans_ccle_melanoma, rowmeans_tumor_melanoma)), decreasing=T)[1:4500] ,]
expr_skinpig_emt_vivo_filt <- expr_skinpig_emt_vivo[order(rowMeans(cbind(rowmeans_ccle_melanoma, rowmeans_tumor_melanoma)), decreasing=T)[1:4500] ,]
# combine the two datasets
expr_skinpig_emt_final <- cbind(expr_skinpig_emt_vitro_filt, expr_skinpig_emt_vivo_filt)
# run PCA
expr_skinpig_emt_pca <- prcomp(t(expr_skinpig_emt_final))
# **************************************************************************
# Plot PCA results - HNSCC
# store pca coordinates in dataframe
expr_emt_episen_plot <- data.frame(expr_emt_episen_pca$x[,1:5])
# calculate program scores
expr_emt_episen_plot$emtII <- colMeans(expr_emt_episen_final[is.element(rownames(expr_emt_episen_final), meta_intersect$emtII_vitro$HNSCC.PEMT),])
expr_emt_episen_plot$episen=colMeans(expr_emt_episen_final[is.element(rownames(expr_emt_episen_final), meta_intersect$episen_vitro$`HNSCC.Epidif.1`),])
# add metadata
expr_emt_episen_plot$type <- rep(c("in vitro", "in vivo"), c(ncol(expr_emt_episen_vitro), ncol(expr_emt_episen_vivo)))
expr_emt_episen_plot$names <- c(rep(episen_emt_ccle, sapply(expr_ccle[episen_emt_ccle],ncol)),rep(episen_emt_tumor, sapply(expr_tumor[episen_emt_tumor],ncol)) )
expr_emt_episen_plot$cell_line <- NA
expr_emt_episen_plot$cell_line[expr_emt_episen_plot$type=="in vitro"] <- expr_emt_episen_plot$names[expr_emt_episen_plot$type=="in vitro"]
expr_emt_episen_plot$cell_line <- gsub("_UPPER_AERODIGESTIVE_TRACT", "", expr_emt_episen_plot$cell_line)
expr_emt_episen_plot$tumor <- NA
expr_emt_episen_plot$tumor[expr_emt_episen_plot$type=="in vivo"] <- expr_emt_episen_plot$names[expr_emt_episen_plot$type=="in vivo"]
# plot pca coordinates
pdf("Output/module4/HNSCC_pca_program.pdf", width = 3.9, height = 3.8)
ggplot(expr_emt_episen_plot, aes(x=PC2, y=PC4+PC5, color=emtII-episen)) +
geom_point( size=0.5) +
scale_color_gradient2(limits=c(-3, 3), midpoint = 0, low= c("#053061" ,"#2166AC" ,"#4393C3" ,"#92C5DE" ), mid= c("#D1E5F0", "gray95", "#FDDBC7"), high = c( "#F4A582", "#D6604D" ,"#B2182B" ,"#67001F") , oob=squish, name="Relative score\n(EMT II - Episen)") +
theme(panel.background = element_blank(), panel.border = element_rect(fill=F), legend.text.align = 0.5, axis.text=element_text(size=12), axis.title = element_text(size=13), legend.title.align = 0.5, legend.position = "top", legend.title = element_text(size=12, vjust = 1), legend.text = element_text(size=11), plot.margin=unit(c(0.1,2,0.1,0.1), "cm"))
dev.off()
pdf("Output/module4/HNSCC_pca_type.pdf", width = 3.95, height = 3.7)
ggplot(expr_emt_episen_plot, aes(x=PC2, y= PC4+PC5, color=type)) +
geom_point( size=0.5) +
scale_color_manual(values=c("goldenrod", "mediumpurple"), name="") +
theme(panel.background = element_blank(), panel.border = element_rect(fill=F), legend.text.align = 0.5, axis.text=element_text(size=12), axis.title = element_text(size=13), legend.title.align = 0.5, legend.position = "top", legend.title = element_text(size=13, vjust = 0.5), legend.text = element_text(size=11), legend.key = element_blank(), plot.margin=unit(c(0.1,2,0.1,0.1), "cm")) +
guides(color=guide_legend(override.aes = list(size=4)))
dev.off()
pdf("Output/module4/HNSCC_pca_cellline.pdf", width = 3.9, height = 3.8)
ggplot(expr_emt_episen_plot, aes(x=PC2, y= PC4 + PC5, color=cell_line)) +
geom_point(data=expr_emt_episen_plot[is.na(expr_emt_episen_plot$cell_line),], aes(x=PC2, y= PC4+PC5), size=0.5, color="gray90") +
geom_point( size=0.5) +
scale_color_brewer(palette = "Set2", name="", breaks=unique(expr_emt_episen_plot$cell_line)[!is.na(unique(expr_emt_episen_plot$cell_line))]) +
theme(panel.background = element_blank(), panel.border = element_rect(fill=F), legend.text.align = 0.5, axis.text=element_text(size=12), axis.title = element_text(size=13), legend.title.align = 0.5, legend.position = "top", legend.title = element_text(size=13, vjust = 0.5), legend.text = element_text(size=11), legend.key = element_blank(), plot.margin=unit(c(0.1,2,0.1,0.1), "cm")) +
guides(color=guide_legend(override.aes = list(size=3), nrow=2) )
dev.off()
pdf("Output/module4/HNSCC_pca_tumor.pdf", width = 3.9, height = 3.8)
ggplot(expr_emt_episen_plot, aes(x=PC2, y= PC4 + PC5, color=tumor)) +
geom_point( size=0.5) +
scale_color_brewer(palette = "Set2", name="", na.value="gray90", breaks=unique(expr_emt_episen_plot$tumor)[!is.na(unique(expr_emt_episen_plot$tumor))], labels=c("HNSCC 18","HNSCC 26", "HNSCC 25", "HNSCC 5", "HNSCC 17" )) +
theme(panel.background = element_blank(), panel.border = element_rect(fill=F), legend.text.align = 0.5, axis.text=element_text(size=12), axis.title = element_text(size=13), legend.title.align = 0.5, legend.position = "top", legend.title = element_text(size=13, vjust = 0.5), legend.text = element_text(size=11), legend.key = element_blank(), plot.margin=unit(c(0.1,2,0.1,0.1), "cm")) +
guides(color=guide_legend(override.aes = list(size=3), nrow=2))
dev.off()
# plot top genes with positive pca loadings
pca_loadings_pos_hnscc <- rev(melt(apply(expr_emt_episen_pca$rotation[,1:5], 2, function(x) names(sort(x, decreasing=T))[1:10]))$value)
pca_loadings_pos_hnscc_plot <- as.matrix(expr_emt_episen_pca$rotation[as.character(pca_loadings_pos_hnscc),1:5] )
rownames(pca_loadings_pos_hnscc_plot) <- NULL
pca_loadings_pos_hnscc_plot <- melt(as.matrix(pca_loadings_pos_hnscc_plot))
pdf("Output/module4/HNSCC_pca_pos.pdf", width = 4.5, height = 7.5)
ggplot(data = pca_loadings_pos_hnscc_plot, aes(x=Var2, y=factor(Var1), fill=value)) +
geom_tile() +
scale_fill_gradient2(limits=c(-0.12, 0.12), midpoint = 0, low= c("#053061" ,"#2166AC" ,"#4393C3" ,"#92C5DE" ), mid= c("#D1E5F0", "gray95", "#FDDBC7"), high = c( "#F4A582", "#D6604D" ,"#B2182B" ,"#67001F"), oob=squish, name="PC loadings")+
scale_color_gradient2(limits=c(-0.12, 0.12), midpoint = 0, low= c("#053061" ,"#2166AC" ,"#4393C3" ,"#92C5DE" ), mid= c("#D1E5F0", "gray95", "#FDDBC7"), high = c( "#F4A582", "#D6604D" ,"#B2182B" ,"#67001F"), oob=squish, name="PC loadings") +
scale_y_discrete(labels=pca_loadings_pos_hnscc) +
geom_hline(yintercept = seq(10.5,40.5,10)) +
geom_vline(xintercept = c(1.5,2.5,3.5, 4.5)) +
scale_x_discrete(expand=c(0,0)) +
labs(y="", x="") +
theme(panel.background = element_blank(), panel.border = element_rect(fill=F), axis.line = element_blank(), legend.title= element_text(size=11), legend.text = element_text(size=10), axis.text = element_text(size=11), legend.text.align = 1, axis.text.y = element_text(face= ifelse(is.element(pca_loadings_pos_hnscc, c(meta_intersect$emtII_vitro$HNSCC.PEMT,meta_intersect$episen_vitro$`HNSCC.Epidif.1`)), "bold","plain")))
dev.off()
# plot top genes with negative pca loadings
pca_loadings_neg_hnscc <- rev(melt(apply(expr_emt_episen_pca$rotation[,1:5], 2, function(x) names(sort(x, decreasing=F))[1:10]))$value)
pca_loadings_neg_hnscc_plot <- as.matrix(expr_emt_episen_pca$rotation[as.character(pca_loadings_neg_hnscc),1:5] )
rownames(pca_loadings_neg_hnscc_plot) <- NULL
pca_loadings_neg_hnscc_plot <- melt(as.matrix(pca_loadings_neg_hnscc_plot))
pdf("Output/module4/HNSCC_pca_neg.pdf", width = 4.5, height = 7.5)
ggplot(data = pca_loadings_neg_hnscc_plot, aes(x=Var2, y=factor(Var1), fill=value)) +
geom_tile() +
scale_fill_gradient2(limits=c(-0.12, 0.12), midpoint = 0, low= c("#053061" ,"#2166AC" ,"#4393C3" ,"#92C5DE" ), mid= c("#D1E5F0", "gray95", "#FDDBC7"), high = c( "#F4A582", "#D6604D" ,"#B2182B" ,"#67001F"), oob=squish, name="PC loadings")+
scale_color_gradient2(limits=c(-0.12, 0.12), midpoint = 0, low= c("#053061" ,"#2166AC" ,"#4393C3" ,"#92C5DE" ), mid= c("#D1E5F0", "gray95", "#FDDBC7"), high = c( "#F4A582", "#D6604D" ,"#B2182B" ,"#67001F"), oob=squish, name="PC loadings") +
scale_y_discrete(labels=pca_loadings_neg_hnscc) +
geom_hline(yintercept = seq(10.5,40.5,10)) +
geom_vline(xintercept = c(1.5,2.5,3.5, 4.5)) +
scale_x_discrete(expand=c(0,0)) +
labs(y="", x="") +
theme(panel.background = element_blank(), panel.border = element_rect(fill=F), axis.line = element_blank(), legend.title= element_text(size=11), legend.text = element_text(size=10), axis.text = element_text(size=11), legend.text.align = 1, axis.text.y = element_text(face= ifelse(is.element(pca_loadings_neg_hnscc, c(meta_intersect$emtII_vitro$HNSCC.PEMT,meta_intersect$episen_vitro$`HNSCC.Epidif.1`)), "bold","plain")))
dev.off()
# **************************************************************************
# Plot PCA results - melanoma
# store pca coordinates in a dataframe
expr_skinpig_emt_plot <- data.frame(expr_skinpig_emt_pca$x[,1:5] )
# calculate program scores
expr_skinpig_emt_plot$emtI <- colMeans(expr_skinpig_emt_final[is.element(rownames(expr_skinpig_emt_final), meta_intersect$emtI_vitro$melanoma.AXL.program),])
expr_skinpig_emt_plot$skinpig <- colMeans(expr_skinpig_emt_final[is.element(rownames(expr_skinpig_emt_final), meta_intersect$skinpig_vitro$melanoma.MITF.program),])
# add additional metadata
expr_skinpig_emt_plot$type <- rep(c("in vitro", "in vivo"), c(ncol(expr_skinpig_emt_vitro), ncol(expr_skinpig_emt_vivo)))
expr_skinpig_emt_plot$names <- c(rep(skinpig_emt_ccle, sapply(expr_ccle[skinpig_emt_ccle],ncol)),rep(skinpig_emt_tumor, sapply(expr_tumor[skinpig_emt_tumor],ncol)) )
expr_skinpig_emt_plot$cell_line <- NA
expr_skinpig_emt_plot$cell_line[expr_skinpig_emt_plot$type=="in vitro"] <- expr_skinpig_emt_plot$names[expr_skinpig_emt_plot$type=="in vitro"]
expr_skinpig_emt_plot$cell_line <- gsub("_SKIN", "", expr_skinpig_emt_plot$cell_line)
expr_skinpig_emt_plot$tumor <- NA
expr_skinpig_emt_plot$tumor[expr_skinpig_emt_plot$type=="in vivo"] <- expr_skinpig_emt_plot$names[expr_skinpig_emt_plot$type=="in vivo"]
# plot pca coordinates
pdf("Output/module4/melanoma_pca_program.pdf", width = 3.9, height = 3.8)
ggplot(expr_skinpig_emt_plot, aes(x=PC2+PC3, y=PC4, color=emtI-skinpig)) +
geom_point( size=0.5) +
scale_color_gradient2(limits=c(-3, 3), midpoint = 0, low= c("#053061" ,"#2166AC" ,"#4393C3" ,"#92C5DE" ), mid= c("#D1E5F0", "gray95", "#FDDBC7"), high = c( "#F4A582", "#D6604D" ,"#B2182B" ,"#67001F") , oob=squish, name="Relative score\n(EMT I - Skinpig)") +
theme(panel.background = element_blank(), panel.border = element_rect(fill=F), legend.text.align = 0.5, axis.text=element_text(size=12), axis.title = element_text(size=13), legend.title.align = 0.5, legend.position = "top", legend.title = element_text(size=12, vjust = 1), legend.text = element_text(size=11), plot.margin=unit(c(0.1,2,0.1,0.1), "cm"))
dev.off()
pdf("Output/module4/melanoma_pca_type.pdf", width = 3.95, height = 3.7)
ggplot(expr_skinpig_emt_plot, aes(x=PC2+PC3, y=PC4, color=type)) +
geom_point( size=0.5) +
scale_color_manual(values=c("goldenrod", "mediumpurple"), name="") +
theme(panel.background = element_blank(), panel.border = element_rect(fill=F), legend.text.align = 0.5, axis.text=element_text(size=12), axis.title = element_text(size=13), legend.title.align = 0.5, legend.position = "top", legend.title = element_text(size=13, vjust = 0.5), legend.text = element_text(size=11), legend.key = element_blank(), plot.margin=unit(c(0.1,2,0.1,0.1), "cm")) +
guides(color=guide_legend(override.aes = list(size=4)))
dev.off()
pdf("Output/module4/melanoma_pca_cellline.pdf", width = 3.9, height = 3.8)
ggplot(expr_skinpig_emt_plot, aes(x=PC2+PC3, y= PC4, color=cell_line)) +
geom_point(data=expr_skinpig_emt_plot[is.na(expr_skinpig_emt_plot$cell_line),], aes(x=PC2+PC3, y= PC4), size=0.5, color="gray90") +
geom_point( size=0.5) +
scale_color_brewer(palette = "Set2", name="", breaks=unique(expr_skinpig_emt_plot$cell_line)[!is.na(unique(expr_skinpig_emt_plot$cell_line))]) +
theme(panel.background = element_blank(), panel.border = element_rect(fill=F), legend.text.align = 0.5, axis.text=element_text(size=12), axis.title = element_text(size=13), legend.title.align = 0.5, legend.position = "top", legend.title = element_text(size=13, vjust = 0.5), legend.text = element_text(size=10.4), legend.key = element_blank(), plot.margin=unit(c(0.1,2,0.1,0.1), "cm")) +
guides(color=guide_legend(override.aes = list(size=3), nrow=2) )
dev.off()
pdf("Output/module4/melanoma_pca_tumor.pdf", width = 3.9, height = 3.8)
ggplot(expr_skinpig_emt_plot, aes(x=PC2+PC3, y= PC4, color=tumor)) +
geom_point( size=0.5) +
scale_color_brewer(palette = "Set2", name="", na.value="gray90", breaks=unique(expr_skinpig_emt_plot$tumor)[!is.na(unique(expr_skinpig_emt_plot$tumor))], labels=c("Mel 59","Mel 71", "Mel 78", "Mel 19", "Mel 80", "Mel 81", "Mel 88", "Mel 89" )) +
theme(panel.background = element_blank(), panel.border = element_rect(fill=F), legend.text.align = 0.5, axis.text=element_text(size=12), axis.title = element_text(size=13), legend.title.align = 0.5, legend.position = "top", legend.title = element_text(size=13, vjust = 0.5), legend.text = element_text(size=11), legend.key = element_blank(), plot.margin=unit(c(0.1,2,0.1,0.1), "cm")) +
guides(color=guide_legend(override.aes = list(size=3), nrow=2))
dev.off()
# plot top genes with positive pca loadings
pca_loadings_pos_mel <- rev(melt(apply(expr_skinpig_emt_pca$rotation[,1:5], 2, function(x) names(sort(x, decreasing=T))[1:10]))$value)
pca_loadings_pos_mel_plot <- as.matrix(expr_mitf_emt_pca$rotation[as.character(pca_loadings_pos_mel),1:5] )
rownames(pca_loadings_pos_mel_plot) <- NULL
pca_loadings_pos_mel_plot <- melt(as.matrix(pca_loadings_pos_mel_plot))
pdf("Output/module4/melanoma_pca_pos.pdf", width = 4.5, height = 7.5)
ggplot(data = pca_loadings_pos_mel_plot, aes(x=Var2, y=factor(Var1), fill=value)) +
geom_tile() +
scale_fill_gradient2(limits=c(-0.12, 0.12), midpoint = 0, low= c("#053061" ,"#2166AC" ,"#4393C3" ,"#92C5DE" ), mid= c("#D1E5F0", "gray95", "#FDDBC7"), high = c( "#F4A582", "#D6604D" ,"#B2182B" ,"#67001F"), oob=squish, name="PC loadings")+
scale_color_gradient2(limits=c(-0.12, 0.12), midpoint = 0, low= c("#053061" ,"#2166AC" ,"#4393C3" ,"#92C5DE" ), mid= c("#D1E5F0", "gray95", "#FDDBC7"), high = c( "#F4A582", "#D6604D" ,"#B2182B" ,"#67001F"), oob=squish, name="PC loadings") +
scale_y_discrete(labels=pca_loadings_pos_mel) +
geom_hline(yintercept = seq(10.5,40.5,10)) +
geom_vline(xintercept = c(1.5,2.5,3.5, 4.5)) +
scale_x_discrete(expand=c(0,0)) +
labs(y="", x="") +
theme(panel.background = element_blank(), panel.border = element_rect(fill=F), axis.line = element_blank(), legend.title= element_text(size=11), legend.text = element_text(size=10), axis.text = element_text(size=11), legend.text.align = 1, axis.text.y = element_text(face= ifelse(is.element(pca_loadings_pos_mel, c(meta_intersect$emtI_vitro$melanoma.AXL.program,meta_intersect$skinpig_vitro$melanoma.MITF.program)), "bold","plain")))
dev.off()
# plot top genes with negative pca loadings
pca_loadings_neg_mel <- rev(melt(apply(expr_skinpig_emt_pca$rotation[,1:5], 2, function(x) names(sort(x, decreasing=F))[1:10]))$value)
pca_loadings_neg_mel_plot <- as.matrix(expr_skinpig_emt_pca$rotation[as.character(pca_loadings_neg_mel),1:5] )
rownames(pca_loadings_neg_mel_plot) <- NULL
pca_loadings_neg_mel_plot <- melt(as.matrix(pca_loadings_neg_mel_plot))
pdf("Output/module4/melanoma_pca_neg.pdf", width = 4.5, height = 7.5)
ggplot(data = pca_loadings_neg_mel_plot, aes(x=Var2, y=factor(Var1), fill=value)) +
geom_tile() +
scale_fill_gradient2(limits=c(-0.12, 0.12), midpoint = 0, low= c("#053061" ,"#2166AC" ,"#4393C3" ,"#92C5DE" ), mid= c("#D1E5F0", "gray95", "#FDDBC7"), high = c( "#F4A582", "#D6604D" ,"#B2182B" ,"#67001F"), oob=squish, name="PC loadings")+
scale_color_gradient2(limits=c(-0.12, 0.12), midpoint = 0, low= c("#053061" ,"#2166AC" ,"#4393C3" ,"#92C5DE" ), mid= c("#D1E5F0", "gray95", "#FDDBC7"), high = c( "#F4A582", "#D6604D" ,"#B2182B" ,"#67001F"), oob=squish, name="PC loadings") +
scale_y_discrete(labels=pca_loadings_neg_mel) +
geom_hline(yintercept = seq(10.5,40.5,10)) +
geom_vline(xintercept = c(1.5,2.5,3.5, 4.5)) +
scale_x_discrete(expand=c(0,0)) +
labs(y="", x="") +
theme(panel.background = element_blank(), panel.border = element_rect(fill=F), axis.line = element_blank(), legend.title= element_text(size=11), legend.text = element_text(size=10), axis.text = element_text(size=11), legend.text.align = 1, axis.text.y = element_text(face= ifelse(is.element(pca_loadings_neg_mel, c(meta_intersect$emtI_vitro$melanoma.AXL.program,meta_intersect$skinpig_vitro$melanoma.MITF.program)), "bold","plain")))
dev.off()