-
Notifications
You must be signed in to change notification settings - Fork 1
/
Copy pathFigure2A-E-TMB-Downsampling.R
280 lines (232 loc) · 14.1 KB
/
Figure2A-E-TMB-Downsampling.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
library(RColorBrewer)
library(Rmisc)
library(purrr)
library(grid)
library(gridExtra)
library(gsheet)
library(ggpubr)
library(cowplot)
library(gridExtra)
source('/ifs/res/taylorlab/chavans/scripts-orphan/multiplot.R')
specify_decimal = function(x, k) format(round(x, k), nsmall=k)
"%ni%" = Negate("%in%")
curr_date = format(Sys.Date(),"%d-%b-%y")
ex_clin = fread('~/tempo-cohort-level/WES_metadata_040620.txt')
dim(ex_clin); head(ex_clin)
ex_clin2 = fread('~/tempo-cohort-level/WES_allTMBs_051020.txt') %>%
select(Tumor_Sample_Barcode, TMBWES_IMgenes) %>%
mutate(TMBWES_IMgenes = ifelse(is.na(TMBWES_IMgenes),0,TMBWES_IMgenes))
ex_clin3 = inner_join(ex_clin, ex_clin2, by = c(Tumor_Sample_Barcode = "Tumor_Sample_Barcode"))
im_clin = fread('~/tempo-cohort-level/IM_metadata_040620.txt') %>%
arrange(desc(TMBIMPACT)) %>%
select(CMO, TMBIMPACT, TMBIMPACT_Downsampled);
dim(im_clin); head(im_clin)
im_ex_clin = inner_join(ex_clin3, im_clin, by = c(Tumor_Sample_Barcode = "CMO"))
im_ex_clin0 = im_ex_clin %>%
select(DMP, TMBWES = TMB, TMBIMPACT, TMBWES_IMgenes, TMBWES_NonIMgenes, TMBIMPACT_Downsampled, Purity_Reviewed) %>%
filter(TMBIMPACT<10, Purity_Reviewed >=0.5) %>%
mutate(adj.depth = TMBIMPACT_Downsampled/TMBIMPACT,
adj.genecontent = TMBWES_NonIMgenes/TMBWES_IMgenes,
adj.TMBIMPACT = TMBIMPACT*adj.depth*adj.genecontent) %>%
select(-c("adj.depth","adj.genecontent", "Purity_Reviewed"))
dim(im_ex_clin0) #679 7
my_comparisons <- list( c("TMBIMPACT","TMBIMPACT_Downsampled"),
c("TMBWES_IMgenes","TMBWES_NonIMgenes"),
c("TMBWES","TMBWES_IMgenes"),
c("TMBIMPACT","TMBWES"),
c("TMBWES","TMBWES_NonIMgenes"),
c("TMBIMPACT","adj.TMBIMPACT"),
c("TMBWES","adj.TMBIMPACT"),
c("TMBIMPACT","TMBWES_IMgenes") )
pdatm = melt(im_ex_clin0, id = c("DMP"))
pdatm = pdatm %>% mutate(group = variable, TMB = value) %>% select(-c(variable,value))
head(pdatm)
mm_tmbs_ = pdatm
mm_tmbs_$group = factor(mm_tmbs_$group, levels = c("TMBIMPACT","TMBIMPACT_Downsampled","TMBWES_IMgenes","TMBWES_NonIMgenes","adj.TMBIMPACT","TMBWES"))
mm_tmbs_ = mm_tmbs_ %>% mutate(seq = ifelse(group %like% "WES", "WES", "IM"))
head(mm_tmbs_)
p14 = ggplot(mm_tmbs_, aes(x=group, y=TMB, fill = seq)) + theme_classic(base_size = 16) +
geom_boxplot() + ylab("TMB in Purity >50% \n(TMBIMPACT <10 Samples Only)") + xlab("") +
scale_fill_manual(values = c('gray35','gray65')) +
theme(axis.text.x = element_text(angle = 90, hjust = 1))
p14 + stat_compare_means(aes(label = ..p.signif..), method = "wilcox.test", comparisons = my_comparisons)
########### For Roslin data TMB boxplot to show adj.TMB
im_dnsmpl_tmb = fread('~/tempo-cohort-level/fixed.downsampledTMB_ForRoslin.txt') %>%
select(DMP, TMBIMPACT_Downsampled = TMBIMPACT.downsampled.fixed) %>%
filter(DMP %in% im_ex_clin$DMP)
dim(im_dnsmpl_tmb) #1413
url <- 'docs.google.com/spreadsheets/d/1ZQCZ-02b8VNNDL05w-FdUiTpIVJ0WAW0EwxCkgX3oM0'
ex_roslin <- read.csv(text=gsheet2text(url, format='csv'), stringsAsFactors=FALSE) %>%
select(DMP = DMPID, TMBWES = TMBExome, TMBIMPACT, TMBWES_IMgenes = TMBExomeIMgenes, TMBWES_NonIMgenes = TMBExome_NonIMgenes, Purity_Reviewed = PurityExome) %>%
filter(DMP %in% im_ex_clin$DMP)
dim(ex_roslin) #1636
im_ex_clin_roslin = inner_join(im_dnsmpl_tmb, ex_roslin, by = 'DMP')
dim(im_ex_clin_roslin) #1413 -- figure out why rest are missing?!
im_ex_clin0_roslin = im_ex_clin_roslin %>%
select(DMP, TMBWES, TMBIMPACT, TMBWES_IMgenes, TMBWES_NonIMgenes, TMBIMPACT_Downsampled, Purity_Reviewed) %>%
filter(TMBIMPACT<10, Purity_Reviewed >=0.5) %>%
mutate(adj.depth = TMBIMPACT_Downsampled/TMBIMPACT,
adj.genecontent = TMBWES_NonIMgenes/TMBWES_IMgenes,
adj.TMBIMPACT = TMBIMPACT*adj.depth*adj.genecontent) %>%
select(-c("adj.depth","adj.genecontent", "Purity_Reviewed"))
dim(im_ex_clin0_roslin) #601 7
pdatm = melt(im_ex_clin0_roslin, id = c("DMP"))
pdatm = pdatm %>% mutate(group = variable, TMB = value) %>% select(-c(variable,value))
head(pdatm)
mm_tmbs_ = pdatm
mm_tmbs_$group = factor(mm_tmbs_$group, levels = c("TMBIMPACT","TMBIMPACT_Downsampled","TMBWES_IMgenes","TMBWES_NonIMgenes","adj.TMBIMPACT","TMBWES"))
mm_tmbs_ = mm_tmbs_ %>% mutate(seq = ifelse(group %like% "WES", "WES", "IM"))
head(mm_tmbs_)
p15 = ggplot(mm_tmbs_, aes(x=group, y=TMB, fill = seq)) + theme_classic(base_size = 16) +
geom_boxplot() + ylab("TMB in Purity >50% \n(TMBIMPACT <10 Samples Only)") + xlab("") +
scale_fill_nejm() +
theme(axis.text.x = element_text(angle = 90, hjust = 1),
plot.margin = unit(c(1,1,1,1), 'lines'))
p15 + stat_compare_means(aes(label = ..p.signif..), method = "wilcox.test", comparisons = my_comparisons)
pdf("~/tempo-cohort-level/Figure2B_TMB_Roslin.pdf", paper = "a4r")
p15 + stat_compare_means(aes(label = ..p.signif..), method = "wilcox.test", comparisons = my_comparisons)
dev.off()
######### For Roslin showing TMB IMPACT vs. WES
im_clin = fread('~/tempo-cohort-level/IM_metadata_040620.txt') %>% select(DMP, TMBIMPACT)
im_ex_clin_roslin1 = inner_join(im_clin, ex_roslin, by = c('DMP')) %>%
select(DMP, TMBIMPACT=TMBIMPACT.x, TMBWES)
dim(im_ex_clin_roslin1) #1636
# Equation
label = lm_eqn(im_ex_clin_roslin,y=im_ex_clin_roslin$TMBIMPACT,x=im_ex_clin_roslin$TMBWES)
label
# R²
r2 = summary(lm(TMBWES~TMBIMPACT,data=im_ex_clin_roslin))$adj.r.squared
r2
#Spearman Correlation
corr = cor.test(im_ex_clin_roslin$TMBWES, im_ex_clin_roslin$TMBIMPACT,method = "spearman")
#print(str(corr))
rho = as.numeric(corr$estimate)
rho
### scatter plot LINEAR scale Original
# Plot
im_ex_clin_roslin_ = im_ex_clin_roslin1 %>%
mutate(TMBIMPACT = ifelse(is.na(TMBIMPACT),0,TMBIMPACT),
TMBIMPACT = TMBIMPACT + 1, TMBWES = TMBWES +1 ) %>%
mutate(grouptmb = ifelse(TMBIMPACT>=20 | TMBWES>=20,'High-TMB','Low-TMB'))
im_ex_clin_roslin_$grouptmb = as.factor(im_ex_clin_roslin_$grouptmb)
overall = ggplot(im_ex_clin_roslin_, aes(x = TMBIMPACT, y = TMBWES)) +
geom_point(pch = 21, size = 2, fill = 'lightgray') +
#scale_fill_manual(values = c('lightgreen','lightgray')) +
#scale_color_manual(values = c('red','blue')) +
geom_abline(intercept = 0, slope = 1, lty = 2) +
geom_smooth(method=lm, se=FALSE, lty=1) +
scale_y_continuous(trans = 'log10', limits = c(1,500), breaks = c(0,1,2,5,10,20,50,100,200,500)) +
scale_x_continuous(trans='log10', limits = c(1,500), breaks = c(0,1,2,5,10,20,50,100,200,500)) +
coord_fixed() + xlab('TMBIMPACT + 1') + ylab('TMBWES + 1') +
labs(title = paste0("R^2 = ",specify_decimal(r2,3),"\n")) +
theme_classic(base_size = 14) +
theme(legend.title = element_blank(), legend.position = 'right', plot.margin = unit(c(1,1,1,1), 'lines'))
#overall
r2_ = summary(lm(TMBWES~TMBIMPACT,data=filter(im_ex_clin_roslin_, grouptmb == 'Low-TMB')))$adj.r.squared
r2_
overall_2 = ggplot(im_ex_clin_roslin_, aes(x = TMBIMPACT, y = TMBWES, fill = grouptmb)) +
geom_point(pch = 21, size = 2) +
scale_fill_manual(values = c('lightgreen','lightgray')) +
scale_color_manual(values = c('red','blue')) +
geom_abline(intercept = 0, slope = 1, lty = 2) +
geom_smooth(method=lm, se=FALSE, lty=1, aes(group=grouptmb, color=grouptmb)) +
scale_y_continuous(trans = 'log10', limits = c(1,500), breaks = c(0,1,2,5,10,20,50,100,200,500)) +
scale_x_continuous(trans='log10', limits = c(1,500), breaks = c(0,1,2,5,10,20,50,100,200,500)) +
coord_fixed() + xlab('TMBIMPACT + 1') + ylab('TMBExome + 1') +
labs(title = paste0("R^2 = ",specify_decimal(r2_,3),"\n")) +
theme_classic(base_size = 14) +
theme(legend.title = element_blank(),legend.position = 'bottom', plot.margin = unit(c(1,1,1,1), 'lines'))
#overall_2
grid.arrange(overall, overall_2, newpage = FALSE, ncol = 2)
# tmb_plot = ggplot(im_ex_clin_roslin,
# aes(x = TMBIMPACT, y = TMBWES)) +
# #facet_wrap(~Algorithm) +
# geom_point(alpha = 0.5, size = 3, shape = 16, color = "#08519c") +
# geom_abline(col = "blue") +
# scale_x_continuous(limits = c(0,150)) +
# scale_y_continuous(limits = c(0,150)) +
# labs(x = "IMPACT TMB", y = "Exome TMB") +
# geom_point(alpha = 0.3, size = 3, shape = 16) +
# geom_abline(col = "black",slope=1, intercept =0) +
# geom_smooth(method=lm, se=FALSE, lty=2) +
# coord_fixed() +
# theme_classic(base_size = 16) +
# theme(axis.text.x = element_text(colour = "blue")) +
# theme(axis.text.y = element_text(colour = "blue")) +
# theme(plot.title = element_text(colour = "blue", size = 14)) +
# labs(title = paste0("Non-Synonymous Coding Mutations-based-TMB","\n",
# " R^2 = ",specify_decimal(r2,3),"\n",
# " Spearman Correlation = ",specify_decimal(rho,3),"\n",
# " IMPACT_TMB = 1.2*Exome_TMB + 1.9"))
pdf("~/tempo-cohort-level/Figure2A_TMB_Compare_Roslin.pdf", paper = "a4r")
grid.arrange(overall, overall_2, newpage = FALSE, ncol = 2)
dev.off()
#########################################################################################################################################################################
url <- 'docs.google.com/spreadsheets/d/1ZQCZ-02b8VNNDL05w-FdUiTpIVJ0WAW0EwxCkgX3oM0'
ex_roslin_purity_depth <- read.csv(text=gsheet2text(url, format='csv'), stringsAsFactors=FALSE)
ex_roslin_purity_depth0 = ex_roslin_purity_depth %>%
select(DMP = DMPID, TMBWES = TMBExome, purity_bin, depth_bin, TMBWESClonal = TMBExomeClonal) %>%
filter(DMP %in% im_ex_clin$DMP)
dim(ex_roslin_purity_depth0) #1636
# > filter(ex_roslin_purity_depth0, is.na(purity_bin))
# DMP TMBWES purity_bin depth_bin TMBWESClonal
# 1 P-0020573-T01-IM6 0 <NA> (100,200] 0
# 2 P-0020769-T01-IM6 0 <NA> (100,200] 0
# 3 P-0021159-T01-IM6 0 <NA> (200,300] 0
# 4 P-0021256-T01-IM6 0 <NA> (100,200] 0
# 5 P-0028349-T01-IM6 0 <NA> (100,200] 0
##Panel B : Effect of purity, and clonality on TMB
#mm_se1 = ddply(mm_, c("purity_bin"), summarise, N = length(TMBExome),mean = mean(TMBExome),sd = sd(TMBExome),se = sd / sqrt(N))
#mm_se = summarySE(mm_, measurevar="TMBExome", groupvars=c("purity_bin"))
mm_se_low = filter(ex_roslin_purity_depth0, !is.na(purity_bin), TMBWES<10) %>% summarySE(., measurevar="TMBWES", groupvars=c("purity_bin"))
mm_se_high = filter(ex_roslin_purity_depth0, !is.na(purity_bin), TMBWES>=10) %>% summarySE(., measurevar="TMBWES", groupvars=c("purity_bin"))
p3 = ggplot(mm_se_low, aes(x=purity_bin, y=TMBWES)) + theme_classic(base_size = 16) +
geom_point(size=5, shape=21, fill="white") + ylab("TMB WES\n(<10)") +
scale_y_continuous(limits=c(0,5),breaks = c(1,2,3,4,5)) +
scale_x_discrete(labels = c(0.1,0.2,0.3,0.4,0.5,0.6,0.7,0.8,0.9,1)) +
geom_errorbar(aes(ymin=TMBWES-ci, ymax=TMBWES+ci), colour="black", width=.1)
p4 = ggplot(mm_se_high, aes(x=purity_bin, y=TMBWES)) + theme_classic(base_size = 16) +
geom_point(size=5, shape=21, fill="white") + ylab("TMB WES\n(>=10)") +
scale_y_continuous(limits=c(10,100),breaks = c(10,20,40,60,80,100)) +
scale_x_discrete(labels = c(0.1,0.2,0.3,0.4,0.5,0.6,0.7,0.8,0.9)) +
geom_errorbar(aes(ymin=TMBWES-ci, ymax=TMBWES+ci), colour="black", width=.1)
grid.arrange(p3,p4, ncol = 2, widths = c(0.5,0.5), newpage = F)
## Clonal TMB
mm_se_low_cl = filter(ex_roslin_purity_depth0, !is.na(purity_bin), TMBWESClonal<10) %>% summarySE(., measurevar="TMBWESClonal", groupvars=c("purity_bin"))
mm_se_high_cl = filter(ex_roslin_purity_depth0, !is.na(purity_bin), TMBWESClonal>=10) %>% summarySE(., measurevar="TMBWESClonal", groupvars=c("purity_bin"))
p5 = ggplot(mm_se_low_cl, aes(x=purity_bin, y=TMBWESClonal)) + theme_classic(base_size = 16) +
geom_point(size=5, shape=21, fill="white") + ylab("Clonal TMB WES\n(<10)") +
scale_y_continuous(limits=c(0,5),breaks = c(1,2,3,4,5)) +
scale_x_discrete(labels = c(0.1,0.2,0.3,0.4,0.5,0.6,0.7,0.8,0.9,1)) +
geom_errorbar(aes(ymin=TMBWESClonal-ci, ymax=TMBWESClonal+ci), colour="black", width=.1)
p6 = ggplot(mm_se_high_cl, aes(x=purity_bin, y=TMBWESClonal)) + theme_classic(base_size = 16) +
geom_point(size=5, shape=21, fill="white") + ylab("Clonal TMB WES\n(>=10)") +
scale_y_continuous(limits=c(0,125),breaks = c(20,40,60,80,100)) +
scale_x_discrete(labels = c(0.1,0.2,0.3,0.4,0.5,0.6,0.7,0.8,0.9)) +
geom_errorbar(aes(ymin=TMBWESClonal-ci, ymax=TMBWESClonal+ci), colour="black", width=.1)
grid.arrange(p5,p6, ncol = 2, widths = c(0.5,0.5), newpage = F)
#########################################################################################################################################################################
#Panel C : Effect of depth on TMB
#Hist with a line
p<-ggplot(ex_roslin_purity_depth0, aes(x=TMBWES)) + theme_classic(base_size = 16) +
geom_histogram(color="black", fill="white", bins = 100) + scale_x_continuous(breaks = c(10,50,100,150,200,250,300,350))
p7 = p + geom_vline(aes(xintercept=10),color="red", lty=2);
p7
mm_se_low = filter(ex_roslin_purity_depth0, TMBWES<10, !is.na(depth_bin)) %>% summarySE(., measurevar="TMBWES", groupvars=c("depth_bin"))
mm_se_high = filter(ex_roslin_purity_depth0, TMBWES>=10, !is.na(depth_bin)) %>% summarySE(., measurevar="TMBWES", groupvars=c("depth_bin"))
#TMB by Sequencing depth (update for Exome depth)
p9 = ggplot(mm_se_low, aes(x=depth_bin, y=TMBWES)) + theme_classic(base_size = 16) +
geom_bar(stat="identity") + ylab("TMB WES\n(<10)") +
#scale_y_continuous(limits=c(0,5),breaks = c(1,2,3,4,5)) +
#scale_x_discrete(labels = c(50,100,150,200,250,300,350)) +
geom_errorbar(aes(ymin=TMBWES-ci, ymax=TMBWES+ci), colour="black", width=.1)
p10 = ggplot(mm_se_high, aes(x=depth_bin, y=TMBWES)) + theme_classic(base_size = 16) +
geom_bar(stat="identity") + ylab("TMB WES\n(>=10)") +
#scale_y_continuous(limits=c(0,120),breaks = c(10,20,40,60,80,100,120)) +
#scale_x_discrete(labels = c(50,100,150,200,250,300,350)) +
geom_errorbar(aes(ymin=TMBWES-ci, ymax=TMBWES+ci), colour="black", width=.1)
pdf('')
grid.arrange(p9,p10, ncol = 2, widths = c(0.5,0.5), newpage = F)
pdf("~/tempo-cohort-level/Figure2CDE_Purity_Depth_TMB_Roslin.pdf", paper = "a4")
grid.arrange(p3, p4, p5, p6, p9,p10, ncol = 2, nrow =3, widths = c(0.5,0.5), newpage = F)
dev.off()