-
Notifications
You must be signed in to change notification settings - Fork 0
/
medicc.R
201 lines (178 loc) · 9.84 KB
/
medicc.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
get_medicc2_tree_and_heatmap <- function(tree_xml_file, cn_profile_file, groups=NULL, title=NA, bp_to_mb=F, vertical=F, ybreaksize=2) {
require(ggplot2)
require(ggtree)
require(RColorBrewer)
require(rphyloxml)
xmltree <- read_phyloxml(tree_xml_file)
tree <- phyloxml2phylo(xmltree)[[1]]
## extract confidence values from xmltree and add to tree
n_nodes <- length(xmltree[[1]]$node.meta)
extract_node_info <- function(node, xmltree) {
node_id <- xmltree[[1]]$edges$id[node]
l <- unlist(xmltree[[1]]$node.meta[node])
out <- as.list(xmltree[[1]]$edges[node,])
toadd <- as.list(l)
out <- append(out, toadd)
out
}
l <- lapply(1:n_nodes, extract_node_info, xmltree)
node_dat <- rbindlist(l, fill=T)
node_dat[isleaf==F, node_label:=name]
node_dat[!is.na(confidence) & !is.na(node_label),node_label_confidence:=paste0(node_label,' (',confidence,')')]
node_dat[is.na(confidence) & !is.na(name), node_label_confidence:=name]
tree$node.label <- node_dat$node_label_confidence[node_dat$isleaf==F]
tree$tip.label[tree$tip.label=='diploid'] <- 'N'
tree$node.label <- gsub('internal_','Int_',tree$node.label)
## plot the tree+confidences
if(vertical==T) {
p1 <- ggtree(tree,layout='rect',right=F) + geom_nodelab(angle=-90, hjust=-0.2, size=2.5, color='blue') + theme(legend.position='none')
if(!is.null(groups)) {
groups[grep('Normal',label), label:='N']
p1 <- p1 %<+% groups
p1 <- p1 + geom_tiplab(aes(color=group), angle=-90, hjust=-0.2) + scale_color_manual(values=group_cols, name='Tissue')
} else {
p1 <- p1 + geom_tiplab()
}
if(!is.na(title)) p1 <- p1 + ggtitle(label=title)
p1 <- p1 + ggplot2::xlim(0, max(p1$data$x)*1.15)
p1 <- p1 + coord_flip()
pd <- as.data.table(p1$data)
pd <- pd[order(y,decreasing=F),]
pd$x <- max(pd$x) - pd$x
pd$y[!is.na(pd$label)] <- seq(1,sum(!is.na(pd$label)))
p1$data <- pd
} else {
p1 <- ggtree(tree,layout='rect') + geom_nodelab(angle=0, hjust=0, size=2.5, color='blue') + theme(legend.position='none')
if(!is.null(groups)) {
groups[grep('Normal',label), label:='N']
p1 <- p1 %<+% groups
p1 <- p1 + geom_tiplab(aes(color=group)) + scale_color_manual(values=group_cols, name='Tissue')
} else {
p1 <- p1 + geom_tiplab()
}
if(!is.na(title)) p1 <- p1 + ggtitle(label=title)
p1 <- p1 + ggplot2::xlim(0, max(p1$data$x)*1.15)
pd <- as.data.table(p1$data)
pd <- pd[order(y),]
pd$y[!is.na(pd$label)] <- seq(1,sum(!is.na(pd$label)))
p1$data <- pd
}
chr <- get_chr_arms()$chr
scna <- fread(cn_profile_file)
scna$chrom <- gsub('chr','',scna$chrom)
if(bp_to_mb) {
scna[,start:=start/1e6]
scna[,end:=end/1e6]
}
scna[,segid:=paste0(chrom,':',start,'-',end)]
scna$segid <- factor(scna$segid, levels=unique(scna$segid))
scna <- merge(scna, chr[,c('chr','global_start'),with=F], by.x='chrom',by.y='chr', all.x=T)
scna[,global_seg_start:=global_start+start]
scna[,global_seg_end:=global_start+end]
scna$sample_id <- gsub('internal_','Int_',scna$sample_id)
scna[sample_id=='diploid', sample_id:='N']
## add in a check for LOH of either allele w.r.t. ancestor node
tmp <- as.data.table(p1$data)
tmp <- tmp[,c('parent','node','label'),with=F]
tmp$label <- sapply(tmp$label, function(s) strsplit(s,' ')[[1]][1])
tmp <- merge(tmp[,c('label','node','parent'),with=F], tmp[,c('label','node'),with=F], by.x='parent', by.y='node', all.x=T)
names(tmp) <- c('parent.node','current.label','current.node','parent.label')
tmp[,c('current.node','parent.node'):=NULL]
tmp <- tmp[!is.na(current.label),]
scna <- merge(scna, tmp, by.x='sample_id', by.y='current.label', all.x=T)
setnames(scna,'parent.label','parent_id')
toadd <- scna[,c('segid','sample_id','cn_a','cn_b'),with=F]
setnames(toadd,c('cn_a','cn_b'),c('cn_a_anc','cn_b_anc'))
scna <- merge(scna, toadd, by.x=c('segid','parent_id'), by.y=c('segid','sample_id'), all.x=T)
scna$is_loh <- F
scna[is.na(cn_a_anc), cn_a_anc:=1]
scna[is.na(cn_b_anc), cn_b_anc:=1]
scna[(cn_a_anc > 0 & cn_a == 0) | (cn_b_anc > 0 & cn_b == 0), is_loh:=T] ## assume that if the MR ancestor node is not available, the most recent ancestor is the normal
scna[is_wgd==F & (is_gain==T & is_loss==F), type:='Gain']
scna[is_wgd==F & (is_gain==F & is_loss==T), type:='Loss']
scna[is_wgd==F & (is_gain==T & is_loss==T), type:='Both']
scna[is_wgd==T & (is_gain==F & is_loss==F), type:='WGD']
scna[is_wgd==T & (is_gain==T & is_loss==F), type:='WGD+Gain']
scna[is_wgd==T & (is_gain==F & is_loss==T), type:='WGD+Loss']
scna[is_wgd==T & (is_gain==T & is_loss==T), type:='WGD+Both']
## get sample order
sample_order <- pd$label[!is.na(pd$label)]
sample_order <- sapply(sample_order, function(s) strsplit(s,' ')[[1]][1], USE.NAMES=F)
scna$sample_id <- factor(scna$sample_id, levels=sample_order)
scna$sample_id_int <- as.integer(scna$sample_id)
scna[is_wgd==T,WGD:='Yes']
scna[is_wgd==F,WGD:='No']
cols <- c('#f9d1d1','#d5e3ef','#b7dfb6','#cecece','#ec5e60','#5e97c6','#5eb75c')
names(cols) <- c('Gain','Loss','Both','WGD','WGD+Gain','WGD+Loss','WGD+Both')
max_cn <- max(c(scna$cn_a),c(scna$cn_b))
y_cn=seq(0, max_cn)
y_breaks <- 0.7*(y_cn / max(y_cn))+0.3/2
tmp <- data.table(ypos=rep(NA, length(y_breaks)*length(sample_order)))
sample_num <- rep(1, length(y_breaks))
global_y_breaks <- y_breaks+rep(0,length(y_breaks))
for(i in 1:(length(sample_order)-1)) {
y_cn <- c(y_cn, seq(0, max_cn))
sample_num <- c(sample_num, rep(i+1, length(y_breaks)))
global_y_breaks <- c(global_y_breaks, rep(i, length(y_breaks)) + y_breaks)
}
tmp$ypos <- global_y_breaks
tmp$cn <- y_cn
tmp$sample_num <- sample_num
if('ypos_a' %in% names(scna)) scna[,ypos_a:=NULL]
if('ypos_b' %in% names(scna)) scna[,ypos_b:=NULL]
scna <- merge(scna, tmp, by.x=c('sample_id_int','cn_a'), by.y=c('sample_num','cn'), all.x=T)
setnames(scna,'ypos','ypos_a')
scna <- merge(scna, tmp, by.x=c('sample_id_int','cn_b'), by.y=c('sample_num','cn'), all.x=T)
setnames(scna,'ypos','ypos_b')
jitter_rows <- which(scna$ypos_a==scna$ypos_b)
scna[jitter_rows, ypos_a:=ypos_a+0.02]
scna[jitter_rows, ypos_b:=ypos_b-0.02]
idvars <- names(scna)
idvars <- idvars[!idvars %in% c('cn_a','cn_b','ypos_a','ypos_b')]
scna_a <- scna[,c(idvars,'cn_a','ypos_a'),with=F]
scna_b <- scna[,c(idvars,'cn_b','ypos_b'),with=F]
setnames(scna_a,c('cn_a','ypos_a'),c('cn','ypos'))
scna_a$allele <- 'A'
setnames(scna_b,c('cn_b','ypos_b'),c('cn','ypos'))
scna_b$allele <- 'B'
scna <- rbind(scna_a, scna_b)
scna$type <- factor(scna$type, levels=c('Gain','Loss','Both','WGD','WGD+Gain','WGD+Loss','WGD+Both'))
scna[is_loh==T, LOH:='LOH']
scna[is_loh==F, LOH:='no LOH']
chr22 <- chr[chr %in% 1:22,]
allele_cols <- c('black','orange')
names(allele_cols) <- c('A','B')
tmp[,ybreaks:=ypos]
tmp <- tmp[tmp$cn %% ybreaksize == 0]
if(vertical) {
p2 <- ggplot(scna) +
scale_y_continuous(limits=c(0, max(scna$sample_id_int)), breaks=tmp$ypos, labels=tmp$cn, expand=c(0,0)) +
scale_x_continuous(breaks=chr22$global_midpoint, labels=chr22$chr, expand=c(0,0)) +
geom_rect(aes(xmin=global_seg_start,xmax=global_seg_end,ymin=sample_id_int-1, ymax=sample_id_int+0, fill=type, linewidth=LOH), color='black') +
geom_segment(aes(x=global_seg_start, xend=global_seg_end, y=ypos, yend=ypos, color=allele), linewidth=0.5) +
geom_hline(yintercept=seq(0,length(sample_order)), color='#bfbfbf', linetype='dashed',size=0.25) +
geom_vline(xintercept=c(0,chr22$global_end),linewidth=0.25,color='#bfbfbf') +
scale_color_manual(values=allele_cols, name='Parental copy') +
scale_fill_manual(values=cols, na.value='white',name='Change from\nparent node') +
scale_linewidth_manual(values=widths, name='LOH since\nparent node?') +
ang::theme_ang(base_size=9) +
coord_flip() +
theme(axis.text.y=element_text(angle=90, size=7,hjust=0.5), axis.line=element_blank(), legend.position='right') + labs(x='Genomic position', y='Copy number')
p <- plot_grid(p1, p2, ncol=1, align='v', rel_heights=c(2, 4), axis='lr')
} else {
p2 <- ggplot(scna) +
scale_y_continuous(limits=c(0, max(scna$sample_id_int)), breaks=tmp$ypos, labels=tmp$cn, expand=c(0,0)) +
scale_x_continuous(breaks=chr22$global_midpoint, labels=chr22$chr, expand=c(0,0)) +
geom_rect(data=scna[is_loh==F], aes(xmin=global_seg_start,xmax=global_seg_end,ymin=sample_id_int-1, ymax=sample_id_int+0, fill=type)) +
geom_rect(data=scna[is_loh==T], aes(xmin=global_seg_start,xmax=global_seg_end,ymin=sample_id_int-1, ymax=sample_id_int+0, fill=type), linewidth=0.35, color='black') +
scale_color_manual(values=allele_cols, name='Parental copy') +
scale_fill_manual(values=cols, na.value='white',name='Change from\nparent node') +
geom_segment(aes(x=global_seg_start, xend=global_seg_end, y=ypos, yend=ypos, color=allele), linewidth=0.5) +
geom_hline(yintercept=seq(0,length(sample_order)), color='#bfbfbf', linetype='dashed',size=0.25) +
geom_vline(xintercept=c(0,chr22$global_end),linewidth=0.25,color='#bfbfbf') +
ang::theme_ang(base_size=9) +
theme(axis.text.y=element_text(angle=90, size=7,hjust=0.5), axis.line=element_blank(), legend.position='right') + labs(x='Genomic position', y='Copy number')
p <- plot_grid(p1, p2, nrow=1, align='h', rel_widths=c(2, 4), axis='tb')
}
p
}