-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathcpgtest.R
162 lines (132 loc) · 6.81 KB
/
cpgtest.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
cpg_processed <- NormalizeData(cpg_processed, ) %>% FindVariableFeatures() %>% ScaleData()
vargenes <- VariableFeatures(object = cpg_processed)[grepl(pattern = "^Ig[klh]", VariableFeatures(object = cpg_processed))]
cpg_processed <- RunPCA(cpg_processed, features = vargenes)
cpg_processed <- RunUMAP(cpg_processed, dims = 1:20)
cpg_processed <- FindNeighbors(cpg_processed)
cpg_processed <- FindClusters(cpg_processed)
ind_clone <- names(which(table(cpg_processed$clone_id)>=3))
ind_clone <- ind_clone[ind_clone != "No_contig"]
distance_calculation_pca <- function(srt, ind){
df <- srt@[email protected]
df <- df[colnames(srt)[srt$clone_id == ind],]
temp <- sapply(1:19, function(dim1){
x <- dim1
y <- dim1 +1
test <- sapply(seq_along(rownames(df)), function(r){
x1 = df[,x][r]
y1 = df[,y][r]
vx = setdiff(df[,x], x1)
vy = setdiff(df[,y],y1 )
x_diff <- sapply(abs(vx - x1), function(x) x^2)
y_diff <- sapply(abs(vy - y1), function(x) x^2)
sqr_dist = x_diff + y_diff
return(sum(sqrt(sqr_dist))/length(vx))
})
return(structure(test, names = rownames(df)))
})
return(temp)
}
perclone_test <- lapply(ind_clone, function(ind){
temp <- distance_calculation_pca(cpg_processed, ind)
return(rowMeans(temp))
})
names(perclone_test) <- ind_clone
df <- rbindlist(lapply(perclone_test, function(x) as.data.frame(x) %>% `colnames<-`("dist")%>% mutate(rep = substr(names(x), 1, 3))), idcol = "clonetype")
head(df)
## ignore replicates
df <- df %>% group_by(clonetype) %>% mutate(median = median(dist))
df$clonetype <- factor(df$clonetype, levels = unique(df$clonetype[order(df$median, decreasing = T)]))
df <- df[order(df$clonetype, decreasing = F), ]
## count in replicates
df <- df %>% group_by(clonetype, rep) %>% mutate(median = median(dist))
df$clonetype <- factor(df$clonetype, levels = unique(df$clonetype[order(df$median, decreasing = T)]))
df <- df[order(df$clonetype, decreasing = F), ]
median_df <- df[, c("clonetype", "median", "rep")] %>% distinct()
median_df$size <- sapply(perclone_test, length)[median_df$clonetype]
median_df$category <- ifelse(median_df$size<8, "3-7", ifelse(median_df$size>7 & median_df$size<13 ,"8-12", "13-21") )
ggplot(median_df,aes(x = median, colour = category)) + geom_density(linetype = rep)
DimPlot(cpg_processed, reduction = "UMAP", cells.highlight = colnames(cpg_processed)[cpg_processed$clone_id=="B_107_6_7_173_1_5"])+
ggtitle("B_107_6_7_173_1_5")
hist(as.numeric(perclone_test[["B_39_2_1_147_1_1"]]), prob = TRUE, breaks = 50, main = paste("B_39_2_1_147_1_1", "\nmedian at", median(perclone_test[["B_39_2_1_147_1_1"]])))
lines(density(perclone_test[["B_39_2_1_147_1_1"]]), col = 4, lwd = 2)
ggplot(df %>% filter(clonetype %in% head(levels(df$clonetype), 10)), aes(x = clonetype, y = dist))+
geom_boxplot()+
theme(axis.text.x = element_text(angle = 90, hjust = 1, vjust = 0.5))+
ggtitle("Top 10 clones with most heterogeneity")
ggplot(df %>% filter(clonetype %in% tail(levels(df$clonetype), 10)), aes(x = clonetype, y = dist))+
geom_boxplot()+
theme(axis.text.x = element_text(angle = 90, hjust = 1, vjust = 0.5))+
ggtitle("Top 10 clones with most homogeneity")
ggplot(rbind(df %>% filter(clonetype %in% head(levels(df$clonetype), 10)), df %>% filter(clonetype %in% tail(levels(df$clonetype), 10))), aes(x = clonetype, y = dist))+
geom_boxplot()+
theme(axis.text.x = element_text(angle = 90, hjust = 1, vjust = 0.5))+
ggtitle("Top and bottom 10 clones with most heterogeneity")
new_cells <- [email protected] %>% filter(clone_id != "No_contig") %>% rownames_to_column("cell_id") %>% select(cell_id, clone_id)
new_cells %>% group_by(clone_id) %>% slice_sample(n = 1)
median_df <- df[, c("clonetype", "median")] %>% distinct()
median_df$size <- sapply(perclone_test, length)[median_df$clonetype]
median_df$category <- ifelse(median_df$size<8, "3-7", ifelse(median_df$size>7 & median_df$size<13 ,"8-12", "13-21") )
f1 <- ggplot(median_df,aes(x = median, colour = category)) + geom_density()
distance_calculation_pca <- function(srt, new_cells, n, perclone_test){
df <- srt@[email protected]
control_distribution <- sapply(3:21, function(n){
controls <- new_cells %>% group_by(clone_id) %>% slice_sample(n = 1) %>% pull(cell_id)
controls <- sample(controls, n)
subdf <- df[controls,]
temp <- sapply(1:19, function(dim1){
x <- dim1
y <- dim1 +1
test <- sapply(seq_along(rownames(subdf)), function(r){
x1 = subdf[,x][r]
y1 = subdf[,y][r]
vx = setdiff(subdf[,x], x1)
vy = setdiff(subdf[,y],y1 )
x_diff <- sapply(abs(vx - x1), function(x) x^2)
y_diff <- sapply(abs(vy - y1), function(x) x^2)
sqr_dist = x_diff + y_diff
return(sum(sqrt(sqr_dist))/length(vx))
})
return(structure(test, names = rownames(subdf)))
})
return(rowMeans(temp))
})
}
df <- srt@[email protected]
control_distribution <- sapply(as.numeric(sapply(perclone_test, length)), function(n){
controls <- sample(colnames(cpg_processed)[cpg_processed$clone_id!="No_contig"], n)
subdf <- df[controls,]
temp <- sapply(1:19, function(dim1){
x <- dim1
y <- dim1 +1
test <- sapply(seq_along(rownames(subdf)), function(r){
x1 = subdf[,x][r]
y1 = subdf[,y][r]
vx = setdiff(subdf[,x], x1)
vy = setdiff(subdf[,y],y1 )
x_diff <- sapply(abs(vx - x1), function(x) x^2)
y_diff <- sapply(abs(vy - y1), function(x) x^2)
sqr_dist = x_diff + y_diff
return(sum(sqrt(sqr_dist))/length(vx))
})
return(structure(test, names = rownames(subdf)))
})
return(rowMeans(temp))
})
names(control_distribution) <- paste("control", names(perclone_test), sep = "_")
contr_df <- rbindlist(lapply(control_distribution, function(x) as.data.frame(x) %>% `colnames<-`("dist")), idcol = "control")
contr_df <- contr_df %>% group_by(control) %>% mutate(median = median(dist))
contr_df$control <- factor(contr_df$control, levels = unique(contr_df$control[order(contr_df$median, decreasing = T)]))
contr_df <- contr_df[order(contr_df$control, decreasing = F), ]
ggplot(head(contr_df,10), aes(x = control, y = dist))+
geom_boxplot()+
theme(axis.text.x = element_text(angle = 90, hjust = 1, vjust = 0.5))+
ggtitle("control clone distance distribution")
median_contr_df <- contr_df[, c("control", "median")] %>% distinct()
median_contr_df$size <- sapply(perclone_test, length)[median_contr_df$control]
median_contr_df$category <- ifelse(median_contr_df$size<8, "3-7", ifelse(median_contr_df$size>7 & median_contr_df$size<13 ,"8-12", "13-21") )
median_df$group <- "real"
median_contr_df$group <- "control"
colnames(median_contr_df) <- colnames(median_df)
all_median_df <- rbind(median_df, median_contr_df)
f2 <- ggplot(all_median_df,aes(x = median, colour = category, linetype = group)) + geom_density()
f2+ggtitle("real vs control intraclone heterogeneity")