tome_code/Section2_trajectory_Step2_permutation.R at main · rkelly712/tome_code · GitHub

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
#############################
#### Section2_trajectory ####
#############################

#### The scripts used for creating cellular trajectories

###########################
#### Step2_permutation ####
###########################

#### performing a permutation (shuffling the Anno labels for each connection) by 1,000 times

library(Seurat)
library(dplyr)
library(ggplot2)
library(Matrix)
library(htmlwidgets)
library(plotly)
library(monocle3)
library(FNN)
source("help_code/help_code.R")

work_path = ""

time_point = paste0("E", c(3.5, 4.5, 5.25, 5.5, 6.25, seq(6.5, 8.25, 0.25), "8.5a", "8.5b", seq(9.5, 13.5, 1)))

args = commandArgs(trailingOnly=TRUE)
kk = as.numeric(args[1])
time_i = time_point[kk]
time_j = time_point[kk+1]
print(time_i)
print(time_j)

emb = readRDS(paste0(work_path, "/", time_i, "_", time_j, "_umap3.rds"))
emb = data.frame(emb)
print(dim(emb))

if(ncol(emb) != 3){
    print(XXX)
}

anno1 = readRDS(paste0(work_path, "/seurat_object_", time_i, ".rds"))
anno1$Anno = as.vector(anno1$cell_state)
anno1 = anno1[,c("day", "Anno")]
anno1$day = "pre"
anno1$stage = time_i
anno2 = readRDS(paste0(work_path, "/seurat_object_", time_j, ".rds"))
anno2$Anno = as.vector(anno2$cell_state)
anno2 = anno2[,c("day", "Anno")]
anno2$day = "nex"
anno2$stage = time_j

permutation_times = 1000
k_neigh = 5

res = list()

for(rep_i in 1:permutation_times){

    anno1$state = anno1$Anno[sample(1:nrow(anno1))]
    anno2$state = anno2$Anno[sample(1:nrow(anno2))]

    anno = rbind(anno1, anno2)
    if(nrow(emb) != nrow(anno)){
        print("Error!")
        print(xxx)
    }
    pd = anno[rownames(emb),]

    emb_sub = emb
    pd_sub = pd

    irlba_pca_res_1 <- emb_sub[as.vector(pd_sub$day)=="pre",]
    irlba_pca_res_2 <- emb_sub[as.vector(pd_sub$day)=="nex",]
    pd_sub1 <- pd_sub[pd_sub$day == "pre",]
    pd_sub2 <- pd_sub[pd_sub$day == "nex",]

    pre_state_min = min(table(as.vector(pd_sub1$state)))

    if (pre_state_min < k_neigh & pre_state_min >= 3){
        k_neigh = pre_state_min
        print(k_neigh)
    }

    if (pre_state_min < 3){
        next
    }

    neighbors <- get.knnx(irlba_pca_res_1, irlba_pca_res_2, k = k_neigh)$nn.index

    tmp1 <- matrix(NA,nrow(neighbors),ncol(neighbors))
    for(i in 1:k_neigh){
        tmp1[,i] <- as.vector(pd_sub1$state)[neighbors[,i]]
    }
    state1 <- names(table(as.vector(pd_sub1$state)))
    state2 <- names(table(as.vector(pd_sub2$state)))

    tmp2 <- matrix(NA,length(state2),length(state1))
    for(i in 1:length(state2)){
        x <- c(tmp1[as.vector(pd_sub2$state)==state2[i],])
        for(j in 1:length(state1)){
            tmp2[i,j] <- sum(x==state1[j])
        }
    }
    tmp2 <- tmp2/apply(tmp2,1,sum)
    tmp2 <- data.frame(tmp2)
    row.names(tmp2) = state2
    names(tmp2) = state1

    res[[rep_i]] = tmp2

}

saveRDS(res, paste0(work_path, "/", time_i, "_", time_j, "_Knn_umap_permutation.rds"))

x = list()
z = NULL
for(cnt in 1:(length(time_point)-1)){
    time_i = time_point[cnt]
    time_j = time_point[cnt+1]

    dat = readRDS(paste0(work_path, "/", time_i, "_", time_j, "_Knn_umap_permutation.rds"))

    permutation_times = 1000
    y = NULL

    for(i in 1:permutation_times){
        y = c(y, as.vector(as.matrix(dat[[i]])))
    }

    x[[cnt]] = y
    z = c(z, y)

    print(paste0(time_i, ":", sum(y > 0.2)/length(y)))
}

print(sum(z >= 0.2)/length(z))
### 0.0089

library(ggplot2)
dat = data.frame(edge_weights_by_permutation = z)

p<-ggplot(dat, aes(x=edge_weights_by_permutation)) +
    geom_histogram(position="identity", alpha=0.5, binwidth=0.01) +
    geom_vline(xintercept = 0.2, colour = "red") +
    theme_classic(base_size = 15)