-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathvisualization.R
157 lines (121 loc) · 6.09 KB
/
visualization.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
# Visualization function for cell interactions
visualization <- function(input_file, output_dir, tr, ir, hr) {
# Load cell data
cell_data <- read.csv(input_file)
cell_type_list <<- sort(unique(cell_data[[3]]))
cell_type_num <- length(cell_type_list)
input_file_name <- basename(tools::file_path_sans_ext(input_file))
filename <- file.path(output_dir, paste0(input_file_name, "_model_TR_", tr, "_IR_", ir, "_HR_", hr, ".Rda"))
load(file = filename)
# Process coefficient intensities
coef_intensity <- coef[1:cell_type_num]
coef_intensity <- ifelse(1:cell_type_num == 1,
coef_intensity[1:cell_type_num],
coef_intensity[1:cell_type_num] + coef_intensity[1])
coef_interaction <- coef[(cell_type_num + 1):length(coef)]
# Process cell type links and nodes
cell_type_links = process_links(coef_interaction, cell_type_num, tr, ir)
cell_type_nodes_list = process_nodes(coef_interaction, cell_type_num)
cell_type_nodes = cell_type_nodes_list[[1]]
cell_type_self_interaction_links = cell_type_nodes_list[[2]]
# Further process nodes and save visualizations
process_and_visualize("pairwise", cell_type_links, cell_type_nodes, input_file_name, output_dir, tr, ir, hr)
process_and_visualize("self", cell_type_self_interaction_links, cell_type_nodes, input_file_name, output_dir, tr, ir, hr)
}
# Initializes an empty data frame with given column names and number of columns
initialize_dataframe <- function(ncols, col_names, nrow=NULL) {
df <- data.frame(matrix(ncol = ncols))
colnames(df) <- col_names
return(df)
}
# Process interactions between different cell types
process_links <- function(coef_interaction, cell_type_num, tr, ir) {
cell_type_links <- initialize_dataframe(4, c('from', 'to', 'weight', 'sign'))
interaction_names <- names(coef_interaction)
idx <- 1
for (i in 1:(cell_type_num-1)){
for (j in (i+1):(cell_type_num)){
for (r in seq(ir, tr, ir)){
current_interaction_name <- paste('InteractionmarkX', cell_type_list[i], 'xX', cell_type_list[j], 'x', r, sep = '')
current_coef <- coef_interaction[grep(current_interaction_name, interaction_names)]
current_sign <- ifelse(sign(current_coef) == 1, 'darkblue', 'darkred')
cell_type_links[idx,] <- c(cell_type_list[i], cell_type_list[j], abs(current_coef), current_sign)
idx <- idx + 1
}
}
}
return(cell_type_links)
}
# Process individual cell types' interactions and calculates influence
process_nodes <- function(coef_interaction, cell_type_num) {
cell_type_nodes <- initialize_dataframe(2, c('cell_type', 'influence'), cell_type_num)
cell_type_self_interaction_links <- initialize_dataframe(4, c('from', 'to', 'weight', 'sign'))
interaction_names <- names(coef_interaction)
idx <- 1
for (i in 1:cell_type_num){
current_interaction_name <- paste('X', cell_type_list[i], sep = '')
current_influence <- 0
for (j in 1:length(coef_interaction)){
if (grepl(current_interaction_name, interaction_names[j], fixed=TRUE)
& grepl(paste('InteractionmarkX', cell_type_list[i], 'xX', cell_type_list[i], sep = ''), interaction_names[j], fixed=TRUE)){
current_sign <- ifelse(sign(coef_interaction[j]) == 1, 'darkblue', 'darkred')
cell_type_self_interaction_links[idx,] <- c(cell_type_list[i], paste(cell_type_list[i], '2', sep='_'), abs(coef_interaction[j]), current_sign)
current_influence <- current_influence + coef_interaction[j]
idx <- idx + 1
}
}
cell_type_nodes[i,] <- c(cell_type_list[i], current_influence)
}
return(list(cell_type_nodes, cell_type_self_interaction_links))
}
process_and_visualize <- function(interaction_type, cell_type_links, cell_type_nodes, input_file_name, output_dir, tr, ir, hr) {
if (interaction_type == 'pairwise'){
x = c(0, -0.951, 0.951, -0.588, 0.588)
y = c(1, 0.309, 0.309, -0.809, -0.809)
cell_type_nodes = cbind(cell_type_nodes, x, y)
} else if (interaction_type == 'self'){
x = c(0, 0, 0, 0, 0)
y = c(4, 3, 2, 1, 0)
cell_type_nodes = cbind(cell_type_nodes, x, y)
cell_type_nodes = cell_type_nodes
cell_type_nodes_copy = cell_type_nodes
cell_type_nodes_copy$cell_type = paste(cell_type_nodes_copy$cell_type,2,sep='_')
cell_type_nodes_copy$x = cell_type_nodes_copy$x + 1
cell_type_nodes = rbind(cell_type_nodes, cell_type_nodes_copy)
}
# Scaling functions
scaling_node_weight <- function(weight) {
positive_weight <- weight - min(weight)
scaled_weight <- scale(positive_weight, center = FALSE, scale = max(positive_weight)/40) + 10
return(scaled_weight)
}
scaling_edge_weight <- function(weight) {
positive_weight <- weight - min(weight)
scaled_weight <- scale(positive_weight, center = FALSE, scale = max(positive_weight)/19.9) + 0.1
return(scaled_weight)
}
# Scale node weights
cell_type_nodes$influence <- scaling_node_weight(as.numeric(cell_type_nodes$influence))
# Scale edge weights
cell_type_links$weight <- scaling_edge_weight(as.numeric(cell_type_links$weight))
# Generate and save pairwise interaction plot
filename <- file.path(output_dir, paste0(input_file_name, "_", interaction_type,"_interaction_TR_", tr, "_IR_", ir, "_HR_", hr, ".png"))
png(filename, width = 3000, height = 2500, res = 300)
net <- graph_from_data_frame(d=cell_type_links, vertices=cell_type_nodes, directed=F)
V(net)$size <- as.numeric(V(net)$influence)
if (interaction_type == "self"){
num_of_vertices <- length(V(net)) / 2
distinct_colors <- rainbow(num_of_vertices)
V(net)$color <- c(distinct_colors, distinct_colors)
} else{
num_of_vertices <- length(V(net))
distinct_colors <- rainbow(num_of_vertices)
V(net)$color <- distinct_colors
}
E(net)$width <- as.numeric(E(net)$weight)
E(net)$edge.color <- E(net)$sign
plot(net, edge.color = E(net)$sign, vertex.label=NA)
legend("topright", legend = V(net)$name[1:num_of_vertices], fill = distinct_colors, cex = 1.5, box.lwd = 2)
garbage <- dev.off()
cat(paste0(interaction_type, " interactions visualized!\n"))
}