-
Notifications
You must be signed in to change notification settings - Fork 0
/
simulate_clusters_identical_sequences_by_age.R
169 lines (145 loc) · 7.02 KB
/
simulate_clusters_identical_sequences_by_age.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
## This script simulates a branching process with mutations
## assuming a Poisson distributed offspring distribution.
## This simulation approach follows that described in
## Tran-Kiem and Bedford 10.1073/pnas.2305299121.
library(tidyverse)
source('../age_analyses/utils_contact_mat.R')
## Load contact matrix
contact_mat_WA <- readRDS('../../results/RR_contacts/contact_mat_decade_overall.rds')
rho_eigenval <- get_max_eigenval(contact_mat_WA)
contact_mat_WA_adj <- contact_mat_WA / rho_eigenval
vec_age_groups <- colnames(contact_mat_WA)
## Load RR of contact between groups
df_RR_overall <- readRDS('../../results/RR_contacts/df_RR_overall.rds')
## Function to simulate clusters of identical sequences with age
simulate_cluster_identical_sequences_with_age <- function(contact_mat_adj = contact_mat_WA_adj,
R0,
p_trans_before_mut,
p_seq,
n_max_gen = 20){
n_age_groups <- nrow(contact_mat_adj)
vec_R0_age <- apply(contact_mat_adj, 1, sum) * R0
# Normalized contact matrix where sum of each row is equal to 1
contact_mat_prop <- contact_mat_adj /
matrix(rep(apply(contact_mat_adj, 1, sum), n_age_groups), ncol = n_age_groups, nrow = n_age_groups, byrow = F)
# Initialization
age_primary <- sample(x = n_age_groups, size = 1) # Draw the age of the primary infector
vec_age_previous_gen <- c(age_primary)
vec_age_new_gen <- NULL
tot_cluster_size <- 1
curr_pop_size <- 1
i_gen <- 0
vec_age_full_cluster <- c(age_primary)
vec_age_sequenced_indiv <- NULL
vec_gen_sequences_indiv <- NULL
if(rbernoulli(n = 1, p = p_seq) == 1){
vec_age_sequenced_indiv <- c(vec_age_sequenced_indiv, age_primary)
vec_gen_sequences_indiv <- c(vec_gen_sequences_indiv, i_gen)
}
# Simulate branching process
while(curr_pop_size > 0 && i_gen < n_max_gen){
# Loop over all infected individuals (previous generation)
for(i_infected in 1:curr_pop_size){
curr_age_infected <- vec_age_previous_gen[i_infected]
curr_R0 <- vec_R0_age[curr_age_infected]
# Draw number of infected individuals
n_new_infected_with_identical_sequences <- rpois(n = 1, lambda = curr_R0 * p_trans_before_mut)
if(n_new_infected_with_identical_sequences > 0){
age_draw <- rmultinom(n = 1, size = n_new_infected_with_identical_sequences, prob = as.numeric(contact_mat_prop[curr_age_infected, ]))
vec_age_infected <- Reduce('c', lapply(1:n_age_groups, FUN = function(i_age){
rep(i_age, age_draw[i_age])
}))
vec_age_full_cluster <- c(vec_age_full_cluster, vec_age_infected)
vec_age_new_gen <- c(vec_age_new_gen, vec_age_infected)
# Are these individuals sequenced?
vec_age_sequenced <- vec_age_infected[rbernoulli(n = n_new_infected_with_identical_sequences, p = p_seq)]
vec_age_sequenced_indiv <- c(vec_age_sequenced_indiv, vec_age_sequenced)
vec_gen_sequences_indiv <- c(vec_gen_sequences_indiv, rep(i_gen, length(vec_age_sequenced)))
}
}
## Update
vec_age_previous_gen <- vec_age_new_gen
curr_pop_size <- length(vec_age_previous_gen)
vec_age_new_gen <- NULL
}
return(
list('vec_age_sequenced_indiv' = vec_age_sequenced_indiv,
'vec_age_full_cluster' = vec_age_full_cluster,
'vec_gen_sequences_indiv' = vec_gen_sequences_indiv
)
)
}
## Function to get pairwise distance from the output of simulate_cluster_identical_sequences_with_age
get_pairwise_distance_from_vec_age <- function(vec_age_sequenced_indiv){
if(is.null(vec_age_sequenced_indiv)){
return(NULL)
} else if(length(vec_age_sequenced_indiv) <= 1) {
return(NULL)
} else {
df_age_indiv <- tibble(id_indiv = 1:length(vec_age_sequenced_indiv),
age = vec_age_sequenced_indiv)
df_pairs <- expand.grid(id_indiv_1 = 1:length(vec_age_sequenced_indiv),
id_indiv_2 = 1:length(vec_age_sequenced_indiv)) %>%
filter(id_indiv_1 != id_indiv_2) %>%
left_join(df_age_indiv, by = c('id_indiv_1' = 'id_indiv')) %>%
rename(age_1 = age) %>%
left_join(df_age_indiv, by = c('id_indiv_2' = 'id_indiv')) %>%
rename(age_2 = age) %>%
group_by(age_1, age_2) %>%
summarise(n_pairs = n()) %>%
ungroup()
return(df_pairs)
}
}
n_clusters <- 1e5 # This takes around 5 minutes for 1e5 clusters
R0 <- 1.2
p_trans_before_mut <- 0.7
p_seq <- 0.1
set.seed(87245)
t0 <- Sys.time()
df_pairs_id_seq <- Reduce('bind_rows', lapply(1:n_clusters, FUN = function(i_cluster){
curr_sim <- simulate_cluster_identical_sequences_with_age(contact_mat_adj = contact_mat_WA_adj,
R0 = R0,
p_trans_before_mut = p_trans_before_mut,
p_seq = p_seq,
n_max_gen = 10)
get_pairwise_distance_from_vec_age(curr_sim$vec_age_sequenced_indiv)
}))
t1 <- Sys.time()
print(t1 - t0)
## Compute RR of identical sequences between age groups
df_RR_id_seq <- df_pairs_id_seq %>%
group_by(age_1, age_2) %>%
summarise(n_pairs_age_1_age_2 = sum(n_pairs)) %>%
group_by(age_1) %>%
mutate(n_pairs_age_1_age_x = sum(n_pairs_age_1_age_2)) %>%
group_by(age_2) %>%
mutate(n_pairs_age_x_age_2 = sum(n_pairs_age_1_age_2)) %>%
ungroup() %>%
mutate(n_pairs_age_x_age_x = sum(n_pairs_age_1_age_2)) %>%
mutate(RR = n_pairs_age_1_age_2/n_pairs_age_1_age_x/n_pairs_age_x_age_2*n_pairs_age_x_age_x) %>%
mutate(age_decade_1 = vec_age_groups[age_1], age_decade_2 = vec_age_groups[age_2])
## Display the relationship between the RR of contacts and the RR of observing identical sequences between groups
plt_comp_RR_id_seq_contacts <- df_RR_id_seq %>%
left_join(df_RR_overall, by = c('age_decade_1' = 'age_decade_i', 'age_decade_2' = 'age_decade_j')) %>%
ggplot(aes(x = RR_contacts, y = RR)) +
geom_smooth(col = 'darkcyan', fill = 'darkcyan',
method = 'loess') +
geom_point() +
scale_x_continuous(trans = 'log', breaks = c(0.5, 0.8, 1.0, 2.0, 3.0, 5., 10., 15.),
name = expression(RR['contacts']),
labels = c('0.5', '0.8', '1', '2', '3', '5', '10', '15')) +
scale_y_continuous(trans = 'log', breaks = seq(0.1, 2.5, 0.1),
name = expression(RR['identical sequences'])) +
theme_classic() +
theme(axis.text = element_text(size = 12),
axis.title = element_text(size = 13))
plot(plt_comp_RR_id_seq_contacts)
pdf('../../figures/supplementary_figures/expected_relationship_RR_id_seq_contacts.pdf',
height = 4, width = 5)
plot(plt_comp_RR_id_seq_contacts)
dev.off()
png('../../figures/supplementary_figures/expected_relationship_RR_id_seq_contacts.png',
height = 4, width = 5, res = 350, units = 'in')
plot(plt_comp_RR_id_seq_contacts)
dev.off()