-
Notifications
You must be signed in to change notification settings - Fork 0
/
code_treatfx_share.R
202 lines (159 loc) · 7.82 KB
/
code_treatfx_share.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
198
199
200
201
202
#### SUMMARY: Evaluation of change in fMRI responses following antidepressant treatment.
#### AUTHOR: Patrick M. Fisher, 2021
####
## RELEVANT PACKAGES
####
library(ggbeeswarm) # beeswarm plots
d.treatment <- read.csv('./pubMaterials/data_demo_aal_treatment.csv', stringsAsFactors = F)
writeLines('**** GROUP SAMPLE SIZES ****')
table(d.treatment$sex[d.treatment$follow.up==F])
# variables of interest
desc.cont <- c('age', # age in years
'rt.faces.all', # reaction time to faces (fmri task)
'rt.shapes.all', # reaction time to shapes (fmri task)
'acc.faces', # accuracy faces (fmri task)
'acc.shapes', # accuracy shapes (fmri task)
'outlier.total', # outlier volumes (fmri)
'outlier.faces', # outlier faces volumes (fmri)
'outlier.shapes', # outlier shapes volumes (fmri)
'hamd6.baseline', # hamd-6 at baseline
'hamd6.w8', # hamd-6 at week 8
'hamd17.baseline', # hamd-17 at baseline
'hamd17.w8', # hamd-17 at week 8
'srt.rt', # simple reaction time (ms, neuropsych)
'srt.std', # simple reaction time st dev (ms, neuropsych)
'srt.acc', # simple reaction time (accuracy, neuropsych)
'cats.mean', # CATS (mean)
'cats.sum' # CATS (total)
)
# clinical measures of interest
clin.measures <- c('hamd6.baseline', 'hamd6.w8', 'hamd17.baseline', 'hamd17.w8')
groups <- c('Baseline','Follow-Up')
####
## Data are foundation for SUPPLEMENTARY TABLE 3
####
writeLines('**** GROUP DESCRIPTIVE INFORMATION ****')
for(g in groups){
if(g=='Baseline'){
row.set <- which(d.treatment$follow.up==F)
} else if(g=='Follow-Up'){
row.set <- which(d.treatment$follow.up==T)
}
txt <- paste(unlist(lapply(desc.cont, function(i){
nomit <- sum(is.na(d.treatment[row.set,i]))
return(paste0(i, ': ',
signif(mean(d.treatment[row.set,i], na.rm=T),4), ' \u00B1 ',
signif(sd(d.treatment[row.set,i], na.rm=T),4), ' [',
signif(median(d.treatment[row.set,i], na.rm=T),4), '; ',
signif(min(d.treatment[row.set,i], na.rm=T),4), '-',
signif(max(d.treatment[row.set,i], na.rm=T),4), ']',
' (nomit: ', nomit, ')\n\t'))
})),collapse='')
writeLines(paste0('**** ', g, ' ****\n\t',txt))
}
txt2 <- paste(unlist(lapply(desc.cont, function(i){
if(i%in%clin.measures){
if(i%in%c('hamd17.baseline','hamd6.baseline')){
currMeasure <- unlist(strsplit(i,'.',fixed=T))[1]
diff.score <- d.treatment[d.treatment$follow.up==F,paste0(currMeasure,'.baseline')]-
d.treatment[d.treatment$follow.up==F,paste0(currMeasure,'.w8')]
ttest <- t.test(diff.score)
return(paste0('Delta-', currMeasure, ': ', signif(ttest$estimate,4), '[',
paste(signif(ttest$conf.int,3),collapse=', '), ']; p = ',
signif(ttest$p.value,3), '\n\t'))
} else {return(NULL)}
} else {
baseline <- sapply(unique(d.treatment$cimbi.id), function(j){
matches <- d.treatment$cimbi.id==j & d.treatment$follow.up==F
return(d.treatment[matches,i])
})
rescan <- sapply(unique(d.treatment$cimbi.id), function(j){
matches <- d.treatment$cimbi.id==j & d.treatment$follow.up==T
return(d.treatment[matches,i])
})
ttest <- t.test(rescan,baseline, paired = T)
return(paste0(i, ': ', signif(ttest$estimate,4), '[',
paste(signif(ttest$conf.int,4),collapse=', '), ']; p = ',
signif(ttest$p.value,4), '\n\t'))
}})),collapse='')
writeLines(paste0('**** GROUP DIFFERENCES: DESCRIPTIVES (PARAMETRIC) ****\n\t', txt2))
txt3 <- paste(unlist(lapply(desc.cont, function(i){
if(i%in%c('age')){
return(NULL)
}
if(i%in%clin.measures){
if(i%in%c('hamd17.baseline','hamd6.baseline')){
currMeasure <- unlist(strsplit(i,'.',fixed=T))[1]
diff.score <- d.treatment[d.treatment$follow.up==F,paste0(currMeasure,'.baseline')]-
d.treatment[d.treatment$follow.up==F,paste0(currMeasure,'.w8')]
l <- wilcox.test(diff.score, mu = 0)
return(paste0('Delta-', currMeasure, ': V=', l$statistic, '; p = ',
signif(l$p.value,3), '\n\t'))
} else {return(NULL)}
} else {
baseline <- sapply(unique(d.treatment$cimbi.id), function(j){
matches <- d.treatment$cimbi.id==j & d.treatment$follow.up==F
return(d.treatment[matches,i])
})
rescan <- sapply(unique(d.treatment$cimbi.id), function(j){
matches <- d.treatment$cimbi.id==j & d.treatment$follow.up==T
return(d.treatment[matches,i])
})
diff <- rescan-baseline
diff <- diff[which(!is.na(diff))] # keep those which are not NA
l <- wilcox.test(diff, mu = 0)
return(paste0(i, ': V=', l$statistic, '; p = ',
signif(l$p.value,3), '\n\t'))
}})),collapse='')
writeLines(paste0('**** GROUP DIFFERENCES: DESCRIPTIVES (NON-PARAMETRIC) ****\n\t', txt3))
####
## TREATMENT EFFECTS - UNIVARIATE (AAL)
####
nregions <- 90
idx <- data.frame(t(sapply(unique(d.treatment$cimbi.id), function(i){
return(c(which(d.treatment$cimbi.id==i & d.treatment$follow.up==F),
which(d.treatment$cimbi.id==i & d.treatment$follow.up==T)))
})))
colnames(idx) <- c('baseline','rescan')
rownames(idx) <- unique(d.treatment$cimbi.id)
stats.out2 <- data.frame(t(sapply(seq(nregions), function(i){
vals.base <- d.treatment[idx$baseline,aal.names[i]]
vals.rescan <- d.treatment[idx$rescan,aal.names[i]]
vals.diff <- vals.rescan-vals.base
ttest <- t.test(vals.diff)
coh.d <- mean(vals.diff)/sd(vals.diff)
return(c(ttest$estimate,
ttest$conf.int,
coh.d,
ttest$p.value,
mean(vals.base),
sd(vals.base),
mean(vals.rescan),
sd(vals.rescan),
mean(vals.diff),
sd(vals.diff)))
})))
colnames(stats.out2) <- c('est','lwr','upr','d', 'p', 'base.mean', 'base.sd', 'rescan.mean', 'rescan.sd', 'diff.mean', 'diff.sd')
stats.out2$regions <- aal.names[seq(nregions)]
stats.out2$p.adj <- p.adjust(stats.out2$p, method='holm')
stats.out2$color <- factor(unlist(lapply(seq(nrow(stats.out2)), function(i){
if(stats.out2[i,'p.adj']<0.05){return('#06E406')}
if(stats.out2[i,'p']<0.05){return('#E49E09')}
return('#000000')
})), levels = c('#06E406','#E49E09','#000000'))
# Figure 2 (see also code_figures.R)
ggplot(stats.out2, aes(x=base.mean,y=rescan.mean,color=color)) +
geom_abline(slope = 1, intercept = 0, lwd = 2, color = 'gray', lty = 2) +
geom_point(show.legend = F, size = 3) +
scale_y_continuous(limits = c(-0.6, 2.1),
breaks = seq(-0.6, 2.1, by = 0.3)) +
scale_x_continuous(limits = c(-0.6, 2.1),
breaks = seq(-0.6, 2.1, by = 0.3)) +
scale_color_manual(breaks = c('#06E406','#E49E09','#000000'),
values = c('#06E406','#E49E09','#000000')) +
labs(x = 'Baseline brain responses (arbitrary units)',
y = 'Week 8 brain responses\n(arbitrary units)') +
theme(axis.text.y = element_text(size=16),
axis.text.x = element_text(size=16),
axis.title.y = element_text(size=20),
axis.title.x = element_text(size=20))