-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy path5-1_GLM_site effect on preds, performance_21.12.10.R
660 lines (476 loc) · 23.3 KB
/
5-1_GLM_site effect on preds, performance_21.12.10.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
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
library(tidyverse)
library(magrittr)
# library(naniar)
# library(skimr)
# library(rstatix)
# library(tigerstats)
# Data: train/validation set - z-scaled
setwd("C:/Users/김보겸/Desktop/Connectome/study-enigma ocd/ENIGMA-OCD/0.Data/Dai result/After piras updated_Adult/T.Dx_D.nonharmo scaled only adult_cv.LOSO_777_F.age.sex_21.11.23")
df_adult_val <- read_csv('train_preds_custom.csv') # 1,068 x 272
#### check target outcome - Dx ####
df_adult_val %>%
xtabs(~Dx, data = ., addNA = T) %>% addmargins()
#### Make Dx_character variable ####
df_adult_val %<>%
mutate(Dx_charac = ifelse(Dx == 1, 'OCD', 'HC')) # 1,336 x 273
# set levels (to reorder)
df_adult_val$Dx_charac %<>%
factor(x = ., levels = c('OCD','HC'))
#### Make Med_12 variable ####
df_adult_val %<>%
mutate(Med_12 = ifelse(Med == 1, 1,
ifelse(Med == 2, 2, NA)))
df_adult_val %>%
xtabs(~ Med_12, data = ., addNA = TRUE)
# Male
df_adult_val %<>%
mutate(Male = ifelse(Sex == 1, 'Male', 'Female')) # 1,336 x 274
#### Set datatype ####
df_adult_val$Sex <- as.factor(df_adult_val$Sex)
df_adult_val$Dx <- as.factor(df_adult_val$Dx)
df_adult_val$Med <- as.factor(df_adult_val$Med)
df_adult_val$Med_12 <- as.factor(df_adult_val$Med_12)
df_adult_val$Anx <- as.factor(df_adult_val$Anx)
df_adult_val$Dep <- as.factor(df_adult_val$Dep)
df_adult_val$CurAnx <- as.factor(df_adult_val$CurAnx)
df_adult_val$CurDep <- as.factor(df_adult_val$CurDep)
df_adult_val$Agr_Check <- as.factor(df_adult_val$Agr_Check)
df_adult_val$Clean <- as.factor(df_adult_val$Clean)
df_adult_val$Ord <- as.factor(df_adult_val$Ord)
df_adult_val$Sex_Rel <- as.factor(df_adult_val$Sex_Rel)
df_adult_val$Hoard <- as.factor(df_adult_val$Hoard)
#### Make subset ####
df_adult_val_ocd <- df_adult_val %>%
filter(Dx_charac == 'OCD')
#### OCD group summary - NA correctly input? ####
# Lifetime diagnosis
# 1. Anxiety
df_adult_val_ocd %<>%
mutate(Anx = ifelse(Anx ==1, 1,
ifelse(Anx == 2, 2,
ifelse(Anx == 0, NA, NA))),
Dep = ifelse(Dep ==1, 1,
ifelse(Dep == 2, 2,
ifelse(Dep == 0, NA, NA))),
CurAnx = ifelse(CurAnx ==1, 1,
ifelse(CurAnx == 2, 2,
ifelse(CurAnx == 0, NA, NA))),
CurDep = ifelse(CurDep ==1, 1,
ifelse(CurDep == 2, 2,
ifelse(CurDep == 0, NA, NA))),
Agr_Check = ifelse(Agr_Check == 0, 0,
ifelse(Agr_Check == 1, 1,
ifelse(Agr_Check ==999, NA, NA))),
Clean = ifelse(Clean == 0, 0,
ifelse(Clean == 1, 1,
ifelse(Clean ==999, NA, NA))),
Ord = ifelse(Ord == 0, 0,
ifelse(Ord == 1, 1,
ifelse(Ord ==999, NA, NA))),
Sex_Rel = ifelse(Sex_Rel == 0, 0,
ifelse(Sex_Rel == 1, 1,
ifelse(Sex_Rel ==999, NA, NA))),
Hoard = ifelse(Hoard == 0, 0,
ifelse(Hoard == 1, 1,
ifelse(Hoard ==999, NA, NA))))
#### OCD group - reassign variable type ####
df_adult_val_ocd$Sex <- as.factor(df_adult_val_ocd$Sex)
df_adult_val_ocd$Dx <- as.factor(df_adult_val_ocd$Dx)
df_adult_val_ocd$Med <- as.factor(df_adult_val_ocd$Med)
df_adult_val_ocd$Med_12 <- as.factor(df_adult_val_ocd$Med_12)
df_adult_val_ocd$Anx <- as.factor(df_adult_val_ocd$Anx)
df_adult_val_ocd$Dep <- as.factor(df_adult_val_ocd$Dep)
df_adult_val_ocd$CurAnx <- as.factor(df_adult_val_ocd$CurAnx)
df_adult_val_ocd$CurDep <- as.factor(df_adult_val_ocd$CurDep)
df_adult_val_ocd$Agr_Check <- as.factor(df_adult_val_ocd$Agr_Check)
df_adult_val_ocd$Clean <- as.factor(df_adult_val_ocd$Clean)
df_adult_val_ocd$Ord <- as.factor(df_adult_val_ocd$Ord)
df_adult_val_ocd$Sex_Rel <- as.factor(df_adult_val_ocd$Sex_Rel)
df_adult_val_ocd$Hoard <- as.factor(df_adult_val_ocd$Hoard)
str(df_adult_val[1:30])
str(df_adult_val_ocd[1:30])
####################################################################################
####################################################################################
#### site effects on predicted probability ####
####################################################################################
####################################################################################
####################################################################################
##########################################################################
# DV: Dx.1
# covariate :
# Demo: Age, Sex, Site
# Dx (if using df_adult_total)
# Clinical variable
# Average DTI
# Result-
###########################################################################
#### Demographic ####
# 1. Age
glm.Age_Cov.Demo.Dx <- df_adult_val %>% lm(Dx.1 ~ Age + Sex + Dx + Site, data = .)
summary(glm.Age_Cov.Demo.Dx)
anova(glm.Age_Cov.Demo.Dx)
# car::vif(glm.Age_Cov.Demo.Dx)
#### Clinical variable in OCD sample ####
# OCD illness severity score
glm.Sev_Cov.Demo.DTI <- df_adult_val_ocd %>%
lm(Dx.1 ~ Sev + Age + Sex + AverageFA + AverageMD + AverageRD + AverageAD + Site, data = .) # with cov
summary(glm.Sev_Cov.Demo.DTI)
anova(glm.Sev_Cov.Demo.DTI)
car::vif(glm.Sev_Cov.Demo.DTI)
# Age at onset
glm.AO_Cov.Demo.DTI <- df_adult_val_ocd %>%
lm(Dx.1 ~ AO + Age + Sex + AverageFA + AverageAD + Site, data = .) # with cov
summary(glm.AO_Cov.Demo.DTI)
anova(glm.AO_Cov.Demo.DTI)
# Duration of illness
glm.Dur_Cov.Demo.DTI <- df_adult_val_ocd %>%
lm(Dx.1 ~ Dur + Age + Sex + AverageFA + AverageAD + Site, data = .) # with cov
summary(glm.Dur_Cov.Demo.DTI)
anova(glm.Dur_Cov.Demo.DTI)
# Medication
glm.Med_12_Cov.Demo.DTI <- lm(data=df_adult_val_ocd, Dx.1 ~ Med_12 + Age + Sex + AverageFA + AverageAD + Site)
summary(glm.Med_12_Cov.Demo.DTI)
anova(glm.Med_12_Cov.Demo.DTI)
# comorbidity
glm.Anx_Cov.Demo.DTI <- lm(data=df_adult_val_ocd, Dx.1 ~ Anx + Age + Sex+ AverageFA+ AverageAD + Site)
summary(glm.Anx_Cov.Demo.DTI)
anova(glm.Anx_Cov.Demo.DTI)
glm.CurAnx_Cov.Demo.DTI <- lm(data=df_adult_val_ocd, Dx.1 ~ CurAnx + Age + Sex+ AverageFA + AverageAD + Site)
summary(glm.CurAnx_Cov.Demo.DTI)
anova(glm.CurAnx_Cov.Demo.DTI)
glm.Dep_Cov.Demo.DTI <- lm(data=df_adult_val_ocd, Dx.1 ~ Dep + Age + Sex + AverageFA + AverageAD + Site)
summary(glm.Dep_Cov.Demo.DTI)
anova(glm.Dep_Cov.Demo.DTI)
glm.CurDep_Cov.Demo.DTI <- lm(data=df_adult_val_ocd, Dx.1 ~ CurDep + Age + Sex + AverageFA +AverageAD + Site)
summary(glm.CurDep_Cov.Demo.DTI)
anova(glm.CurDep_Cov.Demo.DTI)
# Subsymptoms
glm.Agr_Cov.Demo.DTI <- lm(data=df_adult_val_ocd, Dx.1 ~ Agr_Check + Age + Sex + AverageFA + AverageAD + Site)
summary(glm.Agr_Cov.Demo.DTI)
anova(glm.Agr_Cov.Demo.DTI)
glm.Clean_Cov.Demo.DTI <- lm(data=df_adult_val_ocd, Dx.1 ~ Clean + Age + Sex+ AverageFA + AverageAD + Site)
summary(glm.Clean_Cov.Demo.DTI)
anova(glm.Clean_Cov.Demo.DTI)
glm.Sex_Rel_Cov.Demo.DTI <- lm(data=df_adult_val_ocd, Dx.1 ~ Sex_Rel + Age + Sex + AverageFA + AverageAD+ Site)
summary(glm.Sex_Rel_Cov.Demo.DTI)
anova(glm.Sex_Rel_Cov.Demo.DTI)
glm.Hoard_Cov.Demo.DTI <- lm(data=df_adult_val_ocd, Dx.1 ~ Hoard + Age + Sex + AverageFA + AverageAD + Site)
summary(glm.Hoard_Cov.Demo.DTI)
anova(glm.Hoard_Cov.Demo.DTI)
glm.Ord_Cov.Demo.DTI <- lm(data=df_adult_val_ocd, Dx.1 ~ Ord + Age + Sex + AverageFA + AverageAD+ Site)
summary(glm.Ord_Cov.Demo.DTI)
anova(glm.Ord_Cov.Demo.DTI)
###########################################################################
#### Summary - association between clinical variable and Preds #####
###########################################################################
summary(glm.Age_Cov.Demo.Dx)
summary(glm.Sev_Cov.Demo.DTI)
summary(glm.AO_Cov.Demo.DTI)
summary(glm.Dur_Cov.Demo.DTI)
summary(glm.Med_12_Cov.Demo.DTI)
summary(glm.Med_12_Cov.Demo.DTI)
summary(glm.Anx_Cov.Demo.DTI)
summary(glm.CurAnx_Cov.Demo.DTI)
summary(glm.Dep_Cov.Demo.DTI)
summary(glm.CurDep_Cov.Demo.DTI)
summary(glm.Agr_Cov.Demo.DTI)
summary(glm.Clean_Cov.Demo.DTI)
summary(glm.Sex_Rel_Cov.Demo.DTI)
summary(glm.Hoard_Cov.Demo.DTI)
summary(glm.Ord_Cov.Demo.DTI)
###########################################################################
#### Summary - Site effects on preds####
###########################################################################
anova(glm.Age_Cov.Demo.Dx)
anova(glm.Sev_Cov.Demo.DTI)
anova(glm.AO_Cov.Demo.DTI)
anova(glm.Dur_Cov.Demo.DTI)
anova(glm.Med_12_Cov.Demo.DTI)
anova(glm.Med_12_Cov.Demo.DTI)
anova(glm.Anx_Cov.Demo.DTI)
anova(glm.CurAnx_Cov.Demo.DTI)
anova(glm.Dep_Cov.Demo.DTI)
anova(glm.CurDep_Cov.Demo.DTI)
anova(glm.Agr_Cov.Demo.DTI)
anova(glm.Clean_Cov.Demo.DTI)
anova(glm.Sex_Rel_Cov.Demo.DTI)
anova(glm.Hoard_Cov.Demo.DTI)
anova(glm.Ord_Cov.Demo.DTI)
####################################################################################
####################################################################################
####################################################################################
#### site effects on individualized classification performance ####
# Reference code: Hierarchical regression for site effect on {individual prediction result}_fixed effects_0605.R
# Hierarchical glm code is not needed/ since we found the alternative method to show site effect
###########################################################################
####################################################################################
####################################################################################
##########################################################################
# DV:
# covariate :
# Demo: Age, Sex, Site
# Dx (if using df_adult_total)
# Clinical variable
# Average DTI
# Result-
###########################################################################
##################################################################################################
# 1. Dependent variable
# 1.1. list of DV
# 1) True prediction - OCD + HC
# 2) OCD - True positive vs. False Negative
# 3) HC - True Negative vs. False Positive
# 1.2. Process
# 1) add site threshold column
# 1.1) add Dx_predicted column : prediction result - OCD? or HC?
# 2) based on the threshold, add elements of confusion matrix
# 3) create Predictor > all = site dependent >> use site only / current status - do not create. 3.2~3.5
# 3.1) site
# 3.2) Dx ratio: site_pc_ocd
# 3.3) N ratio by site: site_pc_samplesize
# 3.4) Training sample ratio by site: site_pc_trainsize,
# 3.5) Training sample's diagnosis by site: site_pc_train.ocd
# 4) create DV that we will use
# 4.1) True prediction [proba_val] - TN, TP = 1, FN, FP = 0 / False_prediction_total - (reverse)
# 4.2) True prediction_OCD [proba_val_ocd] - TP = 1, FN = 0 / False_prediction_OCD - FN = 1, TP = 0
# 4.3) True prediction_HC [proba_val_hc] - TN = 1, FP =0 / False_prediction_HC - FP = 1, TN = 0
##################################################################################################
# 1.2. Process
# 1) add site threshold column
df_adult_val %<>%
mutate(threshold = ifelse(Site == 'Amsterdam', 0.5606682,
ifelse(Site == 'Bangalore', 0.5467443,
ifelse(Site == 'Capetown', 0.523335,
ifelse(Site == 'Kyoto', 0.5231088,
ifelse(Site == 'Milan', 0.5285721,
ifelse(Site == 'Mountsinai', 0.5327545,
ifelse(Site == 'Munich', 0.5243237,
ifelse(Site == 'Rome', 0.4470824,
ifelse(Site == 'Saopaulo', 0.5288856,
ifelse(Site == 'Seoul', 0.5531548,
ifelse(Site == 'Shanghai', 0.4738248, NA))))))))))))
str(df_adult_val$threshold)
hist(df_adult_val$threshold)
sum(is.na(df_adult_val$threshold)) # 0
df_adult_val %>%
xtabs(~ Site, addNA = TRUE, data = .)
# 1.1) add Dx_predicted column : prediction result - OCD? or HC?
print('This is a confusion matrix. Make the confusion matrix as variable')
df_adult_val %<>%
mutate(Dx_predicted = ifelse(Dx.1 >= threshold, 1, 0))
# confirmation
df_adult_val %>%
xtabs(~ Dx_predicted, data = ., addNA = T)
df_adult_val %>%
xtabs(~ Dx + Dx_predicted, data = ., addNA = T)
print('이게 바로 confusion matrix네. 이걸 변수로 만들어주자. ')
##################################################################################################
# 2) based on the threshold, add elements of confusion matrix
df_adult_val %<>%
mutate(ConfusionMatrix = ifelse(Dx_predicted == 0 & Dx == 0, 'TN',
ifelse(Dx_predicted ==1 & Dx == 1, 'TP',
ifelse(Dx_predicted == 0 & Dx == 1, 'FN',
ifelse(Dx_predicted == 1 & Dx == 0, 'FP', NA)))))
# check
df_adult_val %>%
xtabs(~Site + ConfusionMatrix, addNA = T, data = .)
sum(is.na(df_adult_val$ConfusionMatrix))
##################################################################################################
# 3) Predictor
# 3.1) site
str(df_adult_val$Site) # factor
##################################################################################################
# 4) DV
# 4.1) True prediction [proba_val] - TN, TP = 1, FN, FP = 0 / False_prediction_total - (reverse)
df_adult_val %<>%
mutate(True_prediction_total = ifelse(ConfusionMatrix == 'TN' | ConfusionMatrix == 'TP', 1, 0),
False_prediction_total = ifelse(ConfusionMatrix == 'FN' | ConfusionMatrix == 'FP', 1, 0))
df_adult_val %>%
xtabs(~ ConfusionMatrix + True_prediction_total, data = ., addNA = T) # 1에 TN, TP만 있음을 확인
df_adult_val %>%
xtabs(~ ConfusionMatrix + False_prediction_total, data = ., addNA = T)
# 4.2) True prediction_OCD [proba_val_ocd] - TP = 1, FN = 0 / False_prediction_OCD - FN = 1, TP = 0
df_adult_val_ocd <- df_adult_val %>%
filter(Dx == 1)
table(df_adult_val_ocd$ConfusionMatrix) # there's only FN or TP
df_adult_val_ocd %<>%
mutate(True_prediction_OCD = ifelse(ConfusionMatrix == 'TP', 1, 0))
df_adult_val_ocd %>%
mutate(False_prediction_OCD = ifelse(ConfusionMatrix == 'FN', 1, 0))
df_adult_val_ocd %>%
xtabs(~ Site + True_prediction_OCD, data = .) # check
# 4.3) True prediction_HC [proba_val_hc] - TN = 1, FP =0 / False_prediction_HC - FP = 1, TN = 0
df_adult_val_hc <- df_adult_val %>%
filter(Dx == 0)
table(df_adult_val_hc$ConfusionMatrix) # there's only FP or TN
df_adult_val_hc %<>%
mutate(True_prediction_HC = ifelse(ConfusionMatrix == 'TN', 1, 0))
df_adult_val_hc %<>%
mutate(False_prediction_HC = ifelse(ConfusionMatrix == 'FP', 1, 0))
# check
table(df_adult_val_hc$True_prediction_HC)
table(df_adult_val_hc$False_prediction_HC)
addmargins(xtabs(formula = ~ Site + True_prediction_HC, data = df_adult_val_hc)) #
#### OCD group summary - NA input correctly? ####
# Lifetime diagnosis
# 1. Anxiety
df_adult_val_ocd %<>%
mutate(Anx = ifelse(Anx ==1, 1,
ifelse(Anx == 2, 2,
ifelse(Anx == 0, NA, NA))),
Dep = ifelse(Dep ==1, 1,
ifelse(Dep == 2, 2,
ifelse(Dep == 0, NA, NA))),
CurAnx = ifelse(CurAnx ==1, 1,
ifelse(CurAnx == 2, 2,
ifelse(CurAnx == 0, NA, NA))),
CurDep = ifelse(CurDep ==1, 1,
ifelse(CurDep == 2, 2,
ifelse(CurDep == 0, NA, NA))),
Agr_Check = ifelse(Agr_Check == 0, 0,
ifelse(Agr_Check == 1, 1,
ifelse(Agr_Check ==999, NA, NA))),
Clean = ifelse(Clean == 0, 0,
ifelse(Clean == 1, 1,
ifelse(Clean ==999, NA, NA))),
Ord = ifelse(Ord == 0, 0,
ifelse(Ord == 1, 1,
ifelse(Ord ==999, NA, NA))),
Sex_Rel = ifelse(Sex_Rel == 0, 0,
ifelse(Sex_Rel == 1, 1,
ifelse(Sex_Rel ==999, NA, NA))),
Hoard = ifelse(Hoard == 0, 0,
ifelse(Hoard == 1, 1,
ifelse(Hoard ==999, NA, NA))))
#### OCD group - reassign variable type ####
df_adult_val_ocd$Sex <- as.factor(df_adult_val_ocd$Sex)
df_adult_val_ocd$Dx <- as.factor(df_adult_val_ocd$Dx)
df_adult_val_ocd$Med <- as.factor(df_adult_val_ocd$Med)
df_adult_val_ocd$Med_12 <- as.factor(df_adult_val_ocd$Med_12)
df_adult_val_ocd$Anx <- as.factor(df_adult_val_ocd$Anx)
df_adult_val_ocd$Dep <- as.factor(df_adult_val_ocd$Dep)
df_adult_val_ocd$CurAnx <- as.factor(df_adult_val_ocd$CurAnx)
df_adult_val_ocd$CurDep <- as.factor(df_adult_val_ocd$CurDep)
df_adult_val_ocd$Agr_Check <- as.factor(df_adult_val_ocd$Agr_Check)
df_adult_val_ocd$Clean <- as.factor(df_adult_val_ocd$Clean)
df_adult_val_ocd$Ord <- as.factor(df_adult_val_ocd$Ord)
df_adult_val_ocd$Sex_Rel <- as.factor(df_adult_val_ocd$Sex_Rel)
df_adult_val_ocd$Hoard <- as.factor(df_adult_val_ocd$Hoard)
str(df_adult_val_ocd[1:30])
##################################################################################################
# Modeling ###
# cov:Age + Sex + AverageFA + AverageAD + Sev + Dur+ Med_12
# high correlation among average FA, MD, RD AD > FA, AD only correlation
# site related factors: site only
# why? all 5 vars - site dependent
# glm analysis by group (logistic regression)
# 1. Among OCD:
# cov: Age + Sex + AverageFA + AverageAD + Sev + Dur+ Med_12
# dependent variableL : True_prediction_OCD
# 2. Among HC
# Issue: different results between 'summary' and 'anova' > see the results from summary & use anova for the site effect
# in anova, input order is importatnt
##################################################################################################
### summary
# 1. run a code same as clinical outcome
# see the result firstly (using the summary, same as clinical association with preds)
# 2. use hierarchical model for site effect only
# summarize code as above // => see the results
##########################################
#### clinical variable -> performance effect??
# dataset : df_adult_val_ocd\
# DV: True_prediction_OCD
# in OCD group,correctly OCD prediction?
# IV: clinical variable
# Covariates: Demo (Age, Sex, Site), Average DTI 2 (FA, AD: MD, RD vif > 200 => exclude)
# link fuction: binomial
#### Clinical variable in OCD sample ####
# Sev
glm.Sev_Cov.Demo.DTI <- df_adult_val_ocd %>%
glm(True_prediction_OCD ~ Sev + Age + Sex+ AverageFA + AverageAD + Site , family = binomial(link="logit"), data = .) # with cov
summary(glm.Sev_Cov.Demo.DTI)
anova(glm.Sev_Cov.Demo.DTI, test = 'Chisq') # logisti -> chisq test selection
car::vif(glm.Sev_Cov.Demo.DTI)
# Summary for manuscript
anova(glm.Sev_Cov.Demo.DTI, test = 'Chisq') # logisti -> chisq test selection
# Age at onset
glm.AO_Cov.Demo.DTI <- df_adult_val_ocd %>%
glm(True_prediction_OCD ~ AO + Age + Sex + AverageFA + AverageAD+ Site , data = ., family = binomial(link="logit")) # with cov
summary(glm.AO_Cov.Demo.DTI)
anova(glm.AO_Cov.Demo.DTI, test = 'Chisq')
# Duration of illness
glm.Dur_Cov.Demo.DTI <- df_adult_val_ocd %>%
glm(True_prediction_OCD ~ Dur + Age + Sex + AverageFA + AverageAD+ Site , data = ., family = binomial(link="logit")) # with cov
summary(glm.Dur_Cov.Demo.DTI)
anova(glm.Dur_Cov.Demo.DTI, test = 'Chisq')
# Medication
glm.Med_12_Cov.Demo.DTI <- df_adult_val_ocd %>%
glm(True_prediction_OCD ~ Med_12 + Age + Sex + AverageFA + AverageAD+ Site, data = ., family = binomial(link="logit")) # with cov
summary(glm.Med_12_Cov.Demo.DTI)
anova(glm.Med_12_Cov.Demo.DTI, test = 'Chisq')
# comorbidity
glm.Anx_Cov.Demo.DTI <- df_adult_val_ocd %>%
glm(True_prediction_OCD ~ Anx + Age + Sex + AverageFA + AverageAD+ Site , data = ., family = binomial(link="logit")) # with cov
summary(glm.Anx_Cov.Demo.DTI)
anova(glm.Anx_Cov.Demo.DTI, test = 'Chisq')
glm.CurAnx_Cov.Demo.DTI <- df_adult_val_ocd %>%
glm(True_prediction_OCD ~ CurAnx + Age + Sex + AverageFA + AverageAD+ Site , data = ., family = binomial(link="logit"))
summary(glm.CurAnx_Cov.Demo.DTI)
anova(glm.CurAnx_Cov.Demo.DTI, test = 'Chisq')
glm.Dep_Cov.Demo.DTI <- df_adult_val_ocd %>%
glm(True_prediction_OCD ~ Dep + Age + Sex + AverageFA + AverageAD + Site, data = ., family = binomial(link="logit"))
summary(glm.Dep_Cov.Demo.DTI)
anova(glm.Dep_Cov.Demo.DTI, test = 'Chisq')
glm.CurDep_Cov.Demo.DTI <- df_adult_val_ocd %>%
glm(True_prediction_OCD ~ CurDep + Age + Sex + AverageFA + AverageAD + Site , data = ., family = binomial(link="logit"))
summary(glm.CurDep_Cov.Demo.DTI)
anova(glm.CurDep_Cov.Demo.DTI, test = 'Chisq')
# Subsymptoms
glm.Agr_Cov.Demo.DTI <- df_adult_val_ocd %>%
glm(True_prediction_OCD ~ Agr_Check + Age + Sex + AverageFA + AverageAD + Site, data = ., family = binomial(link="logit"))
summary(glm.Agr_Cov.Demo.DTI)
anova(glm.Agr_Cov.Demo.DTI, test = 'Chisq')
glm.Clean_Cov.Demo.DTI <-df_adult_val_ocd %>%
glm(True_prediction_OCD ~ Clean + Age + Sex + AverageFA + AverageAD + Site, data = ., family = binomial(link="logit"))
summary(glm.Clean_Cov.Demo.DTI)
anova(glm.Clean_Cov.Demo.DTI, test = 'Chisq')
glm.Sex_Rel_Cov.Demo.DTI <- df_adult_val_ocd %>%
glm(True_prediction_OCD ~ Sex_Rel + Age + Sex + AverageFA + AverageAD + Site, data = ., family = binomial(link="logit"))
summary(glm.Sex_Rel_Cov.Demo.DTI)
anova(glm.Sex_Rel_Cov.Demo.DTI, test = 'Chisq')
glm.Hoard_Cov.Demo.DTI <- df_adult_val_ocd %>%
glm(True_prediction_OCD ~ Hoard + Age + Sex + AverageFA + AverageAD + Site, data = ., family = binomial(link="logit"))
summary(glm.Hoard_Cov.Demo.DTI)
anova(glm.Hoard_Cov.Demo.DTI, test = 'Chisq')
glm.Ord_Cov.Demo.DTI <- df_adult_val_ocd %>%
glm(True_prediction_OCD ~ Ord + Age + Sex + AverageFA + AverageAD + Site, data = ., family = binomial(link="logit"))
summary(glm.Ord_Cov.Demo.DTI)
anova(glm.Ord_Cov.Demo.DTI, test = 'Chisq')
##### association between individual performance and clinical variable ####
summary(glm.Sev_Cov.Demo.DTI)
summary(glm.AO_Cov.Demo.DTI)
summary(glm.Dur_Cov.Demo.DTI)
summary(glm.Med_12_Cov.Demo.DTI)
summary(glm.Anx_Cov.Demo.DTI)
summary(glm.CurAnx_Cov.Demo.DTI)
summary(glm.Dep_Cov.Demo.DTI)
summary(glm.CurDep_Cov.Demo.DTI)
summary(glm.Agr_Cov.Demo.DTI)
summary(glm.Clean_Cov.Demo.DTI)
summary(glm.Sex_Rel_Cov.Demo.DTI)
summary(glm.Hoard_Cov.Demo.DTI)
summary(glm.Ord_Cov.Demo.DTI)
##### site effect on individual performance ####
anova(glm.Sev_Cov.Demo.DTI, test = 'Chisq') # logisti -> chisq test selection
anova(glm.AO_Cov.Demo.DTI, test = 'Chisq')
anova(glm.Dur_Cov.Demo.DTI, test = 'Chisq')
anova(glm.Med_12_Cov.Demo.DTI, test = 'Chisq')
anova(glm.Anx_Cov.Demo.DTI, test = 'Chisq')
anova(glm.CurAnx_Cov.Demo.DTI, test = 'Chisq')
anova(glm.Dep_Cov.Demo.DTI, test = 'Chisq')
anova(glm.CurDep_Cov.Demo.DTI, test = 'Chisq')
anova(glm.Agr_Cov.Demo.DTI, test = 'Chisq')
anova(glm.Clean_Cov.Demo.DTI, test = 'Chisq')
anova(glm.Sex_Rel_Cov.Demo.DTI, test = 'Chisq')
anova(glm.Hoard_Cov.Demo.DTI, test = 'Chisq')
anova(glm.Ord_Cov.Demo.DTI, test = 'Chisq')