-
Notifications
You must be signed in to change notification settings - Fork 0
/
GraphForPredictedRisk.R
256 lines (198 loc) · 15.2 KB
/
GraphForPredictedRisk.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
################# Script for Plots of predicted Risks and ORs#########################
######################################################################################################
################### Predicted risk ################################################
##########################################################################################################
########################################## FOR LASSO MODEL ######################################
###predicted risk= expit(logit p risk)
expit<-function(x) {exp(x)/(1+exp(x))}
p<-IPDNMRJAGSresultsLASSO$BUGSoutput$mean$logitp
p<-expit(IPDNMRJAGSresultsLASSO$BUGSoutput$mean$logitp)
####preperation for the graph
###For Dymethyl fumarate- Risk & propability to relapse
DFL<-cbind(Risknew,p[,1])
DFL$Treatment<-"Dimethyl Fumarate"
colnames(DFL)<-c("Risknew","prelapse","Treatment")
##For Glatimarer acetate - Risk & propability to relapse
GAL<-cbind(Risknew,p[,2])
GAL$Treatment<-"Glatiramer Acetate"
colnames(GAL)<-c("Risknew","prelapse","Treatment")
##For Natalizumab acetate - Risk & propability to relapse
NL<-cbind(Risknew,p[,3])
NL$Treatment<-"Natalizumab"
colnames(NL)<-c("Risknew","prelapse","Treatment")
##For Placebo - Risk & propability to relapse
PlL<-cbind(Risknew,p[,4])
PlL$Treatment<-"Placebo"
colnames(PlL)<-c("Risknew","prelapse","Treatment")
##merge data for all the treatments
GraphdataL<-rbind(DFL,GAL,NL,PlL)
###Graph for IPD
ggplot(data=GraphdataL, aes(x=Risknew, y=prelapse, group=Treatment)) +
geom_line()+
geom_point()+
geom_vline(xintercept=0.2197, linetype="dashed", color = "red")+
geom_vline(xintercept=0.6162, linetype="dashed", color = "red")
IPDplotLASSO<-ggplot(GraphdataL, aes(x=Risknew, y=prelapse, group=Treatment)) +
geom_line(aes(color=Treatment))+
geom_point(aes(color=Treatment))+geom_vline(xintercept=0.2197, linetype="dashed", color = "red")+
geom_vline(xintercept=0.6162, linetype="dashed", color = "red")
IPDplotLASSO<-IPDplotLASSO+ xlab("Baseline risk") + ylab("Predicted probability of relapse")
IPDplotLASSO
######################################################################################################
################################# For PreSpecified model #################################################
###predicted risk= expit(logit p risk)
expit<-function(x) {exp(x)/(1+exp(x))}
p<-IPDNMRJAGSresultsPreSpecified$BUGSoutput$mean$logitp
p<-expit(IPDNMRJAGSresultsPreSpecified$BUGSoutput$mean$logitp)
####preperation for the graph
###For Dymethyl fumarate- Risk & propability to relapse
DFF<-cbind(Risknew,p[,1])
DFF$Treatment<-"Dimethyl Fumarate"
colnames(DFF)<-c("Risknew","prelapse","Treatment")
##For Glatimarer acetate - Risk & propability to relapse
GAF<-cbind(Risknew,p[,2])
GAF$Treatment<-"Glatiramer Acetate"
colnames(GAF)<-c("Risknew","prelapse","Treatment")
##For Natalizumab acetate - Risk & propability to relapse
NF<-cbind(Risknew,p[,3])
NF$Treatment<-"Natalizumab"
colnames(NF)<-c("Risknew","prelapse","Treatment")
##For Placebo - Risk & propability to relapse
PlF<-cbind(Risknew,p[,4])
PlF$Treatment<-"Placebo"
colnames(PlF)<-c("Risknew","prelapse","Treatment")
##merge data for all the treatments
GraphdataF<-rbind(DFF,GAF,NF,PlF)
###Graph for IPD
ggplot(data=GraphdataF, aes(x=Risknew, y=prelapse, group=Treatment)) +
geom_line()+
geom_point()+geom_vline(xintercept=0.05113, linetype="dashed", color = "red")+
geom_vline(xintercept=0.72461, linetype="dashed", color = "red")
IPDplotPreSpecified<-ggplot(GraphdataF, aes(x=Risknew, y=prelapse, group=Treatment)) +
geom_line(aes(color=Treatment))+
geom_point(aes(color=Treatment))+geom_vline(xintercept=0.05113, linetype="dashed", color = "red")+
geom_vline(xintercept=0.72461, linetype="dashed", color = "red")
IPDplotPreSpecified<-IPDplotPreSpecified+ xlab("Baseline risk") + ylab("Predicted probability of relapse")
######################################################################################################
################### OR ################################################
##########################################################################################################
##################### LASSO model ################################
pl<-IPDNMRJAGSresultsLASSO$BUGSoutput$mean$logitp
####preperation for the graph
###For Dymethyl fumarate- Risk & propability to relapse
DFL_OR<-cbind(Risknew,exp(pl[,1]-pl[,4]))
DFL_OR$Treatment<-"Dimethyl Fumarate vs Placebo"
colnames(DFL_OR)<-c("Risknew","OR","Comparison")
##For Glatimarer acetate - Risk & propability to relapse
GAL_OR<-cbind(Risknew,exp(pl[,2]-pl[,4]))
GAL_OR$Treatment<-"Glatiramer Acetate vs Placebo"
colnames(GAL_OR)<-c("Risknew","OR","Comparison")
##For Natalizumab acetate - Risk & propability to relapse
NL_OR<-cbind(Risknew,exp(pl[,3]-pl[,4]))
NL_OR$Treatment<-"Natalizumab vs Placebo"
colnames(NL_OR)<-c("Risknew","OR","Comparison")
##For Placebo - Risk & propability to relapse
#PlL_OR<-cbind(Risknew,exp(pl[,4]-pl[,4]))
#PlL_OR$Treatment<-"Placebo"
#colnames(PlL_OR)<-c("Risknew","OR","Treatment")
##merge data for all the treatments
GraphdataL_OR<-rbind(DFL_OR,GAL_OR,NL_OR)
###Graph for IPD
ggplot(data=GraphdataL_OR, aes(x=Risknew, y=OR, group=Comparison)) +
geom_line()+
geom_point() + ylim(0,2) + geom_vline(xintercept=0.2197, linetype="dashed", color = "red") + geom_hline(yintercept=1, color="black") +
geom_vline(xintercept=0.6162, linetype="dashed", color = "red")
IPDplotLASSO_OR<-ggplot(GraphdataL_OR, aes(x=Risknew, y=OR, group=Comparison)) +
geom_line(aes(color=Comparison))+
geom_point(aes(color=Comparison))+ ylim(0,2) +geom_vline(xintercept= 0.2197, linetype="dashed", color = "red")+ geom_hline(yintercept=1, color="black") +
geom_vline(xintercept=0.6162, linetype="dashed", color = "red")
IPDplotLASSO_OR<-IPDplotLASSO_OR+ xlab("Baseline risk") + ylab("OR")
IPDplotLASSO_OR
######################################################################################################
################################# For PreSpecified model #################################################
###predicted risk= expit(logit p risk)
pf<-IPDNMRJAGSresultsPreSpecified$BUGSoutput$mean$logitp
####preperation for the graph
###For Dymethyl fumarate- Risk & propability to relapse
DF_OR<-cbind(Risknew,exp(pf[,1]-pf[,4]))
DF_OR$Treatment<-"Dimethyl Fumarate vs Placebo"
colnames(DF_OR)<-c("Risknew","OR","Comparison")
##For Glatimarer acetate - Risk & propability to relapse
GA_OR<-cbind(Risknew,exp(pf[,2]-pf[,4]))
GA_OR$Treatment<-"Glatiramer Acetate vs Placebo"
colnames(GA_OR)<-c("Risknew","OR","Comparison")
##For Natalizumab acetate - Risk & propability to relapse
N_OR<-cbind(Risknew,exp(pf[,3]-pf[,4]))
N_OR$Treatment<-"Natalizumab vs Placebo"
colnames(N_OR)<-c("Risknew","OR","Comparison")
##For Placebo - Risk & propability to relapse
#Pl_OR<-cbind(Risknew,exp(pf[,4]-pf[,4]))
#Pl_OR$Treatment<-"Placebo"
#colnames(Pl_OR)<-c("Risknew","OR","Treatment")
##merge data for all the treatments
GraphdataF_OR<-rbind(DF_OR,GA_OR,N_OR)
###Graph for IPD
ggplot(data=GraphdataF_OR, aes(x=Risknew, y=OR, group=Comparison)) +
geom_line()+
geom_point()+ ylim(0,2)+geom_vline(xintercept=0.05113, linetype="dashed", color = "red") + geom_hline(yintercept=1, color="black") +
geom_vline(xintercept=0.72461, linetype="dashed", color = "red")
IPDplotPreSpecified_OR<-ggplot(GraphdataF_OR, aes(x=Risknew, y=OR, group=Comparison)) +
geom_line(aes(color=Comparison))+
geom_point(aes(color=Comparison))+ ylim(0,2)+geom_vline(xintercept=0.05113, linetype="dashed", color = "red")+ geom_hline(yintercept=1, color="black") +
geom_vline(xintercept=0.72461, linetype="dashed", color = "red")
IPDplotPreSpecified_OR<-IPDplotPreSpecified_OR+ xlab("Baseline risk") + ylab("OR")
IPDplotPreSpecified_OR
###################### TABLES OF predicted risks by risk-group ###################################
mean1<- summary(GraphdataL$prelapse[which(GraphdataL$Risknew>=0.22 & GraphdataL$Risknew<=0.616 & GraphdataL$Treatment=="Dimethyl Fumarate")])[4]*100
mean2<- summary(GraphdataL$prelapse[which(GraphdataL$Risknew>=0.22 & GraphdataL$Risknew<=0.616 & GraphdataL$Treatment=="Glatiramer Acetate")])[4]*100
mean3<- summary(GraphdataL$prelapse[which(GraphdataL$Risknew>=0.22 & GraphdataL$Risknew<=0.616 & GraphdataL$Treatment=="Natalizumab")])[4]*100
mean1low<- summary(GraphdataL$prelapse[which(GraphdataL$Risknew>=0.22 & GraphdataL$Risknew<0.3 & GraphdataL$Treatment=="Dimethyl Fumarate")])[4]*100
mean2low<- summary(GraphdataL$prelapse[which(GraphdataL$Risknew>=0.22 & GraphdataL$Risknew<0.3 & GraphdataL$Treatment=="Glatiramer Acetate")])[4]*100
mean3low<- summary(GraphdataL$prelapse[which(GraphdataL$Risknew>=0.22 & GraphdataL$Risknew<0.3 & GraphdataL$Treatment=="Natalizumab")])[4]*100
mean1high<- summary(GraphdataL$prelapse[which(GraphdataL$Risknew>0.5 & GraphdataL$Risknew<=0.616 & GraphdataL$Treatment=="Dimethyl Fumarate")])[4]*100
mean2high<- summary(GraphdataL$prelapse[which(GraphdataL$Risknew>0.5 & GraphdataL$Risknew<=0.616 & GraphdataL$Treatment=="Glatiramer Acetate")])[4]*100
mean3high<- summary(GraphdataL$prelapse[which(GraphdataL$Risknew>0.5 & GraphdataL$Risknew<=0.616 & GraphdataL$Treatment=="Natalizumab")])[4]*100
LASSOtable<-data.frame("Average predicted relapse "=c(mean1, mean2, mean3), "Low-isk patients' (<30%) predicted relapse"=c(mean1low,mean2low,mean3low)
, "High-isk patients' (>50%) predicted relapse "=c(mean1high,mean2high,mean3high))
names(LASSOtable)<-c("Average predicted relapse %","Low-isk patients' (<30%) predicted relapse %","High-isk patients' (>50%) predicted relapse %")
rownames(LASSOtable)<-c("Dymethyl fumarate","Glatiramer acetate","Natalizumab")
mean1f<- summary(GraphdataF$prelapse[which(GraphdataF$Risknew>=0.05 & GraphdataF$Risknew<=0.725 & GraphdataF$Treatment=="Dimethyl Fumarate")])[4]*100
mean2f<- summary(GraphdataF$prelapse[which(GraphdataF$Risknew>=0.05 & GraphdataF$Risknew<=0.725 & GraphdataF$Treatment=="Glatiramer Acetate")])[4]*100
mean3f<- summary(GraphdataF$prelapse[which(GraphdataF$Risknew>=0.05 & GraphdataF$Risknew<=0.725 & GraphdataF$Treatment=="Natalizumab")])[4]*100
mean1lowf<- summary(GraphdataF$prelapse[which(GraphdataF$Risknew>=0.05 & GraphdataF$Risknew<0.3 & GraphdataF$Treatment=="Dimethyl Fumarate")])[4]*100
mean2lowf<- summary(GraphdataF$prelapse[which(GraphdataF$Risknew>=0.05 & GraphdataF$Risknew<0.3 & GraphdataF$Treatment=="Glatiramer Acetate")])[4]*100
mean3lowf<- summary(GraphdataF$prelapse[which(GraphdataF$Risknew>=0.05 & GraphdataF$Risknew<0.3 & GraphdataF$Treatment=="Natalizumab")])[4]*100
mean1highf<- summary(GraphdataF$prelapse[which(GraphdataF$Risknew>0.5 & GraphdataF$Risknew<=0.725 & GraphdataF$Treatment=="Dimethyl Fumarate")])[4]*100
mean2highf<- summary(GraphdataF$prelapse[which(GraphdataF$Risknew>0.5 & GraphdataF$Risknew<=0.725 & GraphdataF$Treatment=="Glatiramer Acetate")])[4]*100
mean3highf<- summary(GraphdataF$prelapse[which(GraphdataF$Risknew>0.5 & GraphdataF$Risknew<=0.725 & GraphdataF$Treatment=="Natalizumab")])[4]*100
PreSpecifiedtable<-data.frame("Average predicted relapse"=c(mean1f, mean2f, mean3f), "Low-isk patients' (<30%) predicted relapse"=c(mean1lowf,mean2lowf,mean3lowf)
, "High-isk patients' (>50%) predicted relapse"=c(mean1highf,mean2highf,mean3highf))
names(PreSpecifiedtable)<-c("Average predicted relapse %","Low-isk patients' (<30%) predicted relapse %","High-isk patients' (>50%) predicted relapse %")
rownames(PreSpecifiedtable)<-c("Dymethyl fumarate","Glatiramer acetate","Natalizumab")
###################### TABLES OF ORs by risk-group ###################################
mean1OR<- summary(GraphdataL_OR$OR[which(GraphdataL_OR$Risknew>=0.22 & GraphdataL_OR$Risknew<=0.616 & GraphdataL_OR$Comparison=="Dimethyl Fumarate vs Placebo")])[4]
mean2OR<- summary(GraphdataL_OR$OR[which(GraphdataL_OR$Risknew>=0.22 & GraphdataL_OR$Risknew<=0.616 & GraphdataL_OR$Comparison=="Glatiramer Acetate vs Placebo")])[4]
mean3OR<- summary(GraphdataL_OR$OR[which(GraphdataL_OR$Risknew>=0.22 & GraphdataL_OR$Risknew<=0.616 & GraphdataL_OR$Comparison=="Natalizumab vs Placebo")])[4]
mean1lowOR<- summary(GraphdataL_OR$OR[which(GraphdataL_OR$Risknew>=0.22 & GraphdataL_OR$Risknew<0.3 & GraphdataL_OR$Comparison=="Dimethyl Fumarate vs Placebo")])[4]
mean2lowOR<- summary(GraphdataL_OR$OR[which(GraphdataL_OR$Risknew>=0.22 & GraphdataL_OR$Risknew<0.3 & GraphdataL_OR$Comparison=="Glatiramer Acetate vs Placebo")])[4]
mean3lowOR<- summary(GraphdataL_OR$OR[which(GraphdataL_OR$Risknew>=0.22 & GraphdataL_OR$Risknew<0.3 & GraphdataL_OR$Comparison=="Natalizumab vs Placebo")])[4]
mean1highOR<- summary(GraphdataL_OR$OR[which(GraphdataL_OR$Risknew>0.5 & GraphdataL_OR$Risknew<=0.616 & GraphdataL_OR$Comparison=="Dimethyl Fumarate vs Placebo")])[4]
mean2highOR<- summary(GraphdataL_OR$OR[which(GraphdataL_OR$Risknew>0.5 & GraphdataL_OR$Risknew<=0.616 & GraphdataL_OR$Comparison=="Glatiramer Acetate vs Placebo")])[4]
mean3highOR<- summary(GraphdataL_OR$OR[which(GraphdataL_OR$Risknew>0.5 & GraphdataL_OR$Risknew<=0.616 & GraphdataL_OR$Comparison=="Natalizumab vs Placebo")])[4]
LASSOtableOR<-data.frame("Average OR "=c(mean1OR, mean2OR, mean3OR), "Low-isk patients' (<30%) mean OR"=c(mean1lowOR,mean2lowOR,mean3lowOR)
, "High-isk patients' (>50%) predicted relapse "=c(mean1highOR,mean2highOR,mean3highOR))
names(LASSOtableOR)<-c("Average OR","Low-isk patients' (<30%) OR","High-isk patients' (>50%) OR ")
rownames(LASSOtableOR)<-c("Dymethyl fumarate","Glatiramer acetate","Natalizumab")
mean1ORF<- summary(GraphdataF_OR$OR[which(GraphdataF_OR$Risknew>=0.05 & GraphdataF_OR$Risknew<=0.725 & GraphdataF_OR$Comparison=="Dimethyl Fumarate vs Placebo")])[4]
mean2ORF<- summary(GraphdataF_OR$OR[which(GraphdataF_OR$Risknew>=0.05 & GraphdataF_OR$Risknew<=0.725 & GraphdataF_OR$Comparison=="Glatiramer Acetate vs Placebo")])[4]
mean3ORF<- summary(GraphdataF_OR$OR[which(GraphdataF_OR$Risknew>=0.05 & GraphdataF_OR$Risknew<=0.725 & GraphdataF_OR$Comparison=="Natalizumab vs Placebo")])[4]
mean1lowORF<- summary(GraphdataF_OR$OR[which(GraphdataF_OR$Risknew>=0.05 & GraphdataF_OR$Risknew<0.3 & GraphdataF_OR$Comparison=="Dimethyl Fumarate vs Placebo")])[4]
mean2lowORF<- summary(GraphdataF_OR$OR[which(GraphdataF_OR$Risknew>=0.05 & GraphdataF_OR$Risknew<0.3 & GraphdataF_OR$Comparison=="Glatiramer Acetate vs Placebo")])[4]
mean3lowORF<- summary(GraphdataF_OR$OR[which(GraphdataF_OR$Risknew>=0.05 & GraphdataF_OR$Risknew<0.3 & GraphdataF_OR$Comparison=="Natalizumab vs Placebo")])[4]
mean1highORF<- summary(GraphdataF_OR$OR[which(GraphdataF_OR$Risknew>0.5 & GraphdataF_OR$Risknew<=0.725 & GraphdataF_OR$Comparison=="Dimethyl Fumarate vs Placebo")])[4]
mean2highORF<- summary(GraphdataF_OR$OR[which(GraphdataF_OR$Risknew>0.5 & GraphdataF_OR$Risknew<=0.725 & GraphdataF_OR$Comparison=="Glatiramer Acetate vs Placebo")])[4]
mean3highORF<- summary(GraphdataF_OR$OR[which(GraphdataF_OR$Risknew>0.5 & GraphdataF_OR$Risknew<=0.725 & GraphdataF_OR$Comparison=="Natalizumab vs Placebo")])[4]
PreSpecifiedtableOR<-data.frame("Average OR "=c(mean1ORF, mean2ORF, mean3ORF), "Low-isk patients' (<30%) mean OR"=c(mean1lowORF,mean2lowORF,mean3lowORF)
, "High-isk patients' (>50%) predicted relapse "=c(mean1highORF,mean2highORF,mean3highORF))
names(PreSpecifiedtableOR)<-c("Average OR","Low-isk patients' (<30%) OR","High-isk patients' (>50%) OR ")
rownames(PreSpecifiedtableOR)<-c("Dymethyl fumarate","Glatiramer acetate","Natalizumab")