-
Notifications
You must be signed in to change notification settings - Fork 0
/
main_integrated.m
235 lines (173 loc) · 7.06 KB
/
main_integrated.m
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
% pull sample ids from survival data
[~, ~, ~, ~, sampleid] = preprocessing_survival();
% get indices for training and testing
cvIndices = crossvalind('kfold',sampleid,5);
chosenTestActual = sampleid(cvIndices == 1);
% initialize vectors for cross validation and external validation
% performance
pValues = [];
ciCV = [];
ciEV = [];
cv = [1 2 3 4];
% create a crossval loop
for fold = 2:5
% current Testing and Training indices for this loop
chosenTest = sampleid(cvIndices == fold);
chosenTrain = sampleid(cvIndices ~= fold & cvIndices ~= 1);
mirnaTrain = [];
mirnaTest = [];
mirnaTestActual = [];
% run mirna preprocessing with no inputs
[mirnafinal, cellid, cellid1, mirnaNorm, fieldNamescell] = preprocessing_mirna();
% filter through mirna samples and separate into train and test
for x = 1:length(fieldNamescell)
name = fieldNamescell(x);
name2 = chosenTrain(1);
if ismember(name, chosenTrain) == 1
chosenCol = num2cell(mirnaNorm(:,x));
chosenCol = [fieldNamescell(x); chosenCol];
mirnaTrain = [mirnaTrain chosenCol];
elseif ismember(name, chosenTest) == 1
chosenCol = num2cell(mirnaNorm(:,x));
chosenCol = [fieldNamescell(x); chosenCol];
mirnaTest = [mirnaTest chosenCol];
elseif ismember(name, chosenTestActual) == 1
chosenCol = num2cell(mirnaNorm(:,x));
chosenCol = [fieldNamescell(x); chosenCol];
mirnaTestActual = [mirnaTestActual chosenCol];
end
end
% split mirna into test, train, and validation
fieldNamestrain = mirnaTrain(1,:);
fieldNamestest = mirnaTest(1,:);
fieldNamestestActual = mirnaTestActual(1,:);
mirnaTrainData = cell2mat(mirnaTrain(2:end,:));
mirnaTestData = cell2mat(mirnaTest(2:end,:));
mirnaTestActual = cell2mat(mirnaTestActual(2:end,:));
% call FS, with inputs training
[mirnaTopData, mirnaTopFinal, mirnaTopFeatures, fieldNamestrain, cellid1] = featureselection_mirna(cellid, cellid1, mirnaTrainData, fieldNamestrain);
% call cox with inputs as training and testing
[B, survivalFinal, riskscore, mirnaTest] = cox_mirna(mirnaTopData, mirnaTopFeatures, mirnaTest, fieldNamestrain, cellid1);
B_mirna = B;
survivalFinal_mirna = survivalFinal;
riskscore_mirna = riskscore;
mirnaTest_mirna = mirnaTest;
fieldNamestest_mirna = fieldNamestest;
% RNASEQ
% preprocessing
[mirnafinal, cellid, cellid1, mirnaNorm, fieldNamescell] = preprocessing_rnaseq();
mirnaTrain = [];
mirnaTest = [];
% filter through mirna samples and separate into train and test
for x = 1:length(fieldNamescell)
name = fieldNamescell(x);
name2 = chosenTrain(1);
if ismember(name, chosenTrain) == 1
chosenCol = num2cell(mirnaNorm(:,x));
chosenCol = [fieldNamescell(x); chosenCol];
mirnaTrain = [mirnaTrain chosenCol];
elseif ismember(name, chosenTest) == 1
chosenCol = num2cell(mirnaNorm(:,x));
chosenCol = [fieldNamescell(x); chosenCol];
mirnaTest = [mirnaTest chosenCol];
end
end
% split mirna into test and train
fieldNamestrain = mirnaTrain(1,:);
fieldNamestest = mirnaTest(1,:);
mirnaTrainData = cell2mat(mirnaTrain(2:end,:));
mirnaTestData = cell2mat(mirnaTest(2:end,:));
% fs
[mirnaTopData, mirnaTopFinal, mirnaTopFeatures, fieldNamestrain, cellid1] = featureselection_mirna(cellid, cellid1, mirnaTrainData, fieldNamestrain);
% cox
[B, survivalFinal, riskscore, mirnaTest] = cox_mirna(mirnaTopData, mirnaTopFeatures, mirnaTest, fieldNamestrain, cellid1);
B_rnaseq = B;
survivalFinal_rnaseq = survivalFinal;
riskscore_rnaseq = riskscore;
mirnaTest_rnaseq = mirnaTest;
fieldNamestest_rnaseq = fieldNamestest;
% % only use samples common in both modalities
% if length(riskscore_mirna) > length(riskscore_rnaseq)
% ind = ismember(fieldNamestest_mirna,fieldNamestest_rnaseq);
% riskscore_mirna = riskscore_mirna(ind);
% elseif length(riskscore_rnaseq) > length(riskscore_mirna)
% ind = ismember(fieldNamestest_rnaseq,fieldNamestest_mirna);
% riskscore_rnaseq = riskscore_rnaseq(ind);
% end
% updated code
if length(riskscore_mirna) > length(riskscore_rnaseq)
riskscore_mirna = riskscore_mirna(1:length(riskscore_rnaseq));
elseif length(riskscore_rnaseq) > length(riskscore_mirna)
riskscore_rnaseq = riskscore_rnaseq(1:length(riskscore_mirna));
end
% % old code
% ind = ismember(fieldNamestest_mirna,fieldNamestest_rnaseq);
% riskscore_mirna = riskscore_mirna(ind);
% elseif length(riskscore_rnaseq) > length(riskscore_mirna)
% ind = ismember(fieldNamestest_rnaseq,fieldNamestest_mirna);
% riskscore_rnaseq = riskscore_rnaseq(ind);
% end
% average risk scores
riskscore_integrated = (riskscore_mirna + riskscore_rnaseq)./2;
% split risk scores into high and low risk groups
[testSurvivalFinal, survivalTimeHigh, survivalTimeLow, censoredFlagHigh, censoredFlagLow, p] = kaplanmeier(riskscore_integrated, fieldNamestest, fold);
% plot the survival functions
subplot(1,5,fold-1)
ecdf(survivalTimeHigh, 'Censoring', censoredFlagHigh, 'Function', 'survivor'); % high risk group
hold on
ecdf(survivalTimeLow, 'Censoring', censoredFlagLow,'Function', 'survivor'); % low risk group
str = sprintf('Fold %d : miRNA + mRNA', fold-1);
title(str)
xlabel('Survival Time (days)')
ylabel('Probability of Survival')
legend('High Risk Group', 'Low Risk Group')
txt = ['p = ',num2str(p)];
text(300,0.8,txt)
pValues = [pValues p];
% c index
survivalTimes = survivalFinal(:,4);
survivalTimes = survivalTimes(1:length(riskscore_integrated));
survivalTimes = cell2mat(survivalTimes)';
ci = concordanceIndex(survivalTimes,riskscore_integrated);
% ev vs cv
ciCV = [ciCV ci];
% do the same thing for the actual test data
[testSurvivalFinal, survivalTimeHigh, survivalTimeLow, censoredFlagHigh, censoredFlagLow, p] = kaplanmeier(riskscore_integrated, fieldNamestestActual, fold);
% c-index for external validation
survivalTimes = survivalFinal(:,4);
survivalTimes = survivalTimes(1:length(riskscore_integrated));
survivalTimes = cell2mat(survivalTimes)';
ciTest = concordanceIndex(survivalTimes,riskscore_integrated);
ciEV = [ciEV ciTest];
% plot the survival functions
subplot(2,5,fold + 4)
ecdf(survivalTimeHigh, 'Censoring', censoredFlagHigh, 'Function', 'survivor'); % high risk group
hold on
ecdf(survivalTimeLow, 'Censoring', censoredFlagLow,'Function', 'survivor'); % low risk group
str = sprintf('Fold %d : miRNA + mRNA', fold-1);
title(str)
xlabel('Survival Time (days)')
ylabel('Probability of Survival')
legend('High Risk Group', 'Low Risk Group')
txt = ['p = ',num2str(p)];
text(300,0.8,txt)
end
% plot concordance interval for cross validation
subplot(2,5,5)
scatter(cv, ciCV)
title('CV vs CI')
xlabel('Cross Validation')
ylabel('c-index')
axis([1 4 0 1])
% plot cross validation vs external validation
subplot(2,5, 10)
scatter(ciCV, ciEV, 'filled')
title('Cross Validation vs External Validation')
xlabel('Training Performance')
ylabel('Testing Performance')
axis([0.45 0.55 0.45 0.55])
hold on
p = polyfit(ciCV,ciEV,1);
f = polyval(p,ciCV);
plot(ciCV,f)
hold off