Table 1 Accuracy of different biomarkers across experiments by 10-fold cross-validation
Biomarkers | Dataset | Mean ± Variance | ||||
|---|---|---|---|---|---|---|
GSE18732 | E-MEXP-2559 | GSE20966 | GSE23343 | GSE26887 | ||
Module biomarker (32 genes) | 92.39% | 80% | 80% | 88.24% | 83.33% | 84.79% ± 0.054 |
SVM-RFE | 67.39% | 80% | 85% | 58.82% | 75% | 73.24% ± 0.103 |
PAC | 84.78% | 75% | 60% | 47.06% | 83.33% | 70.03% ± 0.16 |
32 top differentially expressed genes | 73.91% | 75% | 75% | 82.35% | 100% | 81.25% ± 0.11 |
Type 2 diabetes mellitus | 55.43% | 80% | 90% | 35.29% | 91% | 70.34% ± 0.24 |
B cell receptor signalling pathway | 60.87% | 85% | 95% | 70.59% | 83.33% | 78.96% ± 0.133 |
Toll like receptor signalling pathway | 48.91% | 70% | 80% | 88.24% | 83.33% | 74.1% ± 0.156 |
Biosynthesis of unsaturated fatty acids | 55.43% | 80% | 65% | 88.24% | 75% | 72.73% ± 0.128 |
Insulin signalling pathway | 59.78% | 85% | 85% | 58.82% | 100% | 77.72% ± 0.179 |