Table 2.
Prediction ability of three classifiers. These classifiers were build using GSE51618, GSE26966 and the combination of GSE22812, GSE26966, GSE4237, GSE46311, and GSE51618. Prediction ability was assessed by the integration of GSE36314 and GSE2175 as the independent validation dataset.
| Datasets | TP a | FN a | TN a | FP a | ACC b | SEN b | SPE b | MCC b | AUC c |
|---|---|---|---|---|---|---|---|---|---|
| A: GSE51618 | 4 | 4 | 4 | 0 | 0.67 | 0.50 | 1.00 | 0.50 | 0.75 |
| B: GSE26966 | 5 | 3 | 3 | 1 | 0.67 | 0.63 | 0.75 | 0.35 | 0.72 |
| C: DDI Strategy | 7 | 1 | 4 | 0 | 0.92 | 0.88 | 1.00 | 0.84 | 0.97 |
a TP: true positive; TN: true negative; FP: false positive; FN: false negative. b ACC: accuracy; SEN: sensitivity; SPE: specificity; MCC: Matthews correlation coefficient. c AUC: area under the curve.