Table 16 RNA-Seq Data – Macro-averaged Accuracies in % (LOSO-CV with LR Models). \(\emptyset \): Entire gene set. The best overall results are highlighted in bold
From: Systematic detection of predictive gene sets by semantics-based selection
Keyword Set | # Genes | TN | FP | FN | TP | MAA (%) |
|---|---|---|---|---|---|---|
Leukaemia Data Set | ||||||
\(\emptyset \) | 35,087 | 8 | 0 | 1 | 64 | 99.2 |
\(\{\)’Lymph’\(\}\) | 149 | 8 | 0 | 0 | 65 | 100 |
\(\{\)’Blood’\(\}\) | 772 | 8 | 0 | 0 | 65 | 100 |
\(\{\)’RNA’\(\}\) | 5,358 | 8 | 0 | 0 | 65 | 100 |
\(\{\)’Plasma’\(\}\) | 5,685 | 8 | 0 | 0 | 65 | 100 |
Liver Disease Data Set | ||||||
\(\varvec{\emptyset }\) | 40,131 | 26 | 1 | 0 | 216 | 98.1 |
\(\{\)’Gland’\(\}\) | 364 | 26 | 1 | 0 | 216 | 98.1 |
\(\{\)’Lipid’\(\}\) | 1,134 | 26 | 1 | 0 | 216 | 98.1 |
\(\{\)’Hepatocyte’\(\}\) | 130 | 26 | 1 | 1 | 215 | 97.9 |
\(\{\)’Blood’\(\}\) | 768 | 26 | 1 | 1 | 215 | 97.9 |
\(\{\)’RNA’\(\}\) | 5,334 | 26 | 1 | 1 | 215 | 97.9 |
\(\{\)’Liver’\(\}\) | 117 | 26 | 1 | 2 | 214 | 97.7 |