Table 5 Top-performing classifier-modality-configuration combinations for cancer molecular subtype classification.
From: Multi-omics driven computational framework for cancer molecular subtype classification
Cancer | Modality | Configuration | Classifier | ACC | PR | RE | F1 |
|---|---|---|---|---|---|---|---|
ACC | Meth. | HumanMethylation450 | HGB | 1.000 | 1.000 | 1.000 | 1.000 |
BLCA | miRNA | miRNA-HiSeq-gene | RESNET18 | 0.970 | 0.970 | 0.970 | 0.970 |
BRCA | miRNA | miRNA-GA-gene | SVM | 0.880 | 0.810 | 0.880 | 0.830 |
COAD | Exon | GAV2-exon | SVM | 1.000 | 1.000 | 1.000 | 1.000 |
ESCA | CNV | Gistic2-all-data-by-genes | CNN | 1.000 | 1.000 | 1.000 | 1.000 |
GBM | Array | AgilentG4502A-07-2 | XGB | 0.750 | 0.800 | 0.750 | 0.760 |
HNSC | RPPA | RPPA | GNB | 0.710 | 0.710 | 0.710 | 0.700 |
KIRC | RPPA | RPPA | CNN | 0.539 | 0.550 | 0.539 | 0.536 |
KIRP | CNV | Gistic2-all-data-by-genes | DEEPGENE | 0.887 | 0.871 | 0.887 | 0.875 |
LAML | miRNA | GA | SVM | 0.77 | 0.829 | 0.77 | 0.759 |
LGG | CNV | Gistic2-all-thresholded | MLP | 0.660 | 0.580 | 0.660 | 0.610 |
LIHC | miRNA | miRNA-HiSeq-gene | SVM | 0.860 | 0.860 | 0.860 | 0.860 |
LUAD | Meth. | HumanMethylation27 | RESNET101 | 1.000 | 1.000 | 1.000 | 1.000 |
LUSC | Array | HT-HG-U133A | RESNET34 | 0.920 | 0.970 | 0.920 | 0.930 |
PCPG | RNASeq | HiSeqV2-percentile | MLP | 1.000 | 1.000 | 1.000 | 1.000 |
PRAD | RNASeq | HiSeqV2-PANCAN | HGB | 0.990 | 0.990 | 0.990 | 0.990 |
SKCM | CNV | Gistic2-all-thresholded | RNN | 0.460 | 0.410 | 0.460 | 0.430 |
STAD | Meth. | HumanMethylation450 | DEEPGENE | 0.838 | 0.819 | 0.838 | 0.827 |
THCA | Exon | HiSeqV2-exon | HGB | 0.830 | 0.910 | 0.830 | 0.850 |
UCEC | Meth. | HumanMethylation27 | LSTM | 0.881 | 0.937 | 0.881 | 0.900 |
