Research Article
Multiclass Prediction with Partial Least Square Regression for Gene Expression Data: Applications in Breast Cancer Intrinsic Taxonomy
Table 3
Performance of PLS-regression classifiers for independent validation dataset.
| Intrinsic subtype | Basal-like | HER2-enriched | Luminal-A | Luminal-B | Normal breast-like |
| Number of samples | 97 | 94 | 165 | 121 | 56 | PLS-regression | | | | | | Number of gene component | 2 | 1 | 2 | 2 | 1 | -variance explained | 71.1% | 25.5% | 79.9% | 61.7% | 38.1% | -variance explained | 56.9% | 41.6% | 34.5% | 30.9% | 18.1% | Binary LR | | | | | | Adjusted -square | 0.86 | 0.66 | 0.73 | 0.61 | 0.39 | AUC | 0.98 | 0.95 | 0.95 | 0.93 | 0.89 | Accuracy | 96.6% | 90.7% | 88.2% | 86.2% | 90.5% | Sensitivity | 85.6% | 68.1% | 81.8% | 63.6% | 23.2% | Specificity | 99.1% | 95.5% | 91.1% | 92.8% | 98.3% |
|
|
PLS: partial least square, LR: logistic regression, AUC: area under the curve.
|