Network-Based Inference Framework for Identifying Cancer Genes from Gene Expression Data
Figure 2
Comparison of accuracy and robustness between netSAM, -test, and lasso on 100 synthetic datasets. (a) ROC curves: true positive rates against false positive rates. (b) FDR curves: error discovery rates against true positive rates. (c) PPV curves: precision versus recall value. (d) Robustness values (AUC versus SNR) are calculated based on five-fold cross-validation, where standard deviations are shown in error bars.
