Skip to main content

Advertisement

Figure 2 | BMC Bioinformatics

Figure 2

From: Lack of sufficiently strong informative features limits the potential of gene expression analysis as predictive tool for many clinical classification problems

Figure 2

Classifier performance as function of informative sample size and signature strength. Both, the number of spiked probes and the number features included in the predictor model were set to 10. The solid lines indicate the average area above the ROC curve (AAC) from Monte Carlo Cross Validation (MCCV). The smaller the AAC the more accurate the predictor is. The dots represent results from the 20 individual iterations of the analysis performed on the MAQC-II data set (n = 233). The "c" value which is the log2 fold-change takes on the values 0.5, 1.0, 1.2 and 1.5.

Back to article page