Skip to main content
. 2016 Mar;18(2):283–298. doi: 10.1016/j.jmoldx.2015.11.004

Figure 7.

Figure 7

Distinguishing tumor samples from normal tissue based on DNA methylation in endometrial, colon, stomach, lung, and breast samples. Cumulative distribution functions (CDFs) (top panel) and receiver operating characteristic (ROC) curves (bottom panel) are shown. CDFs of normal and tumor samples are in black and gray, respectively, plotted against a logarithmic x axis. ROC curves reveal the point of the maximal sum of sensitivity and specificity (gray dot). Each column contains CDFs and ROC curves corresponding to a different sample measurement, scaled to vary between 0 and 1. A: Mean fraction (percentage per 100) of methylated CpGs per sample, m. B–D: The results for the x, y, and z ratios, respectively, defined in the text. FPR, false-positive rate (ie, 1 − specificity); TPR, true-positive rate (ie, sensitivity).