Clustering-Based Method for Developing a Genomic Copy Number Alteration Signature for Predicting the Metastatic Potential of Prostate Cancer

Figure 4

Receiver-operating characteristic curves showing the performance of our CNA-calling algorithm on the simulated data [13]. Each panel represents a different signal-to-noise ratio and the curves represent varying event widths of the simulated data. The 𝑥-axis represents the false positive rate, and the 𝑦-axis represents the true positive rate. Each curve is generated by testing varying thresholds on 100 simulated chromosomes for the condition specified. The curves are combined using vertical averaging. The dashed line represents the random model.