Figure 5

Comparisons of recall rates between dChip, SAM, and SiPaGene. Experiment 8 was compared to all other experiments with all three tools. A) The mean and standard deviation (SD) of all probe sets identified by the different tools is displayed. Part B) presents the mean and SD of correctly recalled probe sets and part C) the mean and SD of false positive probe sets. The optimized filter strategy implemented in SiPaGene identified the number of probe sets with the highest recall and the lowest error rate. The relatively poor outcome of SAM results from a singular high false positive recall between experiment 8 and 4 (n = 326) indicating that this algorithm may fail in few exceptions.