Figure 10.

Accuracy depending on the similarity score used as threshold in the agglomerative clustering. Thresholds 10-8, 10-7, . . ., 10-3, 0.002, 0.01, 0.05, and 0.1 have been tested for scenario B (see Figure 6(b)) and coverage 60×. The default threshold as computed from the data was approximately 0.002, marked with a *. The minimum and maximum thresholds are marked with a + and ×, respectively. The true and false positive rates for non-overlapping and pairs of overlapping deletions are shown.

Wittler BMC Genomics 2013 14(Suppl 1):S12   doi:10.1186/1471-2164-14-S1-S12