False discovery rate analysis of cnvHiTSeq. (a) FDR of our method for different depths of coverage. FDR determined using segregation analysis on the HapMap CEU trio. (b) FDR of cnvHiTSeq versus the number of detected CNVs. Our method maintains a low FDR even when calling twice as many deletions as competing methods.
Bellos et al. Genome Biology 2012 13:R120 doi:10.1186/gb-2012-13-12-r120