Skip to main content
Fig. 2 | BMC Genomics

Fig. 2

From: Muver, a computational framework for accurately calling accumulated mutations

Fig. 2

Comparison of muver results with other common mutation calling software. a Muver compared to GATK HaplotypeCaller, VarScan, and VarDict mutation calls in a Genome in a Bottle (GiaB) human father-son pair. Presumed false positives are mutations called where none are implied by GiaB genotype calls. Mutation disagreements are mutation calls with positions that agree but mutation identities that disagree with GiaB. Sensitivity is the percent of GiaB genotype differences identified by a given caller, excluding loci with identity disagreements. False positive rates (FPR) are per million base pairs (Mbp− 1) tested by both GiaB and a given caller (i.e. excluding areas of excessively low coverage, etc.). b Comparison of clonal substitution and insertion/deletion calls between muver, MuTect2 and VarScan for Saccharomyces cerevisiae with very low (wild type) and very high (pol3-L612 M msh2Δ) mutation rates. Mutations called by both MuTect2 and VarScan but not by muver are in orange. Mutation calls shared with muver are in green. Mutations called by MuTect2 or VarScan only are in red. Those called by muver but missed by the listed caller are in purple. Statistics from human data in panel A suggest that in panel B most red calls are false positives and most blue calls are false negatives. Mutation rate ratios (pol3-L612 M msh2Δ over wild type) for each caller were calculated from the number of called mutations, the size of the genome, and the number of elapsed cell divisions (see Methods). For comparison, mutation rate ratios from previous URA3 assays are also presented. See Additional file 1: Table S1 for event counts

Back to article page