Skip to main content
Fig. 4 | BMC Medical Genomics

Fig. 4

From: Identification of single nucleotide variants using position-specific error estimation in deep sequencing data

Fig. 4

Benchmarking AmpliSolve calls with Illumina WGS calls. a Venn diagram of mutations on 10 samples sequenced with both Ion Torrent and Illumina platforms and called respectively by AmpliSolve and by MutPlat. Low coverage positions denote mutations excluded by AmpliSolve because poorly covered (< 100 reads on at least one strand, ‘uncallable’ by AmpliSolve). b Scatter plot of VAFs in WGS and AmpliSeq data. Note that all the SNVs not called by AmpliSolve (green point) have some support in the data and are reported in its output (hence they have AF > 0) but are filtered out, mostly because of strand bias. c Same as (b) but for VAFs< 20%. Note that some concordant calls (purple points) have WGS AF = 0; these are real germline variants with no support in the tumor (Methods). For the sake of this comparison, both in (b) and in (c) we don’t consider the 49 mutations at positions of low coverage in Ion Ampliseq data (see (a)) (‘uncallable’ for AmpliSolve)

Back to article page