Skip to main content
. 2016 Jan 4;4:e1508. doi: 10.7717/peerj.1508

Figure 4. Effect of individual filters on the overlap between replicates.

Figure 4

For each filter (names given on the x-axis, with detailed description in Table 1), we looked at (A) the number of mutations removed by a given filter in a given sample on a log scale and (B) the percentage of the WGS–WGA overlap removed by the filter, per sample. The LOH and VAQ filters removed a large number of putative SNVs and portion of the overlap. The <10% filter removed very few putative SNVs and almost none of the overlap. The 10bp-SNV, 10bp-INDEL, and dbSNP filters removed an intermediate number of putative SNVs (100), but only a very small portion of the overlap, making them the best performers on the overlap data.