Fig. 5
From: Characterization of background noise in capture-based targeted sequencing data
The false positive rate. a The error rate of each background allele was calculated and their distribution was plotted, dependent on the total read counts. Data from 19 plasma DNA samples were down-sampled to a designated size of total reads in a range between 2.5 and 50 M. The x-axis denotes the frequency of background alleles, and the y-axis denotes the fraction of alleles with the designated background rate on the x-axis. b–d The fraction of background alleles present at a frequency greater than a given threshold is plotted against the depth of coverage after de-duplication (x-axis). b The effect of coverage depth on the false positive rate is shown in the down-sampled data set generated from 19 plasma DNA samples. c The effect of each DNA shearing condition on the false positive rate was estimated using the down-sampled data set from Fig. 3. a–d indicate the fragmentation conditions, as described in Fig. 3. d A comparison of plasma and fragmented PBL DNA samples