Figure 4.

Results of simulation studies for Experiment 1 (diploid cell lines HCT116 and 184-hTERT-L2). (a) Experiment 1, V-measure, downsampling read depth. The vertical axis indicates the V-measure score obtained by each algorithm for each read depth considered when downsampling the number of variant and reference reads aligned to each SNV. (b) Experiment 1, V-measure, downsampling the number of targets. Box plots of the V-measure scores obtained by applying each algorithm to each simulated data set under each number of targeted SNVs considered. Note that outliers were omitted from the box plots. (c) Experiment 1, V-measure, downsampling the number of samples. Box plots of the V-measure scores obtained by applying each algorithm to each simulated data set for each possible number of samples considered. Note that outliers were omitted from the box plots. (d) Experiment 1, prevalence error, downsampling read depth. Box plots of the absolute errors in estimating cellular prevalence obtained by applying each algorithm to each read depth considered. Note that outliers were omitted from the box plots. (e) Experiment 1, prevalence error, downsampling number of targets. Box plots of the median absolute errors obtained by applying each algorithm to each simulated data set under each number of targeted SNVs considered. Note that outliers were omitted from the box plots. (f) Experiment 1, prevalence error, downsampling number of samples. Box plots of the median absolute errors obtained by applying each algorithm to each simulated data set under each number of samples considered. Note that outliers were omitted from the box plots.