Fig. 3
From: Distinguishing linear and branched evolution given single-cell DNA sequencing data of tumors
Phyolin results of the in silico experiments on a simulated cohort of patients with AML. a A comparison of Phyolin runtimes in seconds between instances with different evolutionary patterns. b A comparison of the distribution of \(\hat{\beta }\) between simulated linear and branched topologies over 10 replications. c The mean value of \(\hat{\beta }\) for each patient along with the standard error. d The distribution of Bayes factor K for each simulated patient separated by the true evolutionary pattern. The horizontal line at \(K=1\) represents the classification criteria (\(K > 1\) is classified as linear). e Relationship between estimated false negative rate and the ancestor–descendant distance. Each point represents the mean value over 10 replications and is labeled by the numerical patient identifier. A linear trend line is shown with a \(95\%\) confidence interval