Figure 1
Overview of the methodology. (a) Schematic of CINner (adapted from9). Cells follow a birth-death process. The probability of division depends on the cell’s fitness, determined based on its copy number profile. When cells divide, new clones may arise according to established CNA probabilities. (b) Selection model (adapted from9). A cell’s fitness depends on its copy numbers and chromosome selection parameters. The fitness increases after a missegregation if the cell gains a chromosome with selection parameter \(>1\), or loses one with selection parameter \(<1\). (c) Application of Approximate Bayesian Computation (ABC) in parameter inference. Parameters are drawn from prior distributions, then statistics are computed from bulk and scDNA-seq samples simulated with CINner. ABC-rf then determines the parameter posterior distributions that can be compared against true values.
