Figure 4
From: SimBA: simulation algorithm to fit extant-population distributions
SimBA simulation results. (a) SimBA hill climbing algorithm for JPT/CHB population with k=10. LD fit, MAF fit, and heatmap of LD for each pair of columns (upper left triangle is the target and lower right triangle is the constructed). The LD fit shows the target “o” and constructed “*” mean r 2 per distance, while the black dots show target and cyan dots constructed r 2 distribution per distance. (b) Constructing three subpopulations with F st constraints \(F_{s_{1}T}=0.2\) and \(F_{s_{2}T}=0.1\). MAF and r 2 constraints from ASW population. Population size n=400 constructed as subpopulations with sizes n 1=200,n 2=100,n 3=100. Here k=7, distances 1–6 and j−1 are used per column j. Stars denote constructing a single population without F st constraints, squares denote the combined population with three subpopulations constructed with F st constraints. Each subpopulation is also shown separately.