Figure 2.—

DAG of the model given in Equation 8. The square node denotes known quantity (i.e., data) and circles represents parameters to be estimated. Lines between nodes represent direct stochastic relationships within the model. The variables within each node correspond to the different model parameters discussed in the text. N is the genetic data, is the vector of inbreeding coefficients, and are, respectively, the actual and ancestral allele frequencies, is the vector of the genetic differentiation coefficient for each local population, and a is the hyperprior determining the shape of the ancestral allele frequencies.