Figure 1.

Overview of the imputation algorithm. The left panel shows the diagram of magicImpute. The right panel shows the directed acyclic graph of the HMM for N offspring at T loci, where the arrows denote probabilistic relationships that are described in the Methods section. See Table 1 for the symbols in the right panel. In the left panel, the second step of founder imputation results in the estimate of $h_t^F$ and the third step of posterior decoding results in the posterior probability of $x_t^o,$ conditional on genotypic data $y_t^o$ and $y_t^F$ for $t=\mathrm{1...}T$ and $o=\mathrm{1...}N.$