The results of our restricted Scenario B comparison are shown in Figure 4, using the same discordance vs. missing genotype percentage format as Figure 3. Note that each panel in this figure follows a different scale. The top two panels (Figure 4A and 4B) share a common set of grid lines, and the bottom two panels (Figure 4C and 4D) share a finer set of grid lines. We omitted the results from the BEAGLE run with 50 iterations since they were only trivially better than the results based on default settings.
