A useful rule for reading ancestry plots is this: matching a population’s broad genetic position by mixing other populations does not necessarily reproduce its ancestry relationships.
A population can sit roughly halfway between two others without being genetically equivalent to their mixture. Combining those two sources may recreate its position on a plot while failing to recreate the pattern of genetic relationships inherited from its history. The rule concerns what a visual match establishes; it does not say that every possible two-source model must fail.
Imagine a child with one Italian parent and one Han Chinese parent. On a genetic ancestry plot, that child might sit between the two parental populations. Now imagine finding someone from an established population at almost the same position. Could the child’s two-source recipe explain that person too?
I tested this using a Hazara individual and simulated children of Italian and Han Chinese parents. The Hazara are an ethnic group whose traditional homeland lies in the mountains of central Afghanistan. They speak Hazaragi, a variety of Persian. The Chinese reference used here consists of Han Chinese sampled in Beijing. [5]
The selected Hazara individual and one simulated child are both roughly halfway between the Italian and Han references on the first two axes of the plot. Yet the same two-source model does not account for their wider ancestry relationships. That is the distinction this experiment demonstrates.