Have a look at the Δm, n, 2 statistic for breakpoint detection

[lenaschimmel]

I was contacted by @maciekboni who recommended the Δm, n, 2 statistic. It looks very relevant for the case of two potential parent lineages.

Still I did not find time to take a closer look, and make up my mind if/how it fits with my mid-term plans of adjusting my algorithms, especially to include the individual mutation prevalences (#15) into the calculation.