Well, going out on a limb here, there should still be some way to give a statistical apprehension of how likely such mutation could be over time. But I'm not aware if there's any feasibility of doing that today.