Re. your question, let me try to refine the argument. For LDLC, there is an actual number of molecules in the sample. We can (hypothetically) count them and always reach the same answer. A genetic risk is a statistical construct, which depends on assumptions. 2/3