I have a general question about admixture dating methods (for SNP data). The available methods like ROLOFF or ALDER model the testing population as being the composite of two source populations (or surrogates for these populations). However, is it also possible to date single ancestry components within a population? Take an admixed population with 3 ancestry components (A,B,C) - is it then possible to date admixture from B by the choice of the surrogate populations alone? However, how should this be done ideally because there are always two of them so do you keep a single surrogate constant and then use another population to date ancestry component A and then B?