African American DNA Research Forum
Re: Testing on Your Own
In Response To: Re: Testing on Your Own ()
Thanks for posting.
I see that OmniPop is doing the calc somewhat differently by first dividing the allele value into 1 (1/(allele value)). This is why the values don't match and also why they are using E+ vs E-. There values are in the millions and they then take the smallest as the likeliest.
When comparing two populations it is fairly straight forward like dividing a pizza into two slices and guessing which slice is bigger. In the case of multiple populations, you have multiple thin slices of pizza. Comparing one slice against the other unfortunately doesn't provide the whole picture. Each slice's space of the total pizza is related the probability (or percent) of the total. (You'll excuse me...I'm hungry)
Now, if I take the populations from OmniPop, divide each value into 1 to get the correct E- value and summarize these into the top 10 geographic populations (with the 10th being OTHER), then the results are quite interesting and enlightening.
Messages In This Thread