> John Harshman wrote:
>>>>On the mt DNA it's a straight line on the female side. All other data is
>>>missed. In other words, "Who's your mama?" is the only question that can be
>>>asked and answered. Under the circumstances mixing in the New World doesn't
>>>mean squat as long as the mt DNA in the person can be localized to some
>>>specific part of Africa as the source. Of course that may be absolutely the
>>>only thing the person inherited from that ancestor.
>>The real problem with this is that human populations mix too much for
>>there to be many of what are called "private alleles". Very few
>>haplotypes are found only in a single ethnic group. They may be found in
>>different frequencies in different ethnics groups, but at most that
>>would give you a probability distribution, stating your chances of
>>having come from this group or that one. If the allele is broadly
>>distributed, that list would be a long one and the chance for each
>>individual ethnic group would be low.
>>Also, a proper estimate of the frequencies in different groups would
>>require large samples of those groups, which mostly hasn't been done.
>>That should be obvious, since many of the haplotypes apparently have
>>never shown up in the sample at all.
> That's what I thought all along.
Of course you did.