B&K base their analyses on a database of 30,000 HV1 sequences, which they claim is sufficient to identify "… idiosyncrasies of data sets that cannot be attributed to natural evolutionary processes." We disagree: 30,000 sequences may sound like a lot, but actually this represents only 0.0005% of the potential HV1 sequences in the human population. Moreover, the existing sequences are nowhere near a random sample, as many areas of the world are under-represented in the studies that have been done to date. Given that 99.9995% of the data is missing, and that large areas of the world are missing in the 0.0005% of the data that does exist, it is hardly surprising that new patterns of variation will be found in new populations.
Annals of Human Genetics (Online Early)
The Patient is Not Dead Yet: Premature Autopsy of a mtDNA Data Set
M. Stoneking and I. Nasidze