Reader Comments

Post a new comment on this article

Referee Comments: Referee 3

Posted by PLOS_ONE_Group on 13 May 2008 at 23:55 GMT

Referee 3's Review:

**********
N.B. These are the comments made by the referee when reviewing an earlier version of this paper. Prior to publication, the manuscript has been revised in light of these comments and to address other editorial requirements.
**********

This paper analyzes mtDNA and non-recombining Y-chromosome (NRY) variation in the Druze, a religious minority in the Near East. The authors find a high frequency and high diversity of mtDNA haplogroup X, which is otherwise found widespread geographically but at much lower frequencies, and suggest this might reflect the ancient structure of Near East populations.

While the paper is generally sound, there are a number of ways in which the analyses could be improved, and there are some alternative interpretations which should be mentioned. Specifically:

p.3, end of first paragraph: the authors claim that the peculiar patter of haplogroup X, namely high diversity and low frequency over a widespread geographic region, suggests that haplogroup X was "geographically concentrated" for much of its history, resulting in high diversity, followed by migrations to many parts of the world. But this is just one possible scenario; it is also possible that haplogroup X was dispersed early on in history, followed by diversification in multiple populations. One way to test these (and other scenarios) would be demographic modeling via simulations, as has been carried out recently in other studies of human population history (e.g., Tishkoff et al., MBE Advance Access pulbication July 26, 2007), to see if the data can distinguish among these various possibilities.

p.4: The Druze are referred to as having a "secret religion," what this means and the relevance for this study should be clarified as it is not clear what this is supposed to mean.

p.8: the formula in the middle of the page is not realistic as it assumes that each lineage contains the same number of individuals (i.e., each of the N lineages contains x individuals; a more general formulation would allow the number of individuals per lineage to vary, such as each of the Ni lineages to have xi individuals. Also at the bottom of this page, the authors estimate the probability that the Druze could have arisen via sampling from a source population, but the calculation is done with the assummption that current haplogroup/lineage frequencies are reasonable proxy estimates of past haplogroup/lineage frequencies. As mentioned in the first comment above, a better approach would be to model demographic history, thereby allowing for differences in ancestral frequencies and changes in population size, to see which historical scenarios can be supported/rejected by the data.

p.10, first complete sentence, "Mitochondrial DNA analysis..." is vague, do the authors mean here the HV1 sequences, complete sequences, or some other analysis?

pp.13-14, again the analytical methods here are all unrealistic as they assume that current haplogroup frequencies are reasonable estimates of ancestral haplogroup frequencies. This may or may not be true, but in any event it is not necessary to make this assumption if instead demographic modeling is carried out.

p.15: the authors make the point that their directed sampling detects higher diversity than a truly random sampling does. While this is true, this observation raises a larger issue that the authors ignore but that deserves some discussion: what is the most appropriate sample to obtain, directed or randome? If the goal is to discover as much diversity as possible, then it would seem that directed sampling is preferable. However, one could also argue that if the goal is to get an unbiased estimate of population diversity, then random sampling is preferable, as directed sampling does give a (upwardly) biased estimate of the diversity. This is an interesting issue (what should be the goal of population studies?) that may not have a simple answer but nonetheless deserves further discussion.

p.16: the authors give coalescence times for the major X subhaplogroups, but the method used to estimate these times assumes population expansions. Again, demographic modelling is preferable, to see if there is indeed a signal of population expansion in the data. Also, the middle of the page suggests that there have been changes in the Near East genetic landscape over time, which further makes unlikely the assumption that current haplogroup frequencies can then be used as estimates of past haplogroup frequencies. This point is also made on p. 17, first sentence of the Conclusions.