Referee 1'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 interesting manuscript provides a comprehensive mitochondrial portrait of non-Ashkenazi Jews worldwide and thus complements an earlier study on mtDNA in Ashkenazi Jews. The topic is certainly of sufficient interest to a wider audience - but, of course, also to mtDNA specialists because of the wealth of information given. In general, the paper is well written, but some technical issues need attention.

I find the haplogroup naming and haplotype reference quite confusing throughout the main text and the supplements. It is certainly no good practice (and not in line with Richards et al. 1998) to give an apparently new name such as X2e1 to a haplotype that is obviously sitting in a nesting of clades within haplogroup X2e. To my understanding, X2e1 should be defined by the rare 16189A transversion (notwithstanding that this mutation has once mutated back in a subclade). Then the next level could become X2e1a as defined by 9380 and the less stable 16134; finally X2e1a1 would then be the clade found in the Libyan and Ethiopian Jews. It is, however, not necessary (and perhaps not even desirable) that every major "Jewish" haplotype became the root of a named clade. It would thus be fine to speak informally of the X2e1a-13789-16311 clade or the X2e1-9380-16134-13789-16311 clade, depending on the resolution that one is willing to name.

In any case, a table should be added to the main text that describes all new sub-haplogroup names along with their characteristic (coding- and control-region) mutations. It cannot be justified expelling all control-region mutations from potential diagnostic status. When previously given names are modified, this should be stated explicitly; references to those papers for nomenclature should be given where the clade names could be traced otherwise. If additional names are used for specific haplotypes where the subclade-defining mutations could not yet be identified; then these new codes should be listed in that table as prospectice candidates of clades to be defined properly in the future.

As to further nomenclature in figures and tables, deletions should be encoded by "d" or "del" rather than "D", since the latter symbol stands for "A or G or T". The asterisk symbol attached to a haplogroup name has been used in a different way in the past. Use another symbol when referring to an mtDNA type found in the Ashkenazi Jews. It is not clear what is meant by a "reversion". Is this a "reversion" relative to the pathway to the rCRS or to the global L root? Wouldn't it be more common to use the "@" symbol for this? In any case symbols are missing in the figure. For example, 60.1T is on the link to R0a1 as well as on the way to R0a1c. In U1a1c the transition pair 309-310 is recorded: this alignment is quite implausible and should rather be turned into a deletion-insertion event affecting the two C stretches around 310.

p.6 Change "this criteria" to "this criterion".

p.20 Haplogroup L encompasses all sampled human mtDNAs; what is meant instead is haplogroups or haplotypes outside M and N.

p.24 typo: "the presencs"

p.43 Change "single nucleotide polymorphisms" to single nucleotide polymorphism".

Figure 1: What is the rationale for this tree, especially in presenting multifurcations where some nesting is known to exist (e.g. within haplogroup U)? Which three lineages were intended to get marked by an asterisk in the figure? I would prefer seeing the few recurrent mutations (such as 3010, 5899+C) underlined. Then it may be clear that this tree (with respect to the selected sites) is the unique most parsimonious tree.

Figure 2: Haplogroup labels should be put into the appropriate places of each phylogeny; otherwise it remains from the figure alone unclear which part e.g. is HV1a2 etc. Why is median-joining used here when the data sets are homoplasy-free and thus supported by the perfect phylogenies shown?

Table 1: What is informally meant by "starlike"?

Table 5: There is some discrepancy between the listed type 16134 16189 16223 16278 16311 and the labeling in Fig.2c: in the latter, 16189 placed on the corresponding link would rather suggest a transition from A to G. However, the table suggests that A at 16189 went back to C or might have been deleted.

Suppl. figure S2: Is this tree provably most parsimonious or tentatively - or what is meant by "a parsimonious tree"?

The English style of the text may need some improvement in a few places. For example, I found the following expressions/sentences a bit odd: "dominated by specific for the subcontinent Hg M sub-branches" (p.12); "shows phylogenetic reconstruction of the Georgian Jewish" (p.13); "carrying transition at 16355. Interestingly, substitution at 4227 was missing" (p.14); "not so far sampled elsewhere in Gujarat" (p.21); "actually accentuates by contrast the magnitude of the founder effect" (p.26); "However, it should be kept in mind, that since both the real initial conditions in terms of Hg and lineage composition, and the subsequent evolutionary process that each community underwent from the time of its establishment are unknown, it is obvious that any lineage present in the contemporary population, irrespective of its current frequency, might have been a founder lineage at time the community was established, while an unknown number of founder lineages might have been lost" (p.30); "Samples Ashkenazi Jews" (Supplement).
In the ms one finds three times "time frame" and six times "timeframe" - opt for one alternative.