Referee 1's review:

This is an interesting paper which examines Y chromosome haplotypes in three separate regions of the island of Sardinia. The authors note that there is a high degree of interindividual variation but a homogenous distribution of the detected variability in samples from the three separate regions. One haplogroup, I-M26, is rare or absent outside Sardinia and is very common (0.37 frequency) throughout the island, consistent with a founder effect. A Bayesian full likelihood analysis has indicated that the time from the most recent common ancestor of I-M26, is 21.0 (16.0-25.5) KYA and that the population began to expand 14.0 (7.8-22.0) KYA. These results suggest a largely pre-Neolithic settlement of the island with little subsequent gene flow from outside populations. The authors conclude that Sardinia is thus, an especially attractive venue for case-control whole genome association scans in common multifactorial diseases.

Specific comments are as follows:

1) The authors should define KYA when they first mention it for readers who do not understand what it means.

2) I am not sure if Nuraghe are a strong argument for a common cultural background of the peoples of Sardinia. Admittedly, they are present all over the island and not anywhere else in neighboring countries. Castles are found all over Europe, but one would likely not say that Europe had a common cultural background at their time of construction; they were built because each group saw their strategic value when facing an opponent with them and thus they propagated throughout the surrounding region. The presence of Nuraghe all over Sardinia could be the result of mimicry by rival tribes on Sardinia, but their absence elsewhere would support their creattion in Sardinia. Additional arguments should be provided to support this assertion.

3) On page 11, paragraph 2, the authors state that due to the presence of Y chromosome haplotypes from eastern and western Europe, analysis in the Sardinians would have a high chance of detecting common disease variants present in other European populations. This argument needs to be hedged, as it is reliant upon the fact that these disease variants arose prior to the populating of Sardinia. Due to their isolation for ~4000 years, and major population expansion in the last 300 years, which will encompass the majority of the historical meiotic events from which common variants can arise, one may find that different risk variants arose in the Sardinian population compared to the rest of Europe, although admittedly it is expected that there will be some variants that do predate the peopling of Sardinia. The question will be: what risk is attributable to those historical variants compared to the more recent variants.

4) The authors state on page 11, paragraph 2, that the Sardinian population has highly informative haplotype splits across the genome. This statement cannot be made based upon their results from their data on the Y chromosome alone, but the authors do not provide the source of data that leads them to draw this conclusion. Is this inferred from data from studies by other groups or these investigators? This statement needs to be backed up by one or more references.

5) The UPGMA tree presented in Figure 2 should be redrawn using the consensus of 1000 bootstraps and the appropriate bootstrap confidence of each branch shown in the figure. This will give the reader an appreciation of the confidence of the pattern that the tree presents.

----------
N.B. These are the general comments made by the reviewer when reviewing this paper in light of which the manuscript was revised. Specific points addressed during revision of the paper are not shown.