Reader Comments

Post a new comment on this article

Referee Comments: Referee 1 (Erich D Jarvis)

Posted by PLOS_ONE_Group on 09 Oct 2007 at 16:54 GMT

Reviewer 1's Review (Erich D Jarvis)

-----

The paper by Li et al on "Rapid evolution of FoxP2 evolution in echolocating bats" is clearly very interesting in that without question they found that bats have a higher prelevance of mutations in the FoxP2 gene than in other vertebrate species, including other vocal learning or echolocating animals. The paper is certainly expected to generate useful discussion on the evolution and function of this gene in vocal communication behavior. Another interesting finding I noted is that a subset of the bat species they examined have one of the mutations that has been reported to be specific to humans and associated with the evolution of language. The authors comment very little on this finding. Overall, the paper is massive effort and I think it will be interesting to a broad audience. However, the paper also has some significant flaws that I believe needs to be fixed and would require a major revision. These problems are in the manner of presentation of the data, limited information given to the reader making some of the data difficult if not impossible to evaluate (and the paper is written too much for the specialists), problems with some of the interpretations, and a problem with distinguishing positive selection from relaxation.

The last problem is probably the most important one. It seems that the authors used the original PAML branch-site model to determine positive selection (PS), which is appropriate here. If true, they should explicitly state this in order to avoid confusion with the modified one that is used more extensively for positive selection (PS) detecting (Zhang et al Mol Biol Evol 2005, 22:2472-2479). With the model they used it is difficult to determine if Foxp2 mutations have had an active role (PS) or a passive role (RSC), because it cannot discriminate RSC from PS effectively. Therefore, it would be helpful if they re-analyze the data using the modified branch-site model for PS detection.

-----

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.