- The question "is this true" is ambiguous and thus I am not sure how to reply. If the question is in regards to our phylogeny estimation, then the answer is certainly "no" because we would never publish phylogenies and state that they are "true".

- Phylogenies are predictions...hypotheses, given the data, taxa and method of tree estimation. All of these factors influence the results and in no portion of our manuscript are the phylogenies treated as anything more than hypotheses.


"Except for the case of the patatine protein (pRF11), the plasmid proteins are always the most divergent sequences and should be reasonably used as outgroup in the trees of figure 3."

-The most ancestral taxon in our analyses, R. bellii, was used to root the trees in Figure 3 because that is the logical method for determining character polarity, and because our robust phylogeny estimate (Figures 1 and 2) provides the topology to explore the relatedness of the pRF genes to the other rickettsial genomes.


"The assertion that there is an affinity of pRF with AR rickettsiae is certainly wrong."

-Again, phylogenies are estimates of historical divergence and we feel that we have done our best with predicting these trees. Claiming that our observations are " certainly wrong" with the absence of an alternative argument does not amount to much. This statement is unprofessional and, in fact, violates the second guideline for posting here, which states "Unsupported assertions or statements should be avoided. Comments must be evidence-based, not authority-based".

-Furthermore, the data are freely available and certainly open to alternative interpretation.


"The way the trees in figure 3 are displayed (R. bellii used as outgroup) is misleading, though most of the trees are probably correct on a topological basis. Upon reanalysis of the plasmid sequences, I have a serious doubt regarding the topology of Fig3 B and F.

-Please see above regarding "correct topologies" and "outgroup selection".

-Regarding "I have a serious doubt regarding the topology of Fig3 B and F", the data are freely available and our analyses should be repeatable by anyone interested in testing our results.

-Finally, the focus on proving the phylogeny estimations and the conclusions drawn from their inference "wrong" is a narrowly focused approach to negating our hypothesis of relatedness between the pRF plasmid and AG rickettsiae. We also offer evidence from gene distribution (Tables 1 and 2) that shows more pRF genes are found in AG rickettsiae genomes than the other groups. Thus we conclude from BOTH phylogeny estimation AND gene distribution that the pRF plasmid likely has had exchanges with putative plasmids or chromosomes of AG rickettsiae.