Skip to main content

Archived Comments for: Structural analysis of polarizing indels: an emerging consensus on the root of the tree of life

Back to article

  1. I am suppose that PyrD polarization in this article is incorrect

    Valery Anisimov, A2iA (Artificial Intelligance & Image Analysis)

    7 December 2010

    For sure this paper propose a new interesting method for the indels polarization based on the structural analysis. But on the same time we should be careful in this point because without clear rules of it implementation sometimes we can come to the wrong conclusions. I think that PyrD polarization made by authors in this paper is exactly the case of applying such muddy logic.
    First of all the authors claims that polarization of the PyrD indel using HemE protein as an outgroup made by Skophammer and others is questionable. I am totally agree with this statement. But then they are also rejected polarization of the PyrD based on the consideration of HisA and HisF proteins as outgroup, claimed that "indel arguments contradict themselves". I do not see a logic in the last statement. Why if polarization based on the one outgroup is wrong it should contradict to results of the another polarization for which authors didn't provide any proof of it incorrectness? I think that most correct decision in this point was to say that probable incorrectness of the PyrD - HemE pair polarization mean that Archaea and Firmicutes can not be excluded from the root and PyrD - HisA (or HisF) polarization (which is out of questions) excluded the root from Gram-negatives and the Actinobacteria. Instead the authors switched from the indel analysis to "another line of reasoning" based on the structural analysis of the PyD proteins. But the method of such analysis in this case also not obvious at all (at least for me). In fact the main statement of the authors on which they based the logic of the final conclusion is that an evolution in the direction monomer -> homodimer -> heterotetramer with emerging a new protein-protein interface at each step is much more probable than reverse one even if indel analysis supporting the second scenario. So what is the reason to think that joining of few proteins in one protein complex during evolution is much more probable than disassembling this protein complex on few subunits? And even more than that, at least in this concrete case I believe that we have some proof that evolution really proceeded in the direction heterotetramer -> homodimer -> monomer not vice versa! In fact the authors themselves made the first step in this proof. They are investigated the distribution of the PyrD heterotetramer - PyrD 1B among bacterial taxons. Initially they said that PyrD 1B (PRK07259) protein was found across the Archaea and Firmicutes. Later on they noted that PyrD 1B is presented also in one subgroup of Chloroflexi. They explained this fact by probable HGT. But I should say that myself request to NCBI data base give me not only 6 strains in the group Chloroflexi, but also 7 strains in the group Aquificae, 9 strains in the group Bacteroides/Chlorobi, 8 strains in the group Deltaproteobacteria and 1 strain on the group Fusobacteria. It is certainly very difficult to explain such representation in the different bacterial groups only by possible HGT. So I consider this as an additional argument that PyrD protein complex was presented (as heterotetramer) already in the genome of LUCA. Or by other words this is one more egg to the bucket of the hypothesis that root of life was inside Firmicutes.

    Competing interests

    None declared

Advertisement