Pharyngula

Posted by: Mike Haubrich, FCD | July 14, 2007 1:14 PM

One aspect of an article I had read on Science Daily which I found interesting is that the Buddenbrockia are useful in studying another parasite, Tetracapsula bryosalmonae:

The new study is also good news for salmon farmers as it may help them to tackle a devastating parasite. Buddenbrockia is closely related to the parasite Tetracapsula bryosalmonae that infects the kidneys and spleens of salmon causing immune cells to multiply out of control destroying these organs and killing the fish. Parasites such as Tetracapsula bryosalmonae are notoriously hard to study because of the difficulty of separating the parasite's DNA from the host's DNA. But because Buddenbrockia sometimes swims free from its host (a moss animal or bryozoan) its DNA can be isolated and analysed: opening up a genetic window onto the mysterious world of such parasites.

I wrote a little bit about this on Saturday discussing the convergent evolution aspect.

Posted by: David Marjanović | July 15, 2007 1:12 PM

Are there sequences for other myxozoans, and if so are they cnidarian too?

Yes and yes, as both papers and both blog posts say.

Posted by: Mikko Kolkkala | July 17, 2007 5:09 AM

Yes, it seems likely that this worm is a Cnidarian. But I would not say this study nails the thing down. I believe more taxon sampling is needed the break down the long branches (how come no other Myxozoan sequences except Buddenbrockia are included in the study? As noted, some Myxozoan sequences reported in genebank are artifacts from the fish hosts.)

Reading "the small print" (supplementary material):

129 protein sequences from 47 animals and 13 outgroups
- Too good to be true? Yes. More than one third of the gene sequences are missing, particularly 56 Buddenbrockia genes (table S3)

Increasing the amount of phylogenetic information by creating chimeric sequences (e.g. human+dog+cow+rat+mouse, table S2)
-Sounds a little bit like cheating to me...

Choosing taxa on the basis of shortest branch lengths
...and this also. What about if these sequences are somehow special?

WAG+Gamma model
- What is the rationale behind choosing the model of the Bayesian analysis? Just some model?

Four other possible phylogenetic positions of Buddenbrockia could not be rejected.
- Which ones? I really cannot understand figure S2. What are those numbers in the tree?

Simple parsimony grouped Buddenbrockia with Nematodes and Platyhelminthes
-I prefer Bayesian inference, but still. CAT-method should account for across-site heterogenities. Perhaps. How does the missing data affect here?