Wednesday, July 13, 2016

What's in a name? Would a Wolbachia by any other name, block RNA viruses as sweetly?

I've ranted about bacterial nomenclature in the past (see here), but this post is specifically about Wolbachia pipientis, and how we should name strain-level variants within a particular host-associated clade (such as wMel, for example).  Currently, there seem to be multiple ways to write these names and to present them in databases or the literature.  For example:

For variants within wMel, we find examples of the following types of strain names:

wMelPop-<insert specific substrain here after the dash>
The stuff that comes after the strain name (wMelPop) sometimes refers to the host insect from which the strain was isolated, or where it currently resides (such as in wMelPop-PYGP) or sometimes refers to how the strain was altered, through sequential passage (such as in wMelPop-CLA).  This variability in the naming schemas used in Wolbachia could lead to confusion down the road.

We are currently adding strain level information to some of the stocks that are now housed in the Bloomington Drosophila Stock Center.  Check out the page here for more. These strains were previously described and published and therefore, the names that appear should match those in the publication (at least for now).  However, when we identify new Wolbachia variants within Drosophila melanogaster, how should they be named?  Should we use the "dash" after the wMel<specificstrain> name? Do you work in Wolbachia? What makes sense to you? Please comment below.

some comments from the twittersphere:
Phenotypes can be dependend on host back ground and environment (e.g. temperature) - not good for naming.

when you refer to strain. Are you using Bloomington's (or source stock centre) stock number, or genetic mutation?


  1. I'll summarize a few thoughts here that I mentioned on Twitter. Generally, I think that sub-names should be reserved for only those variants that have been identified as being extreme in terms of genotype or phenotype relative to other variants within strains, and/or have become model variants for additional analysis.

    The definition of "a new variant" is unclear to me, which makes it difficult to discuss how to give them sub-names. I have hundreds of recently sampled wMel variants from the US and Africa. They vary at sites along the genome and we could find phenotypic variation if we look hard enough. Are each of these a new variant requiring a sub-name simply because they were sampled? Including the origin and interesting genotypic and phenotypic information in the stock center listing would be great, but giving sub-names to all variants will become onerous and those working with many lines at the population level will not use them. In the literature I much prefer an explicit description of the variant being used/evaluated over an uninformative identifier.

    I would like to make one more tangential point. The MLST can be very uninformative for classifying at the strain level. I have examples of Wolbachia infecting sister species that show no variation at the MLST—this holds when looking across many Wolbachia within populations of each host species. However, outside of the typing loci there is substantial variation, and this variation is partitioned perfectly within hosts. Are these different strains? Are variants within hosts that differ from others in terms of SNPs or structural variation deserving of sub-names? What about those that differ very little across the genome, but differ in some interesting phenotypic trait value?

    Sorry to be a pain, and much of this may be irrelevant if only interesting/model variants are maintained by stock centers. I'm mostly concerned about having to apply sub-names to many variants and having to use such names in the literature—this is obviously a specific concern for those sampling and working with many new variants.

  2. I'll echo some of the previously raised points. I think the context dependence of the phenotypes is an important consideration, and probably shouldn't be used in the name. Something like wMelPop would be an exception here, that is quite extreme, and well described/understood. And, because it looks like these are variants that have been transferred into the DrosDel stocks, I worry that even in the lab there could be selection upon the new Wolbachia-host association, and you're left with something different than when you started.

    I think drawing the line between strain/substrain is difficult. I know that even within my isofemale, asexual line of Trichogramma infected with wTpre, there is variation in the Wolbachia population. I would air on the side of being less specific with strain designation, but providing as many details as possible in the stock listing.

    I am happy to see that Wolbachia are being maintained and characterized in stocks!