Best evidence yet that we should do away with the "Candidatus" terminology used for uncultured microbes

Just got this email:

Dear Jonathan , I am relatively new to phylogenetics and have developed a keen interest in bacterial phylogenies. I read one the papers which you have published in Nature on the same topic. I needed some help. Thing is, I have constructed a bacterial phylogeny based on 16s rRNA and have found that Candidatus species differ by a really great extent from one another, and appear to be in completely different clades. I need some way to verify this and was wondering if you could send me the phylogeny you have created in order for me to verify. Thank you

I have been arguing for years that the use of the term “Candidatus” to describe microbes is a bad idea.  This email is exactly the kind of thing I worry about.  Candidatus is a term used “for well characterized but as yet uncultured organisms.” (A nice description of the history of the term is here).  The problem with Candidatus are many – including the confusion such as evidenced by this email – in that people who are not used to dealing with the term have no clue what it means.  But in addition, the constraint that one is not “allowed” to name things unless they have been cultured is silly, outdated, and damaging to the field.  But I can write about that some other time.  For now, this email will suffice …

Some examples of Candidatus names from

Author: Jonathan Eisen

I am an evolutionary biologist and a Professor at U. C. Davis. (see my lab site here). My research focuses on the origin of novelty (how new processes and functions originate). To study this I focus on sequencing and analyzing genomes of organisms, especially microbes and using phylogenomic analysis

4 thoughts on “Best evidence yet that we should do away with the "Candidatus" terminology used for uncultured microbes”

  1. I completely agree. In the Mollicute clade, we have entire tree branches filled with dozens of organisms, many of which have complete sequence published, but are “Candidatus” because no one can culture them in pure culture. The distinction is very outdated, along with the distinction of species by antibody response — distinguishing a few surface lipoproteins instead of looking at complete genome sequence. Not to say that we have a good way of establishing bacterial species boundaries that are all that much better.


  2. I agree to a large extent. However, I think the key sentence is “people who are not used to dealing with the term have no clue what it means.” This is someone with an interest, which is great, but lacking the educational foundation for the task at hand. We see a lot of this in the questions at Biostar: people who appear to be enthusiastic amateurs lacking the required skills/knowledge. It's great that they are interested, but sometimes what you need is to attend the appropriate formal university course.


  3. Wow, you've found another example of one of my pet peeves : overloading names with descriptive information. Names are for naming things, not describing them. If you include metadata in the name, you will increase the risk of a collision, thus undermining the whole point of naming it in the first place.In computer science, this is called a “collision.”

    I made a comic about it a while ago :


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: