Wednesday, August 22, 2012

Referring to 16S surveys as "metagenomics" is misleading and annoying #badomics #OmicMimicry

Aargh.  I am a big fan if of ribosomal RNA based surveys of microbial diversity.  Been doing them for 20+ years and still continue to - even though I have moved on to more genomic/metagenomic based studies.  But it drives me crazy to see rRNA surveys now being called "metagenomics".

Here are some examples of cases where rRNA surveys are referred to as metagenomics:
I found these examples in about five minutes of googling.  I am sure there are many many more.  

Why does this drive me crazy?  Because rRNA surveys focus on a single gene.  They are not gnomicy in any way.  Thus it is misleading to refer to rRNA surveys as "metagenomics".  Why do people do this?  I think it is pretty simple.  Genomics and metagenomics are "hot" topics.  To call what one is doing "metagenomics" makes it sound special.  Well, just like adding an "omic" suffix does not make ones work genomics - falsely labeling work as some kind of "omics" also does not make it genomics.

Enough of this.  If you are doing rRNA surveys of microbial communities - great - I love them.  But do not refer to this work as metagenomics.  If you do, you are being misleading, either accidentally or on purpose.    So I think I need a new category of #badomics - "Omic Mimicry" or something like that ...

Note - this post was spurred on by a Twitter conversation - which is captured below (note - I am certain I have complained about this before but cannot find a record of it ...)


  1. The peer review system has been unable to filter because this mistake is widely spread. It could be better if people looked at definitions more carefully. As Johnathan Eisen says, rRNA surveys focus on a single gene, not genomes. It is important not to miss the ALL
    word in the Genome definition. Not one gene but actually as many as can be found in the DNA from samples.

    Definitions in MESH:

    A collective genome representative of the many organisms existing in a community.
    Year introduced: 2010 (2008)

    The genetic complement of an organism, including all of its GENES, as represented in its DNA, or in some cases, its RNA.
    Year introduced: 1992

    A category of nucleic acid sequences that function as units of heredity and which code for the basic instructions for the development, reproduction, and maintenance of organisms.
    Year introduced: 1965

  2. Yes, this is common enough that I have made this error myself when talking with researchers locally, referring to 16s/ITS surveys as metagenomics.

    I would think editors and reviewers who have enough knowledge could catch this, maybe something like this needs a letter published somewhere (open, of course :). In most circles a blog post might do, but maybe not in this case unfortunately.

  3. Although I prefer metagenomics to not include 16S only surveys, it seems like this linguistic battle has already been lost. Seems everyone refers to 16S as being under the metagenomics umbrella. The crappy thing is when you want to talk about 16S vs real metagenomics; what do you call it? 16S vs WGS?

    1. Everyone is entitled to use language in their own way. However, this battle is not lost since it is so clear that using "metagenomics" to refer to 16S studies is misleading, either on purpose or on accident. It is NOT genomicy in any way. And I will fight the misleading part of the terminology as long as I can ..

  4. I have seen this in textbooks as well. I think the word is evolving in meaning a bit. I am less adamant about it, but it isn't my main focal area of biology. I think the broader term usage is something like DNA sequencing based analysis of ecosystems or communities (as opposed to culture/enrichment studies). 16S based (or other gene marker) survey is a perfectly adequate label, but as you say, not as hot right now. I will make sure that my 16s survey paper is not called metagenomics :).

  5. As far as I'm aware, the Genomic Standards Consortium is trying to advocate for the term "marker gene studies" to refer to any PCR-based amplification of an informative genetic locus (e.g. . I've been adhering to this in our grant applications and newer publications at least...although I must note that us eukaryote people first tried to coin the term "metagenetics" (my paper being an offender ) before it was pointed out that Jo Handelsman coined this term in a completely different context in 2008 ( ). Frustrating to see that the the microbiology community is still lumping rRNA under the metagenomics umbrella, even in 2012 papers!

  6. This comment has been removed by the author.

  7. In my lab, we focus on unicellular marine eukaryotes diversity, and we use the term metabarcoding for our rRNA studies. I don't know who coined it, but I heard it for the first time during a meeting with the ecologist Pierre Taberlet, and he uses it in his papers (see for instance I also heard the term metagenics once.

  8. I think the confusion arose when people started classifying their studies on where the DNA came from, rather than what they were looking for. Thus, anything from a community became 'meta' to differentiate it from axenic culture studies. I don't think it's anyone being deliberately misleading. Beyond the early studies (and now in some newer ones which enable improved quantitative analysis) I would say the field of metagenomics (as you define it) has been a mixed bag of success. Classifying a study as 'metagenomics' is as likely to be met with band-wagon glee as 'oh look, more basepairs' derision.

    1. I think you underestimate the power of buzzwords in appealing to funding agencies and journal editors ...

  9. Seems to me that the original definition by Jo Handelsman was based on cloning of BACs so the term never was a true genomic approach. I phylogenetic reconstruction techniques, such as PICRUSt, advance are 16S studies, marker gene surveys or metagenomics? I agree with Benbo that the 'meta' meaning community is the more revealing part of the word than 'genomics'.

    1. Thank you for this comment. I was thinking exactly the same. I think finally that Metagenomics is not such a bad word for 16S. Certainly the final analysis can take into consideration whole genomes of bacteria from a complex biological sample, which is the definition of metagenomics. However, certainly it is only possible because other people did "real" metagenomics studies before hand, and you can argue that it is only the case for microbes with known whole genomes.


Most recent post

My Ode to Yolo Bypass

Gave my 1st ever talk about Yolo Bypass and my 1st ever talk about Nature Photography. Here it is ...