One of the things I didn’t get a chance to talk about at the Boston Skeptics meeting is how we use evolutionary biology to understand the human microbiome–those microorganisms that live on and in us.
Here’s an example from a paper about Crohn’s disease (italics mine):
It is hypothesized that IBD [inflammatory bowel disease] results from an aberrant immune response against intestinal bacteria that results in inflammatory damage to intestinal tissues. Many clinical and experimental observations strongly implicate intestinal bacteria in the pathogenesis of IBD. However… neither the mechanisms by which bacteria affect the development of IBD nor the disease-specific changes in the intestinal flora have been determined.
So the first step is to compare what microbes healthy and sick people have. In a previous post, I discussed how we can use genetic barcoding to do this. One problem is that many organisms are difficult to culture. Instead, we can PCR amplify a gene (’16S rRNA’) found in all bacteria and archea, and then sequence the copies of the gene. Each species has its own unique version (or versions) of the 16S rRNA gene since this gene evolves very slowly. We can then count the number of times a species shows up, and compare the microbial fauna of healthy and sick people.
So where does the evolutionary biology enter? After all, this is a straightforward ecological problem–that is, we’re just dealing with the abundance and distribution of bacteria (i.e., why are X found here and Y found there?). One reason to use evolutionary biology is to classify strains. Typically, we don’t actually count species because the method we use to sequence the 16S gene only covers part of the gene, so we don’t necessarily have the resolution to classify sequences down to the species level*. Instead, we place species in operational taxonomic units (‘OTUs’) that contain closely related species**.
If you have a sequence that is identical to a known, typed species, that’s obviously straightforward–it’s just a matter of matching the sequence in question to the identical one in a database. However, most sequences lack exact matches because we only have characterized a ridiculously small fraction of the total bacterial diversity. So what do we do then?
Pray for divine revelation of the species taxonomy***.
Well, here’s what the state of art bacterial taxonomic classification system does. It takes all of the sequences in the OTUs that are similar to the ‘query’ sequence and reconstructs an evolutionary history, or phylogeny, of these sequences. The query sequence is placed in whichever OTU it clusters with. The method used to reconstruct the phylogeny of these sequences is identical to that used by biologists to determine the evolutionary history of all sorts of organisms. In other words, if you want 16S classification to assess the human microbiome, all that scary stuff about primates comes along with it. No special pleading for theological convenience allowed (but you can has cheezburger!)
The second way we use phylogenetics–the estimation of evolutionary history–is to answer the question of whether two (or more) microbial communities are different (e.g., gut microbes in people with Crohn’s disease versus those in healthy people). One method uses phylogeny to infer genetic distances between sequences, and then weights these distances based on how many times that particular sequence occurs. We then calculate the total distance between the two microbial communities, and assess if that distance is significantly larger than if we repeated the procedure after randomly assigning sequences to the two communities (if you’re curious, here’s more information).
You have to be careful when using this technique: at a broad enough taxonomic scale, it can breakdown****. But at the family and genus level, it does appear that phylogeny recapitulates ecology (to steal a phrase). In other words, changes in the number of distantly related microbes should be weighted more heavily because distantly related microbes will have very different ecological functions. (Right now, colleagues and I are testing how well this method works in comparison to others).
The primary reason I’m talking about this is that it’s neato biology. But the other reason is that creationist ignoramuses like Michael Egnor and Vox Day constantly claim that evolutionary biology aren’t relevant to biology. The above examples demonstrate how stupid that claim is. Remember, we’re not doing evolutionary biology, we’re doing medical microbiology. But to do that, we need to use tools developed for use in evolutionary biology.
This is why defending evolutionary biology and exposing intelligent design creationism for the useless fraud that it is matters.
*One can sequence regions that give better resolution; however, many groups of bacteria will be missed. There’s a tradeoff between resolution and breadth.
**The taxonomic classifications used by traditional systematists often contain very different levels of genetic variation within the same level of classification. As Julian Parkhill once put it, “Bacterial systematics isn’t.”
***Wouldn’t that be lame for the Intelligent Designer to reveal herself only to help you classify 16S sequences? At the very least, I would want a Middle East peace plan.
****To explain what I mean, there’s an analogous gene, 18S in all eukaryotes. I could use this to calculate a genetic distance between a shrew and an oak, but one is a tree and the other a fuzzy lil’ mammal.