On metagenomics

So, I’ve returned from the ERTC (see previous post), and it was a gas. I think one of the more interesting things about the conference was that it was primarily Chinese and European scientists, and thus gave me some insights into how Americans tend to skew discussions. In any event, one of the coolest talks I saw was by Zhao Liping of Shanghai Jiaotong University (aside: this is where my most excellent graduate student, Xi Chen, came from; whatever you guys are doing there, you’re doing a great job!). Dr. Zhao talked about metagenomics of the gut as a function of diet. He makes a very good case that you are what you eat. Not so much in terms of the food, but in terms of what bacteria the food cultivate over time. To a first approximation, if you eat crap, you cultivate crap bacteria (pathogens), whereas if you eat well, you cultivate more gut-friendly bacteria. This is true even when you account for the genetics of the organism and its ability to utilize different foodstuffs. Some of this has already been reported in Li et al. (2008), PNAS 105:2117.

From a biodefense or intelligence perspective, though, what’s neat is that the bacteria associated with a given individual can be seen as a fingerprint of that individual. As the PNAS paper states: “At the species level, we found structural differences in the Chinese family gut microbiomes and those reported for American volunteers, which is consistent with population microbial cometabolic differences reported in epidemiological studies.” And as the cited epidemiological studies (Dumas et al. (2006), Anal Chem 78:2199) state, such “cross-population differences in urinary metabolites could be related to genetic, dietary, and gut microbial factors.”

Now, it’s perhaps not all that interesting that Chinese and Europeans, say, are different. Well, duh. What is interesting is if you consider the succession of microbial communities in the gut. This is older work, from Milkman amongst others, on “periodic selection.” Our guts are constantly roiling with different bacteria and even different allelic variants of the same bacteria. These bacteria rise to prominence and are extinguished depending on diet, even depending on a single meal. And they are passed between individuals. A family is likely to have more genetically similar gut flora, down to and including the pets of that family.

So, what if one was able to track the habits and associations of individuals based on their gut flora? Would that be useful (see also “On GATTACA and taggants”)? I would previously not have thought so. It would seem that the signal would be buried in so much noise that identification with any surety would have been impossible. But of course that was before we had so very, very much signal. There was another neat talk at the conference, by Yang Huanming, the “Chinese Craig Venter” (no offense to either party) who heads up the super Beijing Genomics Institute (BGI). The sheer scale of the BGI is breathtaking, with over 800 bioinformaticians dedicated to analyzing the data from their hundreds of NextGen sequencing machines (I think it’s not unreasonable to estimate this one Institute as being on the order of 100x University operations). In any event, a landmark event for the BGI was the publication of “a human gut microbial gene catalogue” (Qin et al. (2010), Nature, 464:59), which sampled 124 European individuals. One conclusion is that the genetic diversity of our gut flora dwarfs our own genetic diversity by a factor of ca. 100-fold. Or in other words: forget the 23-and-Me allelic analysis. If you really want to know about someone, ask about the bacteria that live inside them.

Is it possible that you’ll be able to peer into the detailed history, habits, and associations of an individual based on a fine analysis of their metagenome? Who knows? It’s not an unreasonable hypothesis. But I do know that given the scale of sequencing and analysis being carried out by the BGI, that they’ll know the answer long before we in the US do. From a purely Strangelovian perspective, there is most definitely a metagenome gap.


- originally posted on Sunday, October 24th, 2010