Epigenetics Has Become Dangerously Fashionable

The hype around epigenetics, or anything-genetics in general, has been going for a while [1], and it has some very annoying consequences. A lot of the friends I talk to about the subject barely have an inkling of how even basic genetics work, yet tell me about these "crazy things we can do now", except we can't really, nor do we understand as much as a lot of people like to believe. The media and their pop-science are obviously largely responsible for this, but you'll find a lot of researchers using this to promote and fund their research, and really, who can blame them (particularly in academia/research where people will get any funding they can). I do blame some of the social scientists however, who are often quick to draw correlations where there aren't any, if only to get attention.

I work with some absolutely brilliant people, and a lot of them wouldn't touch epigenetics with a ten foot pole. The hype around it means expectations are ridiculous, we're barely scratching the surface of the subject yet many people are already expecting things like authoritative dietary or educational advice. The same is unfortunately true of cancer research and cancer care in general ("that's not how cancers work" is my new go-to phrase...).

In the post-genomics era [2], we're starting to see things like phenomics [3] that could indeed further research into epigenetics, although this is still very new and the real-world applications are still to bear significant results. The most advanced use of phenomics at this point (that I'm aware of) is to monitor and control plant features using mostly image recognition [4]. One particularly interesting application is to line up hundreds of plants (often clones) and apply different conditions to different batches (including radiation to induce random mutations), then have various types of cameras go over the plants and perform feature recognition to try and infer correlations between genotype, phenotype and environment (like these guys do [5]). It's promising, but let's not forget these are plants, not animals, and certainly not humans. It's also rather obvious that this sort of brute-force approach (no matter how amazing I think it is) isn't exactly applicable to anything other than plants at this point ; having rows of bred/cloned humans in extremely controlled conditions could prove a bit tricky...

[1] https://aeon.co/essays/the-dream-of-designing-humans-has-a-l... [2] I find the term a bit pompous, particularly considering we're still very much reliant on genomics and still improving it quite substantially [3] https://en.wikipedia.org/wiki/Phenomics, although it is a bit short on content [4] http://www.sciencedirect.com/science/article/pii/S1369526615... [5] http://www.psi.cz/products/plantscreen-phenotyping/plantscre...