Dr. Jonathan Eisen is a Harvard and Stanford educated scientist, speaker, author, and advisor to microbiome startup uBiome. I was pointed in his direction after catching up with uBiome’s CEO about the burgeoning science behind their biome technology. Between TED talks, interviews, and his ongoing research at University of California at Davis, Jonathan is recognized as a leading expert on the human microbiome. In our conversation I wanted to hone in on the so-called “chicken and egg” problem of using microbiome data, and the best steps forward for science to truly understand this field.

The “chicken and egg” problem was brought up in my conversation with uBiome CEO, Jessica Richman. Essentially, uBiome sells a microbiome sampling kit (used by customers to test the microbiome in their gut or mouth, for example), aims to use information from the human microbiome to help diagnose disease. At the present time, scientists don’t have enough conclusive data to make firm health or disease predictions about microbiome results – and so there is less motive for people to get involved with the research or contribute their own (through uBiome’s process, or anyone else’s). On the other hand, in order to get the kind of data we’d need in order to deliver more conclusive results, we’d need more people involved and more data submitted.

So what’s the science world to do? For Dr. Eisen, much of the concern about limited data in the human microbiome is due to studies with limited samples of people – IE: people of one specific gender or race, or only sampling people living in the exact same geographical area. This is not a failure of the scientific process, but of the process of “sample design” employed by the researchers in many past studies. Dr. Eisen believes that even with the same number of subjects in each study, we can be far better off if microbiome research does better work with selecting more rich and varied groups to sample.

Dr. Eisen isn’t sure as to how long it’ll take in order to gain a firmer foothold in diagnostics and prediction. “Our job with the microbiome may be harder or easier than using genomics information, like with 23 and Me, we just don’t know yet.” He mentions that if scientists can correlate one type of bacteria to a specific auto-immune disease, then application will – at least with respect to that bacterium – be relatively simple. Nothing’s this easy, now, however “We know nothing, lets be clear… we’re just getting our toes wet about testing what matters.” Instead of being an explicit catalyst for an explicit process, the microbiome seems to be one factor among many in the cocktail of health (from genetics to environment to history and beyond).

It’s research and it’s time that will give us that answer, but for Eisen, the research is likely to yield the best results if it has few limits on research approach. This might involve “citizen science” startups like uBiome, and it might also involve research from UC Davis, the MAYO clinic and others. Eisen sees them all as potential contributors.

“I like massive experiments in protocol,” says Eisen, ” I don’t see them as competing at all.” As he mentions, both uBiome and traditional research institutions are aiming to better human health and discover the hidden truth of the microbiome’s nature and function. In his view, our best shot in overcoming the chicken and egg problem is as many approaches as we can muster, done with proper sample design. uBiome will likely continue to gain momentum (even with limited diagnostic implications), traditional institutions will jump onboard more ardently, and the field itself will mature enough to find it’s footholds, make a difference in medicine, and develop to the point of greater application and utility.


TechEmergence conducts direct interviews and consensus analysis with leading experts in machine learning and artificial intelligence. Stay ahead with of the industry with charts, figures, and insights from our unparalleled network, including executives from Facebook, Google, Baidu, Yahoo!, MIT, Stanford and beyond: