The Case of the Curious Microbiome


Meet Siavosh. He’s the data scientist behind last week’s popular post on male vs. female microbiomes.

Siavosh grew up overseas, got his PhD in Physics from Stanford, and is now an amazing part of the uBiome team. His desk is right next to an impressively high window that streams in sunlight and looks out over the city of San Francisco.

During a conversation near this window, we decided to start digging into the microbiomes of the fine folks who work here at uBiome, to see what we find and share it with you. Siavosh bravely offered to go first.

Here’s what we discovered in looking at his gut microbiome data:

Screen Shot 2015-02-23 at 6.32.33 PM


The charts above show that Siavosh’s level of Bacteriodetes is elevated far above what we would expect in the anonymized, aggregate uBiome dataset.

What about the error bars, you might be wondering? Well, the figure below shows that the prevalence of Bacteriodetes is 22% +- 13% in our dataset. In other words, the 68% confidence interval for the mean of Bacteriodetes is 9% to 35 %, and Siavosh has 38%.

Screen Shot 2015-02-23 at 4.32.37 PM

Why is he so different?

Differing levels of bacteria in the gut are known to be affected by a number of factors, including diet, exercise, health conditions like Crohn’s disease and depression, and even where we live and grew up.

One possible explanation for his outlying data is that Firmicutes tend to be more prevalent in obese people, and Bacteriodetes more prevalent in lean people. (He’s on the leaner side.) Growing up in Iran could also have had an effect, as well as the differences between his diet and the traditional Western diet.

We’d like to open this up to our readers too. Can you think of other possible explanations for Siavosh’s gut being enriched with Bacteriodetes? And better yet, would you like to add your microbiome to the mix, and see what mysteries lie in your gut? Find out what’s different about you!



2 thoughts on “The Case of the Curious Microbiome

  1. Pingback: February 27, 2015 | Microbiome Digest – Bik's Picks

  2. Hi,

    I don’t know why Siavosh’s levels are so different. But at a recent meetup about #quantifiedself in Seattle, Richard Sprague told me that my results might be different because my sample size was smaller than normal. Only 27,000 in the raw data, with his average being 60,000.
    Do we know sample size for Siavosh?

Comments are closed.