All my uBiome results in a single table

Here’s another wonderful post by our great friend Richard Sprague. Thanks Richard!

I often read news about a fresh scientific discovery involving the microbiome and immediately wonder if the discovery applies to me.

For example, I recently saw a study from Oregon State University that seemed to find a link between high sugar diets and “cognitive flexibility”, i.e. your ability to adapt and adjust to changing circumstances. The study’s author, Kathy Magnusson, a professor in the OSU College of Veterinary Medicine, found that mice who eat lots of sugar have elevated levels of Clostridiales bacteria, and that this seemed to relate to a slower ability to solve a maze.

Hmmm, I thought — how much Clostridiales do I have?

If you have just one uBiome result, that’s easy: log into http://app.ubiome.com and search for it in the section “All My Bacteria”. (As far as I know there’s no “search” button yet on the uBiome dashboard). But in my experience a single result doesn’t tell you much. You really need at least two and hopefully several uBiome results to see what might be actionable.

In my case, I want to know how my Clostridiales may have changed over time.

I programmed a short Python script to generate a single Excel table with every bacteria I’ve ever found, and then a series of columns with the amount found in each sample. Something like this:

ubiomeExcelMultiTable

The data makes it easy to generate a chart showing how my Clostridiales changes over time:

ubiomeExcelClostridiales

Hmmm, in my case it looks like something happened since last fall to increase my Clostridiales levels. Maybe it was the potato starch I tried in order to hack my sleep? Was it my trip to Central America in February? And of course the biggest question: has the increase affected my cognitive flexibility? I’m not really sure. Whatever happened, the level of Clostridiales seems to have stabilized in the past couple of months.

uBiome has identified more than 900 unique taxa (groups of organisms) in the half-dozen samples I’ve submitted over the past year, and after running this script I have them all laid out on a single page.

Armed with this one spreadsheet I can search anytime for a new microbe and quickly see if I have it now, or if it’s ever been detected in a previous test. Reading news about the microbiome takes on a whole new personal meaning when I can see if the discovery relates to me.

(If you know a little Python, you can make the same spreadsheet with your samples using the ubiome.py Python module on the ubiome-opensource GitHub repository; the script that generated my spreadsheet is there too as an example. Happy exploring!)

New open source uBiome github repository for data analysis tools

uBiome was founded to help all citizens contribute to science. Many uBiome users have been asking if we can make it easier for them to analyze their own data, and now we’re pleased to announce a new Github site to let you do exactly that.

The site has a repository called microbiome-tools (for sharing tools, templates, scripts, or other software or utilities you find useful for analyzing your uBiome data).

To contribute, all you need is a (free) Github account from which you can send pull requests. If you’re brand new to Git or Github and you want to learn more, check out this beginners guide: http://readwrite.com/2013/09/30/understanding-github-a-journey-for-beginners-part-1

Please note that everything in these repositories is being made available under the Creative Commons International license, and by uploading your tools you are agreeing that you own the information and that you agree to share them with everyone.

Share away! Let’s all learn together.

That link again is: https://github.com/ubiome-opensource

Your friends, Alexandra Carmichael (@accarmichael) and Richard Sprague (@sprague)

One Week Change In My Microbiome

The inspiring Richard Sprague joins us again, with a curious finding!

Having done multiple uBiome tests over the past year, I already have a sense of what my “normal” gut biome looks like. Although there is a fair bit of variation (especially after my various experiments, like sleep-hacking or jungle exploration), my results generally fit the range of “healthy omnivore”. But most of my tests are taken several weeks or even months apart, where it can be hard to understand precisely what’s driving the overall differences. How much variation would I see between samples taken just a week apart?

To find out, I sent two gut samples to uBiome, one on April 21 and the other exactly one week later on April 28th. I received the results last week, less than a month after submitting them. (uBiome turnaround times are getting much faster!) The overall picture looks like this:

picture

 

That’s more variation than I expected for such a short time period. What’s driving the changes? Fortunately, the new uBiome web site makes it much easier to compare one sample with another. In my case, it shows the following changes over the week:

picture

These charts show changes in the absolute population of various microbes, which uBiome calculates by dividing the newer sample “count_norm” field by the same field in the earlier sample. Since this tends to give extra weight to the smallest populations of microbes, I prefer to calculate by proportion; in other words, which microbes changed most in overall percentage against my entire microbiome. After downloading the raw data and running it against my open source tools, here’s what I found (at the genus level):

tax_name count_change
Roseburia 41427
Faecalibacterium 33862
Bacteroides 24346
Lachnospira 13601
Lactobacillus 9874

These are all generally considered “good” bacteria, so I’m glad to see the increases. But why the change at all, especially over such a short time period?

Fortunately, I have some additional data.  I regularly track the food I eat using the MyFitnessPal app on my phone. Using a handy data exporter I summarized the macronutrient information like this:

Calories Carbs Fat Protein Cholesterol Sodium Sugars Fiber
Average (month) 1841.7 192.2 102.7 94.9 268.0 2298.5 64.1 15.2
Average (Week) 2242.6 241.9 124.7 108.7 262.3 2814.3 78.3 16.9
Difference from Ave 400.9 49.7 22.0 13.9 -5.7 515.7 14.2 1.6
% Diff from Ave 122% 126% 121% 115% 98% 122% 122% 111%

Looks like I ate more calories than normal that week (that was when the new Chik Fil A opened near us), which explains the higher-than-average numbers for carbs, fat, sugars and the rest.  But there is one unusual result: note that despite my extra appetite (and that Spicy Chicken Deluxe), I ate less dietary cholesterol. Could that explain the increase in those particular microbes?

Of course, this is all extremely speculative, but a quick internet search reveals an intriguing study involving patients with cholesterol gallstones whose microbiomes lost exactly the three microbes that I gained. Is there a link?

Who knows? It was only a week, and it was a pretty small difference. But that’s the fun of experiments like this: “normal” people can make discoveries.  And if I did find evidence of a link between cholesterol and the microbiome, this could have huge implications for the treatment and prevention of heart disease.

Alexandra, you may want to give a heads-up to the Nobel Prize Nominating Committee. 🙂

My Microbiome In The Jungle

by Richard Sprague

How does travel affect your microbiome? In a famous experiment published in 2014, Duke University scientist Lawrence David tracked the daily microbiomes of two people for an entire year and found significant differences when one of the people travelled outside the U.S. Would the same thing happen to me?

According to my latest uBiome results, the answer is yes. My wife and I recently travelled from our home in Seattle to Central America to celebrate her birthday. We spent most of our time in a rural, jungle part of Belize, about a half hour’s drive from Benque, near the border with Guatemala. Besides viewing the fantastic, well-preserved Mayan ruins in the area, we also did horseback riding, cave exploration, and of course plenty of eating.

Here’s a selfie I took in front of the incredibly well-preserved thousand-year-old pyramid at Tikal:

download

 

What did the jungle do to my gut? Here’s the uBiome summary of the test I submitted immediately after returning home:

download (1)

 

The first thing you might notice is the abundance of the phylum Firmicutes compared to “average”, a difference which is often associated with obesity. If this had been my only uBiome test, you might think I should have passed on the birthday cake (especially since 23andme also thinks I’m at risk for obesity). Interestingly, I did gain a couple of pounds during the trip, maybe from all that tasty coconut rice and beans.

But a single result can’t tell you much. Since I’ve been testing myself regularly with uBiome for the past year, I have a good idea of what my “normal” gut biome looks like. Here’s the overall picture through time, including results from my sleep hacking experiments between October and January (see those big chunks of Actinobacteria, colored brown in the graph?):

download (2)

 

From my previous uBiome results, you can see that my Firmicutes levels bounce up and down, but not dramatically, and the jungle doesn’t appear to have changed the major numbers all that much. Incidentally, I soon lost all the extra pounds after returning home to my normal diet.

Seen through time, it’s clear that the most unusual change during my trip was the increase in Proteobacteria, from less than 1% to almost 8%. Even more interesting was that the Lawrence David study found the same thing: a westerner traveling to a developing country sees a sharp rise in Proteobacteria!

Looking more closely, nearly all the new Proteobacteria are members of the class Gammaproteobacteria. Using the Tree Explorer on the uBiome sample dashboard page, I dug further and further until I saw this:

download (3)

 

Most of the new organisms I picked up come from the family Enterobacteriales, of which the majority – about 3.5% of my entire sample – is Cronobacter, a nasty pathogen named after the Greek mythological titan who swallowed his children!

Ouch! Fortunately I never got sick. Why not? The science is just too new to say for sure, but here’s my theory: there is no such thing as “good” or “bad” bacteria. Everything depends on the ratios, on balance among lots of competing germs. In my case – and this is my pure, amateur, unscientific speculation – the Cronobacteria increase might actually have helped my health, by out-eating something else that may have been even worse.

Here are some reasons I suspect this is true:

  • Diversity: Oddly, my gut biome diversity went down slightly. Before the trip, uBiome found 19 unique phyla. Afterwards, there were only 15. You wouldn’t normally expect diversity to drop after exposure to novel microbes from the jungle. But I think there’s an easy explanation:
  • Increase in unidentified organisms: uBiome was able to identify only about 91% of what it found at the phyla level. In my previous tests, they found closer to 95+%. Maybe my apparent drop in diversity was simply a drop in identifiable bacteria. Maybe some of those unknown organisms stimulated the bloom in Cronobacter.
  • Clostridium plunge: I saw huge drops, from 0.66% to 0.18%, in the notorius Clostridum genus, which includes many nasty species (e.g. the infamous antibiotic-resistant C. Difficile). Other pathogens dropped too. Did the change in location precipitate a fight between competing bacterial armies?
  • Parasites: the uBiome test only measures bacteria, so I don’t know the status of other microbes I encountered. If Cronobacteria are normally pathogenic, is it possible that all their toxicity was aimed at a takeover by some other organism?

My personal blog has more details, including how to use my free public uBiome analysis tools to compare these samples further.

The Lawrence David study also showed that the microbiome bounced back to normal again pretty quickly after the person returned back home. Will that happen to me? I submitted another uBiome sample this week and I’ll let you know when I get the results.

 

My uBiome Sleep Hacking Update

by Richard Sprague @sprague

In a post last month about how I use uBiome to hack my sleep, I mentioned the noticeable effect potato starch has on Bifidobacterium, a common gut bacteria with a well-documented role in the production of many important body hormones, including some that affect sleep.

Although I seemed to sleep better, I was a bit worried about continuing my experiment because my uBiome results clearly indicated that, perhaps as a side effect of the increased bifido, some other microbes were disappearing. My solution? Test myself again, then go off potato starch until I see the results. I said I’d report back, so here’s my update.

By the way, if you (or a friend) have tried uBiome in the past, you’ll be pleased that test results now come much more quickly. What took several months earlier last year now took only a few weeks. I think this is because uBiome brought some of their own shiny new equipment in-house. Because of the faster turnaround, it’s much more feasible to do tests like this regularly: test yourself, try something for a few weeks, then test again.

Here’s the overall summary of the four tests I’ve done with uBiome so far:

unnamed

Looks reasonably okay. Nobody really knows what a “good” biome is, but given my otherwise overall good health, it’s nice to see rough consistency across the samples. The most obvious change for both October and January is that large dark red splotch of actinobacteria, of which those sleep-helping bifidobacteriumare a subset. Nothing unexpected.

But what about overall diversity? Most scientists agree that, just as a wider variety of plants and animals is good for the ecology of a forest in the external world, our internal “bacterial forest” is healthier the more unique organisms we host.

How many unique types of organisms did uBiome identify in my samples? Here’s a summary:

Taxa Rank May June Oct Jan
species 165 45 106 142
genus 114 39 69 106
family 74 28 42 83

To understand this table, it’s helpful to know a little about how biologists classify lifeforms, because it’s not enough to simply look at the species count, the way you might begin if you wanted to measure diversity at the zoo. Instead, scientists divide living things into layered groupings: kingdom, phylum, class, order, family, genus, species. Each layer is a subset of the ones above it. So for example, the common housecat is the species Catus, a member of the genus Felis that includes wildcats, which in turn is a member of the family Felidae that includes tigers and jaguars. This continues on up through order Carnivora, class Mammilia, phylum Chordata and kingdom Animalia. All bacterium are members of the kingdom Bacteria, which includes a gazillion other layers of phyla, classes, etc. all the way down to species.

unnamed

The low-cost, revolutionary technology behind uBiome, called 16S rRNA, is pretty accurate at the higher classifications of life, but less so when you get down to the genus or species level. In my case, uBiome assigned species names to only half the organisms in my sample, whereas at the family level they identified almost 95%.

So with that in mind, I’m happy to find that, while there was some disappearance at the less-comprehensive family level and below, I lost only one biological order – and gained six!

Using the free uBiome tools I host at Github, I found the following new orders of life in January:

##   count         tax_name
## 1   564 Flavobacteriales
## 2   137    Legionellales
## 3   120  Xanthomonadales
## 4   120  Fibrobacterales
## 5   103  Haloplasmatales
## 6    86      Rhizobiales

and only lost one:

##   count           tax_name
## 1    25 Desulfuromonadales

That “count”, which corresponds roughly with ten-thousandths of a percentage, is pretty insignificant, so I’d say based on this test I don’t have much to worry about. Other than better sleep, I haven’t noticed anything different about my health: my weight, mood, bowel movements, etc. are all just as good as ever.

One very important note: this science is extremely new and you should never use your uBiome results for more than scientific curiosity. My regular physicals show that I’m in perfect health, so to me this is nothing more than a fun sleep hacking experiment, but if you have serious sleep issues – and especially if you are overweight, or have insulin trouble – this experiment will do you no good and may even be harmful. See a doctor!

Finally, shortly after this test I went on vacation to Central America for two weeks of jungle hiking and eating. Immediately after my return to the US, I sent another sample to uBiome to see how my travel affected my gut. I can’t wait to see the results, and I’ll update you in a future post.

Looking into my mouth microbiome

I’m so excited to give you another of Richard Sprague’s insightful and pioneering posts about his personal microbiome investigation. Thank you for sharing your findings with us, Richard!

The gut biome is interesting enough, but bacteria colonize just about every part of the body, so recently I’ve been studying my uBiome mouth test results. The simple GitHub uBiome utilities I use for analyzing my gut will work for that too, so here’s a short example of how I did it:

First I loaded my uBiome data into two variables, one for each sample: June 2014 (junMouth) and the other from October 2014 (OctMouth), after a visit to my dentist.

Let’s see which species are new in the later (October) sample:

octToJunChange <- span=""> uBiome_sample_unique(OctMouth,junMouth)
##   count                        missing.tax_name
## 1  3640                  bacterium NLAE-zl-P562
## 2  2725                 Streptococcus sanguinis
## 3  2075               Capnocytophaga gingivalis
## 4  1969 Peptostreptococcus sp. oral clone FG014
## 5  1618                 Granulicatella adiacens

One of those species, Streptococcus sanguinis looks interesting. Wikipedia says this:

S. sanguinis is a normal inhabitant of the healthy human mouth where it is particularly found in dental plaque, where it modifies the environment to make it less hospitable for other strains of Streptococcus that cause cavities, such as Streptococcus mutans.

No cavities? Nice! More good news: this quick check confirms that I don’t have any S. mutans:

OctMouth[grepl("Streptococcus",OctMouth$tax_name),]$tax_name
## [1] Streptococcus                      Streptococcus pseudopneumoniae    
## [3] Streptococcus sanguinis            Streptococcus constellatus        
## [5] Streptococcus anginosus group      Streptococcus sp. oral clone GM006
## [7] Streptococcus thermophilus         Streptococcus oralis              
## [9] Streptococcus gordonii            
## 250 Levels: [Eubacterium] sulci ... Veillonellaceae

Then I look at the species that disappeared (went extinct) between the two samples:

junToOctChange <- span=""> uBiome_sample_unique(junMouth,OctMouth)
##   count                        missing.tax_name
## 1  6034                Capnocytophaga granulosa
## 2  4153 Peptostreptococcus sp. oral clone FL008
## 3  3375         Prevotella sp. oral clone ID019
## 4  2691                   Streptococcus rubneri
## 5  1571                       Prevotella buccae

Anything in the genus Capnocytophaga is an opportunistic pathogen, so I say good riddance. Usually they’re fine, but if your immune system dips they can turn bad.

Streptococcus rubneri was discovered a couple years ago, so little is known about it.

Prevotella buccae is more interesting. It seems to be implicated in periodonal disease (yuk!) but that genus is involved too in breaking down proteins and carbohydrates.

Hard to say what’s really going on. Meanwhile, here are the biggest changes (increase) since the first sample:

junToOctCompare <- span=""> uBiome_compare_samples(junMouth,OctMouth)
##                                  tax_name count_change
## 64         Streptococcus pseudopneumoniae        62007
## 68         Veillonella sp. oral taxon 780         8065
## 41                       Neisseria oralis         4693
## 2  Abiotrophia sp. oral clone P4PA_155 P1         2308
## 28                 Granulicatella elegans         1987

Whoah! That first one, Streptococcus pseudopneumoniae, looks nasty! Wikipedia says it may cause pneumonia, though a recent medical journal says more hopefully that it “treads the fine line between commensal and pathogen”

…which is a scientific gobbledygook way of saying nobody has a clue. All the more reason to keep testing, submitting, and getting more data. I just sent two more kits to uBiome, and will let you know more as soon as I get back the results.

(This post was originally published here.)

Hacking my sleep with uBiome

As a very special treat today, Richard Sprague has a guest blog post for you about hacking his sleep, with a surprising discovery. You can also find Richard on Twitter and at his brilliant blog. Let’s give Richard a warm welcome!

Here he is…


About a year ago, I started to notice my sleep becoming less regular. Nothing serious — thankfully, I’ve never had sleep problems — but I was waking up too early, and I didn’t seem to be quite as refreshed. Maybe it’s just a sign of age, I thought, until I read in Martin Blaser’s excellent book Missing Microbes (p.304) that most (80%) of the sleep- and mood-regulating neurotransmitter serotonin is made in the gut. Could my gut microbes affect my sleep?

A few internet searches later led me to evidence that B. infantis modulates tryptophan, the stuff in turkey that urban legends have long blamed for that sleepy feeling you get after Thanksgiving dinner. Seemed like a good target to check, and because I’m a long-time uBiome fan —- I supported them on Indiegogo almost two years ago – my first step was to look at my gut biome results to see my levels of bifidobacterium.

As I suspected, I had no B. infantis, and in fact my overall bifido numbers were pretty low as you can see from this item on my uBiome results sample explorer page:

sprague-1

You may already know about pre-biotics, foods that feed bacteria, as opposed to pro-biotics, which are simply pills or foods that already contain a bunch of (presumably) beneficial microbes. Lately a number of people have noticed that a particular kind of starch, so-called resistant starch is a prebiotic that acts like a yummy smorgasboard for bifido and other bacteria. I followed a protocol that uses plain ole Bob’s Red Mill Potato Starch (easy to find any nice grocery store): just a couple of tablespoons one hour before bedtime to give the starch time to make it to the upper colon where the bifido live.

Whoah! You wouldn’t believe how wonderfully I slept that night. Over 8 hours of rock-solid, uninterrupted sleep, and more vivid dreams than I’ve had in years. It was amazing!

After a few days of this, I submitted another sample to uBiome for testing:

sprague-2

Holy bloom, Batman! That’s just plain stratospheric: up from 0.847% before the potato starch, to over 5.87% afterwards. That 8x improvement it seems clearly explains my much-improved sleep.

But nothing’s free, right? If the bifido are increasing that much, then something else is decreasing. My sleep might be improving, but am I sacrificing some other aspect of my long-term health?

To find out, I wrote a couple of utilities to examine my uBiome results in more detail. One of the things I like most about uBiome is that they give users access to the raw data, which you can convert to work in Excel or another programming environment. If you’re interested in more details, you can read how I did this later, but here are some of the major species that went extinct after I took potato starch:

##   missing.count_norm                   missing.tax_name
## 1               8295        Bifidobacterium tsurumiense
## 2               4650          Subdoligranulum variabile
## 3               2074     Dialister sp. oral clone BS095
## 4                780 Desulfovibrio sp. oral clone BB161
## 5                475        Adlercreutzia equolifaciens
## 6                459               Ruminococcus sp. ID1

What do these microbes do? After a few more internet searches, my answer is: nobody knows! We have some guesses, but so far scientists just haven’t had enough samples from real people to understand much. That’s another reason I hope you’ll submit your samples to uBiome, to increase the number of data points that scientists can work with in hopes of understanding this better.

Meanwhile, my sleep continues to be much better than before I began this experiment, so I’ve stopped taking potato starch and have been substituing a few other foods that seem to affect the microbiome. I’ve already submitted my before-after samples to uBiome and will let you know what I find in a future post.