Human microbiome, Vitamin E for Alzheimer’s, Tweets, $1000 Genome


More on how much of you is microbes

In my final post of last year, I discussed the origin and reliability of my favorite statistic, which is that nine out of ten of the cells in my body, and yours, and everybody’s, are not really yours. They belong to microbes that are essential to your life and your quality of life.

But it’s a new year. And apparently that means new statistics on your cell ratio of resident microbes. This I gleaned from a new report from the American Society for Microbiology, which I learned about from Jonathan Eisen’s Tree of Life blog.  The report is called FAQ: Human Microbiome, and you can get a (free) copy here.

FAQ etc explains the basics, but also goes into detail. There’s a nice clear explanation, for instance, of Koch’s postulates; see p. 12. The report also emphasizes that, while we have grasped the importance of the human microbiome, we know very little about the specifics of what it does in health and disease or about its individual components, normal and abnormal, which can be quite variable. This is a point Eisen emphasizes at Tree of Life, where he often picks apart the many overblown (and false) claims about microbes and health and complains about marketing hype for probiotics.

Anyway, see p. 6 of the FAQ for what I assume is the latest on my beloved statistic. Authors Ann Reid and Shannon Greene think the claim of a 10 to 1 ratio, microbes vs human, is overblown. They say it’s more like 3 microbes for every human cell.

The new number, however, is not based on microbe recounts. The estimate for the human microbiome remains at 100 trillion microbial cells, mostly bacteria and not including viruses. Rather, the revision is due to new estimates of the number of human cells. The human body is now thought to contain not 10 trillion cells but something more like 37 trillion. Hence the revised cell ratio: microbe 3, human 1.  And they say there are an estimated 10-fold fewer fungal cells on/in us than bacteria, so I guess that means we’re home to about 10 trillion fungal cells.

Reid and Greene also say we harbor an estimated 5 times more viruses than bacterial cells, which would bring our viral load up to 500 trillion. Not clear whether that number is supposed to be only the number of free-living (sort of) viruses moving in and out of our cells at any one time or whether it includes all the viruses that have inserted themselves permanently into the human genome over the aeons.

I have written the authors to ask for references that underly the new data and will let you know. I am a bit mournful about this development. The idea of 10 microbes to every human cell is so much cooler than only 3 to 1. I am wondering if I will have to find a new favorite statistic. Or maybe start citing the viral-to-human-cell ratio: 50 to 1.


Tweets don’t measure scientific impact

My pal Beryl Benderly, who writes the Science Careers blog at Science, has posted on a study showing that the number of tweets a scientific paper gets doesn’t seem to be related to its impact as measured by citations.

The authors of the study observe that much-tweeted papers usually either are about some health issue or about funny or surprising findings. That seems accurate to me; tweets are, by and large, not about really serious subjects. There are exceptions, certainly, but the Twitter froth level is very high. Also, it’s my impression that citation records often take time to build, sometimes years. So even if tweets did affect citation numbers, I’m not sure that Twitter has been around long enough for an effect to show up this soon.


Vitamin E slows down Alzheimer’s disease?

The paper in JAMA reports that large doses of vitamin E seem to slow functional decline in Alzheimer’s patients. At Science-Based Medicine, Harriet Hall analyzed the study. She points out that the daily functioning of all the study subjects got worse. What was noteworthy was that the patients taking vitamin E alone got worse more slowly.

This is potentially huge news. Anything that makes it easier for Alzheimer’s patients to carry out normal activities of daily living is potentially huge news, not only for its impact on the lives of afflicted patients and their afflicted families but because something–particularly an inexpensive something like vitamin E–that would slow down the rise in financial costs and custodial care of Alzheimer’s disease even a little bit would save a ton of money.

Hall advised caution given the many unaswered questions remaining. And she was, of course, against impulse purchases of carloads of vitamin E. But on the whole she thought the study was carefully done and encouraging. This is very nearly high praise at Science-Based Medicine, generally a bastion of lambasting.

Gary Schwitzer had a different take at HealthNewsReview, arguing that media reception of the paper was sometimes way out of proportion to the actual findings of the study. The study related to possible improvement of Alzheimer’s symptoms, but he noted that the media often seemed to imply that vitamin E was slowing the disease process itself, which the paper did not claim and for which there is no evidence whatever.

Schwitzer also took media to task for glossing over statistics in the paper suggesting that vitamin E effects on patient behavior, if any, were very small. But expecting journalists, even science journalists who have some understanding of statistics, to explain confidence intervals to their readers is . . . not realistic.

A more important point, which Schwitzer said the stories also ignored, is that the safety of high doses of vitamin E has by no means been established. There are data suggesting the opposite, that they may be dangerous.

The $1000 Genome Machine(s)

The $1000 Genome Machine(s)

At last, here’s the $1000 genome?

The thousand-dollar genome is here at last. At last, that is, it’s here for those who can afford to purchase 10 of Illumina’s sequencers for a million dollars each. Illumina is calling this gang of sequencers that work together the HiSeqX Ten. The company made the announcement the other day, and John Timmer was quick to publish a reality check at Ars Technica.

I don’t know exactly when the $1000 genome became the goal of sequencing technology, but it was well before 2005, when I wrote about how the $1000 genome had just become an official US government project. That was the year NIH committed to spending $32 million to develop technology aimed explicitly at sequencing genomes of individual patients for $1000.

Note that the HiSeqX Ten is not that, not at all. I have no idea whether any of that $32 million, or any taxpayer money at all, went to helping Illumina toward faster sequencing. As Matthew Herper makes clear at Forbes, HiSeqX Ten does not mean consumers can get their genomes sequenced for $1000 any time soon. “There are other costs, like analysis, that matter, and this does not include commercial markup for the providers of genetic tests.”

The government, by the way, is still plugging away at a true $1000 genome.  The latest NIH request for applications closed last fall, a request “to develop novel technologies that will enable extremely low-cost, high quality DNA sequencing.  The goal of this initiative is to reduce the cost of sequencing a mammalian-sized genome to approximately $1000.”  The call urges applicants to explore methods other than those currently in use, especially “high-risk/high-payoff applications.”

Why does this cheer me up, since I suppose this project could well turn out to be a dead-end use of taxpayer money? It would be useful to know how the government-sponsored $1000 Genome Project has fared since 2005. Has it contributed ideas and technology and gotten us closer to the goal in the last almost-decade? I hope so, but whether it has or not, there’s something wonderful about a government that still employs officials who encourage people to go for it. No HiSeqX Ten-type solutions need apply.

I suppose the NIH fantasy is a desktop machine suitable for a doc’s office. Or, who knows, maybe a kitchen table. Which is a very long way from the HiSeqX Ten.


Related Posts Plugin for WordPress, Blogger...
This entry was posted in Uncategorized and tagged , , , , , , , , , , , , , , , , , , , , . Bookmark the permalink.

4 Responses to Human microbiome, Vitamin E for Alzheimer’s, Tweets, $1000 Genome

  1. Jonathan says:

    Tabitha, NIH has played a very significant role in driving down the cost of DNA sequencing since 2005. This page: has a wealth of information on the companies and technologies that have been supported as part of that program, and with the exception of Solexa, all the other sequencing methods benefited from NIH funding during their early development.

  2. Even if the government hadn’t contributed any to Illumina’s sequencing technology since 2007 (which I doubt) Illumina has probably improved its tech far faster than it would have if it were not always dealing with startups nipping at its heels. Offhand, I can’t remember how, but I’ll be some of those have roots in NIH funding.

  3. Tabitha,

    You do a terrific job with On Science Blogs.

    Thanks for mentioning my critique of some of the news coverage of the Vitamin E & Alzheimer’s study.

    You wrote: “expecting journalists, even science journalists who have some understanding of statistics, to explain confidence intervals to their readers is . . . not realistic.”


    Is “margin of error” in political polls beyond the grasp of journalists and the public? It doesn’t seem to be, as this is widely used.

    So why is it unrealistic to expect those covering health care research to examine and report on confidence intervals? This gives people a sense of the “margin of error.” When that interval spans a range of possibilities – from a possible unimpressive or even null finding to “wow” – I argue that it is incumbent on journalists to explain. Or else find a different line of work.

    When health care journalists publish their equivalent of a “Dewey Beats Truman” headline, it results in more than just an embarrassing chuckle. People get hurt.

    This is avoidable ignorance and avoidable harm.

    Thanks for your consideration.

    • Tabitha M. Powledge says:

      I don’t really disagree with you, and it’s certainly true that journalists, especially science journalists, have an obligation to explain things. But as you pointed out in your post, so many of them get the most basic stuff wrong–for example, assuming (or at least implying) that any slowing of Alzheimer’s symptom progression meant that the underlying disease process has been moderated–that it’s close to hopeless to expect them to explain confidence intervals. Or even have a clue about what they are. A measure of my despair. I admire you for continuing to hoist the flag.