Career transition - or destruction? - into complementary and alternative medicine

Funding at NIH is very tight this year, with a new budget that doesn't even keep up with inflation. Luckily, NCCAM (the National Center for Complementary and Alternative Medicine) is still giving out grants aplenty. I've highlighted a few of NCCAM's awards in previous posts - and will do so again - but today I thought I'd look for myself to see what kinds of funding they're offering to eager young scientists.

Well, here's one: a "Complementary and Alternative Medicine Career Transition Award" (PAR-05-129), a grant to encourage scientists to change their focus and start a career in CAM. Think of how rewarding it will be! Scientists who win these prestigious grants will have four years of support as they learn how to confuse patients and obscure the truth, so that they can be paid (sometimes well paid!) for offering sham treatments. And NIH will pay your entire salary all the while - wonderful!

I should explain that "career awards" from NIH are actually a very important mechanism for funding new scientists and for funding scientists who are making a transition to a new field. These awards are used by all the major NIH institutes, so NCCAM is just using one of the standard types of awards, but with a CAM twist, of course. You see, it's hard to find a good position doing CAM research - as NCCAM explains it:
For postdoctoral trainees doing research in CAM, this transition has an added challenge because there are still relatively few CAM departments in research institutions where they might seek positions. Therefore, an important goal of this initiative is to improve the chances of outstanding postdoctoral scientists preparing to do CAM research to obtain positions in either CAM and more conventional departments in research institutions, thus benefiting the CAM research field as a whole.... Providing support to such trainees is crucial to CAM research to ensure a sufficient number of highly trained researchers in this relatively new research field.
This selection has a number of flawed assumptions in it: let's start at the end: why do we need to "ensure a sufficient number" of researchers in a field that doesn't have any solid science to back it up? Scientific research isn't about sustaining any field for its own sake - research that gets results will grow naturally, and research that fails will shrivel away, as it should. Happens all the time. In fact, unless a field has money specially earmarked for it, the peer review process makes it very difficult to obtain funds for poorly designed research or, worse, pseudoscience. This is why we need to eliminate NCCAM.

So what kinds of research does NCCAM want the eager young researcher to pursue? Well, they state that CAM research is in "four major domains: mind-body medicine, biologically based practices, manipulative and body-based practices, and energy medicine." Since all of these "domains" are various forms of quack medicine, many scientists probably don't know exactly what they are - after all, our "conventional" training doesn't cover quackery. (Medical schools, take note! Your curricula need to be expanded.) But NCCAM is ready with some definitions, such as this page for energy medicine, which explains that there are two types of energy fields:
1. Veritable, which can be measured
2. Putative, which have yet to be measured

Yes, you read that right! NIH is funding research into energy fields that cannot be measured! Or to use NCCAM's phrase, that "have defied measurement to date by reproducible methods." NCCAM includes in this category the following types of "energy": qi (Chinese), ki (Japanese), doshas (Ayurveda, see my blogpost on that), prana, etheric energy, fohat, orgone, odic force, mana, and homeopathic resonance.

I couldn't make up better nonsense even if I were trying. But other people believe this stuff, so shouldn't we respect it? No! Does it deserve our scorn and derision? Yes! Scientists are harsh critics even about legitimate science, and we certainly shouldn't go easy on garbage like this. Actually, part of me wonders if someone at NCCAM put together this site as a sort of cry for help - it's like they're asking for someone to point out its ridiculousness.

So, young Jedi scientist, fire up your computer and write a proposal to NCCAM to study the force, or qi, or odic force, or whatever. Real science is much harder - you have to conduct experiments with reproducible results! Why bother when you can just make stuff up instead?

And finally, a postscript to anyone who's read this far and is thinking of applying for one of these NCCAM grants: if you want to try something really subversive, propose to do your postdoc training in my lab. You can "apply rigorous research methods to investigate the efficacy and underlying mechanisms of CAM therapies," exactly as NCCAM requires. I'll work with you to debunk as many claims as we can. And if at the end of a year you determine that it's all bunk, at least some of NCCAM's money will have been well spent.

A tiny drop of ink, a big win for science

Buried amidst the bad news on the budget today is this gem from Rick Weiss of the Washington Post:
It is barely a drop of ink in the gargantuan omnibus spending bill ... a provision that would give the public free access to the results of federally funded biomedical research represents a sweet victory for a coalistion of researchers and activists who lobbied for the language for years.

What this means is that after years of battling the publishers' lobby, patients' advocates and scientists have finally won a victory, even if it's not everything we wanted. What did we win? The language buried in the giant spending bill will require NIH-funded researchers (like me) to make their published research freely available to the public within 12 months of publication.

This is very good news, and the patients' advocates deserve the most credit. Thanks to the Internet, patients have discovered that they can read about the latest research on treatments for their conditions - cancers, infectious diseases, environmental hazards, and more. These same patients have discovered, as scientists have known for years, that most journals won't give you access to anything more than a title and abstract (and sometimes just the title!) unless you pay a heft fee, typically $30-$50 or more per article. These fees are truly unfair, even ridiculous, for work that was funded by the taxpayers in the first place. Patient groups have expressed growing outrage that they can't even read the research that they paid for - research that, in many cases, offers them the only hope they can find for a cure.

A recent statistic showed that a typical NIH-funded publication cost the taxpayers over $150,000. How could it possibly be that private publishers owned the copyrights to any of these studies? But they did, mostly because it's been a long tradition that scientists signed over copyright when they published in a journal, and the funders (usually the government) allowed them to. Believe me, scientists don't want to sign over copyright - it's just that we had little choice until the Internet came along, and more recently the emergence of open-access journals. Even today, though, despite the growth of open access, many of the leading journals in science and medicine are not open access, and their subscription fees are often exhorbitant. [As an aside, the Wash Post reports that the average scientific journal subscription has risen to over $963 per year, a cost that more than doubled since 1995.]

The new system isn't perfect: journals will be allowed to keep results under lock and key for up to 12 months, restricting access. This was a compromise that I hope we can improve upon. The original proposals called for either immediate access or a 6-month delay, but publishers and their lobbyists were able to get this changed to 12-months. Those of us who follow Washington politics know that NIH almost made this their policy over two years ago - but at the time, the lobbyists somehow got to the NIH Director (Zerhouni), who at the very last minute made it "optional." So of course the journals refused to change their practices and nothing changed at that time.

This new free access policy is a big breakthrough. Thanks and kudos to the citizens and scientists - many of them working tirelessly behind the scenes for the past several years - who finally succeeded in opening up scientific research to the public. Now let's work on reducing that 12-month delay to zero.

Time magazine's top 10 discoveries of the year: fooled by Craig Venter

So Time magazine has a list of its choices for "top 10 scientific discoveries of the year". We all love these top-10 lists - they're fun. But they made a huge error on one of them, so I'll make a small, probably futile, effort to correct it here.

Their #2 discovery is "Human mapped", which refers to Craig Venter's publishing his own genome. I blogged about this before, pointing out that PLoS Biology was lowering their standards by publishing that paper, which had virtually no novel scientific content. But Time reports breathlessly that the publication was "the first such genome ever published" of " all the DNA in both sets of chromosomes inherited from each of his [Venter's] parents." But was it?

No - of course it wasn't. If the Time reporter (Krista Mahr) had actually read the paper rather than the press releases, she'd have learned that the paper describes just one set of chromosomes - not two. In fact, Craig published a mishmash of his chromosomes, with bits and pieces of each chromosome merged together in a mosaic. (This is exactly what we did in the original human genome papers, by the way.) The only slight contribution of this year's paper was to fill in some of the gaps, and in some cases to separate the haplotypes (the two chromosomes) enough that, as I wrote in my original blog, they could claim that blocks of 200,000 bases on average came from a single chromosome. What does that mean? It means that in in the genome they published, each chromosome is a composite sequence containing many bits and pieces of both the parental chromosomes. For chromosome 1 there are over 1000 of these chunks, for example, and there is no information (in Venter's paper) about which chunk comes from which parent.

Furthermore, because Venter and colleagues didn't sequence his parents, they have no idea which parent any chunk came from. In other words, Venter did not publish his entire diploid genome: that would imply he published two copies of each chromosome, one from his mother and one from his father. Am I being clear here? Time magazine got this wrong, as did some of the original press reports.

Having worked with Craig for years and observed much of the press hullaballoo that he works so hard to generate, I'm not surprised that Time got this wrong. Craig is the main source of the misinformation: if you go to his institute's website today, as I did, the front page has a big image of a human karyotype (showing two copies of each chromsome), with the text below saying "New individual human diploid sequence." Sure sounds like there are two sequences for each chromosome, doesn't it? And the front page also has a link to a press release about the paper titled, "First individual diploid huma genome published by researchers at J. Craig Venter Institute." The press release goes on to refer to "six billion base pairs from the genome of one person." It also says that "the publicly funded genome ... is a composite version," implying that Craig's is not.

Obviously the Time reporter is reading the press releases: if you go to PLoS Biology and look at the paper, the second sentence of the article says that the genome comprises 2,810 million bases. If anyone is counting, that's just under 3 billion, not 6 billion. The paper also makes it clear that Craig's parents were not sequenced, so there is no way to tell which parts of any chromosome came from which parent.

Clearly Craig is trying to re-write scientific history by press release, and he might even get away with it. I claim that it is scientifically dishonest for him to have these misleading statements on his institute's website. If Time magazine were to look at the actual scientific paper, they'd realize that this item on their top 10 list is sloppy journalism, and an unfortunate mistake.

Let's be clear here: the Venter Institute's own press release says
"Independent sequence and assembly of the six billion base pairs from the genome of one person..."
but the paper reports 2.81 billion bases. That's not six billion. And that is misleading. I think we can get the public excited about science without lying to them - and it's too bad that it succeeded (this time) in getting Time magazine to report a "discovery" that didn't happen.

Damn, I lost another 4000 genes!

Where do those genes keep going? Just a few short years ago, we were arguing about whether the human genome had 100,000 genes or maybe just 80,000. Then we published the human genome ("we" being about 2000 co-authors on 2 papers) and reported that alas, we only had 25,000-30,000 genes. What a let down. The lowly weed Arabidopsis thaliana has 25,000 genes - can't we do better than a plant? It doesn't even have a brain, for pete's sake!

Well, don't look now, but a new study by Michele Clamp and colleagues at MIT's Broad Institute, just published in PNAS, says that the number, which had been hovering at 24,500, should be reduced to 20,500. Their analysis is pretty convincing - they did careful alignments of all human genes to both mouse and dog, and were able to identify several thousand genes that didn't seem to exist in our furry friends. Wait, though, you might say: perhaps these are specific to humans, or at least to primates? No, they thought of that too, and virtually none of the "genes" they propose deleting were found in our chimpanzee cousins either. That means that either we just delete these genes, or accept the rather far-fetched hypothesis that primates have both a "prodigious rate of gene birth" (to allow for over 1100 human-specific genes) and a "ferocious rate of gene death" (to explain why we don't see a similar number of genes shared by humans and chimps but no other species).

They eliminated quite a few other genes on the way to reducing the gene count by 4000, and I must admit their methods are pretty compelling.

So it appears that we have to take ourselves down another peg. We still have more genes than fruit flies, but the gap is getting smaller. I've been telling my students that size doesn't matter - we don't have the biggest genome, we don't have the most genes, we're not the biggest creatures by a wide margin. And lately we've learned that our own bodies have more bacterial cells (by a factor of 10 or more!) than human cells.

We humans want to think we're tops at something - wait, I've got it: we have the biggest egos! But those darn cats, now, I don't know...