welcome

Researchers in the Virtual Cell Program at Harvard have finally published the official release of "little b," a computer language for cell biologists. The summary in the press release is almost impenetrably abstract, so I'll try to explain why this is important.

For years, biologists have been building "wiring diagrams" to explain how information travels through a cell. Usually, these diagrams use cartoonish blocks and blobs to represent proteins, DNA, and RNA, and arrows and notes to indicate how they interact. It's a great teaching tool, but in recent years it's reached its practical limits: cells, it turns out, are just way too complicated to fit into a comprehensible wiring diagram.

The current issue of Science has a report with some reassuring findings for my profession:

Scientists and journalists get along much better than the anecdotal 'horror stories' would lead us to believe, according to new research published today in the journal Science, which has found that 57% of researchers were 'mostly pleased' with their media interaction, while only 6% percent were 'mostly dissatisfied'.

Previous research as well as anecdotal evidence has tended to focus on the negative aspects of scientists' media interaction, but today's survey, based on the responses of 1354 scientists working in the high-profile research fields of epidemiology and stem cell research in the UK, US, France, Germany and Japan, suggests that, for the most part, scientists are comfortable dealing with journalists.

This aligns well with my own experience, but I wasn't sure that the phenomenon was so widespread. When I contact researchers, it's usually for a trade magazine or research journal they already trust, and of course those letters after my name don't hurt either. It's good to know that the lay press gets similarly collegial treatment, because communicating with them means communicating with the general public - which, after all, is paying for much of this cool science.

In the past month or so, I've crossed paths several times with a hype machine that I find increasingly annoying. This growing juggernaut, called "nutraceuticals," is driven by marketers and fueled by scrapings from the fringes of science, and it periodically emits little advertorial globules for mass consumption.

Here's a sample from the most recent one, a press release that was picked up, unskeptically, by wire services and news outlets around the world:

Writing this week in the online, open-access journal Public Library of Science (PLoS) ONE, researchers from industry and academia, including the University of Wisconsin-Madison and the University of Florida, report that low doses of resveratrol -- a natural constituent of grapes, pomegranates, red wine and other foods -- can potentially boost the quality of life by improving heart health in old age.

I'm all in favor of rationalizations for drinking, but as the folks at Spoonful of Medicine point out, wine also includes plenty of substances that aren't so constructive, so these sorts of single-ingredient analyses are always suspect. Worse, though, the data don't actually show what the researchers are trying to claim. They didn't see improved cardiac health, or longer lifespans, or any other clear measure of goodness in the resveratrol-treated mice.

What they saw was a set of changes in gene expression that partly overlaps the changes seen with other interventions that have those benefits. Parsing that, we're left with an arcane laboratory result from a relatively cheap, easy experiment. What's its connection to any meaningful physiological change? Nobody knows. That's why most good journals aren't taking papers like this anymore, because few of these sorts of findings pan out on later investigation. In this case, the answer is even more obvious; better studies have already shown that the benefits of resveratrol - in mice - occur only at doses one would never obtain from natural sources. And do I even need to mention that we have absolutely no relevant human efficacy or safety data?

Unfortunately, these kinds of shenanigans are becoming more commonplace. For example, in late May you might have seen a news blip about the psychogenic properties of frankincense, a famous ingredient in incense. The wire services essentially took the press release as camera-ready copy:

Religious leaders have contended for millennia that burning incense is good for the soul. Now, biologists have learned that it is good for our brains too. In a new study appearing online in The FASEB Journal, an international team of scientists, including researchers from Johns Hopkins University and the Hebrew University in Jerusalem, describe how burning frankincense (resin from the Boswellia plant) activates poorly understood ion channels in the brain to alleviate anxiety or depression. This suggests that an entirely new class of depression and anxiety drugs might be right under our noses.

Once again, the team didn't study the material they're hawking - in this case incense smoke - but a highly purified single compound from it. Injecting high doses of this compound, called incensole, intravenously into mice, they found biochemical and neurological evidence that it might be acting as an antidepressant. While that may be an interesting basic research result, it's nowhere near the lesson the press release wants us to take home. The researchers may have been inspired by a smell from church, but the story they ended up telling the media smells more like a barn.

If you think these are just scientists engaging in the time-honored habit of making their work sound more grant-worthy, I suggest visiting your local supermarket and taking a good look around. In the aisle with the largely unregulated "nutritional supplements," which now undergird a multi-billion-dollar industry, see if you can find pills that contain resveratrol. You probably can, and I'll bet at least some of the bottles cite "authoritative" studies such as the ones I discussed above. Wait a few months, and you'll find similar labeling on incense, and probably even incensole-containing pills.

Now, walk through some other aisles and look at the processed foods. Ketchup, for example, won't trumpet its more abundant but less desirable ingredients, but it will probably advertise that it contains lycopene. That's another natural compound that might be good for you, albeit at levels you'll never get from ketchup. But the label cites authoritative-sounding studies without any respect for their actual data.

Finally, while I'm not a fan of the sorts of conflict-of-interest witchhunts that have come into vogue, I did notice the fine print under the author list on the recent PLoS ONE resveratrol study, which indirectly discloses that it was underwritten by a large nutritional supplement maker. They certainly got their money's worth.

Ever see some new device, and say "wow, why didn't anyone think of that before?" I mean, really say it and mean it - not sarcastically, and not as part of a hackneyed marketing phrase, but because you really, truly wonder how this invention escaped discovery for so long. That doesn't happen very often, because let's face it, most technological development is pretty easy to see coming. The other day, though, I bought one of these:

It's a recording microphone for telephone conversations. Looks like an ear bud, doesn't it? That's because, in contrast to every other ugly hack I've tried for this application, it doesn't plug into the phone wire or suction cup onto the phone receiver. No, it plugs into your ear, quietly interposing itself between the phone receiver (which echoes both sides of the conversation already) and your ear. Get it?

Besides the jaw-droppingly obvious (in retrospect) design concept, it also has nice construction and packaging, with three different sizes of earplugs, a set of adapters for different recording devices, and a classy-looking brushed silvery exterior - and the sound quality is excellent. If you need to record phone calls, you need this device.

Genome Biology just published an interesting paper by a group of prominent protein chemists, plus Jimmy Wales, the creator of Wikipedia. In a nutshell, these folks have developed a new wiki-based protein annotation system, which seems like a wise way to approach the enormous problem of organizing information about these molecules. In an accompanying press release, they explain the work this way:

The source material for WikiProteins comes from a mixture of existing authoritative databases (such as the Unified Medical Language System, UniProtKB/Swiss-Prot, IntAct and GO), supplemented by concepts mined from scientific papers published in public literature databases. The automated data mining identifies ‘facts’ in these available resources, such as protein functions or protein-disease relationships. This process created over one million biomedical concept clouds – called ‘Knowlets’ – around each individual concept. The developers of the site now hope that many researchers will follow their call to annotate, via WikiProteins, the Knowlets for which they are leading experts. The method enables researchers to add data even from sources that are not openly available, such as from journals only accessible via publishers’ databases, immensely enhancing the potential for comprehensive coverage. Each page of text called up via the system is automatically indexed and concepts are connected to the WikiSpace, so that their definition comes up and the information can be edited directly from the page.

I wish they could have accomplished it without feeling the need to coin a new buzzword (Knowlets? Didn't they hang out with the Hobbits?), but hopefully the system will evolve into a useful tool as users edit it. If you're anxious to test drive it, there's a sample page up here.

Over the past few years, a model of scientific journal publishing called "Open access" has generated tremendous buzz. Relying heavily on digital technology, and preaching a gospel of free access to scientific information, open access journals allow anyone, anywhere, to read the latest research without having to pay a hefty subscription fee. It's an attractive idea for a lot of scientists, who pride themselves on sharing their results as widely as possible. A version of Open access publishing has also now become the law of the land for many academic researchers in the US.

I've posted about this trend before (here and here), but a post on Spoonful of Medicine got me thinking about it again:

Most of the comments [on an earlier post] have centered on what I wrote about the fact that open-access publishing is not the only alternative to scientific publishing, but just one of several models. Some people take strong exception to this idea to the point of feeling violated by the fact that we "sell back" the science they produce. Others acknowledge that we provide a filtering service, but point to the fact that the peer-review process is free. And a third group of critics argue that the problem with scientific publishing can be summarized in three words: Nature, Science and Cell. Each of these criticisms deserve some comment, and I'll start with the concept that peer-review is free.

It's worth reading the rest of Juan's post for an insightful interpretation of that "free" peer review. I'm going to take on a slightly different issue, though: who ultimately pays for "Open access" publishing?

First, let's get some terms straight. "Open access" technically refers to the Open Source-like license under which this new crop of scientific articles is published. The idea is that by opening up the copyrights, the publications become more "Free" for readers to use, as well as being "free" of cost. It's been an incredibly successful model for software development, so why not extend it to science?

To most scientists, and many other observers not directly involved in the science publishing business, Open access looks unambiguously double plus good. The standard reasoning is: the public pays for most basic research, so the public is entitled to use the results of that research without having to pay an additional subscription fee. It's a simple, intuitive argument. Like many simple, intuitive arguments, it's also wrong - or at least woefully incomplete.

Okay, so they did a little more legwork than I, but the central concept and a lot of supporting data were there in my piece for Nature Medicine back in 2006. For that matter, I wasn't even the first to pick up on this. What's news, really, is that the general public might actually notice now. In case the day's more salacious news item distracted you, here's a good summary from the Wall Street Journal:

An investigation by the Associated Press found trace amounts of scads of drugs in drinking-water supplies around the country. For a list of what was found in the watersheds of 28 metro areas, click here. Among the water’s offerings were antibiotics, anti-convulsants, mood stabilizers and sex hormones. There were traces of sedatives in water serving the city that never sleeps.

As was the case in 2006, nobody's quite sure what all these drugs in the water really mean. They could be very harmful, or a little bit harmful, or completely harmless. The pharmaceutical industry would like to believe the last, and least likely, of those three scenarios, while environmental extremists would prefer the first, and only slightly more likely one. These chemicals were developed specifically to affect human metabolism, so it's kind of unreasonable to expect that taking small doses of them - every day, for decades - would have no effect. They are extremely dilute, though, and we haven't seen any huge public health catastrophes in population statistics yet, so I'm betting that American tap water is still just about the safest beverage on Earth. Nonetheless, it might not be a bad idea to dump less of this stuff down the toilet.

Dutch clinical psychologists have announced the release of a computer program that allegedly does psychotherapy. I was amused to see them claim that the system, called MindMentor, is the first of its kind:

Two clinical psychologists associated with the Institute for Eclectic Psychology in Holland, Jaap Hollander and Jeffrey Wijnberg, have developed the first robot psychologist, named "MindMentor." MindMentor is an online computer program that helps people solve problems and achieve goals. It has the unique quality, as compared with other on line psychological help systems, of requiring no live human intervention and being completely automated. Said Hollander in an interview with a Dutch radio program: "What made this whole endeavor exciting, was that we suddenly saw a possibility to create an unlimited amount of psychological help."

Unlimited, but not free. If you visit the system (at MindMentor.com), you discover that this psychologist charges about ten bucks an hour. That's cheap for a psychologist, but infinitely more expensive than his competition. If you're an Emacs user, just type "M-x doctor" and enjoy a free session that lasts as long as you like. Maybe the Emacs version is more primitive, but it would be very interesting to see someone do a study to see what its "success" rate is, using the same definition as the Dutch researchers use for their system.

And, of course, I'd be remiss if I didn't ask the obvious question: when androids dream, what is the significance of the electric sheep?

When I started blogging a couple of years ago, I vowed that I would keep the site focused on substance, rather than veering into the infinitely recursive trap of metablogging, or posting essays about posting essays. This introduction shows where that can lead; right now, I'm writing about writing about writing. Who wants to read that?

Nonetheless, science blogs face some peculiar challenges, and I've noticed some of my blogging colleagues struggling with them as well. The crux of the problem is that science bloggers, unlike political, social, or personal bloggers, must confront a genuine information glut. When you boil politics, social groups, or your personal life down to bare facts, the sad truth is that very little is really happening from day to day. A political blogger, for example, simply needs to read the day's top few headlines - which can probably be summarized in a sentence or less - and then spin them according to a defined agenda.

Science is much harder to cover. On any given weekday, I receive upwards of 50 press releases and have access to hundreds more. If I'm motivated, I can also scan the tables of contents of dozens of top journals, and if I were really motivated, I could read full papers from any of them. Unlike most news items, these papers are not mere echoes of each other. Each is a nugget of pure, original information never before known by anyone.

One approach to this deluge is the mile-wide foot-deep strategy. To do that, I would sift the vast crop of current papers through my limited interests and skills, focusing on just the truly significant work that I could understand well enough to explain clearly. That would leave me with more than a dozen original stories a day, so the postings would have to be very brief. There are many science news outlets that do just that, usually with a team of contributors sending out a stream of postings, so dovdox.com would simply become another voice in this wilderness of data.

The other extreme is the pure essay blog, in which a poster cites some new work, then discusses it at length, preferably adding a new dimension to the discussion and bringing up ideas that even the original authors might not have considered. Elio Schaechter at Small Things Considered is a master of this focused approach. He only posts a few times a month on average, and his choices can seem odd sometimes, but nearly every post is interesting and informative.

I've generally leaned toward the essay end of the spectrum, but there's an added complication for me: it feels like work. Writing in-depth coverage of specific research areas is how I've been paying the bills for the past ten years. Occasionally, I'll post something to the blog to plug a story I've recently sent to press, or to add information that wouldn't fit in the final draft, but many of the pieces I write don't really lend themselves to that. Though it may not be obvious, I also practice a fair amount of self-censorship - for every post that seems intemperate or off-topic, there were ten others that were even worse.

My little blog isn't the only one with these challenges. In fact, I was heartened to see that the very smart and talented editorial crew at Nature Medicine is wrestling with exactly the same issues on their recently launched blog. That site got off to a strong start, then went nearly dormant for awhile, and has recently had a resurgence of postings. Besides fluctuating in volume, their postings, like mine, have also been casting about for an editorial voice. Sometimes they're writing about a specific research paper, while other times they're commenting on the state of childcare facilities at conferences.

Market research isn't much help here, either. Reviewing the logs for dovdox.com, I find that the postings that have received the most hits and comments are all over the map. By either measure, the top item by far is Callsign Search Widget Fixed, in which I posted a patch for a very obscure piece of software. Vying for a distant second place are postings about pesticides, microbiology, and a building demolition in New Haven.

Or perhaps that's the answer: there's no telling what will interest readers of this blog, so I might as well post whatever comes to mind. The corollary is that nobody's paying much attention to consistency or editorial voice, so I shouldn't bother with those things, either. In other words, the last thing blog readers want to ponder is ... metablogging.

Continuing my occasional theme of poking fun at the Federal Register (see this post and this one), here's another fascinating bit of insight into our weird republic:

Quarterly Update to Annual Listing of Foreign Government Subsidies on Articles of Cheese Subject to an In-Quota Rate of Duty

Section 702 of the Trade Agreements Act of 1979 (as amended) (``the Act'') requires the Department of Commerce (``the Department'') to determine, in consultation with the Secretary of Agriculture, whether any foreign government is providing a subsidy with respect to any article of cheese subject to an in-quota rate of duty, as defined in section 702(h) of the Act, and to publish an annual list and quarterly updates of the type and amount of those subsidies. We hereby provide the Department's quarterly update of subsidies on articles of cheese that were imported during the period July 1, 2007 through September 30, 2007.

You can read the full text here, but even after doing that, I'm left wondering: exactly what is an "article of cheese"? I'm pretty sure this posting would count as an article about cheese, but I don't receive any foreign subsidies, so hopefully The Act (as amended) won't require The Department to list dovdox.com next year.