On science blogs this week: Limits
LIMITS TO THE LIMITS OF DNA. Geneticists have known it for decades, and those of us who write about genetics have been explaining it for years: genes are not destiny. The most obvious example is "identical" twins. Their genes may be identically structured, but the courses of their lives are not. Living life shapes the way genes work.
So it's a puzzlement why there was so much attention to the twin-study paper published early this month, including a big New York Times piece trumpeting The Limits of DNA. The paper noted that sequencing the genes of individuals isn't going to predict their future health with precision.
I'll get to some links for catching up with the story in a moment, but first I want to point to a post by Erika Check Hayden at the fine group blog Last Word on Nothing because it has some fresh things to say to those of us who write about science. She had written, for the Nature Newsblog, one of the few news items about the twin-study paper that mined skeptical commentary from other scientists. That sent her to Last Word musings on the barriers to covering science well right now.
She describes half a dozen of those barriers, some already all too familiar, such as the pressure to write much more and much faster. But she identifies one barrier in particular that we don't talk about much. The paper is an example, she said, of the increasing number of "big data" studies resting on complex statistical techniques that few science writers comprehend. We're not talking here about the basics every science writer should know, like paying attention to sample sizes and levels of significance and standard deviation and the difference between randomized double-blinded clinical trials and observational studies.
Hayden notes that she would not have recognized the statistical problems in the paper on her own; she needed savvy scientists to point them out. I don't see any way of dealing with this increasing statistical complexity except to spend more time seeking and quizzing outside sources, as she did. Not so easy, and especially not so easy for science journos forced to write much more in much less time.
What helped her out was Twitter. In the comments on her Last Word post, Hayden noted that it was Twitter chatter about the Times piece and the twin-study paper's deficiencies, some of the chatter being from scientists, that helped persuade her that the paper needed writing about in a different way. Read her Storify of the tweets here. A similar Storify was posted by Matthew Herper at Forbes.com.
But note that these tweets, useful as they were for revealing what the story really was, were a reaction after the story was already out there. Twitter did nothing to prevent the initial over-promotion, which left readers with the impression that this was brand new stuff when it isn't.
For catching up on the hype and Hayden's original critique, see Paul Raeburn's analysis of the daily journalism on this paper at the Knight Science Journalism Tracker. For a view from the personal genomics industry, see Luke Jostins at Genomes Unzipped. In addition to legitimate criticisms of lack of novelty and questionable methodology, he argued
to be honest I suspect that a lot of this is a mixture of indignation and sour grapes that this paper, a not particularly original or particularly well done attempt to answer a question that many other people have answered before, got so much press (including a feature in the NYT).
He blames himself and his fellow scientists for not getting the message across, but I'm not sure the responsibility is theirs. I think many in the audience resist hearing the message. We all yearn for simple answers and simple solutions. Readers don't want to know that it's all very complicated and much remains unknown. Editors don't want to know it either.
So I guess you could argue, given the popular and media embrace of genetic determinism with tales of "the gene for" this or that trait, that you can't point out that genes are not destiny too often. The ad world has always known that repetition is the key to selling something.
It's clear — has been for ages — that health is often bound up not with genes per se, but with gene regulation and epigenetics: how and when a gene is turned on or off, and in what tissues, and how much (or little) of its protein it makes, and how that protein is modified before it sets to work, and how a gene's activity is altered by the activity of other genes (including how those other genes are regulated.) And all of these influenced by the complex and mostly unidentified events in the world around us.
A Chinese box if there ever was one. Which is why, at the turn of the century a dozen years ago, when the first "completion" of the Human Genome Project was announced, uber-sequencer Craig Venter observed that it would take most of this century to figure it all out. If anybody knows, he should.
DINOSAURS FROM SPACE. The question here is: how did the alien dinosaurs get into uber-chemist Ronald Breslow's somewhat technical paper on chirality and from there become prominent in the press release? (Chirality is a form of asymmetry in which an object is not identical to its mirror image; that is, it cannot be superposed on the image. The usual analogy is to human hands, which are chiral, anatomical opposites. In chemistry, chirality applies to molecules.)
At Sciencebase, David Bradley wondered if Breslow — an illustrious former president of the American Chemical Society — was maybe trying to spoof the credulous media. David points out
Of course the actual paper on which the release was based was nothing to do with dinosaurs or in fact any kind of alien, although it does mention them and the press release does lead on the dinosaurs! The paper was actually about chemists trying to understand why we have only one handed forms of amino acids and sugars on earth and how this may or may not have underpinned the emergence of life.
At Discoblog, Sarah Zhang imputes pedagogical motives to the ACS publicity machine:
...this press release is really a perverse kind of genius. When was the last time the science media lit up with a very technical paper on the origins of homochirality?
At Dinosaur Tracking (Where Paleontology Meets Pop Culture), Brian Switek is incredulous. He trots us through a couple hundred million years of dinosaur evolution to show that their existence — and their disappearance — was a function of chance calamitous extinctions that occurred uniquely on Earth. The notion that high-IQ dinosaurs could arise on an exoplanet, whether their amino acids and genetic material were right-handed or left-handed, is literally fantastic.
Charlie Petit, at the Knight Science Journalism Tracker, sorts through the press coverage and finds, to his relief, that nobody serious took Breslow's saurian speculation seriously. Charlie's advice:
Note to scientists: Don’t make witty remarks at the ends of your serious papers without being sure any press release writer that chews on it doesn’t get thrown too far off course.
I hope hope hope somebody interviews Breslow soon. I'd love to know what he thought he was thinking.
LIFE FROM SPACE. AND CONTRARIWISE. We're not done with Breslow quite yet. Before he introduced alien dinosaurs, Breslow suggested that meteorites descending on the Earth early in its history could have sprinkled the planet with chiral amino acids that were precursors to life. Which is not at all a crazy notion.
Now comes a post from the Physics arXiv Blog at Technology Review proposing that the seeds of life might have constituted a kind of living feedback loop — and that dinosaurs are involved once again, at least peripherally.
KFC describes a paper from Kyoto Sangyo University in Japan arguing that the enormous asteroid impact that killed off the dinosaurs and much else 65 million years ago ejected billions of tons of rocks and water, perhaps as much mass as the asteroid itself. This mass would have borne many living things into space. Very, very far into space. As far as other places potentially habitable: Jupiter's moon Europa, Saturn's moon Enceladus, and, just maybe, Earth-like exoplanets orbiting other stars.
The Japanese researchers estimate that a thousand rocks from Earth could have reached the red dwarf Gliese 581 in a million years. Gliese 581 is about 20 light years away, and a super-Earth is believed to be orbiting at the edge of its habitable zone.
But it gets better. Assume microbes could survive for the required distances and time they would need to spread throughout the amount of space occupied by the Milky Way. The Milky Way galaxy itself is about 10 billion years old. But
...if life evolved at 25 different sites in the galaxy [10 billion] years ago, then the combined ejecta from these places would now fill the Milky Way...If this scenario has indeed taken place, [the researchers] say: "then the probability is almost one that our solar system is visited by the microorganisms that originated in extra solar system."
DO ANDROIDS DREAM OF ELECTRIC WARS? The Defense Advanced Research Projects Agency (DARPA) has issued another one of its Grand Challenges, this one a call to invent humanoid robots. At Danger Room, Wired's national security blog, Katie Drummond and Noah Shachtman report
Within the next few weeks, Darpa, the Pentagon’s far-out research arm, is expected to launch its contest, which will likely ask roboteers to build a bipedal robot that can do things like drive cars, open doors, traverse rough terrain and show off its fine motor skills, perhaps by repairing busted pipes.
Previous DARPA Challenges have bestowed millions on developers of self-driving cars, apparently with a certain amount of success. A humanoid robot, however, is not expected to be a piece of cake.
ROBOCOPS FOR REAL. In other robot news, also at Danger Room, Spencer Ackerman reports that hundreds of returning robotic veterans of our waning wars will be joining police forces. These will not, apparently, be drone aircraft lobbing rockets at bad guys, but rather are likely to be ground-based machines talented at tasks such as disarming bombs. Or maybe not; a Defense Department spokesperson says the Army is not yet sure just which of its many varieties of robots will be declared war surplus. Good to know that at least some returning veterans will be able to find work. And they probably won't suffer from PTSD either.
A relevant historical note, the marvelously named Cyriaque Lamar tells us at io9 that Hugo Gernsback, one of the founders of science fiction, invented robowarriors nearly a century ago, in 1918. They looked very much like Daleks, although I suppose that might have been the illustrator rather than Gernsback.
io9 is absolutely on top of robonews, this time from the more conventionally named Alasdair Wilkins, who reports that scientists at UC-Davis have built a robotic squirrel to study the tail-wagging signals live squirrels produce when they encounter rattlesnakes. Yes, really. See for yourself; there's a small not- very-clear video with the post.
ONE AND ONE AND ONE IS 202. But here's the very best io9 robovideo: a robot band playing The Beatles' "Come Together." The band is the creation of engineers at Drexel University, who made the instruments out of PVC piping. The song being one of John Lennon's more opaque creations — here are the lyrics, I rest my case — which is saying a lot, I was wondering why the engineers chose it.
Thus I reveal my euphonic ignorance. Checking with Wikipedia, I learn that Rolling Stone has ranked "Come Together" as #202 on its list of 500 Greatest Songs of All Time, and an impressive #9 on its list of 100 Greatest Beatles Songs.
You've heard the robots; now listen to the Real Thing. He got toe jam football indeed.