“imagine what it would feel like to be a sperm” »« But my chariot team is the Greens!

Remedial reading for big-time scienticians

I don’t understand how this happens. You’ve got a good academic position. You’re bringing in reasonable amounts of grant money. You’re publishing in Nature Genetics and Nature Structural and Molecular Biology. And you don’t even understand the basic concepts in your field of study.

For instance, here’s a press release titled “Cause of genetic disorder found in 'dark matter' of DNA”.

For the first time, scientists have used new technology which analyses the whole genome to find the cause of a genetic disease in what was previously referred to as "junk DNA". Pancreatic agenesis results in babies being born without a pancreas, leaving them with a lifetime of diabetes and problems digesting food. In a breakthrough for genetic research, teams led by the University of Exeter Medical School and Imperial College London found that the condition is most commonly caused by mutations in a newly identified gene regulatory element in a remote part of the genome, which can now be explored thanks to advances in genetic sequencing.

Regulatory elements are not and have never been considered junk DNA. The researchers have identified a regulatory region called PTF1A that has allelic variations that cause a failure of the pancreas to form. That’s really interesting! But then you read what they have to say about it, and they are completely oblivious to the literature on genetic structure and gene regulation. Isn’t that something you’d expect them to have studied thoroughly before even proposing this project?

Or how about this press release, “Un-junking junk DNA”. It’s gotten to the point where I just cringe when I see the phrase “junk DNA” in a press release, because it is a sure sign of flamboyant ignorance to come.

"This study provides answers for a decade-old question in biology," explained principal investigator Gene Yeo, PhD, assistant professor of Cellular and Molecular Medicine, member of the Stem Cell Research Program and Institute for Genomic Medicine at UC San Diego, as well as with National University of Singapore. "When the sequence of the human genome was fully assembled, under a decade ago, we learned that less than 3 percent of the entire genome contains information that encodes for proteins. This posed a difficult problem for genome scientists – what is the other 97 percent doing?"

The role of the rest of the genome was largely a mystery and was thus referred to as "junk DNA." Since then sequencing of other, non-human, genomes has allowed scientists to delineate the sequences in the genome that are remarkably preserved across hundreds of millions of years of evolution. It is widely accepted that this evidence of evolutionary constraint implies that, even without coding for protein, certain segments of the genome are vital for life and development.

So many misconceptions. No, noncoding DNA is not synonymous with junk DNA; junk DNA was not so called because its function was mysterious; it is absolutely no surprise that some regions of the genome are vital, even without coding for proteins — haven’t they heard of tRNA or miRNA? Developmental biologists have been yapping for decades about the importance of the switches that control gene regulation…are we just ignored?

I worry that this is a symptom of a serious rot in science education — that we’re turning out great technicians and masters of the arcane art of grant writing who don’t actually understand biology, and in particular have no perspective on what the questions actually are. They may be excellent middle managers, but the comprehension and vision are lacking.

I have a suggestion. If you’re going to do research that leads you to say anything about junk DNA, I urge you to read carefully one or all of the following books: The Origins of Genome Architecture by Michael Lynch; Fundamentals of Molecular Evolution by Dan Graur and Wen-Hsiung Li; or The Logic of Chance: The Nature and Origin of Biological Evolution by Eugene Koonin. Those aren’t lightweight texts — I wouldn’t assign them to your average undergraduate — but hey, you’re a big-time professional scientist. There’s no excuse for not knowing this stuff.

Comments

  1. Dunc says

    Yeah, I was just going to ask how we can distinguish between a PI that doesn’t understand the subject, and a PI that understands all too well what sort of bullshit they need to shovel in order to get some media coverage… I mean, I know pretty much stuff-all about genetics, but even I know that there’s a big difference between “non-coding” and “non-functional”. But I also know that if you want some journalist to write you an article, you need to give them a hook they can understand.

    However, I would still agree that it’s “a symptom of a serious rot in science education” – it’s just that it’s the journalists who I see as exhibiting it.

  2. blf says

    Those aren’t lightweight texts — I wouldn’t assign them to your average undergraduate — but hey, you’re a big-time professional scientist.

    Which means “they” will assign a grad turkey to write a grant proposal for some other grad turkey to read those books and write up a series of LPUs (Least Publishable Units). With the “they” as the PI and first credited “author”.

    More “papers”, lab funding, and the Graduate Students / post-docs might actually learn something. Win, win, win, all around.

    </snark>

  3. says

    DO NOT LET THE PIs OFF THE HOOK.

    Of course they are responsible for the lousy coverage of their work. If nothing else, they should be complaining, and they should be making a stink about poor science media.

    Also, as we’ve discussed here before, often the hype is coming from the PIs themselves, and the media cheerfully gobble it up.

  4. iknklast says

    This isn’t just in genetics, either. In my field (environmental science), students in the masters and doctoral programs where I went to school had so little theoretical science that they couldn’t really interpret the meaning of the data they collected. They knew what the numbers were supposed to be, but when they were off, they didn’t have enough knowledge to work through a list of possible problems. They were really just technicians.

    I felt fortunate that I moved from a theoretical school where I did my masters to an applied school where I did my doctorate. My theoretical school did a great deal of applied science; my applied science program did almost nothing theoretical. I feel like I got the best of both worlds, but the students in the second school were being short changed if they didn’t have the same type of pathway, because they didn’t really understand the science behind their technology.

  5. Dunc says

    If nothing else, they should be complaining, and they should be making a stink about poor science media.

    Fair enough, but I don’t think you can blame them for not doing that in their press releases. “Publish my results! And by the way, your journalism sucks!” is never likely to be a successful tactic.

    Lousy science media is indeed a problem, but as with all problems with the media, it’s difficult to persuade the media to publicise any discussion of it. You very rarely see complaints about the media in the media, for what I would hope are fairly obvious reasons.

    Indeed, this is a huge problem with the media generally. It’s arguably one of the most pressing issues of our times, as most of the other pressing issues of our times involve at least some degree of media failure. How do we, as a society, hold the media to account?

    Also, as we’ve discussed here before, often the hype is coming from the PIs themselves, and the media cheerfully gobble it up.

    Well yes. That was exactly my point – if you want the media to show any interest in your work, you have to coat it with a liberal dose of hype. We could call it “Michio Kaku Syndrome” – the willingness to say any old bollocks in order to get in the media. Unfortunately, it’s fast becoming a key skill.

  6. paulhavlak says

    I’m having trouble digging up the original paper, probably because it’s so recent (even on the Nature Genetics site).

    But even from the press release, it sounds like the terms “dark matter” or “junk DNA” may be fair enough, for this result, for a general audience. The researchers “found that the condition is most commonly caused by mutations in a newly identified gene regulatory element in a remote part of the genome”.

    If by that they mean a trans-regulatory element not found proximal to some protein-coding gene, then damn straight, they have found something interesting in the bulk of DNA that by default would be assumed junk and not subject to sequence-scale selection pressure.

    Doing this in a specific experiment is cool science, regardless of whether it gives solace to those who make exaggerated claims about sequence selection and functionality for the bulk of DNA.

  7. paulhavlak says

    Also, most of Eugene Koonin’s book is surely great, but I find his use of many-worlds interpretations of quantum physics to bootstrap the protein+DNA world — “anthropic selection” building a bridge to biological selection — to be profoundly weird.

    More weird and less scientific/empirical than exaggerating the proportion of a genome that’s functional.

  8. frog says

    Ugh. The media have apparently decided that “junk DNA” means “parts of the DNA where we haven’t yet figured out what it does.”

  9. says

    I worry that this is a symptom of a serious rot in science education — that we’re turning out great technicians and masters of the arcane art of grant writing who don’t actually understand biology, and in particular have no perspective on what the questions actually are.

    I wrote a semi-frivolous piece a while back (link) that looks at a possible partial explanation for problems like this. I proposed that in the pursuit of science (and science education), over reliance on technical jargon leads to a risk of people being fooled by their ability to manipulate familiar technical terms in a convincing or formulaic manner into never taking the time to figure out what those terms stand for. (I cite anecdotal evidence for this.)

    Another thing that makes me cringe, for similar reasons, is when scientists (usually physicists) report that they have no need for philosophy. It is as good as saying that you have no need to think about the meaning of the words you are using. A scandalous attitude for any scientist to take.

  10. Pierce R. Butler says

    Regulatory elements are not and have never been considered junk DNA.

    By a great cosmic coincidence, just this morning I was reading”Revolution Postponed” by Stephen S. Hall in the Oct ’10 issue of Scientific American (an overview of why the great advances predicted “in 10 years” at the 2000 unveiling of the results of the Human Genome Project have not yet appeared), which sayeth:

    The vast areas of DNA that do not code for proteins, once dismissed as ‘junk,’ are now known to conceal important regulatory regions.

    Had Hall (apparently a journalist, not a scientist) cited an example or two, I might consider our esteemed host’s assertion disproved. As it is, the question remains open: somebody get me a grant!

  11. terminus says

    Intrigued by your book list, I checked out the Michael Lynch reference…guess who gave the 1st Editorial Review for his book??? Axel Meyer, from Nature. Maybe he didn’t read it carefully.

  12. chris61 says

    I think the Nature Genetics paper is very clever and so I don’t mind cutting the PR folk some slack for using the term junk DNA. As best as I recollect junk DNA was indeed the term used to describe the vast majority of most animal/plant genomes that doesn’t code for protein, with the understanding that of course some or much of it had a function that we just had no way of identifying. These investigators identified a sequence with a previously unknown function that appears to be important in forming a pancreas. I think that’s pretty impressive.