Scientists cynical use of "Junk DNA"

This blog – like many others I presume – was started to give me a place to vent about a pet peeve. The target of my particular ire is the way that scientists who should know better continue to tout every new paper on the function of non-coding DNA as a new discovery that – GASP – “junk DNA is not really junk afterall“.

The latest example surrounds a paper from my friends and former neighbors Jim Noonan, Shyam Prabhakar and Eddy Rubin published last week Science (I won’t link the paper because it’s not in an open-access journal – another pet peeve…).

Prabhakar et al. Human-specific gain of function in a developmental enhancer. Science 321(5894):1346-50.

The paper reports on a the discovery of a conserved noncoding sequence (named HACNS1) that acts as a developmental enhancer and has evolved extremely rapidly in humans and has gained a strong limb expression domain relative to the orthologous elements from chimpanzee and rhesus macaque. It’s a beautiful piece of work that has both intriguing implications for human evolution and will serve as a paradigm for similar studies in the future.

What bothers me is not the paper, but the press release that accompanied it. Here’s the headline and beginning:

Yale Researchers Find “Junk DNA” May Have Triggered Key Evolutionary Changes in the Thumb and Foot.

New Haven, Conn. — Out of the 3 billion genetic letters that spell out the human genome, Yale scientists have found a handful that may have contributed to the evolutionary changes in human limbs that enabled us to manipulate tools and walk upright.Results from a comparative analysis of the human, chimpanzee, rhesus macaque and other genomes reported in the journal Science suggest our evolution may have been driven not only by sequence changes in genes, but by changes in areas of the genome once thought of as “junk DNA.”

So here’s a fascinating observation about genome evolution, and yet they feel compelled to – once again – peg the story on the discovery that there is actually something going on in “areas of the genome once though of as ‘junk DNA'”.

Of course Noonan and colleagues know better. They work on non-coding DNA precisely because they know it is NOT junk. So why, when it’s time to make a pitch to the local press officer, do they fall back on this old bromide? It obviously appeals to writers – who love it when they can pitch a story as overturning orthodoxy. It seems minor, but pegging it this way leads to some really attrocious misrepresentations of current biological knowledge.

Here are some headlines on news stories that followed the press release:

Who Says It’s ‘Junk DNA’? (Hartford Cuorant)

Enjoy Your Opposable Thumb? Thank your “Junk DNA” (Discover Magazine)

and my personal favorite

Meaningless Genetic Code Helped Form Human Hands (Telegraph)

Why is this such a problem? Well, first it’s just WRONG. We’ve known almost since the dawn of the DNA age that not all DNA is protein-coding, and that there are essential functions encoded in non protein-coding DNA. Unfortunately, for initially practical reasons, a disproportionate amount (surely in excess of 90%) of research has focused on protein-coding genes, fostering the faulty impression – amongst scientists as well as science writers – that the ~3% of the human genome that is protein-coding contains > 90% of the function. And it would be great if scientists who, because they work on non-coding DNA are particularly aware that this view is incorrect, would stop promoting it in the popular press.

A second, and less obvious, problem is that this view has played into the hands of the intelligent design crowd. For reasons that baffle me, smart scientists continue to cite the “fact” that much of the human genome is non-functional as evidence against intelligent design. And every time a new study comes out reporting that “junk DNA” is not junk, the ID’ers jump on it as validation of the predictions of ID. It’s hooey of course, but we needn’t give them the opportunity.

So, I am making it my mission to shame everyone who uses the term “junk DNA” or its equivalent when talking about new research on the function of non-coding DNA. And every year I’m going to give out “JUNKY” awards to the most egregious examples. Send me your candidates!

This entry was posted in NOT junk. Bookmark the permalink. Both comments and trackbacks are currently closed.


  1. Posted September 15, 2008 at 11:10 pm | Permalink

    Is it your position that 90% of the human genome is not junk? If so, how much of it – if any – do you think is junk? It seems that several scientists still hold to the idea that most DNA is junk.

  2. Posted September 16, 2008 at 6:37 am | Permalink

    You’re not the first blogger to throw your hat into the ring on this one. Have you seen what Ryan Gregory has written on the topic (junk DNA search results).

    In the end, it’s a double-edged sword. No study that identifies the function of a non-protein-coding element is overturning a long held paradigm that only protein coding sequences have function. But a large fraction of non-coding sequence in eukaryotic genomes are non-functional (especially in bloated mammalian genomes). Outside of people who spend lots of time thinking about genomes, I don’t think this concept has really set in.

  3. Posted September 16, 2008 at 12:19 pm | Permalink

    I have in fact read everything that Ryan Gregory has written about junk DNA in his blog. Let me identify myself as a software engineer. It’s my job to build complex systems and I have been doing that for 26 years. Ryan Gregory and other biologists are trying to understand a complex system — DNA. I am not trying to suggest that my perspective is superior to theirs but that it is different. Things that might surprise a biologist might be expected by an engineer.

    For example, DNA contains a vast amount of repetition. If a computer scientist was examining the binary code of a software program they would expect find a massive amount repetition and I explain why in my blog. Many biologists on the other have dismissed this repetition as (bloat) junk.

    An engineer could selectively delete 10% of the code from most software programs and the program would run exactly the same until it encountered a statistically unlikely scenario. So a software engineer wouldn’t conclude that just because some DNA can be deleted with no apparent ill effect that is has no function.

    The fact that amount of DNA doesn’t always correspond to the complexity of the organism isn’t a surprise either. It’s common place for engineers to add parallel redundancy to a system to enhance performance. Consider the network topology required to support the Google search engine.

    Any why should there be a universal function for non-coding DNA? I expect that we will discover in time hundreds, maybe thousands of functions performed by non-coding DNA.

  4. Shiran Pasternak
    Posted September 17, 2008 at 1:59 am | Permalink

    Great post. But I think anger at these science articles is a bit misplaced. Most of the educated [non-scientist] population has come to embrace the notion that most of the human genome is “junk DNA” over the past few decades. But coverage of research on non-coding DNA is in fact helping to turn the tide and set the record straight. I do agree that they need to be phrased more carefully, but a headline like Who Says It’s “Junk DNA” and a sentence like areas of the genome once thought of as “junk DNA” help in re-educating lay readers.

  5. Posted September 30, 2008 at 11:06 am | Permalink

    The amazing thing about most of the reporting and much of the scientific commentary on the “junk DNA” question is that it ignores the original observations which provoked the suspicion that a lot of the DNA was junk. The “junk DNA” idea did not come from deletion experiments or the observation that a lot of DNA was “non-coding” (for proteins), it came from (1) the raw observation that genome size varies by *orders of magnitude* among organisms of about the same complexity (e.g., multicellular eukaryotes, vertebrates, vascular plants, etc.), and (2) the fact that the difference in genome size appears to be mostly due to differences in copy number of repetitive elements. (and (3) if the genome has too many functional loci then the chance of a fatal mutation becomes impossibly high).

    No one (who knew their stuff) ever said that *all* non-protein-coding DNA was junk, and there has always been the idea that some of it serves regulatory functions etc. But Facts #1-3 remain true, and the discovery of function for this or that little bit of noncoding DNA won’t change them. The fact will remain that some lungfish have genomes dozens of times larger than humans, the human genome is 10 times larger than the zebrafish, despite all of these critters having about the same number of genes. The fact will also remain that basically identical onion species can have genomes varying in size by 5 or times, ferns can have genomes 50+ times bigger than humans, and most of this variation is due to transposons & other self-copying elements & derivatives.

    The major pattern that does exist is that the genome size correlates strongly with cell volume. This could mean that large cells just tolerate more junk, or it could mean that raw bulk DNA serves some kind of bulk generic function which is most sequence-independent. What it does *not* point to is any kind of hidden “code” or other sequence-specific regulatory function analogous to genes or regulatory elements.

    So perhaps “junk” is a problematic or at least unproven term (Ryan Gregory argues this), but it’s a lot more probable than the naive adaptationist claim that all DNA has some kind of sexy sequence-specific function.

    I think the problem is that a lot of these facts were acquired back in the pre-sequencing days and so a lot of scientists, science journalists, etc. never learn about them.

    (PS Michael — I’m a grad student at Berkeley with John Huelsenbeck, before I came to Cal I worked at the National Center for Science Education fighting the “intelligent design”/creationists. I hope to meet you at some point!)

  6. Tania
    Posted April 27, 2009 at 10:02 pm | Permalink

    Hello Michael! I go to school at UC Irvine, and am studying evolutionary biology right now. Other than because of cell volume, why do you think complex organisms have evolved to have larger genomes, or more transposons? Also, why would ferns for example have a larger genome than humans? Could you please point me in the right direction?

  7. Martin J Sallberg
    Posted November 10, 2011 at 4:19 am | Permalink

    The claim that most DNA must be non-functional because there would otherwise be too many harmful mutations ignores the fact that changing the regulation of gene activity really is far less likely to be fatal than changing the protein manufacturing itself. Quantitative changes in how much of a protein that is produced at each time is likely to give subtle phenotypic changes that are advantageous in some situations but disadvantageous in other situations, but there is almost no way a single regulator mutation could be fatal. A single protein coding mutation has a orders of magnitude more likely to be fatal, so regulating DNA can be allowed to grow orders of magnitude larger than protein-coding DNA. There is functional DNA and there is functional DNA.Adaptionist theory has been empirically vindicated by the discovery that fully half the total DNA difference between humans and chimpanzees are regulatory of gene activity in the brain alone. It is quite likely that numbers of repetitions have regulatory roles, just like numbers of holes had regulatory functions in industry robots controlled by old-fashioned hole card computers. That genome complexity need not necessarily be proportional to organism complexity are much better explained by emergence, that the whole organs are more complex than the sum of the genomes that coded them and the genomes that regulated them. No genome complexity of any organism come anywhere near the organ complexity of a human brain. Neutralists are really guilty of creating a cultural atmosphere where ID nuts think they have won when non-coding DNA turns out to have regulatory functions. The arguments for evolution should, for the sake of scientificness, focus on the necessity of modification of existing structures instead of parsimonous family trees and naive faith in the absolute Newtonian molecular time neutralists seem to believe in.

One Trackback

  • […] post that lets me comment on these issues all at the same time is: Scientists Cynical use of “Junk DNA” at Michael Eisens blog (I know the post is rather old, but it is new to me). Coincidentally […]