Friday, November 08, 2013

Science Journal Blows It Again

This week's issue of Science contains three separate papers analyzing transcription factor binding sites and chromatin modification sites in the genomes of different individuals. If most of these sites are spurious sites that just happen to contain a consensus sequence, then you would expect a lot of variability since the sites are mostly in junk DNA where the sequences make no difference. That's what all three papers found but, of course, they interpret this to mean that the regulatory sites must be responsible for the variation between individuals.

The papers were summarized in the form of a "press release" called a "Perspective." The complete citation is ...
Furey, T.S. and Sethupathy, P. (2013) Genetics Driving Epigenetics. Science 342:705-706. [doi: 10.1126/science.1246755]
These authors are affiliated with several departments at the University of North Carolina in Chapel Hill but, most significantly, they are part of the Carolina Center for Genome Sciences. This strongly suggests that they know something about genomes.

That makes it very difficult to explain the following quotation from their "Perspective" article.
One of the most important discoveries in genetics in the last 10 years is that the vast majority of trait-associated DNA variations occur in regions of the genome that were once labeled as “junk DNA” because they do not code for proteins. We now know that these regions harbor genetic elements that control where, when, and to what extent specific genes are expressed to make functional RNA and protein products. Therefore, most trait-associated DNA variants are thought to alter not the gene itself, but rather, the regulatory elements that control the process of gene expression.
The first thing that jumps out at you is repetition of the silly myth that noncoding DNA is equivalent to junk. Reputable scientists working in a genome group at a reputable university should know better that to propagate such nonsense. I've said it before and I'll keep saying it until it sinks in through very thick skulls: no knowledgeable scientist ever said that all noncoding DNA is junk [Stop Using the Term "Noncoding DNA:" It Doesn't Mean What You Think It Means].

The editors at Science have been alerted to this problem many times. They owe it to the scientific community to stop publishing articles from amateurs who don't understand the junk DNA controversy. Enough is enough.

The second thing that annoys me—and several Sandwalk readers—is the idea that regulatory sequences have just been discovered or that we have just learned that they can actually affect phenotypes by altering the timing and/or location of gene expression. We've known that for decades. The entire field of modern developmental biology is built on the idea that mutations in regulatory sequences can have profound effects on the visible phenotype of an organism. All reputable scientists should know of this work and they should be intimately familiar with bithorax and antennapedia. Books have been written on the subject1 and Nobel Prizes have been awarded. Events such as the Cambrian explosion and other radiations have been explained on the basis of small changes in the regulation of gene expression.

Almost everyone who studied molecular biology expected that the situation in humans would be no different. Phenotypic variation in physical characteristics was bound to be due to both mutations in the coding region AND mutations at regulatory sites. It would have been a surprise to find anything else. The papers just published in Science simply confirm what every reasonable person has expected since the 1980s (at least).


1. e.g. "Endless Forms Most Beautiful" by Sean Carroll and "The Plausibility of Life" by Marc Kirschner and John Gerhart.

21 comments :

  1. As a UNC graduate I find the papers' authors lack of awareness a little embarrassing. In both my introductory genetics and developmental bio class we heavily covered the importance of noncoding DNA and its regulatory capabilities. I certainly hope these authors aren't teaching these courses! Off-topic - our new genomic sciences building is pretty slick looking. The univserity's reserearch greenhouse is located on the roof: https://www.som.com/sites/default/files/uncgenome.jpg

    ReplyDelete
    Replies
    1. It doesn't seem to be an isolated case among the ENCODE-fellows at the UNC or the CCGS. Hold your breath, please:

      Dr. [Jason] Lieb explains, "Everyone has heard the term 'junk DNA.' In many cases, all that really means is that we do not yet understand the function of that DNA. In our project we found that what was previously labeled 'junk DNA' in fact is important for regulating genes. This is something we couldn’t see across the whole genome before because the technology wasn’t available. We and our colleagues coupled two new techniques with sophisticated DNA sequencing technology to test as many different human cell types as possible. We created a high-resolution catalogue of the sites in these cells that are likely to regulate genes, in essence a ‘street level’ view of the cell’s regulatory architecture." (emphasis added)

      T. Furey, by the way, is also an ENCODE fellow, and this explains a lot actually.

      Delete
  2. I'm certainly no top-notch researcher - in fact, I have only done an occasional summer project with students since getting my doctorate in 1998. But my lab in grad school took it for granted that there were regulatory elements in "junk DNA" (to use the Science vernacular) and one of our collaborators, in fact, used such conserved sequences to identify phylogenetic footprints. This was 'old news' ~20 years ago.

    ReplyDelete
  3. @Larry: As a hyper-adaptationist of the natural historian stripe, I find no contradiction between the ‘all-importance’ of adaptation and the fact that the great majority of DNA is Junk. As long as evidence is given weight, the divide is not ideological, but rather one of emphasis. To me, the functional, seemingly adaptive 5% is just more relevant, without denying that the rest also has much relevance.
    Your concern regarding the widespread distortion of the Junk DNA issue among your molecular biology colleagues is well taken. It does smell of ideological blindness, for the purveyors seem clueless of a good deal of basic population genetics and biochemistry, or are faking it. The gratuitous demeaning of Junk DNA by molecular panfunctionalists – gratuitous because the merit of their discoveries is quite unrelated to the functional content of the genome -- has a number of toxic consequences, among which are the muddying of the history of biological discovery and the sabotaging of real conceptual issues such as DNA clocks, C-values, the importance and magnitude of mutational load, and the possibility (unexplored?) that some cell machinery functions to stem harmful effects of Junk DNA. To my mind, ex cathedra pronouncements of the type published by Science border on anti-science.
    Woody Benson

    ReplyDelete
    Replies
    1. To me, the functional, seemingly adaptive 5% is just more relevant, without denying that the rest also has much relevance.

      I understand that position. You are entitled to your opinion as long as you have a good solid understanding of the parts or evolution that don't appeal to you. You should also be tolerant and understanding of those who don't share your opinion. Finally, you should make sure that you understand the null hypothesis whenever you confront a problem in explaining how something evolved.

      To my mind, ex cathedra pronouncements of the type published by Science border on anti-science.

      Exactly. This sort of thing should not be published in a peer-reviewed journal that most people think of as one of the top two in all of science.

      Delete
    2. Larry, One of my comments did not appear on "ASBMB Core Concepts in Biochemistry and Molecular Biology: Matter and Energy Transformation"? Was it too flimsy?

      Delete
  4. It would be good for you to write a letter to the editor to be posted in Science about this. Also, you might recruit some of your colleagues to also write letters to give this matter a broader front.

    ReplyDelete
  5. I left a coment last friday at Science regarding this paper. It wasn't approved...

    ReplyDelete
  6. Furey & Sethupathy refer to methylation as a "layer of gene regulation." Back when I used to work on introns, I used to get annoyed at how splicing was described as a "level of regulation" of gene expression (e.g., an extra level of regulation that prokaryotes lack).

    Apparently, when the evidence shows that X affects Y, biologists feel comfortable describing this by saying that X "regulates" or "controls" Y. I see this all over. In fact, I already saw it twice today and I've only been in the office for just over an hour. Here's the other example:

    “Phenotypic differences between individuals cannot entirely be explained by genetic differences. Considering the transcriptome as a mirror of the sum of regulatory events suggests that nongenetic mechanisms have a profound influence on the phenotype. Epigenetics is responsible for part of this additional layer of control” Heyn, H., et al., DNA methylation contributes to natural human variation. Genome research, 2013. 23(9): p. 1363-72.

    I have a suggestion for a collaborative project to confront this sloppy adaptationist mode of thinking. Let's collect a large number of quotations illustrating the "regulate" meme, wherever it occurs (most common in molecular biology, it seems), then write something up for publication (e.g., BioEssays, following on an editorial calling for a new language of evolutionary biology http://onlinelibrary.wiley.com/doi/10.1002/bies.201190011/abstract).

    We could collect these in a google doc and keep a shared bibliography. Crowd-sourcing this to hundreds of people who are reading papers would make it easy to collect lots of examples.

    Arlin

    ReplyDelete
    Replies
    1. The heart of the problem is how you define "regulate." Just the other day I was having a discussion with a colleague who teaches students that the cleavage of trypsinogen to produce trypsin is a form of regulation. I pointed out that regulation generally refers to something that could be called "modulation." In other words, in order for it to count as regulation is has to be controlled. There must be times when the event occurs and times when it doesn't. In most cases of regulation, the effect is reversible.

      If a gene has introns that are spliced out the same way every time then this is NOT regulation. If a restriction site is methylated all the time to protect it from cleavage then this is NOT regulation. If an enzyme is always bound to ATP then this is NOT regulation.

      I like the article you linked to and I agree that we should have more examples.

      Delete
  7. Someone explain it to me. Because the phrase from the article

    "he vast majority of trait-associated DNA variations occur in regions of the genome that were once labeled as “junk DNA” because they do not code for proteins." (my emphasis)

    doesn't seem to say

    "The first thing that jumps out at you is repetition of the silly myth that noncoding DNA is equivalent to junk."

    Now maybe you are referring to something else in the perspective article, but I don't see the relationship between what the article and the claim.

    ReplyDelete
    Replies
    1. I'm not sure I understood your post correctly. Here's what the article says:


      """One of the most important discoveries in genetics in the last 10 years is that the vast majority of trait-associated DNA variations occur in regions of the genome that were once labeled as “junk DNA” because they do not code for proteins. We now know that these regions harbor genetic elements that control where, when, and to what extent specific genes are expressed to make functional RNA and protein products."""


      What the article implies is that at some point in the past (10 years or so according to them) DNA that was non-coding was considered to be junk DNA. This isn't true. No one ever claimed that non-coding DNA was junk DNA. We have known about structural RNA for many decades (rRNA, tRNA). We have known about regulatory sequences for many decades. Nobel prizes were won for some of these discoveries. All of these are non-coding DNA and none of these were ever claimed to be junk-DNA.

      Delete
    2. OK, I think I get it. Thanks.

      My one comment is that the vast majority of non-scientists do think that non-coding DNA is junk and have for a long time. While scientists may not have claimed it to be true... the state of reality is that many, many people do think it's true.

      That's unfortunate and I think that the perspective article here may help non-scientists come to realize that it isn't true.

      As an aside, we're (and I mean that very loosely) are so busy arguing over the minutia of science, that people are losing sight of the fact that 90% of the population of the US (a made up statistic) have no idea what the argument is about, can't understand it, and don't care.

      I don't know. On the one hand, I want correct science to be portrayed to the public at all times. On the other hand, I don't think that arguing about things that are past is a good way to go. It's really about public perception and I hate that because the public doesn't see things the same way that specialists see things (and yet, non-specialists often claim to be experts).

      I'm willing to give the author of the perspective article the benefit of the doubt here. Of course, I'm speaking from a non-biochem perspective.

      Delete
    3. ogremkv,

      the authors of this article are professional scientists writing for scientists. This kind of mistake is unjustifiable. These people are showing that they have no idea what they are talking about. This is very serious.

      Delete
    4. This being an article published in Science, it's hard to argue that the authors are referring to popular misconceptions rather than scholarly consensus.

      Delete
    5. ogremkv:

      You're welcome to ask questions and clarifications. You'll find that people here at this blog will be happy to to clarify your doubts.

      Delete
  8. Well, looks like my comment and Moran's comment were finaly approved. For a moment I thought Science was keeping them in limbo. That's a start.

    ReplyDelete
  9. I'm on the page with the full text of the article, but can't read the comments. The link "Read Full Text to Comment (2)" just takes me back to the same page. I wonder what's wrong.

    ReplyDelete