Thursday, July 28, 2011

The DNA Hall of Shame



Kalliopi Monoyios is a science illustrator who works at the University of Chicago. She's the artist behind Your Inner Fish and Why Evolution Is true. She has a blog called Symbiartic where she writes about the trials and tribulations of working with scientists making scientific illustrations.

I was especially interested in her recent article about The DNA Hall of Shame.
Confession time. Illustrators are people, too. And by that I mean they bring assumptions to the table at the outset of every project. There’s no avoiding it – no matter how educated and experienced you are, you can’t know it all. That is why it is so critical for researchers and editors to be intimately involved in every draft of the drawings they commission and publish. This may sound like a mega no-brainer, especially if you’re an editor or art director in a field other than in the sciences who is accustomed to working intimately with illustrators to get what you want. But in my experience illustrating two popular non-fiction science books, illustrations are treated as icing on the cake and are glossed over by fact checkers and editors who otherwise comb manuscripts for errors. Illustrating Your Inner Fish is a prime example. The manuscript went through four drafts of revisions with at least two specialized scientific editors. And yet this gaffe made it through:
I know exactly how she feels but even multiple expert reviewers won't save you. When we're writing a textbook we take care to review the text and figures together and we pay just as much attention to the figures as we do to the text. For the most recent edition—about to be sent to the printer—the pages were reviewed by me, my coauthor Marc Perry, our developmental editor Michael Sypes (University of Arizona), content reviewer Barry Ganong (Mansfield University), and accuracy reviewers Scott Lefler (Arizona State University) & Kathleen Nolta (University of Michigan). That's six pairs of expert eyes that look at every page.

All the previous editions went through the same intensive review. Nevertheless, the following incorrect figure—originally drawn by me back in 1992— has been published in four books. It wasn't until Barry Ganong looked at it a few months ago that he recognized the error. Can you spot it? It's subtle, and it may only be apparent to experts on DNA structure, but it's an error nevertheless.1

I add it to The DNA Hall of Shame.



1. I'll post the correct version in a few days.

19 comments :

  1. Each of the DNA chains to which the nucleotides are bonded is identified as "5' DNA", but they should be "3' DNA"?

    ReplyDelete
  2. So, silly me, I missed the handedness thing, probably because I’m not a biochemist. I thought the error was more obvious: the thing labelled “chromosome” is X-shaped. This is a common depiction of a chromosome, but it’s biologically bizarre. The only time genetic material looks like that is during mitosis, and specifically before anaphase.

    We might say, “oh, well it’s not inaccurate, it’s just a little too specific, but it’s only a cartoon so chill.” And if so, I’d say the same thing about the handedness of the cartoon DNA molecule.

    (I left this comment at Symbiaritc as well.)

    ReplyDelete
  3. The illustrated DNA has the wrong kind of twist; the correct twist should be the mirror image of what is shown.

    ReplyDelete
  4. Is the error that the the way you've arranged the series of nucleotides, so that they would form a left hand helix instead of the usual right hand one?

    ReplyDelete
  5. Dale - I don't think that's the problem. All DNA and RNA chains have a 5' end and a 3' end. The 3' end is labeled in the template strand in Dr. Moran's image. The ends appear to be correctly labeled to me, and the growth of the chain is shown in the correct direction, as far as I can tell. I cannot tell what the error is, however. I'm very interested to find out.

    ReplyDelete
  6. Steve - the "metaphase chromosome" (which is what is illustrated) also appears during meiosis.

    ReplyDelete
  7. Z DNA is left-handed, but the illustration doesn't indicate that this is Z DNA, and so it should be right-handed.

    ReplyDelete
  8. jaxkayaker, yes I know. In both cases it's not "a chromosome" but a matched pair (of chromatids, to be exact) linked together. It's worse during meiosis I, when it's two pairs linked together. But it's not a picture of "a chromosome" any more than the DNA molecule is a picture of "a strand."

    ReplyDelete
  9. Arrrrgggghhhh!!!! Thou hast inverted the stereochemistry on the nucleotides in the template strand! The 1' carbon in the strand on the left is R configuration, while the 1' carbon in the strand on the right is S configuration! May the saints have mercy on thee!!

    ReplyDelete
  10. Yes, it is pretty subtle. I think that the template strand shows alpha glucoside bond instead of beta. This is *nothing* in comparison to the ubiquitousness of the left-handed DNA! Sometime it feels that 50% of the book covers that have DNA show it left-handed. E.g., both editions of The Making Of The Fittest.

    ReplyDelete
  11. Very subtle, but I know some chemists who would cringe when they see these things. I wasn't sure if we should avoid spoiling it for others, so the answer is in ROT-13:
    Gur grzcyngr fgenaq vf gur jebat ranagvbzre, nf fubja vg'f Y-QAN.

    ReplyDelete
  12. Steve - I was addressing your claim that "The only time genetic material looks like that is during mitosis, and specifically before anaphase." Genetic material also appears that way during meiosis, thus mitosis is not the ONLY time genetic material looks like that. It also appears X-shaped at metaphase I and metaphase II of meiosis.

    Secondarily, a metaphase chromosome is still called a chromosome, even though it's comprised of a duplicate pair of chromatids, each a chromosome in its own right once they separate at anaphase. This is something of a convention for counting purposes. It's a diagrammatic representation (not a picture) of a metaphase chromosome.

    In addition to the handedness problem, the diagram of the DNA strand is somewhat misleading and not a perfect representation in that a DNA strand isn't a spiral in quite the manner depicted. Each turn has a major groove and a minor groove.

    ReplyDelete
  13. Haha, that is great. Whenever I see pictures of DNA, I check the handedness and whether the major and minor groove are present. It is amazing how many pictures get it wrong, even on genetics textbooks.

    ReplyDelete
  14. I am not sure but I would use the term template strand rather in the context of transcription because in DNA both strands act as templates for newly synthesized strands.

    ReplyDelete
  15. jaxkayaker, I wasn't very clear in my description. If you look at the new strand the label: "5' DNA" can be interpreted as meaning the 5' carbon of the deoxypentose to which the phosphate is bonded. But that phosphate is, in reality, bonded to the 3' carbon. I think the label is ambiguous -- it's meant to indicate that the 5' end of the nascent polynucleotide in that direction, but it could be interpreted in the way I've indicated.

    ReplyDelete
  16. Is it that the oxygen in the chain stage right should be blue? Does I can't remember how the dehydration synthesis works here - which oxygen ends up in the water? None of my intro textbooks in the house go into this level of detail.

    ReplyDelete
  17. Audience is a consideration and whether the errors actually lead to mis-conceptions. Based on the suggested errors I would say that diagram 1 does the job and provides an outline of the organisation of the genetic material within a cell. The intention is not to invite detailed consideration of the handedness or groove structure of DNA and therefore there are no misconceptions. Be honest you must have flicked passed diagrams just like this a thousand times and never stopped to use this has a source to develop your understanding of handedness or grooves in DNA. To the novice the error is not significant, to the expert the error is of no consequence.

    The second diagram on the other hand clearly invites a more detailed consideration and the errors cannot be so readily justified. The student is required to inspect the structure, an invites consideration of the bases phosphate groups,sugars, h-bonding and the rest. This is not a diagram for a passing glance. If there is an error (and I don't see to many suggestions here) then this could cause real problems for the student in a formative learning process and as such any errors do create misconceptions.

    ReplyDelete
  18. I actually find that putting up a question (worth points) that is "wrong" in some way is one of the best way to generate discussion in classrooms that won't talk. For instance asking a muliple choice question with 2 right answers... someone in the room always picks up on it and cries foul, at which point I fake embarrasment and break them up into groups and have them re-write the question. The embarrasment was very real when I stumbled on this tool the first time, but quickly saw how the students' level of discussion went up a notch to figure it out. I have not shown pictures with mistakes, but I think I am going to start doing so.

    ReplyDelete
  19. Having just learned a little more about DNA structure, I thought I'd take another look at this. I can see two things that I thought might be incorrect:

    1. The new strand in this diagram doesn't seem to be forming 5' to 3', but 3' to 5'.

    2. I think both the oxygens in the Phosphodiester bond comes from the dNTP.

    I might be wrong (especially on the 2nd), so feedback would be helpful!

    ReplyDelete