How to spot a badly-drawn DNA helix

By Grant Jacobs 22/07/2013 21


The DNA double helix is one of the icons of our time. You’d wish people would draw it right.

You’d think getting it wrong would make the artist feel like a complete mug, as if they’d drawn the Statue of Liberty holding up a lump of coal instead of a flaming torch.[1]

But the DNA helix in all sorts of places is simply wrong.

You don’t need to be a TV drama-smart science geek to know which artists have screwed up.

Let me show you an easy way. (And impress your non-scientist friends!)

Does it turn the right way?

Look at the DNA with the ‘rungs’ of the ladder horizontal and the helix standing vertically.

LH-w-arrow-80pxRH-w-arrow-80pxLook for the diagonal ‘stripes’ that make up the spiral of the helix, as they cross the front of the helix.

Apply this rule:

From the right downwards, means the helix is right-handed. (Shown to the right.)

From the left downwards, means the helix is left-handedwhich is wrong. (Shown to left.)

(Modified from original image at Georgia Perimeter College webpage. Some mobile phone users with smaller screens may see the left image above the right image.)

So, this is right-handed,

right-DNA-red-blue-ID-10061973
By jscreationzs, published on 22 October 2011; from freedigitalphotos.net, ID: 10061973.

but these are left-handed,

leftie-DNA-brown-ID-100105102
By dream designs, published on 04 October 2012; from freedigitalphotos.net, ID: 100105102.
leftie-DNA-light-blue-ID-10079201
By dream designs, 09 April 2012; freedigitalphotos.net, ID: 10079201.

Flip it over

One way to correct a left-handed DNA cartoon is simply to flip the picture over to it’s left-right mirror image.

No excuses for artists not to put their errors right – that’s click-of-a-button stuff.

A plague of wrongness

Leftie DNA is all over the place, especially in commercial art sites.

leftie-DNA-Obj-Stock-Photo-300px

Unfortunately most commercial artists insist that you pay for their work, even tiny images on-line. So we’ll have to make do with links.[2] The free one to the right is from photosxc.com. Here are more:

ImageEnvision

123RF

VectorJunky

GraphicRiver

I could go on, but you get the picture.

As a more complex example, this pair of chromatids (ID 275311, turbosquid.com) has a mixture of left- and right-handed DNA that the artist has likely created by taking mirror-image of an original.

As I remarked to a reader, I was “a bit disappointed at the DNA” in the logo for Watson and Crick. (That’s not the DNA model they’re famous for!)

It’s even wrong on a web page that aims to show people how to draw the DNA double helix![3]

Some of these illustrations are wrong in other ways too.

To make things confusing you can get left-handed DNA, it’s just not the stuff our genes are made of.[4] There’s a slightly fancier way of telling what is left- or right-handed too.

Let’s try a little more detail.

A little more about DNA

Here’s a model of DNA:

DNA_animation
Source: Wikipedia; public domain by ‘brian0918’.

This model is made from data giving the position[5] of the atoms in a crystal of DNA using X-ray crystallography – one way to take a ‘picture’ of molecule. It’s the real deal, in detail; we can see where every atom is.

Source: UCLA Chemistry Illustrated Glossary of Organic Chemistry.
Source: UCLA Chemistry Illustrated Glossary of Organic Chemistry.

There are several forms of DNA. The best-known are A, B and Z.

The ‘classical’ form, the one that most of DNA our genes are made of is in, is B-form DNA like that in the animated model above and the illustration to the left.

You can see that DNA is made of two helical chains with the DNA bases in the middle.

(The phosphates of the deoxy-ribose sugar-phosphate backbone are orange and red in the rotating model above and black in the illustration to the left.)

B-form DNA has two grooves of different sizes: the major and minor grooves, as marked in the illustration to the left.[6]

Reminds you of keys in music, right?

The major groove is the wider ‘valley’ between the phosphate backbones.

The minor groove is the narrow one.

Source: Wikimedia commons, by Leyo, public domain.
Wikimedia commons, by Leyo, public domain.

Other common mistakes

Making B-form DNA left-handed isn’t the only common error you’ll see.

No major / minor grooves (or, grooves the same size)

Some illustrators draw DNA with two equal-sized grooves, like in the example to the right.

Too many or too few bases per helical turn

B-form DNA typically has 10 base pairs per full helical turn.[7]

There’s a wee trick to this.

If you watch the rotating model shown earlier you’ll see that one base pair gets hidden behind the point where the two helical backbones cross.

So, looked at the right way, for each half helical turn you’ll see four base pairs side on, with the fifth obscured.

Count carefully!

Too skinny or too fat

Very roughly, the length of a full helical turn should be about one and a half times the width of the double helix.

alien-DNA
By renjith krishnan, published on 04 July 2012; Stock Image – image ID: 10090099.

(Alternatively, the width should be about two-thirds of the height of a full helical turn. B-form DNA is roughly 2 nanometers wide for each roughly 3.4 nanometer turn of the double helix.)

So that ‘DNA’ to the right looks like alien DNA, right?

Left-handed – check.

Far too many base pairs per turn – check.

No major / minor grooves – check.

Way too svelte – check!

Even in scientific journals…

cover-mysterious-epigenome-259pxYou might forgive creationists for getting the structure of DNA wrong, like on the cover of a creationist take on epigenetics to the left. Having gotten much else of the science wrong it’s not so startling to see the DNA helix wrong in that context.

Embarrassingly there are also examples of left-handed DNA representations in scientific journals. Almost all of these are works of illustrators. Nevertheless, that’s not really much excuse.

There’s good a comment by scientific illustrator Kalliopi Monoyios in her article, The DNA Hall of Shame, on her blog, Symbiartic, at Scientific American blogs that images may not get the same scrutiny as text despite that they probably should. Looking at the examples in Tom Schneider’s pages you’d have to agree.

These goofs aren’t limited to the ‘lesser’ journals either. (In fact, I wonder if a contributing reason that they occur in the more prominent journals is simply that these publications are more able to afford to outsource illustration work – ?)

Cover-Nature-Methods-10-7-130pxThis month (July 2013) Nature Methods’ cover shows the work of a freelance designer, featuring a stylised blue-and-gold left-handed helix. (The illustration is highlighting an article, Simultaneous DNA amplification and detection using a pH-sensing semiconductor system, describing the development of a DNA-detecting chip – “an integrated chip for real-time amplification and detection of nucleic acid using pH-sensing complementary metal-oxide semiconductor (CMOS) technology”.)

PCB-detail-June-2013-129pxAnother example is in PLoS Computational Biology, in an illustration intended to be a “representation of different elements related to bioinformatics” for the cover of the June 2013 issue.

(Detail shown to left; full image below, taken from the cover information page.)

Cover-PCB-June-2013-640px

cover-nature-genetics-sept-2005-250pxYet another example is in an older Nature Genetics cover (September 2005). It’s a lovely illustration, but the interwound ribbons representing DNA are left-handed.

Tip for illustrators

It’s perhaps best to get a hold a model of real DNA (ideally from the crystal structure databases) as a template to stylise rather than try roll your own. In addition to getting the handedness right, you’ll also get other aspects of the geometry right this way.

There are many websites giving the basics of the geometry of the structure, like this illustration from Florida International University’s Biology Department DNA and hereditary page.

Footnotes

To see a really nice—we could even say sexy—animated model of DNA, try the solid model at wikimedia. I’d put it up here if it weren’t that it’s 2.5Mb in size.

An excellent source of older examples is in Tom Schneider’s Left Handed DNA Hall of Fame.

This article was inspired by noting that one of the logos for famous scientists—for Watson and Crick—features a DNA helix with a problem. Can you see what the problem is?

logo-watson-crick-251px

(Lest anyone forget, the alpha-helix of proteins is also right-handed!)

1. A political satirist might, but that’s another story.

2. I’ve given the names of the sites rather than the artists; I’m not trying to hit on anyone here, but illustrate that this is widespread.

3. I would like to have used this as the example helix, but I haven’t time to sort out if the illustration is free for others to use.

4. I’m simplifying here, this isn’t a textbook!

5. More accurately, x-ray crystallography gives a map of electron density that guides the modelling of the atomic structure.

6. The major and minor grooves are on opposite sides of the DNA base pairs that form the ladder rungs between the two strands.

7. Actually it’s often slightly more that 10 base pairs per helical turn, but that’s getting fussy for cartoons.


Other articles on Code for life:

Animating our DNA*

Coiling bacterial DNA

Epigenetics and 3-D gene structure

Loops to tie a knot in proteins?


21 Responses to “How to spot a badly-drawn DNA helix”

  • Why yes I can spot the problem Grant. Now.

    I was told that Catholics have left handed DNA. Is this true?

  • Ross – Yes, they did too 🙂 In a way that goes to show there’s not a lot of excuse for getting it wrong, I think.

    I tweeted Discover about their error, but no reply (thus far). (It’s also possible their twitter account doesn’t have a person reading anything sent to it.)

  • To be fair to us designers though even when we do get it right someone goes and makes it wrong. Case in point from sciblogs: the southern genes header image contains a rendering of a 3d model of a DNA strand I made aeons ago. But someone has mirrored the original rendering to make the image longer so one of the strands is now wrong.

  • Hi Michael,

    To my viewing whoever made that banner wanted two separate molecules framing the title (rather than extending a strand). They’ve done that by making a mirror image of the right-hand one, which, as you say, is not the right thing to do.

    As you’ll know, the trick to doing more-or-less what they had intended to is copy the original molecule, rotate the copy by 180˚, then shift it into place.

    I don’t know who made the banner by the way – some aren’t made by the blog authors.

    (Now that your first comment has been approved you should be able to comment immediately; first-time comments are held up for approval as an anti-spam measure.)

  • * Spoiler alert. *

    Michael Edmonds,

    That video is posted by a ‘chakraconsciousness’. (Read it aloud.)

    “Originally on the planet all humans had an active 12-strand carbon-based DNA helix configuration and around the Atlantis time period ten of those twelve strands were deactivated…”

    Righty-ho, if you say so kind of territory I think! I stopped there… (Have to get back to writing, anyway.)

    Thanks for the amusement. There’s some incredibly strange stuff out there.

  • Hi Shoo,

    Good on you for correcting it with good grace.

    Strictly-speaking you also need to have major and minor grooves 😉 But that’s harder to explain and draw from scratch so I have sympathies for not going there, although personally I’d prefer it of course!

  • Just a random tidbit while I’m here.

    There are also DNA structures involving more than two DNA strands.

    A third strand of single-stranded DNA or RNA strand can bind within the major groove of double-stranded DNA using what are known as Hoogsten base pairs to form a three-stranded structure. A similar three-stranded conformation is seen in (the three-dimensional structures of) some RNAs (e.g. tRNAs).

    If readers think that is exotic, there are also four-stranded DNA structures like those found in teleomeres.

    These specialised structures are not the ‘usual’ form of DNA, of course.

  • DNA can be left or right handed, it takes either conformation. A drawing won’t be incorrect just because of the handedness of the molecule.

    As for the structural aspect of DNA, that varies among what conformation is taken. So there isn’t only one ‘correct’ way to draw DNA.

    http://www.ncbi.nlm.nih.gov/books/NBK6545/
    Use ctrl + f to find A-DNA, B-DNA, and Z-DNA

    http://biowiki.ucdavis.edu/Genetics/Unit_I%3A_Genes,_Nucleic_Acids,_Genomes_and_Chromosomes/2%3A_Structures_of_nucleic_acids/B-Form,_A-Form,_Z-Form_of_DNA
    For decent pictures of conformations having different structural patterns.

    Also Grant, telomeres are not quadruple-stranded DNA. They are just the ends of chromosomes with repeating structures. I’m not sure where you get this information from but you should use google to check your facts.

  • Estife,

    I’m well aware of the different structural forms of DNA. (My work as a computational biologist started with computational structural biology and I’ve been in this field for 25+ years, etc.)

    Articles intended for the public, etc., have particular target audiences and it’s wise to bear that in mind when reading them.

    You might have seen that in my case the intended audience of the post was mainly illustrators. They likely last did biology in high school, and probably decades earlier. Besides that there’s little sense in confusing them with complex alternatives, what they are wanting to illustrate is not “some form of DNA”, but a very particular form — the iconic form that the public has come to recognise. That’s B -DNA and it doesn’t exist in a left-handed conformation as you know.

    Also, I did briefly mention the different forms in the post itself. Perhaps you missed that? I’m sure the linked article you pointed to mentions that Z-DNA has a quite different helical structure, characterised by the backbone zig-zagging & that’s where the name of that form comes from. It’s not a flip of B-DNA and, perhaps more importantly given what the post was about, I strongly suspect most members of the public wouldn’t even recognise a cartoon of it as DNA!

    You’re welcome to show where I said there was only one way to draw DNA. (It’s hard to imagine I did, especially as I noted the other forms in passing.) It’s moot anyway, as the article is about drawing the iconic form.

    Finally, I think you are aware I was referring to the four-stranded structure the G quadraplex in my passing mention of four-stranded structures as part of teleomeres. The important bit for who I was writing isn’t the detailed story of teleomeres, but just that there are more structures for DNA again that those mentioned earlier as I think is clear from that comment. (As it happens I know something of the back story to these structure models too, as I did my doctorate in contact with a group that proposed the different quadraplexes and later determined an atomic model for one of them – I can’t recall now, but that may have been the first atomic structure of these quadraplexes.)

    Excuse how long it’s taken me to get to your comment, it was buried amongst the spam & I’ve been busy with other things.

  • Hello Grant,

    I know this is an old post but I just found it because I am working on an aprox. drawing of DNA helix.
    I have no biological studies but I was doing just for working with a drawing tool I was programming.
    My target was producing a customizable double helix son it can be then saved as image for documentation or example etc
    After finishing a first approach I realized that it could have so many DNA mistakes that it worthed checking for them.
    This is how I found your blog and it was my first information that DNA helix is mostly dextrogyre but there can be cases of levogyre.
    The DNA helix is on http://veleta.rosety.com
    Thanks for your article

    • Nice work Luis.

      For what it’s worth, computational biologists build ‘cartoons’ of molecules using molecular graphics software either from the atomic co-ordinates of real DNA molecules, or using “ideal” geometry (a bit like you have done).

      I’m not aware of any web-based tools like yours, though. (Not that I’ve looked in recent years!)

      Now that your first comment is approved, you should be able to comment at will. Hope some of the other articles around at Sciblogs are interesting – have fun!

    • If you’re doing paid-for work as an illustrator, you can do your customer wrong 😉

      We’re not talking about ‘pure’ art here 😉 (You could also argue that if you portray it wrongly, you’d be conveying the wrong concept, say ‘“other” DNA’, or ‘not-real DNA’. ‘Other’ here in the sense that ‘alternative’ medicine is pseudo-science, or in the sense you might use in science fiction – you get the idea.)

      Cheers.

Site Meter