{"id":393,"date":"2009-06-29T22:52:42","date_gmt":"2009-06-29T22:52:42","guid":{"rendered":"http:\/\/occamstypewriter.org\/scurry\/2009\/06\/29\/picture_perfect\/"},"modified":"2010-12-26T22:29:47","modified_gmt":"2010-12-26T22:29:47","slug":"picture_perfect","status":"publish","type":"post","link":"https:\/\/occamstypewriter.org\/scurry\/2009\/06\/29\/picture_perfect\/","title":{"rendered":"Picture perfect?"},"content":{"rendered":"

I love pictures. And photographs. Unlike some<\/a>, I can’t paint or draw with any great skill, but in common with many other people I get immense pleasure from photography. I’m no expert and wouldn’t even consider myself a serious hobbyist, but I have been known to enjoy composing and capturing an image with my little compact camera.<\/p>\n

Here’s one I took the weekend before last when we were at Brighton. I now have it as the desktop picture on my computer and it has been making me smile ever since. I like the elemental quality of the scene: earth, water and sky and my children caught in a moment of simple enjoyment.<\/p>\n

\"Brighton<\/a><\/p>\n

Pictures are also important in my work. But the pleasure of their creation is sometimes mixed with pain. Lots of pain.<\/p>\n

In the past few days I have been working on a new paper which will describe the structure of a protein-peptide complex that we have solved in my lab. It’s been tough enough grinding out the words to describe our results, but coming up with the figures has been about a thousand times harder. Hence the well known epigram, I suppose.<\/p>\n

Although I have images of the structure in my mind when I’m writing, I have to lay out the paper in words first to find out what it is I want to say, to identify those parts of the structure that are likely to be the most meaningful for the reader. Once the draft manuscript is complete I can list the figures I’d like to include.<\/p>\n

But this is when the trouble starts.<\/p>\n

Often the list is quite long and has more images than can be accommodated in the finished manuscript. (Colour figures are not cheap, so cost is a major consideration.) But how to condense all that information, without obscuring the message? The meaning of a good figure should leap off the page; ideally the reader shouldn’t even have to consult the legend to appreciate the central message. For structural biologists the principal difficulty is trying to convey detailed three-dimensional information on a two-dimensional page. It requires real artistry, and perhaps that is the solution?<\/p>\n

In the early part of the 20th century Picasso and Braque pioneered cubism, a school of painting that tried to be a way of depicting an object as seen from different viewpoints independently<\/a>. In other words, the artist depicts the subject from a multitude of viewpoints to represent the subject in a greater context<\/a>. Here’s an example:<\/p>\n

\"File:Le<\/p>\n

In the words of the celebrated Australian artist, Rolf Harris, “Can you tell what is it yet?” Fortunately this image does come with a figure legend. It reads, “Le guitariste.”<\/p>\n

But of course!<\/p>\n

I don’t mean to mock: Picasso was clearly a very considerable artist (though I don’t go for everything that he produced). But clearly the aim of the cubist is different to that of the crystallographer*. We work on very different planes of understanding. The idea of being able to present different perspectives in a single image is appealing, though in my work as a scientist I’m looking for literal\u2014rather than conceptual\u2014representations of reality. I guess my problems are a bit more prosaic, but I do still struggle for my art.<\/p>\n

Let me show you what I mean. Here’s an image showing the superposition of the protein structure before (grey) and after (purple) binding of the peptide (orange):<\/p>\n

\"1-CC\"<\/a>
\n_All structural figures made with Warren Delano’s fantastic PyMOL<\/a><\/p>\n

The image is schematic since, rather than depicting all the atoms of the molecule, you see a winding worm tracing the fold of the polypeptide backbone. But I’ve also shown (in sticks) some of the side-chains of the protein that change their conformations when the peptide binds. The aim here was to show global and local conformational changes that resulted from peptide binding.<\/p>\n

The second figure below was an effort to show some of the bonding pattern that locks the peptide and protein together:<\/p>\n

\"2-HB\"<\/a><\/p>\n

The green dashed lines show hydrogen bonds formed between the backbones of the peptide and the protein. There are nine of them in total and they help to zip the peptide into place. The trick here was to find a viewpoint that allowed all nine to be seen. Also, I didn’t want to deviate too much from the viewpoint of the previous figure. I didn’t want to lose the viewer.<\/p>\n

But I wasn’t happy with these images – something gnawed at me overnight. It struck me that the two views weren’t so dissimilar; perhaps they could be combined into a single figure? I thought about how to do that and came up with this:<\/p>\n

\"3-MC\"<\/a><\/p>\n

Here instead of wormy tracery I have depicted the bonds connecting the backbone atoms as sticks. But the zig-zag path of the backbone is untidy – it clutters the view and obscures our perception of the side-chains. This wasn’t the answer.<\/p>\n

So finally I thought to simplify still further. I opted to return to the smooth backbone trace, even though it doesn’t show all the atoms involved in hydrogen bonding to the peptide backbone, because, this doesn’t really matter<\/em>. The key point is to show that the bonds are distributed along almost the entire length of the peptide (and most<\/em> readers will already know that proteins are capable of these interactions).<\/p>\n

\"4-CC-HB\"<\/a><\/p>\n

Sharp-eyed readers may have noticed that I also zoomed in a little – to capture as much detail as possible while still showing enough of the overall structure to preserve the context of the peptide binding site. This is better but I’m still not wholly satisfied with the result. I’m not sure about sticking with the green hydrogen bonds; maybe I should switch to purple or lilac, so that it is clear that the peptide is bound to the purple, rather than to the grey structure?<\/p>\n

This tortuous process occurred over a period of three days. And these weren’t the only figures that I have been wrestling with. Each image demanded a similar level of attention, scrutiny, effort and engendered immense frustration. And in every case there was the wretched struggle to twist and turn the viewpoint so that all the important features remain in view. Like Picasso and Braque, what I really needed was to find a way to show different perspectives in the same image. And I think I’ve found an answer, though it can’t yet be implemented on paper, alas.<\/p>\n

What I really need are moving pictures.<\/p>\n