Elinor Hortie at the University of Sydney (Australia) has written a very interesting essay about CRISPR ‘scissors’, a metaphor she find misleading. From Hortie’s July 4, 2019 essay on The Conversation,

Last week I read an article about CRISPR, the latest tool scientists are using to edit DNA. It was a great piece – well researched, beautifully written, factually accurate. It covered some of the amazing projects scientist are working on using CRISPR, like bringing animals back from extinction and curing diseases. It also gave me the heebies, but not for the reason you might expect.

…

Take CRISPR. It’s most often described as a pair of molecular scissors that can be used to modify DNA, the blueprint for life. And when we read that, I think most of us start imagining something like a child with her Lego bricks strewn in front of her, instruction booklet in one hand, scissors in the other. One set of pictograms, one model; one gene, one disease; one snip, one cure. We’re there in a blink. CRISPR seems like it can work miracles.

I want to stress that the molecular scissors metaphor is pretty damn accurate as far as it goes. But in focusing on the relatively simple relationship between CRISPR and DNA, we miss the far more complicated relationship between DNA and the rest of the body. This metaphor ignores an entire ecosystem of moving parts that are crucial for understanding the awe-inspiring, absolutely insane thing scientists are trying to do when they attempt gene editing.

…

Hortie proposes a different metaphor,

In my research I use CRISPR from time to time. To design experiments and interpret results effectively, I need a solid way to conceptualise what it can (and can’t) do. I do not think of CRISPR as molecular scissors.

Instead I imagine a city. The greater metropolis represents the body, the suburbs are organs, the buildings are cells, the people are proteins, and the internet is DNA.

In this metaphor CRISPR is malware. More precisely, CRISPR is malware that can search for any chosen 20-character line of code and corrupt it. This is not a perfect metaphor by any stretch, but it gets me closer to understanding than almost anything else.

Hortie offers an example from her own work demonstrating how a CRISPR ‘malware’ metaphor/analogy more accurately represents the experience of using the gene-editing system,

As an example, let’s look at Alzheimer’s, one of the diseases CRISPR is being touted to cure. The headlines are usually some variation of “CRISPR to correct Alzheimer’s gene!”, and the molecular scissors analogy is never far behind.

It seems reasonable to me that someone could read those words and assume that chopping away the disease-gene with the DNA-shears should be relatively simple. When the cure doesn’t appear within five years, I can understand why that same person would come to ask me why Big Pharma is holding out (this has happened to me more than once).

Now let’s see how it looks using the malware metaphor. The consensus is that Alzheimer’s manifests when a specific protein goes rogue, causing damage to cells and thereby stopping things from working properly inside the brain. It might have a genetic cause, but it’s complicated. In our allegorical city, what would that look like?

I think riots would come close. Rampaging humans (proteins) destroying houses and property (cells), thereby seriously derailing the normal functioning of a specific suburb (the brain).

And you want to fix that with malware?

It’s hard to predict the domino effect

Can you imagine for a second trying to stop soccer hooligans smashing things on the streets of Buenos Aires by corrupting roughly three words in the FIFA by-laws with what’s essentially a jazzed-up command-F function?

I’m not saying it’s not possible – it absolutely is.

But think of all the prior knowledge you need, and all the pieces that have to fall in place for that to work. You’d have to know that the riots are caused by football fans. You’d have to understand which rule was bothering them (heaven help you if it’s more than one), and if that rule causes drama at every game. You’d have to find a 20-character phrase that, when corrupted, would change how the rule was read, rather than just making a trivial typo.

You’d have to know that the relevant footballers have access to the updated rule book, and you’d have to know there were no other regulations making your chosen rule redundant. You’d have to know there aren’t any similar 20-character phrases anywhere on the internet that might get corrupted at the same time (like in the rules for presidential succession say, or in the nuclear warhead codes). Even then you’d still be rolling the dice.

Even if you stop the riots successfully, which of us really know the long-term consequences of changing the World Game forever?

…

That’s stretching the metaphor as Hortie notes herself later in the essay. And, she’s not the only one concerned about metaphors and CRISPR. There’s a December 8, 2017* article by Rebecca Robbins for STAT news which covers ten analogies/metaphors ranked from worst to best,

… Some of these analogies are better than others. To compile the definitive ranking, I sat down with STAT’s senior science writer Sharon Begley, a wordsmith who has herself compared CRISPR to “1,000 monkeys editing a Word document” and the kind of dog “you can train to retrieve everything from Frisbees to slippers to a cold beer.”

Sharon and I evaluated each of the metaphors we found by considering these three questions: Is it creative? Is it clear? And is it accurate? Below, our rankings of CRISPR analogies, ordered from worst to best:

0. A knockout punch
…

9. The hand of God
…

8. A bomb removal squad
…

It’s a very interesting list with a description of why each does and doesn’t work as an analogy. By the way, ‘scissors’ was not the top analogy. The number one spot went to ‘A Swiss army knife’.

There are many more essays than I would have believed concerning CRISPR and metaphors/analogies. I’m glad to see them as the language we use to describe our work and our world helps us understand it and can constrain us in unexpected ways. Critiques such as Hortie’s and the others can help us to refine the language and to recognize its limitations.

h/t July 4, 2019 news item on phys.org

*”December 8, 0217′ corrected to ‘December 8, 2017’ on Jan.20.21