Crispr is less like molecular scissors and more like molecular malware | Hacker News Reader

Crispr is less like molecular scissors and more like molecular malware

thenextweb.com

115 pointswalterbell28 comments

I don't think that the alzheimers consensus is the amyloid beta hypothesis anymore where it's one protein going rogue. Otherwise interesting article, though I disagree with this idea that its "less" like scissors and "more" like malware. They pretty much are molecular scissors, but they're definitely not malware. They're actually the anti-virus system of a bacteria. Calling them malware doesn't make sense as a metaphor (they arent Infcetious etc) and just confuses things more.

Agreed, I think the disconnect here might be that the author has no idea how malware works or what malware really is.

If the author had said that crispr based gene drives are malware then I can see that metaphor working, but I don't think that was what was the case here. Also, crispr for neuro diseases isn't just knockouts, there's work with single base change Cas forms and RNA targeting crispr, it's a broad range of tools

And also, the scissor analogy actually works better with the riot analogy. You arent going to stop a riot with scissors, but you could totally use some form of malware to censor social media/knock out communications/etc to prevent the riot from being able to organize. Malware is a constructive thing, it allows you to hijack systems you take over, which is much much much more powerful than some blunt scissors.

Yeah I don't think the problem is the metaphor for Crispr, the problem is that we don't really understand most of the stuff it's supposed to be able to cure some day. Crispr is at the end of the day just a tool. If we don't know where to "cut" with Crispr, we can't very well use it to do anything. Now that we know we have the tool to cut, in the future it will probably be very useful, but for all the Crispr hype, it was never really going to be a "nice everything is cured now" like some of the media liked to imply, rather that we have another tool in the box to work towards making use of.

You're not thinking broadly enough. Crispr is already very, very useful for biological research. Thinking just in terms of medicine is far too narrow for such a game-changing biological technology.

No offense intended, but maybe you're not thinking broadly enough? It's a breakthrough technology for biological labs now, sure, but a lab using this to get a bunch of papers or grants or patents really doesn't matter for the general populace in the short term until the research has produced some practical applications.

Replacing an inadequately descriptive term with an incorrect term doesn't really help anyone.

There's an important point in the article about the cascade of unforseen consequences that occur when you start tweaking genes using CRISPR - and that it really isn't the straightforward silver bullet many make it out to be.

But the use of the word malware is there exclusively for the clickbaity headline.

I'd argue that the molecular scissors argument is actually adequately descriptive and holds up to a bit of metaphor overextending. You totally can cut and paste gene sequences using CRISPR. That's the whole point of Homology Directed Repair (HDR). And much like actual sets and crafts, it is a lot easier to make crude cuts all over the place than to elegantly glue pieces back together into a useful end product. HDR still suffers from low efficiency and so on. And the riot analogy for alzheimers is great, but it doesn't require you to give up the scissor analogy either. You arent stopping a riot with scissors, while on the other ha d you totally could stop a riot with malware (shut down every messaging service somehow so that there isn't any way to coordinate).

The most important part of CRISPR (Cas9) is its ability to colocalize with specific, arbitrary DNA sequences. It has a second function, which exists is many many other proteins and is not special at all - to cut double-stranded DNA. And in fact, that it cuts, is not actually desirable most of the time. Some of the most interesting uses of Cas9 actually try to get rid of its ability to cut the DNA (while retaining its ability to home in on specific DNA sequences).

I'm all for metaphors, but don't be afraid to shift them around when you think of a better one. (and 'malware' is not a better one...)

Fair point.

It's actually, literally, a molecular antivirus program.

Much like antivirus software scans your computer for a signature of a computer virus in order to remove it, CRISPR is part of a bacterial mechanism to scan DNA for the signature of an actual virus and snip it out, preventing the virus from using the bacterium to reproduce. Biology is NUTS.

Yes, I think a better analogy would be an antivirus program that simply deletes any file (or more accurately parts of a file) it thinks is suspicious.

.. by no means an optimal antivirus program, but in a sense, it is effective.

I think most negative comments here are a bit harsh. Of course the analogy has its weak points (every analogy has), but what I took away from it is the following:

- the DNA is not the Internet, but rather the Code (capital C) that governs everything in the "city": the buildings, the football fans, the FIFA, everything - including not only how they behave, but also how they are created.

- CRISPR is not malware, but it's (as the article mentions in passing) a search & replace function that allows you to change the Code, but with the described limitations - the biggest of which is that every "edit" can have any number of unintended consequences.

I agree that every analogy has its weak points, but Crispr's actual function in cells is the exact opposite of malware, so it just seems like a very odd analogy to pick. It also creates a sense of fear and distrust that is bad to associate with a tool that the public really doesn't understand as-is.

Of course "malware" is maybe a bad word for it, because it contains "mal" (i.e. "bad"), which is contrary to the (hopefully) beneficial effects of CRISPR. But the fact remains that CRISPR is using an exploit to hack the "Code" I mentioned above, with potentially dangerous results, and if you think about it in these terms, "malware" doesn't seem so far-fetched anymore...

They take a decent technical concept aspect (Crispr is a subversion of existing cells running itself as opposed to something we must do manually every time like scissors), an extended metaphor that makes a bit more sense (bodies as cities of interacting cells - changing one still leaves a whole interacting system to keep in mind) and then proceed to butcher it needlessly instead of going anywhere sensible sadly.

Still their start had promise and raises some interesting semantic questions. Even if Malware is the wrong term for the intended use of the tool it is certainly what it would be flagged as by the immune system because of where it treads - just like Anti-Virus software in many cases. Even so Anti-Virus is such a misleading term for Crispr both from scope collision of computer and physical viruses. Even if it how it works literally it risks confusion.

My (layman) analogy so far was to think of proteins as wires and transistors, cells as microcontrollers, DNA as the program of a single microcontroller and our body as the result of billions of microcontrollers which all run (mostly) the same program, interacting with each other.

At least, this gives some appreciation of the complexity of it all.

Also CRISPR is being used for diagnostics, similarly to PCR. It can detect certain DNA sequences or select regions of the genome for targeted sequencing.

This isn't exactly crispr, it's more an application of Cas9 and friends. These are sometimes modified. The DNA and RNA binding capabilities of these proteins is opening up a lot of new possiblities!

Could it ever be used maliciously? Supposedly some countries are poisoning people via Polonium and inducing heart attacks: https://citizentruth.org/cia-heart-attack-gun/

What's the advantage of that compared to poisoning people with regular poisons or diseases?

Anything can be used maliciously. But it's way harder to make an bioweapon an bioweapon out of crispr than it's worth. And it's not like you'd have any particular advantage from using crispr, I guess you could load a bunch targeting some essential gene into an AAV and spray it on people but it's probably easier just to wait for the flu virus to mutate and do its thing.

Life is tolerant to meddling so even idiots like human beings can meddle with it and it still continues. To a point and then it comes crashing down. Humans however are too stupid and arrogant to realize this simple fact and instead are emboldened at each step that they meddle with it and it solves some problem and does not collapse. However nature also does similar insane experiments and produces plenty of catastrophe on its own, so maybe we can justify human intervention as the lesser of two evils. Queer business in a mostly dead universe which is out to eliminate the little life it has in it, and so are humans.

> However nature also does similar insane experiments and produces plenty of catastrophe on its own

I'm curious about this, any references for me to read about?

Mutations, cancer, bacteria that cause major extinctions, etc.

Well, the Great Oxidation Event might count. https://en.wikipedia.org/wiki/Great_Oxidation_Event

Is there a word for something that keeps coming back and making you laugh 10-20-30 minutes after you've read it? This compact treatise on the human condition has it.

Reply on news.ycombinator.com