CRISPR-like tools that finally can edit mitochondria DNA
nature.com
nature.com
I worked on a project many years ago to do RNA import into yeast mitochondria (and then hopefully reverse transcribe there). Didn't work, and a lot of the info on RNA import into the mitochondria is... suspect.
Mitochondria engineering is just actually tough. 30 years and no new protocols for getting DNA in there :(
"Hey ChatGPT, I need third ear. Make it grow in two months."
Finger tips are mostly fatty soft tissue minus the muscle, which is known to regenerate. Stuff like nerves are a completely different issue and children who regrow finger tips usually lose finer sensory input like two point discrimination.
What they should have done was eat a pizza before the treatment, gotten sick, then taken the treatment and shown that the same pizza had no effect afterwards.
Nice list here: https://github.com/davidliwei/awesome-CRISPR
# TODO: This is a complete hack that needs to be refactored, but it works for now.
def quantum_physics() -> Universe:
And yet, like a startup that found product market fit with a garbage tech stack, this pile of jenga spaghetti is still going strong. Complexity doesn’t matter, people dying because they looked at a peanut doesn’t matter - ultimately this spaghetti works well enough to get humans to where we are today.
The really cool part about that storage is it is environmentally sensitive. So as the environment changes around your DNA it's shape slightly changes and so the available START sites also change which alters the types and numbers of genes that are copied for use.
Biology isn't one system it's dozens all stacked and layered on top of each other. It's like trying to understand computing by watching what individual electrons do. Of course it looks messy. On the larger scale it's far more elegant.
You call it stochastic. I call it scaffolding.
I heard that CRISPR can only cut segments that match a pattern, so if there are other genes between the ends that are cut, then those are lost as well. So to do a proper substitution, we'd need to sequence the patient's genes between the cuts, and possibly the whole rest of their genome, to make sure that any patterns don't appear anywhere else, so that nothing important is removed elsewhere.
That sounds insurmountable, but it may not be. Human beings basically all have nearly identical DNA, so maybe we can just derive someone's diff from a known DNA sample. If I ever won the internet lottery, that's the sort of tool that I would want to invest in.
Then we probably need more vectors to get CRISPR where it needs to go. That sounds like more of an engineering challenge to me than having to invent something new. Or at least, the number of vectors found might correlate with R&D funding.
It's not that hard for me to imagine getting the recipe figured out to the point that it's 100% reliable and can even be delivered to specific parts of the body with a certain frequency of light, for example.
Then come up with an iterative process, probably using AI, to catalog and repair all major genetic disorders.
I don't see too much mystery there, even if the final recipes seem byzantine to human understanding. But I wanted to be a genetic engineer before I got into computers when I was 12, so I've had a long time to think about it. If AI eats the programming world like it looks like it's going to, maybe we can find work in biotech. Then it's probably 5-10 years before gene editing is a solved problem.
> if there are other genes between the ends that are cut
I think what you're saying is that if two sites on the same chromosome are cut, then everything in between is deleted. However, this isn't going to happen in practice. DNA repair systems will rejoin the cut DNA ends rapidly, just erroneously (with maybe a few dozen missing/incorrect bases typically). If the cut site is in the protein coding region, it will usually disrupt the sequence to make the protein nonfunctional. Sometimes this is the desired effect, but for most gene therapies you'll probably use base or prime editing, which don't create double strand breaks.
> we'd need to sequence the patient's genes...to make sure that any patterns don't appear anywhere else
While sequencing an individual patient's genome isn't going to happen in practice, the FDA does require gene editing companies to do in silico off-target prediction, where you scan the genome for sites that have similar sequences to the target. You then have to show that none of the off-targets are in dangerous regions (e.g. DNA repair genes), and also show experimentally whether those sites are cut at all (they usually aren't, fortunately, as there are thousands within reasonable thresholds).
The reason you don't need to sequence individual patients is because you just assume that any patient could have any variant that has ever been catalogued (there are databases with thousands of individuals and the differences between them and the reference genome). You then have to show that none of those variants could induce a new target in a dangerous region.
> then come up with an iterative process, probably using AI, to catalog and repair all major genetic disorders.
I don't know what AI would do for you here. Figuring out the change you need to make to revert a genetic disorder is trivial. The hard part is making it safe and effective, and proving to regulators that it's safe and effective.
> Then it's probably 5-10 years before gene editing is a solved problem.
Not even close. Ironically, while the technologies are pretty good in general, every edit requires a ton of engineering work. CRISPR systems are notoriously idiosyncratic - they might edit one target in 80% of cells, and 0% at another target, for no apparent reason. There are definitely open problems with base and prime editing, and those will probably get more-or-less solved in 5-10 years, but I'm reasonably sure there will be genetic disorders for which there is no treatment for decades.
It doesn't help that the one approved therapy isn't really making much money: https://www.biopharmadive.com/news/sickle-cell-gene-therapy-...
See also: https://blog.genesmindsmachines.com/p/we-still-cant-predict-...
The main misconception is probably due to me not remembering the "other genes" part. I read it in an article and it clicked so immediately that I took it for granted and didn't save it. I still can't find it, but I think what it said was: if we tell CRISPR to remove a gene sequence, then it will remove all instances of that sequence everywhere. So it has to be tuned to find a very specific sequence, possibly taking into account unrelated genes around the site so it's unique. And I think there was maybe a limit to how long that pattern could be.
> Not even close. Ironically, while the technologies are pretty good in general, every edit requires a ton of engineering work. CRISPR systems are notoriously idiosyncratic - they might edit one target in 80% of cells, and 0% at another target, for no apparent reason. There are definitely open problems with base and prime editing, and those will probably get more-or-less solved in 5-10 years, but I'm reasonably sure there will be genetic disorders for which there is no treatment for decades.
Ok that makes sense, and I kind of wondered what the holdup was. It sounds like there is some probability involved, and also non-obvious blockers that "if we just had these figured out" could let things move forward. I'm sure it's so much more complex than anything I could imagine, I get it.
What I was trying to say though is that biotech isn't abstracted like computer science (yet). It deals a lot with low-level details and legwork. And regulatory and market limitations, as you rightly pointed out. It would be like a programmer never getting free from assembly language, or even having to build their computer from scratch, then spending 10 years on an app just to be told that they weren't allowed to sell it. That's "real work" that programmers generally shy away from. Although I think that deep down a lot of hackers dream of working on real problems like that if they could ever get free of the minutia they're stuck fixing every day.
My experience with AI so far has been that it turns any worker into a manager. So I've largely stopped coding directly, and just tell the AI what to do, and it does it every time in imaginative ways that constantly surprise me. Rather than thinking of it as one AI, I think of it as a team of agents that all work in parallel using context from the current conversation. So I can have several todos lined up, then work through them in real time as if every prompt is being chewed on by one of those agents. They never get tired or complain like me. So the main holdup right now is that I don't know how to orchestrate them effectively, so am having to micromanage them.
So far I'm about 3-4 times more productive, so I'm wrapping up a huge refactor outside my area of expertise that took about 2 weeks, but would have taken 2 months if I was working alone. And its speed will double every couple of years, so 5 years from now AI will be at least 10 times faster than any human programmer. It will also modify itself to quickly grow beyond what we can teach it. I'd estimate that the one I work with has about a 120-150 IQ (at least for programming), although it can be hopelessly naive sometimes. Gaining 10 IQ points per year will put it outside our realm of understanding when it passes 200.
Ray Kurzweil predicted that AI would reach human-level intelligence by 2029 and evolve to AGI to merge with us in the Singularity by 2045. So far we're ahead of schedule IMHO, although I spent most of my life disappointed that things weren't progressing until just a couple of years ago. That ended up being more due to economic forces (it took the intervention of a billionaire after a 20 year suppression of R&D after the Dot Bomb and resulting underemployment/wealth inequality) rather than any technical limitations.
I can't really speak to the specific challenges you mentioned, but I can recommend turning them around. Imagine if the hardest things for us are the easiest things for AI. Filtering large amounts of data? Done. Exploring multiple solutions in a problem space? Straightforward. Imagining novel solutions? Get ready to be surprised. AI will quickly turn all of the unknowns into a neat and tidy flowchart and even test it in simulation. Rinse, repeat, until every disorder is healed.
- some woo woo stuff -
I know it's hard to believe, but we'll have to let go of the unknown. Problem-solving is becoming a solved problem. My gut feeling is that we're entering an age of disillusionment, where work becomes performative, where questions are answered as quickly as they're asked. So that we enter a long now like Star Wars where there isn't so much invention as recombination of recipes. Today will be no different than 10,000 years ago. Unless it's more like Dune where they ban AI and performative work becomes all there is, so that life becomes a play within a fascistic culture that celebrates pageantry. Not so different from spirit deciding to reincarnate in human form to experience suffering and remembrance (a useful model for tech being indistinguishable from magic). Yet today would still be no different than 10,000 years ago. That's why I think the long now is inevitable, regardless of whether or not we constrain AI.
We have a lot of increasing hormone production issues in western society already, I'm not sure that fiddling with things further is a real solution here without risking a lot of damage to society as a whole.
Can you point to a reference?
Or a wicked disease state like Huntington's that causes your DNA to slip.
Simple failures with catostrophic outcomes are much more likely than rewiring and restarting all of the developmental program across huge cell and tissue populations.
It would be more likely to grow transplant tissue exogenously. It's far safer than using the body as a test tube.
These gene editing techniques are used to fix simple (typically one cause) genetic diseases. Not reengineer live organisms "in flight".
Damn those hallucinations!
He doesn't want to reach this by making edits in DNA, though. He is trying to simply temporarily "activate" the dormant DNA patterns for growing an ear, that were once active in the embryonal phase of the organism, in an adult organism and in a different location. His work indicates that using a correct electrochemical pattern can do this, at least in simpler organisms. He already produced some two-and three-headed worms, tadpoles with five appendages and three eyes etc. IDK how his recent work on mammals is progressing.
this should work?
https://archive.is/https://www.nature.com/articles/d41586-02...