A layperson's guide to recent advances in gene editing; on undisciplined and irreverent creativity
Godot, Feynman, Bourdain (see: Gödel, Escher, Bach)
Advances in gene editing
Last week I bookmarked this tweet by Patrick Hsu on X for later reading, returned to it, and struggled to understand even at a high level. Later I learned that Patrick Collison (Stripe CEO and cofounder) also cofounded Arc (founders’ list is wild; two Patricks, two Collisons, Ethereum cofounder Vitalik Buterin, Adam Winkel from Jane Street) and also felt the need to explain this groundbreaking research in accessible terms via this tweet (sent to me by Shreya). It’s a good practical example of how text summarization models (with human cross-checking) can help synthesize and translate dense information into something more immediately understandable, and highlight areas to learn more about; a way to reduce the activation energy of initial learning.
[N.B. See Stripe Press for a fun interactive bookshelf reading experience, featuring one of my favorites, the 1986 lecture “You and Your Research” by Richard Hamming.]
Here’s my human paraphrasing of PC’s Claude 3.5 summary. (The last time I took any biology-related course was in high school, so please read at your own risk.)
tl;dr: Bridge recombination is a new gene modification technique proposed by Arc Institute researchers, using “bridge RNA” to insert DNA in-place instead of requiring cuts to DNA as with CRISPR.
PC on X: Since the whole thing is pretty arcane, I fed the blog post (https://arcinstitute.org/news/blog/bridge) to Claude 3.5, and asked it to write an introduction…
For the past few decades, we've been snipping away at genomes with tools like CRISPR, making impressive progress but always constrained by the fundamental nature of our tools: they cut DNA. But what if we could write directly into the genome, inserting whatever we want, wherever we want, without ever making a single cut?
Researchers at the Arc Institute have discovered a new system that does exactly that. They're calling it "bridge recombination”…
Background:
In the late 1990s, we discovered RNA interference (RNAi). This was our first real taste of programmable biology. We could use short RNA sequences to target and shut down specific genes. It was like having a universal remote control for gene expression. Cool, right?
Then came CRISPR in the early 2010s. Suddenly, we could not just turn genes off, but edit them directly… [but] CRISPR works by cutting DNA, and cells don't always repair those cuts exactly the way we want them to.
Both of these systems were revolutionary, but they shared a common limitation: they were destructive
Enter the bridge recombination system. Researchers at Arc Institute, led by Dr. Patrick Hsu, were poking around in the genomes of bacteria, looking at transposable elements. These are sometimes called "jumping genes" because they can cut themselves out of one part of a genome and paste themselves into another. They were particularly interested in a group called IS110 elements. These are about as minimalist as you can get while still being functional – just a gene for the enzyme that does the cutting and pasting… What they found was surprising. When an IS110 element cuts itself out of a genome, those mysterious flanking segments join up to form an RNA molecule. This RNA (which they've dubbed "bridge RNA") folds into two loops, one that binds to the IS110 element itself, and another that binds to the target DNA where the element will insert itself… It's like a set of instructions, telling the enzyme exactly where to insert the DNA and what DNA to insert.
The bridge recombination system [as opposed to CRISPR] doesn't cut the DNA at all. It unzips a small section, inserts the new DNA, and zips it back up again. No broken DNA strands, no relying on error-prone repair mechanisms.
…the real magic is in the programmability… those two loops in the bridge RNA… can be programmed to recognize any DNA sequence. This means you can tell the system to insert any piece of DNA into any location in the genome. It's like having a word processor for DNA.
Here are the Nature papers. As the authors have drawn direct comparisons in utility and precision relative to CRISPR methods, the potential gene therapy, gene editing and wider bioengineering applications (agriculture, etc.) make sense to me.
Interdisciplinary (2024)
Study hard what interests you the most in the most undisciplined, irreverent, and original manner possible. — Richard Feynman
If I could make a film this year pulling together the threads linking events in the world, technology and my personal life and learning philosophy development this year, I’d have to title it Interdisciplinary.
I asked Microsoft Copilot Designer (powered by DALL-E 3), one of the few free tools that will modify images, to generate “interstellar movie poster but it’s interdisciplinary”, and then “interstellar movie poster but replace the word interstellar with the word interdisciplinary”. The best it could do without very obvious typos in huge font was this. I think it’s a reasonable interpretation, placing the subject at the forefront, surrounded by the interdisciplinary practices of astronomy, rocket engineering, and steam and rail automotive(?) technology with a unifying sunbeam backdrop.
Recently I’ve been taking note of when (and as much as possible, how) productive and/or surprising thoughts and connections come to me, especially when I least expect, loosely looking for lifestyle and environment patterns that might help me cultivate and facilitate creative mental space more regularly and consistently.
For example, in the last few days I’ve actually stopped to take a little extra time to take pictures of everything I find beautiful, interesting, worth memorializing in my phone as a keepsake, instead of rushing by and perhaps only taking off-center, motion-blurred photos. (In fact, I have been such a photo purist/minimalist in the past that multiple people have audibly gasped at the low number of photos in my phone.) This is the product of two recent discoveries:
Cold suggesting to me that my phone camera is a good enough camera to capture beautiful photos of the world, and a good way to start practicing basics of photography before I invest in a multi-hundred dollar DSLR / film camera.
Spending a few weeks in the peaceful suburbs with friends I’ve known for the better half of a decade now, and that are like family to me, providing me the comfort and emotional safety to “stop and smell the roses”.
A good illustration of thinking outside the box to apply principles from one field to another is Kris Cao’s blog post from nearly three years ago now, The KLF and how to write a hit song (or paper), on what researchers could learn from the KLF (‘80s British electronic band) and The Manual (proper piece of comedy) they wrote on how to write a chart-topping musical hits. From The Manual:
it is only by following the clear and concise instructions contained in this book that you can realise your childish fantasies of having a Number One hit single in the official U.K… Firstly, you must be skint and on the dole. Anybody with a proper job or tied up with full time education will not have the time to devote to see it through. Also, being on the dole gives you a clearer perspective on how much of society is run. If you are already a musician stop playing your instrument.
Kris writes:
On motivation: I think research is a lot like popular music. Fundamentally, both arenas are battlegrounds of ideas, and your success largely depends on whether or not your ideas get noticed by the community at large or not. From this perspective, there’s a lot of good paper writing advice in the Manual. The core precept is to make your paper as ‘hooky’ as possible – every part of your paper, from the title, to the abstract, to the figures and tables, should be designed to be as informative and memorable as you can.
On structure: By and large, you want to keep to the introduction/methods/experiment/results/conclusion framework as much as possible, as it offers readers a comforting blanket of familiarity when reading your paper. You’re allowed one wildcard section to highlight what’s distinctive about your research, analogous to the breakdown section recommended in the Manual. Until you’ve developed a certain amount of writing mastery, it’s not worth experimenting too much with form.
[N.B. Since this is not a blog of research papers, I am not applying this writing advice down to each note, but do try to structure most of my posts in a consistent and familiar way (by subheading per listed title component).]
In practice, I’m both structuring my time to encourage much more reading and writing, while justifying maximalist consumption of “less productive” media (meme Twitter, The Bear — although the latest season is unfortunately a disappointment) in the name of undisciplined, irreverent, original learning. Watching Ayo Edebiri have a lot of fun on SNL, discovering the many talents of Zack Fox (yes, Tariq on Abbott Elementary) and learning about the creative process behind the song I won’t stop playing this summer, because I have not been avoiding them as “time-wasters” but rather embracing them as opportunities to glean insight on how to stimulate creativity, how to be a compelling public speaker, and how to lead through unapologetic honesty. By consuming and celebrating talent and creativity in all its forms, rather than trying to put in exponentially-less-effective hours with a brain fried by the monotonous and expected, we may unlock productivity in very unexpected ways.
Of course, being able to apply learnings from “down time” to enhance productivity is an insane level of privilege I’ve only been able to grow into the last few years. I think about how to build healthy teams, how to have effective conversations, and how to motivate technological innovation for a living. Google encouraged a limited form of this creativity-facilitating exercise via the (now largely phased out) 20% rule; in an ideal corporate world, I’d build in a wider charter of learning PTO for people to take time off away from their day jobs, up to a year at a time (closer to the professorial sabbatical model, e.g. once every 7 years or so), to learn something entirely different from what they do for a living, and see what they could bring back to their teams.
I’m certainly getting ahead of myself trying to propose a learning PTO model that would work at scale (unlike unlimited PTO — which often works because of the honor code and the idea that when people get unlimited access to anything, it reduces the novelty of that thing and may actually result in them desiring it less — corporate sabbaticals or externships so far appear to be done on an ad hoc basis, tailored to specific talent and interests).
I was chatting with a very senior full stack engineer who comes from a background of construction, woodworking, and contracting. He told me how principles of carpentry in particular translate to architecting and maintaining quality software systems. A carpenter understands the tradeoffs of certain cuts and types of wood and indoor/outdoor applications, and can build a functional piece of furniture with simple tools. Longer-lasting, higher-quality pieces require higher-quality materials, more tailored design, more time and labor, sanding, polishing and finishing, and are more specialized and customized; however, they may be harder to move and re-purpose.
Good carpenters, contractors, and architects both industrial and software have this in common: having a solid vision of what you are building and what it’s for, and being able to improvise with your materials and adjust to time and labor constraints (requiring knowledge of how each component comes together and any inherent flexibility).
The exceptional are inventors, craftspeople, artists. They push the boundary between known and unknown, defining and re-defining realities, ways of working and living and creating. Here’s to the undisciplined, irreverent ones.