Some context: Been waiting for this to come out for a while! Main innovation is leveraging RosettaFold (protein folding neural net) to generate protein backbones via diffusing in 3D space! From backbones, we can generate sequences that would fold into said structures via sequence design algorithms (check out: proteinMPNN, Rosetta FastDesign).
In terms of applications: This is super relevant for our ability to create strongly binding protein binders (ex timely creation of proteins that bind to virus spike proteins), and designing enzyme from scratch!
Prior methods suffered from much lower success rates for generating “good” backbone structures. Extremely exciting!! If you want to learn more, check out the Baker group at UW!
So in essence, if I understand correctly, instead of generating Balenciaga Pope or arrested Trump fake images, we can now dream up fake protein things which may actually be viable for whatever purpose if synthesized in the real world?
Dreaming up a static three dinensional structure does not guarantee that it is stable in a given environment, or that production of this structure in a lab is viable. A huge problem in the space is protein folding–concerned with figuring out how you get from an unfolded linear string of amino acids to this three dimensional structure.
Folding takes into account many variables, and a big chunk of current experimental structure determination is concerned with controlling/adjusting these variables.
So this dreaming up will provide a potential “quicker way” into what a folded protein might look like, but it will not guarantee you that humanity knows how to actually produce it in the real-world.
Disclaimer: someone correct me if I’m wrong. I might be rusty on the latest developments, as I’ve left the field after my PhD.
Indeed there are many pitfalls between a protein sequence and something useful to humanity, but there is reason to believe the technique is capable of generating such proteins:
1) In the paper they express several of their designs and show stability via circular dichroism experiments. They also show size exclusion chromatography results indicating some of the proteins are of the expected size and are not aggregating.
2) Since RFDiffusion and ProteinMPNN, which generates the actually amino acid sequence, are trained using Protein Data Bank (PDB) data, it's reasonable to presume the predicted proteins will be well behaved. To solve a protein structure via say X-ray crystallography, EM, or NMR and deposit it into the PDB requires bucket loads of stable protein. I used several grams of recombinant protein for a X-ray structure I solved. Since the ML models are trained on well behaved proteins, I can believe the generated proteins will also be well behaved.
I like idea (1), where kids don’t have mental space to sort thoughts out. Although I disagree that most kids are too busy for it. It may be that habitual social media use decreases kids’ free time, as they use social media/their phone as soon as they’re free, rather than sit, think, and process thoughts.
I think this is also applicable to adults, although it’s effects could be less since smartphones weren’t prevalent when current adults were children.
I’ve always thought of it as that. More generally, idea could be applied to YouTube, podcasting, etc. One person has ability to broadcast to millions, while the millions can’t broadcast back equally.
The idea is pretty cool! I think it has potential, as when I used instagram, people seemed to enjoy the “closed friends” feature, and also created separate IG accounts to follow close friends.
Thanks! It's been interesting as I used the first prototype to learn how to code. There's been breaks between then and now but that was 18 months ago - a lot of the premises for building Circles have been adopted by IG. Close Friends, Send to Groups, DM focus and easy account switching.
Tragic. It’s always scary to handle needles in a lab. Even with good needle safety protocols, accidents are bound to happen.
There’s this case that my PI showed me of a student who pricked his finger with a needle filled with DCM [1] (which is very common solvent in used in lab). The residual DCM at the tip of the needle caused the damage shown in the pictures.
When your work involves handling tens of needles daily, accidents are bound to happen. I pricked myself once with a needle. Thankfully it was brand new and dry, so nothing happened. It was just the idea of it was terrifying. In an alternate world, I could’ve gone to the ER because of it. Hope someone invents a safer needle.
Just my opinion, but that’s not DCM residue on a needle, that’s someone using a needle to transfer and forcing DCM under their skin.
I drenched myself in DCM and stick myself with needles a few times.
Not to say the lab isn’t dangerous. I knew one person who peppered their abdomen with glass when a perchlorate exploded, scarred their face when a peroxide exploded and burned themselves with an organolithium.
But a simple needle stick with DCM residue won’t do that.
Blunt needles are a thing. They are used for example in electronics as a way to dispense tiny amounts of solder paste or flux on components pads on a PCB. Presumably when transferring solutions from container to container, you don't need a sharp needle, unless the container is closed with a membrane that needs to be pierced.
There are also gloves that protect against puncture. If you're using both sharp needles and non-protecting gloves then it is probably time to revise the procedures.
