Yutong Zhao

I had no idea they came out with a 3rd edition

discovering the parenting joys of jingle bells batman smells

Super excited to preprint our work on developing a Biomolecular Emulator (BioEmu): Scalable emulation of protein equilibrium ensembles with generative deep learning from @MSFTResearch AI for Science. #ML #AI #NeuralNetworks #Biology #AI4Science biorxiv.org/content/10.110…

@FrankNoeBerlin @MSFTResearch But the general question of can we capture most of re-organization dGs with a backbone rep is super interesting, esp if it’s a learnable quantity

@proteneer @MSFTResearch So this is a backbone representation, side chains are marginalised out, therefore not certain if ddG’s of pocket opening can be quantitatively predicted. Do we have any dataset or references? How would you measure this, especially for local changes?

@FrankNoeBerlin @MSFTResearch There’s a couple of datasets/papers on re-org free energies from Schrödinger: T4 lysozyme (Lim et al), Abl, HSP90 (Fajer et al.) pmc.ncbi.nlm.nih.gov/articles/PMC50… pubs.acs.org/doi/abs/10.102…

@davidkmyang TechChem when?

I’m sorry but TechBio is a cringe term and turns people away from the community. I like the movement but this can’t be the long term brand. Many agree in private but I have the courage to say it

Whoever decided that lambda should be a reserved keyword in python obviously never worked on free energy methods

The point of a startup is to make usable technology for others. When you make software, you have to watch at least 10 people use it. Sit next to them and say absolutely nothing. Force yourself to marinate in the failure of your product design. Every version 1 of any software will be absolutely destroyed by first interaction with users. You need to watch your new creation be absolutely misunderstood by users to reform version 1 into the one that actually works. There is only one path: figuring out where the sharp edges, the places people get caught, the assumptions you make as a builder that turn out to be wrong, and then relentlessly sanding it down so that anyone can use it. That is good design. That is the key to a good product. There is no shortcut for this. WATCH USERS AND CRINGE AND THEN FIX IT. SAND DOWN THE EDGES.

Nobel prize for left handed alpha helix

It is turning out to not be the case. "Structure=>function" had seemed easy because there were all these papers solving a key protein's structure on the way to understanding what it does: "ah, these residues here make a pocket and that one moves, and it all makes sense..." 12/

It has been 18 years since the first Falcon flight of Falcon 1 failed

The only status game that matters is if you made something people want

“behind mountains are more mountains.”

Ahead of your time with GPU bling @ImranSHaque ? Tell me about it.

Ahead of my time ht @proteneer

@TimothyDuignan integral of the wavefunction under the born Oppenheimer approximation?

@proteneer I mean this. Is there a fancier term?

This is one of those beautiful ideas that took me a long time to see, but in retrospect seems obvious. If correct, it implies a massive step up in the number of problems molecular simulation can be fruitfully applied to.

Bothered by the lack of long-range interactions in ML potentials? Meet Latent Ewald Summation—our solution to fix "shortfalls" in short-ranged ML potentials for electrostatic and dielectric systems, with only a modest computational cost! arxiv.org/abs/2408.15165