@edunov I benchmarked five cofolding methods against a set of PXR PDB structures and results were not great, despite most of these systems being in the training data for these methods.
Given your results against 2A protease, I bet Pearl could do better!
@edunov Super cool!
A great benchmark for this available right now - the OpenADMET PXR challenge. Highly mobile pocket, with the PDB overrepresented by active configurations.
https://t.co/VfnhTTDWXH
@DaveShapi Software is the horse in this analogy, not the engineer.
I think the better analogy is what did the automobile do to the transportation/civil engineering trades?
Many lower-level tasks going away, but I think more effort will be needed at coordination and systems engineering.
@flyme2mars I like the wispr flow interface where you click when you're done. The "over" signal from walkie/talkie days.
Would be cool to have a glove/wearable to gesture this signal. I'll try to wire one up this week.
@1a1n1d1y Yeah...inability to focus on the main thread, constantly finding marginally relevant side quests to investigate.
I don't need AI for that, that's where I'm elite.
SPLICECRAFT v1.0 IS LIVE!!!!!!!!!!!!
Open your terminal and type in "pipx install splicecraft" if you want to try it out, then spam "splicecraft update" often as I push updates frequently. A labor of love for the community I adore. Enjoy! ๐
@bharath_IGC Nice! I've always suspected this, so excited to read about how you showed it.
And you know Adrian Whitty! My PhD lab was next to his. Great guy for advice and to bounce ideas off of.
Full write-up: methods that worked, methods that didn't, and what I'd do differently next time. Solo build at Minimal Bio LLC on one GPU.
https://t.co/lGWPwazZws
@j0hnparkhill Without a doubt. And for those who will scream "agents expensive!", the solution is the same as it was before: figuring out the standard workflows you can precook and smart engineering. Just more at the agent instead of the code level.
@design_proteins@DdelAlamo The squirrels in my yard laugh at adding chili flake to birdseed. You're making me wonder now if they have a TRPV mutation - now just how to get a sample to sequence...