Jan Domanski (@jandom_random) 's Twitter Profile
Jan Domanski

@jandom_random

Forbes 30U30, ex-CTO @Labstep, ex @DEShawResearch, ex @amazon. Software and drug-discovery at @charmtherapeutx

ID: 707362576762597376

linkhttp://www.jandomanski.com calendar_today09-03-2016 00:29:15

718 Tweet

301 Takipçi

368 Takip Edilen

Paul Graham (@paulg) 's Twitter Profile Photo

Apparently EU users don't get new AI versions till several months after US users, because AI companies have to get regulatory signoff first. Was the cost of overregulation ever so clear? Every user in Europe is several months behind, in a field that changes every several months.

Jan Domanski (@jandom_random) 's Twitter Profile Photo

Very pleased to see this work out, and massive thanks to everyone involved! This works shows the incredible potential of combining the new ML tools with existing biophysical methods, including FEP. Also, an example of *actually* using ML in a real drug-discovery program :)

Jan Domanski (@jandom_random) 's Twitter Profile Photo

A large fraction of #compchem problems could be settled by just two things - believable (industry?) datasets around a property of interest - a suitably expressive software to run multiple tools quickly on above dataset & evaluate 🌶️🌶️🌶️

Jan Domanski (@jandom_random) 's Twitter Profile Photo

Has somebody tried boltz-2 with perturbed “off-pocket” crop? Such that the predicted structure cannot be corrected (eg picked residues are outside the binding pocket) and checked if the affinity head is still working? (It shouldn’t be under such perturbation)