A Nature paper showed RFdiffusion and allied hypermutation systems can design de novo single‑domain antibodies that bind user‑specified epitopes with atomically accurate poses, and structural work by cryo‑EM validated those designs. The study reports successful designs against multiple therapeutically relevant targets and demonstrates that computationally generated sequences can be further optimized for tight binding. Nobel laureate David Baker and collaborators discussed the paper’s advance while cautioning the field that computational antibody design—though now demonstrably feasible at atomic resolution—remains early for broad therapeutic deployment. The work shortens lead timelines by replacing some experimental screening steps, but groups emphasize the need to translate in vitro binding into in vivo safety and pharmacology. For drug discovery teams this marks a step toward design‑first antibody campaigns: de novo generation could reduce screening burden and expand epitope coverage, but will still require downstream validation in cellular and animal models before clinical entry.