Catalytic Motif Scaffolding: Diffusion-Designed Enzyme Backbones

AI protein design tools don't just fold a chain toward a generic shape — for enzyme design, a generative diffusion model has to build a whole scaffold protein around a handful of catalytic residues while holding their relative 3D geometry exact, because that geometry is what actually does the chemistry. This simulator animates that "motif scaffolding" workflow directly: pick a catalytic preset (a Ser-His-Asp triad, a three-ligand metal-binding site, or a Cys-His protease dyad), choose a scaffold length, and watch a backbone denoise from pure coordinate noise into a folded chain that threads through those fixed active-site residues. The guidance-strength slider is the star of the show — it is the same conditioning knob real diffusion-based design pipelines use to trade off backbone plausibility against active-site precision, and the live catalytic-geometry RMSD readout shows exactly what happens to a designed enzyme's function when that guidance is too weak.