IgGM Pipeline: IgGM (Immunoglobulin Generative Model) is a diffusion-based generative model specifically trained on antibody and nanobody sequences and structures. Unlike RFdiffusion which designs the backbone geometry first and then sequences it separately, IgGM jointly generates both the sequence and structure of all three CDR loops (H1, H2, H3) in a single pass, conditioned on the target antigen surface. This co-design of sequence and structure allows IgGM to explore a fundamentally different region of the nanobody design space — it is not constrained by a fixed framework backbone and can discover CDR combinations that may be inaccessible to scaffold-based diffusion approaches. In our pipeline, IgGM was conditioned on the same RBX-1 RING domain structure (PDB 2LGV, residues 40–108) and epitope hotspots (W87–V93), generating 100 diverse nanobody candidates that were validated through the same AlphaFold3 and SAbDab filtering pipeline as the RFantibody designs.