Design of Full-Length Agonist and Antagonist Antibodies for Hard Targets 1. Pythia Labs reports de novo design of full-length human IgG1 antibodies that functionally modulate GPCRs in cells, including both antagonists (e.g., CCR8, CXCR4) and an agonist (MRGPRX2). They frame this as the first demonstration of both GPCR agonism and antagonism by computationally designed full-length IgGs. 2. The core is an end-to-end computational pipeline built around Pythia-1, an all-atom generative model that jointly designs antibody sequence plus 3D all-atom structure under explicit constraints (target structure, chosen epitope, and desired antigen conformational state). Candidate generation and prioritization is designed to run in ~1 day. 3. Scale and outcome: designs were produced for 22 therapeutic targets (15 soluble/ectodomain targets + 7 GPCRs). After a single design round, they screened at most 48 IgGs per target experimentally and found binders with single-digit nM KD for soluble proteins and single-digit nM cell-binding EC50 for multiple GPCRs. 4. A key methodological point is “programmable mechanism”: they choose epitopes and receptor conformations to bias toward antagonism vs agonism. For GPCRs, they explicitly design to inactive-state structures for antagonists (e.g., CCR8) and active-state structures for agonists (MRGPRX2), aiming for state stabilization as the mechanism of action. 5. Example soluble target antagonism: IL11 antibodies were designed to overlap receptor-binding surfaces. Lead IL11_04 bound IL11 with KD ~2.5 nM and inhibited IL11 pathway signaling in a HepG2 STAT3 reporter assay with IC50 ~5.6 nM, outperforming a patent-derived comparator in the same assay conditions. 6. Example soluble target antagonism: β-NGF antibodies were designed to block TrkA engagement. βNGF_18 showed KD ~7.6 nM and antagonized β-NGF/TrkA signaling (IC50 ~6.7 nM in a TrkA reporter assay; ~28 nM in a TF-1 proliferation assay), illustrating translation from binding to functional neutralization. 7. Example GPCR antagonism with pocket engagement: CCR8 designs achieved very strong on-cell binding (best EC50 ~3 nM). Several antibodies antagonized CCL1-induced β-arrestin recruitment, including single-digit nM IC50 designs. Models suggest unusually long LCDR1 loops can reach into the orthosteric pocket, highlighting a design space not typically explored in standard discovery. 8. Example GPCR antagonism with deep HCDR3 penetration: CXCR4_23 showed EC50 ~15.5 nM for cell binding but IC50 ~5.5 nM in a β-arrestin recruitment antagonism assay, and was modeled to penetrate the orthosteric pocket even deeper than a published pocket-binding IgG comparator, emphasizing that predicted binding geometry (not just affinity) can drive function. 9. Example GPCR agonism: MRGPRX2_15 was designed against an active-state receptor structure and produced saturable activation in a β-arrestin recruitment assay (functional EC50 ~213.5 nM), with potency comparable to a natural ligand control in their assay. The model suggests CDRs contact receptor regions used by the ligand, consistent with active-state stabilization. 10. Developability-by-construction: designs were conditioned on human germline frameworks, constrained to human-like CDR length distributions, and filtered to reduce sequence liabilities (e.g., glycosylation motifs, deamidation/isomerization motifs). Experimentally, many designs met typical developability thresholds (expression titer, Tm by DSF, hydrophobicity by HIC, self-association assays) and compared favorably to clinical-stage/FDA-approved controls; the authors also provide full sequences of lead candidates for independent validation. 📜Paper: biorxiv.org/content/10.64898… #ComputationalBiology #ProteinDesign #AntibodyEngineering #TherapeuticAntibodies #GPCR #GenerativeAI #StructuralBiology #DrugDiscovery #Developability #Bioinformatics

We built Pythia-1, a frontier generative model for therapeutic discovery, to expand the reach of antibody therapeutics to hard-to-drug targets. In our latest study, we used it to design full-length human antibodies that activate or inhibit GPCRs in cells. To our knowledge, a first for de novo design. With just one design round and minimal screening, we identified multiple antagonists with single-digit nM inhibitory potency and an agonist with potency comparable to the receptor's natural ligand. Designed antibodies also demonstrated favorable developability profiles. Read the preprint: biorxiv.org/content/10.64898… We're releasing the sequences: pythialabs.com/

Design of Full-Length Agonist and Antagonist Antibodies for Hard Targets

Introducing Pythia-1, a frontier generative model for protein design. In our new study, we used it to rapidly design full-length human agonist and antagonist antibodies for hard-to-drug targets. We’re sharing complete sequences of all lead antibodies to support independent validation and further research. 1/8