Seedlabs

EpitopeLock: De Novo Nanobody Design Service with Structure-Verified Epitope Targeting

A service that designs custom nanobodies to hit a precise spot on a protein structure.

A contract service where a customer supplies an antigen structure and a chosen epitope, and receives de novo-designed VHH nanobody binders that hit exactly that spot — cryo-EM-verified in binding pose and affinity-matured to single-digit nanomolar.

MedicineMonoclonal and Polyclonal Antibodies Research
Microbial Symbiosis and ResistanceBiopharmaceutical antibody discovery (epitope-specific therapeutic and reagent antibody generation)

Concept

A design service for biologics teams that need an antibody to bind one exact surface. The customer supplies a 3D antigen structure and defines the target epitope — a conserved viral stem region, a toxin receptor-binding interface, an active-site-adjacent loop. The service generates de novo VHH (nanobody) backbones with a fine-tuned RFdiffusion network whose CDRs are computationally shaped to that surface, screens the designs by yeast display, and matures hits by continuous hypermutation. Deliverables: a panel of epitope-confirmed VHHs, a cryo-EM structure confirming at least one binder adopts the designed pose, and a matured lead with single-digit nanomolar affinity, convertible to scFv or full IgG format. The single wedge is epitope specificity by design: immunization, patient isolation, and random library panning cannot guarantee which epitope an antibody will hit — this can, which is decisive when only one conserved or functionally critical surface is worth binding.

Why now

Paper [0] demonstrates the complete workflow on real targets: de novo VHHs against four disease-relevant epitopes, with cryo-EM confirming the designed binding pose for influenza haemagglutinin (VHH backbone RMSD 1.45 Å versus the computational design) and C. difficile toxin B, and a high-resolution structure confirming atomic accuracy of the designed CDRs. Initial designs bind at tens to hundreds of nanomolar Kd, and OrthoRep affinity maturation converts them into single-digit nanomolar (even sub-nanomolar) binders that keep the intended epitope selectivity. The same pipeline produced scFvs and a full-length IgG1 (Kd 68 nM), so the service can deliver whichever format the customer needs. This is the first shown route to epitope-specific antibodies generated entirely in silico and then experimentally validated — exactly the capability the service productizes.

Written by GLM · 9 Oct 2026

AI assessment

Backed by 1 paper97

A high-value, technically grounded service that productizes a breakthrough in de novo antibody design to solve the critical industry pain point of epitope-specific binding.

Evidence strength
5/5
The idea directly maps to the full-text results of the cited paper, which demonstrates atomic-level precision (RMSD 1.45 Å) and a complete pipeline from RFdiffusion to cryo-EM validation and affinity maturation.
Market pull
5/5
Biopharma has a massive, urgent need for epitope-specific binders (e.g., targeting conserved viral stems) that cannot be reliably achieved via traditional immunization or panning.
Novelty & moat
5/5
The approach is fundamentally different from existing discovery methods and bears no similarity to the unrelated ideas in the catalog.
Feasibility
4/5
The workflow is already demonstrated in the research; the primary challenge is scaling the yeast display and cryo-EM throughput for a service model.
Wedge clarity
5/5
The wedge is extremely sharp: providing a binder for a specific, user-defined 3D coordinate on an antigen, which is a binary success/fail metric that provides immediate value.
Simplicity / focus
5/5
The idea avoids 'platform' bloat, focusing on a single, clear deliverable: a structure-verified, high-affinity nanobody for a specific epitope.

Scored by Gemma 4 · 9 Oct 2026. Rated against a fixed rubric (evidence, market, novelty, feasibility, wedge, simplicity), relative to the closest ideas already in the catalog. A prior estimate to compare ideas before real-world signal arrives.

Persona discussion

AI personas trained on real people's expertise debate this idea as it evolves.

View the discussion →

Act on this idea

Ideas only matter if someone runs with them. Your message goes straight to the founder's inbox — nothing is stored on our servers.

Who benefits

  • Vir Biotechnologycompany

    Develops haemagglutinin stem-targeting influenza antibodies; epitope-locked de novo design against the conserved stem epitope demonstrated in the paper could yield broader neutralizing candidates faster than immunization-based discovery.

  • Merck & Co.company

    Markets Zinplava (bezlotoxumab) against C. difficile toxin B; the paper's TcdB epitope-specific VHHs, scFvs, and full IgG1 provide a direct route to next-generation anti-toxin programs.

  • Twist Biosciencecompany

    Sells antibody discovery services built on library panning; adding de novo epitope-specific design would differentiate its offering where customers need a guaranteed binding site.

  • NIAID (National Institute of Allergy and Infectious Diseases)organization

    Funds rapid countermeasure development for outbreak pathogens; de novo design against a specified conserved epitope compresses response timelines versus immunization or patient isolation.

  • Sino Biologicalcompany

    Research reagent antibody supplier; epitope-confirmed binders are premium reagents for assay development where binding site determines assay performance.

Research it builds on

  1. Atomically accurate de novo design of antibodies with RFdiffusion
    Nathaniel R. Bennett, Joseph L. Watson, Robert J. Ragotte et al. · 2025 · 305 citations
    All ideas from this paper →
Leave feedback
feasibility