Integrated strategy for Cambody library design
Library design was based on taking native camelid nanobody sequences and computationally filtering them with the aim to remove poorly stable/sticky/ cross-reactive binders that often dominate selections but result in undesirable binders
A critical balance between overall stability and functional variability is important:
High stability = robust production and solubility
Diverse CDRs = essential for discovering binders to a wide range of targets
These approaches have exploited recent computational approaches from the Sormanni lab including those in the following papers:
- Ali et al (2026) Disulphide and sequence-encoded conformational priors guide nanobody structure prediction. BioRXiV. https://doi.org/10.64898/2026.02.13.705647
- Ramon et al (2026) Deep learning assessment of nativeness and pairing likelihood for antibody and nanobody design with AbNatiV2 MAbs18(1):2646361. https://doi.org/10.1080/19420862.2026.2646361
- Ali et al (2025) Improving nanobody structure prediction with self-distillation. BioRXiV. https://doi.org/10.64898/2025.12.01.691162
- Ramon et al (2025) Prediction of protein biophysical traits from limited data: a case study on nanobody thermostability through NanoMelt. MAbs 17(1):2442750. https://doi.org/10.1080/19420862.2024.2442750
Cambody library diversity
NOTE 1: CDR3 length based on AHo numbering. With IMGT numbering the length would be 1 aa longer
NOTE 2: Total diversity = 1 billion sequences for each library
Ongoing development
Community Distribution: Release validated library as open-source tool to research community to democratize nanobody discovery
Iterative Refinement: Utilize screening data in a feedback loop to improve computational models for next-generation library design