AI optimization

Home > Services > Nanobody world > AI optimization
Nanobody Humanization
AI
Optimization

Jotbody’s AI optimization service helps transform promising nanobody binders into stronger, more developable candidates. By combining advanced sequence modeling, structure-guided analysis, and developability prediction, our platform supports the optimization of binding affinity, stability, solubility, expression, and immunogenicity risk—accelerating the path from discovery to functional nanobody development.

Our Solution: JotStruct + JotScreen

We have established an integrated AI platform consisting of JotStruct and JotScreen around the key stages of VHH drug development, which supports core tasks such as head-to-head design, humanization, affinity maturation, and candidate selection.

JotStruct and JotScreen workflow
  • VHH from-scratch design: Design and generate novel VHH candidates per target requirements to accelerate early discovery.
  • VHH humanization optimization: Humanize existing VHH/lead molecules to lower downstream development risks.
  • Affinity maturation & performance optimization: Optimize binding activity, specificity and overall performance to improve candidate quality.
  • Integrated R&D process: Support end-to-end advancement from candidate generation, screening to optimization, boosting efficiency and value.

JotStruct: AI-driven nanobody discovery and optimization platform

JotStruct AI-driven nanobody discovery and optimization workflow
  • Unified Generation & Optimization Framework: Supports integrated development covering De novo design, target-binding molecule generation, and candidate optimization.
  • Structure-Driven Design: Incorporates antigen structure and molecular interaction data to enhance design specificity and efficiency.
  • Multi-Objective Optimization: Balances activity, specificity, and developability to improve the overall quality of candidate molecules.
  • Industrial Application-Oriented: Suitable for multiple R&D stages including early discovery, candidate screening, and downstream optimization. Designs target-specific VHHs against any target within three weeks.

Performance

JotStruct performance data

JotStruct demonstrated excellent computational design capabilities in multi-target testing. It outperformed benchmark methods in success rate, number of successes, and key structural scores. This indicates that JotStruct can stably generate high-quality candidate molecules under complex target conditions and possesses high-throughput design capacity for practical projects.

JotScreen: AI-enabled screening, prioritization, and
optimization of nanobody candidates

JotScreen AI-enabled screening and optimization workflow
  • From existing molecules: Targeted optimization based on current VHH or lead molecules for real-world projects.
  • Humanization & activity balance: Improve humanization while preserving or enhancing binding performance.
  • Multi-round iterative optimization: Enable continuous optimization using experimental data to boost candidate quality.
  • Drug development oriented: Balance affinity, expression, stability and developability to maximize translational value.

Case Study: JotStruct

Denovo design of TGF-β3-specific nanobodies

Denovo design of TGF-beta3-specific nanobodies

JotStruct enables the design of novel nanobodies based on real target requirements. In the TGF-β3 project, the platform successfully obtained candidate molecules with functional potential, among which VHH8 demonstrated better selective blocking performance. This result validates the practical application value of the platform in the discovery of new molecules and function-oriented design.

Single-round Humanization of PD-L1 Nanobody
Yielded ~10x Binding Affinity

Single-round humanization of PD-L1 nanobody
  • JotScreen can efficiently perform humanization and optimization based on existing nanobody lead molecules.
  • In the PD-L1 project, the optimized molecules showed superior performance in both ELISA and SPR tests.
  • The KD value increased from 73.4 pM to 9.48 pM, with an approximately 7.7-fold increase in affinity.
  • This result validates the practical application value of the platform in humanization and performance optimization.

Case Study: JotScreen

JotScreen can directly process large-scale candidates from bulk-seq/ BCR-seq
and efficiently screen for high-performance nanobodies

JotScreen case study workflow from BCR-seq to high-affinity nanobody screening
  • Among the top 24 candidate sequences, 17 showed binding activity, and 3 of them demonstrated strong binding ability.
  • The platform can simultaneously conduct the evaluation of development potential during the candidate screening stage, thereby enhancing the efficiency of selection.
  • JotScreen is suitable for the efficient screening of BCR-seq libraries and AI-generated libraries and has good project adaptability and practical application value.

Case Study: Synergistic application of JotStruct + JotScreen

Screening of 10,000 nanobodies designed by JotStruct using JotScreen

Synergistic application of JotStruct and JotScreen case study workflow

JotStruct & JotScreen can jointly complete the entire process from candidate generation to screening optimization.
The platform identifies early molecules with cross-species binding potential from a large number of design candidates and obtains the candidate molecule LT01 through subsequent optimization.
The experimental results show that this molecule exhibits better in vivo efficacy compared to the control molecule, verifying the practical value of the platform's collaborative application.