Human Antibody Framework Selection via Germline Database
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for humanizing non-human monoclonal antibodies, such as CDR grafting, rely on mathematical best-fit strategies that may not account for biological significance and interchain interactions, and often use mature antibody genes with somatic mutations, leading to potential immunogenicity and incomplete V-J combinations, which complicates the selection of optimal human frameworks for antibody adaptation.
Innovation Solution
A method involving the creation of a virtual human framework database using germline V and J genes to select human antibody frameworks based on similarity and compatibility, constructing a library of peptide sequences for framework regions, and selecting representative frameworks to generate human-adapted antibodies that preserve antigen-binding activity without requiring detailed antibody-antigen complex structure information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If mathematical best-fit strategies are used to select human antibody frameworks, then the selection process is simplified and can be performed without detailed structural information, but the biological significance and interchain interactions are not accounted for, leading to suboptimal antibody adaptation
Solution Approach 1:
The patent segments the framework selection process into multiple evaluation dimensions: individual framework similarity scoring, interchain compatibility assessment, and combined pairing evaluation. This segmentation allows the method to simultaneously consider both ease of selection through systematic scoring and reliability through comprehensive biological factor assessment, resolving the contradiction between simplified process and accurate outcome.
2Measurement precision
If mature antibody genes with somatic mutations are used as frameworks, then the frameworks may show higher sequence similarity to donor antibodies, but they introduce immunogenicity concerns and may contain incomplete V-J combinations
Solution Approach 1:
The patent changes the selection parameter from prioritizing maximum sequence similarity to optimizing a balanced score that considers similarity alongside immunogenicity risk and V-J combination completeness. This parameter change allows the method to select frameworks that achieve sufficient similarity for CDR grafting while avoiding the harmful effects of somatic mutations and incomplete combinations, thus resolving the contradiction between precision and safety.
3Manufacturing precision
If separate best-fit heavy and light chain acceptors are selected independently, then each chain can be optimized individually, but the paired chains may not fit together optimally to conserve binding activity
Solution Approach 1:
The patent merges the individual chain selection processes by introducing an interchain compatibility evaluation step that assesses how well selected heavy and light chain frameworks work together. This combination of individual optimization with joint evaluation ensures both manufacturing precision in chain selection and reliability in their functional interaction, resolving the contradiction between individual and paired optimization.
4Reliability
If comprehensive evaluation of all framework regions and interchain interactions is performed, then the most optimal human-adapted antibodies can be selected, but the complexity of the selection process increases significantly
Solution Approach 1:
The patent replaces complex manual evaluation mechanisms with a computational scoring system that automatically assesses framework similarity, interchain compatibility, and overall adaptation quality. This substitution maintains high reliability in selection by systematically evaluating all relevant factors while reducing process complexity through algorithmic automation, resolving the contradiction between comprehensive evaluation and manageable complexity.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Methods useful for human adapting non-human monoclonal antibodies are disclosed. The methods select candidate human antibody framework sequences from a human germline framework database.