CAR-T Target Pair Selection for AND/NOT-Gated Solid Tumor Therapy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current systems lack effective methods for identifying target pairs for AND-gated and NOT-gated CAR-T cells and generating preclinical candidates in a scaled manner, and there are no computational tools to accurately predict effective multi-target combinations for therapeutics using Bayesian statistics and antibody databases.
Innovation Solution
A system and method using a learning model to identify target pairs for CAR-T therapy by selecting targets highly expressed in cancer and lowly in normal tissues, considering co-expression, dependency, and antibody availability, and generating CAR sequences based on these pairs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional single-target CAR-T approaches are used, then the therapy can be implemented with simpler design, but the specificity and safety against solid tumors are insufficient
Solution Approach 1:
The patent segments the CAR-T therapy into multiple independent target modules (first target and second target) that can be independently selected and combined. This allows the system to evaluate multiple target pairs systematically using computational methods, thereby improving specificity and safety without overwhelming complexity through structured modularity.
Solution Approach 2:
The patent introduces computational tools and algorithms as intermediaries to systematically evaluate and select optimal target pairs. These computational methods mediate between the complex biological requirements for safety/specificity and the practical constraints of CAR design, enabling rigorous multi-target selection without proportional increases in design complexity.
2Measurement precision
If computational tools with Bayesian statistics are used to identify target pairs, then the prediction accuracy for effective multi-target combinations is improved, but the complexity of the computational system increases
Solution Approach 1:
The computational system performs self-service by automatically integrating multiple data sources (antibody databases, expression data, dependency data) and applying Bayesian statistics without requiring manual intervention. This automation maintains high prediction accuracy while managing computational complexity through integrated, self-contained algorithms.
Solution Approach 2:
The computational tool is designed as a universal platform that can evaluate multiple target pairs across different cancer types using the same Bayesian framework. This multi-functionality allows the system to maintain consistent prediction accuracy across diverse applications without proportionally increasing complexity for each specific use case.
3Adaptability or versatility
If multiple data sources including antibody databases are integrated, then the comprehensiveness of target pair evaluation is improved, but the data processing complexity increases
Solution Approach 1:
The patent merges multiple data sources (antibody databases, tumor expression data, normal tissue expression data, dependency data) into a unified computational framework. By combining these diverse data types into a single integrated system that applies Bayesian statistics, the approach achieves comprehensive evaluation while managing complexity through unified data processing architecture.
4Reliability
If AND-gated systems are implemented to improve target specificity, then the safety and efficacy are enhanced, but the difficulty of identifying suitable target pairs increases
Solution Approach 1:
The patent performs preliminary computational evaluation of potential target pairs before actual CAR-T therapy development. By using Bayesian statistics to pre-assess the suitability of target pairs based on multiple criteria (expression patterns, dependency, antibody availability), the system identifies promising candidates in advance, reducing the difficulty of finding suitable targets for AND-gated systems.
Data Source
AI summary
Exemplary systems, methods, and computer-accessible medium are provided for Chimeric Antigen Receptor (CAR)-T therapy. Thus, the exemplary systems, methods, and computer-accessible medium are provided that select a first set of targets highly expressed in cancer and lowly in normal tissues, select a second set of targets that are highly co-expressed in the cancer and lowly in normal tissues, generate an input set by combining the first and second set of targets with dependency data, single-cell heterogeneity, and antibody availability, determine, by a learning model, at least one target pair for use in CAR-T therapy, and apply a cell therapy based on the determined target pair(s).