Autonomous AI-Driven Protein Evolution
CYBELE is BGT’s proprietary AI system for autonomous protein and polypeptide discovery. Designed as a closed-loop evolutionary platform, it combines generative modeling, functional prediction, and empirical verification to optimize for high performance in novel bioactives.
Protein Evolution Architecture
At the core of CYBELE is a continuous feedback system composed of three integrated engines: generative design, functional discrimination, and experimental validation. Together, these components simulate accelerated evolutionary selection in silico and in vitro, enabling iterative refinement of candidate molecules.
Cybele | Generative Engine
Ceres | Empirical Validation System
Mors | Functional Selection Engine
Cybele generates novel peptide and protein sequences using advanced deep learning architectures, exploring expanded combinatorial space beyond natural templates to design structurally diverse molecules with targeted functionality.
Mors | Functional Selection Engine
Ceres | Empirical Validation System
Mors | Functional Selection Engine
Mors evaluates and prioritizes generated sequences through predictive modeling, applying computational selective pressure to identify candidates with the highest likelihood of structural stability and biological activity.
Ceres | Empirical Validation System
Ceres | Empirical Validation System
Ceres | Empirical Validation System
Ceres serves as the platform’s empirical validation framework, where selected candidates undergo structured wet lab testing and experimental data are integrated to refine predictive models and guide subsequent design cycles.
Bio Genetic Technology LLC.
8400 River Road, Ste 2A, North Bergen, New Jersey 07047, United States