BioGeometry Unveils Next-Gen Drug Discovery Engine and Suite, Achieving Tangible Results in AI-Driven Biotech
6 hour ago / Read about 0 minute
Author:小编   

AI protein design is moving beyond "can it bind" to enabling real R&D outcomes with drug-like properties. BioGeometry has launched its next-generation drug discovery engine, GeoFlow V4, along with the agentic drug design orchestration system, GeoFlow Code. GeoFlow V4 achieves international frontier performance in antibody-antigen structure prediction, and its cross-modal confidence scoring enables precise screening of high-quality candidate molecules. GeoFlow Code, an autonomous orchestration system for protein design, can independently plan multiple rounds of exploration and optimization around R&D goals, greatly improving success rates and sequence diversity in complex design tasks. Across 13 publicly disclosed disease-related targets and 15 de novo design tasks, the average wet-lab hit rate reached 34.6%, with positive results obtained across multiple molecule types, including small proteins, cyclic peptides, nanobodies, and IgG antibodies. Many candidates achieved picomolar-level affinity and performed strongly on early developability metrics such as expression level and thermal stability, forming a dry-wet feedback R&D loop that connects "model–molecule–data" and propelling AI pharma from a single-point algorithmic tool toward an industrially operable R&D system.