Evogene Completes First-in-Class AI Foundation Model for Small Molecule Design with 90% Precision
- Evogene Ltd. has completed its generative AI foundation model version 1.0 for small molecule design, developed in collaboration with Google Cloud, achieving approximately 90% precision compared to 29% in traditional GPT AI models.
- The model addresses the core challenge of identifying novel small molecules that simultaneously meet multiple complex product criteria while being patentable, essential for both pharmaceutical and agriculture applications.
- Built on a dataset of approximately 38 billion molecular structures and trained using Google Cloud's advanced AI infrastructure, the model expands ChemPass AI capabilities to generate truly novel molecular structures.
- Development is already underway on version 2.0 with enhanced flexibility for multi-parameter optimization, incorporating customized parameters tailored to therapeutic contexts and specific agriculture requirements.