ChemPass AI

ChemPass AI™ is Evogene's proprietary generative AI technology engine designed for multi-parameter molecular design and optimization. Built to serve the pharmaceutical and agricultural chemistry industries, it bridges advanced AI-driven molecule design with practical, commercially viable product outcomes — helping organizations translate scientific potential into real-world success.

Underpinned by a foundation model trained on 38 billion molecules, ChemPass AI™ replaces slow, iterative trial-and-error discovery workflows with a streamlined, AI-first approach that simultaneously accounts for multiple molecular parameters. The result is faster, smarter drug and agrochemical development with a higher cumulative probability of commercial success.

Core Advantages

• Multi-parameter optimization: Unlike conventional AI engines, ChemPass AI™ optimizes potency, selectivity, stability, toxicity, scaffold design, ADME properties, and synthesizability all at once, yielding molecules with superior performance and real-world viability.
• Designed for potency and precision: By fusing advanced AI design with targeted experimental validation, ChemPass AI™ produces molecules with higher predicted efficacy and binding precision.
• Novelty beyond existing libraries: Built on a 38 billion-molecule chemical universe, ChemPass AI™ generates novel structures that provide a decisive intellectual property and competitive edge in pharma and ag-chem markets.

Molecular Parameters Addressed Simultaneously

• Potency and binding affinity
• Synthetic accessibility
• Solubility
• Chemical stability
• IP novelty
• Molecular weight
• Lipophilicity (LogP)
• Metabolic stability
• Hydrogen bond donors and acceptors
• Scaffold design
• Selectivity
• ML models (QSAR)

Discovery Workflow: How ChemPass AI™ Works

1. Target Preprocessing: Literature search, data preprocessing, and target modeling are performed to establish a strong biological foundation.
2. Hit Screening: Potential hits are identified using the PointHit module, with partner or internal expertise used to purchase hits and test activity and selectivity.
3. Hit-to-Lead: Elaborated analogs are identified using the ActiveSearch module, while partners purchase and test analogs for activity and selectivity.
4. Lead Optimization: Novel leads are generated using LeadOptGPT, followed by synthesis and experimental validation by partner or internal teams.

This closed-loop system dramatically shortens time-to-lead, lowers development costs, and delivers highly optimized molecules ready for downstream validation and commercialization.

Technological Collaboration with Google Cloud

• First collaboration — First-in-class foundation model: A proprietary generative AI model for novel molecular candidates was developed based on 38 billion structures, addressing multiple parameters simultaneously and achieving 90% precision, compared to approximately 30% in traditional GPT models.
• Second collaboration — Advanced AI agents integration: AI agents have been integrated into ChemPass AI™ using Google Cloud Vertex AI to decrease manual errors, accelerate design-make-test-analyze cycles, and enable scalable generation of novel small molecules with improved probability of commercial success.

ChemPass AI™ is deployed in collaboration with Google Cloud, enabling it to scale real-world innovation across the pharmaceutical and agricultural industries — from generative molecular design to autonomous discovery workflows. The platform is positioned as a next-generation solution for organizations seeking to accelerate and de-risk small-molecule discovery at scale.