NOVA (SN68) Developed Drug Optimization Agent with Chutes’ Security and Onepot’s Chemistry

Metanova Labs announced the development of a Hit-to-Lead Optimization Agent on NOVA (SN68).

Built with Chutes (SN64) and Onepot AI, the system is designed to help researchers refine promising molecules while protecting confidential research data.

The system incorporates an algorithm developed by a winner of NOVA’s Blueprint competition, which rewards better strategies for exploring chemical space.

It brings an output from Bittensor’s open competitions into proprietary drug discovery workflows.

Finding a molecule that binds to a therapeutic target is only an initial step. Researchers must improve its activity, limit interactions with unintended targets, and assess properties such as solubility and ease of synthesis.

The agent supports this process by helping users compare strategies, explore chemical modifications, and prioritize candidates for testing.

Onepot AI provides access to proprietary databases spanning billions of possible compounds and an automated synthesis platform.

According to the announcement, selected molecules can be synthesized and delivered within five to seven business days, helping researchers move from computational predictions to laboratory testing.

Chutes (SN64) provides confidential inference through verified Trusted Execution Environments. These hardware-isolated environments, combined with encrypted communication and controlled access, are designed to protect sensitive information such as compound structures, targets, and optimization strategies.

Metanova plans to integrate the agent into its internal hit-to-lead program. The agent gives NOVA’s competition-winning algorithms a direct path into real-world drug development.