Terray Therapeutics Unveils EMMI: AI-Driven Platform Revolutionizes Small Molecule Drug Discovery

Terray Therapeutics has launched EMMI, a new AI-native drug discovery platform specifically designed for de novo small molecule drug discovery3.

EMMI (Experimentation Meets Machine Intelligence) integrates proprietary ultra-miniaturized hardware, a highly automated lab, and an industry-leading full-stack AI platform, enabling Terray to explore and optimize novel molecules rapidly3.

The platform utilizes advanced Generate, Predict, and Select models that allow scientists to evaluate millions of candidate molecules and choose the most impactful ones for experimental validation in each cycle of drug discovery3.

Terray’s AI models are powered by a foundation model for chemistry and leverage an in-house dataset of over 13 billion target–molecule measurements, generated via proprietary microarray technology3.

EMMI's selection AI models optimize which molecules to synthesize and test, reducing the time and cost of iterative drug optimization by 3x compared to industry standards3.

Terray published its first generative method using latent diffusion models directly linked to its COATI chemical embedding in 2024, marking a technical advance in generative chemistry3.

The core of Terray’s approach is the iterative loop between AI and experimental testing:
generated molecules are predicted and scored for activity, then selected for synthesis, with new experimental data feeding back to continually improve the models23.

Terray’s methods and platform are being deployed both in their internal pipeline—focused on immunology—and in cross-industry partnerships, such as their collaboration with Gilead for multi-target small molecule drug development13.

Terray believes that high-quality, scaled, precise data is essential for unlocking generative AI’s true potential in molecular design, and their platform is generating over 1 billion new experimental measurements per quarter23.

Industry observers cite Terray as a pioneer in generative AI-driven drug discovery, with its platforms significantly narrowing the gap between computational molecule generation and practical, actionable drug development112.

Sources:

1. https://www.gilead.com/news/news-details/2024/gilead-and-terray-therapeutics-announce-multi-target-research-collaboration-to-discover-and-develop-novel-small-molecule-therapies

2. https://www.nvidia.com/en-us/customer-stories/generative-ai-for-small-molecule-drug-discovery/

3. https://www.businesswire.com/news/home/20251117222688/en/Terray-Announces-EMMI-an-AI-Native-Drug-Discovery-Platform-for-de-Novo-Small-Molecule-Drug-Discovery

11. https://www.statnews.com/2024/12/03/who-to-know-series-biotech-companies-using-ai-small-molecules/