- The collaboration aims to accelerate the adoption of scientific AI in the life sciences by producing AI-enabled use cases at an industrial scale.
- The work combines the foundational elements of scientific AI: compute, AI-native scientific data, domain-specific models, and scientific use case expertise.
- TetraScience aims to deliver scientific and economic impact across the life sciences value chain powered by the Tetra Scientific Data and AI Cloud and NVIDIA BioNeMo platform.
TetraScience today announced it is collaborating with NVIDIA to accelerate innovation in the $1.5 trillion life sciences industry by bringing standardization and scalability to the development and implementation of scientific AI.
The biopharma industry is racing to unlock AI’s potential across thousands of scientific use cases. However, millions of silos of proprietary and unstructured data and a project-based DIY approach block any ability to produce AI-ready datasets. Achieving success in scientific AI depends on eliminating silos and assembling the four essential components of a scientific AI stack: massive computational power, advanced models, sophisticated scientific ontologies, and deep scientific use case expertise.
TetraScience’s collaboration with NVIDIA makes such a reference scientific AI stack tangible for global drug companies looking to accelerate and optimize scientific workflows across the pharmaceutical value chain. The Tetra Scientific Data and AI Cloud provides a robust pathway to industrial-strength scientific AI through collaboratively optimized solutions with NVIDIA CUDA-X libraries, AI frameworks, and AI models. The Tetra Cloud is purpose-built to transform proprietary and unstructured scientific data into AI-native data and AI-enabled use cases. This solid and open data foundation will now tap into the value of NVIDIA’s accelerated computing infrastructure, domain-specific large language models, and deep learning capabilities.
“Scientific AI is the key to unlocking safer and more effective therapeutics as well as improving the economics of the life sciences industry,” said Patrick Grady, CEO of TetraScience. “By integrating TetraScience’s full-stack data, use case, and AI capabilities with the NVIDIA CUDA accelerated compute platform and domain-specific models, we intend to industrialize the production of AI-enabled use cases to help these vital organizations derive immediate scientific and business value.”
“One of the most significant ways AI can benefit the pharma and life sciences industry is by enhancing complex scientific experiments and lab data, reducing time and costs for managing them,” said Kimberly Powell, vice president of healthcare at NVIDIA. “We’re working with TetraScience to innovate and integrate generative AI and agentic workflows for knowledge extraction, relationship discovery, and reasoning, helping the industry get the most out of its data.”
TetraScience will work together with NVIDIA to engage with leading biopharmaceutical companies to establish a scientific AI use-case roadmap across drug discovery, development, manufacturing, and QC, prioritizing those with the highest impact. Lead clone selection, for example, is a critical but complex and time-consuming step in developing the cell lines that produce biologic drugs such as vaccines and monoclonal antibodies. Traditionally, it takes more than eight months to identify high-producing, stable clones. Models trained on AI-native Tetra Data could shrink that time by 80% by better predicting which are the “super clones.” TetraScience’s new Lead Clone Assistant data app uses the NVIDIA VISTA-2D model to analyze and classify hundreds of cellular images in minutes. It also applies the Geneformer model, accelerated and supported in the NVIDIA BioNeMo framework, to unearth gene expression differences between high-performing and problematic subpopulations, revealing details often missed in population-level studies.
Explore AITechPark for the latest advancements in AI, IOT, Cybersecurity, AITech News, and insightful updates from industry experts!
