Natera has announced a collaboration with NVIDIA to drive the adoption of AI foundation models across the spectrum of precision medicine. The Natera-NVIDIA AI collaboration will accelerate the development of large-scale, multimodal AI systems capable of analyzing complex medical data.
Under the agreement, Natera will combine its longitudinal and multi-time point in vivo datasets with proprietary machine learning models, while NVIDIA will provide high-performance computing platforms, accelerated software libraries, and AI frameworks. The companies aim to build the infrastructure needed to train and deploy large AI models that support more precise diagnostics, deeper biological analysis, and tools for personalized treatment decisions.
Natera’s AI platform is built on one of the world’s largest multimodal, longitudinal oncology datasets. Recent results show strong performance in immunotherapy response prediction and neoantigen discovery. NeoSelect™, Natera’s neoantigen prioritization algorithm, outperformed 25 established prediction tools. NeoPredict, which estimates immunotherapy response, showed more than a two-fold improvement in hazard ratios compared with tumor mutational burden, the current clinical benchmark.
The AI models are being applied to drug target discovery, therapy response prediction, and identification of new prognostic biomarkers. Since unveiling the platform, Natera has reported increased engagement from pharmaceutical partners, including new data collaborations and growth in its data services business.
“Our vision is to build a computational foundation for biological intelligence,” said Matthew Rabinowitz, co-founder and executive chairman of Natera. “By combining Natera’s extensive AI-ready longitudinal data and machine learning with NVIDIA’s computer and software stacks, we can train at unprecedented scale. This collaboration enables AI systems that interpret biology and begin to reason across different data types, marking a meaningful step forward in personalized diagnostics and therapeutic insights.”
Under the Natera-NVIDIA AI collaboration, the company will leverage several NVIDIA platforms and software, such as Parabricks for accelerating genomic workflows, BioNeMo for training at massive scale, and the NeMo Agent Toolkit for creating multi-agent AI systems to be used in research and clinical settings. Initial engineering results have demonstrated average run-time and cost reductions of 75.7% and 59.2%, respectively, using Parabricks and 45.5% faster training by integrating with NVIDIA Transformer Engine.
“AI is advancing every aspect of precision medicine, enhancing the speed and accuracy of biological discovery for better outcomes,” said Rory Kelleher, senior director and global head of business development for life sciences at NVIDIA. “Paired with NVIDIA’s accelerated computing and advanced AI frameworks, Natera’s unique data and platform is helping to unlock better diagnostics, deeper biological insights, and personalized treatments.”
Natera’s global testing and research operations generate large volumes of data each year, supporting the training of increasingly complex AI models. By combining these datasets with NVIDIA’s computing platforms, the companies aim to deliver more accurate, data-driven insights across precision oncology, organ health, and related fields.