Freenome has announced an expanded initiative to enhance its artificial intelligence and deep learning capabilities through a strategic collaboration with NVIDIA, aimed at advancing its blood-based multi-cancer detection tests. The partnership leverages NVIDIA’s accelerated computing technology and specialized AI frameworks to improve the speed, accuracy, and scalability of Freenome’s analytical processes. At the core of the initiative is the refinement and scaling of Freenome’s proprietary Fragment-Level Deep Learning (FLDL) model, which analyzes cell-free DNA (cfDNA) to detect early cancer signals. The model is designed to identify subtle, cancer-specific patterns in blood samples, a task made difficult by the vast and complex nature of the data. A single blood draw can generate millions of tumor-derived cfDNA fragments and billions of DNA base pairs, creating significant challenges for traditional training methods. With NVIDIA’s high-performance computing infrastructure, Freenome can now overcome data loading and training bottlenecks, enabling faster model development and more efficient inference as test volumes grow. A key component of the collaboration is the development of an open-source foundation model focused on cfDNA methylation patterns. Using NVIDIA’s BioNeMo framework—designed specifically for biomolecular data—Freenome will build a model capable of understanding and contextualizing epigenetic signals in cfDNA. The resulting model will be made publicly available, allowing researchers across the genomics and oncology communities to use and build upon it, accelerating broader scientific progress in early cancer detection. Additionally, the partnership aims to empower healthcare organizations (HCOs) by providing them with AI-powered tools to analyze de-identified, real-world data collected during clinical trials. Freenome is working with NVIDIA to create accelerated data interrogation systems that enable HCOs to conduct advanced research using longitudinal, multimodal datasets. This will support the development of new diagnostic and prognostic tools, ultimately improving patient outcomes. Freenome’s first test, SimpleScreen™ CRC, a blood-based screening for colorectal cancer, is currently under review by the U.S. Food and Drug Administration, with a decision expected in the second half of 2026. The company also plans to launch additional blood-based tests for lung and other cancers in 2026, all built on a shared, automated laboratory workflow. Freenome’s approach combines multiomics data—spanning DNA methylation, RNA, proteins, and other analytes—with advanced AI to detect cancer at its earliest, most treatable stages. The company believes that no single technology can identify all cancers due to the disease’s complexity, but a multi-layered, AI-driven strategy offers a powerful path forward. The collaboration with NVIDIA is a critical step in scaling this vision and bringing life-saving early detection to millions.