Network Bio, a startup building artificial intelligence models for use in personalized medicine, launched Aug. 19 with $50 million. The company also announced a partnership with Nvidia to build a model based on cell-free RNA sequences and a $30 million codevelopment and licensing agreement with an unnamed Fortune 100 healthcare company.
Investors include Section 32, Thiel Bio, Founders Fund, Breyer Capital, Blue Venture Fund, and JSL Health Capital, among others.
In a statement, Network Bio said it plans to use data from academic medical center biobanks to develop disease-specific AI models. The company has collaborations with Mass General Brigham, the University of Pennsylvania, and the University of Colorado Anschutz.
“No single academic medical center can capture the full complexity of human disease,” Kimberly Muller, chief innovation officer at CU Anschutz Innovations, said in a statement. “By bringing together biobanks from leading institutions with industry partners, Network Bio is creating a research resource that reflects diverse patient populations and supports discoveries that can translate into real-world clinical practice.”
Network Bio is based in Palo Alto, California and cofounded by Hani Goodarzi, a researcher at the Arc Institute, and CEO Asad Ali Ahmad and Chief Business Officer Raphael Potter, both formerly of nference, an AI-based clinical data software company.
Network’s Nvidia collaboration aims to build the first foundation model trained on cfRNA.
Network will combine its datasets from patient-derived tissues and paired blood samples linked to longitudinal clinical outcomes with Nvidia’s accelerated computing, Parabricks genomic software suite, and BioNeMo Recipes for model training.
Specifically the company plans to “scale training of Nexus, a self-supervised transformer trained on cfRNA expression profiles.” It’ll also try to speed up cfRNA bioinformatics, including secondary analysis, alignment, quantification, and quality control.