Bioptimus announced the launch of its Spatial Tissue Embedding Learning Atlas (STELA), a multinational initiative to create what it describes as the world's largest clinically linked spatial biology atlas. The Paris-based AI company partnered with 10x Genomics (NASDAQ: TXG) and Broad Clinical Labs for the project.

STELA aims to profile up to 100,000 patient tissue specimens across the United States, Europe, and Asia, focusing initially on oncology and immunology tissue. The initiative will use 10x Genomics' Xenium spatial transcriptomics platform as its foundational technology.

The atlas will integrate spatial transcriptomics, histopathology imaging, multi-omics data including genomics and proteomics, and longitudinal clinical records. Bioptimus states this represents approximately a 20-fold increase in scale over existing spatial biology atlases.

Participating hospitals and research institutions will contribute samples under standardized protocols and receive access to spatial characterization and foundation model capabilities in return. The data will support Bioptimus' M-Optimus model, which the company describes as a world model of biology.

"STELA is the fuel to power M-Optimus, allowing us to map the intricate interactions between cells and tissues, across indications, at unprecedented scales," said Jean Philippe Vert, co-founder and CEO of Bioptimus.

The collaboration with Broad Clinical Laboratories includes a multi-year agreement for generating spatial biology data at scale. Broad will leverage its high-throughput spatial biology capabilities to process biological samples, while both organizations will co-develop AI-driven quality control metrics and predictive tools.

10x Genomics CEO Serge Saxonov said the initiative will "generate foundational datasets that allow researchers to connect the underlying biology with disease outcomes." Bioptimus reports that its models are currently used by 16 of the top 20 pharmaceutical companies.