Wedoany.com Report on Feb 7th, PharosAI and 10x Genomics recently announced a partnership for a cancer spatial data project. This initiative has received £18.9 million (approximately €21.8 million) in funding from the UK government's "Research Venture Catalyst" program, along with additional support from other partners. The program will initially focus on breast cancer research before expanding to lung and pancreatic cancers, and is expected to continue until 2027.
This collaboration aims to make cancer data more suitable for computational research. PharosAI stated that its goal is not only to transform tumor tissue into structured datasets but also to integrate AI models with analytical tools, ultimately making the resources available to a broader research and innovation community.
The partnership will primarily utilize existing tumor samples stored in UK National Health Service (NHS) hospitals, analyzing them through 10x Genomics' Xenium platform. This platform allows for the direct measurement of gene expression within intact tissue sections, preserving the structure of cancer cells, immune cells, and surrounding tissue while capturing molecular readouts. Xenium supports reproducible analysis of large numbers of archived samples, and its custom gene panels can be aligned with the molecular profiles of specific cancers.
PharosAI plans to concurrently develop AI models and analytical tools, enabling the generated datasets to be queried and reused without extensive processing, thereby accelerating large-scale cancer research. The final datasets, models, and tools will form a shared research resource, providing secure access to external academic and industry users and supporting multiple future studies rather than a single predefined problem.
This initiative aligns with the current trend of building reusable cancer data infrastructure. Similar projects include Owkin's MOSAIC project, the Human Tumor Atlas Network (HTAN) launched by the US National Cancer Institute, and the UK's PathLAKE project. These efforts are all dedicated to integrating multimodal data to facilitate large-scale computational analysis.
As computational and AI methods mature, access to large, consistent, and well-structured datasets has become a key bottleneck. PharosAI's project seeks to address this challenge by combining NHS pathology archives with multimodal analysis and developing supporting tools. The project's success will depend on data quality, standardization, and the practical application of its outputs beyond the founding institutions.
