By Diagnostics World Staff
February 6, 2026 | PharosAI, a research consortium of four UK institutions, plans to use the 10x Genomics Xenium spatial platform to build one of the world's most comprehensive multimodal cancer datasets and pair them with advanced AI models and analytical capabilities. They plan to democratize access to these datasets, ensuring they are securely available to a broad community of researchers and innovators.
Backed by £18.9 million from the UK Government's Research Ventures Catalyst and additional support from charities and industry partners, PharosAI will deliver a nexus of large-scale, AI-ready datasets, computational models and analytical tools to accelerate earlier diagnosis, enable precision therapies and speed the discovery of new treatments. The program will run through 2027.
PharosAI brings together King's College London, Queen Mary University of London, Guy's and St Thomas' NHS Foundation Trust and Barts Health NHS Trust in a joint venture that combines clinical expertise, academic leadership and advanced technologies. By leveraging 10x Genomics' Xenium spatial platform, PharosAI intends to transform decades of archived NHS cancer samples into high-resolution, multimodal datasets that could be used by researchers, clinicians, and innovators worldwide to power AI models for diagnosis and drug discovery to improve patient care and outcomes.
"AI has the potential to transform cancer care, but its progress has been held back by fragmented and inaccessible data," said Professor Anita Grigoriadis, CEO of PharosAI and Professor of Molecular and Digital Pathology at King's College London in a press release about the project. "PharosAI changes that—creating rich, multimodal datasets and pairing them with cutting-edge AI models and analytical tools, powered by 10x Genomics' advanced spatial technologies to enable earlier diagnoses, more targeted therapies and better outcomes for patients. By uniting universities, NHS Trusts, charities and industry, we are building an ecosystem for breakthrough discoveries."
The initiative will launch with a major focus on breast cancer, before expanding to lung and pancreatic cancer, with plans to include thousands of clinical tissue samples. By combining genomic, transcriptomic, imaging and spatial biology data, and linking them with a suite of custom AI models and analytical capabilities, PharosAI will produce AI-ready datasets capable of revealing patterns that have remained hidden until now. With Xenium providing spatial insights and advanced AI models uncovering new biology, researchers will be able to ask bigger questions and move faster toward the discovery of new diagnostics and therapies.
To support a study of this scale, PharosAI required a spatial platform capable of delivering reproducible results across thousands of archival tissues. Xenium's high-throughput workflow, combined with the ability to generate custom-designed gene panels in collaboration with 10x teams, enabled the consortium to configure assays aligned with the molecular features most relevant to specific cancers.
"Turning biological insight into impact requires scaled generation of consistent, high-quality data," said Serge Saxonov, Chief Executive Officer and Co-founder of 10x Genomics in the same statement. "With Xenium, PharosAI can convert decades of archived NHS cancer samples into robust, high-quality, spatially resolved datasets. This foundation is what makes it possible to move earlier, more precisely and with greater confidence toward the next generation of breakthroughs in cancer research."