Cancer Grand Challenges names first AURORA teams

Cancer Grand Challenges describes AURORA as a $10 million program intended to help scientists speed up cancer discovery.[1] The initiative is led by Marinka Zitnik at Harvard Medical School.[1] The organization says its AI co-scientists will help researchers navigate data and evidence, generate ideas, and design and prioritize experiments.[1] The system draws on scientific literature, experimental data, curated knowledge bases and specialized research tools.[1] Cancer Research UK is funding AURORA through the Cancer Grand Challenges initiative, which it co-founded with the U.S. National Cancer Institute.[1]

Cancer Grand Challenges is providing $4 million to support the three selected teams.[1] In March 2026, the initiative awarded $6 million in Cancer Research UK funding to a team led by Zitnik.[1] The organization says AURORA’s agentic system can plan and carry out tasks toward a research goal while adapting under human oversight.[1] Researchers will set the scientific direction, assess AI recommendations and decide which experiments to pursue.[1] Experimental findings will feed back into AURORA to help refine researchers’ understanding.[1]

The NexTGen team will use AURORA to identify and prioritize CAR-T therapy strategies, including targets, antigen combinations and ways to overcome treatment resistance.[1] PROMINENT will investigate why cells carrying cancer-driving mutations can remain normal, and what causes them to become cancerous.[1] MATCHMAKERS will develop interacting AI co-scientists to identify tumor antigens likely to prompt an immune response and T-cell states able to attack tumors.[1] Together, the projects put the system to work on different questions in cancer biology and treatment research.[2][1]

The three teams were selected after scientific review by a Cancer Grand Challenges committee and external AI experts.[1] The review considered scientific importance, fit with AURORA’s objectives, human-AI interaction, and routes to experimental or clinical validation.[1] Cancer Grand Challenges says feedback from the teams’ experiments will be used both to improve the shared AI system and to evaluate its performance.[1]