Generalizable AI predicts immunotherapy outcomes across cancers and treatments

Wanxiang Shen / nature - Nature Medicine, Published online: 03 July 2026; doi:10.1038/s41591-026-04502-7COMPASS is a pan-cancer foundation model that predicts immunotherapy response, across cancer types and treatments, from bulk tumor transcriptomes.

AI Summary: Researchers unveiled a generalizable AI tool that predicts which patients will respond to immunotherapy across multiple cancer types and treatment regimens. By integrating diverse clinical and molecular data, the model helps stratify likely responders and could reduce exposure to costly, ineffective checkpoint therapy—promising smarter patient selection, faster trials, and fewer frustrated oncologists.

#biotech #drugdevelopment #digitalhealth #oncology #cancerresearch #clinicaltrials #healthit #diagnostics

35 hrs / medicalxpress


Back to Top / Sat, July 4, 2026, 7:21 pm / permalink 26388 / 2 stories in 2 hrs /




Related Stories


ALTAIR trial probes ctDNA‑guided post‑adjuvant therapy in CRC / 21 days

Deep‑learning pathomics platform may predict immunotherapy response in lung cancer / 2 months

Adjuvant selpercatinib sharply cuts recurrence in RET‑fusion lung cancer / 4 wks

DESTINY‑Lung03: T‑DXd shows activity in HER2 lung cancer; triplet stalls / 5 wks

Single‑cycle pembrolizumab yields striking responses in dMMR colon cancer / 7 wks

AI face photos may predict cancer survival years earlier / 2 months

EMERALD‑3 tests tremelimumab/durvalumab combos for embolisation‑eligible HCC / 20 hrs

StackHealth RSS


You can now follow topics by RSS - browse the complete list of topics, people, and organizations. Or, try Wildlife, FDA, Healthcare, Weight Loss and look for the RSS link.





NorthFeed Inc. Terms and Conditions / Privacy Policy

Disclaimer: The information provided on this website is intended for general informational purposes only. While we strive for accuracy, we do not guarantee the completeness or reliability of the content. Users are encouraged to verify all details independently. We accept no liability for errors, omissions, or any decisions made based on this information.