Science and Technology News

New AI Cancer Framework

August 15 , 2026 10 hrs 0 min 12 0

Researchers from S. N. Bose National Centre for Basic Sciences (SNBNCBS) and Ashoka University have developed an Artificial Intelligence (AI) framework to identify hidden cancer stem-like cells.
The new framework, ACSCeND (AI-based Cancer Stem-like Cell Profiler and Neoplasm Deconvoluter), identifies three developmental states: pluripotent-like, multipotent-like and unipotent-like.
It can analyse conventional bulk tumour RNA (Ribonucleic Acid) sequencing data, allowing hidden cell populations to be studied across thousands of patient samples.
Researchers applied ACSCeND to more than 25,000 tumour samples from databases including TCGA (The Cancer Genome Atlas) and PRECOG.
Tumours with highly potent, pluripotent-like cancer stem cells were associated with poorer survival, higher tumour recurrence and reduced response to immunotherapy.
The framework could help identify patients at higher risk of relapse and support the development of precision cancer therapies.