AI Tool SIDISH Identifies High-Risk Cells in Aggressive Cancers, Aiding Targeted Therapy Development
Researchers have developed an artificial intelligence tool named SIDISH, which stands for semi-supervised iterative deep learning for identifying single-cell high-risk populations. This tool is designed to identify small groups of cells that are most responsible for driving aggressive cancers. By focusing on these high-risk cells, SIDISH offers a more precise approach to designing targeted therapies, as it highlights which cells within a tumor are most strongly linked to poor patient outcomes. The tool was tested on pancreatic, breast, and lung cancers, successfully identifying high-risk cells using tumor samples from patients. SIDISH connects detailed single-cell data with patient outcomes, a challenge that has long existed in cancer research. The tool can also simulate how these high-risk cells respond to specific genetic changes, potentially identifying promising drug targets. 
