AI model predicts breast-cancer drug responses in clinical test

September 9, 2026 at 05:00 AM1 min

This digest was compiled by AI from multiple sources — links to the originals are below.

Researchers developed an AI-based virtual cell model that predicts how triple-negative breast cancer cells respond to drugs. The model showed promising results in identifying effective drugs using biopsies from patients. The study was published in Nature on 9 September.

Key Facts

The AI model was trained on more than 38 million protein measurements from 18 breast cancer cell lines, 16 of which were triple-negative.
Researchers treated the cells with 63 FDA-approved antitumour drugs and 59 drug combinations.
The team measured levels of 5,585 protein groups before treatment and after 6, 24 and 48 hours of drug therapy.
Triple-negative breast cancer accounts for 15–20% of breast cancer cases.

Virtual Cell Model

The model uses proteomics data to simulate how individual tumour cells respond to various drugs. It was developed by researchers at Westlake University in Hangzhou, China, led by proteomics specialist Tiannan Guo. Guo stated that this is the first time a virtual cell model has been tested in a clinical scenario. The model has a focused goal in drug discovery for triple-negative breast cancer rather than a comprehensive simulation of cellular behaviour.

Training Data

The training data set included protein measurements from 18 breast cancer cell lines, 16 of which were triple-negative. The cells were treated with 63 FDA-approved antitumour drugs and 59 drug combinations. Protein levels of 5,585 protein groups were measured before treatment and at 6, 24 and 48 hours after drug therapy. Systems biologist Hani Goodarzi at the Arc Institute in Palo Alto, California, noted the unique scale of the proteomics data. Goodarzi added that collecting data at multiple time points is crucial to capture dynamic changes in cells.