AI breakthrough maps tumour cell diversity to personalise cancer treatment

Scientists in Australia and the U.S. have developed an AI tool that could transform cancer care by revealing the hidden diversity of tumour cells and guiding more targeted therapies.

A team of Australian and American researchers has unveiled a powerful new artificial intelligence (AI) tool designed to decode the complex biology of tumours and improve cancer treatment outcomes. The innovation addresses tumour heterogeneity—variations among cancer cells within the same tumour—that has long hindered the effectiveness of standard therapies.

Developed jointly by Sydney’s Garvan Institute of Medical Research and the Yale School of Medicine, the tool, known as AAnet, uses deep learning to analyse gene activity at the single-cell level. According to the researchers, AAnet can distinguish five distinct cell types within a tumour, each with unique behaviours and potential to spread, offering far more precision than traditional diagnostic methods.

“Heterogeneity is a problem because currently we treat tumours as if they are made up of the same cell,” said Associate Professor Christine Chaffer from the Garvan Institute, who co-led the study. “This means we give one therapy that kills most cells in the tumour by targeting a particular mechanism. But not all cancer cells may share that mechanism.”

By identifying the biological traits of different cancer cell populations, AAnet allows clinicians to tailor combination therapies that target all types of tumour cells at once, potentially reducing treatment resistance and relapse.

Co-developer Associate Professor Smita Krishnaswamy of Yale University described AAnet as the first tool capable of simplifying the complexity of tumour cells into actionable categories. She said it could pave the way for a new era of precision oncology where treatments are personalised to the unique cellular makeup of each patient’s tumour.

Validated initially in breast cancer, AAnet has also shown promise for other cancer types and even autoimmune diseases. The team plans to integrate AI insights with conventional diagnostics in clinical settings, offering personalised treatment plans based on the specific cell types within each tumour.

The findings were published in the journal Cancer Discovery, marking a significant step toward more effective, individualised cancer care.

Copied