AI detects DNA repair defects in cancer cells

Researchers at King’s College London have created HRDPath, an artificial intelligence system that can identify signs of a defect in DNA repair within cancer cells. This development, published in the journal Nature Communications on October 21, 2026, may help tailor cancer treatments to the specific features of each tumor. However, the technology is still in the experimental stage and requires further research before it can be used in medical settings.

DNA repair weakness linked to cancer

Some cancer cells have a reduced ability to repair DNA damage, a condition known as homologous recombination deficiency (HRD). This weakness is often associated with mutations in the BRCA1 or BRCA2 genes, which are known to increase the risk of certain cancers. HRDPath aims to detect this condition by analyzing high-resolution digital images of tumor tissues, using patterns that indicate HRD without direct DNA analysis.

Focus on ovarian cancer

The study, led by Dr. Heba Sailem, a computational pathology expert at King’s College London, centers on high-grade serous tubo-ovarian carcinoma, a particularly aggressive type of cancer affecting the ovaries or fallopian tubes. This form of cancer is often difficult to treat and may benefit from more personalized approaches.

AI could improve cancer treatment

Dr. Sailem explained that HRDPath reveals hidden information in microscopic images of cancerous tissues, which could change how treatments are adapted for individual patients. By using AI to detect DNA repair issues, the method could make precision cancer treatment faster, more affordable, and more widely available to women affected by the disease.