New approach to decoding mutational signatures improves prediction of prognosis in ovarian cancer

As cancer develops, it leaves behind tell-tale marks on the DNA of tumour cells, revealing how the disease progresses and how it might respond to treatment. However, some of these patterns—known as mutational signatures—are hard to detect, which limits their use for predicting patient outcomes or deciding on the best therapy.

Now, researchers at IRB Barcelona, led by Dr. Fran Supek, Group Leader at IRB Barcelona and Professor at the Biotech Research & Innovation Centre (BRIC), University of Copenhagen, have developed a new approach that combines two kinds of genetic changes to better identify faults in the DNA repair systems of cancer cells. Published in Genome Medicine, the method improves predictions of survival in ovarian cancer and may also apply to other tumour types.

“This work is about reading the history written in the DNA of cancer cells more precisely, so we can understand why some patients respond better to treatment than others, and help doctors make more informed decisions”, says Dr. Fran Supek.

Looking deeper into cancer’s DNA patterns

Instead of focusing on just one type of genetic change, as most studies have done until now, researchers Patricia Ferrer‑Torres, Dr. Iván Galván‑Femenía and Dr. Fran Supek analysed two kinds simultaneously: tiny DNA letter changes and small insertions or deletions. By integrating this information, they gained a much clearer view of the genetic scars left by faulty DNA repair processes.

The study looked at data from more than 380 whole-genome sequences of ovarian cancer collected from four international cohorts. Using their new method, the researchers were able to spot distinct DNA patterns linked to how well patients responded to standard treatments like platinum-based chemotherapy.

Better predictions for patients—and wider potential

One of the key findings was that these DNA patterns could more accurately predict which ovarian cancer patients were likely to have longer survival, outperforming other commonly used genetic markers.

Importantly, the same approach also showed promise in other cancers, including pancreatic, breast, and prostate tumours. “What makes this method powerful is that it doesn’t rely on any prior knowledge of the tumour’s gene alterations—it learns directly from the data, meaning it could be applied to many different types of cancer,” reflects Patricia Ferrer‑Torres, first author of the study and a PhD student at IRB Barcelona.

This research highlights how looking at the cancer genome in a more integrated way can provide valuable tools for personalised oncology. By better identifying DNA repair problems in tumours, doctors could more precisely tailor treatments and determine who might benefit most from certain therapies.

Related article:

Joint inference of mutational signatures from indels and single-nucleotide substitutions reveals prognostic impact of DNA repair deficiencies
Patricia Ferrer-Torres, Iván Galván-Femenía & Fran Supek
Genome Medicine (2025) DOI: 10.1186/s13073-025-01497-7