Researchers at the University of Navarra have developed RNACOREX, a new open-source software tool that helps uncover the complex genetic networks driving cancer. Designed to identify regulatory interactions between molecules such as microRNAs (miRNAs) and messenger RNA (mRNA), RNACOREX offers new opportunities for cancer survival analysis and precision oncology.
The software was created by scientists at the Institute of Data Science and Artificial Intelligence (DATAI), part of the Cancer Center Clínica Universidad de Navarra, and validated using data from 13 tumor types included in The Cancer Genome Atlas (TCGA). The study describing RNACOREX was published in PLOS Computational Biology.
From big data to interpretable biology
Modern cancer research generates enormous volumes of genomic data, but translating that data into clinically meaningful insights remains challenging. Conventional analytical approaches often miss subtle yet critical molecular interactions or produce results that are difficult to interpret.
RNACOREX addresses this gap by simultaneously analyzing thousands of molecules and ranking the most biologically relevant miRNA–mRNA interactions. By integrating curated international databases with real gene-expression data, the software builds increasingly complex regulatory networks that reflect how genes interact within tumors.
According to Rubén Armañanzas, head of DATAI’s Digital Medicine Laboratory and one of the study’s lead authors, understanding these networks is essential. “The breakdown of regulatory communication between molecules is a hallmark of cancer. RNACOREX helps distinguish meaningful biological signals from noise, something that has been notoriously difficult with existing tools,” he explains.
Predictive power with transparency
To evaluate its performance, the research team applied RNACOREX to genomic and clinical data from cancers including breast, colon, lung, stomach, melanoma, and head and neck tumors. The results showed that RNACOREX can predict patient survival with accuracy comparable to advanced AI models.
The key difference, however, lies in explainability. “Many AI systems function as black boxes,” says first author Aitor Oviedo-Madrid. “RNACOREX delivers similar predictive performance, but also provides clear, interpretable explanations of which molecular interactions drive those predictions.”
Beyond survival prediction, the tool identifies regulatory networks linked to clinical outcomes, reveals molecular patterns shared across different tumor types, and highlights individual molecules that may be of particular diagnostic or therapeutic interest. This makes RNACOREX valuable not only for data scientists, but also for clinicians and biomedical researchers seeking actionable insights.
Open source and future-ready
RNACOREX is freely available as open-source software via GitHub and the Python Package Index (PyPI). It includes automated database downloads to simplify implementation in research environments, lowering barriers to adoption.
As the use of AI and omics data in healthcare accelerates, the developers position RNACOREX as a transparent alternative to opaque machine-learning models. Ongoing development at the University of Navarra includes adding pathway analysis and additional interaction layers to further improve biological interpretability.
The project reflects a broader trend toward interdisciplinary collaboration, combining biomedicine, artificial intelligence, and data science to advance personalized and precision cancer care.
