To understand the biology of diseased organs researchers can use different types of molecular data. One of the biggest computational challenges at the moment is integrating these multiple data types. A new computational method jointly analyses different types of molecular data and disentangles the sources of disease variability to guide personalised treatment.