In a study recently published in Nature Communications, researchers from Berlin, Potsdam, and Jena present a new method for analyzing the epigenome. The machine-learning method identifies differentially methylated DNA regions without sample labels – a prerequisite for many existing algorithms. This makes it possible to identify previously hidden biological patterns as well as new subgroups of cells or diseases. Analyses of blood cells, pancreatic cancer, and brain tumors confirms known biological relationships and reveals new regulatory processes. The metilene3 method opens up new possibilities for better understanding disease mechanisms and identifying potential biomarkers.