Cancer tumors are like snowflakes—no two ever share the same genetic mutations. Their unique characteristics make them difficult to categorize and treat. A new approach proposed by Trey Ideker and his team at the University of California, San Diego, might offer a solution. Their approach, called network-based stratification (NBS), identifies cancer subtypes by how different mutations in different cancer patients affect the same biological networks, such as genetic pathways. As proof of principle, they applied the method to ovarian, uterine and lung cancer data to obtain biological and clinical information about mutation profiles. Such cancer subtyping shows promise in helping to develop more effective, personalized treatments.