Many diseases such as cancer, diabetes and heart disease have a genetic component. We have recently developed methods to define high level genetic interactions in these diseases. This allows us to define “genetic signatures” that predict a person’s risk of developing disease, or if they already have the disease, the outcomes of that disease. For example, we can define a group of people at high risk of having a heart attack or stroke.We can also diagnose people who, after diagnosis of melanoma, will be likely to die within five years (Fig 1); or people with diabetes who are likely to develop complications such as heart disease or kidney failure. We would like to extend these methods to other diseases, such as breast cancer, and to other disease outcomes of cardiovascular disease and diabetes. The limitation in us doing so is the computing power we have available.