During the last few years, most major disease phenotypes have been studied through genome-wide association mapping. This has been a remarkably successful enterprise, resulting in the discovery of more than 500 validated SNP-phenotype associations. Yet, these associations do not explain all the heritability of many common genetic diseases and it is not at all clear how they exert their effect at the cell level. Furthermore, current genome-wide association studies are designed to only find common polymorphisms associated with diseases, and they have very limited ability to detect rare variants and variants that interact with each other to cause a given disease phenotype. In the near future disease mapping projects will have access to full genome sequencing. Our limitations will therefore not be lack of information about the genetic differences but our ability to analyse very high dimensional data in a statistically powerful way, manage the data, and last, but not least, to interpret these in relation to complex phenotypes. Next generation disease mapping therefore calls for methods that can handle complete genetic information and relate it to complex biological information on the disease phenotypes. We propose to do research into new methods that can find more variants contributing to disease by explicitly modelling their interaction and combine the statistical signal contributed by several rare variants. To do this we need methods that can handle the large scale sequencing data and we need systems level approaches that make explicit models of each phenotype (disease) under consideration. The outcome should be a more thorough understanding of the cellular changes that lead to disease and how certain genetic variants contribute.