Many diseases are multi-genetic in the sense that only certain combinations of SNPs (and associated genes) will cause the disease. The big technological challenge is that a huge number of SNP combinations must be tested to identify groups of patients with common SNP profiles. Nevertheless, we have succeeded in identifying certain SNPs causing bipolar disorder. This work has been carried out in a joint research project between Array Technology A/S and the Institute of Neuro Science in Copenhagen. Gert will present the results of the project and the use of nested arrays for representing complex SNP patterns – and some benchmarks with the new PEACH operator in Dyalog.