Skilled use of DNA polymorphisms as a tool for polygenic cancers

Association studies are assumed to be an efficient method of deciding whether a gene or its variant is important for cancer. Sequencing data on 30 000 human genes suggest that an average gene contains one to two single nucleotide polymorphisms (SNP), and high through-put technologies have become available for fast genotyping. Because no functional data are available for most SNPs, the result of the large-scale genotyping effort will be a huge amount of data of unknown biological significance. We discuss here the approaches in study design and reporting that will reduce the spread of false positive data and optimize scientific progress in the genotyping field.