Exome sequencing studies bring the promise of comprehensive testing of coding variation in an unbiased manner. However, we expect that initial studies will be underpowered, and we have highlighted a number of technical issues that could bias the interpretation and analysis of rare variant data, especially novel variants. We expect that thousands of exomes are going to be required to achieve sufficient statistical power to robustly detect associations of rare variation with complex traits. Issues we discussed in this Perspective are also relevant to future whole-genome sequencing studies, in which the analysis of protein-coding variation will remain the same as in the case of exomes.
