Children with autism do not overimitate

population expansion is that most alleles are rare, and are not interrogated by standard commercially available microarrays. The full extent of what is missed became apparent from recent population scale re-sequencing projects: only 13% of variants with a frequency of less than 0.5% had been described previously. If rare variants make a substantial contribution to your disease of interest, beware! GWAS won’t find them. You may also have read that GWAS doesn’t work because GWAS loci cannot account for much of the known or estimated heritability of a trait (‘missing heritability’). For instance, despite finding 180 loci that influence height, these loci account for just 10% of the variation. But, this does not take account of all those SNPs that don’t make the significance threshold. They can’t simply be ignored, but what to do with them? Peter Visscher has an answer, using an approach routine in plant and animal genetics. Examining the effect of all SNPs, regardless of statistical significance, almost half of height’s phenotypic variance can be explained by common SNPs. So is there a ‘missing heritability’ problem? Well, we still can’t explain all the variance.