Abstract

Over the recent years, genome-wide association studies provided new insights into the genetic architecture of multiple complex multi-factorial disorders. A natural next step to follow-up these findings at the molecular level is to correlate disease associated variants with quantitative gene expression RNA data. A widely used approach to achieve that goal consists of measuring RNA expression level in large samples of unrelated individuals with available genotype data. However, the large sample size requirement for this design may be impractical when dealing with tissue types or cell lines that are difficult to obtain in large quantities, for example brain tissues or induced pluripotent stem cell lines. A dual approach to answer the same question that is less limited by sample size requirement is allele specific expression (ASE). ASE correlates genetic factors with uneven RNA gene expression of both haplotypes for nearby genes. Informative measurements are made within rather than between individual and are therefore unaffected by the between individual noise not explained by genetic factors. This elegant and powerful design maximizes the per sample information. Here, we show how the combination of high throughput DNA sequencing and sequence capture provides a powerful tool for quantitative ASE analysis for up to 200 target genes typically selected from genome-wide association analysis. The key challenge to increase the accuracy of this approach is a better understanding of the experimental and in silico biases generated by high troughput DNA /RNA sequencing.