After the identification of a genetic locus associated with disease, the next step is to determine whether this variant alters the expression of transcripts within the region. In Panel A, a disease-associated region of the genome contains four genes. Although it is clear that genetic variability at a locus may affect distal genes (and even those on other chromosomes), in most instances the most proximal genes are investigated. In this case, expression of all four genes would be assessed, but one is shown for clarity. The single-nucleotide polymorphisms (SNPs) across this region form haplotypes that confer risk (red) or protection (blue) against disease. In Panel B, three splice forms are known to exist for the gene of interest: forms 1, 2, and 3, which differ according to their inclusion or removal of exons 3 and 5. In Panel C, genotyping of the risk variants in human tissues and expression analysis of the three splice forms in the same tissues allow a test of association between the genotype or haplotype and expression level. In this example, the risk haplotype is associated with decreasing levels of form 1 of the gene in the brain, has no measurable effect in the heart, and increases expression of form 3 in the liver. If no genotype-expression association were observed in the other three genes in the region, it would be reasonable to suggest that this is the pathogenically relevant transcript; if the disease of interest is neurologic, it would be reasonable to hypothesize that the risk is mediated by reduced levels of form 1 messenger RNA.