Understanding how the DNA sequence influences gene expression across the different tissues of the human body is essential to unveil the molecular mechanisms of many disorders. The identification of genes that are specifically expressed by particular tissues has allowed us to develop a deeper molecular appreciation for how tissues and organs function, as well as to identify the contribution of tissues and cell types to particular traits and diseases. Although associations between genetic variants and diseases have been established by GWA studies, the causal tissues or cell types are still uncertain or unknown for many diseases. Combining our knowledge of tissue specific gene expression with GWAS data would provide the potential to identify relevant tissues and cell types, enhancing our knowledge of disease. In addition, as these diseases can be expected to affect fitness and survival to reproduction, they are one possible way by which evolutionary forces can shape the human genome; in this case these genes will be the mediating factors. More generally, our ability to read DNA means that we can now identify regions of the genome that show signs of evolutionary pressure, for instance genomic regions that have remained unchanged between different species that have diverged millions of years ago as mutations are highly deleterious; and regions of the genome where geography causes large changes, as the genome adapts to local conditions. This thesis aims to bring together all these aspects: which regions are specifically active in particular tissues, which regions are implicated in the development of particular diseases, and which regions have been under selection. In this way we hope to gain a better understanding of how selection has shaped the human genome. We examined the consequences of natural selection on gene expression across different human tissues and traits, as well as the heterogeneity of gene expression in a single tissue, the atherosclerotic plaque, and one disease, atherosclerosis.