Seminar Summary - Many mutations that occur in genomes are deleterious, resulting in a decrease in reproductive fitness. While such mutations are important for evolution and complex traits, much remains to be discovered. In this talk, I will discuss our work on estimating fitness effects of deleterious mutations using genetic variation data from natural populations. First, I will describe our computational approach, FitDadi, to estimate the distribution of fitness effects (DFE) for amino acid changing mutations in humans and other species. Analysis of how the DFE differs across species provides insights about how the DFE itself evolves. I also will describe our work inferring the DFE of mutations in noncoding regions of the human genome. Next, I provide a strategy to validate population genetic models of selection and dominance by testing whether they predict observed levels of inbreeding load and match observed genetic data. Lastly, I will discuss how insights gained from studying deleterious mutations may help identify small populations at risk of inbreeding depression.