In my latest preprint on BioRxiv, I take a look at how interactions and interference between selected mutations shape patterns of diversity and linkage disequilibrium and whether we can learn about those interactions from population genetic data. Here, I extend an approach based on moment equations for the diffusion approximation to include general selection models (epistasis, dominance, or any combination). This lets us explore how LD decays with recombination distance for any interactive model. I also look at LD in genic regions in human populations. The main takeaway is there’s a lot of heterogeneity of both selection strengths and interactions within genes, with opposing effects between missense mutations inside vs outside of annotated features and conserved elements. Plenty more analyses and discussion in the paper. Please reach out with any comments or questions!