It would be great to have a feature that allows the user to specify a distribution from which the selection coefficients for a gene are drawn. This is different from the distribution of selection coefficients across sites, as the selective effect would be constant across the region, just variable between regions. In particular, if you want to simulate 10,000 genes (independently, say), with each gene having a fixed selection coefficient drawn from a normal distribution with mean 0 and variance 10. Ideally, one could specify up to 3 distribution (drawing from one of the distributions with a specified probability).
Anonymous
