Response to Comment on “Evidence of Abundant Purifying Selection in Humans for Recently Acquired Regulatory Functions”

Green and Ewing propose corrections to our methodology, which we incorporate and extend here. The improved methodology supports our initial conclusion of extensive lineage-specific constraint concentrated in ENCODE elements. We clarify that our estimate is dependent on the constrained and neutral references used, which can further increase the number of nucleotides involved, because a particularly stringent definition was initially used. In our initial Report (1), we found reducedgenetic diversity at noncoding genomic re-gions that have not been conserved across mammals but are biochemically active, suggest-ing that some fraction of these regions has ex-perienced lineage-specific purifying selection, and proposed a method for estimating the pro-portion under human constraint (PUC). Green and Ewing (2) suggest that our PUC estimate is inflated by several technical artifacts and that a reduction in SNP density cannot be reliably dis-tinguished from a reduction in mutation rate. I