We now tested ZIPBN on different percentages of zeros (0%, 2

We thank reviewers for their thoughtful comments and provide responses to, in our opinion, the most prominent ones. We agree that filtering out genes with >70% zeros makes the task easier. Causal sufficiency should have been stated explicitly. That being said, it will be very interesting to extend our current work in this direction. We agree that Chickering's paper is indeed very relevant to this paper and should MCMC iterations whereas Chickering's requires a large sample size.