for many things there is simply no good animal model so you are going to need to alter somethings DNA to get a good approximation of what's going on. Sure you

for many things there is simply no good animal model so you are going to need to alter somethings DNA to get a good approximation of what's going on. Sure you can also alter the DNA of a primate but the downsides are A) we know less about primate systems than mice and B) they cost a lot more. This is not to say a primate or even a zebra finches may not be a better animal model for many diseases, just that we share a lot of DNA with mice so there is a lot of relevant information we can gain from continuing to study them. Which when combined with their low cost means it's reasonable to spend more time and money looking at mice than our closer relatives. PS: There is a vary vary large but finite set of useful drugs we can discover. As a first pass approximation being able to consider 30x the candidates at the cost of doubling our false positive and false negative rates is probably worth it (for now). Eventually, I suspect we will be able to accurately simulate many diseases in software based on human DNA or even an individuals DNA at which point we may be able to revisit a lot of these false negatives for lower cost. Between now and then we will probably start looking at a larger array of animal models based on our understanding of an ever more detailed understanding of DNA. But, until we can actually debug diseases the old shotgun approach has a lot of merit.