Mice breed fast and die quick, which lets you run a lot of trials much faster than you could in humans or primates. They're also "close enough" that most treatments will have similar effects in both - most treatments that work in mice but not in humans usually fail due to bad side-effects, or due to having the same direct effect as in mice but that not causing the follow-on effects that are what is actually desired.
There are vastly more compounds and potential treatments that are harmful or do absolutely nothing than there are that are beneficial - being able to filter out nonworking treatments quickly is very important to actually finding the stuff that does work.
> most treatments will have similar effects in both
If so, that would answer the question and undercut what I was responding to.
> being able to filter out nonworking treatments quickly is very important to actually finding the stuff that does work.
Assuming that the filter is accurate. If not, we could close avenues that would be fruitful and open avenues that will be not, wasting decades and billions of dollars.
A filter doesn't need to be perfect to be worthwhile - just to pull some made-up numbers out of nowhere, if 1/100 options are worth investigating, and your filter gives you true positives 70% of the time and false positives 30% of the time, using that filter gives you about twice as much of a return on your investment as skipping the filter and investigating everything.
There are vastly more compounds and potential treatments that are harmful or do absolutely nothing than there are that are beneficial - being able to filter out nonworking treatments quickly is very important to actually finding the stuff that does work.