A friend of mine works on simulating mouse organs. We had some pretty great talks over beer about modeling and cross model validation. Eventually he'd rant about all models are wrong but some models are useful.

If I remember correctly, they had some success in rejecting animal studies. Things like, we don't know what this will do overall, but we're pretty sure the kidneys or liver will fail like X, so don't bother with the wet lab. Or possibly try the wet lab at small scale so they can validate the model.

We're far far away from full simulation. But we know a whole lot about mice. I internalized those conversations as, my friend writes type checkers for drugs against mouse hardware.