I understand that the drug target in the two species need to be very similar, so their shouldn't be any sequence variation within the binding site. What else to you look at?
there are many factors to consider before extrapolating data from animals to human. they include body size, toxicokinetic and total regional deposition and deposit per surface area.
There is no way to really predict how things will go in humans! Certainly in vivo models are much better than just "jumping" form in vitro to patients (and this has been done!!).
There are so many possible confounding factors! I think you may look at plasma concentrations of your drug that have a biological effect, and then see if youc an achieve those plasma concentrations in humans.
Going forward one step at a time is the best way to do things!