I have question about MPO multi-parameter optimisation in drug discovery.
My opinion is that designed molecule’s biological activity and ADMET profile are not always correlate.
So, to design molecules that has good profile, I think it is important to get data about molecules not only biological activity but also phys-chem and ADMET profile even if a molecule does not have high potency about primary target.
Water-flall cascade system, for example check biological activity at first, then check in vitro phy-schem, and ADMET profile, next in vivo PK….. works good for filter low potency molecules but we lost chance to get any molecular property information. And also need more time I think.
I proposed parallel approach that check biological activity and HTS PADMET data at the same time.
Is that system waste of resources ?
I know it’s difficult to do MPO in drug discovery, but we have to do that.
I’m still finding the answer….
Hmm…
Is life worth living? 
I personally like the concept of ligand efficiency as an answer to this question. Efficiency metrics combine potency and molecular properties to have a unifying metric for MPO problem.
There is a nice explanation here:
https://caz.lab.uic.edu/discovery/index.html