Informatics aiding drug discovery: ADME evaluation


William L. Jorgensen, Department of Chemistry, Department of Chemistry, Yale University, New Haven, CT 06520-8107
Computational tools have been developed for the rapid prediction of properties of organic molecules that are relevant to their potential as drugs. A small number of physically significant descriptors, especially surface area components and hydrogen-bonding potentials, are computed from an input three-dimensional structure. Simple linear regression equations have been developed from experimental datasets using these descriptors for the accurate prediction of a variety of properties including aqueous solubility, octanol/water partition coefficient, free energy of hydration, Caco-2 and MDCK cell permeabilities, serum protein binding, and brain/blood partitioning. Other developments include a rule-based system for prediction of primary metabolites, while a more quantitative approach has been used in the computation of pKa values (acidities). The algoritms have formed the basis of the QikProp program.

Prediction of Drug Solubility from Structure. W. L. Jorgensen and E. M. Duffy, Adv. Drug Delivery Reviews, 54, 355-366 (2002).


Herman Skolnik Award Symposium
2:00 PM-4:20 PM, Tuesday, August 24, 2004 Pennsylvania Convention Center -- 110A&B, Oral

Division of Chemical Information

The 228th ACS National Meeting, in Philadelphia, PA, August 22-26, 2004