Derivation of QSARs using 3D structural models of protein-ligand complexes

Rebecca C. Wade

European Molecular Biology Laboratory, Meyerhofstr. 1, 69117 Heidelberg, Germany

With the accelerating pace of protein structure determination and the advent of structural genomics programmes, it will be more and more common to know the 3D protein structures of drug design targets. This will increase the need to be able to derive QSARs that are based on information in protein 3D structures. With the goal of deriving QSARs in an objective and quantitative way from 3D models of protein-ligand complexes, we developed the COMparative BINding Energy (COMBINE) Analysis method (1,2). In this talk, I will describe recent developments and applications of COMBINE analysis.

To carry out a COMBINE analysis, a series of protein-ligand complexes of known activity is modelled and energy-minimized. The molecular mechanics interaction energy between protein and ligand is computed for each complex and partitioned into terms according to the physical nature of the interaction (e.g. Coulombic and Lennard-Jones) and the location in each molecule (e.g. on a per residue basis). Then a PLS analysis is performed to derive a QSAR model identifying the most important energy terms for the activity. This QSAR can be used to make predictions for new molecules or for protein mutants whose complexes are modelled in the same way. In addition to intermolecular interaction energies, other energy terms can be treated in the derivation of the COMBINE QSAR, including changes upon binding in terms describing intramolecular energy, conformational entropy and desolvation energy. COMBINE analysis has been applied to derive QSAR models for enzyme inhibition (2-6), enzyme-substrate binding (7-8) and protein receptor-DNA binding (9).


(1) Wade,R.C., Ortiz,A.R. and Gago,F. Perspectives in Drug Discovery and Design (1998) 9, 19-34.
(2) Ortiz,A.R., Pisabarro,M.T., Gago,F. and Wade,R.C. J. Med. Chem. (1995) 38, 2681-2691.
(3) Ortiz,A.R., Pastor, M., Palomer,A., Cruciani,G., Gago,F. and Wade,R.C. J. Med. Chem. (1997) 40, 1136-1148,4168.
(4) Perez, C., Pastor, M., Ortiz, A.R., Gago, F. J Med. Chem. (1998) 41, 836-52.
(5) Pastor, M., Perez, C., Gago, F. J Mol Graph Model (1997) 15, 364-71, 389.
(6) Wang,T. and Wade,R.C. COMBINE analysis of influenza neuraminidase inhibitors, in preparation.
(7) Lozano, J.J., Pastor,M., Cruciani,G., Gaedt,K., Centeno,N.B., Gago,F. and Sanz, F. J. Comput. Aided Mol. Des. (2000) 14, 341-353.
(8) Kmunicek,J. Luengo,S., Gago,F., Ortiz,A.R., Wade,R.C. and Damborsky,J. COMBINE analysis of the substrates of haloalkane dehalogenase from Xanthobacter autotrophicus GJ10, in preparation.
(9) Tomic,S., Nilsson,L. and Wade,R.C. J. Med. Chem. (2000) 43, 1780-1792.

In ``Rational Approaches to Drug Design: 13th European Symposium on Quantitative Structure-Activity Relationships''
Eds. Holtje, H-D., Sippl,W. (2001), Prous Science S.A., Barcelona, pp23-28.