Cheminformatic models to predict binding affinities to human serum albumin

Models to predict binding affinities to human serum albumin (HSA) should be very useful in the pharmaceutical industry to speed up the design of new c ompounds, especially as far as pharmacokinetics is concerned-We have experi mentally determined through high-performace affinity chromatography the bin ding affinities to HSA of 95 diverse drugs and druglike compounds. These da ta have allowed us the derivation of quantitative structure-activity relati onship models to predict binding affinities to HSA of new compounds on the basis of their structure. Simple linear, one-variable models have been deri ved for specific families of compounds (r(2) > 0.80; q(2) > 0.62): beta -ad renergic antagonists, steroids, COX inhibitors, and tricyclic antidepressan ts. Also, global models have been derived to be applicable to the' whole me dicinal chemical space by using the full database of HSA binding constants described above. For this aim, a genetic algorithm has been used to exhaust ively search and select for multivariate and nonlinear equations, starting from a large pool of molecular descriptors. The resulting models display go od fits to the experimental data (r(2) greater than or equal to 0.78; LOF l ess than or equal to 0.12). In addition, both internal (cross validation an d randomization) and external validation tests have demonstrated that these models have good predictive power (q(2) greater than or equal to 0.73; PRE SS/SSY less than or equal to 0.23; r(2) greater than or equal to 0.82 for t he external set). Statistical analysis of the equation populations indicate s that hydrophobicity (as measured by the ClogP) is the most important vari able determining the binding extent to HSA. In addition, structural factors (especially the topological 6 chi (ring) index and some Jurs descriptors) also frequently appear as descriptors in the best equations. Therefore, bin ding to HSA turns out to be determined by a combination of hydrophobic forc es together with some modulating shape factors. This agrees with X-ray stru ctures of HSA alone or bound to ligands, where the binding pockets of both sites I and II are composed mainly of hydrophobic residues.