3D QSAR pharmacophore model based on diverse IKKβ inhibitors.

The inhibitor kappaB kinase β (IKKβ) is a serine-threonine protein kinase that is critically involved in the activation of the transcription factor nuclear factor kappa B (NF-κB) in response to various inflammatory stimuli. IKKβ-selective inhibitors could prove useful for the treatment of inflammatory diseases. In the absence of structural information, a ligand-based approach can serve as an alternative to the virtual screening of large databases. We have developed a 3D QSAR pharmacophore model based on 23 IKKβ inhibitors with 3 nM ≤ IC(50) ≤ 50000 nM. A four-feature pharmacophore containing a hydrophobic (Hy) feature, two ring aromatic (RA) features, and a hydrogen bond donor (D) feature was constructed. It yielded a correlation coefficient of 0.93 with experimentally determined activity data, and a correlation coefficient of 0.77 with training set activity data. The best hypothesis, Hypo 1, was validated by estimating the activities of 136 compounds in a test set. As well as the correlation analysis and test set activity estimation, a Fisher's validation test was conducted at the 95% confidence level. The pharmacophore model's specificity and selectivity were determined in an exhaustive enrichment study.