Pharmacophore-based virtual screening and Bayesian model for the identification of potential human leukotriene A4 hydrolase inhibitors.

Leukotriene A4 hydrolase (LTA4H), an enzyme involved in the conversion of LTA4 to LTB4, is an emerging and important anti-inflammatory target. This study demonstrates the development of quantitative pharmacophore hypothesis and Bayesian model and their applications in identification of potential human LTA4H (hLTA4H) inhibitors. The best hypothesis with a high correlation coefficient value of 0.951 was validated using different methods including a test set containing 136 compounds. It was further used as a three-dimensional query in database searching to retrieve virtual leads for hLTA4H inhibition. Molecular docking study was employed to identify the compounds that bind the active site with high affinity. Developed Bayesian model suggested molecular features favoring and not favoring the inhibition of hLTA4H.