Pharmacophore modelling: A continuously evolving tool for computational drug design

In the latest two or three years progressive applications of pharmacophore modeling continue to appear in literature. Pharmacophore based parallel screening, for instance, has been introduced in 2006. Moreover, in 2008, a survey discussing the prospective impact of virtual screening techniques in the discovery of bioactive natural products has been published. Finally, virtual screening techniques from the drug discovery field are beginning to be used for profiling the bioactivity of chemicals(especially those of potential environmental concern) with the aim of prioritizing compounds for further testing using more complex systems and reducing and ultimately replacing the use of animals in regulatory testing. Pharmacophore modeling might be extremely helpful to allow full achievement of all the above mentioned goals. In this contribution we report a couple of case studies where pharmacophore generation and handling played a pivotal role. In particular, in the first example, the development of a novel computational pre-screening approach to be used as an in silico filtering tool for natural products is described, applied to the estrogen receptor-? subtype. In the second study, differently, the validation of a preexisting pharmacophore by the prediction of the antifungal activities of new azole compounds is discussed. In this case, it comes to light the importance and utility of adding excluded volumes to a pharmacophore, to increase its predictivity.