Fragment‐Based Approaches in Virtual Screening

Abstract

IntroductionFragment-based screening (FBS) as a method to identify starting points for smallmolecule drug discovery has become increasingly popular in recent years [1]. As a complementary and alternative approach to the traditional high-throughput screening (HTS), FBS starts from molecules whose molecular weights typically range from 150 to 250 Da. This approach is more routinely undertaken in therapeutic areas such as infectious disease, for which conventional screening has usually resulted in low hit rates [2]. The theoretical basis regarding the benefits of the fragment linking approach has already been reported about 30 years ago [3]. Interestingly, the importance of molecular fragments has been first recognized and exploited by computational approaches [4][5][6], which share similar aspects with experimental techniques that have been developed afterward, such as structure-activity relationship by nuclear magnetic resonance (NMR) [7]. More recently, fragment-based drug discovery strategies have been developed using X-ray crystallography [8], NMR spectroscopy [9], surface plasmon resonance [10], mass spectrometry [11,12], substrate activity screening (where the fragments are substrates later converted into inhibitors [13-15]), and tethering [16,17]. Experimental techniques for fragmentbased drug discovery have been discussed in previous reviews that contain a large number of applications [1,[18][19][20][21][22]. Successful in vitro screening campaigns have been reported for several targets, and a nonexhaustive list includes b-secretase [10,18,23,24], several protein kinases [25][26][27][28][29], DNA gyrase [30], caspase [31, 32], anthrax lethal factor [33], and phosphodiesterase [34]. The advantages of FBS over HTS have been reported [1], the main benefit being that a larger degree of chemical space can be explored using fragment libraries. HTS handles compounds up to millions but this amount only covers a tiny amount of the chemical space accessible to the small druglike molecules, which is estimated in the range of 10 60 -10 100 [35,36]. Fragment-based approaches, however, allow sampling of a much larger amount of chemical diversity using a much smaller number of starting molecules. It has been estimated that the number of stable and synthetically Virtual Screening. Edited by C. Sotriffer

Evaluation of docking performance in a blinded virtual screening of fragment-like trypsin inhibitors

“…The use of fragment-based approaches in drug design and particularly in virtual screening has known an explosive development in recent years [1][2][3][4][5][6][7]. In some instances, this approach has been successfully used as a less expensive alternative to the NMR and X-ray crystallography techniques for the assessment of binding site druggability [8].…”