Current State-of-the-art for Virtual Screening and Docking Methods - New Developments in Medicinal Chemistry - page 13

confronted with challenges and limitations. Good solubility of fragments, high
quality of target protein and expensive detection equipment are required. The
theoretical approach can significantly improve the success and efficiency of lead
discovery and optimization. The protocol is used in parallel or independently with
experimental FBDD [150-229].
The key step in drug discovery is the identification of small molecules that
selectively bind to a biological target. Fragment based drug design (FBDD)
constructs novel lead structures from small molecular fragments taking
advantages of both random screening and structure-based drug design (SBDD).
FBDD uses fragment screening (hundreds to thousands of small low molecular
weight fragments) to identify weak binders of the desired target. The hit-to-lead
optimization process may involve a combination of fragment linking, fragment
evolution, fragment optimization as well as fragment self-assembly.
FBDD has advantages including generation of high chemical diversity, sampling large
chemical space, high hit rates and high ligand efficiency (LE = -logIC 50 /number of
heavy atoms). Challenges include coverage of a larger fraction of total diversity space
and suitable classes of targets. It is also necessary to take more into account ligand
specificity or selectivity, changes of geometries and key interactions of original
fragments upon evolving into lead compounds. Future trends should include more
efficiency in selecting proper linkers to bridge fragments, addition of adequate
fragments and prediction of binding modes of new molecules.
The computational FBDD starts with a fragment library design, which can provide
various filters, physicochemical properties, chemical diversity analysis, synthetic
accessibility, solubility predictions and fragment information. The fragment
screening stage can include pre-filtering of fragment library (fragment docking),
virtual hit identification (fragment docking), and hit expansion (substructure or
similarity search). The fragment optimization stage involves building fragment
hits into novel ligands ( de novo drug design), prediction of binding pose
(molecular docking, molecular dynamics), and prediction of binding affinity
(scoring functions). Further stages for candidate leads can include structure-based