Virtual Screening

Abstract

Virtual screening can be considered as in silico analog to high‐throughput screening. It is a knowledge‐driven approach toward finding new lead structures for drug discovery. There are a large variety of different virtual screening approaches and strategies available today. They include ligand‐based approaches such as compound similarity searching, pharmacophore searching, and applying predictive models derived using machine learning techniques as well as structure‐based approaches such as docking and scoring that use information about the protein target structures. Triaging hit sets using target‐independent filters such as drug‐likeness and diversity is an additional component of virtual screening. Using different virtual screening approaches in synergy has become more prevalent. Virtual screening applications have been combined with modern drug discovery technologies such as fragment‐based screening, combinatorial chemistry, and scaffold hopping methods to increase their effectiveness as lead identification tools.