Successful rate gain by combination of multiple techniques

The computational methodologies are usually divided in two main areas: Ligand-Based (LB) and Structure-Based (SB) approaches. The first one is used when the x-ray crystal of the biological target is not available and no computational model can be obtained by homology with known structures. The LB methods attempt to extract the relevant information from the chemical structure and, when available, even correlate this with the biological activity value. The needing in this case is that the starting data set should cover different chemical scaffolds, span among a wide range of consistent activity values, and consist in a quite large number of compounds. By these, pharmacophoric models can be developed and used as filters in a Virtual Screening campaign of e.g. commercially available compounds. On the opposite, when no bioactive ligands are known for a specific target but the x-ray crystal structure is known, the SB approach is used. In this case, the docking procedure is the most common approach, and consist in trying to put every potential virtual hit in the binding site of the target structure and estimate a score of interaction. Now, in the lucky case where both LB and SB approaches are possible, the combination of LB and SB methods in a workflow enhance the success rate of finding a new bioactive molecule.