Find related Matched Pairs
A pair of molecules are Matched Pairs if they differ in structure at a single substitutable position; that is, they have the same scaffold but different R groups. The concept of Matched Pairs has several desirable features which have made it increasingly popular in recent years. It builds on the idea that changes in activities or physical properties are easier to predict than their absolute values. Unlike traditional cheminformatics similarity measures, such as fingerprints or distance in descriptor-space, Matched Pair similarity is straightforward, interpretable and in line with traditional MedChem thinking.
MPSearch rapidly finds all molecules in a database that are related to a query by a Matched Pair relationship. Given a database of chemical structures, the software generates a Matched Pair index according to user-specified criteria. After this, searching the index for molecules that are Matched Pairs of a query takes a fraction of a second.
Aside from activity analysis and prediction, MPSearch could also be useful for synthesis planning. If several analogs of the same series need to be synthesised, MPSearch will retrieve all analogs of the query previously synthesised without the user having to specify all of the different R group positions.