Exploring neurotherapeutic space: how many neurological drugs exist (or could exist)?

Since high-throughput screening of compound libraries (virtual or real) against druggable targets is increasingly being used to discover therapies for brain disorders, it is crucial to ascertain if such screening methods adequately explore 'neurotherapeutic space (i.e. the total number of molecules that are or could be neuroactive)'. We present an approach to providing an estimate of the size of neurotherapeutic space.

Molecular modelling and statistical calculations were used to determine the number of molecules, which exist or could exist, with the necessary physicochemical and structural properties to be neurologically active drugs.

Assuming eight fundamental types of drug-receptor interactions, five different functional groups per type of interaction and five different molecular platforms for each functional group array, we calculated the total number of molecules that could be contained within a 7 Å radius sphere, used to define neuroactive chemical space. This calculation revealed that there are 6 × 10(15) molecules that could be neurological drugs.

Clearly, when it comes to exploring neurochemical space, we are still in our infancy and conventional high-throughput screening provides only a very limited sampling of the neuroactive chemical space that is available to neurotherapeutic compounds.