Development of an ensemble of topological models as a new strategy for the development of new anticonvulsant agents for the treatment of refractory epilepsy

ALAN TALEVI; MAURICIO E. DI IANNI; LUIS E. BRUNO-BLANCH

Eilat

Otro; 4th Eilat International Educational Course: Pharmacological Treatment of Epilepsy; 2011

The International League Against Epilepsy – Comission of European Affairs y The Hebrew University of Jerusalem

Recent research has shown a significant association between refractory epilepsy and over-expression of ATP-dependent efflux transporters from the ABC superfamily in the blood-brain barrier (BBB), epileptogenic tissue and astroglial cells [1,2]. Pgp is a multiple-domain transmembrane protein responsible for efflux of xenobiotics from many tissues associated to drug metabolization and elimination (e.g. liver and kidney) and barrier function (such as the gut walls or the BBB). Its over-expression due to genetic or environmental issues limits the amount of drug that reaches the Central Nervous System target cells, preventing achievement of therapeutic concentrations of the drug in the site of action. Moreover, Pgp is characterized by wide substrate specificity, thus being accountable for multiple drug resistance. Here we present the development of computational models based on topological descriptors to identify potential Pgp-substrates, aimed to be applied as an efficient secondary filter in virtual screening campaigns to identify new drugs capable of overcoming Pgp-mediated resistance. Receiver Operating Characteristic (ROC) curves show that combination of individual models, through data fusion, in a 3-model ensemble, allows attaining higher areas under the curve and an overall better behavior in terms of sensitivity or specificity.