Enaminones, the enamines of beta-dicarbonyl compounds, have been utilized for many years as synthetic intermediates and of late, in pharmaceutical development as anticonvulsants. We have synthesized a large variety of novel enaminones with anticonvulsant activities confirmed by the NIH Anticonvulsant Screening Program. These enaminones represent an attractive area for antiepileptic drug discovery using the cheminformatics approaches. This study employs 63 enaminones compounds with anticonvulsant activities on both rat and mice. Predictive and robust models were generated using extensive combinatorial QSAR workflow. Among all the methods applied, Random Forest performed better than other methods (SA-kNN, GA-kNN, SVM and PLS) with an overall R2 value for external validation set of 0.434 using CDK descriptors. These models will be further used for virtual screening of large chemical libraries to identify novel lead compounds. In addition, we explored the correlation between the dual set of animal activities using the advanced Multi-task Learning (MTL) algorithms.

Cheminfomatic modeling of enaminones anticonvulsants with dual set of animal activities

243rd National Meeting (2012)

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.