A structure–activity relationship survey of histone deacetylase (HDAC) inhibitors

• A new analysis method (rank-based ant system feature selection (RBAS-FS) algorithm)
• A large dataset (about 400 molecules)
• High prediction ability
• Good correlation parameters
• HDACi is a hot topic in medicinal chemistry nowadays.

Histone deacetylase inhibitors have gained a great deal of attention recently for the treatment of cancers and inflammatory diseases. So design of new inhibitors is of great importance in pharmaceutical industries and labs. As synthesis is a costly and resource intensive process, estimation of the compound's property or activity before synthesis is desired. To this end, computational methods such as quantitative structure–activity relationships can be used to predict activity or properties of the molecules of interest.Here, we elaborated a novel rank-based ant system to generate a QSAR model for the prediction of histone deacetylase inhibition activity. The dataset used for the modeling exercise comprised of 314 molecules collated from the original literature. The model was validated by predicting the enzyme inhibition of 79 compounds as the external validation set. The model had high prediction power characterized by the R2 value of 0.75 for all compounds and the RMSE value of 0.51 for the external test set.