Abstract
Aldo‐ketoreductase 1C3 (AKR1C3) is an attractive target for prostate cancer drug discovery. Three pharmacophore models were generated from three AKR1C3 PDB crystal structures based on their crystal ligand molecules. The three pharmacophores generated from three different AKR1C3 were with 10 features, 5 features and 6 features. Pharmacophore based database screening on Pubchem, CHEMBL, Zinc and MCULE databases were performed to get overall 847 molecules and docked into the active sites of three AKR1C3 protein crystal structures. From the orientation and the docking scores, ten well fitted molecules were selected for the ADMET predictions. From the ADME toxicity studies, three molecules violating the toxicity rules and the three structures were identified from one pharmacophore. The other seven molecules identified from other two pharmacophore models were showing better values that drug like molecules. A 100 nano seconds of the simulations for best six protein ligand complexes revealed the stability of the molecules in the active site of protein. The NADPH co‐crystal in the crystal structure is guiding the inhibitor molecule in the active site of the protein.
The molecules bound AKR1C3 protein crystal structures were used to generate three different pharmacophore models. Pharmacophore based database screening was done on Pubchem, CHEMBL, Zinc and MCULE Docking based virtual screening and further evaluated through the ADMET properties. The compounds with the best ADME and toxicity properties have been retrieved and molecular dynamics simulations was performed to understand the stability of the molecules in the active site.