MOLECULAR DOCKING AND IN SILICO ADMET-BASED PRIORITIZATION OF NATURAL COMPOUNDS AGAINST SELECTED PROTEIN TARGETS

Abstract

Natural products offer a source of a vast array of molecules for early-stage drug discovery, but screening of large libraries is expensive and time consuming. In this study, 20 natural compounds were screened against five protein targets using an integrated in silico workflow and all the compounds were prioritized by molecular docking, protein–ligand interaction analysis, drug-likeness analysis, RMSD assessment and ADMET prediction. There were 300 docking records produced (3 runs x 20 compounds x 5 targets). Docking scores ranged from −9.65 to −6.65 kcal/mol, with an overall mean of −8.39 ± 0.61 kcal/mol. The overall mean docking scores were the lowest for curcumin (−9.20 kcal/mol), followed by quercetin (−9.12), luteolin (−9.04), epigallocatechin gallate (−8.94), and kaempferol (−8.76). Target wise profile of Target Protein 3 was the most favorable (−8.67 kcal/mol). The mean number of hydrogen bonds, hydrophobic contacts and π-interactions per record were 1.46, 5.36 and 1.47, respectively. RMSD values ranged from 0.62 to 2.09 Å (mean 1.37 ± 0.44 Å). 90.0% of records were predicted to have high gastrointestinal absorption and 90.0% were predicted to be low in toxicity. If you use Integrated prioritization, 16.7% of records are high priority. The top compounds should be considered to be the ones that were prioritized computationally and not necessarily the ones that are considered a therapeutic lead.