De novo in silico screening of natural products for antidiabetic drug discovery: ADMET profiling, molecular docking, and molecular dynamics simulations

Epigenetic dysfunction which has implicated disease conditions such as diabetes highlights the urgency for the discovery of novel therapeutic alternatives. The rising global incidences of diabetes and the limitations of existing treatments further exacerbate the quest for novel antidiabetic agents' discovery. This study leverages computational approaches to screen selected bioactive natural product phytoconstituents for their potential anti-diabetic properties. Utilizing pharmaceutical profiling, ADMET (Absorption, Distribution, Metabolism, Excretion, and Toxicity) predictions, molecular docking, and molecular dynamics (MD) simulations, the drug-likeness and binding affinity of these natural compounds against human pancreatic amylase was investigated. Out of the total 24,316 ZINC compounds screened for their binding scores with amylase, ZINC85593620, ZINC85593668, and ZINC85490447 came top. The compounds had higher binding scores than the standards (acarbose and ranirestat) with ZINC85593620 having the highest docking score of - 12.162 kcal/mol and interacted with key amino acid residues such as TRP 59, ILE 148, and ASP 197. Further validation through MD simulations reveals that all the compounds showed minimal fluctuations relative to the standards indicating strong and stable binding interactions suggesting potential effective inhibition of the enzyme. ZINC85593620 and ZINC85593668 showed promising distribution and availability characteristics for amylase inhibition. Overall, the compounds displayed potential amylase inhibition which underscores their use as promising natural products in developing new antidiabetic drugs. Further experimental validations are recommended to offer a potential solution to the pressing need for safer and more effective antidiabetic therapies.