Integration of virtual screening of phytoecdysteroids as androgen receptor inhibitors by 3D-QSAR Model, CoMFA, molecular docking and ADMET analysis: An extensive and interactive machine learning.

Ecdysteroids, a class of naturally isolated polyhydroxylated sterols, stands at a very good place in the pharmaceutical industry from their medicinal point of views like anti-inflammatory, neuroprotective, anti-microbial, anti-diabetic, antioxidant, and anti-tumor effects. Due to their excellent antioxidant and anti-microbial potential, ecdysteroids have extensive use in skin products, especially derma creams. To monitor the best anti-acne phytoecdysteroids, here made use of different computational approaches, by using the rapid, easy, cost-effective and high throughput method to screen and identify ecdysteroids as androgen receptor inhibitors. 3D-QSAR study was carried out on a dataset of ecdysteroids by using comparative molecular field analysis (CoMFA) to determine the factors responsible for the activity of compounds. Statistically a cross-validated (q2) 0.1457 and regression coefficient (r2) 0.9713 indicated the best model. Contour map results showed the influence of steric effect to enhance activity. A molecular docking analysis was done to further find out the binding sites and their anti-acne potential against three crystal structured macromolecules (PDB ID: 2REQ, 2BAC, 4EM0). Docking results were further evaluated by prime MM-GBSA analysis and findings confirmed the accuracy. Toxicity by ADMET assessment was carried out and M2 was found as lead druglike with best anti-acne activity against Propionium acnes GehA lipase bacteria after passing all filters. This research study is novel because it is representing first effort to explore ecdysteroids class for their high therapeutic output as androgen receptor inhibitor by using computational tools and expectedly led to novel scaffold for androgen receptor inhibitor. This is a novel and new approach to investigate the ecdysteroids for first time for their practical applications. [Display omitted] • 3D-QSAR study by using Comparative molecular field analysis (CoMFA) determines the factors responsible for the activity of compounds. • Statistically a cross-validated (q2) 0.1457 and regression coefficient (r2) 0.9713 indicated the best model fit with seven compounds as active ligands. • Screened 7 ligands showed high binding affinity in range of − 7.8 to − 11.8 kcal/mol against receptors (PDB ID: 2REQ, 2BAC) and − 5.5 to − 7.5 kcal/mol against receptor (PDB ID: 4EM0) • 2 compounds (M2 & R6) of best binding affinity also fulfill the criteria of Lipinski rule of five. • M2 can serve as an effective lead compound in upcoming exploration for the inhibition of acne. [ABSTRACT FROM AUTHOR]