<![CDATA[Type 2 diabetes mellitus is a chronic metabolic disorder requiring long-term therapy, yet current synthetic PPARγ agonists like thiazolidinediones are often associated with serious adverse effects. Therefore, identifying natural alternatives from sources such as Swietenia mahagoni is essential to provide effective therapy with potentially lower toxicity profiles. This study employed an in silico machine learning approach using SkelSpheres descriptors to predict the IC₅₀ values of compounds derived from the seeds of Swietenia mahagoni against PPARγ, followed by molecular docking validation using Molegro Virtual Docker (MVD). The predictive model for PPARγ agonists demonstrated acceptable validity (R²-test = 0.5308; accuracy = 84.01%). Four compounds from S. mahagoni showed predicted IC₅₀ values below 1 µM (0.0973–0.9527 µM), categorized as “Predicted Excellent activity.” Docking simulations revealed that the bibenzyl derivative 2-Carboxy-3,5-Dihydroxy-4-Geranylbibenzyl (CID: 25135579) and β,β-Carotene tetrol (CID: 23258402) exhibited binding affinities comparable to the control ligand thiazolidinedione, with Rerank Scores of-114.991 and -109.764 kJ/mol, respectively. In conclusion, the bibenzyl derivative and carotene tetrol from S. mahagoni represent promising natural candidates for PPARγ agonists, providing a strong rationale for further in vitro and in vivo investigations as potential antidiabetic agents.]]>
Copyrights © 2026
