Development of a Docking-Based QSAR and ADMET Analysis of White Radish (Raphanus sativus L.) Compounds as Potential Anti Insomnia Agents

Raphanus sativus L. (white radish) contains various bioactive compounds that may have potential as anti-insomnia agents. This study aimed to evaluate the structure–activity relationship and pharmacokinetic properties of white radish compounds using Quantitative Structure–Activity Relationship (QSAR) modeling and ADMET prediction. Fourteen compounds previously identified through molecular docking studies against the GABA receptor were analyzed. Binding affinity values from previous docking studies were used as the dependent variable, while lipophilicity (logP), topological polar surface area (TPSA), and hydrogen bond donor (HBD) were selected as molecular descriptors. QSAR modeling was performed using multiple linear regression and validated through leave-one-out cross validation (LOOCV), multicollinearity analysis, and residual diagnostics. The final QSAR model showed good statistical performance with an adjusted R² of 0.820 and Q²LOO of 0.756. LogP and HBD were identified as the most influential descriptors, whereas TPSA did not significantly contribute to the model. ADMET prediction revealed considerable variation in pharmacokinetic properties among the investigated compounds. Glucobrassicin exhibited the strongest predicted binding affinity, while ledol and nerolidol demonstrated more favorable pharmacokinetic profiles, including blood–brain barrier permeability and high gastrointestinal absorption. These findings suggest that glucobrassicin, ledol, and nerolidol are promising candidates for further development as natural product-based anti-insomnia agents.