Mechanistic QSAR analysis to predict the binding affinity of diverse heterocycles as selective cannabinoid 2 receptor inhibitor

Abstract

CB2R are fascinating targets for neuropathic pain and mood disorders because of their improved biological characteristics. Experimental data on 1296 cannabinoid-2 receptor inhibitors with different structural properties were used to develop a QSAR model following OECD guidelines. This study selected the best-predicted model (80:20 splitting ratio) with fitting parameters, such as R2:0.78; F:623.6, Internal validation parameters, such as Q2Loo:0.78; CCCcv: 0.87 and external validation parameters, such as R2ext:0.77; Q2F1:0.7730; Q2F2:0.7730; Q2F3:0.76; CCCext:0.87. Following this, another QSAR model was developed by using a 50:50 split ratio for thetraining and the prediction sets, which were then swapped to evaluate the robustness of the built QSAR model by the 50:50 ratio, which also gives a deeper understanding of the chemical space. In addition, we have confirmed the QSAR result with pharmacophore modelling, and supported by molecular docking, MD simulation, MMGBSA and ADME studies. Thus, this work may enable cannabinoid 2 receptor inhibsitor development.

Keywords

• Cannabinoid 2 receptor
• QSAR
• GA-MLR
• pharmacophore modeling
• Molecular dynamic simulation
• MMGBSA