Abstract: In this study,a three-dimensional quantitative structure-activity relationship (3D-QSAR) model of GPR35 receptor agonists was established by comparative molecular field analysis (CoMFA) and comparative molecular similarity indices analysis (CoMSIA) ,the quantitative relationship between the structure of the agonists and its biological activities was determined.In the CoMFA model,which was employed to calculate the half effective concentration,the final LOO (Leave-one-out) cross-validation coefficient (q²) is 0.559,the non-cross validation coefficient (r²) is 0.887,standard deviation (SE) is 0.382,and in the CoMSIA model,the final decision coefficient (q²) is 0.627,non-cross validation coefficient (r²) is 0.915,and standard deviation (SE) is 0.365.In the CoMFA model,which was employed to calculate the half inhibitory concentration,the final LOO (Leave-one-out) cross-validation coefficient (q²) is 0.600,non-cross validation coefficient (r²) is 0.903,standard deviation (SE) is 0.375,and in the CoMSIA model,the final decision coefficient (q²) is 0.566,non-cross validation coefficient (r²) is 0.914,and standard deviation (SE) is 0.378.One satisfactory 3D-QSAR model possessing predictive capability was obtained,EC50and IC₅₀ values which predicted by 3D-QSAR model are almost same with experimental data.In this experiment,according to the predicted results,the structures of the ligand molecules were analyzed.The 3D-QSAR model can be applied to design new coumarin derivatives potent GPR35 agonists.

Key words: GPR35 receptor, coumarin compounds, three-dimensional quantitative structure-activity relationship, comparing the molecular force field analysis, comparative molecular similarity index analysis