Effect of structure of dihydrobenzothiophenes on the activity of LRRK2 inhibitors

Familial Parkinson’s disease cases have recently been associated with the leucine rich repeat kinase2 (LRRK2) gene. It has been hypothesized that inhibition of the LRRK2 protein may have the potential to alter disease pathogenesis, thus LRRK2 kinase inhibitors are potentially useful in the treatment of PD [1-2]. The quantitative structure activity relationship (QSAR) of novel dihydrobenzothiophene series of potent, selective, LRRK2 inhibitors was studied. The suitable set of the molecular descriptors was calculated and the important descriptors using genetic algorithm (GA) were selected. Nowadays, GA is well-known as an interesting and more widely used variable selection method [3]. In the present work the data set was selected from the literature [1]. The total of 1481 descriptors was calculated for each molecule using Dragon software, version 2.1 and descriptors with constant or almost constant values for all molecules were eliminated. Then among the 605 remaining descriptors the most significant molecular descriptors were identified using the genetic algorithm method. In a QSAR study, generally, the quality of a model is expressed by its fitting and prediction ability [4]. In order to build and test the model, a data set of 46 compounds was separated into a training set of 37 compounds, which was used to build model and a test set of 9 compounds, which was applied to evaluate the built model. The GA–MLR analysis led to the derivation of one model with four descriptors (BEHe5, GATS1p, Mor06m, HATS1v), and the results are as follows:R2 = 0.694, R2adj= 0.656 , Q2LOO = 0.616, RMSEtr = 0.351, RMSEcv = 0.332, MAEcv = 0.393 , MAEtr =0.292 , RMSEext = 0.523 , MAEext =0.396The value of Q2LOOvalue (0.616) suggests that there is a good internal validation.