Xây dựng mô hình in silico và ứng dụng sàng lọc ảo các chất có khả năng ức chế acetylcholinesterase

In this study, the 2D-QSAR and QSAR classification models were built on 715 compounds that can inhibit acetylcholinesterase (AChE) activity, with the partial least square (PLS) and the binary algorithms. Two 2D-QSAR models on tacrine and aporphine structure groups were generated with R2 = 0.86, RMSE = 0.45 and R2 = 0.93, RMSE = 0.33, respectively. Two Binary QSAR models were built based on IC50 thresholds of 1.0 and 10 µM with the accuracy of 0.82 and 0.90. Virtual screening from 57,423 structures of the Traditional Chinese medicine database by Binary QSAR (IC50 < 10 µM) and two 2D-QSAR models resulted in 80 potential AChE inhibitors having tacrine structure and 121 potential AChE inhibitors having aporphine structure.
Keyword: Alzheimer’s disease (AD), acetylcholinesterase, 2D-QSAR, binary QSAR, in silico.