Toward a general predictive QSAR model for gamma-secretase inhibitors

Area of Science:

• Medicinal Chemistry
• Computational Chemistry
• Drug Discovery

Background:

• Gamma secretase (GS) is a critical drug target for Alzheimer disease and cancer due to its role in amyloid precursor protein and notch protein processing.
• The lack of a three-dimensional structure for GS necessitates novel prediction and screening methods for inhibitors.

Purpose of the Study:

• To develop quantitative structure-activity relationship (QSAR) models for predicting gamma secretase (GS) inhibitors.
• To identify key molecular descriptors influencing GS inhibitor potency.
• To facilitate the rational design of novel GS inhibitors.

Main Methods:

• Quantitative structure-activity relationship (QSAR) studies were performed on 233 compounds from the ChEMBL database.
• Continuous QSAR models (Partial Least Squares regression, Neural Network) and categorical QSAR models (Neural Network, Linear Discriminant Analysis) were developed.
• Structure-activity relationship (SAR) analysis was conducted in conjunction with QSAR descriptors.

Main Results:

• QSAR models were developed to identify descriptors crucial for GS inhibitor potency.
• Electronegative substitutions on aryl rings (PEOE3) were found to be significant in determining inhibitor potency.
• Replacing acyclic amines with N-substituted cyclic amines, along with increased aliphatic rings (sssN_Cnt), enhances GS inhibitor potency.

Conclusions:

• Statistically significant QSAR models were established, enhancing the understanding of compounds targeting GS.
• The findings provide insights for the rational design of novel and potent gamma secretase inhibitors.
• This research aids in the discovery of new therapeutic agents for Alzheimer disease and cancer.