Quantitative structure-activity relationship study of serotonin (5-HT7) receptor inhibitors using modified ant colony...

Area of Science:

• Medicinal Chemistry
• Computational Chemistry
• Pharmacology

Background:

• Serotonin 5-HT(7) receptor antagonists are potential therapeutic agents.
• Quinazolinone derivatives show promise as inhibitors of the 5-HT(7) receptor.
• Predictive modeling is crucial for efficient drug discovery.

Purpose of the Study:

• To develop a robust Quantitative Structure-Activity Relationship (QSAR) model for predicting the inhibitory activity of novel quinazolinone derivatives against the serotonin 5-HT(7) receptor.
• To identify key molecular descriptors influencing the inhibitory activity.
• To evaluate the performance of a modified Ant Colony Optimization (ACO) and Adaptive Neuro-Fuzzy Inference System (ANFIS) model.

Main Methods:

• Quantitative Structure-Activity Relationship (QSAR) modeling.
• Modified Ant Colony Optimization (ACO) algorithm for variable selection.
• Adaptive Neuro-Fuzzy Inference System (ANFIS) for activity prediction.
• Shuffling cross-validation, leave-one-out, leave-multiple-out, and Y-randomization techniques for model validation.

Main Results:

• The modified ACO-ANFIS model identified key descriptors (Q(max), Se, Hy, PJI3, DELS) influencing inhibitory activity.
• The model achieved a high statistical performance with R(2) = 0.775 and RMSE = 0.360.
• The model demonstrated robustness and predictive ability through rigorous validation techniques.
• The proposed ACO-ANFIS model outperformed stepwise MLR-ANFIS and GA-PLS-ANFIS methods.

Conclusions:

• The developed modified ACO-ANFIS model is a reliable tool for predicting the 5-HT(7) receptor inhibitory activity of quinazolinone derivatives.
• The identified descriptors provide valuable insights into the structural requirements for potent inhibition.
• This approach facilitates the design of novel and effective quinazolinone-based therapeutics targeting the 5-HT(7) receptor.