Quantitative Structure-Activity Relationship Study of Antioxidant Tripeptides Based on Model Population Analysis

Area of Science:

• Biochemistry
• Computational Chemistry
• Food Science

Background:

• Antioxidant peptides offer health benefits, driving research interest.
• Understanding structure-activity relationships (SAR) is crucial but incomplete.
• Tripeptides are a key focus for antioxidant peptide research.

Purpose of the Study:

• To build and improve quantitative structure-activity relationship (QSAR) models for antioxidant tripeptides.
• To investigate the impact of model population analysis (MPA) on QSAR model performance.
• To enhance the prediction accuracy of antioxidant activity in peptides.

Main Methods:

• Development of QSAR models using ferric thiocyanate (FTC) and ferric-reducing antioxidant power (FRAP) datasets.
• Inclusion of sixteen amino acid descriptors in model construction.
• Application of model population analysis (MPA) to refine QSAR models.

Main Results:

• QSAR models were developed for 214 (FTC) and 172 (FRAP) antioxidant tripeptides.
• MPA application significantly improved model prediction performance.
• Cross-validated R² increased from 0.6170 to 0.7471 (FTC) and 0.4878 to 0.6088 (FRAP).

Conclusions:

• Integrating diverse amino acid descriptors enriches QSAR model building.
• MPA effectively extracts relevant information for enhanced prediction accuracy.
• This approach advances the understanding and prediction of antioxidant peptide functionality.