QSAR Differential Model for Prediction of SIRT1 Modulation using Monte Carlo Method

Area of Science:

• Medicinal Chemistry
• Computational Chemistry
• Pharmacology

Background:

• Silent Information Regulator 1 (SIRT1) modulators show promise for treating cardiovascular, metabolic, inflammatory, and age-related diseases.
• Understanding the structural basis of SIRT1 modulation is crucial for developing effective therapeutics.

Purpose of the Study:

• To construct and validate quantitative structure-activity relationship (QSAR) models for both SIRT1 activators and inhibitors.
• To identify key molecular features responsible for SIRT1 activation and inhibition.

Main Methods:

• Utilized CORAL software and Monte Carlo optimization with SMILES notation to build differential QSAR models.
• Employed a dataset split into sub-training, calibration, and test sets, further validated with a prediction set.
• Performed mechanistic interpretation to define structure-activity relationships.

Main Results:

• Developed statistically significant QSAR models for SIRT1 modulators.
• The best model achieved high predictive performance with sensitivity of 1.0000, specificity of 0.8889, accuracy of 0.9524, and Matthews' correlation coefficient of 0.9058 on the validation set.
• Identified specific structural features critical for SIRT1 activation and inhibition.

Conclusions:

• The developed QSAR models provide valuable insights into the structural requirements for SIRT1 modulation.
• These findings can guide the rational design of novel SIRT1-targeting drugs for various therapeutic applications.
• The study successfully defined structural features influencing SIRT1 activation and inhibition.