Consensus QSAR model for identifying novel H5N1 inhibitors.

Area of Science:

• Computational Chemistry
• Virology
• Drug Discovery

Background:

• Neuraminidase is a critical target for influenza virus drugs.
• Emerging drug resistance and new structural insights necessitate novel neuraminidase 1 (N1) inhibitors.

Purpose of the Study:

• To develop and validate a robust quantitative structure-activity relationship (QSAR) model for predicting N1 inhibitors.
• To establish a defined applicability domain (AD) for the QSAR model to ensure reliable predictions.

Main Methods:

• Developed a consensus QSAR model using a large, diverse dataset of published N1 inhibitors.
• Validated the model using an external dataset to assess performance and prevent overfitting.
• Screened the ZINC chemical library against the developed QSAR model.

Main Results:

• The consensus QSAR model demonstrated high sensitivity, specificity, and accuracy with low false positive and false discovery rates.
• Model performance on training and external validation sets was comparable, indicating robustness.
• Screening identified 64,772 probable N1 inhibitors and excluded 173,674 compounds outside the model's AD.

Conclusions:

• The developed consensus QSAR model is a reliable tool for identifying novel N1 inhibitors.
• The defined applicability domain enhances the model's accuracy and prevents misapplication.
• The model is accessible via free software (PaDEL-DDPredictor), facilitating drug discovery efforts.