From virtual screening to bench: A dual-validation framework for drug repurposing against PI3K

Area of Science:

• Medicinal Chemistry
• Computational Drug Discovery
• Pharmacology

Background:

• Virtual screening is crucial for identifying novel drug candidates efficiently.
• The phosphoinositide 3-kinase (PI3K) pathway is implicated in various cancers.
• Drug repurposing offers a faster route to therapeutic development.

Purpose of the Study:

• To develop an applicability domain-based virtual screening strategy for PI3K inhibitors.
• To identify novel drug-like molecules and repurpose existing drugs targeting PI3K-α.
• To enhance predictive accuracy using Quantitative Structure-Activity Relationship (QSAR) modeling.

Main Methods:

• Applicability domain-based virtual screening of large compound libraries (Enamine).
• Development of a QSAR model for PI3K ligands using Light Gradient Boosting Machine (LGBM).
• Screening of FDA-approved drugs for PI3K-α inhibitory activity.
• Pharmacophore mapping and wet-lab validation of identified compounds.

Main Results:

• A QSAR model for PI3K ligands achieved an R2 of 0.799.
• Three novel small molecules with favorable drug-like properties were identified from Enamine libraries.
• One FDA-approved drug demonstrated PI3K-α inhibitory activity, confirmed by wet-lab evaluation.
• The integrated approach identified promising PI3K-α inhibitors for drug discovery and repurposing.

Conclusions:

• Applicability domain-based virtual screening combined with QSAR modeling is effective for identifying PI3K inhibitors.
• The study successfully identified novel PI3K-α inhibitors and validated a repurposed drug.
• This integrated framework accelerates drug discovery and repurposing efforts for PI3K-targeted therapies.