LIGSIFT

LIGSIFT performs structural alignment and virtual screening of small molecules using Gaussian molecular shape overlays and a size-independent scoring function that assesses the statistical significance of molecular matches for computer-aided drug discovery.

Key Features:

• Structural alignment: Performs rapid and precise shape-based alignment of small molecules for ligand comparison.
• Gaussian molecular shape overlay: Uses a Gaussian molecular shape overlay technique to compute molecular shape similarity.
• Size-independent scoring function: Implements a size-independent scoring function that evaluates the statistical significance of molecular matches and mitigates size-dependence issues associated with the Tanimoto coefficient.
• Benchmarking: Validated on compounds associated with 40 protein targets from the Directory of Useful Decoys (DUD) and evaluated using area under the ROC curve (AUC), Enrichment Factor (EF), and Hit Rate (HR).

Scientific Applications:

• Virtual screening: Facilitates virtual screening to identify potential therapeutic small molecules in computer-aided drug discovery.
• Hit prioritization: Ranks and prioritizes candidate ligands by statistical significance to improve enrichment and hit rates.
• Benchmarking and comparison: Supports benchmarking and method comparison using datasets such as the Directory of Useful Decoys (DUD) with metrics AUC, EF, and HR.

Methodology:

Computational methods explicitly include Gaussian molecular shape overlay for shape-based ligand alignment and a size-independent scoring function that computes statistical significance of matches; benchmarking was performed on 40 protein targets from the Directory of Useful Decoys (DUD) and evaluated using AUC, Enrichment Factor (EF), and Hit Rate (HR).