PheSA: An Open-Source Tool for Pharmacophore-Enhanced Shape Alignment.

Area of Science:

• Computational chemistry
• cheminformatics
• drug discovery

Background:

• Structure-based drug design (SBDD) relies on computational tools for screening and molecular alignment.
• Existing tools may lack flexibility or optimal performance in identifying potential drug candidates.
• Open-source solutions are crucial for accessibility and further development in the field.

Purpose of the Study:

• To introduce PheSA, an open-source pharmacophore- and shape-based screening and molecular alignment tool.
• To evaluate PheSA's performance in ligand-based screening, alignment refinement, and receptor-guided shape docking.
• To investigate the impact of different similarity metrics on screening enrichment.

Main Methods:

• Development of the PheSA algorithm within the OpenChemLib framework.
• Implementation of standard ligand-based screening and flexible alignment refinement.
• Incorporation of receptor-guided shape docking capabilities.
• Benchmark studies using datasets like DUD-E to assess screening enrichment and pose prediction.

Main Results:

• PheSA demonstrates high flexibility for various SBDD use cases.
• The use of the asymmetric Tversky similarity metric significantly improved screening enrichment rates compared to symmetric Tanimoto.
• PheSA achieved screening enrichment performance comparable to commercial methods on the DUD-E benchmark.
• The receptor-guided algorithm showed effective pose prediction capabilities.

Conclusions:

• PheSA is a powerful and versatile open-source tool for structure-based drug design.
• The Tversky metric offers an advantage for optimizing screening enrichment in pharmacophore-based approaches.
• PheSA provides a competitive, open-source alternative to commercial software for drug discovery screening.