Shape-Aware Synthon Search (SASS) for Virtual Screening of Synthon-Based Chemical Spaces

Area of Science:

• Computational chemistry
• Cheminformatics
• Drug discovery

Background:

• Virtual screening is crucial for identifying drug candidates from large chemical libraries.
• Exhaustive screening of billions of compounds is computationally intensive.
• Synthon-based approaches are needed for navigating larger chemical spaces.

Purpose of the Study:

• To introduce Shape-Aware Synthon Search (SASS), a novel virtual screening method.
• To enable efficient navigation of ultralarge, synthon-based virtual chemical spaces.
• To accelerate large-scale, shape-based virtual screening for drug discovery.

Main Methods:

• Developed SASS, a synthon-based virtual screening technique.
• Implemented shape similarity searches within the synthon space.
• Compared SASS performance against exhaustive screening methods.

Main Results:

• SASS replicates exhaustive search results with high recall for various query molecules.
• The method significantly reduces the number of scored products compared to exhaustive screening.
• SASS demonstrates high efficiency in navigating ultralarge, combinatorial chemical spaces.

Conclusions:

• SASS offers a significant acceleration for large-scale, shape-based virtual screening.
• The method provides a computationally efficient alternative to exhaustive screening.
• SASS enhances the identification of high-quality drug candidates in early-stage drug discovery.