CLIP: similarity searching of 3D databases using clique detection

Area of Science:

• Computational chemistry
• Cheminformatics
• Drug discovery

Background:

• 3D similarity searching is crucial for virtual screening in drug discovery.
• Existing methods may not efficiently handle the geometric and distance tolerances required for pharmacophore matching.

Purpose of the Study:

• To describe a novel program, CLIP (Candidate Ligand Identification Program), for efficient 3D similarity searching.
• To present a modified similarity coefficient that accounts for distance tolerances in pharmacophore matching.

Main Methods:

• Utilized the Bron-Kerbosch clique detection algorithm for substructure matching.
• Characterized molecular structures by the geometric arrangement of pharmacophore points.
• Employed modified Simpson and Tanimoto association coefficients for similarity calculations, incorporating distance tolerance.

Main Results:

• Demonstrated the effectiveness of CLIP in virtual screening using HIV assay data.
• Showcased the computational efficiency of the CLIP program for large-scale searching.

Conclusions:

• CLIP provides an effective and efficient approach for 3D similarity searching in virtual screening.
• The modified similarity coefficients enhance the accuracy of pharmacophore matching by considering distance tolerances.