CAESAR: a new conformer generation algorithm based on recursive buildup and local rotational symmetry consideration.

Area of Science:

• Computational Chemistry
• Molecular Modeling
• Drug Discovery

Background:

• Accurate molecular conformation is crucial for understanding molecular interactions and predicting biological activity.
• Existing conformer generation methods can be computationally intensive and may struggle with molecules possessing high symmetry.

Purpose of the Study:

• To introduce and validate a novel, highly efficient conformer search algorithm named CAESAR (Conformer Algorithm based on Energy Screening and Recursive Buildup).
• To demonstrate CAESAR's superior speed and accuracy in generating relevant molecular conformers compared to existing methods.

Main Methods:

• Development of a divide-and-conquer and recursive conformer build-up approach.
• Integration of local rotational symmetry considerations to eliminate duplicate conformers.
• Implementation within Discovery Studio 1.7 and validation against Catalyst/FAST.

Main Results:

• CAESAR is 5-20 times faster than Catalyst/FAST across various datasets.
• Significant speedups were observed for molecules with high topological symmetry or requiring extensive conformer sampling.
• CAESAR demonstrated slightly improved reproduction of receptor-bound X-ray conformations from the Protein Data Bank (PDB) compared to Catalyst/FAST.
• The algorithm effectively covers pharmacophore space, performing as well as or better than Catalyst/FAST.

Conclusions:

• CAESAR represents a significant advancement in efficient and accurate conformer generation.
• The algorithm's speed and quality make it a valuable tool for computational chemistry and drug discovery.
• CAESAR's ability to handle complex molecules and its improved accuracy enhance its utility in molecular modeling applications.