Indexing Ultrafast Shape-Based Descriptors in MongoDB to Identify TLR4 Pathway Agonists.

Area of Science:

• Computational chemistry
• Cheminformatics
• Drug discovery

Background:

• Molecular shape is crucial for biological activity, influencing how molecules bind to protein targets.
• Traditional shape comparison methods are computationally intensive, limiting the screening of large compound libraries.
• Existing algorithms for molecular shape comparison often require significant computational resources.

Purpose of the Study:

• To develop an ultrafast, shape-based search workflow for screening massive compound collections.
• To implement a method that avoids storing precomputed molecular conformers, reducing computational overhead.
• To accelerate the identification of novel bioactive molecules through efficient shape similarity searches.

Main Methods:

• Utilized open-source software packages, including the Open Drug Discovery Toolkit (ODDT) and RDKit.
• Implemented a workflow for rapid conformer generation and matching based on shape descriptors.
• Developed indexing strategies for transformed shape-based descriptors to expedite search processes.
• Integrated MongoDB for optional storage of descriptors for future large-scale searches.

Main Results:

• Achieved ultrafast conformer generation and matching without extensive storage requirements.
• Demonstrated a shape similarity search on over 70 million compounds in under 8 seconds.
• Successfully identified novel structural motifs of potential TLR4 agonists active in cell lines.

Conclusions:

• The developed workflow enables highly efficient, shape-based virtual screening of exceptionally large chemical datasets.
• This method significantly accelerates the identification of potential drug candidates with desired binding properties.
• The approach holds promise for discovering new therapeutic agents by rapidly exploring vast chemical spaces.