r-BRICS - A Revised BRICS Module That Breaks Ring Structures and Carbon Chains.

Area of Science:

• Computational chemistry
• Cheminformatics

Background:

• Molecular fragmentation is crucial for machine learning, molecular modeling, and drug discovery.
• Existing tools like BRICS often produce large fragments, limiting their utility for specific tasks.
• Challenges include intact long aliphatic chains, connected/fused rings, and nitrogen-containing groups.

Purpose of the Study:

• To develop a novel molecular fragmentation method for increased flexibility.
• To address limitations of existing fragmentation tools, particularly BRICS.
• To generate smaller, more numerous, and unique molecular fragments.

Main Methods:

• Development of a revised BRICS (r-BRICS) algorithm.
• Application of r-BRICS to a diverse range of molecules.
• Comparison of fragment size and diversity generated by r-BRICS versus BRICS.

Main Results:

• r-BRICS successfully generates smaller molecular fragments compared to BRICS.
• The revised method allows for more localized fragment assessments.
• r-BRICS produces a fragment database with a significantly higher number of unique small fragments.

Conclusions:

• r-BRICS provides a more flexible and versatile approach to molecular fragmentation.
• The generated smaller fragments are beneficial for fragment-based machine learning, coarse-graining, and ligand-protein interaction studies.
• The increased diversity of small fragments from r-BRICS shows promise for fragment-based drug discovery.