Searching for substructures in fragment spaces.

Area of Science:

• Computational chemistry
• Cheminformatics
• Drug discovery

Background:

• Drug development requires selecting compounds with specific structural features from vast molecular datasets.
• Existing search methods struggle with the infinite nature of molecular space and complex structural queries spanning multiple fragments.
• Fragment spaces (FSs) offer a feasible representation for large combinatorial datasets but pose challenges for search algorithms.

Purpose of the Study:

• To develop a method for efficiently searching fragment spaces (FSs) to identify substructures without explicit product enumeration.
• To overcome the time and storage limitations associated with searching large combinatorial molecular datasets.
• To enable the discovery of novel compounds for drug lead generation.

Main Methods:

• A novel method is presented that splits target substructures into smaller sub-substructures.
• These sub-substructures are mapped onto molecular fragments within the FSs, respecting fragment connectivity rules.
• The method avoids the combinatorial explosion of explicit product enumeration.

Main Results:

• The method successfully identified substructures in FSs across three drug discovery scenarios.
• Searches within FSs were completed in seconds.
• The approach discovered novel compounds not present in existing databases like PubChem.

Conclusions:

• This method provides an efficient and scalable solution for searching complex fragment spaces in drug discovery.
• It accelerates the identification of potential drug lead structures by uncovering novel chemical entities.
• The technique overcomes limitations of traditional methods, enabling exploration of previously inaccessible molecular regions.