Integrating computational and mixture-based screening of combinatorial libraries.

Area of Science:

• Medicinal Chemistry
• Computational Chemistry
• Pharmacology

Background:

• Mixture-based synthetic combinatorial library (MB-SCL) screening is a key method for identifying structure-activity relationships (SAR) and drug leads.
• Virtual screening (VS) is a powerful computational tool increasingly utilized in drug discovery with documented successes.
• Integrating experimental MB-SCL screening with computational VS approaches remains underexplored.

Purpose of the Study:

• To combine experimental data from MB-SCL screening with molecular similarity methods.
• To develop a novel approach for analyzing and categorizing compounds within a library.
• To demonstrate the retrieval of active compounds from MB-SCLs using computational strategies.

Main Methods:

• Screening of a bicyclic guanidine MB-SCL against the κ-opioid receptor.
• Application of molecular similarity analyses to experimental activity data.
• Integration of activity data and similarity metrics into a biometric analysis-similarity map.

Main Results:

• The biometric analysis-similarity map effectively categorized compounds into actives, activity cliffs, low similarity, and missed hits.
• An active compound with an IC(50) of 309 nM was identified within the 'missed hits' region.
• This demonstrates the capability of computational approaches to uncover active molecules within MB-SCLs.

Conclusions:

• The integration of MB-SCL screening with molecular similarity analysis provides a powerful strategy for hit identification.
• The developed biometric analysis-similarity map facilitates the categorization and discovery of active compounds.
• This generalizable approach can be applied to diverse MB-SCLs in drug discovery programs.