fmcsR

fmcsR identifies flexible maximum common substructures (FMCSs) between small molecules to detect structural similarity for applications in drug discovery and chemical genomics.

Key Features:

• Mismatch Tolerance: Allows configurable atom or bond mismatches to compute flexible MCSs that can be larger and more informative than strict MCS results.
• Efficient Algorithm Implementation: Implements the core FMCS algorithm in C++ to improve computational performance and time efficiency across a range of compound sizes.
• R Integration and Analytical Utilities: Provides functions for pairwise compound comparisons, structure similarity searching, clustering, and visualization of MCSs within the R environment.
• Virtual Screening Performance: Demonstrates improved overall and early-stage enrichment of active compounds in similarity-based virtual screening when using FMCS-based methods.

Scientific Applications:

• Drug Discovery: Enables identification of shared substructures among small molecules to support lead identification and SAR analysis.
• Chemical Genomics: Facilitates comparison of compound libraries to elucidate relationships between chemical structure and biological function.
• Activity Prediction: Supports prediction of compound activity by capturing nuanced structural similarities that strict MCS methods may miss.
• Virtual Screening: Improves early enrichment of actives in similarity search results to prioritize candidates in screening workflows.

Methodology:

Computes flexible maximum common substructures with user-configurable atom/bond mismatches; core algorithm implemented in C++; provides routines for pairwise comparisons, similarity searching, clustering, and MCS visualization.