Find Core Toolbox

Find Core Toolbox identifies core atom sets for NMR-based biomolecular structure determination by implementing the FindCore and FindCore2 algorithms in MATLAB to improve precision of structural superimpositions and RMSD-based assessments.

Key Features:

• FindCore and FindCore2 algorithms: Implements the FindCore and FindCore2 algorithms to determine core atom sets for structural comparison.
• Identification of Core Atoms: Selects core atoms used for root-mean-square deviation (RMSD) calculations to assess precision of NMR-derived structures.
• Partitioning into RMSD-Stable Domains: Partitions core atom sets into RMSD-stable domains representing locally well-defined regions for localized superimpositions.
• Epsilon-Value Calculation: Computes an epsilon-value using a kurtosis-based criterion as a complementary precision metric to RMSD.
• Expanded FindCore Methodology: Provides an Expanded FindCore mode to generate broader core atom sets for model validation contexts such as CASP.
• Superimposition-Independent Core Identification: Applies a superimposition-independent approach to identify core atoms irrespective of global alignment.
• Exclusion of Poorly Defined Atoms: Optimizes core atom sets by excluding poorly defined atoms to ensure well-characterized regions are used in precision assessments.
• MATLAB implementation: Implemented in MATLAB.

Scientific Applications:

• NMR Structure Validation: Assesses precision of NMR-derived structural ensembles via focused superimposition and RMSD calculations on well-defined atomic subsets.
• CASP Competition: Enables validation of predicted protein models against experimental NMR structures by emphasizing well-defined core regions and Expanded FindCore assessments.
• Structural Bioinformatics and Biology: Applies to analyses requiring identification of reliably defined atomic subsets for comparison and interpretation of biomolecular structures.

Methodology:

Implements a superimposition-independent approach to identify core atom sets, partitions cores into RMSD-stable domains, calculates RMSD and an epsilon-value using a kurtosis-based criterion, optimizes sets by excluding poorly defined atoms, and provides an Expanded FindCore option for broader core selection; the methods are implemented in MATLAB.