MD-LOVIs

MD-LOVIs calculates molecular descriptors using LOcal Vertex Invariants (LOVIs) and aggregation operators to generate global and local molecular descriptor indices for chemico-biological and medicinal informatics analyses.

Key Features:

• Local Vertex Invariant-Based Descriptors: Computes molecular descriptors derived from LOcal Vertex Invariants (LOVIs) and related molecular maps.
• Aggregation Operator Framework: Applies aggregation operators (AOs) to generalize descriptor calculations beyond traditional summation of atomic or fragment contributions.
• Global and Local Descriptor Generation: Produces both molecule-level and atom- or group-level descriptor indices.
• Chemical Structure Processing via CDK: Utilizes the Chemical Development Kit (CDK) library to manage chemical structures and compute atomic weights.
• Entropy-Based Descriptor Evaluation: Evaluates descriptor information content using Shannon entropy analysis.
• QSAR Descriptor Integration: Generates descriptors applied in Quantitative Structure–Activity Relationship (QSAR) modeling using multiple linear regression combined with genetic algorithms.

Scientific Applications:

• QSAR Modeling: Supports construction of Quantitative Structure–Activity Relationship models using molecular descriptor indices.
• Chemico-Biological Informatics: Enables computational analysis of chemical structures for biological activity studies.
• Medicinal Chemistry Research: Facilitates molecular descriptor analysis for drug discovery and toxicity prediction.

Methodology:

MD-LOVIs computes molecular descriptors using LOcal Vertex Invariants and related maps, applies aggregation operators to combine atomic or fragment contributions, processes molecular structures with the Chemical Development Kit (CDK), evaluates descriptor information content using Shannon entropy, and generates descriptors used in QSAR models built with multiple linear regression and genetic algorithms.