Exploring anticancer drug structures through vertex based resolving parameters.

Area of Science:

• Cheminformatics and computational chemistry.
• Application of graph theory in molecular modeling.

Background:

• Quantitative Structure-Property Relationship (QSPR) studies face challenges in generating molecular descriptors that capture structural information.
• Classical degree-based descriptors have limitations in distinguishing structurally similar molecules.

Purpose of the Study:

• To introduce a novel class of vertex-based topological indices derived from metric dimension theory.
• To enhance the capacity of molecular descriptors to separate structurally similar molecular graphs.
• To assess the predictive performance of these new descriptors in QSPR modeling.

Main Methods:

• Developed new vertex-based topological indices utilizing resolving degrees from metric dimension theory.
• Calculated these descriptors for a set of anticancer drugs.
• Correlated the descriptors with physicochemical properties like boiling point, molar volume, and enthalpy of vaporization.
• Evaluated predictive performance using linear and cubic regression models.

Main Results:

• The new resolving-degree-based descriptors demonstrated statistically significant and robust correlations with studied physicochemical properties.
• Cubic regression models, prioritizing accuracy, better represented nonlinear structural effects compared to interpretable linear models.
• These novel descriptors showed superior predictive power for molecular structure compared to existing methods.

Conclusions:

• Resolving-degree-based topological indices offer enhanced predictive capabilities for QSPR modeling.
• These metrics show promise for applications in computational drug design and cheminformatics.
• The developed descriptors provide a more nuanced understanding of structure-property relationships.