A Exploring Topological Indices and QSPR Models for Anti-Cancer Drugs

Structure–Property Correlation through Graph-Theoretic Modeling

Résumé

This research examines the utilization of topological indices (TIs) and Quantitative Structure–
Property Relationship (QSPR) models to delineate the physicochemical features of five anti-cancer pharmaceuticals:
Adriamycin, Carboplatin, Carmustine, Ellence, and Hydroxyurea. Employing chemical graph
representations, M-polynomials (M-P) and neighborhood M-polynomials (NM-P) were formulated to calculate
a set of degree-based (DB) and neighborhood degree sum-based (NDSB) topological indices (TIs). After that,
linear, quadratic, and cubic regression models were used to find relationships between these indices and ten
important physicochemical parameters. The analysis shows that several TIs, such the Second Zagreb index,
Forgotten topological index, Harmonic index, and Symmetric division index, can be used to forecast parameters
like molar weight, half-life, polar surface area, density, and refractive index. The findings indicate that
quadratic and cubic models typically surpass linear models in prediction accuracy, as evidenced by low RMSE
values that validate strong structure–property correlations. The results show that polynomial-derived DB and
NDSB TIs are an effective way to study structure-property connections, which can help with rational drug
design and the creation of new anti-cancer treatments.