A theoretical and quartic regression-based QSPR model for predicting physicochemical properties of anti-hepatitis compounds via Van and R topological indices

Abstract The present study develops a theoretical–computational model to explore the relationship between molecular topology and physicochemical properties of anti-hepatitis compounds. Within the framework of graph theory and mathematical chemistry, two degree-based descriptors, namely the Van and R indices, were employed to describe molecular structures. A quartic regression–based quantitative structure–property relationship (QSPR) approach was utilized to evaluate how these indices predict measurable physical parameters such as boiling point, enthalpy, flash point, molar refraction, molar volume, and molecular weight. The proposed models were statistically validated using experimental data, showing high correlation coefficients (R 2 values between 0.91 and 0.98), confirming strong predictive capability. The results demonstrate that the topological indices encode relevant information about molecular connectivity and can effectively predict macroscopic physicochemical behaviors. Overall, the proposed quartic regression–based QSPR framework, integrating degree-based Van and R topological indices and validated through experimental data, provides a mathematically robust and statistically reliable approach for predicting physicochemical properties of anti-hepatitis compounds.