Degree-Based Molecular Descriptors as Predictors of Physicochemical Properties of Antibiotic Compounds

Predicting the physicochemical properties of pharmaceutical compounds through mathematical frameworks o ers a powerful and cost-e ective alternative to laboratory experimentation. In this study, ten degree-based molecular descriptors are applied to eight clinically important antibiotic drugs within a quantitative structure property relationship (QSPR) framework. Descriptor values are computed via the edge partitioning technique, and simple linear regression is employed to model seven key physicochemical properties: boiling point, vapour pressure, ash point, enthalpy of vaporisation, surface tension, molar volume, and molar refractivity. The resulting correlation coe cients reveal strong positive associations between the majority of descriptors and the target properties, with molar refractivity and boiling point exhibiting the highest predictive accuracy. The statistically estimated values show excellent agreement with experimentally reported data, con rming the validity and reliability of the proposed approach. These ndings establish degree-based molecular descriptors as e cient, economical, and computationally accessible tools for characterising the physicochemical behaviour of antibiotic compounds, with promising implications for rational drug design and pharmaceutical property prediction.