1Department of Mathematics, Navkis College of Engineering (Affiliated to Visvesvaraya Technological University, Belagavi), Hassan - 573217, India
2,3Department of Mathematics, Adichunchanagiri Institute of Technology (Affiliated to Visvesvaraya Technological University, Belagavi), Chikkamagaluru - 577102, India
*Corresponding Author: Vinaya K U, Department of Mathematics, Navkis College of Engineering, Hassan - 573217, India
Topological indices are a set of graph invariants that carry information about the structure of a molecule. In this study, we used a set of + and – signs for vertices and edges in chemical structures of 16 antimicrobial compounds in such a way that the number of + signs and – signs are balanced. Labeled topological indices were calculated for all compounds, and multiple regression models in different forms – linear, quadratic, cubic, exponential, and logarithmic – for six physicochemical properties – molecular weight (MW), polarizability, heavy atom count (HAC), boiling point (BP), molar refractivity (MR), and molar volume (MV) – are constructed. The results indicate that the labeled topological indices offer significant predictive ability. Comparison of results is performed by statistical parameters such as coefficient of determination (R²), mean absolute error (MAE), root mean square error (RMSE), and mean absolute percentage error (MAPE). The signed approach shows higher sensitivity for electronic effects in molecules for quantitative structure-property relationship predictions.
Keywords: Cordial labelling, Labeled Topological Indices, Physicochemical Properties, Regression Models.
How to cite this article: Vinaya K U, Shrikanth A S, Mitha V U.; Prediction of Physicochemical Properties of Antimicrobial Compounds Using Labeled Topological Indices and Regression Models. Int J Drug Deliv Technol. 2026;16(11s): 85-94; DOI: 10.25258/ijddt.16.11s.9
Source of support: Nil.
Conflict of interest: Nil