Predictive power of parametric temperature based topological indices with applications in structure property modeling of anti tuberculosis drugs

Tuberculosis remains a serious global health issue, with increasing resistance to existing drugs creating an urgent need for new treatment strategies. This study applies cheminformatics and quantitative structure-property relationship (QSPR) modeling to support anti-tuberculosis drug discovery. Specifically, it focuses on two parametric temperature-based topological indices, which are graph-theoretic molecular descriptors derived from chemical structure. These indices are evaluated through discrete optimization, statistical and computational analysis to determine their predictive accuracy for key physicochemical properties such as enthalpy of formation and boiling point. The study uses 22 benzenoid hydrocarbons as test cases to assess the performance of the proposed indices across various parameter values. Optimal parameter values are identified for which the indices show the strongest correlation with experimental data. The optimized descriptors are then applied to model physicochemical properties of 13 commonly used anti-tuberculosis drugs. High correlation coefficients demonstrate the strong predictive power of the indices in estimating molecular behavior. The results suggest that these temperature-based descriptors can serve as effective tools for QSPR modeling and drug development, offering a bridge between molecular topology and practical pharmaceutical applications.