On QSPR Modeling and SAW-Based Ranking of Laryngeal Cancer Drugs Using Temperature-Based Topological Indices and Support Vector Regression

QSPR modeling offers an effective tool for predicting the physicochemical properties of pharmaceutical molecules by applying molecular descriptors. In this research paper, a new QSPR method based on the temperature-based topological indices in graph theory is presented for analyzing selected drugs that are relevant to laryngeal cancer. Graph theory is used as a mathematical tool in representing the molecular structure and extracting structural information from molecular graphs. New temperature-based topological indices are introduced as molecular descriptors in order to extract structural and thermal properties of the molecules. The temperature-based topological descriptors have been used in conjunction with a support vector regression algorithm to predict various physicochemical properties of the drugs. According to the obtained results, there was a reasonable agreement between the experimentally determined values and predicted ones. Hence, the ability of the proposed molecular descriptors in capturing structural information was revealed. Moreover, a mathematical ranking of the drugs based on their physicochemical properties and calculated molecular descriptors is carried out through a simple additive weighting (SAW) technique. Also, the WASPAS technique has been used for obtaining the hybrid multicriteria ranking, which will be used for comparing purposes, whereas the linear regression technique has been applied as a basic predictive model. It is important to note that the obtained SAW ranking was a structural property–based one and should not be considered in terms of therapeutic effect.