On Degree‐Based Topological Indices and QSPR Analysis of Some Novel Drugs Used in the Treatment of Cervical Cancer Using Machine Learning With XGBoost and Regression Models

ABSTRACT Topological indices (TIs), derived from molecular graph theory, serve as valuable tools in the evaluation of drug efficacy for cervical cancer treatment. These indices capture essential structural information that aids in predicting therapeutic potential and assessing toxicity, both of which are critical for the development of safe and effective anticancer agents. This study introduces a Quantitative Structure‐Property Relationship (QSPR)‐based approach that integrates TIs with molecular features to estimate key physicochemical properties, including boiling point (BP), melting point (MP), enthalpy of vaporization (EV), flash point (FP), index of refraction (IR), molar refractivity (MR), molar volume (MV), polarizability (P), surface tension (ST) and polar surface area (PSA). A linear regression (LR) model is constructed using these features and benchmarked against an Extreme Gradient Boosting (XGBoost) model. Comparative analysis reveals that the XGBoost model significantly enhances prediction accuracy over traditional regression methods. The findings highlight the effectiveness of combining machine learning techniques with topological descriptors in enhancing predictive modeling for drug discovery in cervical cancer.