Regression Models in QSPR Analysis for Blood Cancer Treatment: A Connection Number Approach

Topological indices (TIs) are valuable tools in the study of blood cancer drugs, offering insights into molecular structure and associated properties critical for assessing efficacy and toxicity. This study introduces Connection Number-Based Topological Indices (CNTIs), designed to model and predict important physiochemical characteristics of anti-cancer compounds. Using edge-partitioning techniques, we analyze five clinically relevant blood cancer drugs—Gilteritinib, Ivosidenib, Daunorubicin, Glasdegib, and Palbociclib—by calculating CNTIs and several standard topological indices. A quantitative structure–property relationship (QSPR) model is developed through linear regression to predict properties such as boiling point, flash point, molar volume, molecular weight, and complexity. The results demonstrate that CNTIs are reliable and effective in capturing structure–property relationships, underscoring their potential application in drug design and cancer treatment research.