QSPR modeling of angiotensin-converting enzyme inhibitors based on degree-based and reverse degree-based topological indices

Abstract Angiotensin-converting enzyme (ACE) inhibitors constitute a widely used group of medications for the clinical management of hypertension, coronary artery disease, heart failure, and stroke. In this study, we present a quantitative structure–property relationship (QSPR) framework for 21 ACE inhibitors that employs novel topological indices based on molecular graphs, including a generalized reverse degree $$\\mathcal {R}_k$$ evaluated over the inhibitors maximum degree, which ensures admissibility across heterogeneous molecular graphs. Degree-based and reverse degree-based indices are incorporated to capture essential local and extended structural characteristics of the molecules. Regression modeling was performed using the topological indices that demonstrated strong statistical associations with the investigated physicochemical properties, with the objective of assessing their predictive capability for estimating the corresponding physicochemical attributes. We tested the robustness and utility of our models using leave-one-out cross-validation (LOO-CV). In addition, the external validation was performed using an independent test set of ACE inhibitors. The external validation results confirmed the predictive reliability of the proposed descriptors. The findings demonstrate that QSPR models become more predictive and coherent when reverse degree-based indicators are added.