Abstract
Early identification of ischemic heart failure (IHF) is critical for improving patient prognosis and clinical outcomes. However, effective diagnostic biomarkers and targeted therapeutic strategies for IHF remain limited. Total flavonoids from Dracocephalum moldavica L. (TFDM) exert potential cardioprotective effects; however, the molecular mechanisms by which TFDM acts against IHF have not been fully elucidated. Therefore, this study aims to identify diagnostic biomarkers for IHF and explore the potential therapeutic mechanism of TFDM targeting these key genes. Given the small sample size (n = 17) of the clinical dataset, LASSO regression and Random Forest were employed due to their superior performance in feature selection, noise reduction, and stability in small-sample scenarios. In this study, we screened key characteristic genes of IHF through bioinformatics analysis and further investigated the binding potential between these key genes and active components of TFDM using molecular docking, thus providing new targets for the early diagnosis of IHF and new evidence for the intervention mechanism of TFDM in IHF.