Integrated bioinformatics and machine learning strategies reveal PRDX6 as the key ferroptosis-associated molecular biosignature of heart failure.

Abstract

Heart failure (HF) is the leading cause of death and public health problems in the global population. This study aimed to identify and validate ferroptosis-related biomarkers associated with HF in clinical medicine using bioinformatics and machine learning strategies. Weighted co-expression network analysis (WGCNA) was applied to screen the module genes and analyze their biological functions and pathways. Ferroptosis-associated genes (FAG) in HF were determined and then machine learning algorithms were used for screening. Next, multiple external independent microarrays were used to verify molecular biosignature. Simultaneously, CIBERSORT was applied to estimate the immune infiltration landscape. Combined with the results of the WGCNA, 25 FAGs were determined and 6 FAMBs were selected by machine learning strategies. In addition, Peroxiredoxin 6 (PRDX6) was finally selected as the key ferroptosis-associated molecular biological feature based on multiple verifications of independent data sets. From the results of the infiltration and enrichment analysis, we believed that PRDX6, as a protective biomarker related to ferroptosis in HF, may help provide new ideas in the immunotherapy of HF.