Using artificial intelligence and promoter-level transcriptome analysis to identify a biomarker as a possible prognostic predictor of cardiac complications in male patients with Fabry disease

Abstract

Fabry disease is the most frequently occurring form of lysosomal disease in Japan, and is characterized by a wide variety of conditions. Primarily, the three major types of concerns associated with Fabry disease observed during adulthood that must be prevented are central nervous system, renal, and cardiac complications. Cardiac complications, such as cardiomyopathy, cardiac muscle fibrosis, and severe arrhythmia, are the most common mortality causes in patients with Fabry disease. To predict cardiac complications of Fabry disease, we extracted RNA from the venous blood of patients for cap analysis of gene expression (CAGE), performed likelihood ratio tests for each RNA expression dataset obtained from individuals with and without cardiac complications, and analyzed the correlation between cardiac functional factors observed using magnetic resonance imaging data extracted using artificial intelligence algorithms and RNA expression. Our findings showed that CHN1 expression was significantly higher in male Fabry disease patients with cardiac complications and that it could be associated with many cardiac functional factors. CHN1 encodes a GTPase-activating protein, chimerin 1, which is specific to the GTP-binding protein Rac (involved in oxidative stress generation and the promotion of myocardial fibrosis). Thus, CHN1 is a potential predictive biomarker of cardiac complications in Fabry disease; however, further studies are required to confirm this observation.