Abstract 4143994: Unraveling the Genetic Overlaps in Heart Failure Subtypes: A Systems Biology Approach

Background: Heart failure (HF), a progressive condition with a complex etiopathology causes significant morbidity, mortality, and reduced quality of life. HF affects 64 million people worldwide and imposes a global healthcare burden exceeding $346 billion annually. Currently, HF is classified into two major clinical subtypes based on cardiac contractility: HF with reduced ejection fraction (HFrEF) and with preserved ejection fraction (HFpEF). Given the complex pathophysiology of HF, we utilized systems biology tools applied to gene-level results from genome-wide association studies (GWAS) of HF, HFrEF, and HFpEF to identify underlying molecular pathways underlying these phenotypes. Methods: We utilized the union of three SNP-to-gene mapping methods: SNP-nearest gene, MAGMA, and H-MAGMA which uses Hi-C datasets from cardiovascular, lung, and liver tissues. Additionally, we built a gene-gene multiplex network from protein-protein interactions, transcription factor-target relationships, and predictive expression networks created with explainable-Artificial Intelligence (AI) algorithms from cardiac/metabolic layers, aorta, atrial appendage, left ventricle, coronary artery, whole blood plus lymphocytes, lung, and liver. The multiplex network was then used by the systems biology tools Gene set Refinement through Interacting Networks (GRIN) to refine the genes assigned from SNPs, while Multiplex Embedding of Networks For Team-Based Omics Research (MENTOR) grouped genes into functional clusters using hierarchical clustering of network embeddings. Results: Multiple SNP-to-gene assignment methods identified 51 genes for HF, 6 for HFpEF, and 109 for HFrEF. GRIN retained 42 connected genes for HF, 4 genes for HFpEF, and 78 genes for HFrEF. Ultimately, 119 unique genes were identified in total, with 86 of them being retained by GRIN. MENTOR clustering identified key biological processes affected by these genes including cytoskeletal function, extracellular matrix composition, inflammatory cytokine activity, and post-transcriptional modification of mRNA. Conclusion: Utilizing systems biology approaches, we were able to identify biologically connected genes and distinct molecular pathways underlying HF, HFrEF, and HFpEF.