Screening for immune-related biomarkers associated with myasthenia gravis and dilated cardiomyopathy based on bioinformatics analysis and machine learning

Guiting Zhou; Shushu Wang; Liwen Lin; Kachun Lu; Zhichao Lin; Ziyan Zhang; Yuling Zhang; Danling Cheng; KaMan Szeto; Rui Peng; Chuanjin Luo

doi:10.1016/j.heliyon.2024.e28446

Screening for immune-related biomarkers associated with myasthenia gravis and dilated cardiomyopathy based on bioinformatics analysis and machine learning

Heliyon. 2024 Mar 20;10(7):e28446. doi: 10.1016/j.heliyon.2024.e28446. eCollection 2024 Apr 15.

Authors

Guiting Zhou¹, Shushu Wang¹, Liwen Lin¹, Kachun Lu¹, Zhichao Lin¹, Ziyan Zhang², Yuling Zhang¹, Danling Cheng¹, KaMan Szeto¹, Rui Peng³, Chuanjin Luo^{4

5}

Affiliations

¹ The First Clinical Medical College, Guangzhou University of Chinese Medicine, Guangzhou, China.
² Zhongshan Traditional Chinese Medicine Hospital, Zhongshan, China.
³ Shunde Hospital of Guangzhou University of Chinese Medicine, Foshan, China.
⁴ Cardiology Center, The First Affiliated Hospital of Guangzhou University of Chinese Medicine, Guangzhou, China.
⁵ Guangdong Clinical Research Academy of Chinese Medicine, Guangzhou, China.

Abstract

Background: We aim to investigate genes associated with myasthenia gravis (MG), specifically those potentially implicated in the pathogenesis of dilated cardiomyopathy (DCM). Additionally, we seek to identify potential biomarkers for diagnosing myasthenia gravis co-occurring with DCM.

Methods: We obtained two expression profiling datasets related to DCM and MG from the Gene Expression Omnibus (GEO). Subsequently, we conducted differential gene expression analysis and weighted gene co-expression network analysis (WGCNA) on these datasets. The genes exhibiting differential expression common to both DCM and MG were employed for protein-protein interaction (PPI), Gene Ontology (GO) enrichment analysis, and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis. Additionally, machine learning techniques were employed to identify potential biomarkers and develop a diagnostic nomogram for predicting MG-associated DCM. Subsequently, the machine learning results underwent validation using an external dataset. Finally, gene set enrichment analysis (GSEA) and machine algorithm analysis were conducted on pivotal model genes to further elucidate their potential mechanisms in MG-associated DCM.

Results: In our analysis of both DCM and MG datasets, we identified 2641 critical module genes and 11 differentially expressed genes shared between the two conditions. Enrichment analysis disclosed that these 11 genes primarily pertain to inflammation and immune regulation. Connectivity map (CMAP) analysis pinpointed SB-216763 as a potential drug for DCM treatment. The results from machine learning indicated the substantial diagnostic value of midline 1 interacting protein1 (MID1IP1) and PI3K-interacting protein 1 (PIK3IP1) in MG-associated DCM. These two hub genes were chosen as candidate biomarkers and employed to formulate a diagnostic nomogram with optimal diagnostic performance through machine learning. Simultaneously, single-gene GSEA results and immune cell infiltration analysis unveiled immune dysregulation in both DCM and MG, with MID1IP1 and PIK3IP1 showing significant associations with invasive immune cells.

Conclusion: We have elucidated the inflammatory and immune pathways associated with MG-related DCM and formulated a diagnostic nomogram for DCM utilizing MID1IP1/PIK3IP1. This contribution offers novel insights for prospective diagnostic approaches and therapeutic interventions in the context of MG coexisting with DCM.

Keywords: Diagnostic value; Dilated cardiomyopathy; Immune cell infiltration; Machine learning algorithms; Myasthenia gravis.