Identification of driver genes in lupus nephritis based on comprehensive bioinformatics and machine learning

Zheng Wang; Danni Hu; Guangchang Pei; Rui Zeng; Ying Yao

doi:10.3389/fimmu.2023.1288699

Identification of driver genes in lupus nephritis based on comprehensive bioinformatics and machine learning

Front Immunol. 2023 Dec 7:14:1288699. doi: 10.3389/fimmu.2023.1288699. eCollection 2023.

Authors

Zheng Wang¹, Danni Hu¹, Guangchang Pei¹, Rui Zeng^{1

2

3

4}, Ying Yao^{1

5}

Affiliations

¹ Division of Nephrology, Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, China.
² Key Laboratory of Organ Transplantation, Ministry of Education, Wuhan, China.
³ NHC Key Laboratory of Organ Transplantation, Chinese Academy of Medical Sciences, Wuhan, China.
⁴ Key Laboratory of Organ Transplantation, Chinese Academy of Medical Sciences, Wuhan, China.
⁵ Department of Nutrition, Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, China.

Abstract

Background: Lupus nephritis (LN) is a common and severe glomerulonephritis that often occurs as an organ manifestation of systemic lupus erythematosus (SLE). However, the complex pathological mechanisms associated with LN have hindered the progress of targeted therapies.

Methods: We analyzed glomerular tissues from 133 patients with LN and 51 normal controls using data obtained from the GEO database. Differentially expressed genes (DEGs) were identified and subjected to enrichment analysis. Weighted gene co-expression network analysis (WGCNA) was utilized to identify key gene modules. The least absolute shrinkage and selection operator (LASSO) and random forest were used to identify hub genes. We also analyzed immune cell infiltration using CIBERSORT. Additionally, we investigated the relationships between hub genes and clinicopathological features, as well as examined the distribution and expression of hub genes in the kidney.

Results: A total of 270 DEGs were identified in LN. Using weighted gene co-expression network analysis (WGCNA), we clustered these DEGs into 14 modules. Among them, the turquoise module displayed a significant correlation with LN (cor=0.88, p<0.0001). Machine learning techniques identified four hub genes, namely CD53 (AUC=0.995), TGFBI (AUC=0.997), MS4A6A (AUC=0.994), and HERC6 (AUC=0.999), which are involved in inflammation response and immune activation. CIBERSORT analysis suggested that these hub genes may contribute to immune cell infiltration. Furthermore, these hub genes exhibited strong correlations with the classification, renal function, and proteinuria of LN. Interestingly, the highest hub gene expression score was observed in macrophages.

Conclusion: CD53, TGFBI, MS4A6A, and HERC6 have emerged as promising candidate driver genes for LN. These hub genes hold the potential to offer valuable insights into the molecular diagnosis and treatment of LN.

Keywords: Lupus nephritis; WGCNA; bioinformatics; immune infiltration; machine learning.

Publication types

Research Support, Non-U.S. Gov't

MeSH terms

Computational Biology
Humans
Kidney
Lupus Erythematosus, Systemic*
Lupus Nephritis* / genetics
Machine Learning

Grants and funding

The author(s) declare financial support was received for the research, authorship, and/or publication of this article. This work was supported by the National Natural Science Foundation of China (grant no. 82170701, 81974086, 81974087, 82070739, 81770684).