AUTHOR=Su Jiaming , Peng Jing , Wang Lin , Xie Huidi , Zhou Ying , Chen Haimin , Shi Yang , Guo Yan , Zheng Yicheng , Guo Yuxin , Dong Zhaoxi , Zhang Xianhui , Liu Hongfang 

TITLE=Identification of endoplasmic reticulum stress-related biomarkers of diabetes nephropathy based on bioinformatics and machine learning

JOURNAL=Frontiers in Endocrinology

VOLUME=Volume 14 - 2023

YEAR=2023

URL=https://www.frontiersin.org/journals/endocrinology/articles/10.3389/fendo.2023.1206154

DOI=10.3389/fendo.2023.1206154

ISSN=1664-2392

ABSTRACT=Backgrounds: Diabetes nephropathy (DN) is a growing public health concern worldwide. Renal dysfunction and impairment in DN are intimately linked to ER stress and its related signaling pathways. Nevertheless, the underlying mechanisms and biomarkers for this function of ER stress in the DN remain unknown.
Methods: Microarray datasets were retrieved from the Gene Expression Omnibus (GEO) database, and ER stress-related genes (ERSRGs) were downloaded from the MSigDB and GeneCards databases. We identified hub ERSRGs for DN progression by intersecting ERSRGs with differentially expressed genes and significant genes in WGCNA, followed by a functional analysis. After analyzing hub ERSRGs with three machine learning techniques and taking the intersection, we performed external validation as well as developed a DN diagnostic model based on the characteristic genes. Immune infiltration was performed using CIBERSORT. Moreover, patients with DN were then categorized using a consensus clustering approach. Eventually, the candidate ERSRGs-specific small-molecule compounds were defined by CMAP.
Results: Several biological pathways driving pathological injury to DN and disordered levels of immune infiltration were revealed in the DN microarray datasets and strongly related to deregulated ERSRGs by bioinformatics multi-chip integration. Moreover, CDKN1B, EGR1, FKBP5, GDF15, and MARCKS were identified as ER stress signature genes associated with DN by machine learning algorithms, demonstrating their potential as DN biomarkers. 
Conclusions: Our research sheds fresh light on the function of ER stress in DN pathophysiology and the development of early diagnostic and ER stress-related treatment targets in patients with DN.