Analysis of signature genes and association with immune cells infiltration in pediatric septic shock

Abstract

BackgroundEarly diagnosis of septic shock in children is critical for prognosis. This study committed to investigate the signature genes and their connection with immune cells in pediatric septic shock.MethodsWe screened a dataset of children with septic shock from the GEO database and analyzed differentially expressed genes (DEGs). Functional enrichment analysis was performed for these DEGs. Weighted gene co-expression network analysis (WCGNA) was used to screen the key modules. Least absolute shrinkage and selection operator (LASSO) and random forest analysis were finally applied to identify the signature genes. Then gene set enrichment analysis (GSEA) was exerted to explore the signaling pathways related to the hub genes. And the immune cells infiltration was subsequently classified via using CIBERSORT.ResultsA total of 534 DEGs were screened from GSE26440. The data then was clustered into 17 modules via WGCNA, which MEgrey module was significantly related to pediatric septic shock (cor=−0.62, p<0.0001). LASSO and random forest algorithms were applied to select the signature genes, containing UPP1, S100A9, KIF1B, S100A12, SLC26A8. The receiver operating characteristic curve (ROC) of these signature genes was 0.965, 0.977, 0.984, 0.991 and 0.989, respectively, which were verified in the external dataset from GSE13904. GSEA analysis showed these signature genes involve in positively correlated fructose and mannose metabolism and starch and sucrose metabolism signaling pathway. CIBERSORT suggested these signature genes may participate in immune cells infiltration.ConclusionUPP1, S100A9, KIF1B, S100A12, SLC26A8 emerge remarkable diagnostic performance in pediatric septic shock and involved in immune cells infiltration.

Use of transcriptomics for diagnosis of infections and sepsis in children: A narrative review

“…Fan et al 27 . recently analysed a dataset of children with septic shock compared to healthy controls to search DEGs and they confirmed the results observed by Liang et al 24 …”

“…25 Similarly, Xie et al 26 found 160 upregulated genes and 61 downregulated genes in children with sepsis; the upregulated genes included sucrose metabolism, osteoclast differentiation and tumour necrosis factor (TNF) pathway. 26 Fan et al 27 recently analysed a dataset of children with septic shock compared to healthy controls to search DEGs and they confirmed the results observed by Liang et al 24 ; they also found an activation of genes related to secretory granule lumen, cytoplasmatic vesicle lumen, vesicle lumen and carbohydrate binding and endopeptidase activity. These genes were in patients with shock sepsis and not upregulated in healthy controls, suggesting a potential role in forecasting children with shock sepsis.…”

“…; they also found an activation of genes related to secretory granule lumen, cytoplasmatic vesicle lumen, vesicle lumen and carbohydrate binding and endopeptidase activity. These genes were upregulated in patients with shock sepsis and not upregulated in healthy controls, suggesting a potential role in forecasting children with shock sepsis 27 …”

“…These genes were in patients with shock sepsis and not upregulated in healthy controls, suggesting a potential role in forecasting children with shock sepsis. 27 Zhou et al 28 proposed a 10-core genes panel in paediatric sepsis (MALT1, NLRC3, MAPK14, LTB4R, ITGAM, DOK3, HCK, PRKCD, SIRPA, LCK) extrapolated from three datasets of sepsis as a possible biochemical diagnostic panel to diagnose sepsis to integrate traditional biochemical tests. 28 Genome-base techniques were proposed also as a method to differentiate sepsis from other conditions with a massive activation of inflammation not related to infection, such as INSI.…”

Novel prognostic biomarker TBC1D1 is associated with immunotherapy resistance in gliomas

“…We employed the GSEA computational method to assess the statistical significance of a preselected gene set (15)(16)(17)(18). Following correlation analysis, we generated an initial list of gene categories.…”