Analysis of characteristic genes and ceRNA regulation mechanism of endometriosis based on full transcriptional sequencing

Abstract

Background: Endometriosis is a common gynecological disorder that usually causes infertility, pelvic pain, and ovarian masses. This study aimed to mine the characteristic genes of endometriosis, and explore the regulatory mechanism and potential therapeutic drugs based on whole transcriptome sequencing data and resources from public databases, providing a theoretical basis for the diagnosis and treatment of endometriosis.Methods: The transcriptome data of the five eutopic (EU) and ectopic (EC) endometrium samples were obtained from Beijing Obstetrics and Gynecology Hospital, Beijing, China, and dinified as the own data set. The expression and clinical data of EC and EU samples in GSE25628 and GSE7305 datasets were obtained from the GEO database (https://www.ncbi.nlm.nih.gov/gds). Differential gene expression analysis and weighted gene co-expression network analysis (WGCNA) were used to identify the endometriosis-related differentially expressed genes. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses were conducted by the “clusterProfiler” R package. Then, characteristic genes for endometriosis were identified by the least absolute shrinkage and selection operator (LASSO) and support vector machine recursive feature elimination (SVM-RFE) algorithm. The expression of characteristic genes was verified by quantitative reverse transcription polymerase chain reaction (qRT-PCR) and western-blot. The receiver operating characteristic (ROC) curve was used to evaluate the discriminatory ability of characteristic genes. We assessed the abundance of infiltrating immune cells in each sample using MCP-counter and ImmuCellAI algorithms. The competitive endogenous RNA (ceRNA) regulatory network of characteristic genes was created by Cytoscape and potential targeting drugs were obtained in the CTD database.Results: 44 endometriosis-related differentially expressed genes were obtained from GSE25628 and the own dataset. Subsequently, LASSO and SVM-RFE algorithms identified four characteristic genes, namely ACLY, PTGFR, ADH1B, and MYOM1. The results of RT-PCR and western-blot were consistent with those of sequencing. The result of ROC curves indicated that the characteristic genes had powerful abilities in distinguishing EC samples from EU samples. Infiltrating immune cells analysis suggested that there was a certain difference in immune microenvironment between EC and EU samples. The characteristic genes were significantly correlated with specific differential immune cells between EC and EU samples. Then, a ceRNA regulatory network of characteristic genes was constructed and showed a total of 7, 11, 11, and 1 miRNA associated with ACLY, ADH1B, PTGFR, and MYOM1, respectively. Finally, we constructed a gene-compound network and mined 30 drugs targeting ACLY, 33 drugs targeting ADH1B, 13 drugs targeting MYOM1, and 12 drugs targeting PTGFR.Conclusion: Comprehensive bioinformatic analysis was used to identify characteristic genes, and explore ceRNA regulatory network and potential therapeutic agents for endometriosis. Altogether, these findings provide new insights into the diagnosis and treatment of endometriosis.

SKIL/SnoN attenuates TGF-β1/SMAD signaling-dependent collagen synthesis in hepatic fibrosis

“…135 DE-circRNAs were found in the GSE134146 dataset, including 62 upregulated circRNAs and 73 downregulated circRNAs. We then took the intersection of DE-miRNAs and miRNAs targeting SKIL genes and retained the miRNA–mRNA relationship pairs with opposite expression trends [ 20 ]. The intersection of DE-circRNAs and circRNAs targeting the corresponding miRNAs was taken, and circRNA–miRNA relationship pairs with opposite trends were retained [ 20 ].…”

“…We then took the intersection of DE-miRNAs and miRNAs targeting SKIL genes and retained the miRNA–mRNA relationship pairs with opposite expression trends [ 20 ]. The intersection of DE-circRNAs and circRNAs targeting the corresponding miRNAs was taken, and circRNA–miRNA relationship pairs with opposite trends were retained [ 20 ]. Finally, a ceRNA network with six nodes (1 mRNA, 2 miRNAs, and 3 circRNAs) and six edges was generated ( Figure 6A ).…”

Identification of Inflammatory Gene in the Congenital Heart Surgery Patients following Cardiopulmonary Bypass via the Way of WGCNA and Machine Learning Algorithms

“…In this study, we searched for the gene and signaling pathway between CPB and inflammation through comprehensive bioinformatics analysis [11]. Initially, we got the GSE143780 dataset from the Gene Expression Omnibus (GEO) database and performed gene coexpression network analysis (WGCNA) and differential analysis.…”