An integration analysis of mRNAs and miRNAs microarray data to identify key regulators for ovarian endometriosis based on competing endogenous RNAs

This study aimed to uncover effects of non-coding RNA transcripts on ovarian endometriosis (OEM) development. Two transcription datasets (GSE105764 and GSE105765) about OME were downloaded from Gene Expression Omnibus (GEO) database and the differentially expressed mRNAs, lncRNAs and miRNAs (DEmRNAs, DElncRNAs and DEmiRNAs) between OEM cases and controls were identified followed by protein-protein interaction analysis. Then, co-expression analysis was conducted and DEmiRNAs-DEmRNAs as well as DElncRNAs-DEmiRNAs pairs were predicted to construct the ceRNA network followed by sub-ceRNA network associated with OEM extraction. Functional analyses of DEmRNAs in ceRNA and sub-module network and the survival analysis were also performed to evaluate the correlation of key regulators and OV outcomes. Totally, 1910 DEmRNAs, 158 DElncRNAs and 118 DEmiRNAs were screened between OEM cases and controls and the functional analyses of DEmRNAs showed that they were significantly enriched in cell adhesion. Furthermore, there were 505 nodes in PPI network and ceRNA network included 762 interaction pairs among 357 DEmRNAs, 28 DElncRNAs and 24 DEmiRNAs. The KEGG analysis suggested several genes including FOXO1, STAT5A and RUNX1 were predominately associated with pathways in cancer while IL15 was primarily enriched in cytokine-cytokine receptor interaction pathway. Importantly, these two pathways were also found to be implicated with OEM development. Finally, survival analysis implied that overexpression of ZFPM2-AS1 had a good clinical outcome while the under-expression levels of FOXO1, JUP, STAT5A, RUNX1 and PRKG1-AS1 exhibited a better prognosis for OV. FOXO1, STAT5A, RUNX1 and IL15, PRKG1-AS1 and ZFPM2-AS1 were promising diagnostic makers for OME.