Weighted gene correlation network analysis reveals novel regulatory modules associated with recurrent early pregnancy loss

At present, the etiology and pathogenesis of recurrent early pregnancy loss (REPL) are not completely clear. Therefore, identifying the underlying diagnostic and prognostic biomarkers of REPL can provide new ideas for the diagnosis and treatment of REPL. The chip data of REPL (GSE63901) were downloaded from the NCBI Gene Expression Omnibus (GEO) database. Weighted Gene Co-Expression Network Analysis (WGCNA) was used to construct a co-expression module for studying the relationship between gene modules and clinical features. In addition, functional analysis of hub genes in modules of interest was performed. A total of 23 co-expression modules were identified, two of which were most significantly associated with three clinical features. The MEbrown module was positively correlated with cyclin E level and the out-of-phase trait while the MEred module was positively correlated with the effect of progesterone. We identified 17 hub genes in the MEred module. The functional enrichment analysis indicated that such hub genes were mainly involved in pathways related to cellular defense response and natural killer (NK) cell-mediated cytotoxicity. In the MEbrown module, we identified 19 hub genes, which were mainly enriched in cell adhesion molecule production, regulation of cellular response to growth factor stimulus, epithelial cell proliferation, and transforming growth factor-β (TGF-β) signaling pathway. In addition, the hub genes were validated by using other datasets and three true hub genes were finally obtained, namely DOCK2 for the MEred module, and TRMT44 and ERVMER34-1 for the MEbrown module. In conclusion, our results screened potential biomarkers that might contribute to the diagnosis and treatment of REPL.