Bioinformatics methods in biomarkers of preeclampsia and associated potential drug applications

BACKGROUND: Preeclampsia is a pregnancy-related condition that causes high blood pressure and proteinuria after 20 weeks of pregnancy. It is linked to increased maternal mortality, organ malfunction, and foetal development limitation. In this view, there is a need critical to identify biomarkers for the early detection of preeclampsia. The objective of this study is to discover critical genes and explore medications for preeclampsia treatment that may influence these genes. METHODS: Four datasets, including GSE10588, GSE25906, GSE48424 and GSE60438 were retrieved from the Gene Expression Omnibus database. The GSE10588, GSE25906, and GSE48424 datasets were then removed the batch effect using the “sva” R package and merged into a complete dataset. The differentially expressed genes (DEGs) were identified using the “limma” R package. The potential small-molecule agents for the treatment of PE was further screened using the Connective Map (CMAP) drug database based on the DEGs. Further, Weight gene Co-expression network (WGNCA) analysis was performed to identified gene module associated with preeclampsia, hub genes were then identified using the logistic regression analysis. Finally, the immune cell infiltration level of genes was evaluated through the single sample gene set enrichment analysis (ssGSEA). RESULTS: A total of 681 DEGs (376 down-regulated and 305 up-regulated genes) were identified between normal and preeclampsia samples. Then, Dexamethasone, Prednisone, Rimexolone, Piretanide, Trazodone, Buflomedil, Scoulerin, Irinotecan, and Camptothecin drugs were screened based on these DEGs through the CMAP database. Two modules including yellow and brown modules were the most associated with disease through the WGCNA analysis. KEGG analysis revealed that the chemokine signaling pathway, Th1 and Th2 cell differentiation, B cell receptor signalling pathway and oxytocin signalling pathway were significantly enriched in these modules. Moreover, two key genes, PLEK and LEP were evaluated using the univariate and multivariate logistic regression analysis from the hub modules. These two genes were further validated in the external validation cohort GSE60438 and qRT-PCR experiment. Finally, we evaluated the relationship between immune cell and two genes. CONCLUSION: In conclusion, the present study investigated key genes associated with PE pathogenesis that may contribute to identifying potential biomarkers, therapeutic agents and developing personalized treatment for PE. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1186/s12864-022-08937-3.