AUTHOR=Long Qing , Wang Rui , Feng Maoyang , Zhao Xinling , Liu Yilin , Ma Xiao , Yu Lei , Li Shujun , Guo Zeyi , Zhu Yun , Teng Zhaowei , Zeng Yong 

TITLE=Construction and Analysis of a Diagnostic Model Based on Differential Expression Genes in Patients With Major Depressive Disorder

JOURNAL=Frontiers in Psychiatry

VOLUME=12

YEAR=2021

URL=https://www.frontiersin.org/journals/psychiatry/articles/10.3389/fpsyt.2021.762683

DOI=10.3389/fpsyt.2021.762683

ISSN=1664-0640

ABSTRACT=<p><bold>Background:</bold> Major depressive disorder (MDD) is a common and severe psychiatric disorder with a heavy burden on the individual and society. However, the prevalence varies significantly owing to the lack of auxiliary diagnostic biomarkers. To identify the shared differential expression genes (DEGs) with potential diagnostic value in both the hippocampus and whole blood, a systematic and integrated bioinformatics analysis was carried out.</p><p><bold>Methods:</bold> Two datasets from the Gene Expression Omnibus database (GSE53987 and GSE98793) were downloaded and analyzed separately. A weighted gene co-expression network analysis was performed to construct the co-expression gene network of DEGs from GSE53987, and the most disease-related module was extracted. The shared DEGs from the module and GSE98793 were identified using a Venn diagram. Functional pathway prediction was used to identify the most disease-related DEGs. Finally, several DEGs were chosen, and their potential diagnostic value was determined by receiver operating characteristic curve analysis.</p><p><bold>Results:</bold> After weighted gene co-expression network analysis, the most MDD-related module (MEgrey) was identified, and 623 DEGs were extracted from this module. The intersection between MEgrey and GSE98793 was calculated, and 163 common DEGs were identified. The co-expression network of 163 DEGs from these was then reconstructed. All hub genes were identified based on the connective degree of the reconstructed co-expression network. Based on the results of functional pathway enrichment, 17 candidate hub genes were identified. Finally, logistic regression and receiver operating characteristic curves showed that three candidate hub genes (<italic>CEP350, SMAD5</italic>, and <italic>HSPG2</italic>) had relatively high auxiliary value in the diagnosis of MDD.</p><p><bold>Conclusion:</bold> Our results showed that the combination of <italic>CEP350, SMAD5</italic>, and <italic>HSPG2</italic> has a relatively high diagnostic value for MDD. Pathway enrichment analysis also showed that these genes may play an important role in the pathogenesis of MDD. These results suggest a potentially important role for this gene combination in clinical practice.</p>