Cuproptosis-related genes prediction feature and immune microenvironment in major depressive disorder

Abstract

Background: Major depressive disorder (MDD) is a severe, unpredictable, ill-cured, relapsing neuropsychiatric disorder. A recently identified type of death called cuproptosis has been linked to a number of illnesses. However, the influence of cuproptosis-related genes in MDD has not been comprehensively assessed in prior study.

Aim: This investigation intends to shed light on the predictive value of cuproptosis-related genes for MDD and the immunological microenvironment.

Methods: GSE38206, GSE76826, GSE9653 databases were used to analyze cuproptosis regulators and immune characteristics. To find the genes that were differently expressed, weighted gene co-expression network analysis was employed. We calculated the effectiveness of the random forest model, generalized linear model, and limit gradient lifting to arrive at the best machine prediction model. Nomogram, calibration curve, and decision curve analysis were used to show the anticipated MDD's accuracy.

Results: This study found that there were activated immune responses and cuproptosis-related genes that were dysregulated in people with MDD compared to healthy controls. Considering the test performance of the learned model and validation on subsequent datasets, the RF model (including OSBPL8, VBP1, MTM1, ELK3, and SLC39A6) was considered to have the best discriminative performance. (AUC = 0.875).

Conclusion: Our study constructed a prediction model to predict MDD risk and clarified the potential connection between cuproptosis and MDD.

Keywords: Cuproptosis; Gene; Major depressive disorder; Predictive model; RF model.

Fig. 1

The flow chart of this study.

Fig. 2

19 CRGs (cuproptosis-related genes) express themselves differently in people with MDD and healthy people. (A)The boxplots showed the expression of 19 CRGs. (B)Heatmap differential level is reflected by CRGs with differential expressions.(C)The locations of each chromosome's 19 associated CRGs. (D,E)Diagram of the gene connection network and correlation study of nine CRGs with varied levels of expression. Positive and negative correlations are represented by the colors red and green, respectively. (F)The differences in immunological infiltration between MDD and healthy individual controls were displayed in boxplots. (G) Investigation of the relationships between infiltrating immune cells and 12 differentially expressed CRGs. *p < 0.05, ***p < 0.001, ****p < 0.0001. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

Fig. 3

Molecular and immunological traits that distinguish the two cuproptosis clusters. (A)The heatmap displayed the expression of the two cuproptosis clusters. (B)The differential level in the boxplots is reflected in CRGs with differential expressions. (C)22 immune cells that infiltrate between the two cuproptosis clusters in according to their abundance. (D)The variations in immune infiltration between the two cuproptosis clusters were displayed using boxplots. (E,F)Pathway variations across several CRGs clusters were analyzed using Gene Set Variation Analysis (GSVA). *p < 0.05, ***p < 0.001, ****p < 0.0001.

Fig. 4

Using WCGNA (weighted gene co-expression network analysis), gene modules were screened and co-expression networks were built in two clusters associated with cuproptosis and MDD. The choice of soft threshold power is in (A,E). (B,F) Co-expression module cluster tree dendrogram. Different hues stand for various co-expression modules. (C,G) A heatmap showing the correlations between the 10 modules. (D,H) Analysis of the correlation between the clinical status and the module eigengenes. Each column denotes a clinical status, whereas each row denotes a module.

Fig. 5

Building and assessing RF, GLM, and XGB machine models. (A) The intersection of the MDD module-related genes and the cuproptosis cluster is depicted in a Venn diagram using data from the GSE38206 dataset. (B)A list of significant eigengenes for the RF, GLM, and XGB machine models. (C) Boxplots are used to display residuals for each machine learning model. The root mean square error (RMSE) of the residuals is represented by the red dots. (D) Three machine learning models' ROC analyses. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

Fig. 6

Validation of the RF model based on five genes. (A) Building a nomogram based on the 5-gene RF model to estimate the risk of MDD clusters. (B,C) Building a calibration curve (B) and performing a decision curve analysis (DCA) (C) to determine how well the nomogram model predicts the future. 5-fold cross-validation was used to analyze the receiver operating characteristic curve (ROC) of the 5-gene-based RF model in the datasets GSE39653 (D) and GSE76826 (E).

Figure S1

Identification of cuproptosis-related molecular clusters in CRDs. (A) Consensus clustering matrix when k = 2.(B)Representative cumulative distribution function (CDF) curves ,(C) CDF delta area curves , (D)the score of consensus clustering,(E) difference between the two clusters was determined by PCA(Principal Component Analysis).