Classification and biomarker gene selection of pyroptosis-related gene expression in psoriasis using a random forest algorithm

Abstract

Background: Psoriasis is a chronic and immune-mediated skin disorder that currently has no cure. Pyroptosis has been proved to be involved in the pathogenesis and progression of psoriasis. However, the role pyroptosis plays in psoriasis remains elusive. Methods: RNA-sequencing data of psoriasis patients were obtained from the Gene Expression Omnibus (GEO) database, and differentially expressed pyroptosis-related genes (PRGs) between psoriasis patients and normal individuals were obtained. A principal component analysis (PCA) was conducted to determine whether PRGs could be used to distinguish the samples. PRG and immune cell correlation was also investigated. Subsequently, a novel diagnostic model comprising PRGs for psoriasis was constructed using a random forest algorithm (ntree = 400). A receiver operating characteristic (ROC) analysis was used to evaluate the classification performance through both internal and external validation. Consensus clustering analysis was used to investigate whether there was a difference in biological functions within PRG-based subtypes. Finally, the expression of the kernel PRGs were validated in vivo by qRT-PCR. Results: We identified a total of 39 PRGs, which could distinguish psoriasis samples from normal samples. The process of T cell CD4 memory activated and mast cells resting were correlated with PRGs. Ten PRGs, IL-1β, AIM2, CASP5, DHX9, CASP4, CYCS, CASP1, GZMB, CHMP2B, and CASP8, were subsequently screened using a random forest diagnostic model. ROC analysis revealed that our model has good diagnostic performance in both internal validation (area under the curve [AUC] = 0.930 [95% CI 0.877-0.984]) and external validation (mean AUC = 0.852). PRG subtypes indicated differences in metabolic processes and the MAPK signaling pathway. Finally, the qRT-PCR results demonstrated the apparent dysregulation of PRGs in psoriasis, especially AIM2 and GZMB. Conclusion: Pyroptosis may play a crucial role in psoriasis and could provide new insights into the diagnosis and underlying mechanisms of psoriasis.

Keywords: machine learning; psoriasis; pyroptosis; pyroptosis-related genes; random forest algorithm.

FIGURE 1

Overview of pyroptosis regulatory pathways.

FIGURE 2

Article framework and workflow. PRGs, pyroptosis-related genes. PPI, protein–protein interaction. DE, differentially expressed.

FIGURE 3

PRGs correlation and interaction analysis, and DE PRGs identification between psoriasis and normal samples (A). Spearman correlation analysis of the 39 differentially expressed PRGs; blue represents a positive correlation, red represents a negative correlation, and the darker the color, the stronger the correlation. (B). Protein–protein interaction (PPI) analysis of the 39 differentially expressed PRGs. (C). Heatmap of the 39 differentially expressed PRGs in psoriasis and normal samples. (D). Principal component analysis (PCA) of PRGs.

FIGURE 4

Heatmap of the differentially expressed PRGs in various immune cells. (A). PRG correlation map with immune cells, where the horizontal axis represents PRGs and the vertical axis represents immune cells. (B) The correlation map marked with absolute value. Asterisks represent levels of significance *p < 0.05.

FIGURE 5

A random forest algorithm was used for ranking the importance of prognostic PRGs. (A,B) The importance of PRGs using the scores returned by the random forest model. (C) Parameter optimization was initially performed using randomly generated parameter sets, ntree = 400 was selected in the modeling, and stability was achieved when 100 random samples were taken, three quarters of the dataset was grouped as the training set and the remaining one quarter as the validation set in each repetition. (D) The receiver operating characteristic (ROC) curve (area under the curve (AUC) = 0.93) was generated by cross-internal validation. (E) ROC curve on independent external validation, from left to right are GSE117239, GSE109248, and GSE14905, respectively.

FIGURE 6

Consensus clustering analysis of PRGs. (A) The cumulative distribution function (CDF) curve for each category number k compared with k -1.(B) Delta area curve of consensus clustering, indicating the relative change in area under the CDF. The horizontal axis represents the category number k, and the vertical axis represents the relative change in area under the CDF curve. (C) A heatmap showing the consensus clustering solution (k = 3) for 39 PRGs in three clusters. (D) The PCA results show these three clusters were significantly differentiated, especially clusters 1 and 2.

FIGURE 7

Comparison of biological functions analysis between clusters 1 and 2. (A) The secondary classification of GO, including biological processes, molecular functions, and cellular components. (B) Bubble map of the GO analysis, ranked by p-value. (C) Bubble map of KEGG signaling pathways, ranked by Q-value. GO, Gene Ontology. KEGG, Kyoto Encyclopedia of Genes and Genomes.

FIGURE 8

mRNA levels of the candidate PRGs of imiquimod (IMQ)-induced psoriasis-like mice compared with the control group (n = 5). The data are expressed as means ± SD. Four skin lesions in each group were included in the analysis. *p < 0.05, **p < 0.01, compared with the control group.