Table 1.
Summarizes the RA datasets included in our study.
Fig 1.
The flowchart of this study.
Fig 2.
Screening of the cellular senescence-associated signature genes.
(A) The Volcano plots showing significantly DEGs (P-value < 0.05, |log2 Fold Change| > 0.5) in RA vs. normal samples. (B) The overlapping genes between DEGs and cellular senescence-associated genes. (C) The dot plots show the accuracy of model in each time. (D) Partial likelihood deviance for the LASSO regression.
Fig 3.
(A) GO functional analysis of the cellular senescence related DEGs between RA and control. (B) KEGG pathway analysis of the cellular senescence related DEGs between RA and control.
Fig 4.
The ROC curve analysis to assess the predictive ability of the CSscore model in training dataset (GSE55457).
Fig 5.
Validate the predictive value of the SCscore model in four independent external cohorts.
(A) The ROC curve of CSscore model in GSE12021. (B) The ROC curve of CSscore model in GSE55235. (C) The ROC curve of CSscore model in GSE77298. (D) The ROC curve of CSscore model in GSE178557.
Fig 6.
Compare the infiltration of immune cells between RA and normal samples.
The infiltration of immune cells was estimated by (A) CIBERSORT (B) ssGSEA (C) xCell.
Fig 7.
Correlation analysis between the expression level of immune checkpoint related genes.
Fig 8.
Consensus clustering analysis.
(A) The samples were clustered into three subtypes based on the expression of cellular senescence-related DEGs by consensus clustering method. (B) The expression heatmap of cellular senescence-related DEGs between three subtypes samples. (C) The expression boxplot of immune checkpoint related genes between three subtypes samples. (D) The infiltration boxplot of immune cells between three subtypes samples.
Fig 9.
Association between signature genes and drug sensitivity.
(A) Correlation analysis between the expression level of signature genes and the IC50 values of drugs. (B) The AutoDock results of gene CYR61 and drug ibuprofen. (C) The AutoDock results of gene CYR 61 and drug indometacin. (D) The AutoDock results of gene DHX9 and drug ibuprofen. (E) The AutoDock results of gene DHX9 and drug indometacin.