Skip to main content
Fig. 5 | Arthritis Research & Therapy

Fig. 5

From: Identification and verification of a novel signature that combines cuproptosis-related genes with ferroptosis-related genes in osteoarthritis using bioinformatics analysis and experimental validation

Fig. 5

Machine learning-based potential biomarker screening. a SVM-RFE model with the optimal error rate when the number of signature genes was 58. b LASSO regression model. c Random forest model and the top 20 genes in terms of importance. d The final biomarkers screened using three machine learning algorithms

Back to article page