Fig. 3
Diagnostic performance of m6A-miRNAs signature in different clinical conditions and cancer types. A Differences in the output strength of m6A-miRNAs signature between female and male samples visualized by violin plot. B Correlation between m6A-miRNAs output strength and participant age using spearman analysis. Yellow represented the density of samples at different m6A-miRNAs output strength. Blue represented the density of samples at different age. Cor, correlation coefficient. C The ability of m6A-miRNAs signature in distinguishing cancer types was determined by calculating the area under curve. The radar chart summarized the AUC of each cancer type. Red polyline represented the AUC value of m6A-miRNAs in distinguishing each cancer type from non-cancer controls. Blue polyline represented the AUC value of m6A-miRNAs in distinguishing each cancer type from all the mixed samples with cancer and non-cancer. D Summary of AUC, sensitivity, specificity and accuracy of m6A-miRNAs in distinguishing each cancer type from non-cancer controls. E Summary of AUC, sensitivity, specificity and accuracy of m6A-miRNAs in distinguishing each cancer type from all the mixed samples with cancer and non-cancer. F The ROC curve showing the diagnostic value of m6A-miRNAs signature in early gastric cancer. G The ROC curve showing the ability of m6A-miRNAs in distinguishing patients with hepatocellular carcinoma (HCC) and patients with hepatitis/liver cirrhosis. H Principal component analysis for the 18 candidate m6A target miRNAs in HCC and hepatitis/liver cirrhosis samples. Two independent clusters were identified. Red dot, HCC samples; Blue dot, hepatitis/liver cirrhosis samples. I The density of HCC samples and hepatitis/liver cirrhosis at different m6A-miRNAs output strength