Fig. 2
Construction of serum m6A-miRNAs signature for cancer detection. A Differences in the output strength of m6A-miRNAs signature between cancer and non-cancer control samples visualized by violin plot. B Differences in the output strength of m6A-miRNAs signature among different cancer types. The red lines represented median value of each group. C The ROC curve showing the diagnostic performance of m6A-miRNAs signature in distinguishing cancer from non-cancer controls in the training cohort. The area under the curve, specificity, sensitivity and accuracy were also calculated. D The diagnostic performance of m6A-miRNAs signature was validated in the internal validation cohort using ROC curve. E The diagnostic performance of m6A-miRNAs signature was investigated in the combined training and validation cohort using ROC curve. F Correlation between m6A-miRNAs output strength and 18 candidate m6A target miRNAs expression using spearman analysis. The red connection represented positive correlation and green connection represented negative correlation. The color depth of line represented the correlation coefficient. G, H In a wide range of decision threshold probability, the difference of net benefit between m6A-miRNA and other serum biomarkers using the decision curve analysis (DCA) in the training cohort (G) and the validation cohort (H)