Fig. 2: Differentially expressed cfRNAs from the 28wkGA group of the discovery cohort.

A The histogram of the p-values from the univariable logistic regression of individual cfRNA models. The p-value was obtained from the Kolmogorov–Smirnov test (2-sided). B The volcano plot showing the degree of statistical significance (p-value) against the effect size (fold change) of the expression difference between the case and the control. The original p-values on the y-axis and the fold changes on the x-axis were obtained from edgeR (F-test). The 6 cfRNAs having the lowest p-value were indicated. After applying Benjamini-Hochberg correction method for multiple tests, those with ≤ 0.05 were indicated in dark-coloured dots, otherwise in pale-coloured dots. C The Venn diagram showing the number cfRNAs that overlap among the four selection criteria. D The heatmap showing the relative level of the 17 shared RNAs identified in the Venn diagram. The z-score represents the normalised Count Per Million (on log2-scale) values.