Fig. 2: Selecting small ncRNAs in FF-sEVs as biomarkers.

a Flowchart of the process for the small ncRNA selection. b Volcano plot of the small ncRNAs. The log2FC, P values, and IfcSE were calculated using the DEseq2 package (ver. 1.36.0). c The heatmap analysis for 430 differentially expressed ncRNAs in FF-sEVs. The normalized data were converted to base 10 logarithms and z-score. P means samples in pregnant groups and N means samples in non-pregnant groups. d The principal component analysis of the 430 ncRNAs. Blue and red circles indicate the non-pregnant FF and pregnant FF, respectively. PC means principal component. e Dot plots showing the performance of the 430 differentially expressed ncRNAs, and ROC curves for miR-16-2-3p, miR-378a-3p, and miR-483-5p. Red and blue circles indicate the upregulated and downregulated miRNAs, respectively. Red and blue circles indicate the upregulated and downregulated miRNAs, respectively. Yellow and green triangles indicate the upregulated and downregulated piRNAs, respectively. AUC means area under the curve. f Correlation diagram of the 14 small ncRNAs. g ROC curve of the combination of the three small ncRNAs (miR-16-2-3p, miR-378a-3p, and miR-483-5p). h Dot plot of the propensity score calculated using the three small ncRNAs (miR-16-2-3p, miR-378a-3p, and miR-483-5p). **P < 0.01. The data were analyzed by the Welch’s t test. The pregnant group: n = 8, non-pregnant group: n = 12, respectively, biologically independent samples.