. 2021 Apr 26;16(4):e0250642. doi: 10.1371/journal.pone.0250642

Table 4. Logistic regression miRNAs combination models for IBC diagnosis.

EV-derived miRNAs		B	SE	95% CI	OR	Percentage Accuracy in classification (PAC)	Nagelkerke R² (%)	p-value
miR-222-3p and let-7a-5p panel	miR-222-3p	0.811	0.27	[0.28, 1.34]	2.25	94.7%	83.8%	< 0.01
miR-222-3p and let-7a-5p panel	let-7a-5p	-1.211	0.35	[-1.89, -0.53]	0.30	94.7%	83.8%	< 0.001
Let-7a-5p and miR-181b-5p panel	let-7a-5p	-0.915	0.29	[-1.48, -0.35]	0.40	89.5%	78.1%	< 0.001
Let-7a-5p and miR-181b-5p panel	miR-181b-5p	0.442	0.15	[0.15, 0.74]	1.56	89.5%	78.1%	< 0.001
miR-181b-5p and miRNA-222-3p panel	miR-181b-5p	0.543	0.17	[0.21, 0.88]	1.72	82.5	41.2%	< 0.001
miR-181b-5p and miRNA-222-3p panel	miRNA-222-3p	-0.061	0.17	[-0.40, 0.28]	0.94	82.5	41.2%	0.724

B: Beta regression coefficient; SE: Standard Error; 95% CI: confidence interval; OR: Odds Ratio.