. 2019 Sep 4;14(9):e0222024. doi: 10.1371/journal.pone.0222024

Table 2. Construction of prediction models using miRNAs.

Number of miRNAs	Sensitivity	Specificity	Accuracy	AUC
1^a	0.92	0.27	0.50	0.62
2^b	0.79	0.53	0.62	0.72
3^c	0.77	0.60	0.66	0.74
5^d	0.83	0.62	0.69	0.76
6^e	0.69	0.76	0.73	0.77
7^f	0.69	0.77	0.74	0.78
8^g	0.69	0.82	0.76	0.79

a. (-0.0239)*miR-575–0.4624.

b. (-0.04785)*miR-575+(-0.01639)*miR-3160-5p -0.18945.

c. (-0.004077)*miR-296-3p+(-0.09225)*miR-575+(-0.050573)*miR-3160-5p +0.377653.

d. (-0.004538)*miR-8089+(-0.090765)*miR-296-3p+(-0.184948)*miR-575+(-0.00378)*miR-4710+(-0.106025)*miR-3160-5p +2.096777.

e. (-0.056646)*miR-8089+(-0.002592)*miR-4755-3p+(-0.115385)*miR-296-3p+(-0.220128)*miR-575+(-0.015284)*miR-4710+(-0.11774)*miR-3160-5p +3.061642.

f. (-0.002883)*miR-4483+(-0.079497)*miR-8089+(-0.005242)*miR-4755-3p+(-0.12544)*miR-296-3p+(-0.237057)*miR-575+(-0.020395)*miR-4710+(-0.122488)*miR-3160-5p +3.51243.

g. (-0.01801)*miR-4483+(-0.14291)*miR-8089+(-0.01134)*miR-4755-3p+(-0.14277)*miR-296-3p+(-0.28373)*miR-575+(-0.03178)*miR-4710+(-0.01287)*miR-5698+(-0.13961)*miR-3160-5p +4.80275.