Table 6.
Regression models to estimate KD by using miRNA-derived parameters in cats.
| Raw level | Normalized to uCRE | Normalized to uRNA | Ratio |
let-7b, let-7b/miR-21a* |
let-7b, let-7b/miR-21a, ratio |
|||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Intercept | 5.991 | −0.372 | −1.969 | 10.361 | 4.679 | 2280.474 | ||||||
| Prediction value | let-7b | −2.70E-02 | let-7b | −1.529 | let-7b | −0.140 | let-7b/ | −1564.320 | let-7b | −0.003 | let-7b | −3.535 |
| miR-21a | ||||||||||||
| let-7f | 6.80E-04 | miR-21a | 0.057 | let-7f | 0.006 | let-7f/ | 70.384 | let-7b/ | −244.323 | miR-486/ | 130.008 | |
| miR-22 | miR-21a | miR-191 | ||||||||||
| miR-21a | 2.10E-04 | miR-22 | −0.018 | miR-21a | 0.004 | miR-26a/ | −0.052 | let-7b/ | −228195.714 | |||
| miR-10a | miR-21a | |||||||||||
| miR-22 | −5.70E-05 | miR-26a | −0.065 | miR-22 | −0.001 | miR-486/ | 1.175 | miR-27b/ | 3176.557 | |||
| miR-191 | miR-22 | |||||||||||
| miR-26a | −5.20E-03 | miR-146a | −2.391 | miR-26a | −0.028 | |||||||
| miR-146a | −0.051 | |||||||||||
| miR-181a | −0.011 | |||||||||||
| miR-486a | 0.027 | |||||||||||
| AUC | 0.992 | 0.984 | 0.989 | 0.998 | 0.951 | 1.000 | ||||||
| Contribution (R2) | 0.835 | 0.724 | 0.851 | 0.924 | 0.583 | 1.000 | ||||||
| AICc | 28.672 | 44.841 | 38.004 | 17.816 | 54.754 | 10.909 | ||||||
| BIC | 40.500 | 58.249 | 57.587 | 27.671 | 61.746 | 21.383 | ||||||
NR, normal renal function (n = 43). KD, kidney disease (n = 45). Stepwise regression analysis followed by binary logistic regression analysis. Objective variables were 0 (NR) and 1 (KD). AUC, area under the curve in receiver operating characteristic (ROC) curve. AICc, Akaike information criterion. BIC, Bayesian information criterion. * let-7b and let-7b/miR-21a is directly analyzed by binary logistic regression analysis.