Table 3.
Identification of gene expression signature by univariate and multivariate Cox regression analysis.
| NO. | Gene | Univariate Analysis * | Multivariate Analysis ** | ||||
|---|---|---|---|---|---|---|---|
| HR | 95%CI | p | HR | 95%CI | Coef. | ||
| 1 | UBE2C | 1.145 | 1.033–1.270 | 0.010 | --- | --- | --- |
| 2 | TPX2 | 1.226 | 1.089–1.381 | 0.001 | --- | --- | --- |
| 3 | PBK | 1.264 | 1.100–1.453 | 0.001 | --- | --- | --- |
| 4 | MELK | 1.233 | 1.069–1.422 | 0.004 | --- | --- | --- |
| 5 | TTK | 1.247 | 1.053–1.477 | 0.010 | 0.630 | 0.341–1.165 | −0.462 |
| 6 | KIF11 | 1.358 | 1.148–1.608 | <0.001 | --- | --- | --- |
| 7 | TOP2A | 1.178 | 1.043–1.331 | 0.008 | --- | --- | --- |
| 8 | HMMR | 1.472 | 1.243–1.742 | <0.001 | 1.883 | 1.153–3.074 | 0.633 |
| 9 | RRM2 | 1.298 | 1.128–1.493 | <0.001 | --- | --- | --- |
| 10 | ASPM | 1.409 | 1.169–1.698 | <0.001 | 0.577 | 0.287–1.159 | −0.550 |
| 11 | CENPF | 1.293 | 1.112–1.503 | 0.001 | --- | --- | --- |
| 12 | CDCA8 | 1.206 | 1.039–1.401 | 0.014 | 0.270 | 0.100–0.730 | −1.309 |
| 13 | KIF2C | 1.234 | 1.074–1.417 | 0.003 | 3.281 | 1.232–8.738 | 1.188 |
| 14 | AURKB | 1.188 | 1.039–1.358 | 0.012 | --- | --- | --- |
| 15 | CCNB2 | 1.258 | 1.085–1.458 | 0.002 | 0.622 | 0.329–1.178 | −0.474 |
| 16 | KIF20A | 1.350 | 1.136–1.605 | 0.001 | --- | --- | --- |
| 17 | MKI67 | 1.309 | 1.129–1.518 | <0.001 | 1.768 | 1.103–2.835 | 0.570 |
| 18 | CCNA2 | 1.328 | 1.150–1.533 | <0.001 | 1.622 | 0.889–2.959 | 0.484 |
| 19 | CCNB1 | 1.321 | 1.136–1.535 | <0.001 | --- | --- | --- |
| 20 | NUSAP1 | 1.293 | 1.107–1.511 | 0.001 | --- | --- | --- |
* The 20 DEGs were significantly associated with overall survival (p < 0.05) using univariate Cox regression analysis. ** Then, a least absolute shrinkage and selection operator (LASSO) regression on these 20 DEGs was performed to identify the most informative gene set for survival prediction. Finally, eight genes marked in gray in the table were selected to perform multivariate Cox regression analysis and to generate a prognostic risk model according to their respective regression coefficients. Abbreviations: HR, hazard ratio; CI, confidence interval; p, p-value; Coef., coefficient.