Table 2.
Dose prediction of candidate protein markers.
Modela | Predicted dose (Mean ± SD, Gy) | coefficient b | R2 (p value)b |
MAEc | ||
---|---|---|---|---|---|---|
Actual irradiated dose | ||||||
0 Gy | 1 Gy | 2 Gy | ||||
Dose ~ FDXR + ACTN1 | 0.003 ± 0.050 | 1.06 ± 0.17 | 1.90 ± 0.24 | 0.95 ± 0.06 | 0.958 (<0.0001) | 0.13 |
Dose ~ FDXR + ACTN1 + DDB2 | 0.001 ± 0.034 | 1.06 ± 0.14 | 1.91 ± 0.26 | 0.96 ± 0.06 | 0.959 (<0.0001) | 0.12 |
Dose ~ FDXR + ACTN1 + BAX | 0.001 ± 0.046 | 1.07 ± 0.16 | 1.90 ± 0.23 | 0.95 ± 0.06 | 0.958 (<0.0001) | 0.13 |
Dose ~ FDXR + ACTN1 + DDB2 + BAX | 0.13 ± 0.20 | 1.05 ± 0.14 | 1.91 ± 0.26 | 0.90 ± 0.06 | 0.942 (<0.0001) | 0.15 |
aData obtained from proteomic analysis were used for investigating dose prediction. Ranks and specific values of all models were provided in Supplementary Table S4.
bCoefficient, R2 and p value was obtained by linear regression analysis for predicted vs. actual doses.
cMAE values were used as indicators to compare the difference between actual irradiated and predicted dose.