Table 4.
Performance of regression algorithms applied to the concentration–standard deviation relationships. IBR, ibrutinib. DIB, dihydrodiol ibrutinib. IBR + DIB, sum of IBR and DIB concentrations (active moiety). NSSR, normalized sum of squared residuals. OLS, unweighted linear least squares. 2nd LS, unweighted 2nd-degree least squares. Siegel, Theil’s regression with the Siegel estimator. Theil, Theil’s regression.
| Algorithm | NSSR | Slope | Intercept | ||||||
|---|---|---|---|---|---|---|---|---|---|
| IBR | DIB | IBR + DIB | IBR | DIB | IBR + DIB | IBR | DIB | IBR + DIB | |
| Theil | 1.876 | 3.567 | 3.386 | 0.0479 | 0.0418 | 0.0387 | 0.06635 | 0.5308 | 0.4115 |
| Siegel | 2.352 | 2.516 | 4.682 | 0.0472 | 0.0438 | 0.0385 | 0.05559 | 0.6814 | 0.3526 |
| OLS | 106.9 | 4.428 | 1.986 | 0.0480 | 0.0342 | 0.0373 | −0.1285 | 1.970 | 0.6084 |
| 2nd LS | 1.667 | 2.615 | 1.934 | 0.0457 * | 0.0447 * | 0.0359 * | 0.08408 | 1.071 | 0.8606 |
* Linear coefficients are shown.