Table 2.
The results of multiple linear regression (MLR) analysis for the compound in series 1-7 with different type of descriptors
| Equation | Cell line | QSAR Model | n | R2 a | SE b | RMScv c | Q2 d | F e |
|---|---|---|---|---|---|---|---|---|
| 1 | L1210 | Y= 9.037 (±1.56) Molecular density +3.693 (±1.435) HNar -0.007 (±0.003) Heat of formation - 9.712(±3.910) | 19 | 0.70 | 0.38 | 0.44 | 0.50 | 11.25 |
| 2 | CEM | Y=7.352 (±0.913) Molecular density + 0.062 (±0.009) Hydration energy +89.840 (±34.561) X4A - 12.102(±4.537) | 19 | 0.85 | 0.25 | 0.29 | 0.77 | 29.11 |
| 3 | Molt/C8 | Y=7.507 (±1.155) Molecular density -0.493 (±0.084) Homo + 0.09 (±0.031) Dipole X + 6.039(±1.647) | 19 | 0.81 | 0.31 | 0.34 | 0.72 | 21.31 |
a)
Squared correlation coefficient;
b)
Standard error of regression;
c)
Root mean square of cross validation;
d)
Leave-one-out cross-validation correlation coefficient (R2CV);
e)
F-statistics.Squared correlation coefficient (R2) was calculated using following equation: 1-(Σ (y experimental – y calculated by model)2/Σ (y experimental – y average. experimental)2)) Equation for calculation of Standard error of regression: [SE2= Σ ((y experimental – y calculated)2/n - p') (n= number of samples, p'= number of model parameters)]