Table 6. The evaluation of QSAR models which were constructed by a partial least squares (PLS) algorithm.
Cross validation | Non-cross validation | Fraction | ||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|
CoMFA* | CoMSIA | ONC | q2 | r2 | SEE | F | S | E | H | D | A | |
ONC | 6 | S | 6 | 0.596 | 0.702 | 0.559 | 19.63 | 1.000 | - | - | - | |
q2cv | 0.682 | H | 6 | 0.544 | 0.784 | 0.475 | 30.31 | - | - | 1.000 | - | |
r2 | 0.796 | D | 6 | 0.454 | 0.523 | 0.707 | 9.13 | - | - | - | 1.000 | |
SEE | 0.462 | A | 6 | 0.531 | 0.756 | 0.506 | 25.78 | - | - | - | - | 1.000 |
F | 32.57 | SH | 6 | 0.525 | 0.807 | 0.450 | 34.85 | 0.376 | - | 0.624 | - | - |
SD | 6 | 0.503 | 0.701 | 0.560 | 19.50 | 0.627 | - | - | 0.373 | - | ||
SA | 6 | 0.578 | 0.785 | 0.475 | 30.41 | 0.436 | - | - | - | 0.564 | ||
HD | 6 | 0.512 | 0.769 | 0.492 | 27.70 | - | - | 0.666 | 0.334 | - | ||
HA | 6 | 0.562 | 0.802 | 0.455 | 33.78 | - | - | 0.547 | - | 0.453 | ||
DA | 6 | 0.499 | 0.808 | 0.449 | 35.00 | - | - | - | 0.312 | 0.688 | ||
SHD | 6 | 0.638 | 0.776 | 0.484 | 28.90 | 0.209 | - | 0.514 | 0.276 | - | ||
SHA* | 6 | 0.676 | 0.812 | 0.444 | 35.99 | 0.235 | - | 0.434 | - | 0.331 | ||
SDA | 6 | 0.623 | 0.765 | 0.496 | 27.15 | 0.324 | - | - | 0.266 | 0.410 | ||
HDA | 6 | 0.624 | 0.773 | 0.487 | 28.43 | - | - | 0.445 | 0.252 | 0.302 | ||
SEHD | 6 | 0.652 | 0.776 | 0.484 | 28.90 | 0.209 | 0.000 | 0.514 | 0.276 | - | ||
SEHA | 6 | 0.673 | 0.812 | 0.444 | 35.99 | 0.235 | 0.000 | 0.434 | - | 0.331 | ||
SEDA | 6 | 0.631 | 0.765 | 0.496 | 27.15 | 0.324 | 0.000 | - | 0.266 | 0.410 | ||
SHDA | 6 | 0.625 | 0.777 | 0.483 | 29.05 | 0.163 | - | 0.328 | 0.231 | 0.279 | ||
EHDA | 6 | 0.627 | 0.773 | 0.773 | 28.43 | - | 0.000 | 0.445 | 0.252 | 0.302 | ||
SEHDA | 6 | 0.634 | 0.777 | 0.483 | 29.05 | 0.163 | 0.000 | 0.328 | 0.231 | 0.279 |
The above abbreviations represent:
A: Acceptor. D: Donor. E: Electrostatic. F: F-test value. H: Hydrophobic. ONC: Optimal number of components. PLS: partial least squares. S: Steric. SEE: Standard error of estimate. *: Optimum prediction model.