Skip to main content

Advertisement

Table 6 Performance of PLS and random forest models on the aryl-amine sets.

From: An investigation into pharmaceutically relevant mutagenicity data and the influence on Ames predictive potential

   Set A, Training Set A, Test Set B, Training Set B, Test
PLS with NitFormE PC1 0.76 ± 0.02 0.63 ± 0.08 0.80 ± 0.02 0.78 ± 0.05
PLS with 9 descriptors PC1 0.68 ± 0.03 0.66 ± 0.08 0.77 ± 0.02 0.76 ± 0.04
Random Forest With NitFormE    0.62 ± 0.01   0.855 ± 0.003
Random Forest With 9 descriptors   0.682 ± 0.008   0.844 ± 0.004
Nitrenium Formation Energy   0.72 ± 0.04 0.71 ± 0.09 0.78 ± 0.02 0.77 ± 0.05
  1. The performance of statistical models for Set A and Set B, over 100 random samples of the data. For the random forest model, the performance is for the data when it was out-of-bag in the construction of the trees and for PLS, it is for a 30% test sample when the model is trained with the other 70%.