Number of errors from the 6-genes randomly selected signatures on the “AD” validation test set.

The Random forest algorithm was used as classifier, for each signature 10 runs with different seeds were done. We used the WEKA software implementation, and the algorithm was allowed to generate 150 trees. The best and worst signatures are highlighted in bold text. This result shows what it is expected, that a 6-signature, when the biomarkers are randomly chosen, is performing significantly worse than the panel of 18 biomarkers selected by Ray et. al. Now the best result (81.5%) is worse than the average result of a random 18-signature (86%).