Figure 1

Proteomic Training and Test (AD and MCI). The Training Set is composed by 43 AD samples and 40 NDC samples. The Test Set AD is composed by 42 AD samples and 50 non-AD samples (9 NDC and 11 OD). The Test Set MCI is composed by 47 samples diagnosed with a mild cognitive impairment and with known final follow-up (2-6 years). The Training Set is fed as input to the l ₁ l _2FS framework (grey box), which splits the data in K subsplits (light blue boxes), evaluating the relevant variables and the classification error for each one. Performance of the classifier is then tested on the Test Set AD and Test Set MCI using only the 21 proteins selected in the training phase. The test error is decomposed in True Positives (TP), False Negatives (FN), False Positives (FP), True Negatives (TN).