Figure 1.
Receiver operating characteristic (ROC curves) obtained for the 4 selected predictors, PAFIG, SALSA, Fold Amyloid and Waltz, and Leave-One-Out cross validated MetAmyl on the training dataset.
Table 1.
Area Under the Curve (AUC) based on the training dataset.
Table 2.
Prediction performances, based on the training dataset are given for the 10 compared predictors.
Table 3.
Evaluation of the performance of the tool MetAmyl on a subset of 33 proteins of the amylome.
Table 4.
Evaluation of the performance of the tool MetAmyl on scrambled sequences from the 17- amino acid N-terminal segment of the Huntingtin protein.
Figure 2.
Metamyl predictions applied to human fibrinogen-.
The effect of mutations is reported on a diagram where each column represents the difference of scores between the mutant and the corresponding wild-type sequence. The analysis is limited to mutations affecting the fragment of 80 amino acids found in amyloid fibrils, which is the region 500–580 of the mature protein. In red are variants involved in renal amyloidosis. In blue are non-pathological variants.