Figure 3From: On the amyloid datasets used for training PAFIG how (not) to extend the experimental dataset of hexapeptidesHeatmap of merged consensus matrix for two clusters. Heatmap showing the similarity within of the two groups of methods: Waltz (denoted by Wxxx, where, for example, Waltz at pH = 2.6 and threshold = 79, was denoted by W2679, and Waltz at pH = 7 and threshold = 79 was denoted as W779), AmylPred (AP), and FoldAmyloid (Fxxxx). Pafig (P) was clustered as a quite different approach. Similarity was calculated using merged distance matrix over different clustering algorithms (agnes, pam, hclust, kmeans, and diana).Back to article page