Figure 2
From: Clustering protein environments for function prediction: finding PROSITE motifs in 3D
Distribution of f i over all the FEATURE vectors used. The weight used in the F-distance, f i , is related to the information content of feature i, and weights the distance computed between two FEATURE vectors. The values of f i are calculated from the distribution of ones and zeros in feature i over the entire two million FEATURE vectors used. This heat map shows the values of f i over all 44 features in each of the 6 shells.