Fig. 1

Selection of peptide parameters for CUPP clustering. The performance of CUPP clustering and prediction for GH30 using various peptide lengths and number of ambiguous amino acids within them (written as the length and the number of ambiguous amino acids separated by “x”). Each dot represents the average of CUPP group precision/family sensitivity for a c_clust ranging from 3 to 14 with corresponding error bars indicating the standard deviation. The gray areas indicate settings which are relatively less favorable. Peptides with a length of eight with two ambiguous amino acids were selected as the optimal for GH30 and applied for all CAZy families