Skip to main content
. 2010 Dec 3;11(12):4991–5008. doi: 10.3390/ijms11124991

Figure 1.

Figure 1.

Prediction accuracies for the clustered and isolated types of mucin-type O-glycosylation in various sequences varying in length (window size, Ws) from three to 55. Amino acid sequence or composition information was used as the input to SVM. The crosses and circles indicate the prediction accuracies obtained by using the sequence information and composition information, respectively. The clustered and isolated types are shown in red and blue, respectively.