Figure 1.
IFS curves for hydroxyproline dataset and hydroxylysine dataset.
Each curve shows that prediction accuracies of the 500 predictors evaluated by the jackknife cross-validation.
Figure 2.
Distribution of the three kinds of features and distribution of 13 positions of the peptides in feature set A and B for hydroxyproline dataset.
Legend “Distributive” means that the frequency of each kind of features are calculated according to the proportion of each kind of features in the 6,345 features (e.g., in dataset A, there should be 478 () amino acid factors, 21 (
) conservation, and 1 (
) disorder); while legend “Resultant” represents the frequency of each kind of features in the dataset (A or B).
Figure 3.
Distribution of the 5 feature clusters of the AAindex and distribution of conservation of 20 amino acids in the feature set A and B for hydroxyproline dataset.
Figure 4.
Distribution of the three kinds of features and distribution of 13 positions of the peptides in feature set A and B for hydroxylysine dataset.
Legend “Distributive” means that the frequency of each kind of features are calculated according to the proportion of each kind of features in the 6,345 features (e.g., in dataset A, there should be 478 () amino acid factors, 21 (
) conservation, and 1 (
) disorder); while legend “Resultant” represents the frequency of each kind of features in the dataset (A or B).
Figure 5.
Distribution of the 5 feature clusters of the AAindex and distribution of conservation of 20 amino acids in the feature set A and B for hydroxylysine dataset.