Model performance by learning technique and feature mapping method for correlations and mean squared error on the entire dataset and by 10-fold cross validation.

Learning Techniques: Artificial Neural Network (ANN), General Linear Model (GLM), Support Vector Machine (SVM).
Mapping methods: Position Specific Base Composition (PSBC), Thermodynamic (THER), N-Grams of length 2 though 5 (NG25), Guide Strand Structure Features (GSSF), Guide Strand Secondary Structure (GSSS), Positions specific base compositions plus N-Grams of length 1 through 3 (P+13), Positions specific base compositions plus N-Grams of length 2 through 5 (P+25) the combination of each of the methods PSBC, THER, NG25, GSSF and GSSS (ALL).
R  =  Pearson correlation coefficient, of model predicted activities to observed activities.
MSE  =  Mean Squared Error of model predicted activities to observed activities.
column maxima for R and minima for MSE are in bold for 10-fold cross validations, same values bolded in Table 7.