DNA binding specificities with the two component model. A: The 2D table of interactions for amino acid position 3 with base position 8 obtained from the data in Figure 3A for all cases except R with G or A (and normalized to a sum of 1000). B: The 2D table of interactions for amino acid position 3 with base position 8 for the cases with R and G or A (normalized to 1000). C: The predicted binding probabilities for the entire dataset using the two component model. The elements for the cases of R with G or A are obtained by the outer product of the matrix from B with the R/G,A elements of the matrix in Figure 3B. The rest of the elements are obtained from the outer product of A with the remaining elements of the matrix from Figure 3B.