Are there any computational tools to predict the binding affinity to specific DNA motifs from protein domain sequence information?
$\begingroup$
$\endgroup$
Those that come up from a Google search ("predict DNA binding from protein domain tool") seem pretty well suited to your question.
Depending on your computational/statistical know-how you might also find these papers relevant
-
1$\begingroup$ The OP is not asking to find which residues interact with DNA, but which DNA motifs are preferred by a given binding protein. $\endgroup$ – Erik Garrison Feb 18 '13 at 19:24