How to plot the decision boundary when the feature vectors are letters

Question

I am doing classification where each sample is a sequence of letters A C T G(a DNA sequence). For example:

Sample 1: A C T G A Sample 2: C T G A A

How can I pl