An Idea: Spam Detection using DNA

I have an idea of Spam detection using DNA (Hybridization or Sequence Alignment). Like represent each English alphabet by a codon. Lets say we represent ‘a’ by AUG and ‘@’ by AUT. Now we can convert the incoming message into a DNA sequence and search for patterns such as ‘V-i-a-g-r-a’ or ‘C@sin0’ via standard principles of sequence alignment or Hybridization.

There are 3 factors:
1. No of words recognized
2. If hybridization is used, Temperature at which words are recognized
3. How good is a match

Then using the combination of these 3 factors, we can determine if a message should be classified as spam or not.

Any comments, suggestions or questions?