We are constructing a new amino acid alphabet, rather then yet another reduced alphabet, we will make an extended alphabet. For the evaluation we would like to perform MSAs with benchmark data but we need an MSA program that can handle a new matrix, a matrix of for instance 40 by 40 instead of 20 by 20. In addition, our scheme does indeed result in a nested alphabet. It will have all 20 AAs but for instance D will be element of 1 ={DE} and 2={DENQ} (hence 1 is element of 2 as well.

Any idea what we could use/easily adjust for this purpose?

Thanks a bunch!