create artificial DNA alignment from protein alignment

[westram]

Motivation: enable basic "next-relative" features

Options:

• translation / codon start
  • specify manually
  • use existing (from species)
  • do not specify (=use all) ⇒ DNA seq will degenerate more
• degeneration modes:
  1. for each amino acid, mix all codons translating into the AA using IUPAC.
     (will result in "correct" DNA sequences, but usefulness for relative search will be low - due to IUPAC/'N's)
  2. always use same codon for each protein.
     (DNA sequences will look accurate (i.e. have no IUPACs) where they are fakes, but relative search will ml lead to better results)