Character-state weighting for cladistic analysis of protein-coding DNA sequences