Available methods to simulate nucleotide or amino acid data typically use Markov models to simulate each position independently. These approaches are not appropriate to assess the performance of combinatorial and probabilistic methods that look for coevolving positions in nucleotide or amino acid sequences. |