for n > 1 motifs there should be (n * (n - 1)) / 2 correlations, obviously there are none for only 1 motif