Some remarks on the use of TaqI to detect highly repetitive DNA sequences in human chromosomes