The average mutual information (AMI) has been claimed to be a strong genome signature in some literatures. The range of k values is an important parameter in AMI but no standard range of k value is yet proposed. We introduce a new growth threshold (GT) equation in Growing Self-Organising Maps (GSOM) to identify the best k range for clustering prokaryotic sequence fragments of 10 kb. However, the results using the best k range of AMI were still worse than our previously published results using oligonucleotide frequencies. These experiments showed that the newly proposed GT equation makes GSOM able to efficiently and effectively analyse different data features for the same data.