The data used in this paper are sequenced bacterial and archaeal genomes that were available as of September 2013. In total, 2670 prokaryotic genomes along with their annotation information were downloaded from GenBank (ftp://ftp.ncbi.nlm. nih.gov/genbank/genomes/Bacteria). We want to state that theoretically, using a much larger data set may have the possibility to introduce bias in some rare cases although this issue does not appear in this work The corresponding genomic length and GC content information of all these prokaryotic genomes are presented in Table S1.