Figure 1

Mononucleotide repeats are less frequent in coding regions than in intergenic regions in the genome of M. tuberculosis . The figure shows the ratio between total number of observed and expected mononucleotide repeats in coding (filled circles and solid line) and intergenic regions (open circles and dashed line). The expected values have been calculated under the null model that nucleotides are randomly distributed in each of these two compartments, and occur at the frequencies observed in each compartment. For repeats of length three to six nucleotides, the under-representation of mononucleotide repeats is significantly stronger in coding regions than in intergenic regions (at p < 0.0001, chi square test).