Unsupervised Two-Way Clustering of Metagenomic Sequences

Figure 1

Distribution of dimers and pentamers across 50,000 reads sampled from the genome of Haemophilus influenzae (only a few distributions are shown). (a) Distribution of dimers tends to Gaussian, two groups can be observed. (b) Distribution of pentamers tends to Poisson, three groups are seen.

(a) Distribution of dimers across reads

(b) Distribution of pentamers across reads