Table 1.

The samples present in each of the clusters identified by the K-means analysis with K = 6. This was chosen because the silhouette analysis suggested that six clusters were the most appropriate (Figure 3). There were 33 human, 9 terrestrial animal, 10 mat community, 42 open water, 20 reef water, 60 coastal water, 5 deep water, 7 fresh water, 15 hypersaline, and 6 hot spring samples in total.

Cluster	Number of metagenomes	Original metagenome classification
1	52	31 Human
		5 Terrestrial animals
		6 Mat community
		Water samples:
		4 Open marine
		3 Coral reef
		2 Coastal marine
		1 Fresh
2	1	1 Coral reef water sample
3	1	1 Coral reef water sample
4	3	1 Human
		1 Fresh water
		1 Coral reef water
5	149	4 Mat
		4 Terrestrial animals
		1 Human
		Water samples:
		56 Coastal marine
		5 Deep marine
		15 Solar evaporation ponds
		6 Hydrothermal spring
		38 Open mainre
		13 Coral reef
		7 Freshwater
6	6	Water samples:
		2 Coastal marine
		3 Freshwater
		1 Coral reef