Fig. 3: Shared viral proteins between geographically distant hydrothermal vents.

A The bar plot shows the percent of shared protein clusters, and the bottom matrix shows the identity of the sites with shared protein clusters (filled, colored circles). The percent of shared protein clusters was calculated as the number of shared protein clusters divided by the smallest number of total protein clusters for a group, multiplied by 100. The leftmost bar plot shows the total number of protein clusters per site. The black line through the matrix separates deposit and diffuse flow (DF) samples from plume samples. Sites with fewer than fifteen shared protein clusters were removed. All clusters are reported in Supplementary Data 10. B A heatmap of the annotations of proteins shared between different hydrothermal vents or sample types. The annotations are the best hit of KEGG, VOG, and Pfam databases and were assigned functional categories using a custom Python script (see methods). The normalized counts were obtained by dividing the number of proteins per site by the total number of annotated, clustered proteins at that site. The dendrogram was produced by hierarchical clustering using the correlation distance “Spearman”.