Fig. 1.

The presence–absence tree of NCLDVs and distribution of the 26 most shared protein clusters in NCLDVs. The tree is based on the presence–absence of the 3464 protein clusters. The protein clusters were made as follows. The tree was reconstructed from binary data with GTR2 model [95] with an ascertainment bias correction model [96] and 1000 ultrafast bootstraps [97] using IQ-tree [98]. Branches are marked with a black dot if the branch is supported at >95 %. The presence and number of protein cluster members are shown in a heatmap for the 26 most shared protein clusters that are present in more than six virus families or groups.