Fig. 1

Phylogenetic tree and domain architectures of the SAMD9 family proteins. Domain identities and positions are based on multiple searches with the representative sequences, for which the architectures are shown, as queries to search the CDD, HHpred, Pfam, and SMART databases (see Additional file 1: Table S1) and multiple alignments with human SAMD9 as well as with the sequences within the collapsed branches. The multiple alignment used to build the tree was trimmed down to the conserved core, from the SIR2-like domain through the OB-fold domain (the first upstream copy in the sequences with two OB-fold domains). GenBank identifiers and species names are shown for all domain architectures. Amino acid sequences of the SAMD9 family members were collected from Non-redundant protein sequences database (nr) at the NCBI using BLASTP. Distant homologs containing only a single domain with significant similarity tp SAMD9 proteins were discarded. Sequences were clustered into putative orthologous sets, and highly similar sequences were purged using BLASTClust with –L 0.9 –S 0.9 parameters (https://www.ncbi.nlm.nih.gov/Web/Newsltr/Spring04/blastlab.html). Domain architectures were analyzed by batch CDD search with Expect Value threshold 1.0 at https://www.ncbi.nlm.nih.gov/Structure/bwrpsb/bwrpsb.cgi [33] followed by manual inspection and HHpred [20]. Multiple sequence alignments were constructed using MUSCLE with default parameters [34], and the phylogenetic tree was constructed using FastTree with the WAG evolutionary model and discrete gamma model with 20 rate categories [35]. The numbers at internal branches indicate bootstrap support (percent)