Myotubularin Protein Domain Architecture. Candidate myotubularin homologue sequences were obtained by searching the protein datasets of fully sequenced eukaryotic genomes, as detailed in Methods. PH-GRAM, myotubularin phosphatase, and other structural domains were identified as detailed in Methods. To conserve space, organisms are identified in the Figure as genus names only. Organisms are grouped in this figure by taxonomic category as in Figure 1. The domain architecture of human MTMR2 is given for orientation purposes. The organisms are as follows: Homo sapiens, Giardia lamblia, Leishmania major, Trypanosoma brucei, Dictyostelium discoideum, Entamoeba histolytica, Monosiga brevicollis, Trichoplax adhaerens, Nematostella vectensis, Aspergillus niger, Fusarium graminearum, Saccharomyces cerevisiae, Schizosaccharomyces pombe, Laccaria bicolor, Ustilago maydis, Cryptosporidium hominis, Paramecium tetraurelia, Tetrahymena thermophila, Thalassiosira pseudonana, Arabidopsis thaliana, Oryza sativa, Populus trichocarpa, Sorghum bicolor, Vitis vinifera, Physcomitrella patens. The complete set of sequences illustrated in the Figure, along with database accession numbers, is available as Additional File 5. As detailed in Methods, some incomplete candidate myotubularin sequences with annotation mistakes were corrected with additional sequence, and are denoted with the suffix "C" in the Figure. Sequence "Mbrevi5R3" was assembled manually from individual genomic sequence reads unincorporated into scaffolds, through use of TBLASTN against Monosiga genomic DNA with Nematostella myotubularin homologue query sequences.