Identification of I-UmaI homologs.

1NCBI TBLASTN 2.2.26+ search (July 2012) against the nr data base with the I-UmaI protein sequence as query. Shown are the top 10 hits.2B: Basidiomycota; G: Glomeromycota; C: Chytridiomycota; A: Ascomycota.3Length in amino acids. See Figure S4 for the ORF regions derived from the indicated accession numbers. In brackets: the number of premature stop codons (UAA, UAG) within the predicted ORFs. UGA has been converted to Trp according to the mitochondrial codon usage; for FN377860 and AF404306, each one nucleotide (positions 465 and 301, respectively) has been removed to provide for a continuous ORF; nd: not determined for FQ311469 due to 7 frameshift mutations within the predicted ORF region (data not shown).4According to Pfam; see alignment in Figure S4.5According to Rfam; see reference [18] for I-UmaI.6See Figure S3; for HQ540074, the minimum target sequence on the left of the intron insertion site is fully conserved, with positions 348–401 in HQ540074 corresponding to 3360–3413 in the U. maydis 521 mtDNA sequence. The right border could not be aligned due to the limited sequence available in HQ540074.