A map of conserved motifs for each stem in each tRNA isotype. The summary is based on the sequence alignments of tRNAs of H. sapiens, M. musculus, and B. mori shown in Supplementary Figure S1. The four main stems of the tRNA secondary structure are abbreviated as follows: the amino acid (AA), the dihydrouridine (D), the anticodon (AC), and the thymine (T) stems. Motifs are shown in the 5′ to 3′ direction for the first strand encountered in the secondary structure, in monospaced font for ease of alignment across rows. All bases follow standard IUPAC conventions, but K is intended to signify ‘G or U’, rather than the standard ‘G or T’. The GGGG(R) motif, specific to only Ala- and Cys-tRNAs, is highlighted in bold. (M) denotes that the motifs are relatively constant across Mammalia. Only five tRNAs have four nucleotides in the variable loop: Gly, Glu, Gln, Asp, and His. Leu (with a preference for 5′YYY…GGG3′) and Ser (with a preference for 5′GGG…CCC3′) have a long variable loop with an additional helix (long-arm tRNAs) (see Supplementary Figure S2 where tRNAs with introns are also indicated, the insertion site is always between nucleotides 37 and 38 (1). The amino acids are organized by decreasing number of isoacceptor families with the color code indicating the strength of the codon-anticodon triplet (blue, high GC-content, red, high AU-content, and black in-between). Note that in the 2-codon boxes NNY only one tRNA is used for decoding (with G34 in the anticodon triplet) and in the 2-codon boxes NNR two tRNAs are used (with both C34 and U34 in the anticodon triplet)