Table 2.

Analysis of mutations within regions of putative secondary structure

^aOne sequence from each subfamily was considered, excluding those sequences with insertions or deletions. In particular, elements with known insertions of SDR elements (e.g., SDR1.3) were excluded. In the case of SDR1, only those instances that were detected according to the number of mismatches were considered.

^bAll pairs of positions in the set of aligned sequences were considered, whenever one nucleotide of the pair was a G and the other a C or one was an A and the other a T. Double mutations were defined as those in which both nucleotides of a pair under consideration differed from the nucleotides at the same positions in the canonical sequence for the SDR family. Single mutations were defined as those in which only one of the two nucleotides differed from the canonical sequence.

^cPairs of nucleotides were considered structural if they corresponded to a pair identified in Figure 1 as possibly participating in base-pairing. Otherwise, the pair was considered nonstructural.

^dDouble mutations were considered compensatory if the resulting pair consisted of a G and a C or an A and a T. Pairs suffering single mutations were counted as either leading to a GT pair, an AC pair, or all other pairs. The actual count is the number of instances of the type of mutation under consideration that were observed in structural pairs. The expected count is the number calculated for structural pairs using the mutation-class frequencies in nonstructural pairs. A χ² test was used to assess significance. One and two asterisks indicate significance to the P < 0.05 and P < 0.01 levels, respectively. Significant differences are all increases (shown in bold) except where indicated.