Figure S1. Flanking Sequence Nucleotide Frequency Distributions Illustrating Three Classes of Patterning with Neither Strong Periodicity Nor Asymmetry
From little structure of any kind (A) to complicated aperiodic clustering (C). Plots are as described in Figure 3.
(1.7 MB TIF).
Figure S2. Flanking Sequence Frequency Distributions for Four Dinucleotide Motif–Cassette Combinations Further to Those Shown in Figure 4
Plots are as described in Figure 4.
(3.0 MB TIF).
Figure S3. Dinucleotide Flanking Sequence Patterning in Interspersed Repeats and Unique Sequence DNA
Figure depicts equivalent patterns of asymmetry in AT dinucleotide frequencies for the commonest cassette type, (T/A), around microsatellites in unique sequence DNA (A), LINE/L1 elements (B), and SINE/Alu elements (C). Plotting conventions are the same as for Figure 4.
(1.4 MB TIF).
Figure S4. Dependence of Differences among Homologous Loci on Location of Microsatellite
Block position is relative to the central microsatellite (not shown).
(A and B) Proportion of exact matches (with 95% binomial confidence intervals) and average number of matches, excluding exact matches (± standard error), with block position around (AC)2–3 microsatellites (n = 4,593).
(C and D) As (A and B) but for (AC)4+ microsatellites (n = 356). Average proportion of exact matches and number of matches, calculated separately for 5′ and 3′ blocks around (AC)2–3 microsatellites, are shown by a black line in (A) and (C), and (B) and (D), respectively. Average percentage match rather than average match is plotted in (B) and (D) because overlapping blocks were truncated to exclude overlapping regions from the analysis, with the result that not all blocks contained 20 bases.
(2.9 MB TIF).
Figure S5. Mean Proportion of Mismatches along Homologous Flanking Sequences
The proportion of mismatches occurring at a given base in a flanking sequence are averaged over (AC)2–3 microsatellite loci (A) and over (AC)4+ microsatellite loci (B). Shown ± standard error. The microsatellite at base position 0 is not shown. Expectation, calculated separately for 5′ and 3′ sequences around (AC)2–3 microsatellites, is shown by a black line in both plots.
(3.4 MB TIF).
Work was funded under a Natural Environment Research Council studentship. We are grateful for help with computing facilities from David Judge (Department of Genetics, Cambridge University).
Conflicts of interest. The authors have declared that no conflicts of interest exist.
Author contributions. EJV and WA conceived and designed the experiments. EJV performed the experiments and analyzed the data. EJV and WA wrote the paper.