Figure 1

Due to the high collinearity of the local abundance of the different di-, tri-, and tetra-nucleotide motifs, two principal components were extracted for each motif length that explained most of the variation in local recombination rates (see text). For the analysis of di-nucleotides, shown here, both principal components differed significantly (p < 0.05) among genome intervals that exhibited low (first quartile), medium (second and third quartile), and high (forth quartile) average local recombination rates. Individual data are shown here as small circles and the 95% confidence interval or the group means as the larger solid ellipses. The principal component analysis of tri- and tetra-nucleotides revealed almost identical patterns.