Results of the genome-scan, polymorphisms, observed counts and effects, results of single point analysis.

<p>(A) Results of genome-scan for QTL affecting genome-wide jungle usage (GJU) using a method that simultaneously extracts linkage and LD signal <a href="http://www.plosgenetics.org/article/info:doi/10.1371/journal.pgen.1002854#pgen.1002854-Druet1" target="_blank">[34]</a>. The red and blue horizontal lines mark the genome-wide significant and suggestive thresholds determined by permutation testing. The position of two pairs of PRDM9 parologues, respectively on BTA1 and BTAX, are highlighted. (B) Polymorphisms detected in the ZF array of the <i>PRDM9-XB</i> paralogue. Nine SNP are labeled ss1 to ss9, while the length polymorphism corresponding to the loss of two ZF domains is labeled VNTR. For each SNP, we define the position and the nature of the nucleotide substitution in the corresponding ZF domain (labeled in top row). For non-synonymous substitutions, we also define position and nature of the amino-acid substitution. Ss9 is synonymous and labeled as such (SYN). We represent the eight detected haplotypes sorted by VNTR genotype: haplotypes with 22 ZFs above, haplotypes with 20 ZFs below. For each length class, the top row represents the corresponding ZF domains labeled A, B, C, … I, and colored accordingly. ZFs differing by ‘or ‘’ (f.i. D, D′ and D″) have an amino-acid similarity of ≥92%. The triplet of amino-acids at DNA binding positions −1,3 and 6 ^f.i.41 are shown above each ZF. The frequency of the eight haplotypes in the HF population are given in the column on the right. (C) Observed counts (blue bars), and effects on genome-wide jungle usage (log₁₀ of GJU) (red circles) ± standard error (black vertical line) for the eight <i>PRDM9-XB</i> haplotypes defined in (B). (D) Result of single point analysis on GJU for BTAX. The ten <i>PRDM9-XB</i> variants are highlighted in red.