Fig. 8From: Weighted likelihood inference of genomic autozygosity patterns in dense genotype dataDistribution of differential ROA signals between subgroups in the ITU and PJL. Manhattan plots showing for each window the log 10 (P) of pairwise comparisons of per-individual wLOD scores in the two subgroups present in the (a) ITU (580 SNV window) and (b) PJL (610 SNV window). In each plot, P-values for the comparison testing whether wLOD scores in cluster A are greater than those in cluster B (Additional file 1: Figure S25) are shown on top with P-values for the reverse comparison shown below. P-values represent the proportion of genome-wide maximum Wilcoxon rank-sum test statistics observed in 1000 permutations of group labels that exceed the Wilcoxon rank-sum test statistic obtained with the true labels [152]. Windows with P > 0.05 are shown in black and those with P ≤ 0.05 are shown in orange. The horizontal grey dashed line denotes P = 0.05 and while the grey dotted line denotes P = 0.01Back to article page