Extended Data Fig. 1: Identifying inversion polymorphisms based on population genomic signatures.
From: Chromosomal inversion polymorphisms shape the genomic landscape of deer mice
For each identified inversion polymorphism, the following signatures of inversions are shown (colors correspond to focal population or population-pair in which inversion was identified, see legend): (1) Local PCA performed with lostruct, where each dot represents a 100-kb window. Distances between local PCA maps are represented by the MDS1 axis, with outlier windows highlighted in color. (2) Clustering of samples by PCA for entire outlier region found with local PCA, assigned using k-means clustering. (3) Heterozygosity (percent of sites that are heterozygous) of outlier region for samples by cluster assignments from PCA above. Boxes indicate upper and lower quartiles; center line represents median; whiskers extend to minimum and maximum values within 1.5x interquartile range; points show outliers beyond whiskers. Sample sizes are shown below the x-axis for each cluster. (4) LD for chromosomes harboring the example inversions, shown as mean r2 values for paired windows across each chromosome. Upper triangle shows mean r2 values including all samples from PCA clustering. Lower triangle shows mean r2 values for only the more common homozygote genotype as determined in PCA clustering. Colored bars highlight outlier region from local PCA. Scales for r2 values provided. (5) Recombination rates in cM/Mb shown for lab-born inversion heterozygotes. Outlier region found in local PCA is highlighted. Five inversions have missing data since inversion heterozygotes were not measured in the lab. (6) FST between homozygous genotypes (clusters 1 and 3 from PCA and heterozygosity plots). Outlier regions found in local PCA are highlighted. Note that the discontinuity for inv23.0 is likely due to reference genome mis-assembly.
