Sorting duplicated loci disentangles complexities of polyploid genomes masked by genotyping by sequencing

Abstract

Many plants and animals of polyploid origin are currently enjoying a genomics explosion enabled by modern sequencing and genotyping technologies. However, routine filtering of duplicated loci in most studies using genotyping by sequencing introduces an unacceptable, but often overlooked, bias when detecting selection. Retained duplicates from ancient whole-genome duplications (WGDs) may be found throughout genomes, whereas retained duplicates from recent WGDs are concentrated at distal ends of some chromosome arms. Additionally, segmental duplicates can be found at distal ends or nearly anywhere in a genome. Evidence shows that these duplications facilitate adaptation through one of two pathways: neo-functionalization or increased gene expression. Filtering duplicates removes distal ends of some chromosomes, and distal ends are especially known to harbour adaptively important genes. Thus, filtering of duplicated loci impoverishes the interpretation of genomic data as signals from contiguous duplicated genes are ignored. We review existing strategies to genotype and map duplicated loci; we focus in detail on an overlooked strategy of using gynogenetic haploids (1N) as a part of new genotyping by sequencing studies. We provide guidelines on how to use this haploid strategy for studies on polyploid-origin vertebrates including how it can be used to screen duplicated loci in natural populations. We conclude by discussing areas of research that will benefit from better inclusion of polyploid loci; we particularly stress the sometimes overlooked fact that basing genomic studies on dense maps provides value added in the form of locating and annotating outlier loci or colocating outliers into islands of divergence.

Salmonid chromosome evolution as revealed by a novel method for comparing RADseq linkage maps

“…Comparisons of linkage group correspondence and synteny between species were investigated using available high-density linkage maps. This included maps generated with haploid crosses for mapping regions exhibiting residual tetraploidy (Limborg et al 2016), although in most cases, we only retained the nonduplicated loci due to problems in pairing these markers (described in workflow below). Some of these maps also contain centromere information, including Chinook Salmon (Brieuc et al 2014), Coho Salmon (Kodama et al 2014), Sockeye Salmon (Everett et al 2012; Limborg et al 2015), and Chum Salmon (Waples et al 2016).…”

Association mapping reveals candidate loci for resistance and anaemic response to an emerging temperature‐driven parasitic disease in a wild salmonid fish

“…Although the species differ in the chromosomes numbers (2n; S. trutta : 80, Atlantic salmon: 54–58), the fine‐scale LD patterns are expected to be congruent because of the high synteny between the two species (Leitwein et al., ). Finally, (iv), salmonid fishes have undergone a recent genome duplication ~60–80 million years ago (Ss4R) (Macqueen & Johnston, ) and at least 10% of the Atlantic salmon genome is still in a tetraploid state (Allendorf et al., ; Lien et al., ; Limborg, Seeb, & Seeb, ). This results in residual tetrasomic inheritance that retards divergence of duplicated loci at the distal ends of some homeologous chromosomes (Figure a; in Limborg et al., ).…”

“…Finally, (iv), salmonid fishes have undergone a recent genome duplication ~60–80 million years ago (Ss4R) (Macqueen & Johnston, ) and at least 10% of the Atlantic salmon genome is still in a tetraploid state (Allendorf et al., ; Lien et al., ; Limborg, Seeb, & Seeb, ). This results in residual tetrasomic inheritance that retards divergence of duplicated loci at the distal ends of some homeologous chromosomes (Figure a; in Limborg et al., ). Our analysis consisted of several steps to exclude duplicated regions/loci (maximization of uniquely mapped reads, exclusion of loci deviating from HWE, exclusion of dRAD contigs that mapped to same positions in Atlantic salmon genome) as our GWAS analysis pipeline assumes Mendelian segregation of diploid loci.…”

Genetic mapping in polyploids

“…Genotyping polyploids is in general more complex and less reliable than genotyping diploids (Mason, 2015). This could be caused by off-target amplification due to duplication of marker sequences at other loci (Limborg et al, 2016), but is also due in large part to the difficulty in correctly estimating the number of copies of a marker allele, also known as marker dosage assignment (Chapter 2). One of the standard practices when assigning dosage is to compare marker scores between parents and offspring.…”