Single Nucleotide Polymorphism (SNP) Discovery in Rainbow Trout using Restriction Site Associated DNA (RAD) Sequencing of Doubled Haploids and Assessment of Polymorphism in a Population Survey

Background:  Our goal is to produce a high-throughput SNP genotyping platform for genomic analyses in rainbow trout that will enable fine mapping of QTL, whole genome association studies, genomic selection for improved aquaculture production traits, and genetic analyses of wild populations that aid better fisheries management.  Salmonid genomes are considered to be in a semi-tetraploid state as a result of an evolutionarily recent genome duplication event, therefore complicating application of traditional molecular genetic approaches.   As a result, DNA or RNA sequence assemblies often produce contigs that contain paralogous sequences.  This situation complicates single nucleotide polymorphism (SNP) discovery in rainbow trout as many putative SNPs are actually paralogous sequence variants (PSVs) and not simple allelic variants.  In an effort to reduce the number of PSVs we sampled nearly 100% homozygous doubled haploid (DH) lines that represent a wide geographic range of rainbow trout populations.  In addition, we surveyed 18 populations to characterize the degree of genetic diversity within and between multiple commercial populations, wild populations, and resource populations primarily used for research in the U.S.A. and in France.

Results:  We employed restriction-site associated DNA (RAD) technology to generate a large SNPs data set from deep sequencing of a panel of 11 DH lines.  The data set is composed of 145,168 high-quality putative SNPs that were genotyped in at least 9 of the 11 lines, of which 71,446 (49%) had minor allele frequencies (MAF) of at least 18% (i.e. at least 2 of the 11 lines).  Our populations’ survey revealed a wide range of shared RAD SNPs between every possible pair of populations ranging from 5% to 55% while the range of SNPs from any given population shared with at least one other population was between 42% and 89%.  On average, approximately 33% of the RAD SNPs identified in the populations’ survey were also found in the dataset from the 11 DH lines.

Conclusions:  A large dataset of high quality RAD SNPs that passed our rigorous analysis pipeline was generated from a panel of 11 DH rainbow trout lines as a resource for a SNP chip.  The RADs marker system was found to be very useful for assessing genetic diversity within a single rainbow trout population or between a pair of populations, but not for comparing more than two populations.  The generally low percentage of RAD SNPs shared between populations in a pairwise comparison suggests that for a SNP chip to be useful for a wide range of genetic analyses in rainbow trout it will be crucial to select the most common SNPs with relatively high MAF.