Structural variants and tandem repeats in the founder individuals of four F2 pig crosses and implications to F2 GWAS results

Abstract

Background Structural variants and tandem repeats are relevant sources of genomic variation that are not routinely analyzed in genome wide association studies mainly due to challenging identification and genotyping. Here, we profiled these variants via state-of-the-art strategies in the founder animals of four F2 pig crosses using whole-genome sequence data (20x coverage). The variants were compared at a founder level with the commonly screened SNPs and small indels. At the F2 level, we carried out an association study using imputed structural variants and tandem repeats with four growth and carcass traits followed by a comparison with a previously conducted SNPs and small indels based association study. Results A total of 13,201 high confidence structural variants and 103,730 polymorphic tandem repeats (with a repeat length of 2-20 bp) were profiled in the founders. We observed a moderate to high (r from 0.48 to 0.57) level of co-localization between SNPs or small indels and structural variants or tandem repeats. In the association step 56.56% of the significant variants were not in high LD with significantly associated SNPs and small indels identified for the same traits in the earlier study and thus presumably not tagged in case of a standard association study. For the four growth and carcass traits investigated, many of the already proposed candidate genes in our previous studies were confirmed and additional ones were identified. Interestingly, a common pattern on how structural variants or tandem repeats regulate the phenotypic traits emerged. Many of the significant variants were embedded or nearby long non-coding RNAs drawing attention to their functional importance. Through which specific mechanisms the identified long non-coding RNAs and their associated structural variants or tandem repeats contribute to quantitative trait variation will need further investigation. Conclusions The current study provides insights into the characteristics of structural variants and tandem repeats and their role in association studies. A systematic incorporation of these variants into genome wide association studies is advised. While not of immediate interest for genomic prediction purposes, this will be particularly beneficial for elucidating biological mechanisms driving the complex trait variation.

Structural variation and eQTL analysis in two experimental populations of chickens divergently selected for feather pecking behavior

“…SVs were called as described by Blaj et al . [21] with slightly different settings: A high confidence SV call set was produced from the output of three variant callers: smove v0.2.6 (Brent, P. (2018). Smoove.…”

“…RefSeq assembly) and used to call SNPs and short (< 50bp) insertions and deletions (InDels) in our previous study [2]. SVs were called as described by Blaj et al [21] with slightly different settings: A high confidence SV call set was produced from the output of three variant callers: smove v0.2.6 (Brent, P. (2018). Smoove.…”

Research progress and applications of genome‐wide association study in farm animals

“…Over a decade of improvements in sequencing technology, whole-genome sequencing and modified reduced genome sequencing (e.g., RAD, RRGS, and GBS) have occurred and SNPs are the most common markers in recent studies [89][90][91]. Currently, multiple genomic variant types, including copy number variations (CNVs) [92][93][94], insertion-deletion (InDel) [84,95], structure variations (SV) [96][97][98], haplotype [99], presence/absence variations (PAVs) based on pan-genome [100], and even epigenomic markers [101] have gained increasing attention through GWAS. Multiple variant profiles have been generated for multiple breeds [102,103].…”

“…In EWAS, association analysis links DNA methylation to phenotypes. Though it has been PRIM2, COL21A1, HMGCLL1, etc [98] a…”

Structural variation and eQTL analysis in two experimental populations of chickens divergently selected for feather-pecking behavior

“…This was achieved with the genome analysis toolkit (GATK) v. 4.0, according to the best practice guideline of the broad institute [25]. SVs were called as described by Blaj et al [26] with slightly different settings: A high-confidence SV call set was produced from the output of three variant callers: smove v. 0.2.6 (Brent, P. (2018) Smoove. https:// brentp.…”