Revealing misassembled segments in the bovine reference genome by high resolution linkage disequilibrium scan

Utsunomiya, Adam T.H. [UNESP]; Santos, Daniel J.A. [UNESP]; Boison, Solomon A.; Utsunomiya, Yuri T. [UNESP]; Milanesi, Marco [UNESP]; Bickhart, Derek M.; Ajmone-Marsan, Paolo; Sï¿½lkner, Johann; Garcia, Josï¿½ F. [UNESP]; da Fonseca, Ricardo [UNESP]; da Silva, Marcos V.G.B.

Revealing misassembled segments in the bovine reference genome by high resolution linkage disequilibrium scan

Arquivos

2-s2.0-84984993613.pdf (2.05 MB)

Data

2016-09-05

Autores

Utsunomiya, Adam T.H.

Tipo

Artigo

Direito de acesso

Acesso aberto

Resumo

Background: Misassembly signatures, created by shuffling the order of sequences while assembling a genome, can be detected by the unexpected behavior of marker linkage disequilibrium (LD) decay. We developed a heuristic process to identify misassembly signatures, applied it to the bovine reference genome assembly (UMDv3.1) and presented the consequences of misassemblies in two case studies. Results: We identified 2,906 single nucleotide polymorphism (SNP) markers presenting unexpected LD decay behavior in 626 putative misassembled contigs, which comprised less than 1 % of the whole genome. Although this represents a small fraction of the reference sequence, these poorly assembled segments can lead to severe implications to local genome context. For instance, we showed that one of the misassembled regions mapped to the POLL locus, which affected the annotation of positional candidate genes in a GWAS case study for polledness in Nellore (Bos indicus beef cattle). Additionally, we found that poorly performing markers in imputation mapped to putative misassembled regions, and that correction of marker positions based on LD was capable to recover imputation accuracy. Conclusions: This heuristic approach can be useful to cross validate reference assemblies and to filter out markers located at low confidence genomic regions before conducting downstream analyses.