De novo assembly of the cattle reference genome with single-molecule sequencing.

Benjamin D Rosen ,Wilson Nandolo,Sergey Koren,Elizabeth Tseng,Christine Couldrey,Wenli Li,Jinna Hoffman,Arang Rhie,Christian Dreischer,Jay Ghurye,Aleksey V Zimin ,Darren E Hagen ,John B Cole ,Brenda M Murdoch ,Derek M Bickhart ,Timothy P L Smith ,Wai Yee Low ,Adam M Phillippy ,Christine G Elsik ,Richard J Hall ,Steven G Schroeder ,Françoise Thibaud‐Nissen ,Robert D Schnabel ,John A Hammond ,Stephanie Mckay ,Troy N Rowan ,T G Mcdaneld ,George E Liu ,Curtis P Van Tassell ,John C Schwartz ,Sebastian J Schultheiß ,W M Snelling ,Juan F Medrano

doi:10.1093/gigascience/giaa021

Abstract

BackgroundMajor advances in selection progress for cattle have been made following the introduction of genomic tools over the past 10–12 years. These tools depend upon the Bos taurus reference genome (UMD3.1.1), which was created using now-outdated technologies and is hindered by a variety of deficiencies and inaccuracies.ResultsWe present the new reference genome for cattle, ARS-UCD1.2, based on the same animal as the original to facilitate transfer and interpretation of results obtained from the earlier version, but applying a combination of modern technologies in a de novo assembly to increase continuity, accuracy, and completeness. The assembly includes 2.7 Gb and is >250× more continuous than the original assembly, with contig N50 >25 Mb and L50 of 32. We also greatly expanded supporting RNA-based data for annotation that identifies 30,396 total genes (21,039 protein coding). The new reference assembly is accessible in annotated form for public use.ConclusionsWe demonstrate that improved continuity of assembled sequence warrants the adoption of ARS-UCD1.2 as the new cattle reference genome and that increased assembly accuracy will benefit future research on this species.

Highlights

Major advances in selection progress for cattle have been made following the introduction of genomic tools over the past 10–12 years
We present the new reference genome for cattle, ARS-UCD1.2, based on the same animal as the original to facilitate transfer and interpretation of results obtained from the earlier version, but applying a combination of modern technologies in a de novo assembly to increase continuity, accuracy, and completeness
We demonstrate that improved continuity of assembled sequence warrants the adoption of ARS-UCD1.2 as the new cattle reference genome and that increased assembly accuracy will benefit future research on this species

Summary

Genome sequencing

The original Hereford assembly used blood as the source of DNA, leading to difficulties in assembling specific genomic regions. The Falcon assembly and Chicago library read pairs were used as input data for HiRise [10], a software pipeline for using Chicago data to scaffold genomes. Long reads were used to close gaps between contigs, resulting in 2,511 scaffolds, with an N50 of 63 Mb and L50 of 16. The IrysView v2.5.1 software package (BioNano Genomics, San Diego, CA) was used to map the assembly scaffolds to the optical map contigs. Pearson correlation coefficients between scaffold marker alignment order and genetic map marker order were used to calculate the most probable scaffold order and orientation. Another round of polishing was undertaken with Arrow with the SMRT Analysis 3.1.1 software package. The closing of gaps between contigs increased the contig N50 from 12 to 21 Mb and reduced the number of gaps in the genome to 459

Manual curation

RNA sequencing

Annotation

Findings

Conclusions