Optical mapping compendium of structural variants across global cattle breeds

Abstract

36 Structural variants (SV) have been linked to important bovine disease phenotypes, but due to 37 the difficulty of their accurate detection with standard sequencing approaches, their role in 38 shaping important traits across cattle breeds is largely unexplored. Optical mapping is an 39 alternative approach for mapping SVs that has been shown to have higher sensitivity than 40 DNA sequencing approaches. The aim of this project was to use optical mapping to develop a 41 high-quality database of structural variation across cattle breeds from different geographical 42 regions, to enable further study of SVs in cattle. 43 To do this we generated 100X Bionano optical mapping data for 18 cattle of nine different 44 ancestries, three continents and both cattle sub-species. In total we identified 13,457 SVs, of 45 which 1,200 putatively overlap coding regions. This resource provides a high-quality set of 46 optical mapping-based SV calls that can be used across studies, from validating DNA 47 sequencing-based SV calls to prioritising candidate functional variants in genetic association 48 studies and expanding our understanding of the role of SVs in cattle evolution. 49 bioRxiv preprint doi: https://doi.org/10.1101/2022.05.05.490773; this version posted May 16, 2022. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under a CC-BY 4.0 International license. 50 Background & Summary 51 Structural variants (SV) are a heterogeneous class of genetic variants involving large 52 fragments of the genome (>50bp)1. These variants include genomic insertions and deletions 53 (InDels), inversions, duplications, translocations and more complex rearrangements 2. Single 54 nucleotide polymorphisms (SNPs) have been the primary focus of studies trying to map 55 genetic loci underlying important cattle phenotypes. However, there are multiple lines of 56 evidence suggesting SVs likely underlie many important cattle traits. As many as 25-29% of all 57 protein truncating events are thought to be caused by an SV in humans1 and notably, despite 58 being less well studied, SVs have already been tied to key livestock phenotypes. For example, 59 a duplication of the CIITA class II major histocompatibility complex transactivator gene in 60 cattle has been tied to resistance to intestinal nematodes3 and a 12Kb copy number variant 61 has been linked to mastitis in cattle4. Chromosomal translocations and duplications have been 62 linked to skin pigmentation, a phenotype closely tied to environmental adaptation, and SVs 63 across livestock species have been linked to phenotypes such as olfaction or resistance to 64 adenocarcinoma-causing viruses5. Importantly SVs are responsible for approximately 5-10 65 times as many heritable nucleotide sequence differences between individuals than SNPs 6. 66 Unlike SNPs, that only effect a single basepair, and most often far from coding regions, SVs 67 effect large regions and potentially multiple genes. Consequently, although smaller in 68 number, any given novel SV event is more likely to have a phenotypic consequence. 69 The two most popular methods used to detect SVs are high-throughput sequencing (HTS) and 70 array comparative genomic hybridisation (aCGH), both of which have been applied to 71 European cattle7–9, but with few studies performed in other cattle breeds10–12. Each 72 technology has advantages and limitations. aCGH, for example, involves measuring binding 73 to probes covering the reference genome, and therefore it can only detect relative copy 74 number changes between sample pairs and cannot for example detect novel insertions. 75 Resolution is also limited. A major advantage of HTS approaches is that theoretically they can 76 detect SVs at base-pair resolution. However, accurate calling of SVs from HTS data has proven 77 to be difficult for a number of reasons including poor reference assemblies, chimeric reads, 78 aligners penalising reads that don’t match the reference and the difficulties of sequencing 79 and mapping to repetitive regions. This is exemplified by the generally poor agreement 80 between SV callers even when run across the same samples13,14. Approaches using long reads bioRxiv preprint doi: https://doi.org/10.1101/2022.05.05.490773; this version posted May 16, 2022. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under a CC-BY 4.0 International license. 81 and de novo assembly can still have true positive rates as low as 77%, even when using 82 simulated data15. 83 84 Optical mapping (OM), a light microscope-based method that labels and physically locates 85 specific motifs in the genome16, offers an alternative protocol to accurately detect large SVs. 86 OM molecules can be consistently hundreds of Kb long, allowing for the detection of complex 87 rearrangements undetectable using HTS. Despite the limitation of not being able to detect 88 the actual sequence of the identified SVs, as well as missing smaller SVs, OM has a very high 89 sensitivity and specificity, allowing for the generation of high-quality catalogues of SVs in 90 individuals17. A study in humans successfully used OM reads to identify SVs in a total of 26 91 genomes revealing population-specific patterns of structural variation18. 92 In this study, we generated the first catalogue of cattle OM data for 18 animals from 9 93 different global breeds, and three continents, to better characterise common SVs across the 94 cattle pan-genome. This data is a particularly valuable resource of SVs for the cattle species 95 to intersect with other datasets, for example, for the validation of SV calls from other 96 approaches. 97