Genotyping common, large structural variations in 5,202 genomes using pangenomes, the Giraffe mapper, and the vg toolkit

Sirén, Jouni; Monlong, Jean; Chang, Xian; Novak, Adam M.; Eizenga, Jordan M.; Markello, Charles; Sibbesen, Jonas Andreas; Hickey, Glenn; Chang, Pi-Chuan; Carroll, Andrew; Gupta, Namrata; Gabriel, Stacey; Blackwell, Thomas W.; Taylor, Kent D.; Rich, Stephen S.; Rotter, Jerome I.; Haussler, David; Garrison, Erik; Paten, Benedict

doi:10.1101/2020.12.04.412486

Cited by 32 publications

(40 citation statements)

References 61 publications

(125 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Our workflow provides tools to determine the origin of non-reference bases, derive structural variations from multi-assembly graphs, predict putatively novel genes and append the novel sequences linearly to a reference genome. We anticipate that the latter will become obsolete as soon as accurate and fast base-level alignment and split-read graph mapping enables the full-suite of genome analyses from a reference graph 48 .…”

Section: Discussionmentioning

confidence: 99%

Novel functional sequences uncovered through a bovine multi-assembly graph

Crysnanto

Leonard

Fang

et al. 2021

Preprint

View full text Add to dashboard Cite

Linear reference genomes are typically assembled from single individuals. They are unable to reflect the genetic diversity of populations and lack millions of bases. To overcome such limitations and make non-reference sequences amenable to genetic investigations, we build a multi-assembly graph from six reference-quality assemblies from taurine cattle and their close relatives. We uncover 70,329,827 bases that are missing in the bovine linear reference genome. The missing sequences encode novel transcripts that are differentially expressed between individual animals. Reads which were previously poorly or unmapped against the bovine reference genome now align accurately to the non-reference sequences. We show that the non-reference sequences contain polymorphic sites that segregate within and between breeds of cattle. Our efforts to uncover novel functional sequences from a multi-assembly graph pave the way towards the transition to a more representative bovine reference genome.

show abstract

Section: Discussionmentioning

confidence: 99%

Novel functional sequences uncovered through a bovine multi-assembly graph

Crysnanto

Leonard

Fang

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…Short reads are aligned to the graph along the path of best fit, facilitating genotyping even in structurally complex and repetitive regions. Informed by a large catalog of candidate SV alleles discovered by long-read sequencing, graph genotyping thus permits the study of variants that would be difficult or impossible to discover with short-read data alone (Sibbesen et al, 2018;Chen et al, 2019;Hickey et al, 2020;Sirén et al, 2020).…”

Section: Graph Genotyping Of Structural Variationmentioning

confidence: 99%

Local adaptation and archaic introgression shape global diversity at human structural variant loci

Shu

Sherman

Taylor

et al. 2021

Preprint

View full text Add to dashboard Cite

Large genomic insertions, deletions, and inversions are a potent source of functional and fitness-altering variation, but are challenging to resolve with short-read DNA sequencing alone. While recent long-read sequencing technologies have greatly expanded the catalog of structural variants (SVs), their costs have so far precluded their application at population scales. Given these limitations, the role of SVs in human adaptation remains poorly characterized. Here, we used a graph-based approach to genotype 107,866 long-read-discovered SVs in short-read sequencing data from diverse human populations. We then applied an admixture-aware method to scan these SVs for patterns of population-specific frequency differentiation—a signature of local adaptation. We identified 220 SVs exhibiting extreme frequency differentiation, including several SVs that were among the lead variants at their corresponding loci. The top two signatures traced to separate insertion and deletion polymorphisms at the immunoglobulin heavy chain locus, together tagging a 325 Kbp haplotype that swept to high frequency and was subsequently fragmented by recombination. Alleles defining this haplotype are nearly fixed (60-95%) in certain Southeast Asian populations, but are rare or absent from other global populations composing the 1000 Genomes Project. Further investigation revealed that the haplotype closely matches with sequences observed in two of three high-coverage Neanderthal genomes, providing strong evidence of a Neanderthal-introgressed origin. This extraordinary episode of positive selection, which we infer to have occurred between 1700 and 8400 years ago, corroborates the role of immune-related genes as prominent targets of adaptive archaic introgression. Our study demonstrates how combining recent advances in genome sequencing, genotyping algorithms, and population genetic methods can reveal signatures of key evolutionary events that remained hidden within poorly resolved regions of the genome.

show abstract

“…Note that for , this use case is not directly supported, as it is designed to genotype structural variants only -we include the results anyway as this outperformed . We also note that during the finalisation of this paper, a new caller based on (named [8]) was released, which we have not tested here. Third, we show how locally defined alternate references allow accessing small variants on top of diverged forms of a dimorphic gene in P. falciparum .…”

Section: Discussionmentioning

confidence: 99%

“…There are data structures that in principle can genotype alternate alleles which include both long structural variants and SNPs -some implementations include , , , [4–7]. All of these are based on graph representations of one form or another ranging from genotyping a whole-genome de Bruijn graph (), mapping all reads to a whole-genome Directed Acyclic Graph (DAG) of informative k-mers (), mapping all reads to a wholegenome graph of minimizing k-mers and matched haplotype index (/ [8]) or remapping premapped reads either to local DAGs of SNPs and indels off the reference (), or to graphs built from structural variant breakpoints ( [9]). These all reduce the impact of reference bias, and allow cohort genotyping at consistent sites, but all of them struggle with the issue of representation.…”

Section: Introductionmentioning

confidence: 99%

Enabling multiscale variation analysis with genome graphs

Letcher

Hunt

Iqbal

2021

Preprint

View full text Add to dashboard Cite

Background: Standard approaches to characterising genetic variation revolve around mapping reads to a reference genome and describing variants in terms of differences from the reference; this is based on the assumption that these differences will be small and provides a simple coordinate system. However this fails, and the coordinates break down, when there are diverged haplotypes at a locus (e.g. one haplotype contains a multi-kilobase deletion, a second contains a few SNPs, and a third is highly diverged with hundreds of SNPs). To handle these, we need to model genetic variation that occurs at different length-scales (SNPs to large structural variants) and that occurs on alternate backgrounds. We refer to these together as multiscale variation. Results: We model the genome as a directed acyclic graph consisting of successive hierarchical subgraphs ("sites") that naturally incorporate multiscale variation, and introduce an algorithm for genotyping, implemented in the software gramtools. This enables variant calling on different sequence backgrounds. In addition to producing regular VCF files, we introduce a JSON file format based on VCF, which records variant site relationships and alternate sequence backgrounds. We show two applications. First, we benchmark gramtools against existing state-of-the-art methods in joint genotyping 17 M. tuberculosis samples at long deletions and the overlapping small variants that segregate in a cohort of 1,017 genomes. Second, in 706 African and SE Asian P. falciparum genomes, we analyse a dimorphic surface antigen gene which possesses variation on two diverged backgrounds which appeared to not recombine. This generates the first map of variation on both backgrounds, revealing patterns of recombination that were previously unknown. Conclusions: We need new approaches to be able to jointly analyse SNP and structural variation in cohorts, and even more to handle variants on different genetic backgrounds. We have demonstrated that by modelling with a directed, acyclic and locally hierarchical genome graph, we can apply new algorithms to accurately genotype dense variation at multiple scales. We also propose a generalisation of VCF for accessing multiscale variation in genome graphs, which we hope will be of wide utility.

show abstract

Genotyping common, large structural variations in 5,202 genomes using pangenomes, the Giraffe mapper, and the vg toolkit

Cited by 32 publications

References 61 publications

Novel functional sequences uncovered through a bovine multi-assembly graph

Novel functional sequences uncovered through a bovine multi-assembly graph

Local adaptation and archaic introgression shape global diversity at human structural variant loci

Enabling multiscale variation analysis with genome graphs

Contact Info

Product

Resources

About