Pangenome (cattle)

JBrowse v5 beta

The tutorials target the JBrowse v5 beta, and the v4.3.0 release on the download page lacks some of what they show. To install the beta of JBrowse Web, run npm install -g @jbrowse/cli@next, then jbrowse create jbrowse2 --branch v5.0.0-beta.13. Desktop beta builds are coming soon.

The bovine super-pangenome builds twelve assemblies into one graph on the ARS-UCD1.2 cattle reference, a Hereford that is itself one of the twelve: taurine and indicine breeds, yak, bison and gaur. Each assembly is a named path through the graph, and vg deconstruct turns those paths into a VCF, so one locus reads both as a graph of where sequence is present and absent and as a callset listing which assemblies have it. We:

read a whole chromosome with one node per variant region
at HSPA1A, compare the graph's list of alleles with the callset
find three published breed and species variants in the callset

Every step starts from the graph's page on genomes.jbrowse.org, which launches the JBrowse 5 development build until JBrowse 5 ships.

Experimental

The graph view is a beta plugin. We welcome your feedback.

Prerequisites

the GraphGenomeView plugin, for the GraphTrack below. In your own JBrowse, install it from the plugin store; the portal page already loads it
python3 and htslib (bgzip, tabix), to build the OMIA track
vg, to deconstruct the graph into a callset

Where the data comes from

Show full URLs

The OMIA build script fetches OMIA's dump, and the track configs below read the hosted files by URL.

OMIA's database dump, the source of the curated variant track: omia.org/…/omia.sql.gzhttps://omia.org/static/omia.sql.gz

Read by URL (no download needed)

the segment and link index: jbrowse.org/…/bovine-arsucd12-minigraph.segs.bed.gzhttps://jbrowse.org/demos/bovine_pangenome/bovine-arsucd12-minigraph.segs.bed.gz and jbrowse.org/…/bovine-arsucd12-minigraph.links.bed.gzhttps://jbrowse.org/demos/bovine_pangenome/bovine-arsucd12-minigraph.links.bed.gz
the bubble index, where assemblies' paths split and rejoin: jbrowse.org/…/bovine-arsucd12-minigraph.bubbles.bed.gzhttps://jbrowse.org/demos/bovine_pangenome/bovine-arsucd12-minigraph.bubbles.bed.gz
the allele inventory, one row per alternative sequence at a bubble: jbrowse.org/…/bovine-arsucd12-minigraph.alleles.bed.gzhttps://jbrowse.org/demos/bovine_pangenome/bovine-arsucd12-minigraph.alleles.bed.gz
the overview used for whole chromosomes, with variants under 10 kb folded into the reference: jbrowse.org/…/bovine-arsucd12-minigraph.fold10000.segs.bed.gzhttps://jbrowse.org/demos/bovine_pangenome/bovine-arsucd12-minigraph.fold10000.segs.bed.gz and jbrowse.org/…/bovine-arsucd12-minigraph.fold10000.links.bed.gzhttps://jbrowse.org/demos/bovine_pangenome/bovine-arsucd12-minigraph.fold10000.links.bed.gz

Hosting your own graph describes each file and how to produce it.

Load the genome and the graph

We'll load the reference the graph is anchored to, then the graph track. Swap the prefix in uri for your own gfa-to-tabix build, which writes the tabix-indexed segments and links and the overview.

Config file

Goes in the assemblies array of config.json. See Assemblies.

{
  "name": "bosTau9",
  "aliases": ["ARS-UCD1.2"],
  "uri": "https://hgdownload.soe.ucsc.edu/goldenPath/bosTau9/bigZips/bosTau9.2bit",
  "refNameAliases": {
    "uri": "https://jbrowse.org/ucsc/bosTau9/bosTau9.chromAlias.txt"
  }
}

CLI (add-assembly)

jbrowse add-assembly https://hgdownload.soe.ucsc.edu/goldenPath/bosTau9/bigZips/bosTau9.2bit \
  --name bosTau9 \
  --alias ARS-UCD1.2 \
  --refNameAliases https://jbrowse.org/ucsc/bosTau9/bosTau9.chromAlias.txt

Config file

Goes in the tracks array of config.json. See Tracks.

{
  "type": "GraphTrack",
  "trackId": "bovine_minigraph_segments",
  "name": "Bovine super-pangenome (rGFA segments)",
  "assemblyNames": ["bosTau9"],
  "adapter": {
    "type": "RgfaTabixAdapter",
    "uri": "https://jbrowse.org/demos/bovine_pangenome/bovine-arsucd12-minigraph",
    "coarse": {
      "uri": "https://jbrowse.org/demos/bovine_pangenome/bovine-arsucd12-minigraph.fold10000",
      "aboveBpPerPx": 1000
    }
  },
  "displays": [
    { "type": "LinearGraphDisplay" },
    { "type": "LinearBasicDisplay" }
  ]
}

CLI (add-track-json)

jbrowse add-track-json '{
  "type": "GraphTrack",
  "trackId": "bovine_minigraph_segments",
  "name": "Bovine super-pangenome (rGFA segments)",
  "assemblyNames": ["bosTau9"],
  "adapter": {
    "type": "RgfaTabixAdapter",
    "uri": "https://jbrowse.org/demos/bovine_pangenome/bovine-arsucd12-minigraph",
    "coarse": {
      "uri": "https://jbrowse.org/demos/bovine_pangenome/bovine-arsucd12-minigraph.fold10000",
      "aboveBpPerPx": 1000
    }
  },
  "displays": [
    { "type": "LinearGraphDisplay" },
    { "type": "LinearBasicDisplay" }
  ]
}'

In the app

In JBrowse Desktop, or in any running JBrowse Web session, open a view on this track’s assembly, then File → Open track..., choose Add track from pasted JSON, and paste:

{
  "type": "GraphTrack",
  "trackId": "bovine_minigraph_segments",
  "name": "Bovine super-pangenome (rGFA segments)",
  "assemblyNames": ["bosTau9"],
  "adapter": {
    "type": "RgfaTabixAdapter",
    "uri": "https://jbrowse.org/demos/bovine_pangenome/bovine-arsucd12-minigraph",
    "coarse": {
      "uri": "https://jbrowse.org/demos/bovine_pangenome/bovine-arsucd12-minigraph.fold10000",
      "aboveBpPerPx": 1000
    }
  },
  "displays": [
    { "type": "LinearGraphDisplay" },
    { "type": "LinearBasicDisplay" }
  ]
}

Overview of chr23 with one node per variant region

Click chr23 on the Whole chromosomes line of the portal page. The whole chromosome opens with the graph drawn as one node per bubble, a stretch where the assemblies' paths split and rejoin.

Chromosome 23 on ARS-UCD1.2, one axis: RefSeq genes, a curve of how many segments each bubble holds (higher means more variation), and the graph with one node per bubble. The curve peaks over BoLA.

BoLA, the cattle major histocompatibility complex (immune genes that vary a lot between breeds), contains the heat shock gene HSPA1A, the subject of the next section.

HSPA1A in the graph and the callset

ARS-UCD1.2 lacks an 11 kb segment beside the heat shock gene HSPA1A, which contains HSPA1B, its near-identical copy. Leonard et al. (2022) recovered it in every assembly they built. The graph lists that segment in the allele inventory. These graphs record no build order, so its firstSeenIn column names the first assembly in a fixed list, which need not contain the segment.

The callset gives a genotype per assembly. We ran vg deconstruct once per chromosome over the same graph:

# -p: vg's PackedGraph format, which deconstruct reads
vg convert -g chr1.gfa -p > chr1.vg
# -p: the reference path to decompose against
# -a: nested snarls too, so a bubble inside a bubble gets a record
vg deconstruct -p chr1 -a -t 8 chr1.vg > chr1.vcf

The VCF's CHROM column is the -p path name, which has to equal the assembly's refName. Its sample names are the three-letter codes on each assembly's path, and the sample table's first column has to match them. A mismatch leaves the row unlabelled and raises no error:

name	breed	lineage
ANG	Angus	taurine
BIS	Bison	bison

In grammar-of-graphics terms, rowColor in the track config is a color scale on each row's label bar, mapping the sample table's lineage column through domain to range. The config also sets:

rows.domain lists the cattle breeds above the wild species, and rows.labels writes each breed beside its row
unit: "haplotype" draws one row per assembly, each being one haplotype, with a second alternate allele in a separate color
showVariantLane draws each call once in a lane above the rows, across the reference span it replaces, labelled with its VCF ID (the graph nodes that bound it) and the allele change

Config file

Goes in the tracks array of config.json. See Tracks.

{
  "type": "VariantTrack",
  "trackId": "bovine_pangenome_vcf",
  "name": "Bovine super-pangenome variants (12 assemblies)",
  "assemblyNames": ["bosTau9"],
  "adapter": {
    "type": "VcfTabixAdapter",
    "uri": "https://jbrowse.org/demos/bovine_pangenome/bovine-arsucd12-minigraph.vcf.gz",
    "samplesTsvLocation": {
      "uri": "https://jbrowse.org/demos/bovine_pangenome/bovine-arsucd12-minigraph.samples.tsv"
    }
  },
  "displays": [
    {
      "type": "LinearMultiSampleVariantDisplay",
      "unit": "haplotype",
      "showVariantLane": true,
      "rows": {
        "domain": [
          "ANG",
          "BSW",
          "HIG",
          "OBV",
          "PIE",
          "SIM",
          "BRA",
          "NEL",
          "GAU",
          "BIS",
          "YAK"
        ],
        "labels": {
          "ANG": "Angus",
          "BSW": "Brown Swiss",
          "HIG": "Highland",
          "OBV": "Original Braunvieh",
          "PIE": "Piedmontese",
          "SIM": "Simmental",
          "BRA": "Brahman",
          "NEL": "Nellore",
          "GAU": "Gaur",
          "BIS": "Bison",
          "YAK": "Yak"
        }
      },
      "rowColor": {
        "field": "lineage",
        "domain": ["taurine", "indicine", "gaur", "bison", "yak"],
        "range": ["#0072B2", "#E69F00", "#009E73", "#CC79A7", "#D55E00"]
      }
    }
  ]
}

CLI (add-track-json)

jbrowse add-track-json '{
  "type": "VariantTrack",
  "trackId": "bovine_pangenome_vcf",
  "name": "Bovine super-pangenome variants (12 assemblies)",
  "assemblyNames": ["bosTau9"],
  "adapter": {
    "type": "VcfTabixAdapter",
    "uri": "https://jbrowse.org/demos/bovine_pangenome/bovine-arsucd12-minigraph.vcf.gz",
    "samplesTsvLocation": {
      "uri": "https://jbrowse.org/demos/bovine_pangenome/bovine-arsucd12-minigraph.samples.tsv"
    }
  },
  "displays": [
    {
      "type": "LinearMultiSampleVariantDisplay",
      "unit": "haplotype",
      "showVariantLane": true,
      "rows": {
        "domain": [
          "ANG",
          "BSW",
          "HIG",
          "OBV",
          "PIE",
          "SIM",
          "BRA",
          "NEL",
          "GAU",
          "BIS",
          "YAK"
        ],
        "labels": {
          "ANG": "Angus",
          "BSW": "Brown Swiss",
          "HIG": "Highland",
          "OBV": "Original Braunvieh",
          "PIE": "Piedmontese",
          "SIM": "Simmental",
          "BRA": "Brahman",
          "NEL": "Nellore",
          "GAU": "Gaur",
          "BIS": "Bison",
          "YAK": "Yak"
        }
      },
      "rowColor": {
        "field": "lineage",
        "domain": ["taurine", "indicine", "gaur", "bison", "yak"],
        "range": ["#0072B2", "#E69F00", "#009E73", "#CC79A7", "#D55E00"]
      }
    }
  ]
}'

In the app

In JBrowse Desktop, or in any running JBrowse Web session, open a view on this track’s assembly, then File → Open track..., choose Add track from pasted JSON, and paste:

{
  "type": "VariantTrack",
  "trackId": "bovine_pangenome_vcf",
  "name": "Bovine super-pangenome variants (12 assemblies)",
  "assemblyNames": ["bosTau9"],
  "adapter": {
    "type": "VcfTabixAdapter",
    "uri": "https://jbrowse.org/demos/bovine_pangenome/bovine-arsucd12-minigraph.vcf.gz",
    "samplesTsvLocation": {
      "uri": "https://jbrowse.org/demos/bovine_pangenome/bovine-arsucd12-minigraph.samples.tsv"
    }
  },
  "displays": [
    {
      "type": "LinearMultiSampleVariantDisplay",
      "unit": "haplotype",
      "showVariantLane": true,
      "rows": {
        "domain": [
          "ANG",
          "BSW",
          "HIG",
          "OBV",
          "PIE",
          "SIM",
          "BRA",
          "NEL",
          "GAU",
          "BIS",
          "YAK"
        ],
        "labels": {
          "ANG": "Angus",
          "BSW": "Brown Swiss",
          "HIG": "Highland",
          "OBV": "Original Braunvieh",
          "PIE": "Piedmontese",
          "SIM": "Simmental",
          "BRA": "Brahman",
          "NEL": "Nellore",
          "GAU": "Gaur",
          "BIS": "Bison",
          "YAK": "Yak"
        }
      },
      "rowColor": {
        "field": "lineage",
        "domain": ["taurine", "indicine", "gaur", "bison", "yak"],
        "range": ["#0072B2", "#E69F00", "#009E73", "#CC79A7", "#D55E00"]
      }
    }
  ]
}

Open in JBrowse

Open this track in JBrowse Web ↗

Open this track in JBrowse Desktop ↗ (JBrowse Desktop 5.0+)

In the chr23 view the portal opened, type chr23:27,508,000-27,536,000 to draw the graph around HSPA1A. Then:

Turn on the callset and the allele inventory in the track selector.
Open Graph settings... in the graph track's menu and set Bubble spread to Compress lengths. An insertion has no reference span to draw along, and the portal opens the graph force-directed, which draws it by its shape.

HSPA1A on ARS-UCD1.2: RefSeq genes, the deconstructed callset with one row per assembly, the allele inventory, and the graph track. The variant lane over the rows marks the insertion beside HSPA1A. Every row but the yak has the insertion, which the inventory lists once, and the graph draws it as the charcoal loop off the backbone at HSPA1A.

The yak row has the reference allele, and Leonard et al. built no yak assembly, so their result does not cover it.

Published variants in the callset

Each locus below is a structural variant a paper reported in one of these breeds or species. The rows without it are the control.

The Celtic polled allele

Angus cattle are born without horns. Medugorac et al. (2012) traced the Celtic form of polledness to a 202 bp duplication and insertion on chromosome 1, which OMIA, the Online Mendelian Inheritance in Animals database, curates as OMIA 000483-9913.

build_omia_cattle_variants.sh writes omia_cattle_variants.gff3.gz from OMIA's nightly database dump. It keeps the cattle records published on ARS-UCD1.2 or ARS-UCD1.3, which share every chromosome's coordinates:

curl -fO https://raw.githubusercontent.com/GMOD/jbrowse-components/main/scripts/build_omia_cattle_variants.sh
bash build_omia_cattle_variants.sh

Add the records as a track:

Config file

Goes in the tracks array of config.json. See Tracks.

{
  "trackId": "omia_cattle_variants",
  "name": "OMIA causal variants (cattle)",
  "uri": "omia_cattle_variants.gff3.gz",
  "assemblyNames": ["bosTau9"],
  "displayDefaults": {
    "labels": {
      "description": "jexl:feature.inheritance"
    }
  }
}

CLI (add-track)

jbrowse add-track omia_cattle_variants.gff3.gz \
  --trackId omia_cattle_variants \
  --name "OMIA causal variants (cattle)" \
  --assemblyNames bosTau9 \
  --displayDefaults '{"labels":{"description":"jexl:feature.inheritance"}}' \
  --load copy

In the app

In JBrowse Desktop, or in any running JBrowse Web session, open a view on this track’s assembly, then File → Open track..., choose Add track from pasted JSON, and paste:

{
  "trackId": "omia_cattle_variants",
  "name": "OMIA causal variants (cattle)",
  "uri": "omia_cattle_variants.gff3.gz",
  "assemblyNames": ["bosTau9"],
  "displayDefaults": {
    "labels": {
      "description": "jexl:feature.inheritance"
    }
  }
}

omia_cattle_variants.gff3.gz is relative to a config.json. Replace it with its URL or its path on this computer.

Then open chr1:2,428,800-2,429,800. The figure shades chr1:2,429,109-2,429,320, the stretch OMIA's record says the allele duplicates.

The POLLED locus on ARS-UCD1.2: OMIA's record of the Celtic polled allele, and the callset. The Angus row alone has the insertion, at the right edge of the shaded sequence it copies; every other row is reference.

OMIA also records the Friesian polled allele, an 80 kb duplication 200 kb further along, which Holstein cattle have. The panel has no Holstein, and the callset holds nothing that size there.

A repeat upstream of KIT in white-headed cattle

Simmental and Hereford cattle have white heads. Milia et al. (2025) tied the trait to a 14.3 kb segment repeated in tandem upstream of KIT: white-headed breeds have extra copies, color-headed breeds a deletion. The Hereford reference holds one collapsed copy. Open chr6:70,080,000-70,180,000, and put the graph track back on the reference's coordinates under the callset with Anchored under the Layout row of its menu.

Upstream of KIT on ARS-UCD1.2: RefSeq genes, the callset and the graph track. The Simmental row has a distinct allele across the repeat and every other row has the deletion, which the graph draws as the dashed arc skipping the repeat.

The larger box at the left of the Simmental row is an insertion the length of the 14.3 kb segment: Simmental has one more copy than the Hereford reference.

A deletion of TAS2R46 in gaur

TAS2R46 encodes a bitter taste receptor. Leonard et al. (2022) found a 17 kb deletion in gaur that removes it. Open chr5:98,575,000-98,615,000.

TAS2R46 on ARS-UCD1.2: RefSeq genes, the callset and the graph track. The gaur row has the deletion, which the graph draws as the dashed arc over TAS2R46, and four cattle rows have a different allele across the same span.

Angus, Piedmontese, Brahman and Nellore, two taurine and two indicine breeds, share that allele, and only the gaur's deletes the span. The insertion boxed in each of the four rows, between TAS2R46 and the next gene, is most of the difference.

Building the bovine graph files

Pangenome (hosting your own graph) turns a graph into the files above with one command, gfa-to-tabix build. The published bovine graphs need three steps beyond it, which build_bovine_pangenome.sh runs:

Join the autosomes. The archive holds one graph per autosome, each numbering its segments from 1, so the script renumbers and concatenates them.
Recover rGFA tags. The graphs state their coordinates in P lines, and gfa_paths_to_rgfa.py converts them back into SN/SO/SR tags (each segment's contig, offset and rank).
Deconstruct the callset with the vg deconstruct call above.

The script stops if the reference path does not reproduce bosTau9's chromosome lengths, and a README.txt beside the hosted files records the source and every modification.

Leonard et al. also built pggb and Minigraph-Cactus graphs of the same twelve assemblies, base-level GFAs with path lines and no rGFA tags. gfa-to-tabix build writes an SM:Z: tag per segment listing the assemblies that pass through it.

Citations

Leonard AS, Crysnanto D, Fang ZH, Heaton MP, Vander Ley BL, Herrera C, Bollwein H, Bickhart DM, Kuhn KL, Smith TPL, Rosen BD, Pausch H. Structural variant-based pangenome construction has low sensitivity to variability of haplotype-resolved bovine assemblies. Nature Communications. 2022;13:3012. https://doi.org/10.1038/s41467-022-30680-2
Leonard AS, Crysnanto D, Mapel XM, Bhati M, Pausch H. Graph construction method impacts variation representation and analyses in a bovine super-pangenome. Genome Biology. 2023;24:124. https://doi.org/10.1186/s13059-023-02969-y
Li H, Feng X, Chu C. The design and construction of reference pangenome graphs with minigraph. Genome Biology. 2020;21:265. https://doi.org/10.1186/s13059-020-02168-z
Medugorac I, Seichter D, Graf A, Russ I, Blum H, Göpel KH, Rothammer S, Förster M, Krebs S. Bovine polledness: an autosomal dominant trait with allelic heterogeneity. PLoS ONE. 2012;7(6):e39477. https://doi.org/10.1371/journal.pone.0039477
Milia S, Leonard AS, Mapel XM, Bernal Ulloa SM, Drögemüller C, Pausch H. Taurine pangenome uncovers a segmental duplication upstream of KIT associated with depigmentation in white-headed cattle. Genome Research. 2025;35(4):1041-1052. https://doi.org/10.1101/gr.279064.124

Feedback on this tutorial is welcome: contact us.

Pangenome (cattle)

Prerequisites#

Where the data comes from#

Load the genome and the graph#

Overview of chr23 with one node per variant region#

HSPA1A in the graph and the callset#

Published variants in the callset#

The Celtic polled allele#

A repeat upstream of KIT in white-headed cattle#

A deletion of TAS2R46 in gaur#

Building the bovine graph files#

See also#

Citations#

Prerequisites

Where the data comes from

Load the genome and the graph

Overview of chr23 with one node per variant region

HSPA1A in the graph and the callset

Published variants in the callset

The Celtic polled allele

A repeat upstream of KIT in white-headed cattle

A deletion of TAS2R46 in gaur

Building the bovine graph files

See also

Citations