Pangenome (HPRC) part 1: the graph's alleles and the haplotypes that have them

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.20. Desktop beta builds for Linux and macOS are on the releases page; the Windows build waits on code signing.

The Human Pangenome Reference Consortium's release 2 builds 464 human haplotypes into a pangenome graph; the phased VCF used here lists 462 of them. Where the haplotypes differ, the graph forms a bubble, and each route through a bubble is an allele. HPRC writes each haplotype's route through every bubble as a phased VCF, so a record lists which haplotypes have which allele. We open the graph and the VCF from the HPRC page on genomes.jbrowse.org as tracks of a linear view of GRCh38, at the complement genes C4 and beside HLA-DRB5 in MHC class II. We:

launch tracks from the HPRC page, and open a whole chromosome as a graph track with one node per bubble
open C4 as a force-directed graph track and name its backbone, bubbles and alleles
open the assembly an allele beside HLA-DRB5 came from with Open in
load the phased VCF as a variant track with one row per haplotype, and cluster the rows
load HPRC's multiple alignment as a MAF track at C4

Experimental

The graph view is a beta plugin, and the HPRC page launches it on the JBrowse 5 development build until JBrowse 5 ships. We welcome your feedback.

The graph and the page also leave real variation out, at some of the best-known loci in the genome. Read what they do not show before trusting an empty or simple picture.

Prerequisites

the GraphGenomeView plugin. In your own JBrowse, install it from the plugin store; every launch from the HPRC page already loads it
for the alignment on your own data: taffy

The data

Nothing to download: the HPRC page's launches read the files of HPRC release 2, and our tabix index of its graph, by URL. Hosting your own graph builds the same files for a graph of your own.

The files

What it is	File
The structural-variant graph (`sv.gfa`) the index files are cut from	hprc-v2.1-mc-grch38.sv.gfa.gz
The phased VCF, one record per allele, 462 haplotypes	hprc-v2.1-mc-grch38.pgbi.vcf.gz
The multiple alignment of every haplotype against GRCh38	hprc-v2.1-mc-grch38.full.maf.gz

Launching tracks from the HPRC page

Open the HPRC page, which opens on its first example. Click another Examples entry, or type a gene symbol or a region such as chr6:32,510,001-32,600,000 into the Gene or region box and press Show. The page draws a map of the window: GRCh38's genes, then one row per structural form the haplotypes fall into, commonest first, each with its share of the haplotypes aligned there.

The HPRC page's map of its HLA / MHC example. The boxed button opens the window in JBrowse. Open the HPRC page ↗ (large dataset — this session takes a while to load)

Open in JBrowse opens the window in a JBrowse tab. Clicking a row draws that form's largest change as a graph under the map, cut from the graph's haplotype walks, and the button then opens that form's haplotype as a lane. More about this window unfolds a table of each change with its carriers, and every launch:

Graph: the region drawn as a graph, with the structural variants each haplotype has under it
Variants: the structural variants each haplotype has, over the genes
Haplotypes: a lane for each structural form, commonest first, and more lanes for the forms more haplotypes carry
BandageJS: the same haplotypes' walks in BandageJS

A window up to 150 kb wide adds two rows under the genes in every launch: the known alleles in it, and the GWAS Catalog's hits. A window over 150 kb offers the graph alone. The page keeps the window in its address, so a link such as ?region=chr6:32,510,001-32,600,000 opens the same map.

A known allele: CFHR3–CFHR1

Some windows hold an allele with a published story. Click the CFH / CFHR example. Above the map, the page names the CFHR3–CFHR1 deletion, says what it does, counts its carriers in the panel and cites Hughes et al. 2006, who tied it to a lower risk of age-related macular degeneration. The map calls the change by the same name.

Press Open in JBrowse. Under the genes:

Known alleles spans the deletion and names it
GWAS Catalog (NHGRI-EBI) marks the trait associations in the window. Click a hit for its trait, paper and p-value; the two advanced age-related macular degeneration hits sit one on either side of the deletion
the matrix under the graph marks the deletion in each carrier's row

Overview of chr1 with one node per variant region

Open chr1 on the HPRC page and press Open in JBrowse. Zoomed out past the graph adapter's coarse level, JBrowse draws one node per bubble, so the whole chromosome fits. A curve of segments per bubble (how many pieces of sequence each bubble holds) shows how much the haplotypes disagree at each locus.

GRCh38 chr1 with the cytogenetic bands and two tracks built from the graph's bubbles: segments per bubble as a blue curve, and one node per bubble as gold blocks.

Hosting your own graph builds this overview and writes the coarse slot.

The C4 locus as a graph

Click the HLA / MHC example on the HPRC page, then Open in JBrowse. The Graph launch opens genes, the GWAS Catalog's hits, the graph in the force-directed layout, and under it a matrix of the structural variants each haplotype has, one row per haplotype. Then:

Hide the GWAS Catalog and the variant matrix in the track selector, leaving the genes over the graph.
Type C4A in the location box, pick the chr6 hit, and zoom out twice to take in C4B.

C4 as a force-directed graph under the hg38 genes. The dashed arc skips one copy of the C4-CYP21-TNX module.

The graph has four kinds of object:

the backbone is GRCh38's path, colored red to magenta by reference position
a bubble is a place where haplotypes disagree
an allele is a haplotype's sequence in place of the reference, drawn in charcoal because the reference path does not visit it
an edge that skips sequence is a deletion, drawn as a dashed arc

Opening the haplotype the HLA-DRB5 allele came from

Type chr6:32,500,000-32,560,000, a window in MHC class II, a cluster of immune-system HLA genes. Beside HLA-DRB5, a charcoal node is an allele much shorter than the backbone stretch it hangs across. Left-click it. Its details give contributingHaplotype, NA20809#2: the graph's SN tag records which assembly minigraph first took the allele from.

Open in appears when the session has an assembly named or aliased sample#haplotype, the graph's name for it. The launch's config declares every release 2 haplotype that way. To declare one, load its contig sizes (or FASTA) under a display name, with the graph's name as an alias:

Config file

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

{
  "name": "NA20809.2",
  "aliases": ["NA20809#2"],
  "uri": "https://jbrowse.org/pangenome/hprc-grch38/NA20809.2.chrom.sizes"
}

CLI (add-assembly)

jbrowse add-assembly https://jbrowse.org/pangenome/hprc-grch38/NA20809.2.chrom.sizes \
  --name NA20809.2 \
  --alias "NA20809#2"

Right-click the allele and take Open in NA20809.2. A second linear view opens below the first, on that haplotype's chromosome 6, framed on the allele with its CAT (Comparative Annotation Toolkit) gene annotation, which part 2 describes. Zoom out a few steps.

MHC class II with an allele's right-click menu open, and the view Open in NA20809.2 opens.

The haplotype's CAT annotation has HLA-DRB9 and HLA-DRB6 either side of the allele and no HLA-DRB5, the gene it hangs across on hg38.

Listing every haplotype with the HLA-DRB5 allele

The graph names one source assembly per allele. HPRC's phased VCF lists every haplotype's allele at every bubble: a record is one allele written against GRCh38, and each sample's genotype holds a 1 for each of its haplotypes with that allele. We'll add it to the graph launch's session:

Config file

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

{
  "type": "VariantTrack",
  "trackId": "hprc2_pgbi_grch38",
  "name": "HPRC release 2 phased VCF, one record per allele (462 haplotypes)",
  "assemblyNames": ["hg38"],
  "adapter": {
    "type": "VcfTabixAdapter",
    "uri": "https://s3-us-west-2.amazonaws.com/human-pangenomics/pangenomes/freeze/release2/minigraph-cactus/v2.1/hprc-v2.1-mc-grch38/hprc-v2.1-mc-grch38.pgbi.vcf.gz",
    "fetchSizeLimit": 20000000
  },
  "displays": [
    {
      "type": "LinearVariantDisplay",
      "rows": "haplotype",
      "filter": ["jexl:alleleLength(feature)>=50"]
    }
  ]
}

CLI (add-track-json)

jbrowse add-track-json '{
  "type": "VariantTrack",
  "trackId": "hprc2_pgbi_grch38",
  "name": "HPRC release 2 phased VCF, one record per allele (462 haplotypes)",
  "assemblyNames": ["hg38"],
  "adapter": {
    "type": "VcfTabixAdapter",
    "uri": "https://s3-us-west-2.amazonaws.com/human-pangenomics/pangenomes/freeze/release2/minigraph-cactus/v2.1/hprc-v2.1-mc-grch38/hprc-v2.1-mc-grch38.pgbi.vcf.gz",
    "fetchSizeLimit": 20000000
  },
  "displays": [
    {
      "type": "LinearVariantDisplay",
      "rows": "haplotype",
      "filter": ["jexl:alleleLength(feature)>=50"]
    }
  ]
}'

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": "hprc2_pgbi_grch38",
  "name": "HPRC release 2 phased VCF, one record per allele (462 haplotypes)",
  "assemblyNames": ["hg38"],
  "adapter": {
    "type": "VcfTabixAdapter",
    "uri": "https://s3-us-west-2.amazonaws.com/human-pangenomics/pangenomes/freeze/release2/minigraph-cactus/v2.1/hprc-v2.1-mc-grch38/hprc-v2.1-mc-grch38.pgbi.vcf.gz",
    "fetchSizeLimit": 20000000
  },
  "displays": [
    {
      "type": "LinearVariantDisplay",
      "rows": "haplotype",
      "filter": ["jexl:alleleLength(feature)>=50"]
    }
  ]
}

Open in JBrowse

Open this track in JBrowse Web ↗

Open in JBrowse Desktop ↗

Needs JBrowse Desktop 5.0 or newer. If nothing opens, copy the link and paste it into Desktop's Open JBrowse Web link..., on the start screen or under File → Session.

https://jbrowse.org/code/jb2/main/?config=https%3A%2F%2Fjbrowse.org%2Fucsc%2Fhg38%2Fconfig.json&session=spec-%7B%22sessionTracks%22%3A%5B%7B%22type%22%3A%22VariantTrack%22%2C%22trackId%22%3A%22hprc2_pgbi_grch38%22%2C%22name%22%3A%22HPRC+release+2+phased+VCF%2C+one+record+per+allele+%28462+haplotypes%29%22%2C%22assemblyNames%22%3A%5B%22hg38%22%5D%2C%22adapter%22%3A%7B%22type%22%3A%22VcfTabixAdapter%22%2C%22uri%22%3A%22https%3A%2F%2Fs3-us-west-2.amazonaws.com%2Fhuman-pangenomics%2Fpangenomes%2Ffreeze%2Frelease2%2Fminigraph-cactus%2Fv2.1%2Fhprc-v2.1-mc-grch38%2Fhprc-v2.1-mc-grch38.pgbi.vcf.gz%22%2C%22fetchSizeLimit%22%3A20000000%7D%2C%22displays%22%3A%5B%7B%22type%22%3A%22LinearVariantDisplay%22%2C%22rows%22%3A%22haplotype%22%2C%22filter%22%3A%5B%22jexl%3AalleleLength%28feature%29%3E%3D50%22%5D%7D%5D%7D%5D%2C%22views%22%3A%5B%7B%22type%22%3A%22LinearGenomeView%22%2C%22assembly%22%3A%22hg38%22%2C%22loc%22%3A%22chr6%3A32%2C500%2C000-32%2C560%2C000%22%2C%22tracks%22%3A%5B%22hprc2_pgbi_grch38%22%5D%7D%5D%7D&sessionName=JBrowse+docs%3A+HPRC+release+2+phased+VCF%2C+one+record+per+allele+%28462+haplotypes%29

rows: "haplotype" draws one row per haplotype, named <sample> HP1, <sample> HP2.
filter drops records, and alleleLength(feature)>=50 keeps alleles of 50 bp and up, the size of the bubbles minigraph draws. alleleLength is the length of the record's longest allele, so it counts an insertion's inserted bases.
fetchSizeLimit raises the download cap past its default. The VCF spells out each inserted allele, which puts the MHC class II window over the default.
Filter by → Edit filters... in the track menu shows the filter and changes it.
Advanced → Edit plot... in the track menu shows rows and filter as text and applies edits live.

Take Clustering → Cluster rows by genotype... in the track menu and press Run clustering. The rows reorder so haplotypes with the same alleles sit together, with a dendrogram beside them.

The VCF at MHC class II clustered from the track menu.

Video

Open the session the video starts in ↗, then take the steps it shows:

Cluster the haplotypes by genotype from the track menu
Haplotypes with the same alleles now gather into blocks

Steps

Open this view in JBrowse Web ↗ ·

Open in JBrowse Desktop ↗

Needs JBrowse Desktop 5.0 or newer. If nothing opens, copy the link and paste it into Desktop's Open JBrowse Web link..., on the start screen or under File → Session.

https://jbrowse.org/code/jb2/main/?config=test_data%2Fgraphgenomeview%2Fhprc.json&session=spec-%7B%22views%22%3A%5B%7B%22type%22%3A%22LinearGenomeView%22%2C%22assembly%22%3A%22hg38%22%2C%22loc%22%3A%22chr6%3A32%2C510%2C000-32%2C600%2C000%22%2C%22tracks%22%3A%5B%7B%22trackId%22%3A%22hg38_ncbiRefSeq_ucsc%22%2C%22geneGlyphMode%22%3A%22representative%22%2C%22displayMode%22%3A%22compact%22%2C%22height%22%3A60%7D%2C%7B%22trackId%22%3A%22hprc_minigraph_segments%22%2C%22type%22%3A%22LinearBasicDisplay%22%2C%22labels%22%3A%7B%22name%22%3A%7B%22show%22%3Afalse%7D%2C%22description%22%3A%7B%22show%22%3Afalse%7D%7D%2C%22heightMode%22%3A%22grow%22%2C%22encoding%22%3A%7B%22color%22%3A%7B%22value%22%3A%22jexl%3Aget%28feature%2C%27rank%27%29%3E0+%3F+%27rgb%2860%2C65%2C72%29%27+%3A+%27hsl%28%27+%2B+min%28300%2C+max%280%2C+%28%28get%28feature%2C%27start%27%29%2Bget%28feature%2C%27end%27%29%29%2F2+-+32510000%29+%2F+90000+*+300%29%29+%2B+%27%2C70%25%2C50%25%29%27%22%7D%7D%7D%2C%7B%22trackId%22%3A%22hprc2_pgbi_grch38%22%2C%22height%22%3A340%2C%22filter%22%3A%5B%22jexl%3AalleleLength%28feature%29%3E%3D50%22%5D%7D%5D%7D%5D%7D&sessionName=JBrowse+docs%3A+pangenome%2Fhprc_cluster_callset

Quickest: open the view, then File → Open track... for your files.

From scratch:

Open your genome: Open new genome in JBrowse Desktop, or jbrowse add-assembly for JBrowse Web.This figure uses hg38.
Add your track: File → Open track..., or check it in the track selector. Needs a bgzip-compressed, tabix-indexed GFF3 (.gff.gz + .tbi).This figure uses “NCBI RefSeq genes (hg38)”.
1. Track menu → Gene glyph → Representative transcript
2. Track menu → Feature height → Compact
Add the next track the same way.This figure uses “HPRC v2.1 graph (rGFA segments)”.
1. Track menu → Show → Names (unchecked)
2. Track menu → Show → Descriptions (unchecked)
3. Track menu → Feature height → Track sizing → Fixed feature height + autogrow track height
4. Track menu → Settings → colorA per-feature expression. The Color by radios write a solid color, a field's palette, and the strand colors only, so an expression is authored on the track config.
Add the next track the same way. Needs a bgzip-compressed, tabix-indexed VCF (.vcf.gz + .tbi).This figure uses “HPRC release 2 phased VCF, one record per allele (462 haplotypes)”.
1. Track menu → Filter by → add the expressionThis figure filters on alleleLength(feature)>=50.
Go to your region in the location box.This figure is at chr6:32,510,000-32,600,000.

Agent

Paste into a coding agent, with your goal on the first line.

My goal: [write yours, e.g. this view of my own files, or these track settings in my config.json]

Start from this JBrowse view.
- Config it loads: https://jbrowse.org/code/jb2/main/test_data/graphgenomeview/hprc.json
- Tracks:
  - NCBI RefSeq genes (hg38) (hg38_ncbiRefSeq_ucsc), FeatureTrack, a bgzip-compressed, tabix-indexed GFF3 (.gff.gz + .tbi), drawn as LinearBasicDisplay
  - HPRC v2.1 graph (rGFA segments) (hprc_minigraph_segments), GraphTrack, drawn as LinearBasicDisplay
  - HPRC release 2 phased VCF, one record per allele (462 haplotypes) (hprc2_pgbi_grch38), VariantTrack, a bgzip-compressed, tabix-indexed VCF (.vcf.gz + .tbi), drawn as LinearVariantDisplay

Docs: https://jbrowse.org/jb2/docs/agents_authoring/ (writing a config), https://jbrowse.org/jb2/docs/agents/ (driving a running JBrowse). Use the jbrowse-authoring, jbrowse-hosted-data and jbrowse-capture skills if you have them. Check any config with `npx @jbrowse/cli validate`, then render it and look: a config that loads is not yet a track that drew.

The session spec that draws the view:
{
  "views": [
    {
      "type": "LinearGenomeView",
      "assembly": "hg38",
      "loc": "chr6:32,510,000-32,600,000",
      "tracks": [
        {
          "trackId": "hg38_ncbiRefSeq_ucsc",
          "geneGlyphMode": "representative",
          "displayMode": "compact",
          "height": 60
        },
        {
          "trackId": "hprc_minigraph_segments",
          "type": "LinearBasicDisplay",
          "labels": {
            "name": {
              "show": false
            },
            "description": {
              "show": false
            }
          },
          "heightMode": "grow",
          "encoding": {
            "color": {
              "value": "jexl:get(feature,'rank')>0 ? 'rgb(60,65,72)' : 'hsl(' + min(300, max(0, ((get(feature,'start')+get(feature,'end'))/2 - 32510000) / 90000 * 300)) + ',70%,50%)'"
            }
          }
        },
        {
          "trackId": "hprc2_pgbi_grch38",
          "height": 340,
          "filter": [
            "jexl:alleleLength(feature)>=50"
          ]
        }
      ]
    }
  ]
}

Redraw the figure headlessly

cat > session.json <<'JSON'
{
  "views": [
    {
      "type": "LinearGenomeView",
      "assembly": "hg38",
      "loc": "chr6:32,510,000-32,600,000",
      "tracks": [
        {
          "trackId": "hg38_ncbiRefSeq_ucsc",
          "geneGlyphMode": "representative",
          "displayMode": "compact",
          "height": 60
        },
        {
          "trackId": "hprc_minigraph_segments",
          "type": "LinearBasicDisplay",
          "labels": {
            "name": {
              "show": false
            },
            "description": {
              "show": false
            }
          },
          "heightMode": "grow",
          "encoding": {
            "color": {
              "value": "jexl:get(feature,'rank')>0 ? 'rgb(60,65,72)' : 'hsl(' + min(300, max(0, ((get(feature,'start')+get(feature,'end'))/2 - 32510000) / 90000 * 300)) + ',70%,50%)'"
            }
          }
        },
        {
          "trackId": "hprc2_pgbi_grch38",
          "height": 340,
          "filter": [
            "jexl:alleleLength(feature)>=50"
          ]
        }
      ]
    }
  ]
}
JSON

npx @jbrowse/capture --instance https://jbrowse.org/code/jb2/main/ \
  --config https://jbrowse.org/code/jb2/main/test_data/graphgenomeview/hprc.json \
  --width 1920 --height 880 \
  --spec session.json -o figure.png

More: Using JBrowse with AI agents.

CLI

Adds these tracks to your config.json, styled as in the figure. Point each uri at your file (replace YOUR_FILE).

jbrowse add-track-json '{
  "type": "FeatureTrack",
  "trackId": "hg38_ncbiRefSeq_ucsc",
  "name": "NCBI RefSeq genes (hg38)",
  "assemblyNames": [
    "hg38"
  ],
  "adapter": {
    "type": "Gff3TabixAdapter",
    "uri": "YOUR_FILE"
  },
  "displayDefaults": {
    "geneGlyphMode": "representative",
    "displayMode": "compact",
    "height": 60
  }
}'

jbrowse add-track-json '{
  "type": "VariantTrack",
  "trackId": "hprc2_pgbi_grch38",
  "name": "HPRC release 2 phased VCF, one record per allele (462 haplotypes)",
  "assemblyNames": [
    "hg38"
  ],
  "adapter": {
    "type": "VcfTabixAdapter",
    "uri": "YOUR_FILE"
  },
  "displayDefaults": {
    "height": 340,
    "filter": [
      "jexl:alleleLength(feature)>=50"
    ]
  }
}'

Run beside config.json, or add --out <dir>. Needs hg38 in the config (jbrowse add-assembly).

Session

The session spec behind the figure, over this config.

{
  "views": [
    {
      "type": "LinearGenomeView",
      "assembly": "hg38",
      "loc": "chr6:32,510,000-32,600,000",
      "tracks": [
        {
          "trackId": "hg38_ncbiRefSeq_ucsc",
          "geneGlyphMode": "representative",
          "displayMode": "compact",
          "height": 60
        },
        {
          "trackId": "hprc_minigraph_segments",
          "type": "LinearBasicDisplay",
          "labels": {
            "name": {
              "show": false
            },
            "description": {
              "show": false
            }
          },
          "heightMode": "grow",
          "encoding": {
            "color": {
              "value": "jexl:get(feature,'rank')>0 ? 'rgb(60,65,72)' : 'hsl(' + min(300, max(0, ((get(feature,'start')+get(feature,'end'))/2 - 32510000) / 90000 * 300)) + ',70%,50%)'"
            }
          }
        },
        {
          "trackId": "hprc2_pgbi_grch38",
          "height": 340,
          "filter": [
            "jexl:alleleLength(feature)>=50"
          ]
        }
      ]
    }
  ]
}

Right-click the charcoal allele beside HLA-DRB5 in the force-directed graph and take Highlight in hg38: a band marks the span of GRCh38 it replaces, which covers most of HLA-DRB5.

The VCF clustered by genotype over the force-directed graph. The arrow runs from the HLA-DRB5 allele's band to the same allele in the graph.

The VCF holds the HLA-DRB5 allele as several records, one for each of its sequences, which differ by a base or two. The blue block under the band is the haplotypes with any of the four, plus a smaller block with an 86 kb deletion that removes the whole stretch.

Reading C4's alleles as aligned bases

HPRC also publishes its alignment as hprc-v2.1-mc-grch38.full.maf.gz, a multiple alignment with one row of bases per haplotype. A .tai index beside the large file makes a locus one ranged read:

Config file

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

{
  "type": "MafTrack",
  "trackId": "hprc_v2_1_mc_grch38",
  "name": "HPRC release 2 pangenome alignment (464 haplotypes)",
  "assemblyNames": ["hg38"],
  "adapter": {
    "type": "BgzipMafAdapter",
    "uri": "https://s3-us-west-2.amazonaws.com/human-pangenomics/pangenomes/freeze/release2/minigraph-cactus/v2.1/hprc-v2.1-mc-grch38/hprc-v2.1-mc-grch38.full.maf.gz"
  }
}

CLI (add-track-json)

jbrowse add-track-json '{
  "type": "MafTrack",
  "trackId": "hprc_v2_1_mc_grch38",
  "name": "HPRC release 2 pangenome alignment (464 haplotypes)",
  "assemblyNames": ["hg38"],
  "adapter": {
    "type": "BgzipMafAdapter",
    "uri": "https://s3-us-west-2.amazonaws.com/human-pangenomics/pangenomes/freeze/release2/minigraph-cactus/v2.1/hprc-v2.1-mc-grch38/hprc-v2.1-mc-grch38.full.maf.gz"
  }
}'

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": "MafTrack",
  "trackId": "hprc_v2_1_mc_grch38",
  "name": "HPRC release 2 pangenome alignment (464 haplotypes)",
  "assemblyNames": ["hg38"],
  "adapter": {
    "type": "BgzipMafAdapter",
    "uri": "https://s3-us-west-2.amazonaws.com/human-pangenomics/pangenomes/freeze/release2/minigraph-cactus/v2.1/hprc-v2.1-mc-grch38/hprc-v2.1-mc-grch38.full.maf.gz"
  }
}

Open in JBrowse

Open this track in JBrowse Web ↗

Open in JBrowse Desktop ↗

Needs JBrowse Desktop 5.0 or newer. If nothing opens, copy the link and paste it into Desktop's Open JBrowse Web link..., on the start screen or under File → Session.

https://jbrowse.org/code/jb2/main/?config=https%3A%2F%2Fjbrowse.org%2Fucsc%2Fhg38%2Fconfig.json&session=spec-%7B%22sessionTracks%22%3A%5B%7B%22type%22%3A%22MafTrack%22%2C%22trackId%22%3A%22hprc_v2_1_mc_grch38%22%2C%22name%22%3A%22HPRC+release+2+pangenome+alignment+%28464+haplotypes%29%22%2C%22assemblyNames%22%3A%5B%22hg38%22%5D%2C%22adapter%22%3A%7B%22type%22%3A%22BgzipMafAdapter%22%2C%22uri%22%3A%22https%3A%2F%2Fs3-us-west-2.amazonaws.com%2Fhuman-pangenomics%2Fpangenomes%2Ffreeze%2Frelease2%2Fminigraph-cactus%2Fv2.1%2Fhprc-v2.1-mc-grch38%2Fhprc-v2.1-mc-grch38.full.maf.gz%22%7D%7D%5D%2C%22views%22%3A%5B%7B%22type%22%3A%22LinearGenomeView%22%2C%22assembly%22%3A%22hg38%22%2C%22loc%22%3A%22chr6%3A31%2C980%2C000-32%2C050%2C000%22%2C%22tracks%22%3A%5B%22hprc_v2_1_mc_grch38%22%5D%7D%5D%7D&sessionName=JBrowse+docs%3A+HPRC+release+2+pangenome+alignment+%28464+haplotypes%29

The uri shorthand resolves the sibling .tai. For a MAF of your own, taffy writes the index beside it:

# writes alignment.maf.gz.tai, which the adapter finds by its name
taffy index -i alignment.maf.gz

Go back to the C4 window, chr6:31,980,000-32,050,000, with the genes, the VCF and this alignment over the graph track. A white stretch in an alignment row is sequence that haplotype lacks. Then:

Clustering → Cluster rows by genotype... on the VCF orders its rows.
Clustering → Cluster rows by identity... on the alignment clusters over the window in view, because HPRC's file ships no guide tree.
Reset row order restores the order in the file.

C4 with the RefSeq genes, the VCF, alignment rows and the graph track. The arrow runs from the shaded band to the graph segment holding it.

The VCF's widest blue block is one record, a deletion of a whole C4-CYP21-TNX module, which is the dashed arc in the C4 graph. Every alignment row in the white block has that deletion in the VCF. The narrower blocks are haplotypes that lack the 6.4 kb HERV-K (endogenous retrovirus) in a C4 gene's intron, which GRCh38 carries: the difference between a long and a short C4.

What the graph and the page do not show

The HPRC page's examples are loci where at least one launch draws something readable. At a locus of your own, a picture that looks simple or empty can mislead in these cases:

Near-identical duplications draw as a single thread. The graph puts the copies of SMN1 / SMN2, SRGAP2 and the 8p23.1 beta-defensin cluster on one path, so those examples offer no Graph launch. A gene-level graph shows copy number there.
The graph does not align every haplotype everywhere. A haplotype the graph does not align across a window gets no lane, and the page counts them. At FLNA / EMD those are the haplotypes without an X chromosome; elsewhere the sequence exists in the assembly and the graph leaves it out.
The VCF has no record for the largest events. HPRC drops the biggest sites from the VCF and keeps the smaller sites nested inside them, so the Variants launch cannot show the UGT2B17 deletion. The page's map reads each haplotype's forms from the graph itself, so it draws the deletion and names it.
The antibody loci hold the cell lines' deletions. Many HPRC assemblies come from lymphoblastoid cell lines, B cells whose V(D)J recombination deletes hundreds of kb of IGH, IGK and IGL (Rodriguez et al. 2021, IGLoo for IGH). In the walk file the lanes read, many haplotypes jump over chunks of IGH and IGK, fewer over IGL, and none over the T-cell receptor loci TRA and TRB. Check a deletion there against the donor's germline data before reporting it.
A typed locus is checked for structure only. The page reads the window you type, and where the haplotypes share one structure it says so, as do the Graph and Variants launches. A window can still land where the graph merges duplicated copies into one thread or the VCF has no record, and the page offers the launch anyway.
No single launch shows the whole locus. The graph shows which sequence exists, the VCF which haplotypes have it, and the lanes how a few haplotypes are arranged.

Reproduce it end to end

Hosting your own graph builds the index files and the bubbles track in one command that runs on any rGFA. Run on hprc-v2.1-mc-grch38.sv.gfa.gz, it writes the files we host, and README.txt beside them records their provenance. For a graph of your own, minigraph can list each assembly's route through every bubble; see Which haplotypes pass through each segment.

External links

Li H. The rGFA format and gfatools, which define the SN/SO/SR tags this page opens the graph by, and whose bubble subcommand calls the bubbles it reads.
taffy, which writes the .tai index that makes the alignment addressable by locus.
ggbandage, which draws the C4 graph from the same hosted files as ggplot2 layers in R.
ggtubemap, which draws HLA-A's haplotypes as a tube map in ggplot2.

Citations

HPRC release 2, the release this page opens: the Minigraph-Cactus graph, the multiple alignment and the VCFs vg deconstruct writes from the graph.

Feedback on this tutorial is welcome: contact us.

Pangenome (HPRC) part 1: the graph's alleles and the haplotypes that have them

Prerequisites#

The data#

Launching tracks from the HPRC page#

A known allele: CFHR3–CFHR1#

Overview of chr1 with one node per variant region#

The C4 locus as a graph#

Opening the haplotype the HLA-DRB5 allele came from#

Listing every haplotype with the HLA-DRB5 allele#

Reading C4's alleles as aligned bases#

What the graph and the page do not show#

Reproduce it end to end#

See also#

External links#

Citations#

Prerequisites

The data

Launching tracks from the HPRC page

A known allele: CFHR3–CFHR1

Overview of chr1 with one node per variant region

The C4 locus as a graph

Opening the haplotype the HLA-DRB5 allele came from

Listing every haplotype with the HLA-DRB5 allele

Reading C4's alleles as aligned bases

What the graph and the page do not show

Reproduce it end to end

See also

External links

Citations