Pangenome (HPRC) part 1, reading the graph
2 videos in this tutorial
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.8. Desktop beta builds are coming soon.
A pangenome graph records what a set of genomes share and where they diverge. Sequence that one person carries and the reference lacks is therefore an object in the file. We open the Human Pangenome Reference Consortium's release 2 graph at a locus and draw it beside GRCh38. We then follow one allele from the drawing back to the GRCh38 coordinates where it attaches, and then to the haplotype it came from. Part 2 reads who carries that sequence from the rest of the release on the same axis. Part 3 moves the same haplotypes off that axis and onto the haplotypes' assembly coordinates.
The graph view is a beta plugin, and this tutorial covers experimental ideas. Where a step below says the view works a particular way today, the step describes a current limit of the view. We welcome your feedback.
Prerequisites
- a JBrowse 5 build, Web or Desktop, with the GraphGenomeView plugin loaded; the graph track needs the plugin, and every other track here is a URL you can paste
- htslib (
bgzip,tabix), for the two commands that query the hosted indexes from the command line, andgfatoolsfor the command that built them; both are for Reproduce it end to end and the other windows, and the route itself runs none - nothing at all, to follow the first five steps: the HPRC page on genomes.jbrowse.org opens this graph at twenty loci, from the same files and the same build as the track below, so a session it opens has the graph track in it and the plugin loaded
The steps below build no graph file. Preparing your own graph builds one and lists what that takes.
Where the data comes from
The data is HPRC release 2. We read its Minigraph-Cactus graph through small tabix projections that we host beside it, and fetch the contributing assemblies from their original archives.
- the SV-resolution graph (
sv.gfa), which is the minigraph backbone we build our rGFA tabix projections from: 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.sv.gfa.gz - the CAT gene annotation index, one GFF3 per haplotype: https://raw.githubusercontent.com/human-pangenomics/hprc_intermediate_assembly/main/data_tables/annotation/cat/cat_genes_hprc_r2_v1.3.index.csv
- every release 2 assembly as a UCSC GenArk hub, a 2bit and an alias file per haplotype; this is haplotype 2 of NA20809, the one the graph opens below: https://hgdownload.soe.ucsc.edu/hubs/GCA/044/166/615/GCA_044166615.1/
- that haplotype's CAT annotation, whole genome, from the index above: https://s3-us-west-2.amazonaws.com/human-pangenomics/working/HPRC/NA20809/assemblies/release2/annotation/cat/NA20809_hap2_hprc_r2_v1.0.1_cat_v1.3.gff3.gz
- our rGFA projections, with the exact build recorded beside them: https://jbrowse.org/demos/hprc/README.txt
All three pages read the release's v2.1/ build, so node ids agree across the
pages. The one exception is the alignment that
part 3
opens, which only the v2.0/ build carries.
The route
At the MHC class II locus, many haplotypes replace 12 kb of GRCh38 with 1.8 kb of different sequence. The graph stores that replacement as one object with a length, attached to the reference on either side of the 12 kb it replaces. We find that object on the drawing, read where it sits on GRCh38, and then load the haplotype it came from to see what the replacement does to the genes there. Six steps, each one section:
- Start in hg38 with its genes and nothing else.
- Add the graph track and go to the MHC class II window.
- Cut that window out as a graph and learn the four words that describe it.
- Lay the graph out on GRCh38 coordinates to find the allele.
- Take the allele back to its GRCh38 interval and read which haplotype contributed it.
- Load that haplotype and open the allele on its own chromosome.
The GraphGenomeView plugin
GraphGenomeView is beta and not in the
plugin store yet, so it loads by URL. In
JBrowse Web that is a plugins array at the top level of config.json, beside
assemblies and tracks (see
configuring plugins):
{
"plugins": [
{
"name": "GraphGenomeView",
"esmUrl": "https://jbrowse.org/demos/graphgenomeviewer/jbrowse-plugin-graphgenomeviewer.esm.js"
}
]
}On JBrowse Desktop, install it once from the start
screen at Global plugins... → Add custom plugin, putting that esmUrl under
Advanced options in ESM build URL and leaving the two fields above it
empty.
The plugin reads two things core first exported in v5.0.0-beta.1, so it needs a JBrowse 5 build. The graph genome view guide covers the view's layouts, colors and menus on a smaller graph than this one.
Start
The route starts in a session holding hg38, its RefSeq genes and nothing of the pangenome. If your JBrowse already has hg38 with a gene track, open it there and skip to the next step. With nothing installed, press graph on the HLA / MHC row of the hosted page instead: the session it opens already holds the track the next step adds and a cut of the class II window, so pick up at reading the cut. Otherwise these two configs are the assembly and the gene track every figure on this page draws:
{
"name": "hg38",
"aliases": ["GRCh38"],
"uri": "https://jbrowse.org/genomes/GRCh38/fasta/hg38.prefix.fa.gz",
"refNameAliases": {
"uri": "https://s3.amazonaws.com/jbrowse.org/genomes/GRCh38/hg38_aliases.txt"
},
"cytobands": {
"uri": "https://s3.amazonaws.com/jbrowse.org/genomes/GRCh38/cytoBand.txt"
},
"geneticCodes": { "chrM": 2 }
}jbrowse add-assembly https://jbrowse.org/genomes/GRCh38/fasta/hg38.prefix.fa.gz \
--name hg38 \
--alias GRCh38 \
--refNameAliases https://s3.amazonaws.com/jbrowse.org/genomes/GRCh38/hg38_aliases.txt \
--config '{"cytobands":{"uri":"https://s3.amazonaws.com/jbrowse.org/genomes/GRCh38/cytoBand.txt"},"geneticCodes":{"chrM":2}}'{
"type": "FeatureTrack",
"trackId": "hg38_ncbiRefSeq_ucsc",
"name": "NCBI RefSeq genes (hg38)",
"assemblyNames": ["hg38"],
"adapter": {
"type": "Gff3TabixAdapter",
"uri": "https://jbrowse.org/ucsc/hg38/ncbiRefSeq.gff.gz",
"csi": true
}
}jbrowse add-track-json '{
"type": "FeatureTrack",
"trackId": "hg38_ncbiRefSeq_ucsc",
"name": "NCBI RefSeq genes (hg38)",
"assemblyNames": ["hg38"],
"adapter": {
"type": "Gff3TabixAdapter",
"uri": "https://jbrowse.org/ucsc/hg38/ncbiRefSeq.gff.gz",
"csi": true
}
}'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": "FeatureTrack",
"trackId": "hg38_ncbiRefSeq_ucsc",
"name": "NCBI RefSeq genes (hg38)",
"assemblyNames": ["hg38"],
"adapter": {
"type": "Gff3TabixAdapter",
"uri": "https://jbrowse.org/ucsc/hg38/ncbiRefSeq.gff.gz",
"csi": true
}
}The session below is that starting state, open on a 300 kb stretch of the MHC with the gene track showing. The clip under the cut was filmed from this session, and the link beside it opens the same state on a hosted JBrowse with the plugin already loaded, which is the quickest way to follow the page without installing anything.
{
"defaultSession": {
"name": "HPRC part 1, before the graph",
"views": [
{
"type": "LinearGenomeView",
"assembly": "hg38",
"loc": "chr6:32,400,000-32,700,000",
"tracks": ["hg38_ncbiRefSeq_ucsc"]
}
]
}
}jbrowse set-default-session --session - << 'EOF'
{
"name": "HPRC part 1, before the graph",
"views": [
{
"type": "LinearGenomeView",
"assembly": "hg38",
"loc": "chr6:32,400,000-32,700,000",
"tracks": ["hg38_ncbiRefSeq_ucsc"]
}
]
}
EOFAdd the graph track
JBrowse reads two tabix-indexed BED projections of the graph, which we host. A
FeatureTrack pointed at their shared prefix downloads only the region in view.
The adapter resolves <uri>.segs.bed.gz, <uri>.links.bed.gz, and both .tbi
files:
{
"type": "FeatureTrack",
"trackId": "hprc_minigraph_segments",
"name": "HPRC release 2 graph (rGFA segments)",
"assemblyNames": ["hg38"],
"adapter": {
"type": "RgfaTabixAdapter",
"uri": "https://jbrowse.org/demos/hprc/hprc-v2.1-mc-grch38",
"assemblyNameToPanSN": { "hg38": "GRCh38" }
},
"displayDefaults": { "showLabels": "none" }
}jbrowse add-track-json '{
"type": "FeatureTrack",
"trackId": "hprc_minigraph_segments",
"name": "HPRC release 2 graph (rGFA segments)",
"assemblyNames": ["hg38"],
"adapter": {
"type": "RgfaTabixAdapter",
"uri": "https://jbrowse.org/demos/hprc/hprc-v2.1-mc-grch38",
"assemblyNameToPanSN": { "hg38": "GRCh38" }
},
"displayDefaults": { "showLabels": "none" }
}'In JBrowse Desktop, or in any running JBrowse Web session, open a view on this track’s assembly, then File → Open track..., choose Add pangenome graph track, and fill in:
- File type:
rGFA segments (tabix BED pair) - Path:
https://jbrowse.org/demos/hprc/hprc-v2.1-mc-grch38.segs.bed.gz - Sample name in the graph:
GRCh38 - Track name:
HPRC release 2 graph (rGFA segments) - Assembly:
hg38
Change these two fields to load a different graph:
urinames a prefix. The adapter appends.segs.bed.gzand.links.bed.gzand reads the.tbibeside each, so one string points at the pair. The first cut of a session downloads both indexes, about 9 MB.assemblyNameToPanSNmaps your assembly name to the name the graph uses. The session calls the referencehg38, and the graph calls itGRCh38#0#chr6. JBrowse cannot infer the sample part of that name.
showLabels is off because a segment's name is its GFA id. Nothing sets a
color: cutting a graph from the lane paints the lane in the graph's own colors,
red at the start of the cut window to magenta at its end, so a block above and
its node below share a hue.
With the track showing, type chr6:32,500,000-32,560,000 into the location box
and press Enter. That is the MHC class II window every figure below is cut from,
and the lane draws one block per graph segment across it: the GRCh38 backbone
tiles the reference, because every segment in sv.gfa carries the rGFA tags
that place it. Where the lane stacks blocks on top of each other, haplotypes
disagree with the reference.
Cut the window out as a graph
The graph draws a window at a time. Open the segments track's menu and pick Launch → Graph genome view (this region). A graph pane opens under the linear view holding the window on screen, and the lane above takes the graph's colors. Two other ways to choose a window:
- Drag across the ruler and choose Graph genome view (this selection). If the selection is wider than the view will draw, the menu greys the item out and shows the limit.
- Right-click one segment in the lane to cut the graph around that segment.
A pane keeps the window it was cut from while the linear view moves on. At segment resolution the layout stays legible up to around a hundred kilobases, the view refuses a cut past 5 Mb, and part 2's bubble tier draws anything wider.
The clip below runs the route from this page's own start: the track added through the add-track form, the window typed in, the cut, the re-layout of the next step and the highlight of the one after.
Reading what you cut
Four words describe the drawing. The chain of segments running across it is the backbone, which is GRCh38's path through the graph. A bubble is a place where that chain opens out and closes again, at one locus where the haplotypes disagree. An allele is a loop inside a bubble, a stretch of sequence some haplotype carries in place of the reference sequence. A deletion is an edge, a dashed arc from one backbone segment to another that skips the segments between them.
Hover any node in your cut: the tooltip gives its length and its rank, and rank 0 is the backbone. The figure labels all four shapes on a smaller cut, the C4 locus half a megabase towards the centromere, which holds few enough nodes to label. Your MHC class II cut has the same parts and many more of them.
A force layout has no x axis to share with the linear view, so color links the two. The graph opens colored by Reference position, which ramps hue across the window the subgraph was cut from. A backbone (rank 0) segment takes a hue from that ramp. An allele (rank>0) segment sits on another assembly's sequence and has no GRCh38 position, so it draws flat charcoal. Hovering a node bands its interval across the lanes above, and hovering a block in the lane lights its node.
The view also draws the session's gene track onto the backbone. Each gene's exons are dark stretches along the reference nodes that carry them, with the gene name pinned under the backbone. Every bubble in the hosted index draws as a halo along its nodes, labelled with what it is. View menu → Settings turns either off (Genes on the backbone, Mark bubbles), and the guide covers what a label click does.
Lay it out on GRCh38 coordinates
Open the Layout dropdown in the graph pane's toolbar and pick Anchored. Every x is now a GRCh38 coordinate, the reference row is at the top, each lower row is one rank, and each allele hangs below the point it attaches at. A deletion draws as a dashed jump along the reference row, and the drawing lines up with the segments lane above it. The guide sets out what the five layouts put on each axis.
Both halves ring the allele the route follows: the charcoal node hanging under HLA-DRB5, near the middle of the window. Hover it and the tooltip gives its length, 1.8 kb, and the haplotype it came from. The anchored half labels it by net size as a deletion, because the allele is shorter than the 12 kb of backbone it replaces. The rank rows are what the anchored layout adds: the order minigraph added the assembly that first contributed each stretch of sequence.
From the allele back to GRCh38
Right-click the ringed node for two actions whose result stays after the hover ends:
- Highlight in hg38 marks its reference interval in the linear view beside the graph and leaves it there.
- Open in hg38 scrolls that view to it.
Take Highlight in hg38. A band appears in the linear view across the 12 kb the allele attaches over, and that band covers most of HLA-DRB5. The band stays through the rest of this page and part 2 reads its lanes against it.
What the menu offers depends on which segment you clicked, because rGFA records
each segment's source sequence (SN) and offset (SO):
- a backbone (rank 0) segment sits on GRCh38, so you get its exact coordinates there.
- an allele (rank>0) segment sits on one haplotype's sequence, e.g.
NA20809#2#CM094351.1. That coordinate is also exact, but no loaded assembly can open it yet. You get the GRCh38 interval between the two backbone segments where the allele leaves and rejoins the backbone. A hover highlights the same span.
Now left-click the same node. The details panel opens on the right, and
contributingAssembly names the first assembly to contribute the segment, with
contributingHaplotype giving the sample#haplotype part of its rGFA name:
NA20809#2. The callset records who else carries the sequence, and
part 2
reads one genotype per haplotype at the site under the band. The graph's own
Launch menu does the same as the node menu for the whole window the graph
was cut from.
Load the haplotype that contributed it
UCSC GenArk hosts every release 2
assembly as a 2bit and an alias file. GenArk names the sequences by the same
GenBank accessions the graph uses: NA20809#2#CM094351.1 in the graph is
CM094351.1 in the 2bit. A haplotype therefore loads with no name translation,
and the alias file's ucsc column adds chr6. The
HPRC sample table lists the accession
for every haplotype; the one the details panel named is GCA_044166615.1.
{
"name": "NA20809.2",
"displayName": "NA20809 haplotype 2 (HPRC release 2, GCA_044166615.1)",
"uri": "https://hgdownload.soe.ucsc.edu/hubs/GCA/044/166/615/GCA_044166615.1/GCA_044166615.1.2bit",
"refNameAliases": {
"uri": "https://hgdownload.soe.ucsc.edu/hubs/GCA/044/166/615/GCA_044166615.1/GCA_044166615.1.chromAlias.txt"
}
}jbrowse add-assembly https://hgdownload.soe.ucsc.edu/hubs/GCA/044/166/615/GCA_044166615.1/GCA_044166615.1.2bit \
--name NA20809.2 \
--displayName "NA20809 haplotype 2 (HPRC release 2, GCA_044166615.1)" \
--refNameAliases https://hgdownload.soe.ucsc.edu/hubs/GCA/044/166/615/GCA_044166615.1/GCA_044166615.1.chromAlias.txtIn JBrowse Desktop, Open new genome on the start screen (or File → Open genome... in a session), then Open from a URL and paste:
https://hgdownload.soe.ucsc.edu/hubs/GCA/044/166/615/GCA_044166615.1/GCA_044166615.1.2bitJBrowse reads the format off the file name. Then fill in:
- Genome name:
NA20809.2 - Assembly display name (under More options):
NA20809 haplotype 2 (HPRC release 2, GCA_044166615.1) - refName aliases (under More options):
https://hgdownload.soe.ucsc.edu/hubs/GCA/044/166/615/GCA_044166615.1/GCA_044166615.1.chromAlias.txt
A launched view shows the session's annotation for the assembly it opens. The
CAT index's location column names one GFF3 per haplotype. Each GFF3 covers the
whole genome and has no index, so slice out the window once and index the slice:
# haplotype 2 of NA20809, from the index's location column; chr6 is CM094351.1 there
curl -sSL https://s3-us-west-2.amazonaws.com/human-pangenomics/working/HPRC/NA20809/assemblies/release2/annotation/cat/NA20809_hap2_hprc_r2_v1.0.1_cat_v1.3.gff3.gz \
| gzip -dc \
| awk -F'\t' '/^#/ {next} $1=="CM094351.1" && $5>=32300000 && $4<=32800000' \
| sort -k1,1 -k4,4n > hprc_mhc_NA20809.2.genes.gff3
bgzip hprc_mhc_NA20809.2.genes.gff3
tabix -p gff hprc_mhc_NA20809.2.genes.gff3.gz{
"type": "FeatureTrack",
"trackId": "NA20809.2_cat_genes",
"name": "CAT genes (NA20809 haplotype 2)",
"assemblyNames": ["NA20809.2"],
"adapter": {
"type": "Gff3TabixAdapter",
"uri": "hprc_mhc_NA20809.2.genes.gff3.gz"
}
}jbrowse add-track hprc_mhc_NA20809.2.genes.gff3.gz \
--trackId NA20809.2_cat_genes \
--name "CAT genes (NA20809 haplotype 2)" \
--assemblyNames NA20809.2 \
--load copyIn 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": "FeatureTrack",
"trackId": "NA20809.2_cat_genes",
"name": "CAT genes (NA20809 haplotype 2)",
"assemblyNames": ["NA20809.2"],
"adapter": {
"type": "Gff3TabixAdapter",
"uri": "hprc_mhc_NA20809.2.genes.gff3.gz"
}
}hprc_mhc_NA20809.2.genes.gff3.gz is relative to a config.json. Replace it with its URL or its path on this computer.
The segments track also draws on the haplotype once assemblyNames lists the
haplotype and assemblyNameToPanSN names the graph haplotype that the entry
stands for. The view needs no other link between the two. A graph cut from the
track reads the same map to open a node contributed by NA20809#2 on
NA20809.2, so the assembly can be given any name. hg38 stays first, because
the view cuts a graph on the first assembly its track names. The config below
replaces the track above, with the same trackId and
one more assembly:
{
"type": "FeatureTrack",
"trackId": "hprc_minigraph_segments",
"name": "HPRC release 2 graph (rGFA segments)",
"assemblyNames": ["hg38", "NA20809.2"],
"adapter": {
"type": "RgfaTabixAdapter",
"uri": "https://jbrowse.org/demos/hprc/hprc-v2.1-mc-grch38",
"assemblyNameToPanSN": { "hg38": "GRCh38", "NA20809.2": "NA20809#2" }
},
"displayDefaults": {
"color": "jexl:feature.rank==0 ? 'rgb(52,152,219)' : 'rgb(237,137,44)'"
}
}jbrowse add-track-json '{
"type": "FeatureTrack",
"trackId": "hprc_minigraph_segments",
"name": "HPRC release 2 graph (rGFA segments)",
"assemblyNames": ["hg38", "NA20809.2"],
"adapter": {
"type": "RgfaTabixAdapter",
"uri": "https://jbrowse.org/demos/hprc/hprc-v2.1-mc-grch38",
"assemblyNameToPanSN": { "hg38": "GRCh38", "NA20809.2": "NA20809#2" }
},
"displayDefaults": {
"color": "jexl:feature.rank==0 ? '\''rgb(52,152,219)'\'' : '\''rgb(237,137,44)'\''"
}
}'In JBrowse Desktop, or in any running JBrowse Web session, open a view on this track’s assembly, then File → Open track..., choose Add pangenome graph track, and fill in:
- File type:
rGFA segments (tabix BED pair) - Path:
https://jbrowse.org/demos/hprc/hprc-v2.1-mc-grch38.segs.bed.gz - Sample name in the graph:
GRCh38 - Track name:
HPRC release 2 graph (rGFA segments) - Assembly:
hg38
Cut chr6:32,500,000-32,560,000 as a graph again and right-click the same 1.8
kb allele. The menu now has Open in NA20809.2, which adds a linear view of
that haplotype's chr6 framed on the allele, with its CAT genes and the segments
the graph records as contributed by it. Zoom that view out a few steps for the
genes around it.
On GRCh38 the ringed allele attaches across the 12 kb of backbone the band
marks, and that 12 kb covers most of HLA-DRB5. On NA20809 haplotype 2 the same
1.8 kb sits in the gap between HLA-DRB9 and HLA-DRB6, and that haplotype's
annotation has no HLA-DRB5 model at all. The two views show one structural
difference from each side. The same steps work for any of the 464 haplotypes:
read contributingHaplotype off the node, find its accession in the sample
table, and paste the two configs with the names changed.
Other windows to cut
The route works on any window the view will draw. Each locus below is small enough at segment resolution, and each shows a shape the MHC window does not:
| Locus | Window | What the cut holds |
|---|---|---|
| C4 | chr6:31,980,000-32,050,000 | one copy of a tandem module as a dashed arc |
| KIR | chr19:54,750,000-54,840,000 | a gene cluster that varies in gene count |
| LPA KIV-2 | chr6:160,525,000-160,655,000 | a repeat array as a knot of loops |
| CFH cluster | chr1:196,700,000-196,900,000 | a two-gene deletion as one edge |
| AMY1 | chr1:103,500,000-103,850,000 | a copy-number bubble wider than its genes |
Every node and every deletion arc has a size. Extra sequence is a node, drawn as a tube. Missing sequence is an edge, drawn dashed and near-black, outside the color ramp, and hovering an edge gives the interval and the bp it removes. Read a deletion on the anchored layout, where the arc spans exactly the sequence it removes.
Part 4 cuts the LPA KIV-2 window from the table as a graph, opens its repeat array and lifts one haplotype's walk out of it.
An edge carries no sequence, so it names no donor haplotype and opens only on GRCh38. The two haplotype rows above the graph are release 2's CAT annotation on each haplotype's own assembly, loaded the way the haplotype step loads NA20809.2, over alignments that part 3 reads out of the graph.
The view cuts only the window, so an allele whose interior falls outside the
window draws as a short arm off the backbone. When a cut comes back as a single
tangle, widen it until the backbone chain is in frame. The amylase window in the
table is wider than its bubble for that reason. The hosted bubble index says
what any window holds, and tabix reads it over HTTP; the bubble spanning
AMY1A and AMY1B is the first row:
tabix https://jbrowse.org/demos/hprc/hprc-v2.1-mc-grch38.bubbles.bed.gz \
'GRCh38#0#chr1:103,690,000-103,780,000' | cut -f1-8 | head -1
# GRCh38#0#chr1 103611080 103732636 95 269401 1 26889 316616The columns after the span are segments, paths, the inversion flag, and the
lengths of the shortest and longest allele the bubble holds. Length stands in
for copy number here, because gfatools bubble and the rGFA tags record the
distinct sequence a bubble can hold. The .gbz carries one walk per haplotype,
so it holds a count per haplotype, and
reading that count is a query.
Reproduce it end to end
The track above reads two files we built from HPRC's
sv.gfa.gz. One command over any rGFA builds the first: gfa2bed -m reads the
SN/SO/SR tags and puts each segment at a reference coordinate.
# one row per segment: stableName, start, end, segmentId, rank
gfatools gfa2bed -m your-graph.gfa.gz | sort -k1,1 -k2,2n | bgzip > your-graph.segs.bed.gz
tabix -f -p bed your-graph.segs.bed.gzThe links index comes from the same script's second pass over the L-lines.
build_rgfa_tabix.sh
runs both, and
Preparing your own graph walks
through the script and
lists which builder each graph format needs.
Pointed at hprc-v2.1-mc-grch38.sv.gfa.gz, it writes the files we host, and
README.txt beside those files
records their provenance.
Part 2 and part 3 carry the scripts behind the rest of the release's files.
See also
- Pangenome (HPRC) part 2, who carries what
- Pangenome (HPRC) part 3, every haplotype in its own coordinates
- Pangenome (HPRC) part 4, the graph as the picture
- Pangenome (preparing your own graph)
- The HPRC pangenome on genomes.jbrowse.org
- Pangenome (Minigraph-Cactus)
- Pangenome (pggb)
- Graph genome view
References
- HPRC release 2, the release this page opens: the Minigraph-Cactus graph and the assemblies it was built from.
- Li H.
The rGFA format and
gfatools, which define the
SN/SO/SRtags this page opens the graph by.
Feedback on this tutorial is welcome: contact us.