Quantitative track

BigWig and BedGraph files store genome-wide quantitative signals (read depth, ChIP-seq enrichment, conservation scores). JBrowse renders them as an XY plot, a density heatmap, a line, or a scatter plot, switchable from the track menu's Plot type submenu. One track draws one signal or a whole cohort of them.

Rendering types

The track menu's Plot type submenu (backed by the display's defaultRendering slot) offers these styles:

XY plot - filled bar chart
Density - a heatmap row, compact for many signals at once
Line (step) - the tops of the bars as a stepped line
Line (interpolated) - midpoint to midpoint, smoother for sparse signals
Scatter - individual points, for sparse data

The same BigWig rendered in every plot type at once (XY plot, Density, Line (step), Line (interpolated), and Scatter), so the styles can be compared directly. Switch a track between them from its Plot type menu. Open this view in JBrowse ↗

Score options

Plot type and Resolution are top-level track menu items; the rest of the settings below are grouped under Score.

Autoscale type

The Y-axis range (the display's scales.y.autoscale slot). All three rescale to the region in view and differ in how they treat outliers:

Local - the plain min and max, so one anomalous position flattens the rest
Local (99th percentile) - clips the outermost 1% of each sign
Local ± 3σ - three standard deviations of the local signal, a harder clip when the spikes are very tall

Summary score mode

Zoomed out, a BigWig serves precomputed summary bins, and Summary score mode picks which statistic a pixel draws: Minimum, Maximum, Average, or Whiskers (summaryScoreMode). Whiskers shows all three. An XY plot nests a darker average bar inside the lighter min-to-max range. A line plot fills min to max as a translucent band behind the average line, stepped or interpolated to match the line. Density mode draws the average, since it maps score to color.

A narrow peak fades out across a whole chromosome when averaged over a wide bin. Maximum keeps it visible.

Other score options

Scale type - linear or log Y axis
Resolution - overrides the resolution chosen from the view width
Set min/max score - pins the Y axis, for comparison across samples
Reference lines - a dashed line across the plot at each value you name, each with an optional label and colour

Colors

Edit color... opens the colour spec the config file holds. A CSS color string paints the whole plot in it; { "field": "score", "scale": "threshold" } is the two-sided plot, one colour below the cut and one above; a ramp ({ "field": "score", "scale": "linear", "scheme": "viridis" }) is what density mode fades through; and { "field": "source" } gives each signal in the track a palette entry of its own. The quantitative track configuration guide writes the same object into a config file.

Many signals in one track

A MultiQuantitativeTrack combines several quantitative signals (typically BigWig files) into one track on a shared Y axis, and opens with one row per signal. Plot type holds the five plot styles twice, once under Multi-row and once under Overlapping, so one click picks both how a signal is drawn and whether it gets a row of its own. A track with a single signal has no layout to choose and lists the five styles on their own.

The track menu lists the available plot types. Open this view in JBrowse ↗

Each row keeps the colour its subtrack was configured with. Sources sharing one plot box take a palette entry each instead, so the overlaid plots can be told apart.

An outlier on one signal can blow out the shared Y axis. The "Local ± 3σ" autoscale type clips to three standard deviations of the visible data for a more readable view, or pin the min and max from the track menu.

Twelve per-cell-type BigWigs from a 5k PBMC scATAC dataset as one multi-quantitative track, over CD8A and MS4A1 in one discontinuous view. CD8A is carried by the CD8, MAIT and NK rows and MS4A1 by the two B rows, on one shared scale. Open this view in JBrowse ↗ (large dataset — this session takes a while to load)

Adding a multi-quantitative track

Three ways to create one:

The "Add a track" form lets you paste a list of BigWig URLs, or open multiple BigWig files from your machine
The track selector lets you multi-select existing tracks and combine them into a multi-quantitative track, which is how a set of per-cell-type BigWigs becomes one stacked track (see the single-cell ATAC pseudobulk tutorial)
Hand-edit the config, described in the quantitative track configuration guide

The 'Add a track' form's workflow selector (red callout) lets you reach the multi-quantitative workflow, where you can paste a list of BigWig URLs or open multiple BigWig files from disk. Open this view in JBrowse ↗

Loading bedMethyl as a multi-quantitative track

modkit pileup produces a bedMethyl file, a tab-separated BED format where each row reports the methylation fraction at a single CpG position for one modification type (e.g. 5mC or 5hmC). It loads as BedTabixAdapter and naturally maps to MultiQuantitativeTrack, with one subtrack per modification type; see Loading bedMethyl as a multi-quantitative track for generating the file and the adapter config. For the per-read view of the same modified-base calls, see Color by base modifications on the alignments track.

Clustering rows by score

Reorder rows by signal similarity, via Clustering → Cluster rows by score... in the track menu. Auto mode samples signal values at each pixel across the visible region to build the matrix. See Clustering rows for the modes, the dendrogram, and how to share a result in a session URL.

Clustering a multi-quantitative track. Top: the 'Cluster rows by score' dialog with its auto/manual mode options. Bottom: after clustering, rows are reordered by signal similarity. Open this view in JBrowse ↗

Sorting rows by score at one position

Right-click a row at the column you want to rank on and choose Sort rows by score here. This reorders the rows by the score each carries at that base, highest at the top, so a cohort reads top-to-bottom at a candidate locus. Clustering orders the rows by the whole region in view; this orders them by a single column.

Reset row order puts the rows back in the order they were loaded in. It appears in the same right-click menu and in the track menu, and it undoes a sort, a clustering run, and a hand-arranged order alike. Where a track should open on a particular order, the facet slot's domain lists the sources that lead — the rest keep the adapter's order, and the reset returns to that order rather than past it.

A session can persist the sort with sortRowsBy, the way runClustering persists a clustering run — see LinearWiggleDisplay for both fields.

Viewing whole-genome coverage for CNV profiling

For a chromosome-scale view of copy-number changes:

Open the BigWig track
Show all regions in the assembly
Set Autoscale type to Local ± 3σ to clip outlier spikes
Increase the Resolution until the profile looks smooth
Drag the bottom edge of the track down to make it taller

Whole-genome CNV coverage profile from a BigWig file. Each chromosome is shown as a separate region; the signal represents read depth normalized by the pipeline. Copy-number gains appear as elevated signal; losses as depressed signal. Open this view in JBrowse ↗ (large dataset — this session takes a while to load)

Do it yourself

The steps behind the figure above, run against your own data.

Open your genome: on the JBrowse Desktop start screen click Open new genome (or Show all available genomes to pick a hosted one).This figure uses hg19, loaded from a bgzip-compressed FASTA (.fa.gz + .fai + .gzi).
Add your own track: File → Open track..., then paste a URL or choose a local file. This one needs a BigWig file (.bw).This figure uses “COLO829 MinION tumor (coverage)”.
Track menu → Score → Autoscale type → Local ± 3σ
Track menu → Settings → numStdDevHow many standard deviations the "Local ± σ" autoscale spans. Nothing sets it from a menu — it only reads back out, as the σ in that option's own label — so it is set on the config.
Track menu → Plot type → Scatter
Track menu → ResolutionStep with Coarser and Finer, 2× per click; the caption between them reads the current bin size. Higher fetches finer bins, and this figure uses 5×.
Track menu → Score → Summary score mode → Average
Type your region of interest into the location box and press Enter.This figure is at chr10.

In Desktop

Open this view in JBrowse Desktop ↗

Opens JBrowse Desktop (JBrowse Desktop 5.0+) at this view and saves it as a reopenable session — swap in your own files afterwards.

Nothing happens? Paste this link into Desktop's Open JBrowse Web link... (start screen, or File → Session):

https://jbrowse.org/code/jb2/main/?config=test_data%2Fconfig_demo.json&session=spec-%7B%22views%22%3A%5B%7B%22type%22%3A%22LinearGenomeView%22%2C%22assembly%22%3A%22hg19%22%2C%22loc%22%3A%22chr10%22%2C%22tracks%22%3A%5B%7B%22trackId%22%3A%22colo_tumor%22%2C%22type%22%3A%22LinearWiggleDisplay%22%2C%22scales%22%3A%7B%22y%22%3A%7B%22autoscale%22%3A%22localsd%22%2C%22numStdDev%22%3A3%7D%7D%2C%22defaultRendering%22%3A%22scatter%22%2C%22resolution%22%3A5%2C%22summaryScoreMode%22%3A%22avg%22%7D%5D%7D%5D%7D&sessionName=JBrowse+docs%3A+bigwig%2Fwhole_genome_coverage

Session spec

This session spec pastes after &session=spec- on any JBrowse Web instance.

{
  "views": [
    {
      "type": "LinearGenomeView",
      "assembly": "hg19",
      "loc": "chr10",
      "tracks": [
        {
          "trackId": "colo_tumor",
          "type": "LinearWiggleDisplay",
          "scales": {
            "y": {
              "autoscale": "localsd",
              "numStdDev": 3
            }
          },
          "defaultRendering": "scatter",
          "resolution": 5,
          "summaryScoreMode": "avg"
        }
      ]
    }
  ]
}

Loaded against config: test_data/config_demo.json

In a notebook

The same view with jbrowse-anywidget, from the config the figure loads. Swap a track's uri for your own file.

from jbrowse_anywidget import JBrowseApp, fetch_hub

config = fetch_hub("https://jbrowse.org/code/jb2/main/test_data/config_demo.json")
app = JBrowseApp(
    assemblies=config["assemblies"],
    tracks=config.get("tracks", []),
    views=[
        {
            "type": "LinearGenomeView",
            "assembly": "hg19",
            "loc": "chr10",
            "tracks": [
                {
                    "trackId": "colo_tumor",
                    "type": "LinearWiggleDisplay",
                    "scales": {
                        "y": {
                            "autoscale": "localsd",
                            "numStdDev": 3
                        }
                    },
                    "defaultRendering": "scatter",
                    "resolution": 5,
                    "summaryScoreMode": "avg"
                }
            ]
        }
    ],
)
app  # display the widget

With an agent

Hand this to a coding agent. It rebuilds the figure above headlessly, and the session file is then the thing to edit — swap an adapter uri for your own file and rerun.

cat > session.json <<'JSON'
{
  "views": [
    {
      "type": "LinearGenomeView",
      "assembly": "hg19",
      "loc": "chr10",
      "tracks": [
        {
          "trackId": "colo_tumor",
          "type": "LinearWiggleDisplay",
          "scales": {
            "y": {
              "autoscale": "localsd",
              "numStdDev": 3
            }
          },
          "defaultRendering": "scatter",
          "resolution": 5,
          "summaryScoreMode": "avg"
        }
      ]
    }
  ]
}
JSON

npx @jbrowse/capture --instance https://jbrowse.org/code/jb2/main/ \
  --config https://jbrowse.org/code/jb2/main/test_data/config_demo.json \
  --session session.json -o figure.png

Using JBrowse with AI agents covers the rest: hosted genomes to point it at, and why waiting for the render is the part that goes wrong.

For tumor vs normal on one Y axis, put both signals in one track; for a whole cohort, the TCGA cohort copy number tutorial.

Coverage is also shaped by GC content, mappability, repeats and PCR bias, so not every dip or spike is a copy-number change.

Quantitative track

Rendering types#

Score options#

Autoscale type#

Summary score mode#

Other score options#

Colors#

Many signals in one track#

Adding a multi-quantitative track#

Loading bedMethyl as a multi-quantitative track#

Clustering rows by score#

Sorting rows by score at one position#

Viewing whole-genome coverage for CNV profiling#

See also#