Jupyter Widget Usage

Note: Python 2.7 is dropping support January 1, 2020. For this reason, Mango no longer supports Python 2.

The Mango widgets are Jupyter widgets built using pileup.js. The widgets support visualizations for alignments, features, variants, and genotypes in a Jupyter Notebook and Jupyter lab version >2.0.

Installation for Jupyter notebook

First, install and enable bdgenomics.mango.pileup, a Jupyter Widget:

pip install bdgenomics.mango.pileup

jupyter nbextension install --py --user bdgenomics.mango.pileup
jupyter nbextension install --py --user widgetsnbextension

jupyter nbextension enable --py --user widgetsnbextension
jupyter nbextension enable --py --user bdgenomics.mango.pileup

Note: If you are using an conda environment, install extensions using --sys-prefix:

jupyter nbextension install --py --sys-prefix bdgenomics.mango.pileup
jupyter nbextension install --py --sys-prefix widgetsnbextension

jupyter nbextension enable --py --sys-prefix widgetsnbextension
jupyter nbextension enable --py --sys-prefix bdgenomics.mango.pileup

This will install the bdgenomics.mango.pileup extension into your current conda environment.

Installation for Jupyter lab

To use the Mango widgets in Jupyter lab, you will need the following requirements:

pip install bdgenomics.mango.pileup
jupyter labextension install @jupyter-widgets/jupyterlab-manager # install the Jupyter widgets extension
jupyter labextension install bdgenomics.mango.pileup

These tutorials show how to create a Jupyter pileup.js widget. An example notebook can be found in the Mango Github repository.

Pileup Example

This example shows how to visualize alignments through a Jupyter widget.

# imports
import bdgenomics.mango.pileup as pileup
from bdgenomics.mango.pileup.track import *
import pandas as pd

# read in JSON
readsJson = pd.read_json("./data/alignments.ga4gh.chr17.1-250.json")
GA4GHAlignmentJson = readsJson.to_json()

# make pileup track
tracks=[Track(viz="pileup", label="my Reads", source=pileup.sources.GA4GHAlignmentJson(GA4GHAlignmentJson))]

# render tracks in widget
reads = pileup.PileupViewer(chrom="chr17", start=1, stop=100, reference="hg19", tracks=tracks)
reads
../_images/pileupWidget.png

Variant Example

This example shows how to visualize variants through a Jupyter widget.

# make variant track
tracks=[Track(viz="variants", label="my Variants", source=pileup.sources.VcfDataSource("<path_to_file>/my_vcf.vcf"))]

# render tracks in widget
variants = pileup.PileupViewer(chrom="chr1", start=10436, stop=10564, reference="hg19", tracks=tracks)
variants
../_images/variantWidget.png

Feature Example

This example shows how to visualize features through a Jupyter widget.

featuresJson = pd.read_json("./data/features.ga4gh.chr1.120000-125000.json")
GA4GHFeatureJson = featuresJson.to_json()

# make feature track
tracks=[Track(viz="features", label="my Features", source=pileup.sources.GA4GHFeatureJson(GA4GHFeatureJson))]

# render tracks in widget
features = pileup.PileupViewer(chrom="chr1", start=120000, stop=121000, reference="hg19", tracks=tracks)
features
../_images/featureWidget.png

Genotype Example

This example shows how to visualize genotypes through a Jupyter widget.

# make genotype track
tracks=[Track(viz="genotypes", label="my Genotypes", source=pileup.sources.VcfDataSource("<path_to_file>/my_vcf.vcf"))]

# render tracks in widget
genotypes = pileup.PileupViewer(chrom="chr22", start=21340030, stop=21340150, reference="hg19", tracks=tracks)
genotypes
../_images/genotypeWidget.png