Mango is a scalable genomics visualization tool built on top of the ADAM genomics processing engine. Apache 2 licensed.
Mango consists of a notebook and browser form factor, allowing users to visualize reads, variants, and features in a GUI or programmable interface. The Mango tools use Pileup.js for interactive scrolling at genomic loci.
Mango documentation is hosted at readthedocs. Documentation for Mango includes instructions for running the Mango notebook and the Mango browser locally, remotely, and in the cloud (Amazon EMR and Google Cloud Engine). Documentation also provides Python API documentation for the Mango notebook.
The Mango tools are published to Maven central. Corresponding python modules are published to Pypi. Readthedocs provides instructions on how to install the Mango tools from the most recent distribution.
Instructions for running the Mango tools from source can be found at readthedocs.
You will need to have Maven installed in order to build mango. Mango browser also requires npm > 3.10.10.
Note: The default configuration is for Hadoop 2.7.3. If building against a different version of Hadoop, please edit the build configuration in the
<properties>section of thepom.xmlfile.
$ git clone https://github.com/bigdatagenomics/mango.git
$ cd mango
$ mvn clean package -DskipTests
If using the Mango notebook, we recommend setting up a virtual environment to install required python modules.
To configure your python environment for the Mango notebook, refer to instructions for Building for Python.
The Mango notebook is a set of Python APIs and Jupyter widgets for loading and manipulating genomic data in a Jupyter notebook environment. The Mango APIs can be used for loading and visualizing raw features, variants and alignments, as well as calculating and viewing aggregate information.
Mango can also be run through the notebook form.
./bin/mango-notebook
In the jupyter UI, navigate to example-files/notebooks to view example notebooks.
Stated above, the Mango notebook provides python APIs for loading and visualizing genomic data in a Jupyter notebook environment. API information for the Mango notebook can be found in readthedocs.
The Mango Documentation includes instructions for running the Mango notebook on Amazon AWS. These instructions can be leveraged to visualize genomic datasets staged in Amazon S3, and include example notebooks for exploring the 1000 Genomes Dataset on Amazon S3. If using the Mango Docker instances, an example notebook can be found on the GitHub.
The Mango Documentation includes instructions for running the Mango notebook on Google Cloud Engine. These instructions can be leveraged to visualize genomic datasets staged publically. If using the Mango Docker instances, an example notebook can be found on the GitHub.
The Mango browser provides a GUI to visualize genomic data stored remotely or in the cloud.
The Mango browser uses IntervalRDDs to perform fast indexed lookups on interval-keyed data.
mango is packaged via appassembler and includes all necessary dependencies.
The Mango repository includes example scripts to run the Mango browser on small example files.
To run the example files in the Mango browser, run:
./example-files/browser-scripts/run-example.sh
to see a demonstration of chromosome 17, region 7500000-7515000.
Now view the mango genomics browser at localhost:8080 or the port specified:
View the visualization at: 8080
Quit at: /quit
For help launching the script, run bin/mango-submit -h
$ bin/mango-submit -h
Using SPARK_SUBMIT=/Applications/spark-1.6.1-bin-hadoop2.4/bin/spark-submit
reference : The reference file to view, required
-cacheSize N : Bp to cache on driver.
-coverage VAL : A list of coverage files to view, separated by commas (,)
-discover : This turns on discovery mode on start up.
-features VAL : The feature files to view, separated by commas (,)
-genes VAL : Gene URL.
-h (-help, --help, -?) : Print help
-parquet_block_size N : Parquet block size (default = 128mb)
-parquet_compression_codec [UNCOMPRESSED | SNAPPY | GZIP | LZO] : Parquet compression codec
-parquet_disable_dictionary : Disable dictionary encoding
-parquet_logging_level VAL : Parquet logging level (default = severe)
-parquet_page_size N : Parquet page size (default = 1mb)
-port N : The port to bind to for visualization. The default is 8080.
-prefetchSize N : Bp to prefetch in executors.
-preload VAL : Chromosomes to prefetch, separated by commas (,).
-print_metrics : Print metrics to the log on completion
-reads VAL : A list of reads files to view, separated by commas (,)
-show_genotypes : Shows genotypes if available in variant files.
-test : For debugging purposes.
-variants VAL : A list of variants files to view, separated by commas (,). Vcf files require a
corresponding tbi index.
The Mango Documentation includes instructions for running the Mango browser on Amazon AWS. These instructions can be leveraged to visualize genomic datasets staged in Amazon S3.
The Mango Widgets are standalone Jupyter widgets for interacting with Pileup.js in a Jupyter notebook environment. API documentation for the Mango widgets can be found in readthedocs.