Skip to content

Latest commit

 

History

History
95 lines (67 loc) · 3.43 KB

README.md

File metadata and controls

95 lines (67 loc) · 3.43 KB
Raw

pipeline

A pipeline combining odgi - component_segmentation - Schematize on Docker image or CWL

Installation

Docker is needed before running.

git clone https://github.com/graph-genome/pipeline
cd pipeline
docker build -t pipeline .

Usage

Running on CWL on example data

pip install arvados-cwl-runner

cwltool --cachedir $PWD/cache --parallel graph-genome-previz.cwl example_plain.yml
# for local execution

# or
arvados-cwl-runner graph-genome-previz.cwl example_arvados.yml

Running on Docker

Suppose that the input file is "data.gfa".

cp /pass/to/your/data.gfa .
docker run -ti --rm --publish=3000:3000 --volume=`pwd`:/usr/src/app/data pipeline data/data.gfa
docker run -ti --rm --publish=3000:3000 --volume=`pwd`:/usr/src/app/data pipeline data/data.gfa -w 10000 
  # With -w argument you can change the bin width.
docker run -ti --rm --publish=3000:3000 --volume=`pwd`:/usr/src/app/data pipeline data/data.gfa -w 10000 -s Sn
  # With -s argument you can change the sort option.

Access to http://localhost:3000/. The production build of Schematize is running.

Running PathIndex Server

Pathindex server works on the same container of Schematize at port 3010. Users need to specify the host of the server.

docker run -ti --rm \
  --publish=3000:3000 \ # For Schematize server 
  --publish=3010:3010 \ # For odgi server (*)
  --volume=`pwd`:/usr/src/app/data pipeline data/data.gfa -w 10000 -s Sn \
  --port 3010 \ # The host's port to expose the odgi server, the same as the host port of (*).
  --host localhost # The host name to expose the odgi server.

If you change the server to example.com:3020 to expose odgi server, then

docker run -ti --rm \
  --publish=3000:3000 \ # For Schematize server
  --publish=3020:3010 \ # For odgi server (*)
  --volume=`pwd`:/usr/src/app/data pipeline data/data.gfa -w 10000 -s Sn \
  --port 3020 \ # The host's port to expose the odgi server, the same as the host port of (*). 
  --host "example.com" # The host name to expose the odgi server.

Customization

You can change the options on odgi / Schematize.

  • gfa name (first argument, mandatory)
  • -w: the bin width on odgi (optional, default: 1000)
  • -s: the sort option on odgi sort (optional, default: bSnSnS)
  • -t: the threads option on odgi (optional, default: 12)
  • -c: the cells-per-file option on component_segmentation (optional)
  • -i: the host of odgi index (optional, default: localhost)

The full list of the argument is as follows:

docker run -ti --rm --publish=3000:3000 --volume=`pwd`:/usr/src/app/data pipeline -h

Support Development

git clone https://github.com/graph-genome/component_segmentation # For debugging component_segmentation
git clone https://github.com/graph-genome/Schematize # For debugging Schematize
docker run -d --publish=3000:3000 --publish=3010:3010 --volume=`pwd`:/usr/src/app/data --volume=`pwd`/Schematize:/usr/src/app/Schematize --volume=`pwd`/component_segmentation:/usr/src/app/component_segmentation pipeline data/data.gfa -w 1000 -s s -c 10000

Then, the pipeline is running through cloned component_segmentation and Schematize. Docker container is failed, but the output json file is stored on Schematize directory. Therefore just run yarn start on Schematize directory works.