tengfei/bamstats repository overview

⁠Reason to use public docker registry

• Tons of published thrid party apps developed and maintained by the author or offiical organization directly
• Global mirror
• "The place" to develop, maintain and/or publish your script and work officially
• Could be pulled and used on many platform directly
• Better integration with other environment, e.g. your github account.

⁠Prerequisite

• User have dockerhub account
• (optional) user have github account
• (optinoal) User have local docker client for testing purpose

⁠Case 1: Use existing public docker image that contains the tools you need

1. Use dockerhub⁠ for example, search for keyword like "fastqc"
2. Find a source you can trust: official build and maintained by author, or third repos, for example, you can find official GATK 4.1 docker image broadinstitute/gatk:4.1.4.0 on dockerhub or ENCODE chip-seq container at quay.io/encode-dcc/chip-seq-pipeline:v1.3.6 at quay.
3. For testing purpose, we can use biocontainers/fastqc:v0.11.8dfsg-2-deb_cv1 which read a fastq and output qc report.

⁠Case 2: Build your own tool quickly based on official or good quality docker image

1. Choose base image: for example "ubuntu:18.04" or "bioconda/bioconda-utils-build-env"
2. Create Dockerfile to explain how the image will be created (bunch of bash script)

• Install third party app directly following their manual (samtool example⁠ or more complicated arcasHLA example⁠)
• Or add your own script to your docker image (IEHC WGBS bamstats.py exmaple⁠ )

1. (optional) Build docker image locally and test the image make sure you can execute the script

• code example to build

docker build --tag bamstats:1.0 .

• code example to run your docker image interactively

docker run -ti bioconda/bioconda-utils-build-env /bin/bash

1. Push to dockerhub

• Method 1 (autobuild)：Link your github repo that contaning dockerfile for autobuild on dockerhub (official tutorial⁠) you can find test example here linked on docker hub⁠ for autobuild
• Method 2 (local): use docker client command line to push it manually. Example

docker push bamstats:1.0