The Nextflow pipeline relies on the Docker containers in this repo to run its
processes. This repo houses the Dockerfiles that creates the Docker containers containing the bioinformatics tools required by the Nextflow scripts, e.g. the preeclampsia Nextflow script can be run with Docker containers, see Mirvie's nf-cfrna Repo).

Docker must be installed. Run docker --version to see if Docker is already installed. If not, follow the instructions at https://docs.docker.com/get-docker/ to install it for your OS.

There are 11 Docker containers required by the nf-cfrna Nextflow pipeline.

One Docker container is required by the BCLConvert Nextflow pipeline.

The Dockerfiles need to be constructed into images before containers can be run. There are two options for building the images in this repo.

The build script will build all the containers needed for the nf-crna pipeline and mirbclconvert, starting first with mirbase from which all other images are built. The build script will first remove images that match the name and tag prior to build images tagged by the commit of this repo.

NOTES:

• mirrseqc requires a significant amount of memory to build, so set the memory requirements for the docker engine to 4.0 GB of memory.
• mirstar will need 32 GB (> 16 GB) to run in the pipeline.

1. Clone this repo.

git clone https://gitlab.com/Mirvie/docker-bioinfo-tools.git

2. Run the build script, which will delete the latest image if it exists prior to building the new image.

sh build.sh

3. Check that the images are created with the correct tags pulled from the commit.

$ docker image ls
REPOSITORY       TAG              IMAGE ID       CREATED        SIZE
mirmultiqc       f232966          18b9b1d3bb72   2 days ago     2.34GB
mirtrimmomatic   f232966          8674feb4ef6c   2 days ago     2.1GB
mirstar          f232966          09899c2c209b   2 days ago     2.03GB
mirsamtools      f232966          ff4ca83b10d3   2 days ago     2.01GB
mirrseqc         f232966          1d5daee4c9a8   2 days ago     2.48GB
mirpicard        f232966          950734544895   2 days ago     2.79GB
mirhtseq         f232966          77aad0acbd29   2 days ago     3.18GB
mirfastqc        f232966          02bb7fc839a8   2 days ago     2.3GB
mirpandas        f232966          f8b2d0263755   2 days ago     2.24GB
mircheckfastq    f232966          bc64d4ee8e90   2 days ago     3.13GB
mirbase          f232966          645458a5771d   2 days ago     1.13GB

1. The first image to build is the base image mirbase from which the other images build from. Replace with the desired tag name, e.g. the commit tag or latest:

% cd docker-bioinfo-tools/mirbase
% docker build -t mirbase:<tag> - < Dockerfile

2. Confirm that the mirbase image was created:

 % docker image ls
REPOSITORY          TAG       IMAGE ID       CREATED          SIZE
mirbase             latest    781f7ea3283c   17 seconds ago   791MB

If using Docker Desktop on a Mac, the images can also be see from the Images on disk page accessed by navigating via Images from the sidebar.

3. Build the remaining images by navigating to the appropriate image_name folder in this repo and building the Dockerfile:

% cd docker-bioinfo-tools/<image_name>
% docker build -t <image_name>:latest - < Dockerfile

If there are local files that need to be copied into the container, these files will need to be in the same dir as the Dockerfile and you will need to pass the full path of the enclosing dir to docker build:

% docker build -t <image_name>:latest /PATH/TO/DOCKERFILE/PARENT/DIR

1. To run a container in interactive mode, stepping into a bash shell, execute the following command:

docker run -it <image_name> bash

Alternatively in Docker Desktop, click the Run button next to each image via the Images on disk page.

2. Confirm the image (example here is with the mirsamtools image) is running:

% docker ps
CONTAINER ID   IMAGE                COMMAND       CREATED              STATUS              PORTS     NAMES
372065f80e4a   mirsamtools:latest   "/bin/bash"   About a minute ago   Up About a minute             exciting_greider

Using any user that has docker run privileges, run pytest from the root of the repository to run all the tests; test results are output to stdout based on the test config settings in pytest.ini. If you don't have pytest installed you can create a conda environment by running conda create -n pytestenv pytest and then activating it with conda activate pytestenv.

After running pytest to confirm all tests pass, to append the results of the tests to the clinical fetal sex test log on marge run pytest >> /mnt/storage/clinical/fetal_sex/test/docker-bioinfo-tools_test_log.txt.