and container friends!
This repository implements the SnakeMake tutorial workflow and uses the Scientific Filesystem (SCIF) to provide a reproducible research environment. It has been extended to not just include Docker and Singularity, but (as many container technologies as @vsoch can figure out) to demonstrate the variable use for SCIF. To get started, clone this repository and build the containers per the instructions below.
The following container technologies don't have user friendly build clients available (but are used by admins successfully in their respective locations).
- Inception: I couldn't figure out how to use it, my issue is posted here.
We will take the following format in each section, to show the same command for each container technology, followed by the output that each results in.
This repository contains the following:
- README.md: is what you are reading, which has the complete walk through of building and running each container. This is the intended use case for this work, as an interactive exercise.
- Container Build Recipes: Singularity and Dockerfile are build recipes for Docker and Singularity containers.
- Scientific Applications: are installed in each container using the same scientific filesystem recipe, snakemake_tutorial.scif
- Workflow Is provided by using Snakemake. There is a Snakefile to define rules and steps, and a small config.yaml file to define sample inputs in the data folder.
- configs is a folder of configuration files for runc, one of the container technologies.
- run_comparison.sh will run the pipeline for all but Shifter, which needs a Shifter box. This assumes the software for each container is installed and containers are already built, unless the variable
DOBUILDis set to 1. The results for each run are moved to results. - measure_comparison.sh uses the time command and valgrind to collect runtime metrics, from within the containers.
- For Shifter we (manually) went through [run_shifter.sh] to collect both metrics and running because it required running from inside a shifter box.
- asciinema is a totally unecessary folder of asciinemas recorded and text to cat to the screen to make it fun :)
It is recommended to approach this learning exercise reading through the seconds below.
Singularity
sudo singularity build snakemake.simg Singularity
Docker
docker build -t vanessa/snakemake.scif .
Charliecloud
ch-build -t snakemake.ch .
ch-docker2tar snakemake.ch /var/tmp
ch-tar2dir /var/tmp/snakemake.ch.tar.gz /var/tmp
runc
mkdir -p /tmp/runc/rootfs
sudo singularity build --sandbox /tmp/runc/rootfs snakemake
# I'm not sure if this is needed, but I did it.
sudo chown -R vanessa /tmp/runc/rootfs
cd /tmp/runc
runc spec --rootless --bundle /tmp/runc
ls /tmp/runc
config.json rootfs
At this point we need to customize the config.json to allow for specific mounts,
and add the /opt/conda/bin to the $PATH to find the scif executable.
"env": ["PATH=/opt/conda/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin",
"TERM=xterm"
],
...
{
"destination": "/scif/data",
"type": "bind",
"options": ["bind"],
"source": "/home/vanessa/Documents/Dropbox/Code/scif/examples/snakemake.scif/data"
},
Shifter
The user should know this is the "shifter in a box" which is likely much harder to use than a cluster installation. I don't have shifter at my insitution so this was the approach I chose to test. Also note that for this example we assume the Docker box to be our host, and so we clone the repository into it (as opposed to a double bind from host --> Shifter box --> shifter container).
docker run -it --rm --privileged scanon/shifterbox:2
# We need git to clone the repo!
yum install -y git
We can then start the gateway and database (I believe it is MongoDB). This works for me about half the time, and usually on the second go. But you can try again.
/src/start.sh
su - auser
Now we can pull the container that has our analysis stuffs in it:
shifterimg pull vanessa/snakemake.scif
If you get an error about contacting the gateway, it's a race condition likely, and you should exit the user, as the superuser remove the lock for MongoDB, and try again:
$ exit
$ rm -rf /data/db/
$ /src/start.sh
su - auser
shifterimg pull vanessa/snakemake.scif
After you use shifterimg pull and have the image, also as the user,
let's get our data (it's the same repository, but we want it inside the shifter
box to mimic binding it to the container).
su - auser
git clone https://github.com/sci-f/snakemake.scif
cd snakemake.scif/
If you intend to run / test the shifter box, leave it open! We will continue these steps in the sections below with the pulled container.
The interesting thing about the above is that we are building many of the containers by way of starting with Docker layers. This reinforces the idea that the underlying guts of a container are the file contents. Docker delivers the contents nicely in layers, so it's a starting point to dump them into another base, and then interact. Do all roads lead to, or start from Docker layers? Is the power in the allmighty tarball? I'll let you decided.
For each of the examples below, we will show examples running commands inside and outside of the containers.
The Snakefile specifies rules that should be executed. State a target file and Snakemake will build a directed acyclic graph (DAG) from the rules until the target file is reached. The rules are then executed to create the target file. The Snakefile hides the details of the environment provided by SCIF in the container as you can see from the commands below. All commands can be executed from outside the container or from within the container.
In all cases we bind the example dataset to the data directory provided by SCIF. SCIF provides the environment in which the workflow steps are executed. The workflow steps themself and therefore the interaction with SCIF is implemented in the Snakefile.
Inside container, Singularity
singularity shell --bind data/:/scif/data snakemake.simg
$ scif run snakemake all
Inside container, Docker
To run the whole workflow we need the Snakefile which is not part of the container. To access the Snakefile from within the Docker container, we need an additional mount.
docker run -v $PWD/data:/scif/data:z -v $PWD:/working_dir -it -w /working_dir --entrypoint /bin/bash vanessa/snakemake.scif
$ scif run snakemake all
Inside container, Charliecloud
ch-run --cd $PWD -b data:/scif/data /var/tmp/snakemake.ch -- /bin/bash
export PATH=/opt/conda/bin:$PATH
$ scif run snakemake all
Inside container, runc
The complete config file can be seen in [configs/runc-config-interactive.json]. This interactive version uses a shell entrypoint, and then runs scif. If the Snakemake and config.yaml aren't found in the data directory, they are copied there from the container snakemake application.
runc --root /tmp/runc run --bundle /tmp/runc snakmake.runc
# which scif
/opt/conda/bin/scif
# scif run snakemake all
[snakemake] executing /bin/bash /scif/apps/snakemake/scif/runscript all
...
Complete log: /scif/data/.snakemake/log/2018-02-25T180544.725734.snakemake.log
Inside container, Shifter
Note that we are working in a container that has had the build step done (as shown above to pull) and we have cloned our snakemake repo.
$ echo $PWD
/home/auser/snakemake.scif
shifter --image=vanessa/snakemake.scif --volume `pwd`/data:/scif/data bash
$ scif run snakemake all
Outside Container, Singularity
singularity run --bind data/:/scif/data snakemake.simg run snakemake all
Outside Container, Docker
Because Docker does not mount the current working directory, we mount it to a the SCIF data directory of the SCIF snakemake app. The snakemake app will execute snakemake in /scif/data/snakemake, thus snakemake can find the Snakefile.
docker run -v $PWD/data:/scif/data:z -v $PWD:/scif/data/snakemake:z -it vanessa/snakemake.scif run snakemake all
Outside Container, Charliecloud
ch-run --cd $PWD -b data:/scif/data /var/tmp/snakemake.ch -- /opt/conda/bin/scif run snakemake all
Outside container, runc
This is a very easy switch to do! All that is needed is to define the entrypoint to the container via the configuration file to be the scif entrypoint, and we can even "hard code" the application to run as snakemake. This modified config file can be found in [configs/runc-config-noninteractive.json].
runc --root /tmp/runc run --bundle /tmp/runc snakmake.runc
[snakemake] executing /bin/bash /scif/apps/snakemake/scif/runscript all
...
Complete log: /scif/data/.snakemake/log/2018-02-25T180544.725734.snakemake.log
So it comes down to changing:
"args": [
"sh"
],
to
"args": [
"scif", "run", "snakemake", "all"
],
I can imagine that I would have some strategy to either manage a small collection of config files to use per rootfs, or a method to produce the one I need on demand. I suspect this is what many popular container technologies are already doing and I just haven't realized it :)
Outside container, Shifter
Note that we are working in a container that has had the build step done (as shown above to pull and clone the repository)
$ shifter --image=vanessa/snakemake.scif --volume `pwd`/data:/scif/data --workdir /scif/data /opt/conda/bin/scif run snakemake all
The whole workflow should finish within 5 seconds ... too fast to realize what is happening. A visualization of the whole workflow we just ran would be nice ...
The SCIF app graphviz_create_dag generates a directed acyclic graph of the workflow execution plan. This is a feature of Snakemake, but needs Graphviz as additional dependency. Therefore Graphviz is installed in the appinstall section of this app.
rm -r data/calls/ /data/mapped_reads/ data/sorted_reads/ data/report.html
Now we can generate the graph of the execution plan. The plan is computed from the Snakefile in our current working directory and the target file we specify. In this case we want to create the final report file
report.html. The directed acyclic graph of the workflow should be saved atdata/dag.svg. The commands for Singularity and Docker are as follows:
Singularity
singularity run --bind data/:/scif/data snakemake.simg run graphviz_create_dag $PWD report.html dag.svg
Docker
In contrast to Singularity, Docker cannot access the current working directory. Therefore we mount the current working directory to the SCIF data folder of the graphviz_create_dag app.
docker run -v $PWD/data:/scif/data:z -v $PWD:/scif/data/graphviz_create_dag:z -it vanessa/snakemake.scif run graphviz_create_dag /scif/data/graphviz_create_dag report.html dag.svg
Charliecloud
ch-run --cd $PWD -b data:/scif/data /var/tmp/snakemake.ch -- /opt/conda/bin/scif run graphviz_create_dag $PWD report.html dag.svg
runc This is the same deal as before, we are just going to use a config to specify a different entry point and arguments to generate the dag. Since you've done this twice, I'll just show you the modified command:
"args": [
"scif", "run", "graphviz_create_dag", "/scif/data", "report.html","dag.svg"
],
and then run! Note that if you do this without changing permission, you would get an error:
process:10): Pango-WARNING **: error opening config file '/root/.config/pango/pangorc': Permission denied
and this would need to be resolved by specifying a different file, or changing the permission of the file from the initial build.
runc --root /tmp/runc run --bundle /tmp/runc snakmake.runc
[graphviz_create_dag] executing /bin/bash /scif/apps/graphviz_create_dag/scif/runscript /scif/data report.html dag.svg
Shifter
shifter --image=vanessa/snakemake.scif --volume `pwd`/data:/scif/data --workdir /scif/data /opt/conda/bin/scif run graphviz_create_dag `pwd` report.html dag.svg
> [graphviz_create_dag] executing /bin/bash /scif/apps/graphviz_create_dag/scif/runscript /home/fbartusch/github/snakemake_tutorial report.html dag.svg
> Building DAG of jobs...
Instead of running the whole workflow at once with Snakemake, we can also use the SCIF environment to execute computations. The following examples also illustrate special SCIF syntax for working with environment variables or piping output to a file. First, we map the reads with bwa mem to the reference genome.
Inside container, Singularity
Note the use of [e] so we can easily pass the environment variable to the SCIF. Otherwise, it would be evaluated on the host.
singularity shell --bind data/:/scif/data snakemake.simg
mkdir /scif/data/mapped_reads
scif run bwa mem -o [e]SCIF_DATA/mapped_reads/A.sam [e]SCIF_DATA/genome.fa [e]SCIF_DATA/samples/A.fastq
Inside container, Docker
docker run -v $PWD/data:/scif/data:z -it --entrypoint /bin/bash vanessa/snakemake.scif
mkdir -p /scif/data/mapped_reads
scif run bwa mem -o [e]SCIF_DATA/mapped_reads/A.sam [e]SCIF_DATA/genome.fa [e]SCIF_DATA/samples/A.fastq
Inside container, Charliecloud
ch-run --cd $PWD -b data:/scif/data /var/tmp/snakemake.ch -- /opt/conda/bin/scif shell
/opt/conda/bin/scif run bwa mem -o [e]SCIF_DATA/mapped_reads/A.sam [e]SCIF_DATA/genome.fa [e]SCIF_DATA/samples/A.fastq
Inside container, runc Here we are shelling inside, so we are using the first config.json with args as a list with "sh".
runc --root /tmp/runc run --bundle /tmp/runc snakmake.runc
scif run bwa mem -o [e]SCIF_DATA/mapped_reads/A.sam [e]SCIF_DATA/genome.fa [e]SCIF_DATA/samples/A.fastq
Inside container, Shifter
We are still working in a container that has had the build step done:
$ echo $PWD
/home/auser/snakemake.scif
shifter --image=vanessa/snakemake.scif --volume `pwd`/data:/scif/data bash
$ scif run bwa mem -o [e]SCIF_DATA/mapped_reads/A.sam [e]SCIF_DATA/genome.fa [e]SCIF_DATA/samples/A.fastq
Outside container, Singularity
singularity run --bind data:/scif/data snakemake.simg run bwa mem -o [e]SCIF_DATA/mapped_reads/A.sam [e]SCIF_DATA/genome.fa [e]SCIF_DATA/samples/A.fastq
Outside the container, Docker
docker run -v $PWD/data:/scif/data:z vanessa/snakemake.scif run bwa mem -o [e]SCIF_DATA/mapped_reads/A.sam [e]SCIF_DATA/genome.fa [e]SCIF_DATA/samples/A.fastq
Outside the container, Charliecloud
ch-run --cd $PWD -b data:/scif/data /var/tmp/snakemake.ch -- /opt/conda/bin/scif run bwa mem -o [e]SCIF_DATA/mapped_reads/A.sam [e]SCIF_DATA/genome.fa [e]SCIF_DATA/samples/A.fastq
Outside container, runc
The modified "args" in the config.json is:
"args": [
"scif", "run", "bwa", "mem", "-o", "[e]SCIF_DATA/mapped_reads/A.sam", "[e]SCIF_DATA/genome.fa", "[e]SCIF_DATA/samples/A.fastq"
],
and then run!
runc --root /tmp/runc run --bundle /tmp/runc snakmake.runc
Outside container, Shifter
$ shifter --image=vanessa/snakemake.scif --volume `pwd`/data:/scif/data --workdir /scif/data /opt/conda/bin/scif run bwa mem -o [e]SCIF_DATA/mapped_reads/A.sam [e]SCIF_DATA/genome.fa [e]SCIF_DATA/samples/A.fastq
Inside the container, Singularity
Note the use of [out] as a substitute for >. If you wanted to use > you could put the entire thing in quotes.
singularity shell --bind data/:/scif/data snakemake.simg
scif run samtools view -Sb [e]SCIF_DATA/mapped_reads/A.sam [out] [e]SCIF_DATA/mapped_reads/A.bam # or
scif run samtools 'view -Sb $SCIF_DATA/mapped_reads/A.sam > $SCIF_DATA/mapped_reads/A.bam'
> [samtools] executing /bin/bash /scif/apps/samtools/scif/runscript view -Sb $SCIF_DATA/mapped_reads/A.sam > $SCIF_DATA/mapped_reads/A.bam
Inside the container, Docker
docker run -v $PWD/data:/scif/data:z -it --entrypoint /bin/bash vanessa/snakemake.scif
scif run samtools view -Sb [e]SCIF_DATA/mapped_reads/A.sam [out] [e]SCIF_DATA/mapped_reads/A.bam # or
scif run samtools 'view -Sb $SCIF_DATA/mapped_reads/A.sam > $SCIF_DATA/mapped_reads/A.bam'
> [samtools] executing /bin/bash /scif/apps/samtools/scif/runscript view -Sb $SCIF_DATA/mapped_reads/A.sam > $SCIF_DATA/mapped_reads/A.bam
Inside the container, Charliecloud
ch-run --cd $PWD -b data:/scif/data /var/tmp/snakemake.ch -- /opt/conda/bin/scif shell
/opt/conda/bin/scif run samtools view -Sb [e]SCIF_DATA/mapped_reads/A.sam [out] [e]SCIF_DATA/mapped_reads/A.bam
Inside container, runc Here we are again shelling inside, so we are using the first config.json with args as a list with "sh".
runc --root /tmp/runc run --bundle /tmp/runc snakmake.runc
scif run samtools view -Sb [e]SCIF_DATA/mapped_reads/A.sam [out] [e]SCIF_DATA/mapped_reads/A.bam
Inside container, Shifter
We are still working in a container that has had the build step done:
$ echo $PWD
/home/auser/snakemake.scif
shifter --image=vanessa/snakemake.scif --volume `pwd`/data:/scif/data bash
$ scif run samtools view -Sb [e]SCIF_DATA/mapped_reads/A.sam [out] [e]SCIF_DATA/mapped_reads/A.bam
Outside the container, Singularity
singularity run --bind data:/scif/data snakemake.simg run samtools view -Sb [e]SCIF_DATA/mapped_reads/A.sam [out] [e]SCIF_DATA/mapped_reads/A.bam # or
singularity run --bind data:/scif/data snakemake.simg run samtools 'view -Sb $SCIF_DATA/mapped_reads/A.sam > $SCIF_DATA/mapped_reads/A.bam'
Outside the container, Docker
docker run -v $PWD/data:/scif/data vanessa/snakemake.scif run samtools view -Sb [e]SCIF_DATA/mapped_reads/A.sam [out] [e]SCIF_DATA/mapped_reads/A.bam # or
docker run -v $PWD/data:/scif/data vanessa/snakemake.scif run samtools 'view -Sb $SCIF_DATA/mapped_reads/A.sam > $SCIF_DATA/mapped_reads/A.bam'
Outside the container, Charliecloud
ch-run --cd $PWD -b data:/scif/data /var/tmp/snakemake.ch -- /opt/conda/bin/scif run samtools view -Sb [e]SCIF_DATA/mapped_reads/A.sam [out] [e]SCIF_DATA/mapped_reads/A.bam
Outside container, runc Change the config.json args to:
"args": [
"scif", "run", "samtools", "view", "-Sb", "[e]SCIF_DATA/mapped_reads/A.sam", "[out]", "[e]SCIF_DATA/mapped_reads/A.bam"
],
and then run
runc --root /tmp/runc run --bundle /tmp/runc snakmake.runc
Outside container, Shifter
We are still working in a container that has had the build step done:
$ shifter --image=vanessa/snakemake.scif --volume `pwd`/data:/scif/data --workdir /scif/data /opt/conda/bin/scif run samtools view -Sb [e]SCIF_DATA/mapped_reads/A.sam [out] [e]SCIF_DATA/mapped_reads/A.bam
This can be done for Docker, Singularity, Charliecloud, or runc, just with different commands to shell into the container. Note that for runc, the args in the config.json needs to be:
"args": [
"scif", "pyshell"
],
docker run -it -v $PWD/data:/scif/data:z vanessa/snakemake.scif pyshell
singularity run --bind data/:/scif/data snakemake.simg pyshell
ch-run /var/tmp/snakemake.ch -- /opt/conda/bin/scif pyshell
runc --root /tmp/runc run --bundle /tmp/runc snakmake.runc
shifter --image=vanessa/snakemake.scif --volume `pwd`/data:/scif/data scif pyshell
Found configurations for 4 scif apps
bwa
graphviz_create_dag
samtools
snakemake
[scif] /scif bwa | graphviz_create_dag | samtools | snakemake
Python 3.6.2 |Anaconda, Inc.| (default, Sep 22 2017, 02:03:08)
[GCC 7.2.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
(InteractiveConsole)
>>>
# Activate bwa
client.activate('samtools')
# Environment variables active
client.environment
# Run bwa interactively
args = ['mem', '-o', '[e]SCIF_DATA/mapped_reads/a.sam', '[e]SCIF_DATA/genome.fa', '[e]SCIF_DATA/samples/A.fastq']
client.run('bwa', args=args)
# Run sam--bam interactively
args = ["view", "-Sb", "/scif/data/mapped_reads/r1_subset.sam", ">", "/scif/data/mapped_reads/r1_subset.bam"]
client.run('samtools', args=args)
This repo is also used as example on the SCIF GitHub repo, so take a look:
<script src="https://code.jquery.com/jquery-3.3.1.slim.min.js"></script> <script src="docs/living-doc.js"></script>