Setup

Learning outcomes

After having completed this chapter you will be able to:

• Use the command line to:
  • Make a directory
  • Change file permissions to ‘executable’
  • Run a bash script
  • Pipe data from and to a file or other executable
• Program a loop in bash

Cloud serverDockerconda

Exercises

First login

If you are participating in this course with a teacher, you have received a link and a password. Copy-paste the link (including the port, e.g.: http://12.345.678.91:10002) in your browser. This should result in the following page:

Info

The link gives you access to a web version of Visual Studio Code. This is a powerful code editor that you can also use a local application on your computer.

Type in the password that was provided to you by the teacher. Now let’s open the terminal. You can do that with Ctrl+`. Or by clicking Application menu > Terminal > New Terminal:

For a.o. efficiency and reproducibility it makes sense to execute your commands from a script. With use of the ‘new file’ button:

Material

• Instructions to install docker
• Instructions to set up to container

Exercises

First login

Docker can be used to run an entire isolated environment in a container. This means that we can run the software with all its dependencies required for this course locally in your computer. Independent of your operating system.

In the video below there’s a tutorial on how to set up a docker container for this course. Note that you will need administrator rights, and that if you are using Windows, you need the latest version of Windows 10.

The command to run the environment required for this course looks like this (in a terminal):

Modify the script

Modify the path after -v to the working directory on your computer before running it.

docker run \
--rm \
-p 8443:8443 \
-e PUID=1000 \
-e PGID=1000 \
-e DEFAULT_WORKSPACE=/config/workdir \
-v $PWD:/config/workdir \
geertvangeest/ngs-variants-vscode

If this command has run successfully, use your browser to navigate to port 8443 on your local machine:

http://127.0.0.1:8443

The option -v mounts a local directory in your computer to the directory /config/workdir in the docker container. In that way, you have files available both in the container and on your computer. Use this directory on your computer to e.g. visualise data with IGV. Change the first path to a path on your computer that you want to use as a working directory.

Don’t mount directly in the home dir

Don’t directly mount your local directory to the home directory (/root). This will lead to unexpected behaviour.

The part geertvangeest/ngs-variants-vscode is the image we are going to load into the container. The image contains all the information about software and dependencies needed for this course. When you run this command for the first time it will download the image. Once it’s on your computer, it will start immediately.

If you have a conda installation on your local computer, you can install the required software using conda.

You can build the environment from environment.yml

Generate the conda environment like this:

conda env create --name ngs-tools -f environment.yml

The yaml file probably only works for Linux systems

If you want to use the conda environment on a different OS, use:

conda create -n ngs-tools python=3.8

conda activate ngs-tools

conda install -y -c bioconda \
samtools \
bwa \
snpeff \
gatk4 \
r-base

This will create the conda environment ngs-tools

Activate it like so:

conda activate ngs-tools

After successful installation and activating the environment all the software required to do the exercises should be available.

A UNIX command line interface (CLI) refresher

Most bioinformatics software are UNIX based and are executed through the CLI. When working with NGS data, it is therefore convenient to improve your knowledge on UNIX. For this course, we need basic understanding of UNIX CLI, so here are some exercises to refresh your memory.

If you need some reminders of the commands, here’s a link to a UNIX command line cheat sheet:

UNIX cheat sheet

Make a new directory

Make a directory scripts within ~/workdir and make it your current directory.

cd ~/workdir
mkdir scripts
cd scripts

File permissions

Generate an empty script in your newly made directory ~/workdir/scripts like this:

touch new_script.sh

Add a command to this script that writes “SIB courses are great!” (or something you can better relate to.. ) to stdout, and try to run it.

#!/usr/bin/env bash

echo "SIB courses are great!"

Usually, you can run it like this:

./new_script.sh

But there’s an error:

bash: ./new_script.sh: Permission denied

Why is there an error?

ls -lh new_script.sh

-rw-r--r--  1 user  group    51B Nov 11 16:21 new_script.sh

Make the script executable for yourself, and run it.

chmod u+x new_script.sh

-rwxr--r--  1 user  group    51B Nov 11 16:21 new_script.sh

./new_script.sh

More on chmod and file permissions here.

Redirection: > and |

In the root directory (go there like this: cd /) there are a range of system directories and files. Write the names of all directories and files to a file called system_dirs.txt in your working directory.

ls / > ~/workdir/system_dirs.txt

The command wc -l counts the number of lines, and can read from stdin. Make a one-liner with a pipe | symbol to find out how many system directories and files there are.

ls / | wc -l

Variables

Store system_dirs.txt as variable (like this: VAR=variable), and use wc -l on that variable to count the number of lines in the file.

FILE=~/workdir/system_dirs.txt
wc -l $FILE

shell scripts

Make a shell script that automatically counts the number of system directories and files.

#!/usr/bin/env bash
cd /
ls | wc -l

Loops

If you want to run the same command on a range of arguments, it’s not very convenient to type the command for each individual argument. For example, you could write dog, fox, bird to stdout in a script like this:

#!/usr/bin/env bash

echo dog
echo fox
echo bird

However, if you want to change the command (add an option for example), you would have to change it for all the three command calls. Amongst others for that reason, you want to write the command only once. You can do this with a for-loop, like this:

#!/usr/bin/env bash

ANIMALS="dog fox bird"

for animal in $ANIMALS
do
  echo $animal
done

Which results in:

dog
fox
bird

Write a shell script that removes all the letters “e” from a list of words.

word="test"
echo $word | tr -d "e"

tst

#!/usr/bin/env bash

WORDLIST="here is a list of words resulting in a sentence"

for word in $WORDLIST
do
  echo "'$word' with e's removed looks like:"
  echo $word | tr -d "e"
done

'here' with e's removed looks like:
hr
'is' with e's removed looks like:
is
'a' with e's removed looks like:
a
'list' with e's removed looks like:
list
'of' with e's removed looks like:
of
'words' with e's removed looks like:
words
'resulting' with e's removed looks like:
rsulting
'in' with e's removed looks like:
in
'a' with e's removed looks like:
a
'sentence' with e's removed looks like:
sntnc

Like you might be used to in R or python you can also loop over lines in files. This can be convenient if you have for example a set of parameters in each line of a file.

Create a tab-delimited file animals.txt with the following contents:

dog retrieves   4
fox jumps   4
bird    flies   2

With unix shell you can loop over the lines of that file and store each column as a variable. Below, the three columns in the tab delimited file are stored in the variables $animal, $behaviour and $leg_number:

cat animals.txt | while read animal behaviour leg_number
do
    #something here
done

Exercise: Modify the script in such a way that it writes the strings that are stored in the variables at each line to stdout.

cat animals.txt | while read animal behaviour leg_number
do
    echo "The $animal $behaviour, and has $leg_number legs" 
done

Questions & Answers