Linux crash course
Ing. Stanislav Smatana
In the beginning there was Richard Stallman
Wish I had a free OS !
● In the 80s everything was proprietary
● R. Stallman founded free software
foundation
● At the core of the project was unix-like
operating system called GNU (GNU is not
Unix)
But one piece was missing
Engine ! (kernel)
Meet Linus Torvalds
● Started his Unix-like kernel as a small hobby
project 1991
● This kernel became known as Linux (guess why)
● Now he is the head of the Linux Foundation and
BDFL of Linux
● His little project blew out of proportions ...
And so GNU/Linux was born
GNU
Linux
GNU/Linux
Ubuntu
Archlinux
Biolinux
Distributions
Let’s jump into the terminal
Current
directory
Username
Computer
name
Like in a file browser, terminal is always opened in some
directory, which is called the current working directory.
Let’s try a bunch of commands
● ls - list current directory
● pwd - print current directory
● mkdir <name> - create directory “name”
● cd <name> - set current directory to “name”
● echo “text” - prints “text”
● cat <name> - print contents of file “name”
Keys
Redirection
● command > file - stores output of
command in file
● ↑↓ - command history
● tab - command completion
● Ctr + c - kills current command
(does not copy !)
● Ctrl + r - search in history
Commands
Type in command and hit enter !
The file system
Absolute path and relative path
Absolute path
/home/matt/mp3/supersong.mp3
slash succession of
directories
Relative path (to the CWD !!)
CWD = /home/matt
mp3/supersong.mp3
No slash !!
Relative path special strings
. - current directory
.. - parent directory
CWD = /home/bill
../matt/mp3/supersong.mp3
File-related commands
● pwd - print current working directory
● ls - list contents of current directory
● cd <path> - change current directory to path
● mkdir <path> - create directory
● rm <path> - remove file
● rmdir <path> - remove directory (only if it is empty)
● cp <path1> <path2> - copy file in path1 into path2
● mv <path1> <path2> - move file in path1 into path2
● cat <path> - print contents of file in path
● head <path> - print n first lines of a file in a path
● tail <path> - print last n lines of a file in a path
wget https://raw.githubusercontent.com/dwyl/english-words/master/words.txt
Complex commands
$ command [options] [arguments]
Options
● Modify behaviour of a
command
● Either short (-l) or (--long)
● Can be given in any order
● E.g. ls -l
Arguments
● Typically paths
● Their order is important !
● E.g. cat my_file.txt
Specific for every command ! Help can be usually found by running
command -h, command --help or man command.
The mighty pipe - counting english words with ing
wget https://raw.githubusercontent.com/dwyl/english-words/master/words.txt
head words.txt
cat words.txt | wc -l
cat words.txt | grep “ing” | wc -l
1. Download list of words
2. Inspect
3. Count all words
4. Count only words containing “ing”
command1 | command2
Automation - creating a script
#! /bin/bash
#This is comment and it is ignored
wget https://raw.githubusercontent.com/dwyl/english-words/master/words.txt
cat words.txt | wc -l
cat words.txt | grep "ing" | wc -l
Save as script.sh and run using command bash script.sh
Improving our little script - variables
NAME=”Peter”
echo “$NAME”
echo “Hello $NAME”
Variables allow you to save values under a given name.
NAME=$(whoami)
echo “$NAME”
echo “Hello $NAME”
Outputs of commands can be saved as variables
Improved script
#! /bin/bash
wget https://raw.githubusercontent.com/dwyl/english-words/master/words.txt 2> /dev/null
NUM_WORDS=$(cat words.txt | wc -l)
NUM_WORDS_WITH_ING=$(cat words.txt | grep "ing" | wc -l )
echo "There are $NUM_WORDS english words and out of them, $NUM_WORDS_WITH_ING end with ing !"
Looping
for file in *.txt
do
…
done
For loop allows you to perform a set of operations on all specified files.
File specification
Commands to execute
Counting reads in sequence files
https://filesender.cesnet.cz/?s=download&token=dfbab33a-05a8-c4b6-70f1-05721c0576e4
1. Download tar.gz archive with sequence files from the link on the bottom of this slide
(do this through browser, not through terminal)
2. Run “tar xvfz sequences.tar.gz” in terminal to unpack files
3. You should see files a.fasta and b.fasta
Counting sequences in files - script
#! /bin/bash
for f in *.fasta
do
N_SEQUENCES=$(cat $f | grep ">" | wc -l)
echo "$f $N_SEQUENCES"
done
Other useful utilities
● top / htop - show currently running processes
● mc - file manager
● nano - text editor
● ssh user@server - connect terminal to remote server
● less - make long output scrollable
● sort
● sed
● awk
● tar
● gunzip
● chmod
● chown
IF control structure
If condition
then
…
fi
If [ $file -d ]
then
echo “It’s a directory !”
fi
Enables program to decide, what path of computation will be taken
based on previous computational result.
Other types of loops
Some advice
● Write steps of your bioinformatic analyses into scripts, otherwise you will
forget what have you done with your data.
● Comment scripts. You will be surprised how quickly you forget what your code
means !
● Name your files consistently.
Other resources
● https://stackoverflow.com/questions/68372/what-is-your-single-most-favorite-c
ommand-line-trick-using-bash
● http://www.proccli.com/2012/01/useful-bash-tricks
● https://www.thegeekstuff.com/2010/08/bash-shell-builtin-commands/