This section contains some basic linux information and tips for setting up an environment for bioinformatics, but imo the bare essentials to get started withour prior programming knowledge. It contains elementary information so experienced users should skip this.

This is by no means exaustive, and not the best resource available, it contains enough to get started and is the ‘obvious’ stuff I wish I knew when I was starting out.

Open a terminal

To open a terminal on a mac press and hold cmd then press space to open spotlight. Type ‘terminal’. This opens a window in your home folder. For instructions on how to do this on a windows machine see here.

The terminal is just an alternative way to access the filing system on your computer without using the graphical user interface (GUI). So instead of seeing icons which represent folders and programs we just see lists. Whenever we open a terminal window for the first time we are located in the home directory of our computer.

Navigation

To carry out the tasks we need to via the command line we need to navigate between directories. To find out where we are at any given time we type:

pwd

Which prints the current working directory to the screen. This is the address for our home directory. To list
all the directories and files in our home folder we can type:

ls

We can see this directory contains twelve sub-directories, with a / after their name, and one file. Note here the directories are in blue but this coulour scheme may be different, or non-existant, on different machines.

To move between folders we use the command cd.

cd Desktop

Takes us to the Desktop directory. Typing ls again lists the contents of the Programs directory.

To move back one file in the file hierarchy we can type:

cd ..

And at any time if we want to return to the home directory from anywhere in the files hierarchy we type:

cd

# OR

cd ~/   

There are two options here we can just type cd or use the tilde which is command line shorthand for the home directory.

Relative and absolute paths

From the Desktop directory we can move forward using either relative or the absolute paths. The relative path requires you to type the adresss of your target directory relative to where you currently are in the file hierarchy. So if you are currently located in the /Users/Darren/Desktop/Programs/ directory, to move to the Homer directory, you simply have to type:

cd /Homer

Typing the absolute path would do the same thing from your currently location:

cd /Users/Darren/Desktop/Programs/Homer

However, only the absolute pathw would take you to that particular folder from anywhere in the file hierarchy, as the relative path can only access directories relative to the directory you are located.

Autocompletion

A useful tool when using the command line is the autocomplete fuction. Rather than typing out long directory addresses all the time we only need to type the first letter of any directoy in the current folder and press the tab key to auto complete the name of that directory. This function is case senstive.

If you have multiple directories starting with the same letter you need to type enough letters until you can uniquely identify the file of interest.

Note that you can autocomplete through multiple directories consecutively if you know the names of the directories you intend to move through in advance. If you need to see a list of the files in a particular directory that start with the same letter, press tab twice.

Assigning variables

Variables are used to assign information in programming. In bash we assign variables using the = sign. The variable is on the left and the information we want to assign is on the right. It works similarly to algebra.

x=1

After you have assigned 1 to x any time you type echo $x the computer will interpret x as 1. Note that this is only true in the current session. In bash the $ sign must proceed a call to any variable and spaces are not allowed at either sided of =.

The ‘echo’ statement prints whatever comes next to the screen.

Hidden files

When we type ls on the command line we are not presented with a list all the files in the current directory. Most working directories contain a series of ‘hidden’ files that are there to run background processes for specific programs. They are not immediatly accessible as, generally, altering these files is not recommended.

We can view these files by adding the -a flag to the ls call.

ls -a

Note that there are several files that begin with ., these are the hidden files. The most important one for our purposes is the file called .bash_profile.

Subsetting fastq files for testing

Often when debugging a script we want to subset our fastq files so that we can significantly reduce the processing time for every step in our script rather than run all the data through at once.

# read the compressed forward fasq file, 
# piping the result in sed to extract the 
# first 4 million lines (equivalent to 1M sequences, 
# as one sequence entry always span 4 lines)
# and finally compressing it as a new file

zcat pair_1.fastq.gz | sed -n 1,20000000p | gzip -c > pair-subset_1.fastq.gz

# same for read 2 (the reverse read) if your data is Paired-End
zcat pair_2.fastq.gz | sed -n 1,20000000p | gzip -c > pair-subset_2.fastq.gz

Move on to Scope, or back to Index.