====== A non-comprehensive list of bash commands ======

I have listed a number of common bash commands that come in handy when working in a shell.\\
You can try the [[http://applbio.biologie.uni-frankfurt.de/mscmbw/lib/exe/fetch.php?media=wiki:mbw:shell_basic_commands.txt|following basic commands]] on the [[http://applbio.biologie.uni-frankfurt.de/mscmbw/lib/exe/fetch.php?media=wiki:mbw:ejemplo.txt|example file]] to understand what they are doing.

Feel free to complement your notes on the {{ :general:computerenvironment:bash_spellbook.xlsx |compilation of basic commands}}. 
===== Slightly more advanced bash =====

=== AWK ===

This command is also considered a programming language on its own. It is particularly useful when you need to process the elements of a table. The basic syntax is as follows:

<code>awk -F "\t" '{print $2 "\t" $1} file.txt</code>

  * -F to indicate the delimiter of your table as tabs (default is space).
  * '{print}' to select with "$" the column number you want in the output.
  * Text added to the output must be written inside quotation marks; in this case, the text is just the addition of a new tab.
  * You can also process columns by mathematical operations. For instance, '{print $1 + $2}'


<code>awk -F "\t" '$3 > 10 {print $0}'</code>

  * You can indicate you want only the rows in which a column meets certain condition. For example, the column 3 requires a value greater than 10.
      * Equal ==
      * Non equal =!
      * > Greater than
  * '{print $0}' will print all the columns of the table

----

=== Sort ===

<code>sort -g -u file.txt</code>

  * -r to sort in descending order
  * -g to sort by number
  * -k1,1 to sort by a specific column, in this example, first column only.
  * -u make values unique
  * -t ' ' to select a non-default field separator (default is tab, and in this case I changed to a space) 

----

=== Translate ===

<code>tr '[ATGC]' '[TACG]' | rev</code>

  * This is a trick if you are working on the complementary strand.
  * It will convert each "A" into "T" and so on.
  * rev will make everything read backwards.

----

=== sed (slightly more advanced) ===

<code>sed -n -e '/AAA/,/BBB/ p' file.txt</code>

  * This will find AAA, and keep all the lines in a file until it reaches BBB. Pro tip: use this one to extract a sequence in a multi-line fasta.
  * Note that using variables inside a sed command requires double quotation marks " instead of single '.

===== Working with lists and tables =====

Play with the following sample files.{{ :general:computerenvironment:comm_join_example.tar.gz |}}

=== comm ===

To compare contents of both files (in this case, the identifiers of the first column of the two files):

<code>comm <() <()</code>

  * Within each "<()" we place the command of the input to compare.
  * These should be sorted out
  * I use the second part of the command (the sed) to adjust the output to have the correct number of columns

<code>comm <(cut -f1 1_table.txt | sort) <(cut -f1 2_table.txt | sort) | sed -e 's/$/\t\t/' | cut -f1,2,3</code>

  * Output:
  * First column: identifiers exclusive of the table in input 1
  * Second column: identifiers exclusive of the table in input 2
  * Third column: identifiers present in both tables

----


=== join ===

Join two tables based on a column in common.

<code>join -t $'\t' <() <()</code>

  * Input within the "<()" must be sorted out.
  * I highly recommend using only the first column with the identifiers to join.
  * If one of the tables has repeated identifiers, the output will generate all combinations possible. 
  * The standard output will display only lines with columns in common. We can add option -a1 or -a2 to also include the entries of one of the tables, with no joined values from other. Do not use both.

<code>join -t $'\t' -a1 <(sort -k1,1 1_table.txt) <(sort -k1,1 2_table.txt)</code>