how to delete particular number or letter before an symbols (|) using grep command

Question

I have N number of sequence in a file like this

>668|CAX10446 conserved hypothetical protein 
MSIRGVGGNGNSRIPSHNGDGSNRRSQNTKNKVEDRVRSLYSSRSNENRESPYAVVDVSSMIESTPTSGE
TTRASRGVFSRFQRGLGRVADKVRRAVQRAWSSVSIRRSSATRATESRSSSRTARGASSGYKEYSPSAAR

>669|CAX10447 conserved protein 
MSIRGVGGNGNSRIPSHNGDGSNRRSQNTKNKVEDRVRSLYSSRSNENRESPYAVVDVSSMIESTPTSGE
TTRASRGVFSRFQRGLGRVADKVRRAVQRAWSSVSIRRSSATRATESRSSSRTARGASSGYKEYSPSAAR

or like this

>gi|76167242|gb|AAX50250.1| hypothetical protein CTA_0001 
MSIRGVGGNGNSRIPSHNGDGSNRRSQNTKNKVEDRVRSLYSSRSNENRESPYAVVDVSSMIESTPTSGE
TTRASRGVFSRFQRGLGRVADKVRRAVQRAWSSVSIRR

Now, I need a grep command to delete a particular word or number in first line.

Output should be:

>CAX10446 conserved hypothetical protein 
MSIRGVGGNGNSRIPSHNGDGSNRRSQNTKNKVEDRVRSLYSSRSNENRESPYAVVDVSSMIESTPTSGE
TTRASRGVFSRFQRGLGRVADKVRRAVQRAWSSVSIRRSSATRATESRSSSRTARGASSGYKEYSPSAAR

or

>AAX50250.1| hypothetical protein CTA_0001 [Chlamydia trachomatis A/HAR-13]
MSIRGVGGNGNSRIPSHNGDGSNRRSQNTKNKVEDRVRSLYSSRSNENRESPYAVVDVSSMIESTPTSGE
TTRASRGVFSRFQRGLGRVADKVRRAVQRAWSSVSIRRSSATRAAESRS
SSRTARGASSGYREYSPSAAR

I need a grep command to do this action, I used sed but it removed the first line, the ID, so it's difficult to identify which sequence is which.

Answer 1

It looks like you want to remove everything before the last | character. However, your second example has a | after the ID and before the hypothetical protein. Presumably, in that case, you want to remove everything before the last | that is not followed by a space.

Here's one way:

$ sed 's/| / /; s/^[^ ]*|/>/' file.pep 
>CAX10446 conserved hypothetical protein 
MSIRGVGGNGNSRIPSHNGDGSNRRSQNTKNKVEDRVRSLYSSRSNENRESPYAVVDVSSMIESTPTSGE
TTRASRGVFSRFQRGLGRVADKVRRAVQRAWSSVSIRRSSATRATESRSSSRTARGASSGYKEYSPSAAR
>AAX50250.1 hypothetical protein CTA_0001 
MSIRGVGGNGNSRIPSHNGDGSNRRSQNTKNKVEDRVRSLYSSRSNENRESPYAVVDVSSMIESTPTSGE
TTRASRGVFSRFQRGLGRVADKVRRAVQRAWSSVSIRR

The s/| / /; removes any | followed by a space. This will remove the | from AAX50250.1|. The s/^[^ ]*|/>/ looks for a > at the beginning of the line (^>) and removes the longest string of non-space ([^ ]) characters until the last | on the line.

Answer 2

If you want to do this with grep. You should read 3 line after each empty line and copy it to variable. You should cut last part of variable.

The parser will do;

First reverse string

Second cut string with | and get first

Third reverse string

echo $yourstring |rev |cut -d'|' -f1|rev

BTW this will not remove, only shows you the part. If you want to save output redirect it to file.

echo $yourstring |rev |cut -d'|' -f1|rev > yournewfile