Sort header from the multifasta sequnce file into only gene name

Question

[k.kathirvel93]

I have multifasta protein sequences with long headings, but i want to exclude everything and keep only gene name which appears after 'GN= '. Can anyone help with this pls....

sp|P0AGM2|YICG_ECOLI UPF0126 inner membrane protein YicG OS=Escherichia coli (strain K12) OX=83333 GN=yicG PE=1 SV=1

MLLHILYLVGITAEAMTGALAAGRRRMDTFGVIIIATATAIGGGSVRDILLGHYPLGWVK HPEYVIIVATAAVLTTIVAPVMPYLRKVFLVLDALGLVVFSIIGAQVALDMGHGPIIAVV AAVTTGVFGGVLRDMFCKRIPLVFQKELYAGVSFASAVLYIALQHYVSNHDVVIISTLVF GFFARLLALRLKLGLPVFYYSHEGH

sp|P64442|YCEO_ECOLI Uncharacterized protein YceO OS=Escherichia coli (strain K12) OX=83333 GN=yceO PE=1 SV=1

MRPFLQEYLMRRLLHYLINNIREHLMLYLFLWGLLAIMDLIYVFYF

I want output like this (with > symbol)

yicG

MLLHILYLVGITAEAMTGALAAGRRRMDTFGVIIIATATAIGGGSVRDILLGHYPLGWVK HPEYVIIVATAAVLTTIVAPVMPYLRKVFLVLDALGLVVFSIIGAQVALDMGHGPIIAVV AAVTTGVFGGVLRDMFCKRIPLVFQKELYAGVSFASAVLYIALQHYVSNHDVVIISTLVF GFFARLLALRLKLGLPVFYYSHEGH

yceO

MRPFLQEYLMRRLLHYLINNIREHLMLYLFLWGLLAIMDLIYVFYF

Note : Sorry > symbol was there in all the fasta headers but its not appearing in biostars ( May be i don't know how to post)

Thanks in advance

[cpad0112]

you can also try this:

sed  '/^>/ s/.*\sGN=\(.*\)\sPE.*/>\1/g' test.fa

Answer 1

[gayachit]

Very rough solution but should work:

sed -e 's/.*GN=/>/' -e 's/PE=.*//g' your_file.txt > new_file.txt