How to remove the words and just keep the needed in a line?

[wmseq]

Hi every one,
there are a lot lines in a fasta file as follows:

>gi|319965812|ref|NW_003378297.1|Apis mellifera strain DH4 unplaced genomic scaffold, Amel_4.5 GroupUn22, whole genome shotgun sequence

I just want to keep ">" and "GroupUn??", and delete other in each line of them, so that It should be like the following.

>GroupUn22

Any one could help me?

Thanks a lot!!

Richard

[mdalessio]

I may not 100% understand what you are doing, but you can just edit FASTA in a text editor like gedit and just save it

[wmseq]

Thank you for your response!
I am creating the reference genome according to the gff file.

[RDW]

perl -pe 's/^>.+(GroupUn.+),.+/>$1/' in.fasta >out.fasta

[gringer]

er, if you want precisely what you asked for, then:

cat input.fasta | perl -pe 's/^>.*(GroupUn..).*$/>$1/' > output.fasta

... but you probably don't want to do that, because that's just standard regular expression stuff that anyone with Perl experience should be able to work out themselves. It is unclear from your request what you would like to do that extends beyond this.