Help with formating fasta headers

[asperjelly]

Hi all,

I'm fairly new at all of this - trained as a benchtop scientist and learning the bioinformatics on the fly, so sorry if this is a relatively naive question. But I need some help re-formatting fasta headers in a multi-fasta file. I would like to change the headers that look like this:

>ref|O74455|ID=6PGL_SCHPO|MODRES=|NCBITAXID=284812|Probable 6-phosphogluconolactonase (6PGL)
MSVYSFSDVSLVAKALGAFVKEKSEASIKRHGVFTLALSGGSLPKVLAEGLAQQRGIEFS

Into the format for blast2go which is >ref|UniqueID|SeqDescription. So I want the above to look like this:

>ref|O74455|Probable 6-phosphogluconolactonase (6PGL)
MSVYSFSDVSLVAKALGAFVKEKSEASIKRHGVFTLALSGGSLPKVLAEGLAQQRGIEFS

I'm assuming something like sed can do this relatively quickly, I just can't seem to figure out how to remove the fields in-between. Thanks in advance for any help you can offer.

- Jon

[GenoMax]

See if this works.

$ cut -f1,2,6 -d"|" inputfile > outputfile

[asperjelly]

fantastic - thank you. I knew there had to be a simple solution.