This is a test version of Biostars. For the public version, visit https://www.biostars.org.
Modifying fasta header

I have strange fasta headers like this for some good number of sequences,

>gi|61221638|sp|P0A366.1| >gi|61221640|sp|P0A368.1|CR1AA_BACTE RecName: Full= protein synthase>gi|40267|emb|CAA31886.1| unnamed protein product
HTSIDGRPINQGNFSATMSSGSNLQSGSFRTVGFTTPFNFSNGSSVFTLSAHVFNSGNEVYIDRIEFVPAEVTFEAEYDLERAQKAVNELFTSSNQIGLKTDVTDYHIDQVSNLVECLSDEFCLDEKQELSEKVKHAKRLSDERNLLQDPNFRGINRQLDRGWRGSTDITIQGGDDVFKENYVTLLGTFDECYPTYLYQKIDESKLKAYTRYQLRGYIEDSQDLEIYLIRYNAKHETVNVPGTGSLWPLSAQSPIGKCGEPNRCAPHLEWNPDLDCSCRDGEKCAHHSHHFSLDIDVGCTDLNEDLGVWVIFKIKTQDGHARLGNLEFLEEKPLVGEALARVKRAEKKWRDKREKLEWETNIVYKEAKESVDALFVNSQYDQLQADTNIAMIHAADKRVHSI

I would like to replace the other (>gi) in the fasta header to blank or ;. Can anyone suggest how to do it. I have many such sequences in a big fasta file.

awk perl sed unix python
• 3,860 views • link
updated by Ram Ram • written by empyrean999 empyrean999

1 answer

If the symbol before the second > is a space, just:

cut -d ' ' -f 1 in.fasta > out.fasta

But imho, if your fasta headers are broken, you'd better check your upstream workflow

• 0 views • link
updated by Ram Ram • written by Pierre Lindenbaum Pierre Lindenbaum

The space thing worked. Thanks for the tip. I just substituted in vim using space . Unfortunately i cannot change the upstream flow as this itself is starting file which was given to me.

• 1 views • link
written by empyrean999 empyrean999

You should at least inform whoever gave it to you that what they gave you was not valid fasta. This kind of error is just a waste of everyone's time and should be fixed.

• 0 views • link
written by Neilfws Neilfws

Log in to answer this question.

More posts like this