remove fasta file heading

Question

[ravihansa82]

Dear friends I need a help from you to remove some part of fasta file heading.

For example I need to keep only gene id (ENSG00000026652) so that I need to remove all other parts right to the pipe

Since I am having about 4000 sequences I need to do this with the help of some programming

>ENSG00000026652|ENST00000437165;ENST00000320285;ENST00000366911;ENST00000436279;ENST00000366905
TTTGACACTTGCATAGCTGTTAGTAATTCTGCAATGTGCTTGGGCTATTTGTGAGCACAT
GTATTTTTCCTTTATAGATTATAAACATCTAAAGAACAAGGTTAACCCAGGAGTCAAGTA
AATAGTTAAATATTATTTTGACAATGGCTGTAATAGTGGACATTTGAAAGGAATACACCT
CAGTATTTTGAAATTGAAATAATTTTCTAGATCCTGGCATTTCTGGACTTTCAACAGCCC

Desired output

>ENSG00000026652
TTTGACACTTGCATAGCTGTTAGTAATTCTGCAATGTGCTTGGGCTATTTGTGAGCACAT
GTATTTTTCCTTTATAGATTATAAACATCTAAAGAACAAGGTTAACCCAGGAGTCAAGTA
AATAGTTAAATATTATTTTGACAATGGCTGTAATAGTGGACATTTGAAAGGAATACACCT
CAGTATTTTGAAATTGAAATAATTTTCTAGATCCTGGCATTTCTGGACTTTCAACAGCCC

I am bit familiar with the perl so can somebody help me to do this?

Answer 1

[Jorjial]

If you now that the part you want to delete starts with "|", you can use sed in the terminal:

sed 's/|.*$//' file.fasta > new_file.fasta

What you are doing is to delete all the characters from "|" to the end of the line.

I hope this helps

[ravihansa82]

thank you... Jorjial that has perfectly worked....

[xiachongjing]

Hello, I wonder how to use this command to delete all the character from"|" to the left end of the line? Thanks!

[Misz Laydy]

Thanks a lot ! Its work very well for my multifasta file that Ive been looking for two days.

Answer 2

[edrezen]

Hello,

Not a perl answer, but you can do it with the 'gawk' command (or 'awk' if you don't have gawk)

If the separator is only the pipe character, you can do the following:

gawk 'BEGIN{FS="|"}{print $1}' reads.fa

[PoGibas]

You should get rid of cat part.

[Devon Ryan]

I just did that for edrezen

Answer 3

[5heikki]

cut -f1 -d "|" input > output

[xiachongjing]

Hello, how can I use this command to remove strings all the left from "|"? Thank you.

[5heikki]

You mean keep everything right from pipe?

cut -f2- -d "|" input > output

You should:

man cut

To see:

Numbers or number ranges may be fol-
 lowed by a dash, which selects all fields or columns from the last number to the end of the line.

[xiachongjing]

Thank you 5heikki. It works, but not as what I expected. I am sorry I did make it clear. What I want is everything right from pipe and the ">". For example, my header is: >PacBio|Sequence0001 what I want is: >Sequence0001

Thanks so much for your response.

[5heikki]

Can't do that with cut.

awk 'BEGIN{FS="|"}{if(/^>/){print ">"$2}else{print $0}}' input > output

Answer 4

[Devon Ryan]

Because I think perl is a terrible language, here's how to do it with biopython:

#!/usr/bin/env python
from Bio import SeqIO
import re

of = open("foobar.fa", "w")
for record in SeqIO.parse("foo.fa", "fasta") :
    matches = re.search("(ENSG[\d]+)", record.id)
    if(matches != None) :
        record.id = matches.group(1);
        record.description=""
    SeqIO.write(record, of, "fasta")
of.close()

You could easily translate that to either perl or bioperl. You could also do this with a sed or awk one-liner.

[rwn]

Out of interest, why do you think Perl is a terrible language?

[Devon Ryan]

It makes it too easy for people to write opaque unsupportable code.