This is a test version of Biostars. For the public version, visit https://www.biostars.org.
HOw to merge multifasta sequence into a single sequence having only one header?

I have a multifasta sequence file. I want to merge all the sequences together to create a single sequence file. I men that the ">IDs" in the sequences be removed to create a super sequence. THis would take much time doing mannualy.

how can it be done in linux

THanks

fasta merge
• 14,739 views • link
updated by Charles Plessy Charles Plessy • written by majeedaasim • majeedaasim

@majeedaasim please choose the accept answer option if it works for you, It will help us motivated. Good Luck!

• 0 views • link
written by mittu1602 mittu1602

4 answers

Using the union command from the EMBOSS package:

$ cat test.fasta 
>seq1
AAAATTGGG
>seq2
GGCCCTTTT
>seq3
AAATGGGG

$ union -filter test.fasta
>seq1
AAAATTGGGGGCCCTTTTAAATGGGG
• 0 views • link
written by Charles Plessy Charles Plessy

grep -v "^>" test.fasta | awk 'BEGIN { ORS=""; print ">My_New_Sequence_name\n" } { print }' > new​.fasta

test.fasta
>seq1
AAAATTGGG
>seq2
GGCCCTTTT
>seq3
AAATGGGG

new.fasta
>My_New_Sequence_name
AAAATTGGGGGCCCTTTTAAATGGGG
• 0 views • link
written by mittu1602 mittu1602

Hi,

Would this work on a mac osx?

• 0 views • link
written by cssulliv • cssulliv

Not necessarily. MacOS ships with a non standard version of grep (I.e. not GNU coreutils). Consequently, the syntax often isn't 100% transferable. It may work, but that's not something you can rely on. You can however download and install the 'proper' coreutils via HomeBrew or MacPorts.

• 0 views • link
written by Joe Joe 
cat multifasta.fa | sed -e '1!{/^>.*/d;}' | sed  ':a;N;$!ba;s/\n//2g' > output.fa

E.g:

$ cat ~/test/seqs.fasta
>tpg|Magnaporthiopsis_incrustans|JF414846
ACTGTAGTAGCTACGATCGATCAGATGATCACGTAGCATCGATCGATCATCGACTAGTAGATCACTCGACATAGATCCACATCAATAGATCATCATCATCATAATCGATCACTAGCAGCNNNNNN
>tpg|Pyricularia_pennisetigena|AB818016
NNNNNNGCAAGNTTCATGACGATGTAGAATGGCTTATCGAAGGGAGCAGGCCAGGGATTGAGGTCCGTCTCACGGGTTGGCTTCACTCCCCCACTGCCAGCCCTCTTGCTGCAACTCCACCAGAA
>tpg|Inocybe_sororia|EU525947
NNNAACCANGCCGCGACGGCGGTGCGATCGGGAAACGCGGCGGTGGCGGAGGAATCGGCCATCCTTCACCATATCGGCCAAGGATTGTGGTTCCTGTAGGGCTCGCGCAGCCCAGGACGCGCNNN


$ cat ~/test/seqs.fasta | sed -e '1!{/^>.*/d;}' | sed  ':a;N;$!ba;s/\n//2g'
>tpg|Magnaporthiopsis_incrustans|JF414846
ACTGTAGTAGCTACGATCGATCAGATGATCACGTAGCATCGATCGATCATCGACTAGTAGATCACTCGACATAGATCCACATCAATAGATCATCATCATCATAATCGATCACTAGCAGCNNNNNNNNNNNNGCAAGNTTCATGACGATGTAGAATGGCTTATCGAAGGGAGCAGGCCAGGGATTGAGGTCCGTCTCACGGGTTGGCTTCACTCCCCCACTGCCAGCCCTCTTGCTGCAACTCCACCAGAANNNAACCANGCCGCGACGGCGGTGCGATCGGGAAACGCGGCGGTGGCGGAGGAATCGGCCATCCTTCACCATATCGGCCAAGGATTGTGGTTCCTGTAGGGCTCGCGCAGCCCAGGACGCGCNNN

(retains just the header of the first seq in the multifasta)

Bonus:

If you also want to hard line-wrap the fasta to 80 chars (or whatever), the command becomes;

cat $1 | sed -e '1!{/^>.*/d;}' | sed ':a;N;$!ba;s/\n//2g' | sed '1!s/.\{80\}/&\n/g'
• 0 views • link
written by Joe Joe

Can you keep it the last file .gz? Thanks

• 0 views • link
written by psschlogl • psschlogl

You can pipe the output of the command to gzip - just tell it to use STDIN as the data source.

• 0 views • link
written by Joe Joe

grep -v '^>' in.fa > out.fa

if in.fa =

>chr1
ttttccccaaaagggg
>chr2
ACTGACTGnnnnACTG
>chr3.1
ACTGACTGaaaac
>chr3.2
ACTGACTGaaaacc
>chr3.3
ACTGACTGaaaaccc
>chr4
ACTGnnnn
>chr5
nnACTG

then out.fa becomes:

ttttccccaaaagggg
ACTGACTGnnnnACTG
ACTGACTGaaaac
ACTGACTGaaaacc
ACTGACTGaaaaccc
ACTGnnnn
nnACTG
• 0 views • link
written by yhoogstrate yhoogstrate

Log in to answer this question.

More posts like this