change headers from fasta files

Question

[erickfqqa]

Hello, I'm trying to trim the next fasta headers

  >ID:CHARACTERS | [Genus specie] | strain | gene_name | length | NCBI_ID | other | other | other|
>ID:CHARACTERS | [Genus specie | strain | gene_name | length | NCBI_ID | other | other | other|

And I 'd like to trim them like the following example

>ID:CHARACTERS | [Genus specie] | strain | gene_name |

the ID:CHARATERS, [Genus specie], strain and gene_name are variables accross the headers, but always are separated by "space|space"

[h.mon]

Your fasta headers are identical, this will create problems for most downstream tools. You can do what you want with sed:

sed "s/ length | NCBI_ID | other | other | other|//" file.fasta

[WouterDeCoster]

I'm trying to trim the next fasta headers

And can you elaborate on what you tried and how that didn't deliver the result you had in mind?

Answer 1

[Joe]

This works if you aren't bothered about keeping the last pipe symbol...

cat myseqs.fa | cut -d '|' -f -4
>ID:CHARACTERS | [Genus specie] | strain | gene_name
TCCACGATCGAATAAATGTGCGATTAGCACCTGTAGAACCATACAAGCTAAGCCGCCTCACGGGCATGTTAACGAGATTA
>ID:CHARACTERS | [Genus specie | strain | gene_name
TATTAATTTTAGTAAAACATAGCGTCGGAGACGGGCCGACACCAACGATCCGTTACCCCACATGCACCGGAGTAAGCGAC

I'll let you exert some effort to keep the last delimiter if you want it ...

[Joe]

I'll add that you really should use the search function as this type of question is Biostar's 'public enemy number one'.

You're lucky it's easy and I'm sat in front of a terminal though ;)

Answer 2

[cpad0112]

$ cat test.fa 
>ID:CHARACTERS | [Genus specie | strain | gene_name | length | NCBI_ID | other | other | other|
atgc

$ sed '/^>/ s/|\s\w\+\s*//3g' test.fa 
>ID:CHARACTERS | [Genus specie | strain | gene_name |
atgc