extract sequence problem

Question

[mxlsherry1992]

Hi, I have a fasta file like this

>TRINITY_DN100000_c1_g1::TRINITY_DN100000_c1_g1_i3::g.3039::m.3039 TRINITY_DN100000_c1_g1::TRINITY_DN100000_c1_g1_i3::g.3039  ORF type:complete len:100 (-) TRINITY_DN100000_c1_g1_i3:1027-1326(-)
MVWIKFRGLHRVLTSTPLVKSGKTPSQTWAFLDISVELIVFLFLNVHKSPMPHFKIYSEA
FSEEWSLLWLQYSRHLIQKPKPWQIKIELLHLCCCNRLC*
>TRINITY_DN100000_c1_g6::TRINITY_DN100000_c1_g6_i2::g.84365::m.84365 TRINITY_DN100000_c1_g6::TRINITY_DN100000_c1_g6_i2::g.84365  ORF type:complete len:112 (-) TRINITY_DN100000_c1_g6_i2:379-714(-)
MEMMQEIIPFAREMLSARPSKGTMKVYLVGGTFAVLGIVSGMVEAACSLFPEQEESTLTK
LMEDCLTVTAQNQEPQTFIPEDDEQDAEMEAKAKDLPMFRQRRMSFRAHAS*

If I want to simplify it like this:

>TRINITY_DN100000_c1_g1_i3
MVWIKFRGLHRVLTSTPLVKSGKTPSQTWAFLDISVELIVFLFLNVHKSPMPHFKIYSEA
FSEEWSLLWLQYSRHLIQKPKPWQIKIELLHLCCCNRLC*
>TRINITY_DN100000_c1_g6_i2
MEMMQEIIPFAREMLSARPSKGTMKVYLVGGTFAVLGIVSGMVEAACSLFPEQEESTLTK
LMEDCLTVTAQNQEPQTFIPEDDEQDAEMEAKAKDLPMFRQRRMSFRAHAS*

what command should I use? I know python can easily solve this, but if there is any simple command could do that..

[cpad0112]

$ sed '/^>/ s/.*::\(.*\)::.*::.*::.*::.*/>\1/g' test.fa   

>TRINITY_DN100000_c1_g1_i3
MVWIKFRGLHRVLTSTPLVKSGKTPSQTWAFLDISVELIVFLFLNVHKSPMPHFKIYSEA
FSEEWSLLWLQYSRHLIQKPKPWQIKIELLHLCCCNRLC*
>TRINITY_DN100000_c1_g6_i2
MEMMQEIIPFAREMLSARPSKGTMKVYLVGGTFAVLGIVSGMVEAACSLFPEQEESTLTK
LMEDCLTVTAQNQEPQTFIPEDDEQDAEMEAKAKDLPMFRQRRMSFRAHAS*

Answer 1

[stcatpang]

try 's/::.*//' input > output

[mxlsherry1992]

Hi, thanks, so it is sed 's/::.*//' input > output , it can keep the first header ">TRINITY_DN100000_c1_g1", if we want to keep the second header ">TRINITY_DN100000_c1_g1_i3", if you have any other modified?

Thank you!!!