mjhjmtu

I'm trying to rearrange primer3_core output.

For example:

SEQUENCE_ID=ID_1
PRIMER_LEFT_0_SEQUENCE=ACGTGTAGCGGTTCAGACG
PRIMER_RIGHT_0_SEQUENCE=ACCATGCATGATCCATCCAGG
PRIMER_LEFT_1_SEQUENCE=CACAGCCACAGCAGCACAC
PRIMER_RIGHT_1_SEQUENCE=ATGCAGGTGATCAAGTTACGCC
=
SEQUENCE_ID=ID_2
PRIMER_LEFT_0_SEQUENCE=CACAGCCACAGCAGCACAC
PRIMER_RIGHT_0_SEQUENCE=GCAGGTGATCAAGTTACGCCATT
=

So, it's possible that each ID would have a different number of primers it produces, anywhere from 0-20.

The output would look like this:

ID_1 ACGTGTAGCGGTTCAGACG
ID_1 ACCATGCATGATCCATCCAGG
ID_1 CACAGCCACAGCAGCACAC
ID_1 ATGCAGGTGATCAAGTTACGCC
ID_2 CACAGCCACAGCAGCACAC
ID_2 GCAGGTGATCAAGTTACGCCATT

awk -F= '$0 ~ "^SEQUENCE" SEQ=$2 $0 !~ "^SEQUENCE" print SEQ" "$2 ' filename

Use awk and use = as the field delimiter. Where then line begins with SEQUENCE, set the SEQ variable equal to the second delimited piece. For all other instances, print SEQ along with the second delimited piece of data

Awk approach:

awk -F'=' '/^SEQUENCE_ID/ s = $2 /^PRIMER/ print s, $2 ' file

The output:

ID_1 ACGTGTAGCGGTTCAGACG
ID_1 ACCATGCATGATCCATCCAGG
ID_1 CACAGCCACAGCAGCACAC
ID_1 ATGCAGGTGATCAAGTTACGCC
ID_2 CACAGCCACAGCAGCACAC
ID_2 GCAGGTGATCAAGTTACGCCATT

Using a sed script:

# delete lines starting with '='
/^=/d

# handle sequence ID lines
/^SEQUENCE_ID=/
 # remove everything up to and including the '='
 s///
 # put the sequence ID in the hold space
 h
 # delete the pattern space and continue with next line
 d


# handle primer lines
/^PRIMER.*=/
 # remove everything up to and including the '='
 s///
 # append a newline and the sequence ID from the hold space to the pattern space
 G
 # swap the two bits of the pattern space around, deleting the newline
 s/^(.*)n(.*)$/2 1/

Testing it:

$ sed -f script.sed file
ID_1 ACGTGTAGCGGTTCAGACG
ID_1 ACCATGCATGATCCATCCAGG
ID_1 CACAGCCACAGCAGCACAC
ID_1 ATGCAGGTGATCAAGTTACGCC
ID_2 CACAGCCACAGCAGCACAC
ID_2 GCAGGTGATCAAGTTACGCCATT

Without a separate script file:

$ sed -e '/^=/d' -e '/^SEQUENCE_ID=/s///;h;d;' -e '/^PRIMER.*=/s///;G;s/^(.*)n(.*)$/2 1/;' file
ID_1 ACGTGTAGCGGTTCAGACG
ID_1 ACCATGCATGATCCATCCAGG
ID_1 CACAGCCACAGCAGCACAC
ID_1 ATGCAGGTGATCAAGTTACGCC
ID_2 CACAGCCACAGCAGCACAC
ID_2 GCAGGTGATCAAGTTACGCCATT

Shorter variant:

$ sed -n -e '/^SEQUENCE_ID=/s///;h;' -e '/^PRIMER.*=/s///;G;s/^(.*)n(.*)$/2 1/p;' file