This is a test version of Biostars. For the public version, visit https://www.biostars.org.
bioawk looping to retrieve fasta sequences by seq ID pattern

Hi, I have a list of partial seq IDs, and a fasta file, and I want to get a fasta file that contains all the sequences that match the IDs in the list.

Example:

list_ID.txt

AF306937.1
S56354.1

Requested output: proteins_output.fasta

lcl|AF306937.1_prot_AAG42454.1_1 MGMDPIALQAGFDLLGDGRPETLWLGIGTLLMIIGTFYFIAQGWGVTDKEAREYYAITILVPGIASAAYLAMFFGIGVTEVELASGAVLDIYYARYADWLFTTPLLLLDLALLAKVDRVSIGTLIGVDALMIVTGLIGALSKTPLARYTWWLFSTIAFLFVLYYLLTSLRSAAAQRSEEVQSTFNTLTALVAVLWTAYPILWIVGTEGAGVVGLGVETLAFMVLDVTAKVGFGFALLRSRAILGETEAPEPSAGT

lcl|S56354.1_prot_AAB19870.2_1 MDPIALQAGFDLLNDGRPETLWLGIGTLLMLIGTFYFIARGWGVTDKEAREYYAITILVPGIASAAYLAMFFGIGVTEVELASGTVLDIYYARYADWLFTTPLLLLDLALLAKVDRVTIGTLIGVDALMIVTGLIGALSKTPLARYTWWLFSTIAFLFVLYYLLTSLRSAAAKRSEEVRSTFNTLTALVAVLWTAYPILWIVGTEGAGVVGLGIETLAFMVLDVTAKVGFGFVLLRSRAILGETEAPEPSAGADASAAD

I wrote this loop for bioawk, but it isn't quite working as I want.

When I run this loop, I get an empty "proteins_outout.fasta" file:

while read p; do
   bioawk -c fastx '$name ~ /$p/ { print ">"$name"\n"$seq"\n"; }' proteins_db.fasta > proteins_outout.fasta; 
done <list_ID.txt

Although when I try to run the bioawk line for a single seq ID, I do get the right output:

bioawk -c fastx '$name ~ /AF306937.1/ { print ">"$name"\n"$seq"\n"; }'
proteins_db.fasta

lcl|AF306937.1_prot_AAG42454.1_1 MGMDPIALQAGFDLLGDGRPETLWLGIGTLLMIIGTFYFIAQGWGVTDKEAREYYAITILVPGIASAAYLAMFFGIGVTEVELASGAVLDIYYARYADWLFTTPLLLLDLALLAKVDRVSIGTLIGVDALMIVTGLIGALSKTPLARYTWWLFSTIAFLFVLYYLLTSLRSAAAQRSEEVQSTFNTLTALVAVLWTAYPILWIVGTEGAGVVGLGVETLAFMVLDVTAKVGFGFALLRSRAILGETEAPEPSAGT

Why is my loop not working as I want?

Would appreciate any help!

Thanks

regex fasta bioawk loop
• 2,275 views • link
updated by ATpoint ATpoint • written by shaharbra • shaharbra

1 answer

I think a) you have to pass the shell variable $p to bioawk properly using -v and b) you are currently overwriting your output file in every iteration as you use > rather than >>:

while read p
    do
    bioawk -v var="${p}" -c fastx '$name ~ var {print ">"$name"\n"$seq}' proteins_db.fasta >> proteins_outout.fasta
    done < list_ID.txt
• 0 views • link
written by ATpoint ATpoint

Yessssss thank you! I didn't know about -v for passing shell variables to bioawk. Working perfectly now!

• 0 views • link
written by shaharbra • shaharbra

Cool! The -v is actually from "standard" awk, and afaik bioawk can do all awk things plus bioinformatics-specific ones.

• 0 views • link
written by ATpoint ATpoint

Log in to answer this question.

More posts like this