Piping fx2tab to grep

[desmodus1984]

Hi,
I am interesting in finding genes under selection. 1st, I use Orthofinder to find the orthogroups using protein data.
The issue is that the selection needs codon alignments.
Thus, what I figured out is - first getting the headers with fx2tab from the Orthogroups MSA, and then using grep to retrieve the CDS from a database. I have been trying to minimize file generation.

As of now, I figured I need to save the headers to a file, and then passing this file to grep.

for f in `cat test_SCO.tsv`; do echo -e "\n$f"; seqkit fx2tab -n MSA/${f}.fa > ${f}.IDS | seqkit grep -f ${f}.IDS CDS.db.fasta | seqkit replace -p '_.+'  > ${f}.CDS.ren.fa ; echo -e "Done $f\n" ; done

Can I pipe fx2tab to grep, or do I unavoidable have to create header file?

Thank you

[shenwei356]

Use process substitution to avoid creating a tmp file.

seqkit grep -f <(seqkit fx2tab -n MSA/${f}.fa) CDS.db.fasta | seqkit replace -p '_.+'  > ${f}.CDS.ren.fa