Hi, Fabiano Menegidio ,

You can try query esearch by taxon ID:

"-query txid$taxonid[Organism:exp]"

Replace "$taxonid" with the target taxon ID, which is 7711 for chordata, according to https://www.ncbi.nlm.nih.gov/taxonomy)

David Aciole Barbosa , the question is not just the taxa, but the gene at all taxa.

Below I download all proteins at a given taxa.

esearch -db “protein” -query “txid10239[Organism]” | efetch -format fasta > output.fasta

Would just a new query with cox1 work?

esearch -db “protein” -query “txid10239 [Organism]” -query “cox1” | efetch -format fasta > output.fasta

Or do I need to use an AND? And will I be able to download by the term cox1?

Answering my own question:

esearch -db "protein" -query "txid7993[Organism] AND cox1" | efetch -format fasta > cox1_aa.fasta

Fabiano Menegidio, Yup... That's what I meant, use taxID as filter.

Is it actually protein sequences you needed, not gene?

You can also try adding efilter:

esearch -db protein -query "target protein" | efilter -query txid"target taxa"[Organism:exp] | efetch -format fasta

EDIT: Also, I guess it would be recommended to limit results by

"cytochrome oxidase subunit 1"[Protein Name]

as you may find unrelated results such as "LOW QUALITY PROTEIN: surfeit locus protein 1 [Astyanax mexicanus]", which has "Cox1" in its Region description:

Region 80..295 /region_name="SURF1" /note="SURF1 superfamily. Surf1/Shy1 has been implicated in the posttranslational steps of the biogenesis of the mitochondrially-encoded Cox1 subunit of cytochrome c oxidase (complex IV). Cytochrome c oxidase (complex IV), the terminal electron-transferring...; cd06662" /db_xref="CDD:119401"