Skip to contents

Fetch NCBI references for many accessions using batched EFetch requests

Source: R/blast_ref_utils.R
fetch_blast_refs.Rd

Reduces NCBI round-trips by fetching all accessions' GFF3 and FASTA in one request each (plus one taxonomy request per unique taxid), then writing one `ref_<accession>/` subdirectory per accession (annotations CSV, sequence, genetic code, JSON) under out_dir. The concatenated responses are split by accession key, so a scrambled batch can never misassign a record. Any accession the batch cannot produce (missing/invalid chunk, or a failed batched HTTP call) falls back to a per-accession [fetch_blast_ref()] call. A fully failed accession produces no output dir (its samples are flagged downstream), matching the old one-accession-per-task behavior.

Usage

fetch_blast_refs(
  accessions,
  out_dir = ".",
  blast_species = NULL,
  blast_evalue = NULL
)

Arguments

accessions

character vector of NCBI accessions

out_dir

directory to create the per-accession `ref_<accession>/` dirs in

blast_species, blast_evalue

unused (per-sample metadata is stamped downstream); kept for signature parity with [fetch_blast_ref()].