Undesired added bonus w ncbi_searcher using id (var.)

[mpnelsen]

Not trying to be a troublemaker, but I've encountered another minor issue with the the ncbi_searcher function. I'm using the txid number and it's retrieving the accessions of interest, but it's also retrieving accessions for varieties, forms, etc of the same species (which actually have different txid numbers). I know I can find filter the output and just extract the sequences i'm interested in (filter by species name or else get taxid numbers for each accession and then filter), but it seems like it would be more efficient to refine the initial search to only retrieve accessions matching that txid.

Example:

require(traits)
search.res<-ncbi_searcher(id=60451,getrelated=FALSE,entrez_query="internal transcribed spacer",verbose=FALSE,seqrange="350:3500")
sort(unique(search.res$taxon))
[1] "Tuber excavatum" "Tuber excavatum var. intermedium"
[3] "Tuber excavatum var. longisporum" "Tuber excavatum var. sulphureum"

I'm wondering if perhaps the [ORGN] modifier is automatically used in the ncbi_searcher search, which also grabs varieties (with diff txid numbers), etc.?

For instance, on NCBI...a search using:
txid60451[ORGN] AND internal transcribed spacer
----retrieves varieties

but,

txid60451 AND internal transcribed spacer
------or
txid60451[ORGN:noexp] AND internal transcribed spacer
------only retrieve the txid of interest

Would it be possible to modify the search option in ncbi_searcher to only retrieve accessions for the txid of interest? Thanks much for your help!

[sckott]

hi, not troublemaking at all - super useful in fact! there's nothing better than a user finding a potential problem

this is what we're using https://github.com/ropensci/traits/blob/master/R/ncbi_searcher.R#L124 the porgn - makes sense if searching by taxonomic ID to only get records that match that ID exactly

[sckott]

@mpnelsen can you reinstall from github, then try again?

see the new fuzzy parameter. releavant change is extracted here:

query_term <- if (fuzzy) {
    sprintf("xXarbitraryXx[porgn:__txid%s] AND %s[SLEN]", id, seqrange)
 } else {
    sprintf("txid%s AND %s[SLEN]", id, seqrange)
}

We needed to retain a fuzzy option since the getrelated parameter would have done nothing if we only allowed txid<ID> searches

[mpnelsen]

works perfectly! thanks again.

[sckott]

👍