getFASTAFromUniProt and getSeqFromUniProt errors

[Amirreza-Mousavi]

id = c('P00750', 'P00751', 'P00752')

getFASTAFromUniProt(id) 
#gives "" "" ""

from R documentations of getFASTAFromUniProt does not run properly. the second line gives an empty character string, which will result in errors in other functions dependent on getFASTAFromUniProt such as getSeqFromUniProt.

Also,

id = c('P00750', 'P00751', 'P00752')

getSeqFromUniProt(id)
#ERROR

leads to error.

(Error in FUN(X[[i]], ...) : no line starting with a > character found)

which is obviously resulted from problems in getFASTAFromUniProt.

It seems that getURLAsynchronous in the implementation of getFASTAFromUniProt is causing the problem somehow. It's better to replace it with some equivalent base-R function.

[askarlupka]

I'm having the same issue with getSeqFromUniProt. I believe the URL has changed on uniprot's end. In the getFASTAFromUniProt function getURLAsynchronous is directed to a website. However I think that has recently been updated to, "https://https://rest.uniprot.org/uniprotkb/", id, ".fasta"

> getFASTAFromUniProt
function (id, parallel = 5)
{
fastaURL = paste0("https://www.uniprot.org/uniprot/", id,
".fasta")
fastaTxt = getURLAsynchronous(url = fastaURL, perform = parallel)
return(fastaTxt)
}
<bytecode: 0x7fcc0e8555c0>
<environment: namespace:Rcpi>

Example: "https://www.uniprot.org/uniprot/P00750.fasta" gets redirected to "https://rest.uniprot.org/uniprotkb/P00750.fasta"

[nanxstats]

Thanks @askarlupka and @Amirreza-Mousavi for reporting this.

I've fixed the issue in this repo - you can try it with remotes::install_github("nanxstats/Rcpi").

I also submitted the updated version v1.33.2 (devel) and v1.32.2 (release) to Bioconductor. It may take a few days to fully propagate through their build system.