Closed remcv closed 1 year ago
I can see there is a workaround described by you in a different issue https://github.com/msberends/AMR/issues/54#issuecomment-1120890522
Thanks, and you’re welcome!
Yes, it seems the current release shows some problems with translating user input. Not sure why this didn’t come up in the past three years. Especially for C. auris it’s rather painful that it doesn’t work right.
The good news is that the new algorithm fixes this, which is now available as beta version.
as.mo("Candida auris")
#> Class 'mo'
#> [1] F_CANDD_AURS
# full info:
mo_info("Candida auris")
#> $kingdom
#> [1] "Fungi"
#>
#> $phylum
#> [1] "Ascomycota"
#>
#> $class
#> [1] "Saccharomycetes"
#>
#> $order
#> [1] "Saccharomycetales"
#>
#> $family
#> [1] "Debaryomycetaceae"
#>
#> $genus
#> [1] "Candida"
#>
#> $species
#> [1] "auris"
#>
#> $subspecies
#> [1] ""
#>
#> $status
#> [1] "accepted"
#>
#> $synonyms
#> NULL
#>
#> $gramstain
#> [1] NA
#>
#> $url
#> [1] "https://www.gbif.org/species/3479791"
#>
#> $ref
#> [1] "Satoh et al., 2009"
#>
#> $snomed
#> [1] "3491000146109"
You can install the beta version with:
remotes::install_github("msberends/AMR")
We’re curious for feedback, so please let us know :)
I've installed the beta version and used it. The good news is that the Candida auris problem is solved. The bad news is that now there are more mo uncertainties, including microbes that had a name change and were previously (in 1.8.2) detected as such.
Below is the list of germs I'm trying to map to MOs.
> unique(urinary_isolates$microb)
[1] Enterobacter cloacae Candida tropicalis
[3] Escherichia coli Citrobacter koseri
[5] Proteus mirabilis Klebsiella pneumoniae
[7] Streptococcus agalactiae Enterococcus faecalis
[9] Candida parapsilosis Candida albicans
[11] Pseudomonas aeruginosa Enterococcus faecium
[13] Cronobacter sakazakii Serratia marcescens
[15] Citrobacter freundii Enterococcus spp
[17] Candida non-albicans Klebsiella aerogenes
[19] Klebsiella oxytoca Morganella morganii
[21] Kluyveromyces marxianus Proteus vulgaris
[23] Staphylococcus aureus Acinetobacter baumannii
[25] Providencia stuartii Acinetobacter calcoaceticus/baumannii complex
[27] Candida rugosa Pseudomonas spp
[29] Enterobacter hormaechei Providencia rettgeri
[31] Proteus penneri Candida glabrata
[33] Enterococcus gallinarum Pantoea agglomerans
[35] Stenotrophomonas maltophilia Candida norvegensis
[37] Candida krusei Candida guilliermondii
[39] Citrobacter spp Atelosaccharomyces hudeloi
[41] Citrobacter braakii Candida auris
[43] Cryptococcus laurentii Acinetobacter pittii
[45] Burkholderia cepacia Serratia spp
[47] Enterobacter asburiae Salmonella enterica
[49] Raoultella ornithinolytica Acinetobacter lwoffii
[51] Alcaligenes faecalis Pseudomonas putida
[53] Achromobacter xylosoxidans Serratia odorifera
[55] Candida dubliniensis Citrobacter farmeri
[57] Candida ciferrii Staphylococcus hominis
[59] Citrobacter youngae Enterobacter kobei
[61] Staphylococcus saprophyticus Achromobacter denitrificans
[63] Saprochaete capitata Acinetobacter spp
[65] Proteus hauseri Pantoea dispersa
[67] Klebsiella variicola Sphingomonas paucimobilis
[69] Serratia liquefaciens Candida lusitaniae
[71] Hafnia alvei Enterococcus hirae
[73] Streptococcus gallolyticus Serratia ficaria
[75] Burkholderia contaminans Serratia fonticola
[77] Proteus spp Enterobacter spp
[79] Myroides spp Pseudomonas luteola
That’s right, AMR now supports former taxonomic names with an own ID, for users who do not want to transform old names to currently used names.
For this reason, we added the ‘keepsynonyms’ argument to as.mo()
and all `mo*()` functions. This value defaults to FALSE to reflect the old behaviour (so not much should change for users).
I should check your whole list, but a lot of your microorganisms never had a name change, such as E. cloacae (first on your list). I’ll get back to you!
Thanks, I was not aware of the keep synonyms addition. Indeed, the problems with the old naming appear only for some Candida species. This means that the Candida auris issue is solved by using the beta version 🥳 and the naming by using the keep synonyms.
👍🏼
Hello Matthijs,
Thank you for continuing to maintain and update this great package.
The problem is that as.mo("Candida auris") is wrongly identified as just Candida and is treated as an uncertain result.
> mo_uncertainties() "Candida auris" -> Candida (F_CANDD, 0.143)
I checked the library's microorganism dataset and Candida auris is present there.
AMR package version 1.8.2