2018-06-08 12:06:54 +02:00
# ==================================================================== #
# TITLE #
# Antimicrobial Resistance (AMR) Analysis #
# #
# AUTHORS #
# Berends MS (m.s.berends@umcg.nl), Luz CF (c.f.luz@umcg.nl) #
# #
# LICENCE #
# This program is free software; you can redistribute it and/or modify #
# it under the terms of the GNU General Public License version 2.0, #
# as published by the Free Software Foundation. #
# #
# This program is distributed in the hope that it will be useful, #
# but WITHOUT ANY WARRANTY; without even the implied warranty of #
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the #
# GNU General Public License for more details. #
# ==================================================================== #
2018-08-31 13:36:19 +02:00
#' Transform to microorganism ID
2018-06-08 12:06:54 +02:00
#'
2018-09-24 23:33:29 +02:00
#' Use this function to determine a valid microorganism ID (\code{mo}). Determination is done using Artificial Intelligence (AI) and the complete taxonomic kingdoms \emph{Bacteria}, \emph{Fungi} and \emph{Protozoa} (see Source), so the input can be almost anything: a full name (like \code{"Staphylococcus aureus"}), an abbreviated name (like \code{"S. aureus"}), an abbreviation known in the field (like \code{"MRSA"}), or just a genus. You could also \code{\link{select}} a genus and species column, zie Examples.
2018-09-04 11:33:30 +02:00
#' @param x a character vector or a \code{data.frame} with one or two columns
#' @param Becker a logical to indicate whether \emph{Staphylococci} should be categorised into Coagulase Negative \emph{Staphylococci} ("CoNS") and Coagulase Positive \emph{Staphylococci} ("CoPS") instead of their own species, according to Karsten Becker \emph{et al.} [1].
#'
#' This excludes \emph{Staphylococcus aureus} at default, use \code{Becker = "all"} to also categorise \emph{S. aureus} as "CoPS".
2018-09-24 23:33:29 +02:00
#' @param Lancefield a logical to indicate whether beta-haemolytic \emph{Streptococci} should be categorised into Lancefield groups instead of their own species, according to Rebecca C. Lancefield [2]. These \emph{Streptococci} will be categorised in their first group, e.g. \emph{Streptococcus dysgalactiae} will be group C, although officially it was also categorised into groups G and L.
2018-09-04 11:33:30 +02:00
#'
#' This excludes \emph{Enterococci} at default (who are in group D), use \code{Lancefield = "all"} to also categorise all \emph{Enterococci} as group D.
2018-09-24 23:33:29 +02:00
#' @param allow_uncertain a logical to indicate whether empty results should be checked for only a part of the input string. When results are found, a warning will be given about the uncertainty and the result.
2018-08-31 13:36:19 +02:00
#' @rdname as.mo
#' @aliases mo
#' @keywords mo Becker becker Lancefield lancefield guess
2018-09-24 23:33:29 +02:00
#' @details
#' A microbial ID (class: \code{mo}) typically looks like these examples:\cr
#' \preformatted{
#' Code Full name
#' --------------- --------------------------------------
#' B_KLBSL Klebsiella
#' B_KLBSL_PNE Klebsiella pneumoniae
#' B_KLBSL_PNE_RHI Klebsiella pneumoniae rhinoscleromatis
#' | | | |
#' | | | |
#' | | | ----> subspecies, a 3-4 letter acronym
#' | | ----> species, a 3-4 letter acronym
#' | ----> genus, a 5-7 letter acronym, mostly without vowels
#' ----> taxonomic kingdom, either Bacteria (B), Fungi (F) or Protozoa (P)
#' }
2018-08-01 08:03:31 +02:00
#'
2018-09-04 11:33:30 +02:00
#' Use the \code{\link{mo_property}} functions to get properties based on the returned code, see Examples.
2018-08-28 13:51:13 +02:00
#'
2018-09-24 23:33:29 +02:00
#' This function uses Artificial Intelligence (AI) to help getting more logical results, based on type of input and known prevalence of human pathogens. For example:
2018-07-23 14:14:03 +02:00
#' \itemize{
#' \item{\code{"E. coli"} will return the ID of \emph{Escherichia coli} and not \emph{Entamoeba coli}, although the latter would alphabetically come first}
2018-09-16 16:43:29 +02:00
#' \item{\code{"H. influenzae"} will return the ID of \emph{Haemophilus influenzae} and not \emph{Haematobacter influenzae} for the same reason}
2018-07-23 14:14:03 +02:00
#' \item{Something like \code{"p aer"} will return the ID of \emph{Pseudomonas aeruginosa} and not \emph{Pasteurella aerogenes}}
2018-09-16 16:43:29 +02:00
#' \item{Something like \code{"stau"} or \code{"S aur"} will return the ID of \emph{Staphylococcus aureus} and not \emph{Staphylococcus auricularis}}
2018-07-23 14:14:03 +02:00
#' }
2018-09-24 23:33:29 +02:00
#' This means that looking up human non-pathogenic microorganisms takes a longer time compares to human pathogenic microorganisms.
#'
#' \code{guess_mo} is an alias of \code{as.mo}.
#' @section ITIS:
#' \if{html}{\figure{itis_logo.jpg}{options: height=60px style=margin-bottom:5px} \cr}
#' This \code{AMR} package contains the \strong{complete microbial taxonomic data} from the publicly available Integrated Taxonomic Information System (ITIS, https://www.itis.gov). ITIS is a partnership of U.S., Canadian, and Mexican agencies and taxonomic specialists [3]. The complete taxonomic kingdoms Bacteria, Fungi and Protozoa (from subkingdom to the subspecies level) are included in this package.
# (source as section, so it can be inherited by mo_property:)
#' @section Source:
2018-09-04 11:33:30 +02:00
#' [1] Becker K \emph{et al.} \strong{Coagulase-Negative Staphylococci}. 2014. Clin Microbiol Rev. 27(4): 870– 926. \url{https://dx.doi.org/10.1128/CMR.00109-13}
#'
#' [2] Lancefield RC \strong{A serological differentiation of human and other groups of hemolytic streptococci}. 1933. J Exp Med. 57(4): 571– 95. \url{https://dx.doi.org/10.1084/jem.57.4.571}
2018-09-24 23:33:29 +02:00
#'
#' [3] Integrated Taxonomic Information System (ITIS). Retrieved September 2018. \url{http://www.itis.gov}
2018-06-08 12:06:54 +02:00
#' @export
2018-09-24 23:33:29 +02:00
#' @importFrom dplyr %>% pull left_join
#' @importFrom data.table as.data.table setkey
2018-08-31 13:36:19 +02:00
#' @return Character (vector) with class \code{"mo"}. Unknown values will return \code{NA}.
2018-09-24 23:33:29 +02:00
#' @seealso \code{\link{microorganisms}} for the \code{data.frame} with ITIS content that is being used to determine ID's. \cr
#' The \code{\link{mo_property}} functions (like \code{\link{mo_genus}}, \code{\link{mo_gramstain}}) to get properties based on the returned code.
2018-06-08 12:06:54 +02:00
#' @examples
#' # These examples all return "STAAUR", the ID of S. aureus:
2018-08-31 13:36:19 +02:00
#' as.mo("stau")
#' as.mo("STAU")
#' as.mo("staaur")
#' as.mo("S. aureus")
#' as.mo("S aureus")
#' as.mo("Staphylococcus aureus")
#' as.mo("MRSA") # Methicillin Resistant S. aureus
#' as.mo("VISA") # Vancomycin Intermediate S. aureus
#' as.mo("VRSA") # Vancomycin Resistant S. aureus
2018-09-24 23:33:29 +02:00
#' as.mo(369) # Search on TSN (Taxonomic Serial Number), a unique identifier
#' # for the Integrated Taxonomic Information System (ITIS)
2018-06-08 12:06:54 +02:00
#'
2018-09-05 10:51:46 +02:00
#' as.mo("Streptococcus group A")
#' as.mo("GAS") # Group A Streptococci
#' as.mo("GBS") # Group B Streptococci
#'
2018-08-31 13:36:19 +02:00
#' # guess_mo is an alias of as.mo and works the same
2018-09-24 23:33:29 +02:00
#' guess_mo("S. epidermidis") # will remain species: B_STPHY_EPI
#' guess_mo("S. epidermidis", Becker = TRUE) # will not remain species: B_STPHY_CNS
2018-08-02 13:15:45 +02:00
#'
2018-09-24 23:33:29 +02:00
#' guess_mo("S. pyogenes") # will remain species: B_STRPTC_PYO
#' guess_mo("S. pyogenes", Lancefield = TRUE) # will not remain species: B_STRPTC_GRA
2018-08-02 13:15:45 +02:00
#'
2018-08-31 13:36:19 +02:00
#' # Use mo_* functions to get a specific property based on `mo`
2018-09-24 23:33:29 +02:00
#' Ecoli <- as.mo("E. coli") # returns `B_ESCHR_COL`
2018-08-28 13:51:13 +02:00
#' mo_genus(Ecoli) # returns "Escherichia"
2018-09-24 23:33:29 +02:00
#' mo_gramstain(Ecoli) # returns "Gram negative"
#' # but it uses as.mo internally too, so you could also just use:
#' mo_genus("E. coli") # returns "Escherichia"
#'
2018-08-28 13:51:13 +02:00
#'
2018-06-08 12:06:54 +02:00
#' \dontrun{
2018-08-31 13:36:19 +02:00
#' df$mo <- as.mo(df$microorganism_name)
2018-06-08 12:06:54 +02:00
#'
#' # the select function of tidyverse is also supported:
2018-07-23 14:14:03 +02:00
#' library(dplyr)
2018-08-31 13:36:19 +02:00
#' df$mo <- df %>%
2018-07-23 14:14:03 +02:00
#' select(microorganism_name) %>%
2018-08-31 13:36:19 +02:00
#' guess_mo()
2018-06-08 12:06:54 +02:00
#'
#' # and can even contain 2 columns, which is convenient for genus/species combinations:
2018-08-31 13:36:19 +02:00
#' df$mo <- df %>%
2018-07-23 14:14:03 +02:00
#' select(genus, species) %>%
2018-08-31 13:36:19 +02:00
#' guess_mo()
2018-07-23 14:14:03 +02:00
#'
2018-06-08 12:06:54 +02:00
#' # same result:
2018-07-23 14:14:03 +02:00
#' df <- df %>%
2018-08-31 13:36:19 +02:00
#' mutate(mo = guess_mo(paste(genus, species)))
2018-06-08 12:06:54 +02:00
#' }
2018-09-24 23:33:29 +02:00
as.mo <- function ( x , Becker = FALSE , Lancefield = FALSE , allow_uncertain = FALSE ) {
2018-07-23 14:14:03 +02:00
2018-06-08 12:06:54 +02:00
if ( NCOL ( x ) == 2 ) {
# support tidyverse selection like: df %>% select(colA, colB)
# paste these columns together
x_vector <- vector ( " character" , NROW ( x ) )
for ( i in 1 : NROW ( x ) ) {
x_vector [i ] <- paste ( pull ( x [i , ] , 1 ) , pull ( x [i , ] , 2 ) , sep = " " )
}
x <- x_vector
} else {
if ( NCOL ( x ) > 2 ) {
stop ( ' `x` can be 2 columns at most' , call. = FALSE )
}
2018-09-09 12:11:44 +02:00
x [is.null ( x ) ] <- NA
2018-07-23 14:14:03 +02:00
2018-06-08 12:06:54 +02:00
# support tidyverse selection like: df %>% select(colA)
if ( ! is.vector ( x ) ) {
x <- pull ( x , 1 )
}
}
2018-09-24 23:33:29 +02:00
MOs <- as.data.table ( AMR :: microorganisms )
setkey ( MOs , prevalence , tsn )
MOs_mostprevalent <- MOs [prevalence != 9999 , ]
MOs_allothers <- NULL # will be set later, if needed
MOs_old <- NULL # will be set later, if needed
if ( all ( unique ( x ) %in% MOs [ , mo ] ) ) {
class ( x ) <- " mo"
attr ( x , ' package' ) <- ' AMR'
attr ( x , ' ITIS' ) <- TRUE
return ( x )
}
if ( AMR :: is.mo ( x ) & isTRUE ( attributes ( x ) $ ITIS ) ) {
# check for new mo class, data coming from ITIS
return ( x )
}
2018-08-28 13:51:13 +02:00
failures <- character ( 0 )
x_input <- x
# only check the uniques, which is way faster
x <- unique ( x )
2018-09-24 23:33:29 +02:00
x_backup <- trimws ( x , which = " both" )
x_species <- paste ( x_backup , " species" )
2018-09-08 16:06:47 +02:00
# translate to English for supported languages of mo_property
2018-09-10 15:45:25 +02:00
x <- gsub ( " (Gruppe|gruppe|groep|grupo|gruppo|groupe)" , " group" , x )
2018-09-09 12:11:44 +02:00
# remove 'empty' genus and species values
x <- gsub ( " (no MO)" , " " , x , fixed = TRUE )
2018-06-08 12:06:54 +02:00
# remove dots and other non-text in case of "E. coli" except spaces
2018-09-10 11:40:54 +02:00
x <- gsub ( " [^a-zA-Z0-9/ \\-]+" , " " , x )
2018-06-08 12:06:54 +02:00
# but spaces before and after should be omitted
x <- trimws ( x , which = " both" )
2018-08-28 13:51:13 +02:00
x_trimmed <- x
2018-09-24 23:33:29 +02:00
x_trimmed_species <- paste ( x_trimmed , " species" )
2018-06-08 12:06:54 +02:00
# replace space by regex sign
2018-08-02 13:15:45 +02:00
x_withspaces <- gsub ( " " , " .* " , x , fixed = TRUE )
2018-06-08 12:06:54 +02:00
x <- gsub ( " " , " .*" , x , fixed = TRUE )
2018-08-02 13:15:45 +02:00
# add start en stop regex
2018-06-08 12:06:54 +02:00
x <- paste0 ( ' ^' , x , ' $' )
2018-09-01 21:19:46 +02:00
x_withspaces_all <- x_withspaces
2018-09-06 16:50:03 +02:00
x_withspaces_start <- paste0 ( ' ^' , x_withspaces )
2018-08-02 13:15:45 +02:00
x_withspaces <- paste0 ( ' ^' , x_withspaces , ' $' )
2018-06-08 12:06:54 +02:00
2018-09-16 16:43:29 +02:00
# cat(paste0('x "', x, '"\n'))
2018-09-24 23:33:29 +02:00
# cat(paste0('x_species "', x_species, '"\n'))
2018-09-16 16:43:29 +02:00
# cat(paste0('x_withspaces_all "', x_withspaces_all, '"\n'))
# cat(paste0('x_withspaces_start "', x_withspaces_start, '"\n'))
# cat(paste0('x_withspaces "', x_withspaces, '"\n'))
# cat(paste0('x_backup "', x_backup, '"\n'))
2018-09-24 23:33:29 +02:00
# cat(paste0('x_trimmed "', x_trimmed, '"\n'))
# cat(paste0('x_trimmed_species "', x_trimmed_species, '"\n'))
2018-09-10 11:40:54 +02:00
2018-06-08 12:06:54 +02:00
for ( i in 1 : length ( x ) ) {
2018-09-24 23:33:29 +02:00
if ( identical ( x_trimmed [i ] , " " ) | is.na ( x_trimmed [i ] ) ) {
2018-07-30 00:14:06 +02:00
# empty values
x [i ] <- NA
next
}
2018-06-08 12:06:54 +02:00
2018-09-14 10:31:21 +02:00
# translate known trivial abbreviations to genus + species ----
2018-08-28 13:51:13 +02:00
if ( ! is.na ( x_trimmed [i ] ) ) {
if ( toupper ( x_trimmed [i ] ) == ' MRSA'
| toupper ( x_trimmed [i ] ) == ' VISA'
| toupper ( x_trimmed [i ] ) == ' VRSA' ) {
2018-09-24 23:33:29 +02:00
x [i ] <- ' B_STPHY_AUR'
2018-07-30 01:18:40 +02:00
next
2018-06-08 12:06:54 +02:00
}
2018-08-28 13:51:13 +02:00
if ( toupper ( x_trimmed [i ] ) == ' MRSE' ) {
2018-09-24 23:33:29 +02:00
x [i ] <- ' B_STPHY_EPI'
2018-08-01 08:03:31 +02:00
next
2018-06-08 12:06:54 +02:00
}
2018-08-28 13:51:13 +02:00
if ( toupper ( x_trimmed [i ] ) == ' VRE' ) {
2018-09-24 23:33:29 +02:00
x [i ] <- ' B_ENTRC'
2018-07-30 01:18:40 +02:00
next
2018-06-08 12:06:54 +02:00
}
2018-08-28 13:51:13 +02:00
if ( toupper ( x_trimmed [i ] ) == ' MRPA' ) {
2018-06-08 12:06:54 +02:00
# multi resistant P. aeruginosa
2018-09-24 23:33:29 +02:00
x [i ] <- ' B_PDMNS_AER'
2018-07-30 01:18:40 +02:00
next
2018-06-08 12:06:54 +02:00
}
2018-08-28 13:51:13 +02:00
if ( toupper ( x_trimmed [i ] ) %in% c ( ' PISP' , ' PRSP' , ' VISP' , ' VRSP' ) ) {
2018-09-04 11:33:30 +02:00
# peni I, peni R, vanco I, vanco R: S. pneumoniae
2018-09-24 23:33:29 +02:00
x [i ] <- ' B_STRPTC_PNE'
next
}
if ( toupper ( x_trimmed [i ] ) %like% ' ^G[ABCDFGHK]S$' ) {
x [i ] <- gsub ( " G([ABCDFGHK])S" , " B_STRPTC_GR\\1" , x_trimmed [i ] )
2018-08-01 09:45:11 +02:00
next
2018-06-08 12:06:54 +02:00
}
2018-09-24 23:33:29 +02:00
# CoNS/CoPS in different languages (support for German, Dutch, Spanish, Portuguese) ----
if ( tolower ( x [i ] ) %like% ' [ck]oagulas[ea] negatie?[vf]'
| tolower ( x_trimmed [i ] ) %like% ' [ck]oagulas[ea] negatie?[vf]'
| tolower ( x [i ] ) %like% ' [ck]o?ns[^a-z]?$' ) {
# coerce S. coagulase negative
x [i ] <- ' B_STPHY_CNS'
next
}
if ( tolower ( x [i ] ) %like% ' [ck]oagulas[ea] positie?[vf]'
| tolower ( x_trimmed [i ] ) %like% ' [ck]oagulas[ea] positie?[vf]'
| tolower ( x [i ] ) %like% ' [ck]o?ps[^a-z]?$' ) {
# coerce S. coagulase positive
x [i ] <- ' B_STPHY_CPS'
2018-09-05 10:51:46 +02:00
next
}
2018-06-08 12:06:54 +02:00
}
2018-09-24 23:33:29 +02:00
# FIRST TRY FULLNAMES AND CODES
# if only genus is available, don't select species
if ( all ( ! c ( x [i ] , x_trimmed [i ] ) %like% " " ) ) {
found <- MOs [tolower ( fullname ) %in% tolower ( c ( x_species [i ] , x_trimmed_species [i ] ) ) , mo ]
if ( length ( found ) > 0 ) {
x [i ] <- found [1L ]
next
}
if ( nchar ( x_trimmed [i ] ) > 4 ) {
# not when abbr is esco, stau, klpn, etc.
found <- MOs [tolower ( fullname ) %like% gsub ( " " , " .*" , x_trimmed_species [i ] , fixed = TRUE ) , mo ]
if ( length ( found ) > 0 ) {
x [i ] <- found [1L ]
next
}
}
}
# search for GLIMS code ----
found <- AMR :: microorganisms.umcg [which ( toupper ( AMR :: microorganisms.umcg $ umcg ) == toupper ( x_trimmed [i ] ) ) , ] $ mo
2018-08-02 13:15:45 +02:00
if ( length ( found ) > 0 ) {
2018-09-24 23:33:29 +02:00
x [i ] <- MOs [mo.old == found , mo ] [1L ]
2018-08-02 13:15:45 +02:00
next
}
2018-09-14 11:54:01 +02:00
2018-09-24 23:33:29 +02:00
# TRY FIRST THOUSAND MOST PREVALENT IN HUMAN INFECTIONS ----
found <- MOs_mostprevalent [tolower ( fullname ) %in% tolower ( c ( x_backup [i ] , x_trimmed [i ] ) ) , mo ]
# most probable: is exact match in fullname
if ( length ( found ) > 0 ) {
x [i ] <- found [1L ]
next
}
found <- MOs_mostprevalent [tsn == x_trimmed [i ] , mo ]
# is a valid TSN
if ( length ( found ) > 0 ) {
x [i ] <- found [1L ]
next
}
found <- MOs_mostprevalent [mo == toupper ( x_backup [i ] ) , mo ]
# is a valid mo
if ( length ( found ) > 0 ) {
x [i ] <- found [1L ]
next
}
found <- MOs_mostprevalent [mo.old == toupper ( x_backup [i ] )
| ( substr ( x_backup [i ] , 4 , 6 ) == " SPP" & mo.old == substr ( x_backup [i ] , 1 , 3 ) ) , mo ]
# is a valid old mo
2018-09-10 11:40:54 +02:00
if ( length ( found ) > 0 ) {
x [i ] <- found [1L ]
next
}
2018-09-14 11:54:01 +02:00
2018-09-24 23:33:29 +02:00
# try any match keeping spaces ----
found <- MOs_mostprevalent [fullname %like% x_withspaces [i ] , mo ]
2018-09-16 16:43:29 +02:00
if ( length ( found ) > 0 ) {
x [i ] <- found [1L ]
next
}
2018-09-14 10:31:21 +02:00
# try any match keeping spaces, not ending with $ ----
2018-09-24 23:33:29 +02:00
found <- MOs_mostprevalent [fullname %like% x_withspaces_start [i ] , mo ]
2018-09-06 16:50:03 +02:00
if ( length ( found ) > 0 ) {
x [i ] <- found [1L ]
next
}
2018-08-02 13:15:45 +02:00
2018-09-14 10:31:21 +02:00
# try any match diregarding spaces ----
2018-09-24 23:33:29 +02:00
found <- MOs_mostprevalent [fullname %like% x [i ] , mo ]
2018-07-30 00:14:06 +02:00
if ( length ( found ) > 0 ) {
x [i ] <- found [1L ]
next
}
2018-09-14 10:31:21 +02:00
# try fullname without start and stop regex, to also find subspecies ----
# like "K. pneu rhino" -> "Klebsiella pneumoniae (rhinoscleromatis)" = KLEPNERH
2018-09-24 23:33:29 +02:00
found <- MOs_mostprevalent [fullname %like% x_withspaces_start [i ] , mo ]
2018-09-01 21:19:46 +02:00
if ( length ( found ) > 0 ) {
x [i ] <- found [1L ]
next
}
2018-09-24 23:33:29 +02:00
# try splitting of characters and then find ID ----
# like esco = E. coli, klpn = K. pneumoniae, stau = S. aureus
x_split <- x
x_length <- nchar ( x_trimmed [i ] )
x_split [i ] <- paste0 ( x_trimmed [i ] %>% substr ( 1 , x_length / 2 ) %>% trimws ( ) ,
' .* ' ,
x_trimmed [i ] %>% substr ( ( x_length / 2 ) + 1 , x_length ) %>% trimws ( ) )
found <- MOs_mostprevalent [fullname %like% paste0 ( ' ^' , x_split [i ] ) , mo ]
if ( length ( found ) > 0 ) {
x [i ] <- found [1L ]
next
}
# try any match with text before and after original search string ----
# so "negative rods" will be "GNR"
# if (x_trimmed[i] %like% "^Gram") {
# x_trimmed[i] <- gsub("^Gram", "", x_trimmed[i], ignore.case = TRUE)
# # remove leading and trailing spaces again
# x_trimmed[i] <- trimws(x_trimmed[i], which = "both")
# }
# if (!is.na(x_trimmed[i])) {
# found <- MOs_mostprevalent[fullname %like% x_trimmed[i], mo]
# if (length(found) > 0) {
# x[i] <- found[1L]
# next
# }
# }
# THEN TRY ALL OTHERS ----
if ( is.null ( MOs_allothers ) ) {
MOs_allothers <- MOs [prevalence == 9999 , ]
}
found <- MOs_allothers [tolower ( fullname ) == tolower ( x_backup [i ] ) , mo ]
# most probable: is exact match in fullname
if ( length ( found ) > 0 ) {
x [i ] <- found [1L ]
next
}
found <- MOs_allothers [tolower ( fullname ) == tolower ( x_trimmed [i ] ) , mo ]
# most probable: is exact match in fullname
if ( length ( found ) > 0 ) {
x [i ] <- found [1L ]
next
}
found <- MOs_allothers [tsn == x_trimmed [i ] , mo ]
# is a valid TSN
if ( length ( found ) > 0 ) {
x [i ] <- found [1L ]
next
}
found <- MOs_allothers [mo == toupper ( x_backup [i ] ) , mo ]
# is a valid mo
if ( length ( found ) > 0 ) {
x [i ] <- found [1L ]
next
}
found <- MOs_allothers [mo.old == toupper ( x_backup [i ] ) , mo ]
# is a valid old mo
if ( length ( found ) > 0 ) {
x [i ] <- found [1L ]
next
}
# try any match keeping spaces ----
found <- MOs_allothers [fullname %like% x_withspaces [i ] , mo ]
if ( length ( found ) > 0 ) {
x [i ] <- found [1L ]
next
}
# try any match keeping spaces, not ending with $ ----
found <- MOs_allothers [fullname %like% x_withspaces_start [i ] , mo ]
if ( length ( found ) > 0 ) {
x [i ] <- found [1L ]
next
}
# try any match diregarding spaces ----
found <- MOs_allothers [fullname %like% x [i ] , mo ]
if ( length ( found ) > 0 ) {
x [i ] <- found [1L ]
next
}
# try fullname without start and stop regex, to also find subspecies ----
# like "K. pneu rhino" -> "Klebsiella pneumoniae (rhinoscleromatis)" = KLEPNERH
found <- MOs_allothers [fullname %like% x_withspaces_start [i ] , mo ]
2018-07-30 00:14:06 +02:00
if ( length ( found ) > 0 ) {
x [i ] <- found [1L ]
next
2018-06-08 12:06:54 +02:00
}
2018-07-30 00:14:06 +02:00
2018-09-14 10:31:21 +02:00
# try splitting of characters and then find ID ----
2018-07-30 00:14:06 +02:00
# like esco = E. coli, klpn = K. pneumoniae, stau = S. aureus
x_split <- x
2018-08-28 13:51:13 +02:00
x_length <- nchar ( x_trimmed [i ] )
x_split [i ] <- paste0 ( x_trimmed [i ] %>% substr ( 1 , x_length / 2 ) %>% trimws ( ) ,
2018-07-30 00:14:06 +02:00
' .* ' ,
2018-08-28 13:51:13 +02:00
x_trimmed [i ] %>% substr ( ( x_length / 2 ) + 1 , x_length ) %>% trimws ( ) )
2018-09-24 23:33:29 +02:00
found <- MOs_allothers [fullname %like% paste0 ( ' ^' , x_split [i ] ) , mo ]
2018-07-30 00:14:06 +02:00
if ( length ( found ) > 0 ) {
x [i ] <- found [1L ]
next
2018-06-08 12:06:54 +02:00
}
2018-07-30 00:14:06 +02:00
2018-09-24 23:33:29 +02:00
# # try any match with text before and after original search string ----
# # so "negative rods" will be "GNR"
# if (x_trimmed[i] %like% "^Gram") {
# x_trimmed[i] <- gsub("^Gram", "", x_trimmed[i], ignore.case = TRUE)
# # remove leading and trailing spaces again
# x_trimmed[i] <- trimws(x_trimmed[i], which = "both")
# }
# if (!is.na(x_trimmed[i])) {
# found <- MOs_allothers[fullname %like% x_trimmed[i], mo]
# if (length(found) > 0) {
# x[i] <- found[1L]
# next
# }
# }
# MISCELLANEOUS ----
# look for old taxonomic names ----
if ( is.null ( MOs_old ) ) {
MOs_old <- as.data.table ( microorganisms.old )
setkey ( MOs_old , name , tsn_new )
2018-06-08 12:06:54 +02:00
}
2018-09-24 23:33:29 +02:00
found <- MOs_old [tolower ( name ) == tolower ( x_backup [i ] ) |
tsn == x_trimmed [i ] , ]
if ( NROW ( found ) > 0 ) {
x [i ] <- MOs [tsn == found [1 , tsn_new ] , mo ]
message ( " Note: '" , found [1 , name ] , " ' was renamed to '" ,
MOs [tsn == found [1 , tsn_new ] , fullname ] , " ' by " ,
found [1 , authors ] , " in " , found [1 , year ] )
next
}
# check for uncertain results ----
# (1) try to strip off one element and check the remains
if ( allow_uncertain == TRUE ) {
x_strip <- x_backup [i ] %>% strsplit ( " " ) %>% unlist ( )
x_strip <- x_strip [1 : length ( x_strip ) - 1 ]
x [i ] <- suppressWarnings ( suppressMessages ( as.mo ( x_strip ) ) )
if ( ! is.na ( x [i ] ) ) {
warning ( " Uncertain result: '" , x_backup [i ] , " ' -> '" , MOs [mo == x [i ] , fullname ] , " ' (" , x [i ] , " )" )
2018-07-30 00:14:06 +02:00
next
2018-06-08 12:06:54 +02:00
}
}
2018-09-14 10:31:21 +02:00
# not found ----
2018-07-30 00:14:06 +02:00
x [i ] <- NA_character_
2018-08-28 13:51:13 +02:00
failures <- c ( failures , x_backup [i ] )
2018-07-30 00:14:06 +02:00
2018-06-08 12:06:54 +02:00
}
2018-07-23 14:14:03 +02:00
failures <- failures [ ! failures %in% c ( NA , NULL , NaN ) ]
if ( length ( failures ) > 0 ) {
2018-09-24 23:33:29 +02:00
warning ( " These " , length ( failures ) , " values could not be coerced (try again with allow_uncertain = TRUE):\n" ,
2018-07-23 14:14:03 +02:00
paste ( ' "' , unique ( failures ) , ' "' , sep = " " , collapse = ' , ' ) ,
" ." ,
call. = FALSE )
}
2018-08-28 13:51:13 +02:00
2018-09-14 10:31:21 +02:00
# Becker ----
2018-09-01 21:19:46 +02:00
if ( Becker == TRUE | Becker == " all" ) {
# See Source. It's this figure:
# https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4187637/figure/F3/
CoNS <- MOs %>%
filter ( genus == " Staphylococcus" ,
species %in% c ( " arlettae" , " auricularis" , " capitis" ,
" caprae" , " carnosus" , " cohnii" , " condimenti" ,
" devriesei" , " epidermidis" , " equorum" ,
" fleurettii" , " gallinarum" , " haemolyticus" ,
" hominis" , " jettensis" , " kloosii" , " lentus" ,
" lugdunensis" , " massiliensis" , " microti" ,
" muscae" , " nepalensis" , " pasteuri" , " petrasii" ,
" pettenkoferi" , " piscifermentans" , " rostri" ,
" saccharolyticus" , " saprophyticus" , " sciuri" ,
" stepanovicii" , " simulans" , " succinus" ,
" vitulinus" , " warneri" , " xylosus" ) ) %>%
pull ( mo )
CoPS <- MOs %>%
filter ( genus == " Staphylococcus" ,
species %in% c ( " simiae" , " agnetis" , " chromogenes" ,
" delphini" , " felis" , " lutrae" ,
" hyicus" , " intermedius" ,
" pseudintermedius" , " pseudointermedius" ,
" schleiferi" ) ) %>%
pull ( mo )
2018-09-24 23:33:29 +02:00
x [x %in% CoNS ] <- " B_STPHY_CNS"
x [x %in% CoPS ] <- " B_STPHY_CPS"
2018-09-01 21:19:46 +02:00
if ( Becker == " all" ) {
2018-09-24 23:33:29 +02:00
x [x == " B_STPHY_AUR" ] <- " B_STPHY_CPS"
2018-09-01 21:19:46 +02:00
}
}
2018-09-14 10:31:21 +02:00
# Lancefield ----
2018-09-04 11:33:30 +02:00
if ( Lancefield == TRUE | Lancefield == " all" ) {
2018-09-01 21:19:46 +02:00
# group A
2018-09-24 23:33:29 +02:00
x [x == " B_STRPTC_PYO" ] <- " B_STRPTC_GRA" # S. pyogenes
2018-09-01 21:19:46 +02:00
# group B
2018-09-24 23:33:29 +02:00
x [x == " B_STRPTC_AGA" ] <- " B_STRPTC_GRB" # S. agalactiae
2018-09-01 21:19:46 +02:00
# group C
S_groupC <- MOs %>% filter ( genus == " Streptococcus" ,
species %in% c ( " equisimilis" , " equi" ,
" zooepidemicus" , " dysgalactiae" ) ) %>%
pull ( mo )
2018-09-24 23:33:29 +02:00
x [x %in% S_groupC ] <- " B_STRPTC_GRC" # S. agalactiae
2018-09-04 11:33:30 +02:00
if ( Lancefield == " all" ) {
2018-09-24 23:33:29 +02:00
x [substr ( x , 1 , 7 ) == " B_ENTRC" ] <- " B_STRPTC_GRD" # all Enterococci
2018-09-04 11:33:30 +02:00
}
2018-09-01 21:19:46 +02:00
# group F
2018-09-24 23:33:29 +02:00
x [x == " B_STRPTC_ANG" ] <- " B_STRPTC_GRF" # S. anginosus
2018-09-01 21:19:46 +02:00
# group H
2018-09-24 23:33:29 +02:00
x [x == " B_STRPTC_SAN" ] <- " B_STRPTC_GRH" # S. sanguinis
2018-09-01 21:19:46 +02:00
# group K
2018-09-24 23:33:29 +02:00
x [x == " B_STRPTC_SAL" ] <- " B_STRPTC_GRK" # S. salivarius
2018-09-01 21:19:46 +02:00
}
2018-08-28 13:51:13 +02:00
# left join the found results to the original input values (x_input)
df_found <- data.frame ( input = as.character ( unique ( x_input ) ) ,
found = x ,
stringsAsFactors = FALSE )
df_input <- data.frame ( input = as.character ( x_input ) ,
stringsAsFactors = FALSE )
x <- df_input %>%
left_join ( df_found ,
by = " input" ) %>%
pull ( found )
2018-08-31 13:36:19 +02:00
class ( x ) <- " mo"
2018-07-23 14:14:03 +02:00
attr ( x , ' package' ) <- ' AMR'
2018-09-24 23:33:29 +02:00
attr ( x , ' ITIS' ) <- TRUE
2018-06-08 12:06:54 +02:00
x
}
2018-07-23 14:14:03 +02:00
2018-09-24 23:33:29 +02:00
2018-08-31 13:36:19 +02:00
#' @rdname as.mo
2018-07-23 14:14:03 +02:00
#' @export
2018-08-31 13:36:19 +02:00
is.mo <- function ( x ) {
# bactid for older releases
# remove when is.bactid will be removed
identical ( class ( x ) , " mo" ) | identical ( class ( x ) , " bactid" )
}
#' @rdname as.mo
#' @export
guess_mo <- as.mo
#' @exportMethod print.mo
#' @export
#' @noRd
print.mo <- function ( x , ... ) {
cat ( " Class 'mo'\n" )
print.default ( as.character ( x ) , quote = FALSE )
}
2018-07-23 14:14:03 +02:00
2018-08-31 13:36:19 +02:00
#' @exportMethod as.data.frame.mo
2018-07-23 14:14:03 +02:00
#' @export
2018-08-31 13:36:19 +02:00
#' @noRd
as.data.frame.mo <- function ( x , ... ) {
# same as as.data.frame.character but with removed stringsAsFactors
nm <- paste ( deparse ( substitute ( x ) , width.cutoff = 500L ) ,
collapse = " " )
if ( ! " nm" %in% names ( list ( ... ) ) ) {
as.data.frame.vector ( x , ... , nm = nm )
} else {
as.data.frame.vector ( x , ... )
}
}
#' @exportMethod pull.mo
#' @export
#' @importFrom dplyr pull
#' @noRd
pull.mo <- function ( .data , ... ) {
pull ( as.data.frame ( .data ) , ... )
2018-07-23 14:14:03 +02:00
}