R/Functions_SIRIUS.R
In mjhelf/Mosaic: Metaboseek platform
Defines functions
getSiriusTree
getSirius
runSirius
writeMS
Documented in
getSirius
getSiriusTree
runSirius
writeMS
#' writeMS
#'
#' write an .ms file containing a spectrum to be exported to SIRIUS.
#' will append additional spectrum to file if filename is that of an existing file.
#'
#' @return Returns nothing, but writes a SIRIUS \code{.ms} file.
#'
#' @param filename path of file to write to
#' @param ms2 a matrix with columns mz and intensity, or a list of ms2 spectra
#' in the same format
#' @param ms1 a matrix with columns mz and intensity for a single MS1 spectrum
#' @param parentmz numeric: parent ion m/z value
#' @param comments character: comments
#' @param rt numeric: retention time in seconds
#' @param ion ion type (e.g. \eqn{[M+H]+})
#' @param charge charge (positive or negative integer)
#' @param hashtag hash tag of the ms2 spectrum
#' @param scanindices which scans were averaged into this ms2 spectrum
#'
#' @importFrom digest digest
#' @importFrom data.table fwrite
#'
writeMS <- function(filename,
ms2,
ms1 = NULL,
parentmz,
comments = "",
rt = "",
ion = NULL,
charge = 1,
hashtag = NULL,
scanindices = NULL){
if(is.null(hashtag)){hashtag <- digest(ms2, algo = "xxhash64")}
appending <- file.exists(filename)
write(paste(
if(appending){"\n"}else{paste0('# Generated with MOSAiC version ',
packageVersion("Metaboseek"),', timestamp: ',
strftime(Sys.time(),"%Y%m%d_%H%M%S"))},
paste0(">compound ", hashtag),
paste0(">parentmass ", parentmz),
if(!is.null(ion)){paste0(">ion ", ion)}else{paste0(">charge ", charge)},
if(!is.null(scanindices)){paste0(">scans ", scanindices)}else{""},
paste0(">hash ", hashtag),
paste0("# Comments: ", comments),
paste0(">rt ", rt),
sep = "\n"),
file = filename,
append = appending
)
if(!is.null(ms1)){
write(
paste0(">ms1"),
file = filename,
append = T
)
fwrite(data.table(ms1), filename, append = T, sep = "\t",
row.names = F, col.names = F, quote = T, eol = "\n")
}
if(!is.null(ms2)){
if(!is.list(ms2)){ms2 <- list(ms2)}
lapply(seq(length(ms2)), function(n){
#expect collision energy to be in names, resulting in ">collision20", etc.
write(
if(is.null(names(ms2))){">ms2"}else{paste0(">",names(ms2)[n])},
file = filename,
append = T
)
fwrite(data.table(ms2[[n]]), filename, append = T, sep = "\t",
row.names = F, col.names = F, quote = T, eol = "\n")
})
}
}
#' runSirius
#'
#' run SIRIUS externally for a list of ms2 spectra
#'
#' @param outfolder output folder for the SIRIUS results
#' @param ms1 if not NULL, a list of MS1 spectra
#' @param ms2 a list (or list of lists) of ms2 spectra
#' (matrices with columns mz and intensity), see \code{Details}
#' @param parentmz numeric: parent ion m/z value
#' @param comments character: comments
#' @param rt numeric: retention time in seconds
#' @param fingerid if TRUE, will try to predict structures with CSI:FingerID
#' @param ion ion type (e.g. \eqn{[M+H]+})
#' @param charge charge (positive or negative integer)
#' @param scanindices which scans were averaged into this ms2 spectrum
#' @param sirpath path to SIRIUS executable
#' @param moreOpts character with additional options to be passed to SIRIUS
#' @param force force calculation, even if same results should exist according to indexfile
#' @param config list containing sirius configuration
#'
#' @return Will cause SIRIUS to run a job and save its results in
#' a timestamped subfolder of \code{outfolder}, and will register the job in
#' the \code{index.csv} file in \code{outfolder}.
#'
#' @details
#' \describe{
#' \item{ms2}{can be a list of matrices, or a list of list of matrices.
#' In the latter case, the (optional) names of the nested list items are
#' expected to be in the format "collisionXX", where XX denotes an integer
#' collision energy.}
#' }
#'
#' @importFrom data.table fread fwrite data.table
#' @importFrom digest digest
#'
#
# ms1 = list(matrix(c(285.253723144531, 3128961.25, 286.256988525391, 469789.125),
# dimnames = list(rows = NULL,
# cols = c('mz','intensity')),
# ncol = 2,
# byrow = TRUE))
#
# ms2 = list(list(collision30 = matrix(c(84.0183,114.09156, 143.117874,443266,165303,516731),
# dimnames = list(rows = NULL,
# cols = c('mz','intensity')),
# ncol = 2)))
#
# runSirius( sirpath = "C:/Runspace/sirius-4.8.2-win64/sirius-gui/sirius",
# ms2 = ms2,
# parentmz = 285.253723144531,
# outfolder = "C:/Runspace/sirius-4.8.2-win64/sirius-gui/Metaboseek",
# ms1 = ms1,
# ion = "[M+?]+",
# comments = "",
# rt = "",
# charge= 1,
# fingerid = T,
# scanindices = 111111,
# moreOpts = "",
# config = list(IsotopeSettings.filter = TRUE,
# FormulaSearchDB = c('BIO'),
# Timeout.secondsPerTree = 0,
# FormulaSettings.enforced = "HCNOP[5]S",
# Timeout.secondsPerInstance = 0,
# AdductSettings.detectable = ",",
# UseHeuristic.mzToUseHeuristicOnly = 650,
# AlgorithmProfile = list("orbitrap"), #qtof
# IsotopeMs2Settings = "IGNORE",
# MS2MassDeviation.allowedMassDeviation = '5.0ppm',
# NumberOfCandidatesPerIon = 5,
# UseHeuristic.mzToUseHeuristic = 300,
# FormulaSettings.detectable = ",",
# NumberOfCandidates = 10,
# StructureSearchDB = c('BIO'),
# AdductSettings.fallback = ",",
# RecomputeResults = TRUE),
# force = T)
#'
#'
#' @export
runSirius <- function( sirpath = file.path(.MseekOptions$siriusFolder, "sirius"),
ms2,
parentmz,
outfolder = getwd(),
ms1 = NULL,
ion = "[M+?]+",
comments = "",
rt = "",
charge= 1,
fingerid = T,
scanindices = "",
moreOpts = "",
config = list(IsotopeSettings.filter = TRUE,
FormulaSearchDB = c('ALL_BUT_INSILICO','ALL','BIO',
'METACYC','CHEBI','COCONUT',
'ECOCYCMINE','GNPS','HMDB',
'HSDB','KEGG','KEGGMINE',
'KNAPSACK','MACONDA','MESH',
'NORMAN','UNDP','PLANTCYC',
'PUBCHEM','PUBMED','YMDB',
'YMDBMINE','ZINCBIO'),
Timeout.secondsPerTree = 0,
FormulaSettings.enforced = "HCNOP[5]S",
Timeout.secondsPerInstance = 0,
AdductSettings.detectable = "[[M+K]+,[M+H3N+H]+,[M+Na]+,[M-H4O2+H]+,[M-H2O+H]+,[M+H]+]",
UseHeuristic.mzToUseHeuristicOnly = 650,
AlgorithmProfile = "orbitrap", #qtof
IsotopeMs2Settings = "IGNORE",
MS2MassDeviation.allowedMassDeviation = '5.0ppm',
NumberOfCandidatesPerIon = 1,
UseHeuristic.mzToUseHeuristic = 300,
FormulaSettings.detectable = ",",
NumberOfCandidates = 10,
StructureSearchDB = c('ALL_BUT_INSILICO','ALL','BIO',
'METACYC','CHEBI','COCONUT',
'ECOCYCMINE','GNPS','HMDB',
'HSDB','KEGG','KEGGMINE',
'KNAPSACK','MACONDA','MESH',
'NORMAN','UNDP','PLANTCYC',
'PUBCHEM','PUBMED','YMDB',
'YMDBMINE','ZINCBIO'),
AdductSettings.fallback = "[[M+K]+,[M+Na]+,[M+H]+]",
RecomputeResults = TRUE),
force = T){
#write the MS2 data so sirius can read it
dir.create(outfolder, showWarnings = F)
#hashtag <- lapply(ms2, gethash)
hashtag <- lapply(ms2, digest, algo = "xxhash64")
ts <- strftime(Sys.time(),"%y%m%d_%H%M%S")
#make a data.frame with information on what will be analyzed here
newjobs <- data.frame(mz = parentmz,
stringsAsFactors = F)
newjobs$rt = rt
newjobs$hash = unlist(hashtag)
newjobs$timestamp = ts
newjobs$ion = ion
newjobs$ms1hash = sapply(ms1, function(x){if(is.null(x)){""}else{digest(x,algo = "xxhash64")}})
newjobs$charge = charge
newjobs$fingerid = fingerid
newjobs$moreOpts = paste0(moreOpts)
newjobs$Metaboseek_version = as.character(packageVersion("Metaboseek"))
newjobs$Metaboseek_sirius_revision = 3
newjobs <- cbind(newjobs, as.data.frame(lapply(config,function(x){if(length(x) >1){paste(x, collapse = ",")}else{x}}), stringsAsFactors = FALSE))
newjobs$settingsHash <- apply(newjobs[,c("ion", "ms1hash", "charge",
"fingerid", "moreOpts",
"Metaboseek_sirius_revision",
names(config)),
drop = F], 1, digest,
algo = "xxhash64")
indexfile <- file.path(outfolder, "index.csv")
if(file.exists(indexfile) && !force){
jobindex <- as.data.frame(data.table::fread(indexfile, na.strings = NULL))
# for all columns that are of type logical and only contain NAs, assume they are mutilated empty character strings
# and a victim of type.convert - make them character vectors again
charCols <- sapply(jobindex,typeof) == "logical" & sapply(lapply(jobindex,is.na),all)
if(any(charCols)){
jobindex[,charCols] <- character(nrow(jobindex))
}
checkme <- as.data.frame(rbindlist(list(old = jobindex,
new = newjobs),
fill = T))[,c("mz", "hash",
"settingsHash")]
#To avoid problems with m/z digits possibly not retained when
#exporting and reimporting of siriusIndex to/from index.csv
checkme$mz <- round(checkme$mz, 4)
dups <- duplicated.data.frame(checkme)[seq(nrow(jobindex)+1,nrow(checkme))]
newjobs <- newjobs[!dups,]
}
if(nrow(newjobs) > 0){
filename <- file.path(outfolder,paste0(ts,".ms"))
np <- mapply(writeMS,
filename = filename,
ms1 = ms1,
ms2 = ms2,
parentmz = parentmz,
scanindices = scanindices,
comments = comments,
rt = rt,
hashtag = hashtag,
ion = ion,
charge = charge)
outfolder <- file.path(outfolder,ts)
### config --IsotopeSettings.filter true --FormulaSearchDB --Timeout.secondsPerTree 0 --FormulaSettings.enforced HCNOP[5]S --Timeout.secondsPerInstance 0 --AdductSettings.detectable [[M + K]+, [M + H3N + H]+, [M + Na]+, [M - H4O2 + H]+, [M - H2O + H]+, [M + H]+] --UseHeuristic.mzToUseHeuristicOnly 650 --AlgorithmProfile orbitrap --IsotopeMs2Settings IGNORE --MS2MassDeviation.allowedMassDeviation 5.0ppm --NumberOfCandidatesPerIon 1 --UseHeuristic.mzToUseHeuristic 300 --FormulaSettings.detectable , --NumberOfCandidates 10 --AdductSettings.fallback [[M + K]+, [M + H3N + H]+, [M + Na]+, [M - H4O2 + H]+, [M - H2O + H]+, [M + H]+] --RecomputeResults true formula
## --StructureSearchDB BIO
## --StructureSearchDB PUBCHEM
finger <- if(fingerid){"fingerid"}else{""}
configstring <- sapply(names(config),function(n){
x <- config[[n]]
if(is.logical(x)){
x <- tolower(as.character(x))
}
if(length(x) > 1){
x <- paste(x, collapse = ",")
}
paste0("--", n, " ", x)
})
configstring <- paste0("config ", paste(configstring, collapse = " "))
cmd <- paste0(
sirpath,
' -i "',
filename,
'"',
" -o ",
'"',
outfolder,
'" ',
configstring,
" tree ",
finger,
moreOpts
)
cmd2 <- paste(paste0(sirpath,
' -i "',
outfolder,
'" ftree-export --all --dot --output ',
file.path(outfolder,paste0(seq_len(length(hashtag))-1,"_",ts,"_",unlist(hashtag)), "trees")),
collapse = "\n")
commandfile <- file.path(dirname(outfolder),paste0(ts,".sh"))
if(Sys.info()['sysname'] == "Windows"){
commandfile <- gsub("\\.sh$",".cmd",commandfile)
}
writeLines(c(cmd,cmd2),
commandfile
)
if(Sys.info()['sysname'] == "Windows"){
# commandfile <- paste0("cmd /c ",'"', commandfile,'"')
commandfile <- paste0("powershell start ",'"', commandfile,'"')#,'" -NoNewWindow')
#using powershell because otherwise bat/cmd file won't run if wait = FALSE
# in system()
}
print(commandfile)
system(commandfile,
intern = FALSE,
wait = FALSE)
# system2("cmd", args = commandfile,
#
# wait = FALSE)
#
# shell(commandfile,
# intern = FALSE,
# wait = FALSE,
# translate = FALSE)
if(file.exists(indexfile)){
jobindex <- as.data.frame(data.table::fread(indexfile, na.strings = NULL))
#not appending so that if new columns are in newjobs, there are no problems
fwrite(rbindlist(list(data.table::fread(indexfile, na.strings = NULL),
newjobs),
fill = TRUE),
indexfile,
append = F, sep = ",", row.names = F,
col.names = T, quote = T, eol = "\n")
}else{
fwrite(newjobs,
indexfile,
append = F, sep = ",", row.names = F,
col.names = T, quote = T, eol = "\n")
}
}
}
#' getSirius
#'
#' get SIRIUS results for a given spectrum
#'
#' @param outfolder folder with SIRIUS results
#' @param hash hash tag to look for
#' @param ts timestamp to look for
#'
#' @importFrom data.table fread
#'
#' @return Returns a list with information for a completed SIRIUS job
#' that matches \code{hash} and \code{ts} in \code{outfolder}
#'
#' @details Finds a result folder matching the search terms and reads
#' its contents.
#' \itemize{
#' \item \code{hash} copy of \code{hash}
#' \item \code{timestamp} copy of \code{ts}
#' \item \code{allfiles} character vector with all files in \code{outfolder}
#' \item \code{summary} path of the \code{summary_sirius.csv} file
#' \item \code{summary_expanded} data.frame read in from the
#' \code{summary_sirius.csv} file
#' \item \code{annotations} list of all files in the \code{spectra}
#' subfolder of the result folder
#' \item \code{trees_dot} list of all fragmentation tree \code{.dot} files
#' \item \code{trees_json} list of all fragmentation tree \code{.json} files
#' \item \code{summary_fingerid} path of the \code{summary_csi_fingerid.csv}
#' file
#' \item \code{summary_fingerid_expanded} data.frame read in from the
#' \code{summary_csi_fingerid.csv} file
#' \item \code{fingerids} list of all files in the \code{csi_fingerid}
#' subfolder of the result folder
#' \item \code{fingerprints} list of all files in the \code{fingerprints}
#' subfolder of the result folder
#' }
#'
#' @export
getSirius <- function(outfolder, hash, ts){
res <- list()
res[["hash"]] <- hash
res[["timestamp"]] <- ts
targetfolder <- grep(paste0(hash),
list.dirs(grep(paste0(ts),
list.dirs(file.path(outfolder),
full.names = T,
recursive = F),
value = T),
full.names = T,
recursive = F),
value = T)
res[["allfiles"]] <- list.files(targetfolder, full.names = T, recursive = T)
res[["summary"]] <- grep(paste0("formula_candidates.tsv"),
res[["allfiles"]],
value = T)
res[["summary_expanded"]] <- as.data.frame(fread(res[["summary"]],
stringsAsFactors = F),
stringsAsFactors = F)
res[["annotations"]] <- list.files(file.path(dirname(res[["summary"]]),
"spectra"),
full.names = T, recursive = T)
res[["trees_dot"]] <- grep(".dot$",
list.files(file.path(dirname(res[["summary"]]),
"trees"),
full.names = T, recursive = T),
value = T)
res[["trees_json"]] <- grep(".json$",
list.files(file.path(dirname(res[["summary"]]),
"trees"),
full.names = T, recursive = T),
value = T)
if(length(grep(paste0("structure_candidates.tsv"),
res[["allfiles"]],value = T))){
res[["summary_fingerid"]] <- grep(paste0("structure_candidates.tsv"),
res[["allfiles"]],value = T)
res[["summary_fingerid_expanded"]] <- as.data.frame(fread(res[["summary_fingerid"]],
stringsAsFactors = F),
stringsAsFactors = F)
res[["fingerids"]] <- list.files(file.path(dirname(res[["summary"]]),
"csi_fingerid"),
full.names = T, recursive = T)
res[["fingerprints"]] <- list.files(file.path(dirname(res[["summary"]]),
"fingerprints"),
full.names = T, recursive = T)
}
return(res)
}
#' getSiriusTree
#'
#' get additional SIRIUS results (fragmentation trees) for a spectrum
#'
#' @param paths list generated by \code{\link{getSirius}()}
#' @param formula molecular formula prediction to get details for
#'
#' @return Returns a named list with fragmentation tree objects for plotting,
#' see |code{Details}
#'
#' @details
#' \itemize{
#' \item \code{annotations} the matching file from \code{spectra} subfolder
#' as data.frame
#' \item \code{trees_dot} the matching \code{.dot} tree file, read in by
#' \code{DiagrammeR::grViz}
#' \item \code{trees_json} the matching \code{.json} tree file, read in by
#' \code{jsonlite::read_json}
#' }
#'
#' @importFrom data.table fread
#' @importFrom jsonlite read_json
#'
getSiriusTree <- function(paths, formula){
res <- list()
res[["annotations"]] <- as.data.frame(fread(grep(paste0("",formula,"_"),
paths[["annotations"]],
value = T), stringsAsFactors = F),
stringsAsFactors = F)
res[["trees_dot"]] <- DiagrammeR::grViz(grep(paste0("_",formula,"_"),
paths[["trees_dot"]],
value = T))
res[["trees_json"]] <- jsonlite::read_json(grep(paste0("",formula,"_"),
paths[["trees_json"]],
value = T))
return(res)
}
mjhelf/Mosaic documentation built on April 28, 2022, 11:32 a.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Defines functions getSiriusTree getSirius runSirius writeMS
Documented in getSirius getSiriusTree runSirius writeMS
#' writeMS
#'
#' write an .ms file containing a spectrum to be exported to SIRIUS.
#' will append additional spectrum to file if filename is that of an existing file.
#'
#' @return Returns nothing, but writes a SIRIUS \code{.ms} file.
#'
#' @param filename path of file to write to
#' @param ms2 a matrix with columns mz and intensity, or a list of ms2 spectra
#' in the same format
#' @param ms1 a matrix with columns mz and intensity for a single MS1 spectrum
#' @param parentmz numeric: parent ion m/z value
#' @param comments character: comments
#' @param rt numeric: retention time in seconds
#' @param ion ion type (e.g. \eqn{[M+H]+})
#' @param charge charge (positive or negative integer)
#' @param hashtag hash tag of the ms2 spectrum
#' @param scanindices which scans were averaged into this ms2 spectrum
#'
#' @importFrom digest digest
#' @importFrom data.table fwrite
#'
writeMS <- function(filename,
ms2,
ms1 = NULL,
parentmz,
comments = "",
rt = "",
ion = NULL,
charge = 1,
hashtag = NULL,
scanindices = NULL){
if(is.null(hashtag)){hashtag <- digest(ms2, algo = "xxhash64")}
appending <- file.exists(filename)
write(paste(
if(appending){"\n"}else{paste0('# Generated with MOSAiC version ',
packageVersion("Metaboseek"),', timestamp: ',
strftime(Sys.time(),"%Y%m%d_%H%M%S"))},
paste0(">compound ", hashtag),
paste0(">parentmass ", parentmz),
if(!is.null(ion)){paste0(">ion ", ion)}else{paste0(">charge ", charge)},
if(!is.null(scanindices)){paste0(">scans ", scanindices)}else{""},
paste0(">hash ", hashtag),
paste0("# Comments: ", comments),
paste0(">rt ", rt),
sep = "\n"),
file = filename,
append = appending
)
if(!is.null(ms1)){
write(
paste0(">ms1"),
file = filename,
append = T
)
fwrite(data.table(ms1), filename, append = T, sep = "\t",
row.names = F, col.names = F, quote = T, eol = "\n")
}
if(!is.null(ms2)){
if(!is.list(ms2)){ms2 <- list(ms2)}
lapply(seq(length(ms2)), function(n){
#expect collision energy to be in names, resulting in ">collision20", etc.
write(
if(is.null(names(ms2))){">ms2"}else{paste0(">",names(ms2)[n])},
file = filename,
append = T
)
fwrite(data.table(ms2[[n]]), filename, append = T, sep = "\t",
row.names = F, col.names = F, quote = T, eol = "\n")
})
}
}
#' runSirius
#'
#' run SIRIUS externally for a list of ms2 spectra
#'
#' @param outfolder output folder for the SIRIUS results
#' @param ms1 if not NULL, a list of MS1 spectra
#' @param ms2 a list (or list of lists) of ms2 spectra
#' (matrices with columns mz and intensity), see \code{Details}
#' @param parentmz numeric: parent ion m/z value
#' @param comments character: comments
#' @param rt numeric: retention time in seconds
#' @param fingerid if TRUE, will try to predict structures with CSI:FingerID
#' @param ion ion type (e.g. \eqn{[M+H]+})
#' @param charge charge (positive or negative integer)
#' @param scanindices which scans were averaged into this ms2 spectrum
#' @param sirpath path to SIRIUS executable
#' @param moreOpts character with additional options to be passed to SIRIUS
#' @param force force calculation, even if same results should exist according to indexfile
#' @param config list containing sirius configuration
#'
#' @return Will cause SIRIUS to run a job and save its results in
#' a timestamped subfolder of \code{outfolder}, and will register the job in
#' the \code{index.csv} file in \code{outfolder}.
#'
#' @details
#' \describe{
#' \item{ms2}{can be a list of matrices, or a list of list of matrices.
#' In the latter case, the (optional) names of the nested list items are
#' expected to be in the format "collisionXX", where XX denotes an integer
#' collision energy.}
#' }
#'
#' @importFrom data.table fread fwrite data.table
#' @importFrom digest digest
#'
#
# ms1 = list(matrix(c(285.253723144531, 3128961.25, 286.256988525391, 469789.125),
# dimnames = list(rows = NULL,
# cols = c('mz','intensity')),
# ncol = 2,
# byrow = TRUE))
#
# ms2 = list(list(collision30 = matrix(c(84.0183,114.09156, 143.117874,443266,165303,516731),
# dimnames = list(rows = NULL,
# cols = c('mz','intensity')),
# ncol = 2)))
#
# runSirius( sirpath = "C:/Runspace/sirius-4.8.2-win64/sirius-gui/sirius",
# ms2 = ms2,
# parentmz = 285.253723144531,
# outfolder = "C:/Runspace/sirius-4.8.2-win64/sirius-gui/Metaboseek",
# ms1 = ms1,
# ion = "[M+?]+",
# comments = "",
# rt = "",
# charge= 1,
# fingerid = T,
# scanindices = 111111,
# moreOpts = "",
# config = list(IsotopeSettings.filter = TRUE,
# FormulaSearchDB = c('BIO'),
# Timeout.secondsPerTree = 0,
# FormulaSettings.enforced = "HCNOP[5]S",
# Timeout.secondsPerInstance = 0,
# AdductSettings.detectable = ",",
# UseHeuristic.mzToUseHeuristicOnly = 650,
# AlgorithmProfile = list("orbitrap"), #qtof
# IsotopeMs2Settings = "IGNORE",
# MS2MassDeviation.allowedMassDeviation = '5.0ppm',
# NumberOfCandidatesPerIon = 5,
# UseHeuristic.mzToUseHeuristic = 300,
# FormulaSettings.detectable = ",",
# NumberOfCandidates = 10,
# StructureSearchDB = c('BIO'),
# AdductSettings.fallback = ",",
# RecomputeResults = TRUE),
# force = T)
#'
#'
#' @export
runSirius <- function( sirpath = file.path(.MseekOptions$siriusFolder, "sirius"),
ms2,
parentmz,
outfolder = getwd(),
ms1 = NULL,
ion = "[M+?]+",
comments = "",
rt = "",
charge= 1,
fingerid = T,
scanindices = "",
moreOpts = "",
config = list(IsotopeSettings.filter = TRUE,
FormulaSearchDB = c('ALL_BUT_INSILICO','ALL','BIO',
'METACYC','CHEBI','COCONUT',
'ECOCYCMINE','GNPS','HMDB',
'HSDB','KEGG','KEGGMINE',
'KNAPSACK','MACONDA','MESH',
'NORMAN','UNDP','PLANTCYC',
'PUBCHEM','PUBMED','YMDB',
'YMDBMINE','ZINCBIO'),
Timeout.secondsPerTree = 0,
FormulaSettings.enforced = "HCNOP[5]S",
Timeout.secondsPerInstance = 0,
AdductSettings.detectable = "[[M+K]+,[M+H3N+H]+,[M+Na]+,[M-H4O2+H]+,[M-H2O+H]+,[M+H]+]",
UseHeuristic.mzToUseHeuristicOnly = 650,
AlgorithmProfile = "orbitrap", #qtof
IsotopeMs2Settings = "IGNORE",
MS2MassDeviation.allowedMassDeviation = '5.0ppm',
NumberOfCandidatesPerIon = 1,
UseHeuristic.mzToUseHeuristic = 300,
FormulaSettings.detectable = ",",
NumberOfCandidates = 10,
StructureSearchDB = c('ALL_BUT_INSILICO','ALL','BIO',
'METACYC','CHEBI','COCONUT',
'ECOCYCMINE','GNPS','HMDB',
'HSDB','KEGG','KEGGMINE',
'KNAPSACK','MACONDA','MESH',
'NORMAN','UNDP','PLANTCYC',
'PUBCHEM','PUBMED','YMDB',
'YMDBMINE','ZINCBIO'),
AdductSettings.fallback = "[[M+K]+,[M+Na]+,[M+H]+]",
RecomputeResults = TRUE),
force = T){
#write the MS2 data so sirius can read it
dir.create(outfolder, showWarnings = F)
#hashtag <- lapply(ms2, gethash)
hashtag <- lapply(ms2, digest, algo = "xxhash64")
ts <- strftime(Sys.time(),"%y%m%d_%H%M%S")
#make a data.frame with information on what will be analyzed here
newjobs <- data.frame(mz = parentmz,
stringsAsFactors = F)
newjobs$rt = rt
newjobs$hash = unlist(hashtag)
newjobs$timestamp = ts
newjobs$ion = ion
newjobs$ms1hash = sapply(ms1, function(x){if(is.null(x)){""}else{digest(x,algo = "xxhash64")}})
newjobs$charge = charge
newjobs$fingerid = fingerid
newjobs$moreOpts = paste0(moreOpts)
newjobs$Metaboseek_version = as.character(packageVersion("Metaboseek"))
newjobs$Metaboseek_sirius_revision = 3
newjobs <- cbind(newjobs, as.data.frame(lapply(config,function(x){if(length(x) >1){paste(x, collapse = ",")}else{x}}), stringsAsFactors = FALSE))
newjobs$settingsHash <- apply(newjobs[,c("ion", "ms1hash", "charge",
"fingerid", "moreOpts",
"Metaboseek_sirius_revision",
names(config)),
drop = F], 1, digest,
algo = "xxhash64")
indexfile <- file.path(outfolder, "index.csv")
if(file.exists(indexfile) && !force){
jobindex <- as.data.frame(data.table::fread(indexfile, na.strings = NULL))
# for all columns that are of type logical and only contain NAs, assume they are mutilated empty character strings
# and a victim of type.convert - make them character vectors again
charCols <- sapply(jobindex,typeof) == "logical" & sapply(lapply(jobindex,is.na),all)
if(any(charCols)){
jobindex[,charCols] <- character(nrow(jobindex))
}
checkme <- as.data.frame(rbindlist(list(old = jobindex,
new = newjobs),
fill = T))[,c("mz", "hash",
"settingsHash")]
#To avoid problems with m/z digits possibly not retained when
#exporting and reimporting of siriusIndex to/from index.csv
checkme$mz <- round(checkme$mz, 4)
dups <- duplicated.data.frame(checkme)[seq(nrow(jobindex)+1,nrow(checkme))]
newjobs <- newjobs[!dups,]
}
if(nrow(newjobs) > 0){
filename <- file.path(outfolder,paste0(ts,".ms"))
np <- mapply(writeMS,
filename = filename,
ms1 = ms1,
ms2 = ms2,
parentmz = parentmz,
scanindices = scanindices,
comments = comments,
rt = rt,
hashtag = hashtag,
ion = ion,
charge = charge)
outfolder <- file.path(outfolder,ts)
### config --IsotopeSettings.filter true --FormulaSearchDB --Timeout.secondsPerTree 0 --FormulaSettings.enforced HCNOP[5]S --Timeout.secondsPerInstance 0 --AdductSettings.detectable [[M + K]+, [M + H3N + H]+, [M + Na]+, [M - H4O2 + H]+, [M - H2O + H]+, [M + H]+] --UseHeuristic.mzToUseHeuristicOnly 650 --AlgorithmProfile orbitrap --IsotopeMs2Settings IGNORE --MS2MassDeviation.allowedMassDeviation 5.0ppm --NumberOfCandidatesPerIon 1 --UseHeuristic.mzToUseHeuristic 300 --FormulaSettings.detectable , --NumberOfCandidates 10 --AdductSettings.fallback [[M + K]+, [M + H3N + H]+, [M + Na]+, [M - H4O2 + H]+, [M - H2O + H]+, [M + H]+] --RecomputeResults true formula
## --StructureSearchDB BIO
## --StructureSearchDB PUBCHEM
finger <- if(fingerid){"fingerid"}else{""}
configstring <- sapply(names(config),function(n){
x <- config[[n]]
if(is.logical(x)){
x <- tolower(as.character(x))
}
if(length(x) > 1){
x <- paste(x, collapse = ",")
}
paste0("--", n, " ", x)
})
configstring <- paste0("config ", paste(configstring, collapse = " "))
cmd <- paste0(
sirpath,
' -i "',
filename,
'"',
" -o ",
'"',
outfolder,
'" ',
configstring,
" tree ",
finger,
moreOpts
)
cmd2 <- paste(paste0(sirpath,
' -i "',
outfolder,
'" ftree-export --all --dot --output ',
file.path(outfolder,paste0(seq_len(length(hashtag))-1,"_",ts,"_",unlist(hashtag)), "trees")),
collapse = "\n")
commandfile <- file.path(dirname(outfolder),paste0(ts,".sh"))
if(Sys.info()['sysname'] == "Windows"){
commandfile <- gsub("\\.sh$",".cmd",commandfile)
}
writeLines(c(cmd,cmd2),
commandfile
)
if(Sys.info()['sysname'] == "Windows"){
# commandfile <- paste0("cmd /c ",'"', commandfile,'"')
commandfile <- paste0("powershell start ",'"', commandfile,'"')#,'" -NoNewWindow')
#using powershell because otherwise bat/cmd file won't run if wait = FALSE
# in system()
}
print(commandfile)
system(commandfile,
intern = FALSE,
wait = FALSE)
# system2("cmd", args = commandfile,
#
# wait = FALSE)
#
# shell(commandfile,
# intern = FALSE,
# wait = FALSE,
# translate = FALSE)
if(file.exists(indexfile)){
jobindex <- as.data.frame(data.table::fread(indexfile, na.strings = NULL))
#not appending so that if new columns are in newjobs, there are no problems
fwrite(rbindlist(list(data.table::fread(indexfile, na.strings = NULL),
newjobs),
fill = TRUE),
indexfile,
append = F, sep = ",", row.names = F,
col.names = T, quote = T, eol = "\n")
}else{
fwrite(newjobs,
indexfile,
append = F, sep = ",", row.names = F,
col.names = T, quote = T, eol = "\n")
}
}
}
#' getSirius
#'
#' get SIRIUS results for a given spectrum
#'
#' @param outfolder folder with SIRIUS results
#' @param hash hash tag to look for
#' @param ts timestamp to look for
#'
#' @importFrom data.table fread
#'
#' @return Returns a list with information for a completed SIRIUS job
#' that matches \code{hash} and \code{ts} in \code{outfolder}
#'
#' @details Finds a result folder matching the search terms and reads
#' its contents.
#' \itemize{
#' \item \code{hash} copy of \code{hash}
#' \item \code{timestamp} copy of \code{ts}
#' \item \code{allfiles} character vector with all files in \code{outfolder}
#' \item \code{summary} path of the \code{summary_sirius.csv} file
#' \item \code{summary_expanded} data.frame read in from the
#' \code{summary_sirius.csv} file
#' \item \code{annotations} list of all files in the \code{spectra}
#' subfolder of the result folder
#' \item \code{trees_dot} list of all fragmentation tree \code{.dot} files
#' \item \code{trees_json} list of all fragmentation tree \code{.json} files
#' \item \code{summary_fingerid} path of the \code{summary_csi_fingerid.csv}
#' file
#' \item \code{summary_fingerid_expanded} data.frame read in from the
#' \code{summary_csi_fingerid.csv} file
#' \item \code{fingerids} list of all files in the \code{csi_fingerid}
#' subfolder of the result folder
#' \item \code{fingerprints} list of all files in the \code{fingerprints}
#' subfolder of the result folder
#' }
#'
#' @export
getSirius <- function(outfolder, hash, ts){
res <- list()
res[["hash"]] <- hash
res[["timestamp"]] <- ts
targetfolder <- grep(paste0(hash),
list.dirs(grep(paste0(ts),
list.dirs(file.path(outfolder),
full.names = T,
recursive = F),
value = T),
full.names = T,
recursive = F),
value = T)
res[["allfiles"]] <- list.files(targetfolder, full.names = T, recursive = T)
res[["summary"]] <- grep(paste0("formula_candidates.tsv"),
res[["allfiles"]],
value = T)
res[["summary_expanded"]] <- as.data.frame(fread(res[["summary"]],
stringsAsFactors = F),
stringsAsFactors = F)
res[["annotations"]] <- list.files(file.path(dirname(res[["summary"]]),
"spectra"),
full.names = T, recursive = T)
res[["trees_dot"]] <- grep(".dot$",
list.files(file.path(dirname(res[["summary"]]),
"trees"),
full.names = T, recursive = T),
value = T)
res[["trees_json"]] <- grep(".json$",
list.files(file.path(dirname(res[["summary"]]),
"trees"),
full.names = T, recursive = T),
value = T)
if(length(grep(paste0("structure_candidates.tsv"),
res[["allfiles"]],value = T))){
res[["summary_fingerid"]] <- grep(paste0("structure_candidates.tsv"),
res[["allfiles"]],value = T)
res[["summary_fingerid_expanded"]] <- as.data.frame(fread(res[["summary_fingerid"]],
stringsAsFactors = F),
stringsAsFactors = F)
res[["fingerids"]] <- list.files(file.path(dirname(res[["summary"]]),
"csi_fingerid"),
full.names = T, recursive = T)
res[["fingerprints"]] <- list.files(file.path(dirname(res[["summary"]]),
"fingerprints"),
full.names = T, recursive = T)
}
return(res)
}
#' getSiriusTree
#'
#' get additional SIRIUS results (fragmentation trees) for a spectrum
#'
#' @param paths list generated by \code{\link{getSirius}()}
#' @param formula molecular formula prediction to get details for
#'
#' @return Returns a named list with fragmentation tree objects for plotting,
#' see |code{Details}
#'
#' @details
#' \itemize{
#' \item \code{annotations} the matching file from \code{spectra} subfolder
#' as data.frame
#' \item \code{trees_dot} the matching \code{.dot} tree file, read in by
#' \code{DiagrammeR::grViz}
#' \item \code{trees_json} the matching \code{.json} tree file, read in by
#' \code{jsonlite::read_json}
#' }
#'
#' @importFrom data.table fread
#' @importFrom jsonlite read_json
#'
getSiriusTree <- function(paths, formula){
res <- list()
res[["annotations"]] <- as.data.frame(fread(grep(paste0("",formula,"_"),
paths[["annotations"]],
value = T), stringsAsFactors = F),
stringsAsFactors = F)
res[["trees_dot"]] <- DiagrammeR::grViz(grep(paste0("_",formula,"_"),
paths[["trees_dot"]],
value = T))
res[["trees_json"]] <- jsonlite::read_json(grep(paste0("",formula,"_"),
paths[["trees_json"]],
value = T))
return(res)
}
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.