R/02_lipid_miner.R
In PNNL-Comp-Mass-Spec/Rodin: Lipid Mini-on
Defines functions
lipid.miner
Documented in
lipid.miner
#' lipid.miner()
#'
#' This function performs the text minning and extract lipid information from a <Query> or a <Universe> lipid list. It creates three new distinct R objects in the global environment: <name>.intact, <name>.chain and <name>.allchain
#'
#' @param X is character vector containing the lipids names
#' @param name is character string setting the name of the <name>.intact, <name>.chain and <name>.allchain object created
#' @param TGcollapse.rm is a boolean value. If TRUE, it removes the Triglycerides from the <name>.chain and <name>.allchains files when collapsed
#' @param output.list is a boolean value, which defaults to FALSE. If FALSE, the function returns objects as described below. If TRUE, the function instead returns a list of these same objects (minus the appended prefix '<name>.'), thus allowing the user to set the output of this function to a variable of their naming.
#'
#' @details We recommend to use the clean.lipid.list() function prior to use lipid.miner().
#' @details lipid.miner() mines the lipid names with the following format:
#' @details mainclass(chainLenght:doubleBond/chainLenght:doubleBond/etc.)
#' @details the mainclasses recognized by lipid.miner() are contained in the main.colors object
#' @details Any lipid that don't belong to one of recognized classes
#' @details will be considered as 'Uncategorized'.
#' @details Contact us if you need the inclusion of another lipid class
#'
#' @return This function return 3 objects named :
#' @return <name>.intact, <name>.chain and <name>.allchains
#' @return <name>.intact contains : category, main class, subclass, and various chain info
#' @return <name>.chain contains information regarding the length of the fatty acid chains
#' @return <name>.allchain contains the list of the individual fatty acid
#'
#' @examples lipid.miner(cleaned.queryExample, name= "QueryExample", TGcollapse.rm=TRUE)
#' @examples lipid.miner(cleaned.universeExample, name= "UniverseExample", TGcollapse.rm=TRUE)
#' @author Geremy Clair
#' @export
lipid.miner <- function(X,name="yourname", TGcollapse.rm=TRUE, output.list=FALSE){
if(!is.character(X)){stop("X has to be a character vector")}
if(!(is.character(name)&&length(name)==1)){"The output name is not properly defined"}
if(!exists("TGcollapse.rm")){TGcollapse.rm<-TRUE}
#remove the deuterated standards from the list
X <- X[!grepl("\\(d5\\)", X)]
#Create the output df
df<-data.frame(matrix(NA,ncol = 27, nrow = length(X)))
colnames(df)<- c("Lipid","Category", "Main class", "Sub class", "Total Number of Carbon", "Double Bonds", "Chain 1", "Chain 2", "Chain 3", "Chain 4", "N Carbon Chain1", "N Carbon Chain2", "N Carbon Chain3", "N Carbon Chain4", "DB Chain1", "DB Chain2","DB Chain3","DB Chain4","Lyso","SCFA","MCFA","LCFA","VLCFA", "Saturated", "Monounsaturated", "Diunsaturated", "Polyunsaturated" )
df$Lipid <- X
# Compute the main class
X<- gsub("\\(3'-sulfo)Galbeta-Cer\\b","Sulfatide",X, ignore.case = TRUE)
X<- gsub("M\\(IP\\)2C", "MIP2C", X) #rename M(IP)2C by M(IP)2C #PARENTHESIS PROBLEM FOR M(IP)2C
X<- gsub(".*anandamide\\(", "EA(", X,ignore.case = TRUE)
X<- gsub(".*CoenzymeQ.*", "CoQ(", X,ignore.case = TRUE)
X<- gsub(".*CoQ.*", "CoQ(", X,ignore.case = TRUE)
X<- gsub(".*Sulfatide", "Sulfatide(", X,ignore.case = TRUE)
X<- gsub("Cholesterol", "Cholesterol(", X,ignore.case = TRUE)
X<- gsub("Sulfatide\\(\\(","Sulfatide(",X,ignore.case = TRUE)
X<- gsub("caritine\\(\\(","caritine(",X,ignore.case = TRUE)
X<- gsub("PE-NMe2\\(","PE\\(NMe2\\-",X,ignore.case = TRUE)
X<- gsub("PE-NMe\\(","PE\\(NMe\\-",X,ignore.case = TRUE)
X<- gsub("\\,n\\-3","",X,ignore.case = TRUE)
X<- gsub("\\,n\\-6","",X,ignore.case = TRUE)
X<- gsub("\\,n\\-9","",X,ignore.case = TRUE)
mainclass.list<- gsub("\\(.*", "",X)
mainclass.list<-gsub("(.*)\\[.*", "\\1", mainclass.list)
df$`Main class`<-mainclass.list
# compute the lipid Category based on the main class
#create the different lists
FattyAcyls<-c("EA", "carnitine","FAHFA", "WE" )
Prenol<- c("CoQ")
Sterol<- c("Cholesterol","CE")
Sphingolipids<- c("Cer","CerP","HexCer","SM","GM3","GD","PE-Cer","PI-Cer","Sulfatide","GM1","GM2","LacCer","GalCer","GD1", "GD3","GlcCer", "MIP2C","MIPC","GT3")
Glycerophospholipids<-c("PE", "PC", "PS", "PI", "PG", "PA", "CL","PIP","PIP2", "PIP3")
Glycerolipids<- c("TG","DG","MG","DGDG","MGDG","SQMG", "SQDG", "DGTS/A")
#Fill the category section
df$Category<- rep("Uncategorized", nrow(df))
df$Category[df$`Main class` %in% Sphingolipids]<-"Sphingolipid"
df$Category[df$`Main class` %in% Prenol]<-"Prenol"
df$Category[df$`Main class` %in% Glycerolipids]<-"Glycerolipid"
df$Category[df$`Main class` %in% Sterol]<-"Sterol"
df$Category[df$`Main class` %in% FattyAcyls]<-"FattyAcyls"
df$Category[df$`Main class` %in% Glycerophospholipids]<-"Glycerophospholipid"
# Modify the input to remove unwanted strings and introduce the <oxidized> and <hydroxy>
#remove text after ";"
X<-gsub("\\;.*","",X)
#Remove the text inside brackets from X
X<-gsub("\\[[^\\]]*\\]", "", X)
#transform the (CHO) (OH) and (COOH) into <oxidized>
X<-gsub("\\(CHO\\)", "<oxidized>", X) #CHO
X<-gsub("\\(COOH\\)", "<oxidized>", X) # (COOH)
X<-gsub("\\(12OH\\)","<oxidized>",X) #(12OH[S])
X<-gsub("\\(14OH\\)","<oxidized>",X) #(12OH[S])
X<-gsub("\\(15OH\\)","<oxidized>",X) #(12OH[S])
X<-gsub("\\(5OH\\)","<oxidized>",X) #(5OH[S])
X<-gsub("\\(2OH\\)", "<hydroxy>", X) ##(2OH[S])
#modify the CardioLipids for proper minning
X<- gsub("1\\'\\-\\[","",X)#remove the 1'-[
X<- gsub("\\,3\\'\\-\\[","\\/",X)#remove the 3'-[
X<- gsub("\\]","",X)#remove the ]
#compute the chains, the N of carbon, the N of DB, the total N of C and of Db by mining between the outermost parenthesis
#extract the fatty acid chains
chains.list<-X
chains.list<- gsub("(.*)\\).*", "\\1", chains.list) #after last parenthesis
chains.list<- sub(".*\\(.*?", "", chains.list) #before the first parenthesis
chains.list<-gsub("NMe[2]\\-","",chains.list)
chains.list<-gsub("NMe\\-","",chains.list)
#split using '/' character
chains.list<-as.list(strsplit(as.character(chains.list), '/'))#split into
chains.list<-lapply(chains.list, `length<-`,4)
chains.df <- data.frame(matrix(unlist(chains.list), nrow=length(X),byrow=T))
df$`Chain 1`<- as.character(t(chains.df$X1))
df$`Chain 2`<- as.character(t(chains.df$X2))
df$`Chain 3`<- as.character(t(chains.df$X3))
df$`Chain 4`<- as.character(t(chains.df$X4))
#REMOVE 0:0 in DG and MG in df
DGMG<-c("DG","MG")
df$`Chain 1`[grepl("^0\\:0",df$`Chain 1`)&df$`Main class`%in% DGMG]<-NA
df$`Chain 2`[grepl("^0\\:0",df$`Chain 2`)&df$`Main class`%in% DGMG]<-NA
df$`Chain 3`[grepl("^0\\:0",df$`Chain 3`)&df$`Main class`%in% DGMG]<-NA
df$`Chain 4`[grepl("^0\\:0",df$`Chain 3`)&df$`Main class`%in% DGMG]<-NA
#################################################take care of the lyso removal?
Lyso<-rep(0,length = nrow(df))
for(i in 1:nrow(df)){
if("0:0" %in% df[i,7:10]){
Lyso[i]<-1
}
}
#remove the 0:0 from df
df$`Chain 1`[grepl("^0\\:0",df$`Chain 1`)]<-NA
df$`Chain 2`[grepl("^0\\:0",df$`Chain 2`)]<-NA
df$`Chain 3`[grepl("^0\\:0",df$`Chain 3`)]<-NA
df$`Chain 4`[grepl("^0\\:0",df$`Chain 3`)]<-NA
#add the number in Lyso
df$Lyso<-Lyso
#Bring this back in the chain object
chains<-paste(df$`Chain 1`,"/",df$`Chain 2`,"/",df$`Chain 3`,"/",df$`Chain 4`, sep="")
#remove O- P- d t <ox> <hydroxyle>, etc
chains<- gsub("<oxidized>", "",chains)
chains<- gsub("<hydroxy>", "",chains)
chains<- gsub("5-O-","O-",chains)
chains<- gsub("6-O-","O-",chains)
chains<- gsub("7-O-","O-",chains)
chains<- gsub("8-O-","O-",chains)
chains<- gsub("9-O-","O-",chains)
chains<- gsub("10-O-","O-",chains)
chains<- gsub("11-O-","O-",chains)
chains<- gsub("12-O-","O-",chains)
chains<- gsub("13-O-","O-",chains)
chains<- gsub("O-", "",chains)
chains<- gsub("P-", "",chains)
chains<- gsub("d", "",chains)
chains<- gsub("t", "",chains)
chains<- gsub("h", "",chains)
chains<- gsub("Me\\)", "",chains)
chains<- gsub("m", "",chains)
#chains<- gsub("\\)Me", "",chains)
#remove internal parenthesis
chains.list<- gsub("\\s*\\([^\\)]+\\)","",chains)
chains.list<- gsub("\\(5E,9E","",chains.list)
#split the string
chains.list<-as.list(strsplit(as.character(chains.list), '/'))#split into list
chains.list<-lapply(chains.list, `length<-`,4)
chains.df <- data.frame(matrix(unlist(chains.list), nrow=length(X),byrow=T))
#N carbon Chains 1->4
df$`N Carbon Chain1` <- gsub(":.*","",chains.df$X1)
df$`N Carbon Chain2` <- gsub(":.*","",chains.df$X2)
df$`N Carbon Chain3` <- gsub(":.*","",chains.df$X3)
df$`N Carbon Chain4` <- gsub(":.*","",chains.df$X4)
#DB Chains 1->4
df$`DB Chain1` <- gsub(".*:","",chains.df$X1)
df$`DB Chain2` <- gsub(".*:","",chains.df$X2)
df$`DB Chain3` <- gsub(".*:","",chains.df$X3)
df$`DB Chain4` <- gsub(".*:","",chains.df$X4)
#create a function to make real NAs
make.true.NA <- function(x) if(is.character(x)||is.factor(x)){
is.na(x) <- x=="NA"; x} else {
x}
df[] <- lapply(df, make.true.NA)
#Total number of carbon
suppressWarnings(chains.df$X1<-as.numeric(df$`N Carbon Chain1`))
suppressWarnings(chains.df$X2<-as.numeric(df$`N Carbon Chain2`))
suppressWarnings(chains.df$X3<-as.numeric(df$`N Carbon Chain3`))
suppressWarnings(chains.df$X4<-as.numeric(df$`N Carbon Chain4`))
chains.df$X1[is.na(chains.df$X1)]<-0
chains.df$X2[is.na(chains.df$X2)]<-0
chains.df$X3[is.na(chains.df$X3)]<-0
chains.df$X4[is.na(chains.df$X4)]<-0
df$`Total Number of Carbon`<- rowSums (chains.df, na.rm = T)
#Total number of DB
suppressWarnings(chains.df$X1<-as.numeric(df$`DB Chain1`))
suppressWarnings(chains.df$X2<-as.numeric(df$`DB Chain2`))
suppressWarnings(chains.df$X3<-as.numeric(df$`DB Chain3`))
suppressWarnings(chains.df$X4<-as.numeric(df$`DB Chain4`))
chains.df$X1[is.na(chains.df$X1)]<-0
chains.df$X2[is.na(chains.df$X2)]<-0
chains.df$X3[is.na(chains.df$X3)]<-0
chains.df$X4[is.na(chains.df$X4)]<-0
df$`Double Bonds`<- rowSums (chains.df, na.rm = T)
remove(chains.df)
#comupte the subclass by minning between the parenthesis and adding the subclass to the main class
#O P d t DiEther ox and Lyso
subclass.list<- gsub("<oxidized>", "<@>", X)
subclass.list<- gsub("<hydroxy>", "<#>", subclass.list)
subclass.list<- gsub("NMe[2]\\-", "\\|", subclass.list)
subclass.list<- gsub("NMe\\-", "\\%", subclass.list)
subclass.list<- gsub("(.*)\\).*", "\\1", subclass.list) #after last parenthesis
subclass.list<- sub(".*\\(.*?", "", subclass.list) #before the first parenthesis
subclass.list<- gsub("\\d", "",subclass.list) #remove numbers
subclass.list<- gsub("\\:","",subclass.list)#remove the :
subclass.list<- gsub("\\/","",subclass.list)#remove the /
subclass.list<- gsub("\\(","",subclass.list)#remove the (
subclass.list<- gsub("\\Z","",subclass.list)#remove the Z
subclass.list<- gsub("\\E","",subclass.list)#remove the E
subclass.list<- gsub("\\,","",subclass.list)#remove the ,
subclass.list<- gsub("\\)","",subclass.list)#remove the )
subclass.list<- gsub("\\A","",subclass.list)#remove the )
subclass.list<- gsub("\\-O","O",subclass.list)#remove the -O replace by O
subclass.list<- gsub("OO\\-","diester",subclass.list)#remove the -O replace by O
subclass.list<-gsub( "Cholesterol","", subclass.list)
subclass.list<-gsub( "CSulfatide","", subclass.list)
subclass.list<-gsub( "CoQ","",subclass.list)
#bring back oxidized and hydroxy
subclass.list<-gsub( "<@>","<oxidized>", subclass.list)
subclass.list<-gsub( "<#>","<hydroxy>", subclass.list)
subclass.list<- gsub( "\\|","NMe2\\-", subclass.list)
subclass.list<- gsub( "\\%","NMe\\-", subclass.list)
#create the subclass list
df$`Sub class`<- paste(df$`Main class`,"(",subclass.list,sep="")
#Lyso (in the subclass and Lyso column)
df[df$Lyso>0,4]<-paste0("L",df[df$Lyso>0,4])
# Chains, Calculate the chains properties
#this can probably be simplified for faster calculation
chains.df<-cbind(df$Lipid,df[,7:18],df[,20:27])
chains.df[,6:13][is.na(chains.df[,6:13])]<-"-1"
chains.df[,14:21]<-0
colnames(chains.df)[1]<-"Lipid"
for(i in 6:13){
chains.df[,i]<-as.numeric(chains.df[,i])
}
for (i in 1:nrow(chains.df)){
chains.df$SCFA[i]<-sum(chains.df[i,6:9]>=1 & chains.df[i,6:9]<=5)
chains.df$MCFA[i]<-sum(chains.df[i,6:9]>5 & chains.df[i,6:9]<=12)
chains.df$LCFA[i]<-sum(chains.df[i,6:9]>12 & chains.df[i,6:9]<=21)
chains.df$VLCFA[i]<-sum(chains.df[i,6:9]>=22)
chains.df$Saturated[i]<-sum(chains.df[i,10:13]==0)
chains.df$Monounsaturated[i]<-sum(chains.df[i,10:13]==1)
chains.df$Diunsaturated[i]<-sum(chains.df[i,10:13]==2)
chains.df$Polyunsaturated[i]<-sum(chains.df[i,10:13]>=3)
}
df[,20:27]<-chains.df[,14:21]
if(TGcollapse.rm==TRUE){
#remove the chains that are from collapsed TGs
chains.df<- chains.df[!(chains.df$`N Carbon Chain2`==-1&df$`Main class`=="TG"),]
}
chains.df[chains.df==-1]<- NA
################################################################
#create the last object that combine the existing chains in a vector
allchains<- data.frame(matrix(nrow = 4*nrow(chains.df), ncol=2))
allchains[,1]<- c(paste(chains.df$Lipid,"_Chain1",sep=""),paste(chains.df$Lipid,"_Chain2",sep=""),paste(chains.df$Lipid,"_Chain3",sep=""),paste(chains.df$Lipid,"_Chain4",sep=""))
allchains[,2]<-c(as.character(chains.df$`Chain 1`),as.character(chains.df$`Chain 2`),as.character(chains.df$`Chain 3`),as.character(chains.df$`Chain 4`))
colnames(allchains)<- c("Lipid", "Chain")
allchains<- allchains[!is.na(allchains$Chain),]
#finalize the object chains
chains.df<- cbind(chains.df[,1],chains.df[,14:21])
colnames(chains.df)[1]<- "Lipid"
#if unknown main.class/subclass replace with Uncategorized
df$`Main class`[!df$`Main class` %in% main.colors$MainClass]<-"Uncategorized"
df$`Sub class`[!df$`Sub class` %in% sub.colors$SubClass]<-"Uncategorized"
# Final output
if(output.list == FALSE){
assign(paste(name,".intact",sep=""),df[,1:18], envir = globalenv())
assign(paste(name,".chain",sep=""),data.frame(chains.df), envir = globalenv())
assign(paste(name,".allchains",sep=""),data.frame(allchains),envir = globalenv())
}else{
# output.list == TRUE #
return(list(intact = df[,1:18], chain = data.frame(chains.df), allchains = data.frame(allchains)))
}
}
PNNL-Comp-Mass-Spec/Rodin documentation built on Jan. 28, 2024, 2:12 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 lipid.miner
Documented in lipid.miner
#' lipid.miner()
#'
#' This function performs the text minning and extract lipid information from a <Query> or a <Universe> lipid list. It creates three new distinct R objects in the global environment: <name>.intact, <name>.chain and <name>.allchain
#'
#' @param X is character vector containing the lipids names
#' @param name is character string setting the name of the <name>.intact, <name>.chain and <name>.allchain object created
#' @param TGcollapse.rm is a boolean value. If TRUE, it removes the Triglycerides from the <name>.chain and <name>.allchains files when collapsed
#' @param output.list is a boolean value, which defaults to FALSE. If FALSE, the function returns objects as described below. If TRUE, the function instead returns a list of these same objects (minus the appended prefix '<name>.'), thus allowing the user to set the output of this function to a variable of their naming.
#'
#' @details We recommend to use the clean.lipid.list() function prior to use lipid.miner().
#' @details lipid.miner() mines the lipid names with the following format:
#' @details mainclass(chainLenght:doubleBond/chainLenght:doubleBond/etc.)
#' @details the mainclasses recognized by lipid.miner() are contained in the main.colors object
#' @details Any lipid that don't belong to one of recognized classes
#' @details will be considered as 'Uncategorized'.
#' @details Contact us if you need the inclusion of another lipid class
#'
#' @return This function return 3 objects named :
#' @return <name>.intact, <name>.chain and <name>.allchains
#' @return <name>.intact contains : category, main class, subclass, and various chain info
#' @return <name>.chain contains information regarding the length of the fatty acid chains
#' @return <name>.allchain contains the list of the individual fatty acid
#'
#' @examples lipid.miner(cleaned.queryExample, name= "QueryExample", TGcollapse.rm=TRUE)
#' @examples lipid.miner(cleaned.universeExample, name= "UniverseExample", TGcollapse.rm=TRUE)
#' @author Geremy Clair
#' @export
lipid.miner <- function(X,name="yourname", TGcollapse.rm=TRUE, output.list=FALSE){
if(!is.character(X)){stop("X has to be a character vector")}
if(!(is.character(name)&&length(name)==1)){"The output name is not properly defined"}
if(!exists("TGcollapse.rm")){TGcollapse.rm<-TRUE}
#remove the deuterated standards from the list
X <- X[!grepl("\\(d5\\)", X)]
#Create the output df
df<-data.frame(matrix(NA,ncol = 27, nrow = length(X)))
colnames(df)<- c("Lipid","Category", "Main class", "Sub class", "Total Number of Carbon", "Double Bonds", "Chain 1", "Chain 2", "Chain 3", "Chain 4", "N Carbon Chain1", "N Carbon Chain2", "N Carbon Chain3", "N Carbon Chain4", "DB Chain1", "DB Chain2","DB Chain3","DB Chain4","Lyso","SCFA","MCFA","LCFA","VLCFA", "Saturated", "Monounsaturated", "Diunsaturated", "Polyunsaturated" )
df$Lipid <- X
# Compute the main class
X<- gsub("\\(3'-sulfo)Galbeta-Cer\\b","Sulfatide",X, ignore.case = TRUE)
X<- gsub("M\\(IP\\)2C", "MIP2C", X) #rename M(IP)2C by M(IP)2C #PARENTHESIS PROBLEM FOR M(IP)2C
X<- gsub(".*anandamide\\(", "EA(", X,ignore.case = TRUE)
X<- gsub(".*CoenzymeQ.*", "CoQ(", X,ignore.case = TRUE)
X<- gsub(".*CoQ.*", "CoQ(", X,ignore.case = TRUE)
X<- gsub(".*Sulfatide", "Sulfatide(", X,ignore.case = TRUE)
X<- gsub("Cholesterol", "Cholesterol(", X,ignore.case = TRUE)
X<- gsub("Sulfatide\\(\\(","Sulfatide(",X,ignore.case = TRUE)
X<- gsub("caritine\\(\\(","caritine(",X,ignore.case = TRUE)
X<- gsub("PE-NMe2\\(","PE\\(NMe2\\-",X,ignore.case = TRUE)
X<- gsub("PE-NMe\\(","PE\\(NMe\\-",X,ignore.case = TRUE)
X<- gsub("\\,n\\-3","",X,ignore.case = TRUE)
X<- gsub("\\,n\\-6","",X,ignore.case = TRUE)
X<- gsub("\\,n\\-9","",X,ignore.case = TRUE)
mainclass.list<- gsub("\\(.*", "",X)
mainclass.list<-gsub("(.*)\\[.*", "\\1", mainclass.list)
df$`Main class`<-mainclass.list
# compute the lipid Category based on the main class
#create the different lists
FattyAcyls<-c("EA", "carnitine","FAHFA", "WE" )
Prenol<- c("CoQ")
Sterol<- c("Cholesterol","CE")
Sphingolipids<- c("Cer","CerP","HexCer","SM","GM3","GD","PE-Cer","PI-Cer","Sulfatide","GM1","GM2","LacCer","GalCer","GD1", "GD3","GlcCer", "MIP2C","MIPC","GT3")
Glycerophospholipids<-c("PE", "PC", "PS", "PI", "PG", "PA", "CL","PIP","PIP2", "PIP3")
Glycerolipids<- c("TG","DG","MG","DGDG","MGDG","SQMG", "SQDG", "DGTS/A")
#Fill the category section
df$Category<- rep("Uncategorized", nrow(df))
df$Category[df$`Main class` %in% Sphingolipids]<-"Sphingolipid"
df$Category[df$`Main class` %in% Prenol]<-"Prenol"
df$Category[df$`Main class` %in% Glycerolipids]<-"Glycerolipid"
df$Category[df$`Main class` %in% Sterol]<-"Sterol"
df$Category[df$`Main class` %in% FattyAcyls]<-"FattyAcyls"
df$Category[df$`Main class` %in% Glycerophospholipids]<-"Glycerophospholipid"
# Modify the input to remove unwanted strings and introduce the <oxidized> and <hydroxy>
#remove text after ";"
X<-gsub("\\;.*","",X)
#Remove the text inside brackets from X
X<-gsub("\\[[^\\]]*\\]", "", X)
#transform the (CHO) (OH) and (COOH) into <oxidized>
X<-gsub("\\(CHO\\)", "<oxidized>", X) #CHO
X<-gsub("\\(COOH\\)", "<oxidized>", X) # (COOH)
X<-gsub("\\(12OH\\)","<oxidized>",X) #(12OH[S])
X<-gsub("\\(14OH\\)","<oxidized>",X) #(12OH[S])
X<-gsub("\\(15OH\\)","<oxidized>",X) #(12OH[S])
X<-gsub("\\(5OH\\)","<oxidized>",X) #(5OH[S])
X<-gsub("\\(2OH\\)", "<hydroxy>", X) ##(2OH[S])
#modify the CardioLipids for proper minning
X<- gsub("1\\'\\-\\[","",X)#remove the 1'-[
X<- gsub("\\,3\\'\\-\\[","\\/",X)#remove the 3'-[
X<- gsub("\\]","",X)#remove the ]
#compute the chains, the N of carbon, the N of DB, the total N of C and of Db by mining between the outermost parenthesis
#extract the fatty acid chains
chains.list<-X
chains.list<- gsub("(.*)\\).*", "\\1", chains.list) #after last parenthesis
chains.list<- sub(".*\\(.*?", "", chains.list) #before the first parenthesis
chains.list<-gsub("NMe[2]\\-","",chains.list)
chains.list<-gsub("NMe\\-","",chains.list)
#split using '/' character
chains.list<-as.list(strsplit(as.character(chains.list), '/'))#split into
chains.list<-lapply(chains.list, `length<-`,4)
chains.df <- data.frame(matrix(unlist(chains.list), nrow=length(X),byrow=T))
df$`Chain 1`<- as.character(t(chains.df$X1))
df$`Chain 2`<- as.character(t(chains.df$X2))
df$`Chain 3`<- as.character(t(chains.df$X3))
df$`Chain 4`<- as.character(t(chains.df$X4))
#REMOVE 0:0 in DG and MG in df
DGMG<-c("DG","MG")
df$`Chain 1`[grepl("^0\\:0",df$`Chain 1`)&df$`Main class`%in% DGMG]<-NA
df$`Chain 2`[grepl("^0\\:0",df$`Chain 2`)&df$`Main class`%in% DGMG]<-NA
df$`Chain 3`[grepl("^0\\:0",df$`Chain 3`)&df$`Main class`%in% DGMG]<-NA
df$`Chain 4`[grepl("^0\\:0",df$`Chain 3`)&df$`Main class`%in% DGMG]<-NA
#################################################take care of the lyso removal?
Lyso<-rep(0,length = nrow(df))
for(i in 1:nrow(df)){
if("0:0" %in% df[i,7:10]){
Lyso[i]<-1
}
}
#remove the 0:0 from df
df$`Chain 1`[grepl("^0\\:0",df$`Chain 1`)]<-NA
df$`Chain 2`[grepl("^0\\:0",df$`Chain 2`)]<-NA
df$`Chain 3`[grepl("^0\\:0",df$`Chain 3`)]<-NA
df$`Chain 4`[grepl("^0\\:0",df$`Chain 3`)]<-NA
#add the number in Lyso
df$Lyso<-Lyso
#Bring this back in the chain object
chains<-paste(df$`Chain 1`,"/",df$`Chain 2`,"/",df$`Chain 3`,"/",df$`Chain 4`, sep="")
#remove O- P- d t <ox> <hydroxyle>, etc
chains<- gsub("<oxidized>", "",chains)
chains<- gsub("<hydroxy>", "",chains)
chains<- gsub("5-O-","O-",chains)
chains<- gsub("6-O-","O-",chains)
chains<- gsub("7-O-","O-",chains)
chains<- gsub("8-O-","O-",chains)
chains<- gsub("9-O-","O-",chains)
chains<- gsub("10-O-","O-",chains)
chains<- gsub("11-O-","O-",chains)
chains<- gsub("12-O-","O-",chains)
chains<- gsub("13-O-","O-",chains)
chains<- gsub("O-", "",chains)
chains<- gsub("P-", "",chains)
chains<- gsub("d", "",chains)
chains<- gsub("t", "",chains)
chains<- gsub("h", "",chains)
chains<- gsub("Me\\)", "",chains)
chains<- gsub("m", "",chains)
#chains<- gsub("\\)Me", "",chains)
#remove internal parenthesis
chains.list<- gsub("\\s*\\([^\\)]+\\)","",chains)
chains.list<- gsub("\\(5E,9E","",chains.list)
#split the string
chains.list<-as.list(strsplit(as.character(chains.list), '/'))#split into list
chains.list<-lapply(chains.list, `length<-`,4)
chains.df <- data.frame(matrix(unlist(chains.list), nrow=length(X),byrow=T))
#N carbon Chains 1->4
df$`N Carbon Chain1` <- gsub(":.*","",chains.df$X1)
df$`N Carbon Chain2` <- gsub(":.*","",chains.df$X2)
df$`N Carbon Chain3` <- gsub(":.*","",chains.df$X3)
df$`N Carbon Chain4` <- gsub(":.*","",chains.df$X4)
#DB Chains 1->4
df$`DB Chain1` <- gsub(".*:","",chains.df$X1)
df$`DB Chain2` <- gsub(".*:","",chains.df$X2)
df$`DB Chain3` <- gsub(".*:","",chains.df$X3)
df$`DB Chain4` <- gsub(".*:","",chains.df$X4)
#create a function to make real NAs
make.true.NA <- function(x) if(is.character(x)||is.factor(x)){
is.na(x) <- x=="NA"; x} else {
x}
df[] <- lapply(df, make.true.NA)
#Total number of carbon
suppressWarnings(chains.df$X1<-as.numeric(df$`N Carbon Chain1`))
suppressWarnings(chains.df$X2<-as.numeric(df$`N Carbon Chain2`))
suppressWarnings(chains.df$X3<-as.numeric(df$`N Carbon Chain3`))
suppressWarnings(chains.df$X4<-as.numeric(df$`N Carbon Chain4`))
chains.df$X1[is.na(chains.df$X1)]<-0
chains.df$X2[is.na(chains.df$X2)]<-0
chains.df$X3[is.na(chains.df$X3)]<-0
chains.df$X4[is.na(chains.df$X4)]<-0
df$`Total Number of Carbon`<- rowSums (chains.df, na.rm = T)
#Total number of DB
suppressWarnings(chains.df$X1<-as.numeric(df$`DB Chain1`))
suppressWarnings(chains.df$X2<-as.numeric(df$`DB Chain2`))
suppressWarnings(chains.df$X3<-as.numeric(df$`DB Chain3`))
suppressWarnings(chains.df$X4<-as.numeric(df$`DB Chain4`))
chains.df$X1[is.na(chains.df$X1)]<-0
chains.df$X2[is.na(chains.df$X2)]<-0
chains.df$X3[is.na(chains.df$X3)]<-0
chains.df$X4[is.na(chains.df$X4)]<-0
df$`Double Bonds`<- rowSums (chains.df, na.rm = T)
remove(chains.df)
#comupte the subclass by minning between the parenthesis and adding the subclass to the main class
#O P d t DiEther ox and Lyso
subclass.list<- gsub("<oxidized>", "<@>", X)
subclass.list<- gsub("<hydroxy>", "<#>", subclass.list)
subclass.list<- gsub("NMe[2]\\-", "\\|", subclass.list)
subclass.list<- gsub("NMe\\-", "\\%", subclass.list)
subclass.list<- gsub("(.*)\\).*", "\\1", subclass.list) #after last parenthesis
subclass.list<- sub(".*\\(.*?", "", subclass.list) #before the first parenthesis
subclass.list<- gsub("\\d", "",subclass.list) #remove numbers
subclass.list<- gsub("\\:","",subclass.list)#remove the :
subclass.list<- gsub("\\/","",subclass.list)#remove the /
subclass.list<- gsub("\\(","",subclass.list)#remove the (
subclass.list<- gsub("\\Z","",subclass.list)#remove the Z
subclass.list<- gsub("\\E","",subclass.list)#remove the E
subclass.list<- gsub("\\,","",subclass.list)#remove the ,
subclass.list<- gsub("\\)","",subclass.list)#remove the )
subclass.list<- gsub("\\A","",subclass.list)#remove the )
subclass.list<- gsub("\\-O","O",subclass.list)#remove the -O replace by O
subclass.list<- gsub("OO\\-","diester",subclass.list)#remove the -O replace by O
subclass.list<-gsub( "Cholesterol","", subclass.list)
subclass.list<-gsub( "CSulfatide","", subclass.list)
subclass.list<-gsub( "CoQ","",subclass.list)
#bring back oxidized and hydroxy
subclass.list<-gsub( "<@>","<oxidized>", subclass.list)
subclass.list<-gsub( "<#>","<hydroxy>", subclass.list)
subclass.list<- gsub( "\\|","NMe2\\-", subclass.list)
subclass.list<- gsub( "\\%","NMe\\-", subclass.list)
#create the subclass list
df$`Sub class`<- paste(df$`Main class`,"(",subclass.list,sep="")
#Lyso (in the subclass and Lyso column)
df[df$Lyso>0,4]<-paste0("L",df[df$Lyso>0,4])
# Chains, Calculate the chains properties
#this can probably be simplified for faster calculation
chains.df<-cbind(df$Lipid,df[,7:18],df[,20:27])
chains.df[,6:13][is.na(chains.df[,6:13])]<-"-1"
chains.df[,14:21]<-0
colnames(chains.df)[1]<-"Lipid"
for(i in 6:13){
chains.df[,i]<-as.numeric(chains.df[,i])
}
for (i in 1:nrow(chains.df)){
chains.df$SCFA[i]<-sum(chains.df[i,6:9]>=1 & chains.df[i,6:9]<=5)
chains.df$MCFA[i]<-sum(chains.df[i,6:9]>5 & chains.df[i,6:9]<=12)
chains.df$LCFA[i]<-sum(chains.df[i,6:9]>12 & chains.df[i,6:9]<=21)
chains.df$VLCFA[i]<-sum(chains.df[i,6:9]>=22)
chains.df$Saturated[i]<-sum(chains.df[i,10:13]==0)
chains.df$Monounsaturated[i]<-sum(chains.df[i,10:13]==1)
chains.df$Diunsaturated[i]<-sum(chains.df[i,10:13]==2)
chains.df$Polyunsaturated[i]<-sum(chains.df[i,10:13]>=3)
}
df[,20:27]<-chains.df[,14:21]
if(TGcollapse.rm==TRUE){
#remove the chains that are from collapsed TGs
chains.df<- chains.df[!(chains.df$`N Carbon Chain2`==-1&df$`Main class`=="TG"),]
}
chains.df[chains.df==-1]<- NA
################################################################
#create the last object that combine the existing chains in a vector
allchains<- data.frame(matrix(nrow = 4*nrow(chains.df), ncol=2))
allchains[,1]<- c(paste(chains.df$Lipid,"_Chain1",sep=""),paste(chains.df$Lipid,"_Chain2",sep=""),paste(chains.df$Lipid,"_Chain3",sep=""),paste(chains.df$Lipid,"_Chain4",sep=""))
allchains[,2]<-c(as.character(chains.df$`Chain 1`),as.character(chains.df$`Chain 2`),as.character(chains.df$`Chain 3`),as.character(chains.df$`Chain 4`))
colnames(allchains)<- c("Lipid", "Chain")
allchains<- allchains[!is.na(allchains$Chain),]
#finalize the object chains
chains.df<- cbind(chains.df[,1],chains.df[,14:21])
colnames(chains.df)[1]<- "Lipid"
#if unknown main.class/subclass replace with Uncategorized
df$`Main class`[!df$`Main class` %in% main.colors$MainClass]<-"Uncategorized"
df$`Sub class`[!df$`Sub class` %in% sub.colors$SubClass]<-"Uncategorized"
# Final output
if(output.list == FALSE){
assign(paste(name,".intact",sep=""),df[,1:18], envir = globalenv())
assign(paste(name,".chain",sep=""),data.frame(chains.df), envir = globalenv())
assign(paste(name,".allchains",sep=""),data.frame(allchains),envir = globalenv())
}else{
# output.list == TRUE #
return(list(intact = df[,1:18], chain = data.frame(chains.df), allchains = data.frame(allchains)))
}
}
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.