annotate lib_metams.r @ 1:142fbe102a9d draft

Uploaded version 2.0

# lib_metams.r version 0.99.6
# R function for metaMS runGC under W4M
# author Yann GUITTON CNRS IRISA/LINA Idealg project 2014-2015


##ADDITIONS FROM Y. Guitton
getBPC <- function(file,rtcor=NULL, ...) {
     object <- xcmsRaw(file)
	 sel <- profRange(object, ...)
	 cbind(if (is.null(rtcor)) object@scantime[sel$scanidx] else rtcor ,xcms:::colMax(object@env$profile[sel$massidx,sel$scanidx,drop=FALSE]))
    
}

getBPC2s <- function (files, pdfname="BPCs.pdf", rt = c("raw","corrected"), scanrange=NULL) {
    require(xcms)

        
                            
    #create sampleMetadata, get sampleMetadata and class
    sampleMetadata<-xcms:::phenoDataFromPaths(files)
    class<-class<-as.vector(levels(sampleMetadata[,"class"])) #create phenoData like table
    classnames<-vector("list",length(class))
    for (i in 1:length(class)){
        classnames[[i]]<-which( sampleMetadata[,1]==class[i])
    }
        
    N <- dim(sampleMetadata)[1]
      
  
   
    TIC <- vector("list",N)


    for (j in 1:N) {

        cat(files[j],"\n")
        TIC[[j]] <- getBPC(files[j])
        #good for raw 
        # seems strange for corrected
        #errors if scanrange used in xcmsSetgeneration
        if (!is.null(xcmsSet) && rt == "corrected")
          rtcor <- xcmsSet@rt$corrected[[j]] else
          rtcor <- NULL
          TIC[[j]] <- getBPC(files[j],rtcor=rtcor)
    }

    pdf(pdfname,w=16,h=10)
    cols <- rainbow(N)
    lty = 1:N
    pch = 1:N
    #search for max x and max y in BPCs
    xlim = range(sapply(TIC, function(x) range(x[,1])))
    ylim = range(sapply(TIC, function(x) range(x[,2])))
    ylim = c(-ylim[2], ylim[2])


    ##plot start
    
    if (length(class)>2){
      for (k in 1:(length(class)-1)){
        for (l in (k+1):length(class)){
            print(paste(class[k],"vs",class[l],sep=" ")) 
            plot(0, 0, type="n", xlim = xlim/60, ylim = ylim, main = paste("Base Peak Chromatograms \n","BPCs_",class[k]," vs ",class[l], sep=""), xlab = "Retention Time (min)", ylab = "BPC")
            colvect<-NULL
           for (j in 1:length(classnames[[k]])) {
      
              tic <- TIC[[classnames[[k]][j]]]
              # points(tic[,1]/60, tic[,2], col = cols[i], pch = pch[i], type="l")
              points(tic[,1]/60, tic[,2], col = cols[classnames[[k]][j]], pch = pch[classnames[[k]][j]], type="l")
              colvect<-append(colvect,cols[classnames[[k]][j]])
            }
          for (j in 1:length(classnames[[l]])) {
          # i=class2names[j]
          tic <- TIC[[classnames[[l]][j]]]
          points(tic[,1]/60, -tic[,2], col = cols[classnames[[l]][j]], pch = pch[classnames[[l]][j]], type="l")
          colvect<-append(colvect,cols[classnames[[l]][j]])
          }
          legend("topright",paste(basename(files[c(classnames[[k]],classnames[[l]])])), col = colvect, lty = lty, pch = pch)
        }
      }
    }#end if length >2
    if (length(class)==2){
        k=1
		l=2
        colvect<-NULL
        plot(0, 0, type="n", xlim = xlim/60, ylim = ylim, main = paste("Base Peak Chromatograms \n","BPCs_",class[k],"vs",class[l], sep=""), xlab = "Retention Time (min)", ylab = "BPC")

       for (j in 1:length(classnames[[k]])) {

          tic <- TIC[[classnames[[k]][j]]]
          # points(tic[,1]/60, tic[,2], col = cols[i], pch = pch[i], type="l")
          points(tic[,1]/60, tic[,2], col = cols[classnames[[k]][j]], pch = pch[classnames[[k]][j]], type="l")
          colvect<-append(colvect,cols[classnames[[k]][j]])
        }
      for (j in 1:length(classnames[[l]])) {
          # i=class2names[j]
          tic <- TIC[[classnames[[l]][j]]]
          points(tic[,1]/60, -tic[,2], col = cols[classnames[[l]][j]], pch = pch[classnames[[l]][j]], type="l")
          colvect<-append(colvect,cols[classnames[[l]][j]])
      }
      legend("topright",paste(basename(files[c(classnames[[k]],classnames[[l]])])), col = colvect, lty = lty, pch = pch)

    }#end length ==2
       if (length(class)==1){
        k=1
		ylim = range(sapply(TIC, function(x) range(x[,2])))
        colvect<-NULL
        plot(0, 0, type="n", xlim = xlim/60, ylim = ylim, main = paste("Base Peak Chromatograms \n","BPCs_",class[k], sep=""), xlab = "Retention Time (min)", ylab = "BPC")

       for (j in 1:length(classnames[[k]])) {

          tic <- TIC[[classnames[[k]][j]]]
          # points(tic[,1]/60, tic[,2], col = cols[i], pch = pch[i], type="l")
          points(tic[,1]/60, tic[,2], col = cols[classnames[[k]][j]], pch = pch[classnames[[k]][j]], type="l")
          colvect<-append(colvect,cols[classnames[[k]][j]])
        }
      
      legend("topright",paste(basename(files[c(classnames[[k]])])), col = colvect, lty = lty, pch = pch)

    }#end length ==1
    dev.off()

    # invisible(TIC)
}

getTIC <- function(file,rtcor=NULL) {
     object <- xcmsRaw(file)
     cbind(if (is.null(rtcor)) object@scantime else rtcor, rawEIC(object,mzrange=range(object@env$mz))$intensity)
}

##
##  overlay TIC from all files in current folder or from xcmsSet, create pdf
##
getTIC2s <- function(files, pdfname="TICs.pdf", rt=c("raw","corrected")) {
         
    #create sampleMetadata, get sampleMetadata and class
    sampleMetadata<-xcms:::phenoDataFromPaths(files)
    class<-class<-as.vector(levels(sampleMetadata[,"class"])) #create phenoData like table
    classnames<-vector("list",length(class))
    for (i in 1:length(class)){
        classnames[[i]]<-which( sampleMetadata[,1]==class[i])
    }
        
    N <- dim(sampleMetadata)[1]
    TIC <- vector("list",N)

    for (i in 1:N) {
        cat(files[i],"\n")
        if (!is.null(xcmsSet) && rt == "corrected")
            rtcor <- xcmsSet@rt$corrected[[i]]
        else
            rtcor <- NULL
        TIC[[i]] <- getTIC(files[i],rtcor=rtcor)
    }
 
    pdf(pdfname,w=16,h=10)
    cols <- rainbow(N)
    lty = 1:N
    pch = 1:N
    #search for max x and max y in TICs
    xlim = range(sapply(TIC, function(x) range(x[,1])))
    ylim = range(sapply(TIC, function(x) range(x[,2])))
    ylim = c(-ylim[2], ylim[2])
	  
	  
    ##plot start
    if (length(class)>2){
        for (k in 1:(length(class)-1)){
            for (l in (k+1):length(class)){
                print(paste(class[k],"vs",class[l],sep=" ")) 
                plot(0, 0, type="n", xlim = xlim/60, ylim = ylim, main = paste("Total Ion Chromatograms \n","TICs_",class[k]," vs ",class[l], sep=""), xlab = "Retention Time (min)", ylab = "TIC")
                colvect<-NULL
                for (j in 1:length(classnames[[k]])) {

                    tic <- TIC[[classnames[[k]][j]]]
                    # points(tic[,1]/60, tic[,2], col = cols[i], pch = pch[i], type="l")
                    points(tic[,1]/60, tic[,2], col = cols[classnames[[k]][j]], pch = pch[classnames[[k]][j]], type="l")
                    colvect<-append(colvect,cols[classnames[[k]][j]])
                }
                for (j in 1:length(classnames[[l]])) {
                    # i=class2names[j]
                    tic <- TIC[[classnames[[l]][j]]]
                    points(tic[,1]/60, -tic[,2], col = cols[classnames[[l]][j]], pch = pch[classnames[[l]][j]], type="l")
                    colvect<-append(colvect,cols[classnames[[l]][j]])
                }
                legend("topright",paste(basename(files[c(classnames[[k]],classnames[[l]])])), col = colvect, lty = lty, pch = pch)
            }
        }
    }#end if length >2
    if (length(class)==2){


        k=1
        l=2
		
        plot(0, 0, type="n", xlim = xlim/60, ylim = ylim, main = paste("Total Ion Chromatograms \n","TICs_",class[k],"vs",class[l], sep=""), xlab = "Retention Time (min)", ylab = "TIC")
        colvect<-NULL
        for (j in 1:length(classnames[[k]])) {

            tic <- TIC[[classnames[[k]][j]]]
            # points(tic[,1]/60, tic[,2], col = cols[i], pch = pch[i], type="l")
            points(tic[,1]/60, tic[,2], col = cols[classnames[[k]][j]], pch = pch[classnames[[k]][j]], type="l")
            colvect<-append(colvect,cols[classnames[[k]][j]])
        }
        for (j in 1:length(classnames[[l]])) {
            # i=class2names[j]
            tic <- TIC[[classnames[[l]][j]]]
            points(tic[,1]/60, -tic[,2], col = cols[classnames[[l]][j]], pch = pch[classnames[[l]][j]], type="l")
            colvect<-append(colvect,cols[classnames[[l]][j]])
        }
        legend("topright",paste(basename(files[c(classnames[[k]],classnames[[l]])])), col = colvect, lty = lty, pch = pch)

    }#end length ==2
    if (length(class)==1){
        k=1
        ylim = range(sapply(TIC, function(x) range(x[,2])))
		
        plot(0, 0, type="n", xlim = xlim/60, ylim = ylim, main = paste("Total Ion Chromatograms \n","TICs_",class[k], sep=""), xlab = "Retention Time (min)", ylab = "TIC")
        colvect<-NULL
        for (j in 1:length(classnames[[k]])) {

            tic <- TIC[[classnames[[k]][j]]]
            # points(tic[,1]/60, tic[,2], col = cols[i], pch = pch[i], type="l")
            points(tic[,1]/60, tic[,2], col = cols[classnames[[k]][j]], pch = pch[classnames[[k]][j]], type="l")
            colvect<-append(colvect,cols[classnames[[k]][j]])
        }

        legend("topright",paste(basename(files[c(classnames[[k]])])), col = colvect, lty = lty, pch = pch)

    }#end length ==1
    dev.off()

    # invisible(TIC)
}


##addition for quality control of peak picking
#metaMS EIC and pspectra plotting option
#version 20150512
#only for Galaxy 

plotUnknowns<-function(resGC, unkn=""){


    ##Annotation table each value is a pcgrp associated to the unknown 
    ##NOTE pcgrp index are different between xcmsSet and resGC due to filtering steps in metaMS
    ##R. Wehrens give me some clues on that and we found a correction
 
    mat<-matrix(ncol=length(resGC$xset), nrow=dim(resGC$PeakTable)[1])
     
    for (j in 1: length(resGC$xset)){
        test<-resGC$annotation[[j]]
        print(paste("j=",j))
        for (i in 1:dim(test)[1]){
            if (as.numeric(row.names(test)[i])>dim(mat)[1]){
                next
            } else {
                mat[as.numeric(row.names(test)[i]),j]<-test[i,1]
            }
        }
    }
    colnames(mat)<-colnames(resGC$PeakTable[,c((which(colnames(resGC$PeakTable)=="rt"|colnames(resGC$PeakTable)=="RI")[length(which(colnames(resGC$PeakTable)=="rt"|colnames(resGC$PeakTable)=="RI"))]+1):dim(resGC$PeakTable)[2])])
    
    #debug

    # print(dim(mat))
    # print(mat[1:3,]) 
    # write.table(mat, file="myannotationtable.tsv", sep="\t", row.names=FALSE)
    #correction of annotation matrix due to pcgrp removal by quality check in runGCresult
    #matrix of correspondance between an@pspectra and filtered pspectra from runGC

    allPCGRPs <-
        lapply(1:length(resGC$xset),
            function(i) {
                an <- resGC$xset[[i]]
                huhn <- an@pspectra[which(sapply(an@pspectra, length) >=
                metaSetting(resGC$settings,
                "DBconstruction.minfeat"))]
                matCORR<-cbind(1:length(huhn), match(huhn, an@pspectra))
            })

    if (unkn[1]==""){    
    #plot EIC and spectra for all unknown for comparative purpose
   

        par (mar=c(5, 4, 4, 2) + 0.1)
        for (l in 1:dim(resGC$PeakTable)[1]){ #l=2
            #recordPlot
            perpage=3 #if change change layout also!
            num.plots <- ceiling(dim(mat)[2]/perpage) #three pcgroup per page
            my.plots <- vector(num.plots, mode='list')
            dev.new(width=21/2.54, height=29.7/2.54, file=paste("Unknown_",l,".pdf", sep="")) #A4 pdf
            # par(mfrow=c(perpage,2))
            layout(matrix(c(1,1,2,3,4,4,5,6,7,7,8,9), 6, 2, byrow = TRUE), widths=rep(c(1,1),perpage), heights=rep(c(1,5),perpage))
            # layout.show(6)
            oma.saved <- par("oma")
            par(oma = rep.int(0, 4))
            par(oma = oma.saved)
            o.par <- par(mar = rep.int(0, 4))
            on.exit(par(o.par))
            stop=0 #initialize
            for (i in 1:num.plots) {
                start=stop+1
                stop=start+perpage-1 #
                for (c in start:stop){
                    if (c <=dim(mat)[2]){
                            
                        #get sample name
                        sampname<-basename(resGC$xset[[c]]@xcmsSet@filepaths)

                        #remove .cdf, .mzXML filepattern
                        filepattern <- c("[Cc][Dd][Ff]", "[Nn][Cc]", "([Mm][Zz])?[Xx][Mm][Ll]", 
                                "[Mm][Zz][Dd][Aa][Tt][Aa]", "[Mm][Zz][Mm][Ll]")
                        filepattern <- paste(paste("\\.", filepattern, "$", sep = ""), 
                                collapse = "|")
                        sampname<-gsub(filepattern, "",sampname)
                         
                        title1<-paste("unknown", l,"from",sampname, sep=" ")
                        an<-resGC$xset[[c]]
                         
                        par (mar=c(0, 0, 0, 0) + 0.1)
                        plot.new()
                        box()
                        text(0.5, 0.5, title1, cex=2)
                        if (!is.na(mat[l,c])){
                            pcgrp=allPCGRPs[[c]][which(allPCGRPs[[c]][,1]==mat[l,c]),2]
                            if (pcgrp!=mat[l,c]) print ("pcgrp changed")
                            par (mar=c(3, 2.5, 3, 1.5) + 0.1)
                            plotEICs(an, pspec=pcgrp, maxlabel=2)
                            plotPsSpectrum(an, pspec=pcgrp, maxlabel=2)
                        } else {
                            plot.new()
                            box()
                            text(0.5, 0.5, "NOT FOUND", cex=2)
                            plot.new()
                            box()
                            text(0.5, 0.5, "NOT FOUND", cex=2)
                        }
                    }
                 }
                # my.plots[[i]] <- recordPlot()
            }
            graphics.off()

                # pdf(file=paste("Unknown_",l,".pdf", sep=""), onefile=TRUE)
                # for (my.plot in my.plots) {
                    # replayPlot(my.plot)
                # }
                # my.plots
                # graphics.off()

        }#end  for l
    }#end if unkn=""
    else{
        par (mar=c(5, 4, 4, 2) + 0.1)
        l=unkn
        if (length(l)==1){
            #recordPlot
            perpage=3 #if change change layout also!
            num.plots <- ceiling(dim(mat)[2]/perpage) #three pcgroup per page
            my.plots <- vector(num.plots, mode='list')
            
            dev.new(width=21/2.54, height=29.7/2.54, file=paste("Unknown_",l,".pdf", sep="")) #A4 pdf
            # par(mfrow=c(perpage,2))
            layout(matrix(c(1,1,2,3,4,4,5,6,7,7,8,9), 6, 2, byrow = TRUE), widths=rep(c(1,1),perpage), heights=rep(c(1,5),perpage))
            # layout.show(6)
            oma.saved <- par("oma")
            par(oma = rep.int(0, 4))
            par(oma = oma.saved)
            o.par <- par(mar = rep.int(0, 4))
            on.exit(par(o.par))
            stop=0 #initialize
            for (i in 1:num.plots) {
                start=stop+1
                stop=start+perpage-1 #
                for (c in start:stop){
                    if (c <=dim(mat)[2]){
                            
                        #get sample name
                        sampname<-basename(resGC$xset[[c]]@xcmsSet@filepaths)

                        #remove .cdf, .mzXML filepattern
                        filepattern <- c("[Cc][Dd][Ff]", "[Nn][Cc]", "([Mm][Zz])?[Xx][Mm][Ll]", "[Mm][Zz][Dd][Aa][Tt][Aa]", "[Mm][Zz][Mm][Ll]")
                        filepattern <- paste(paste("\\.", filepattern, "$", sep = ""), collapse = "|")
                        sampname<-gsub(filepattern, "",sampname)
                         
                        title1<-paste("unknown", l,"from",sampname, sep=" ")
                        an<-resGC$xset[[c]]
                         
                        par (mar=c(0, 0, 0, 0) + 0.1)
                        plot.new()
                        box()
                        text(0.5, 0.5, title1, cex=2)
                        if (!is.na(mat[l,c])){
                            pcgrp=allPCGRPs[[c]][which(allPCGRPs[[c]][,1]==mat[l,c]),2]
                            if (pcgrp!=mat[l,c]) print ("pcgrp changed")
                            par (mar=c(3, 2.5, 3, 1.5) + 0.1)
                            plotEICs(an, pspec=pcgrp, maxlabel=2)
                            plotPsSpectrum(an, pspec=pcgrp, maxlabel=2)
                        } else {
                            plot.new()
                            box()
                            text(0.5, 0.5, "NOT FOUND", cex=2)
                            plot.new()
                            box()
                            text(0.5, 0.5, "NOT FOUND", cex=2)
                        }
                    }
                }
                # my.plots[[i]] <- recordPlot()
            }
            graphics.off()

            # pdf(file=paste("Unknown_",l,".pdf", sep=""), onefile=TRUE)
            # for (my.plot in my.plots) {
                # replayPlot(my.plot)
            # }
            # my.plots
            # graphics.off()
        } else {
            par (mar=c(5, 4, 4, 2) + 0.1)
            for (l in 1:length(unkn)){ #l=2
                #recordPlot
                perpage=3 #if change change layout also!
                num.plots <- ceiling(dim(mat)[2]/perpage) #three pcgroup per page
                my.plots <- vector(num.plots, mode='list')
                dev.new(width=21/2.54, height=29.7/2.54, file=paste("Unknown_",unkn[l],".pdf", sep="")) #A4 pdf
                # par(mfrow=c(perpage,2))
                layout(matrix(c(1,1,2,3,4,4,5,6,7,7,8,9), 6, 2, byrow = TRUE), widths=rep(c(1,1),perpage), heights=rep(c(1,5),perpage))
                # layout.show(6)
                oma.saved <- par("oma")
                par(oma = rep.int(0, 4))
                par(oma = oma.saved)
                o.par <- par(mar = rep.int(0, 4))
                on.exit(par(o.par))
                stop=0 #initialize
                for (i in 1:num.plots) {
                    start=stop+1
                    stop=start+perpage-1 #
                    for (c in start:stop){
                        if (c <=dim(mat)[2]){
                        
                            #get sample name
                            sampname<-basename(resGC$xset[[c]]@xcmsSet@filepaths)

                            #remove .cdf, .mzXML filepattern
                            filepattern <- c("[Cc][Dd][Ff]", "[Nn][Cc]", "([Mm][Zz])?[Xx][Mm][Ll]", "[Mm][Zz][Dd][Aa][Tt][Aa]", "[Mm][Zz][Mm][Ll]")
                            filepattern <- paste(paste("\\.", filepattern, "$", sep = ""), collapse = "|")
                            sampname<-gsub(filepattern, "",sampname)
                             
                            title1<-paste("unknown",unkn[l],"from",sampname, sep=" ")
                            an<-resGC$xset[[c]]
                             
                            par (mar=c(0, 0, 0, 0) + 0.1)
                            plot.new()
                            box()
                            text(0.5, 0.5, title1, cex=2)
                            if (!is.na(mat[unkn[l],c])){
                                pcgrp=allPCGRPs[[c]][which(allPCGRPs[[c]][,1]==mat[unkn[l],c]),2]
                                if (pcgrp!=mat[unkn[l],c]) print ("pcgrp changed")
                                par (mar=c(3, 2.5, 3, 1.5) + 0.1)
                                plotEICs(an, pspec=pcgrp, maxlabel=2)
                                plotPsSpectrum(an, pspec=pcgrp, maxlabel=2)
                            } else {
                                plot.new()
                                box()
                                text(0.5, 0.5, "NOT FOUND", cex=2)
                                plot.new()
                                box()
                                text(0.5, 0.5, "NOT FOUND", cex=2)
                            }
                        }
                    }
                    # my.plots[[i]] <- recordPlot()
                }
                graphics.off()

                # pdf(file=paste("Unknown_",unkn[l],".pdf", sep=""), onefile=TRUE)
                # for (my.plot in my.plots) {
                    # replayPlot(my.plot)
                # }
                # my.plots
                # graphics.off()

            }#end  for unkn[l]
        
        }
    
    }
} #end function