comparison AskoR.R @ 2:f7dde9cca31e draft default tip

planemo upload for repository https://github.com/genouest/galaxy-tools/tree/master/tools/askor commit 994ecff7807bb0eb9dac740d67ad822415b0b464

1:ea4a79798dcf

  print(table(ASKO_stat$Expression))
  colnames(ASKOlist$stat.table)[colnames(ASKOlist$stat.table)=="gene"] <- paste("is", "gene", sep="@")                  # header formatting for askomics
  colnames(ASKOlist$stat.table)[colnames(ASKOlist$stat.table)=="contrast"] <- paste("measured_in", "Contrast", sep="@") # header formatting for askomics
  o <- order(ASKOlist$stat.table$FDR)                                                                                   # ordering genes by FDR value
  ASKOlist$stat.table<-ASKOlist$stat.table[o,]
  #
  dir.create(parameters$out_dir)
  write.table(ASKOlist$stat.table,paste0(parameters$out_dir,"/", parameters$organism, contrasko, ".txt"),                                    #
              sep=parameters$sep, col.names = T, row.names = F, quote=FALSE)

  if(parameters$heatmap==TRUE){
    cpm_gstats<-cpm(dge, log=TRUE)[o,][1:parameters$numhigh,]
    heatmap.2(cpm_gstats, cexRow=0.5, cexCol=0.8, scale="row", labCol=dge$samples$Name, xlab=contrast, Rowv = FALSE, dendrogram="col")
  }

  return(ASKOlist)

      count<-read.table(parameters$fileofcount, header=TRUE, sep = "\t", row.names = parameters$col_genes, comment.char = "")
    }
    select_counts<-row.names(samples)
    #countT<-count[,c(parameters$col_counts:length(colnames(count)))]
    countT<-count[,select_counts]
    dge<-DGEList(counts=countT, samples=samples)
    # if(is.null(parameters$select_sample)==FALSE){
    #   slct<-grep(parameters$select_sample, colnames(countT))
    #   countT<-countT[,slct]
    # }

         cex=0.5)
  return(filtered_counts)
}

GEnorm <- function(filtered_GE, parameters){
  filtered_cpm <- cpm(filtered_GE, log=TRUE)   #nouveau calcul Cpm sur donn?es filtr?es, si log=true alors valeurs cpm en log2
  colnames(filtered_cpm)<-rownames(filtered_GE$samples)
  boxplot(filtered_cpm,
          col=filtered_GE$samples$color,         #boxplot des scores cpm non normalis?s
          main="A. Before normalization",
          cex.axis=0.5,

  lcpm<-cpm(dge, log=TRUE)
  colnames(lcpm)<-rownames(dge$samples)
  cormat<-cor(lcpm)
 # color<- colorRampPalette(c("yellow", "white", "green"))(20)
  color<-colorRampPalette(c("black","red","yellow","white"),space="rgb")(28)
  heatmap(cormat, col=color, symm=TRUE,RowSideColors =as.character(dge$samples$color), ColSideColors = as.character(dge$samples$color))
  #MDS
  mds <- cmdscale(dist(t(lcpm)),k=3, eig=TRUE)
  eigs<-round((mds$eig)*100/sum(mds$eig[mds$eig>0]),2)

  mds1<-ggplot(as.data.frame(mds$points), aes(V1, V2, label = rownames(mds$points))) + labs(title="MDS Axes 1 and 2") + geom_point(color =as.character(dge$samples$color) ) + xlab(paste('dim 1 [', eigs[1], '%]')) +ylab(paste('dim 2 [', eigs[2], "%]")) + geom_label_repel(aes(label = rownames(mds$points)), color = 'black',size = 3.5)

2:f7dde9cca31e

  print(table(ASKO_stat$Expression))
  colnames(ASKOlist$stat.table)[colnames(ASKOlist$stat.table)=="gene"] <- paste("is", "gene", sep="@")                  # header formatting for askomics
  colnames(ASKOlist$stat.table)[colnames(ASKOlist$stat.table)=="contrast"] <- paste("measured_in", "Contrast", sep="@") # header formatting for askomics
  o <- order(ASKOlist$stat.table$FDR)                                                                                   # ordering genes by FDR value
  ASKOlist$stat.table<-ASKOlist$stat.table[o,]
  
  #
  dir.create(parameters$out_dir)
  write.table(ASKOlist$stat.table,paste0(parameters$out_dir,"/", parameters$organism, contrasko, ".txt"),                                    #
              sep=parameters$sep, col.names = T, row.names = F, quote=FALSE)
  
  
  if(parameters$heatmap==TRUE){
    numhigh=parameters$numhigh
    if (numhigh>length(o)) {numhigh=length(o)}
    cpm_gstats<-cpm(dge, log=TRUE)[o,][1:numhigh,]
    heatmap.2(cpm_gstats, cexRow=0.5, cexCol=0.8, scale="row", labCol=dge$samples$Name, xlab=contrast, Rowv = FALSE, dendrogram="col")
  }

  return(ASKOlist)

      count<-read.table(parameters$fileofcount, header=TRUE, sep = "\t", row.names = parameters$col_genes, comment.char = "")
    }
    select_counts<-row.names(samples)
    #countT<-count[,c(parameters$col_counts:length(colnames(count)))]
    countT<-count[,select_counts]
    #print(countT)
    dge<-DGEList(counts=countT, samples=samples)
    # if(is.null(parameters$select_sample)==FALSE){
    #   slct<-grep(parameters$select_sample, colnames(countT))
    #   countT<-countT[,slct]
    # }

         cex=0.5)
  return(filtered_counts)
}

GEnorm <- function(filtered_GE, parameters){
  filtered_cpm=log2(1000000*filtered_GE$counts/colSums(filtered_GE$counts))
  #filtered_cpm <- cpm(filtered_GE, log=TRUE, normalized.lib.sizes=TRUE)   #nouveau calcul Cpm sur donn?es filtr?es, si log=true alors valeurs cpm en log2
  colnames(filtered_cpm)<-rownames(filtered_GE$samples)
  boxplot(filtered_cpm,
          col=filtered_GE$samples$color,         #boxplot des scores cpm non normalis?s
          main="A. Before normalization",
          cex.axis=0.5,

  lcpm<-cpm(dge, log=TRUE)
  colnames(lcpm)<-rownames(dge$samples)
  cormat<-cor(lcpm)
 # color<- colorRampPalette(c("yellow", "white", "green"))(20)
  color<-colorRampPalette(c("black","red","yellow","white"),space="rgb")(28)
  heatmap.2(cormat, col=color, symm=TRUE,RowSideColors =as.character(dge$samples$color), ColSideColors = as.character(dge$samples$color))
  #MDS
  mds <- cmdscale(dist(t(lcpm)),k=3, eig=TRUE)
  eigs<-round((mds$eig)*100/sum(mds$eig[mds$eig>0]),2)

  mds1<-ggplot(as.data.frame(mds$points), aes(V1, V2, label = rownames(mds$points))) + labs(title="MDS Axes 1 and 2") + geom_point(color =as.character(dge$samples$color) ) + xlab(paste('dim 1 [', eigs[1], '%]')) +ylab(paste('dim 2 [', eigs[2], "%]")) + geom_label_repel(aes(label = rownames(mds$points)), color = 'black',size = 3.5)