inst/ProstarApp/server/srv_ConvertData.R

callModule(modulePopover,"modulePopover_convertChooseDatafile", 
           data = reactive(list(title = HTML(paste0("<strong><font size=\"4\">Data file</font></strong>")), 
                                content="Select one (.txt, .csv, .tsv, .xls, .xlsx) file.")))

callModule(modulePopover,"modulePopover_convertIdType", 
           data = reactive(list(title = HTML(paste0("<strong><font size=\"4\">ID definition</font></strong>")), 
                                content="If you choose the automatic ID, Prostar will build an index.")))



callModule(modulePopover,"modulePopover_convertProteinID", 
           data = reactive(list(title = HTML(paste0("<strong><font size=\"4\">Select protein IDs</font></strong>")), 
                                content="Select the column containing the parent protein IDs.")))


callModule(modulePopover,"modulePopover_convertDataQuanti", 
           data = reactive(list(title = HTML(paste0("<strong><font size=\"4\">Quantitative data</font></strong>")), 
                                content="Select the columns that are quantitation values by clicking in the field below.")))

callModule(moduleStaticDataTable,"overview_convertData", table2show=reactive({GetDatasetOverview()}))




##--------------------------------------------------------------
## Gestion du slideshow
##--------------------------------------------------------------


output$checkConvertPanel <- renderUI({
  rv$tab1
  rv$pageConvert
  color <- rep("lightgrey",NUM_PAGES_CONVERT)
  
  ##Step 1
  if (rv$pageConvert >= 1){
    res <- !is.null(rv$tab1)
    ifelse(res, color[1] <- "green", color[1] <- "red")
    toggleState(id = "nextBtnConvert", condition = (rv$pageConvert < NUM_PAGES_CONVERT) && res)
    toggleState(id = "prevBtnConvert", condition = rv$pageConvert > 1)
    hide(selector = ".page")
  }
  
  ##Step 2: Choose data ID
  
  if (rv$pageConvert >= 2){
    res1 <- !is.null(input$colnameForID) && ((input$colnameForID == "AutoID") || datasetID_Ok())
    res2 <- !is.null(input$convert_proteinId) && (input$convert_proteinId != "")
    
    ifelse(res1 && res2, color[2] <- "green", color[2] <- "red")
    toggleState(id = "nextBtnConvert", condition = (rv$pageConvert < NUM_PAGES_CONVERT) && res1 && res2)
    toggleState(id = "prevBtnConvert", condition = rv$pageConvert > 1)
    hide(selector = ".page")
  } 
  
  ## Step 3: Choose quantitative data
  if (rv$pageConvert >= 3){
    res <- !is.null(input$eData.box) && checkIdentificationMethod_Ok()
    
    ifelse(res, color[3] <- "green", color[3] <- "red")
    toggleState(id = "nextBtnConvert", condition = (rv$pageConvert < NUM_PAGES_CONVERT) && res)
    toggleState(id = "prevBtnConvert", condition = rv$pageConvert > 1)
    hide(selector = ".page")
  }
  
  if (rv$pageConvert >= 4){
    res <- isTRUE(rv$designChecked$valid)
    ifelse(res, color[4] <- "green", color[4] <- "red")
    toggleState(id = "nextBtnConvert", condition = (rv$pageConvert < NUM_PAGES_CONVERT) && res)
    toggleState(id = "prevBtnConvert", condition = rv$pageConvert > 1)
    hide(selector = ".page")
  }
  
  if (rv$pageConvert >= 5){
    res <- TRUE
    ifelse(!is.null(rv$current.obj), color <- rep("green",NUM_PAGES_CONVERT), color[5] <- "red")
    toggleState(id = "nextBtnConvert", condition = (rv$pageConvert < NUM_PAGES_CONVERT) && res)
    toggleState(id = "prevBtnConvert", condition = rv$pageConvert > 1)
    hide(selector = ".page")
  }
  
  txt <- c("Select file", "Select ID", "Select quantitative data", "Build design", "Convert")
  buildTable(txt, color)
})

NUM_PAGES_CONVERT <- 5

# observe({
#   toggleState(id = "prevBtnConvert", condition = rv$pageConvert > 1)
#  # toggleState(id = "nextBtnConvert", condition = rv$pageConvert < NUM_PAGES_CONVERT)
#   hide(selector = ".page")
# })

navPageConvert <- function(direction) {
  rv$pageConvert <- rv$pageConvert + direction
}

observeEvent(input$prevBtnConvert, navPageConvert(-1))
observeEvent(input$nextBtnConvert, navPageConvert(1))

##---------------------------------------------------------------
##------------------------------------------------------------------



#################################
output$Convert_SelectFile <- renderUI({
  #if (rv$pageConvert != 1){return()}
  
  
  tagList(br(), br(),
          fluidRow(
            column(width=2, modulePopoverUI("modulePopover_convertChooseDatafile")),
            column(width = 10, fileInput("file1", "", 
                                         multiple=FALSE, 
                                         accept=c(".txt", ".tsv", ".csv",".xls", ".xlsx")))),
          #actionButton("loadData2Convert", "Load data file",class = actionBtnClass),
          uiOutput("ManageXlsFiles"),
          # helpText("Hint : before importing quantification 
          #             file data, check the syntax of your text 
          #             file."),
          br(),
          uiOutput("ConvertOptions")
  )
})


output$Convert_DataId <- renderUI({
  #if (rv$pageConvert != 2){return()}
  
  tagList(
    
    br(), br(),
    #uiOutput("helpTextDataID"),
    
    tags$div(
      tags$div( style="display:inline-block; vertical-align: top; padding-right: 100px;",
                uiOutput("id"),
                uiOutput("warningNonUniqueID")
      ),
      tags$div( style="display:inline-block; vertical-align: top;",
                uiOutput("convertChooseProteinID_UI"),
                uiOutput("previewProteinID_UI")
      )
    )
  )
})




output$Convert_ExpFeatData <- renderUI({
  #if (rv$pageConvert != 3){return()}
  
  tagList(
    fluidRow(
      column(width=4,checkboxInput("selectIdent", 
                                   "Select columns for identification method", 
                                   value = FALSE)),
      column(width=4,uiOutput("checkIdentificationTab"))
    ),
    fluidRow(
      column(width=4,uiOutput("eData",width = "400px")),
      column(width=8,DT::dataTableOutput("x1", width='500px'))),
    tags$script(HTML("Shiny.addCustomMessageHandler('unbind-DT', function(id) {
                                   Shiny.unbindAll($('#'+id).find('table').DataTable().table().node());
                                   })"))
  )
})





output$Convert_BuildDesign <- renderUI({
  #if (rv$pageConvert != 4){return()}
  req(input$file1)
  tagList(
    tags$p("If you do not know how to fill the experimental design, you can click
                                  on the '?' next to each design in the list that appear once the conditions 
                                  are checked or got to the ", 
           actionLink("linkToFaq1", "FAQ",style="background-color: white"), 
           " page."),
    fluidRow(
      column(width=6,tags$b("1 - Fill the \"Condition\" column to identify the conditions to compare.")),
      column(width=6,uiOutput("UI_checkConditions")  )
    ),
    fluidRow(
      column(width=6,uiOutput("UI_hierarchicalExp")),
      column(width=6,uiOutput("checkDesign") )
    ),
    hr(),
    selectInput("convert_reorder", "Order by conditions ?",
                choices=c("No"="No", "Yes"="Yes"),
                width="100px"),
    tags$div(
      
      tags$div(style="display:inline-block; vertical-align: top;",
               uiOutput("viewDesign",width="100%")
      ),
      tags$div(style="display:inline-block; vertical-align: top;",
               shinyjs::hidden(div(id = "showExamples", uiOutput("designExamples") ))
      )
    )
    
  )
  
})






output$Convert_Convert <- renderUI({
  #if (rv$pageConvert != 5){return()}
  
  tagList(
    br(), br(),
    
    uiOutput("convertFinalStep"),
    moduleStaticDataTableUI("overview_convertData"),
    uiOutput("conversionDone"),
    p("Once the 'Load' button (above) clicked, you will be automatically redirected to Prostar home page. The dataset will be accessible within Prostar 
    interface and processing menus will be enabled. However, all importing functions ('Open MSnset', 'Demo data' and 'Convert data') will be disabled 
    (because successive dataset loading can make Prostar unstable). To work on another dataset, use first the 'Reload Prostar' functionality from 
    the 'Dataset manager' menu: it will make Prostar restart with a fresh R session where import functions are enabled.")
    
  )
})

output$warningNonUniqueID <- renderUI({
  req(input$colnameForID)
  req(rv$tab1)
  if (input$colnameForID =="AutoID") {return(NULL)  }
  
  t <- (length(as.data.frame(rv$tab1)[, input$colnameForID])
        == length(unique(as.data.frame(rv$tab1)[, input$colnameForID])))
  
  if (!t){
    text <- "<img src=\"images/Problem.png\" height=\"24\"></img><font color=\"red\">
        Warning ! Your ID contains duplicate data.
        Please choose another one."
    
  }
  else {
    text <- "<img src=\"images/Ok.png\" height=\"24\"></img>"
  }
  HTML(text)
})


output$convertChooseProteinID_UI <- renderUI({
  req(rv$tab1)
  
  if (input$typeOfData == "protein") {return(NULL)}
  
  .choices <- c("",colnames(rv$tab1))
  names(.choices) <- c("",colnames(rv$tab1))
  tagList(
    modulePopoverUI("modulePopover_convertProteinID"),
    selectInput("convert_proteinId", 
                "",
                choices =  .choices , selected = character(0))
  )
})




#########################################################
output$id <- renderUI({
  req(rv$tab1)
  
  .choices <- c("AutoID",colnames(rv$tab1))
  names(.choices) <- c("Auto ID",colnames(rv$tab1))
  
  tagList(
    modulePopoverUI("modulePopover_convertIdType"),
    selectInput("colnameForID", label = "", choices = .choices)
  )
  
})



output$ConvertOptions <- renderUI({
  input$file1
  if (is.null(input$file1)){return(NULL)}
  
  tagList(
    radioButtons("typeOfData", 
                 "Is it a peptide or protein dataset ?", 
                 choices=c("peptide dataset" = "peptide", 
                           "protein dataset" = "protein")
    )
    
    ,radioButtons("checkDataLogged", 
                  "Are your data already log-transformed ?", 
                  #width = widthWellPanel, 
                  choices=c("yes (they stay unchanged)" = "yes", 
                            "no (they wil be automatically transformed)"="no"), 
                  selected="no")
    ,br()
    ,checkboxInput("replaceAllZeros", 
                   "Replace all 0 and NaN by NA", 
                   value= TRUE)
  )
})


observeEvent(input$fData.box,ignoreInit = TRUE,{
  
  choices = colnames(rv$tab1)[-which(colnames(rv$tab1) %in% input$fData.box)]
  names(choices) = 
    colnames(rv$tab1)[-which(colnames(rv$tab1) %in% input$fData.box)]
  updateSelectInput(session, "eData.box", 
                    label = "",
                    choices = choices,
                    selected = choices)
  
})




output$helpTextDataID <- renderUI({
  input$typeOfData
  if (is.null(input$typeOfData)){return(NULL)}
  t <- ""
  switch(input$typeOfData,
         protein = {t <- "proteins"},
         peptide = {t <- "peptides"}
  )
  txt <- paste ("Please select among the columns of your data the one that 
                corresponds to a unique ID of the ", t, ".", sep=" ")
  helpText(txt)
  
})



############ Read text file to be imported ######################
observeEvent(c(input$file1,input$XLSsheets),{
  input$XLSsheets
  if (((GetExtension(input$file1$name)== "xls")
       || (GetExtension(input$file1$name) == "xlsx") )
      && is.null(input$XLSsheets)) {return(NULL)  }
  
  authorizedExts <- c("txt","csv", "tsv","xls","xlsx")
  if( is.na(match(GetExtension(input$file1$name), authorizedExts))) {
    shinyjs::info("Warning : this file is not a text nor an Excel file ! 
                  Please choose another one.")
  }
  else {
    # result = tryCatch(
    #   {
    ClearUI()
    ClearMemory()
    ext <- GetExtension(input$file1$name)
    shinyjs::disable("file1")
    switch(ext,
           txt = { rv$tab1 <- read.csv(input$file1$datapath,  header=TRUE, sep="\t", as.is=T)},
           csv = { rv$tab1 <- read.csv(input$file1$datapath,  header=TRUE, sep="\t", as.is=T)},
           tsv = { rv$tab1 <- read.csv(input$file1$datapath,  header=TRUE, sep="\t", as.is=T)},
           xls = { rv$tab1 <- readExcel(input$file1$datapath, ext, sheet=input$XLSsheets)},
           xlsx = {rv$tab1 <- readExcel(input$file1$datapath, ext, sheet=input$XLSsheets)}
    )
    #   }
    #   , warning = function(w) {
    #     shinyjs::info(conditionMessage(w))
    #   }, error = function(e) {
    #     shinyjs::info(paste("Read text file to convert",":",
    #                         conditionMessage(e), 
    #                         sep=" "))
    #   }, finally = {
    #     #cleanup-code 
    #   })
  }
  #shinyjs::disable('file1')
  
})






output$conversionDone <- renderUI({
  rv$current.obj
  if (is.null(rv$current.obj)) { return(NULL)}
  
  h4("The conversion is done. Your dataset has been automatically loaded 
       in memory. Now, you can switch to the Descriptive statistics panel to 
       vizualize your data.")
  
})








#####-------------------------------------------------------
output$ManageXlsFiles <- renderUI({
  req(input$file1)
  
  .ext <- GetExtension(input$file1$name)
  if ((.ext == "xls") || (.ext == "xlsx")){ 
    sheets <- listSheets(input$file1$datapath)
    selectInput("XLSsheets", "sheets", choices = as.list(sheets),
                width='200px')
  }
  
})







###########################################

#
#
##   Quanti data table

##
##
##############################################
output$eData <- renderUI({
  input$file1
  rv$tab1
  if (is.null(rv$tab1)) {return(NULL)  }
  
  choices <- colnames(rv$tab1)
  names(choices) <- colnames(rv$tab1)
  
  tagList(
    modulePopoverUI("modulePopover_convertDataQuanti"),
    selectInput("eData.box",
                label = "",
                choices = choices,
                multiple = TRUE, width='200px',
                size = 20,
                selectize = FALSE)
  )
})





# updateInputs <- function(id, n){
#     for (i in seq_len(n)) {
#         updateSelectInput(paste0(id,i),label=NULL,selected = input[[paste0(id,i)]])
#     }
# }







output$checkIdentificationTab <- renderUI({
  req(input$selectIdent)
  if (!isTRUE(input$selectIdent)){return(NULL)}
  
  shinyValue("colForOriginValue_",length(input$eData.box))
  temp <- shinyValue("colForOriginValue_",length(input$eData.box))
  
  if ((length(which(temp == "None")) == length(temp)))
  {
    img <- "images/Ok.png"
    txt <- "Correct"
  }  else {
    if (length(which(temp == "None")) > 0)
    {
      img <- "images/Problem.png"
      txt <- "The identification method is not appropriately defined for each sample."
    } else {
      if(length(temp) != length(unique(temp))){
        img <- "images/Problem.png"
        txt <- "There are duplicates in identification columns."
      }else { 
        img <- "images/Ok.png"
        txt <- "Correct"
      }
    }
  }
  tags$div(
    tags$div(
      tags$div(style="display:inline-block;",tags$img(src = img, height=25)),
      tags$div(style="display:inline-block;",tags$p(txt))
    )
  )
  
  
})



# reactive dataset
quantiDataTable <- reactive({
 
  req(input$eData.box)
  req(rv$tab1)
  session$sendCustomMessage('unbind-DT', 'x1')
  df <- NULL
  choices <- c("None",colnames(rv$tab1))
  names(choices) <- c("None",colnames(rv$tab1))
  
  if (isTRUE(input$selectIdent)) {
    
    df <- data.frame(input$eData.box,
                     shinyInput(selectInput,
                                "colForOriginValue_",
                                nrow(as.data.frame(input$eData.box)),
                                choices=choices))
    colnames(df) <- c("Sample", "Identification method")
  } else {
    df <- data.frame(Sample = as.data.frame(input$eData.box))
    colnames(df) <- c("Sample")
  }
  df
})



output$x1 <- renderDataTable(
  quantiDataTable(),
  escape=FALSE,
  rownames = FALSE,
  extensions = c('Scroller', 'Buttons'),
  server=FALSE,
  selection='none', 
  class = 'compact',
  options=list(
    preDrawCallback=JS(
      'function() {
            Shiny.unbindAll(this.api().table().node());}'),
    drawCallback= JS(
      'function(settings) {
            Shiny.bindAll(this.api().table().node());}'),
    # rowCallback = JS("function(r,d) {$(r).attr('height', '10px')}"),
    dom = 'Bfrtip',
    autoWidth=TRUE,
    deferRender = TRUE,
    bLengthChange = FALSE,
    scrollX = 200,
    scrollY = 500,
    scroller = TRUE,
    ajax = list(url = dataTableAjax(session, quantiDataTable()))
    
  )
  
)


#observeEvent(shinyValue("colForOriginValue_",nrow(quantiDataTable())),{})


checkIdentificationMethod_Ok <- reactive({
  req(quantiDataTable())
  res <- TRUE
  tmp <- NULL
  if (isTRUE(input$selectIdent)) {
    tmp <- shinyValue("colForOriginValue_",nrow(quantiDataTable()))
    if ((length(grep("None", tmp)) > 0)  || (sum(is.na(tmp)) > 0)){ res <- FALSE }
  } 
  res
  
})


datasetID_Ok <- reactive({
  req(input$colnameForID)
  req(rv$tab1)
  if (input$colnameForID == "AutoID") {t <- TRUE}
  else {
    t <- (length(as.data.frame(rv$tab1)[, input$colnameForID])
          == length(unique(as.data.frame(rv$tab1)[, input$colnameForID])))
  }
  t
})



output$previewProteinID_UI <- renderUI({
  req(input$convert_proteinId)
  if (input$convert_proteinId == "") {return (NULL)}
  
  tagList(
    p(style="color: black;", 'Preview'),
    tableOutput("previewProtID")
  )
  
})



output$previewProtID <- renderTable(
  # req(input$convert_proteinId),
  head(rv$tab1[,input$convert_proteinId]),
  colnames = FALSE
)




output$warningCreateMSnset <- renderUI({
  if (isTRUE(input$selectIdent)){
    colNamesForOriginofValues <- shinyValue("colForOriginValue_",nrow(quantiDataTable()))
    if (length(which(colNamesForOriginofValues == "None")) >0){
      text <- "<font color=\"red\"> Warning: The MSnset cannot be created because the identification 
            method are not fully filled.  <br>"
      HTML(text)
    }
  }
})






#######################################
observeEvent(input$createMSnsetButton,ignoreInit =  TRUE,{
  # if(is.null(input$createMSnsetButton) || (input$createMSnsetButton == 0)) 
  #{return(NULL)}
  
  colNamesForOriginofValues <- NULL
  if (isTRUE(input$selectIdent)) {
    colNamesForOriginofValues <- shinyValue("colForOriginValue_",nrow(quantiDataTable()))
    if (length(which(colNamesForOriginofValues == "None")) >0){ return (NULL)   }
  } 
  
  isolate({
    result = tryCatch(
      {
        ext <- GetExtension(input$file1$name)
        txtTab <-  paste("tab1 <- read.csv(\"", input$file1$name,
                         "\",header=TRUE, sep=\"\t\", as.is=T)",  sep="")
        txtXls <-  paste("tab1 <- read.xlsx(",input$file1$name,
                         ",sheet=", input$XLSsheets,")",sep="")
        switch(ext,
               txt = writeToCommandLogFile(txtTab),
               csv = writeToCommandLogFile(txtTab),
               tsv = writeToCommandLogFile(txtTab),
               xls= writeToCommandLogFile(txtXls),
               xlsx = writeToCommandLogFile(txtXls)
        )
        
        input$filenameToCreate
        rv$tab1
        
        tmp.eData.box <- input$eData.box
        indexForEData <- match(tmp.eData.box, colnames(rv$tab1))
        if (!is.null(rv$newOrder)){
          tmp.eData.box <- tmp.eData.box[rv$newOrder]
          indexForEData <- indexForEData[rv$newOrder]
        }
        
        indexForFData <- seq(1,ncol(rv$tab1))[-indexForEData]
        
       # indexForIDBox <- NULL
       # if (input$colnameForID !="Auto ID") {
      #    indexForIDBox <- match(input$colnameForID, colnames(rv$tab1))
      #  }
        
        
        metadata <- hot_to_r(input$hot)
        logData <- (input$checkDataLogged == "no")
        
        
        indexForOriginOfValue <- NULL
        if (!is.null(colNamesForOriginofValues) && (length(grep("None", colNamesForOriginofValues))==0)  && (sum(is.na(colNamesForOriginofValues)) == 0)){
          for (i in 1:length(tmp.eData.box)){
            indexForOriginOfValue <- c(indexForOriginOfValue, which(colnames(rv$tab1) == input[[paste0("colForOriginValue_", i)]]))
          }
        }
        
        
        versions <- list(Prostar_Version = 
                           installed.packages(lib.loc = Prostar.loc)["Prostar","Version"],
                         DAPAR_Version = 
                           installed.packages(lib.loc = DAPAR.loc)["DAPAR","Version"]
        )
        options(digits=15)
        
        protId <- NULL
        if (input$typeOfData == 'protein')
          protId <- input$colnameForID
        else if(input$typeOfData == 'peptide') 
          protId <- input$convert_proteinId
        
        tmp <- DAPAR::createMSnset(file = rv$tab1, 
                                   metadata, 
                                   indexForEData, 
                                   indexForFData, 
                                   colnameForID = input$colnameForID,
                                   indexForOriginOfValue,
                                   logData, 
                                   input$replaceAllZeros,
                                   pep_prot_data = input$typeOfData,
                                   proteinId =  gsub(".", "_", protId, fixed=TRUE),
                                   versions
        )
        ClearUI()
        ClearMemory()
        rv$current.obj <- tmp
        
        rv$current.obj.name <- input$filenameToCreate
        rv$indexNA <- which(is.na(exprs(rv$current.obj)))
        
        l.params <- list(filename = input$filenameToCreate)
        
        loadObjectInMemoryFromConverter()
        
        updateTabsetPanel(session, "tabImport", selected = "Convert")
        rv$pageConvert <- 5
      }
      , warning = function(w) {
        if (conditionMessage(w) %in% c("NaNs produced", "production de NaN")){
          shinyjs::info(paste("Warning : Your original dataset may contain negative values",
                              "so that they cannot be logged. Please check back the dataset or", 
                              "the log option in the first tab.",
                              sep=" "))
        } else {
          shinyjs::info(paste("Warning in CreateMSnSet",":",
                              conditionMessage(w), 
                              sep=" "))
        }
      }, error = function(e) {
        shinyjs::info(paste("Error :","CreateMSnSet",":",
                            conditionMessage(e), 
                            sep=" "))
      }, finally = {
        #cleanup-code 
      })
    
    
    
  })
})

Try the Prostar package in your browser

Any scripts or data that you put into this service are public.

Prostar documentation built on April 11, 2021, 6 p.m.