inst/src/tabs/gprofiler/svr_gprofiler.R

In LUMC/dgeAnalysis: dgeAnalysis

## Enrichment table from all term data
output[["enrich_table"]] <- DT::renderDataTable({
  tryCatch({
    checkReload()
    enrich <- clean_enrich()
    colnames(enrich) <-
      c(
        "source",
        "enrichedTerm",
        "P.Value",
        "termSize",
        "querySize",
        "geneCount",
        "significant",
        "genes"
      )
    DT::datatable(enrich, options = list(pageLength = 50, scrollX = TRUE))
  }, error = function(err) {
    return(DT::datatable(data.frame(c(
      "No data available in table"
    )), rownames = FALSE, colnames = ""))
  })
})

## Show the default gprofiler plot
output[["enrich_gprofiler"]] <- renderPlotly({
  tryCatch({
    checkReload()
    gostplot(inUse_enrich, capped = FALSE, interactive = TRUE)
  }, error = function(err) {
    return(NULL)
  })
})

## Get a dataframe to use in plots from enrich object
clean_enrich <- reactive({
  tryCatch({
    checkReload()
    enrich_processed <- inUse_enrich$result
    enrich_processed <-
      enrich_processed[c(
        "source",
        "term_name",
        "p_value",
        "term_size",
        "query_size",
        "intersection_size",
        "significant",
        "intersection"
      )]
    rownames(enrich_processed) <- inUse_enrich$result$term_id
    enrich_processed <- enrich_processed[order(enrich_processed$p_value), ]
    enrich_processed
  }, error = function(err) {
    return(NULL)
  })
})

## create barplot with gprofilers enrichment terms
output[["enrich_barplot"]] <- renderPlotly({
  tryCatch({
    checkReload()
    
    ## Get input data
    enrich <- clean_enrich()
    plot_data <- enrich_bar(enrich, input$terms_slider)
    text <- 'paste("Source:", source,
                  "\nTerm:", term_name,
                  "\nGenes:", intersection_size,
                  "\nP-Value:", round(p_value, 5))'
    
    ## Create plot
    ggplotly(
      bar_plot(
        df = plot_data,
        x = "intersection_size",
        y = "term_name",
        text = text,
        fill = "p_value",
        title = "Enrichment barplot",
        xlab = "Number of genes",
        ylab = ""
      ) + labs(fill = "P-Value"),
      tooltip = "text"
    )
  }, error = function(err) {
    return(NULL)
  })
})

## create barplot with the number of DE genes in term
output[["enrich_DEbarplot"]] <- renderPlotly({
  tryCatch({
    checkReload()
    
    ## Get input data
    enrich <- clean_enrich()
    plot_data <- enrich_barDE(enrich, input$DEterms_slider, inUse_deTab)
    text <- 'paste("Term:", name,
                  "\nGenes:", abs(values),
                  "\nRegulation:", ind)'
    
    ## Create plot
    ggplotly(
      bar_plot(
        df = plot_data,
        x = "values",
        y = "name",
        text = text,
        fill = "ind",
        title = "DE genes in terms",
        xlab = "Number of genes",
        ylab = ""
      ),
      tooltip = "text"
    )
  }, error = function(err) {
    return(NULL)
  })
})

## create cnet plot with gprofiler enrichment terms
output[["cnet_plot"]] <- renderPlotly({
  tryCatch({
    checkReload()
    
    ## Get input data
    enrich <- clean_enrich()
    geneSets <- extract_geneSets(enrich,
                                 input$cnet_slider,
                                 input$select_pathway)
    graphData <- cnetPlotly(enrich, geneSets, inUse_deTab)
    plot_data <- cnet_data(inUse_deTab, graphData, (input$cnet_slider + length(input$select_pathway)))
    text <- 'paste("Node:", genes,
                  "\nLog2FC:", round(fc, 2))'
    
    ## Create plot
    ggplotly(
      network_plot(
        df = plot_data,
        text = text,
        label1 = input$cnet_annoP,
        label2 = input$cnet_annoG,
        title = "Gene-Concept Network"
      ) + labs(color = "Log2FC"),
      tooltip = "text"
    )
  }, error = function(err) {
    return(NULL)
  })
})

## Add specific pathway to cnet plot
output[["cnet_select_pathway"]] <- renderUI({
  tryCatch({
    enrich <- clean_enrich()
    selectInput(
      inputId = "select_pathway",
      label = "Add specific pathway:",
      multiple = TRUE,
      choices = c("Click to add pathway" = "", enrich$term_name)
    )
  }, error = function(err) {
    return(NULL)
  })
})

## get all genes of cnet plot
output[["cnet_table"]] <- DT::renderDataTable({
  tryCatch({
    checkReload()
    
    enrich <- clean_enrich()
    geneSets <- extract_geneSets(enrich,
                                 input$cnet_slider,
                                 input$select_pathway)
    
    graphData <- cnetPlotly(enrich, geneSets, inUse_deTab)
    DT::datatable(inUse_deTab[rownames(inUse_deTab) %in% names(V(graphData)),], options = list(pageLength = 15, scrollX = TRUE))
  }, error = function(err) {
    return(DT::datatable(data.frame(c(
      "No data available in table"
    )), rownames = FALSE, colnames = ""))
  })
})


## INFORMATION BOXES

output[["gProfiler2_plot_info"]] <- renderUI({
  infoText <-
    "The gProfiler 2 manhattan plot shows the terms colored by source (database).
        Each term is represented by a circle. The size of a circle represents the term
        size (number of genes). The y-axis shows the adjusted p-value in a (negative)
        log10 scale."
  informationBox(infoText)
})

output[["enrich_barplot_info"]] <- renderUI({
  infoText <-
    "The bar graph shows a sorted list of the most enriched terms. The bar chart is sorted based on
        p-value. The colors of the bars are also generated based on p-value. on the X-axis,
        the amount of genes linked to the pathway is displayed."
  informationBox(infoText)
})

output[["enrich_DEbarplot_info"]] <- renderUI({
  infoText <-
    "The bar graph shows a sorted list of the most enriched terms. The bar chart is sorted based on
        p-value. The plot shows the number of DE genes (up or down) regulated in an enrichment term."
  informationBox(infoText)
})

output[["enrich_cnet_info"]] <- renderUI({
  infoText <-
    "The concept network can show which genes are involved in the most significant terms. The 
        most enriched terms along with all associated genes are collected and displayed in a
        network plot. Some pathways may contain some matching genes. These genes will also
        connected. The color given to genes is based on the log-fold change determined after the
        expression analysis. Ultimately, this plot shows the connection of genes between the most
        important road."
  informationBox(infoText)
})