inst/AmpGram/server.R

In AmpGram: Prediction of Antimicrobial Peptides

library(shiny)
library(ggplot2)
library(AmpGram)
library(AmpGramModel)
library(ranger)
library(DT)
library(shinythemes)
library(rmarkdown)
library(pander)

source("shiny-server-utils.R")

data(AmpGram_model)

options(shiny.maxRequestSize=10*1024^2)

shinyServer(function(input, output) {
  
  prediction <- reactive({
    
    if (!is.null(input[["seq_file"]]))
      input_sequences <- read_txt(input[["seq_file"]][["datapath"]])
    input[["use_area"]]
    isolate({
      if (!is.null(input[["text_area"]]))
        if(input[["text_area"]] != "")
          input_sequences <- read_txt(textConnection(input[["text_area"]]))
    })
    
    if(exists("input_sequences")) {
      if(length(input_sequences) > 50) {
        #dummy error, just to stop further processing
        stop("Too many sequences. Please use AmpGram locally.")
      } else {
        if(any(lengths(input_sequences) < 10)) {
          #dummy error, just to stop further processing
          stop("The minimum length of the sequence is 10 amino acids.")
        } else {
          predict_in_shiny(AmpGram_model, input_sequences)
        }
      }
    } else {
      NULL
    }
  })
  
  decision_table <- reactive({
    if(!is.null(prediction())) {
      pred2df(prediction())
    }
  })
  
  output[["decision_table"]] <- renderDataTable({
    df <- decision_table()
    colnames(df) <- c("Protein name", "AMP probability")
    my_DT(df) %>% 
      formatRound(2, 4) 
    
  })
  
  selected_proteins <- reactive({
    validate(
      need(input[["decision_table_rows_selected"]], 
           "Select at least one row in the Results table")
    )
    
    prediction()[input[["decision_table_rows_selected"]]]
  })
  
  detailed_preds <- reactive(({
    
    selected_pred_data <- selected_proteins()
    
    detailed_pred_list <- lapply(1L:length(selected_pred_data), function(ith_pred_id) {
      ith_pred <- selected_pred_data[[ith_pred_id]]
      
      data.frame(seq_name = names(selected_pred_data)[ith_pred_id],
                 start = 1L:length(ith_pred[["all_mers_pred"]]), 
                 end = 1L:length(ith_pred[["all_mers_pred"]]) + 9, 
                 pred = ith_pred[["all_mers_pred"]],
                 decision = factor(ith_pred[["all_mers_pred"]] > 0.5, 
                                   levels = c("FALSE", "TRUE"),
                                   labels = c("No", "Yes")))
    })
    
    names(detailed_pred_list) <- names(selected_pred_data)
    detailed_pred_list
    
  }))
  
  
  
  output[["detailed_preds"]] <- renderUI({
    detailed_preds_list <- lapply(1L:length(detailed_preds()), function(i) {
      list(plotOutput(paste0("detailed_plot", i)),
           dataTableOutput(paste0("detailed_table", i)))
    })
    c(list(downloadButton("download_long_graph", "Download long output (with graphics)")),
      do.call(tagList, unlist(detailed_preds_list, recursive = FALSE)))
  })
  
  
  for (i in 1L:50) {
    local({
      my_i <- i
      
      output[[paste0("detailed_plot", my_i)]] <- renderPlot({
        #browser()
        plot_single_protein(detailed_preds()[[my_i]])
        })
      output[[paste0("detailed_table", my_i)]] <- renderDataTable(AMP_DT(get_AMPs(selected_proteins()[[my_i]])))
    })
  }
  
  output[["detailed_tab"]] <- renderUI({
    uiOutput("detailed_preds")
  })
  
  output[["dynamic_ui"]] <- renderUI({
    if (!is.null(input[["seq_file"]]))
      input_sequences <- read_txt(input[["seq_file"]][["datapath"]])
    input[["use_area"]]
    isolate({
      if (!is.null(input[["text_area"]]))
        if(input[["text_area"]] != "")
          input_sequences <- read_txt(textConnection(input[["text_area"]]))
    })
    
    if(exists("input_sequences")) {
      fluidRow(
        tags$p(HTML("<h3><A HREF=\"javascript:history.go(0)\">Start a new query</A></h3>"))
      )
    } else {
      fluidRow(
        h4("Exemplary sequences"),
        pre(includeText("prots.txt"))
      )
    }
  })
  
  output[["dynamic_tabset"]] <- renderUI({
    if(is.null(prediction())) {
      
      tabPanel(title = "Sequence input",
               tags$textarea(id = "text_area", style = "width:90%",
                             placeholder="Paste sequences (FASTA format required) here...", 
                             rows = 22, cols = 60, ""),
               tags$p(""),
               actionButton("use_area", "Submit data from field above"),
               tags$p(""),
               fileInput('seq_file', 'Submit .fasta or .txt file:'))
      
      
    } else {
      tabsetPanel(
        tabPanel("Results",
                 tags$h4("Select at least one protein to produce detailed results."), 
                 dataTableOutput("decision_table")
        ),
        tabPanel("Detailed results",
                 uiOutput("detailed_tab")
        )
      )
    }
  })
  
  file_name <- reactive({
    if(is.null(input[["seq_file"]][["name"]])) {
      part_name <- "AmpGram_results"
    } else {
      part_name <- strsplit(input[["seq_file"]][["name"]], ".", fixed = TRUE)[[1]][1]
    }
    part_name
  })
  
  output[["download_long"]] <- downloadHandler(
    filename  = function() { 
      paste0(file_name(), "_pred.txt") 
    },
    content <- function(file) {
      sink(file, type = "output")
      cat("Input file name: ", ifelse(is.null(input[["seq_file"]][["name"]]), "none",
                                      input[["seq_file"]][["name"]]), "\n\n")
      cat(paste0("Date: ", Sys.time()), "\n\n")
      for (i in 1L:length(prediction())) {
        cat("\n\n")
        summary(prediction()[[i]])
        cat("\n\n")
      }
      sink()
    }
  )
  
  output[["download_long_graph"]] <- downloadHandler(
    filename  = function() { 
      paste0(file_name(), "_pred.html") 
    },
    content <- function(file) {
      src <- normalizePath("AmpGram-report.Rmd")
      
      # temporarily switch to the temp dir, in case you do not have write
      # permission to the current working directory
      owd <- setwd(tempdir())
      on.exit(setwd(owd))
      file.copy(src, "AmpGram-report.Rmd")
      out <- render("AmpGram-report.Rmd", output_format = "html_document", file, quiet = TRUE)
      file.rename(out, file)
    }
  )
  
})