R/polya_shiny_app.R
In smaegol/nanotail: Analysis and visualisation of Nanopolish-based polyA lengths estimations
Defines functions
nanoTailApp
Documented in
nanoTailApp
# TODO ShinyApp: Allow to choose columns shown in the table
# TODO ShinyApp: Allow launching on raw polya data.frame, calculate statistics using selected factor in the app
# TODO ShinyApp: Export report
# TODO allow statistics on more than 2 levels of factor (anova?)
# TODO Add withProgress
#' wrapper for NanoTail Shiny interface
#'
#' @param polya_table polyA predictions table. Can be obtained with \link{read_polya_multiple}
#' @param precomputed_polya_statistics precomputed differential adenylation table (obtained with \link{calculate_polya_stats})
#' @param precomputed_annotations precomputed annotations
#'
#' @export
#'
nanoTailApp <- function(polya_table,precomputed_polya_statistics=NA,precomputed_annotations=NA) {
if ( !requireNamespace('shiny',quietly = TRUE) ) {
stop("NanoTail requires 'shiny'. Please install it using
install.packages('shiny')")
}
if ( !requireNamespace('shiny',quietly = TRUE) ) {
stop("NanoTail requires 'shinydashboard'. Please install it using
install.packages('shinydashboard')")
}
if ( !requireNamespace('shinycssloaders',quietly = TRUE) ) {
stop("NanoTail requires 'shinycssloaders'. Please install it using
install.packages('shinycssloaders')")
}
if ( !requireNamespace('plotly',quietly = TRUE) ) {
stop("NanoTail requires 'plotly'. Please install it using
install.packages('plotly')")
}
#check if parameters are provided
if (missing(polya_table)) {
stop("The polyA lengths table (argument polya_table) is missing",
call. = FALSE)
}
assertthat::assert_that(assertive::has_rows(polya_table),msg = "Empty data frame provided as an input (polya_table). Please provide the correct polya_table")
assertthat::assert_that("polya_length" %in% colnames(polya_table),msg = "Input polya_table should contain polya length predictions. Did you provide valid input?")
assertthat::assert_that("transcript" %in% colnames(polya_table),msg = "Input polya_table should contain 'transcript' column. Did you provide valid input?")
assertthat::assert_that("sample_name" %in% colnames(polya_table),msg = "Input polya_table should contain at least `sample_name` column for grouping. If data were read using read_polya_single() function please use read_polya_single(...,sample_name= SAMPLE_NAME ) or read_polya_multiple() to provide metadata")
assertthat::assert_that(is.factor(polya_table$sample_name),msg = "Sample_name column should be a factor")
# Calculate processing statistics
# TODO Move to reactive values
#remove failed reads from polya_table before further analysis
polya_table_passed <- remove_failed_reads(polya_table)
# make sure transcript names are not factor
polya_table_passed$transcript <- as.character(polya_table_passed$transcript)
# Convert polya_table to data.table (if required)
# TODO check if is already datatable or not
polya_table_passed <- data.table::data.table(polya_table_passed)
data.table::setkey(polya_table_passed,"transcript")
number_of_samples = length(levels(polya_table_passed$sample_name))
number_of_transcripts = length(levels(as.factor(polya_table_passed$transcript)))
### !<- colorpicker start ->! ###
## colorpicker based on Zbyszek Pietras Bar Plot app (https://github.com/zetp/Bar_plot_NC/blob/master/app.R)
### here is code to make color scale picker (palette picker) UI input
### it is taken from shinyWidgets website: https://dreamrs.github.io/shinyWidgets/articles/palette_picker.html
# List of palettes
colors_pal <- list()
colors_pal$ggsci_based_colors <- c("npg", "aaas", "nejm", "lancet", "jama",
"jco", "ucscgb", "d3", "locuszoom",
"igv", "uchicago", "startrek", "tron",
"futurama", "rickandmorty", "simpsons",
"gsea")
colors_pal$RColorBrewer_qualitative = rownames(RColorBrewer::brewer.pal.info[RColorBrewer::brewer.pal.info$category=="qual",])
# Get all colors given a palette name(s)
get_brewer_name <- function(name) {
pals <- RColorBrewer::brewer.pal.info[rownames(RColorBrewer::brewer.pal.info) %in% name, ]
if(nrow(pals)>0) {
res <- lapply(
X = seq_len(nrow(pals)),
FUN = function(i) {
RColorBrewer::brewer.pal(n = pals$maxcolors[i], name = rownames(pals)[i])
}
)
unlist(res)
}
else {
eval(parse(text = paste0("ggsci::pal_",name,"()")))(7)
}
}
background_pals <- sapply(unlist(colors_pal, use.names = FALSE), get_brewer_name)
# Calc linear gradient for CSS
linear_gradient <- function(cols) {
x <- round(seq(from = 0, to = 100, length.out = length(cols)+1))
ind <- c(1, rep(seq_along(x)[-c(1, length(x))], each = 2), length(x))
m <- matrix(data = paste0(x[ind], "%"), ncol = 2, byrow = TRUE)
res <- lapply(
X = seq_len(nrow(m)),
FUN = function(i) {
paste(paste(cols[i], m[i, 1]), paste(cols[i], m[i, 2]), sep = ", ")
}
)
res <- unlist(res)
res <- paste(res, collapse = ", ")
paste0("linear-gradient(to right, ", res, ");")
}
background_pals <- unlist(lapply(X = background_pals, FUN = linear_gradient))
colortext_pals <- rep(c("white", "black"), times = sapply(colors_pal, length))
### !<- colorpicker end ->! ###
processing_info_global <- get_nanopolish_processing_info(polya_table,grouping_factor = NA)
processing_info_global_spread <- processing_info_global %>% tidyr::spread(qc_tag,n)
processing_info_per_sample <- get_nanopolish_processing_info(polya_table,grouping_factor = "sample_name")
processing_info_per_sample_spread <- processing_info_per_sample %>% tidyr::spread(qc_tag,n)
polya_summary_table <- summarize_polya(polya_data = polya_table_passed,summary_factors = "sample_name")
if (!is.na(precomputed_polya_statistics)) {
#assertthat::assert_that(is.list(precomputed_polya_statistics),msg="Please provide the output of calculate_polya_stats() as precomputed_polya_statistics")
#assertthat::assert_that("summary" %in% names(precomputed_polya_statistics),msg="Please provide the output of calculate_polya_stats() as precomputed_polya_statistics")
initial_summary_table = precomputed_polya_statistics %>% dplyr::select(transcript,dplyr::ends_with("_counts"),dplyr::ends_with("_polya_median"),p.value,padj,contains("effect_size"))
initial_table_for_volcano <- precomputed_polya_statistics %>% dplyr::select(transcript,fold_change,padj,significance)
}
else {
if (number_of_samples>1) {
initial_summary_table <- polya_summary_table %>% dplyr::select(transcript,sample_name,polya_median) %>% tidyr::spread(sample_name,polya_median)
}
else {
initial_summary_table <- polya_summary_table
}
initial_table_for_volcano <- initial_summary_table
}
initial_summary_table_counts <- polya_summary_table %>% dplyr::select(transcript,sample_name,counts) %>% tidyr::spread(sample_name,counts)
grouping_factor_levels <- polya_table_passed %>% dplyr::select_if(is.factor) %>% colnames
nanotail_shiny_ui <- shinydashboard::dashboardPage(
skin="blue",
# dashboard header definition -----------------------------------------------------------
shinydashboard::dashboardHeader(
title = paste0("NanoTail - Interactive exploration of polyA lengths estimations ",
"done using Nanopore sequencing - version ",
packageVersion("nanotail")),
titleWidth = 900),
# dashboard sidebar definition -----------------------------------------------------------
shinydashboard::dashboardSidebar(
width = 180,
shinydashboard::sidebarMenu(id="nanotail_menu",
shinydashboard::menuItem("QC info", icon = shiny::icon("info-square"), tabName = "basicInfo"),
shinydashboard::menuItem("Global polyA distribution", icon = shiny::icon("chart-line"), tabName = "global_distr"),
shinydashboard::menuItem("Differential adenylation", icon = shiny::icon("dna"), tabName = "diff_polya"),
shinydashboard::menuItem("Differential expression", icon = shiny::icon("dna"), tabName = "diff_exp"),
shinydashboard::menuItem("Annotation_analysis", icon = shiny::icon("dna"), tabName = "annotation_analysis_inactive"),
shinydashboard::menuItem("Plot settings", icon = shiny::icon("settings"), tabName = "plot_settings"),
shinydashboard::menuItem("About", tabName = "dashboard", icon = shiny::icon("info"))
),
shiny::selectInput("groupingFactor","Group by",choices=grouping_factor_levels),
shiny::uiOutput("condition1UI"),
shiny::uiOutput("condition2UI"),
shiny::checkboxInput("plot_only_selected_conditions","Plot only selected conditions",value=FALSE),
shiny::actionButton("clear_selection",
shiny::HTML("Clear selection"),
icon = shiny::icon("spinner"))
),
# dashboard body definition ---------------------------------------------------------
shinydashboard::dashboardBody(
shinydashboard::tabItems(
shinydashboard::tabItem(tabName = "dashboard",
shiny::fillPage(padding = 0, title = NULL, bootstrap = F, theme = NULL,
shiny::wellPanel(style = "background-color: #ffffff; overflow-y:scroll; max-height: 750px;",
shiny::includeMarkdown(system.file("extdata", "about.md",package = "nanotail")))
)),
shinydashboard::tabItem(tabName = "basicInfo",
shiny::fluidRow(
shinydashboard::infoBoxOutput("numberOfSamples"),
shinydashboard::infoBoxOutput("numberOfTranscripts"),
shinydashboard::infoBoxOutput("numberOfAllReads")),
shiny::fluidRow(
shiny::column(1),
shiny::column(7,shiny::fluidRow(shiny::h4("Per sample plot:")),
shiny::fluidRow(
plotly::plotlyOutput("nanopolish_qc_summary_plot"),
shiny::checkboxInput("show_frequency_plot_nanopolioshQC",label = "Show QC as frequencies?",value = TRUE))),
shiny::column(1),
shiny::column(3,shiny::fluidRow(shiny::h4("Information for all samples:")),
shiny::fluidRow(shinydashboard::infoBoxOutput("numberOfPassReads",width = NULL)),
shiny::fluidRow(shinydashboard::infoBoxOutput("numberOfAdapterReads",width = NULL)),
shiny::fluidRow(shinydashboard::infoBoxOutput("numberOfReadLoadFailedReads",width = NULL)),
shiny::fluidRow(shinydashboard::infoBoxOutput("numberOfNoregionReads",width = NULL))))),
shinydashboard::tabItem(tabName = "plot_settings",
shinydashboard::box(
shinyWidgets::pickerInput(inputId = "col_palette", label = "Colour scale",
choices = colors_pal, selected = "Set1", width = "100%",
choicesOpt = list(
content = sprintf(
"<div style='width:100%%;padding:5px;border-radius:4px;background:%s;color:%s'>%s</div>",
unname(background_pals), colortext_pals, names(background_pals)))),
#shiny::uiOutput("Colorblind"),
shiny::sliderInput("scale_limit_low","Plot scale limit low",0,1024,0,10),
shiny::sliderInput("scale_limit_high","Plot scale limit high",50,1024,200,10),
shiny::sliderInput("counts_scatter_low_value","Counts scatter lower counts limit",0,10024,0,50),
shiny::sliderInput("counts_scatter_high_value","Counts scatter high counts limit",100,10024,1000,50),
shiny::selectInput("center_values_for_distribution_plot",choices=c("none","median","mean","gm_mean","mode"),selected = "median",label = "Show vertical lines with centered values on density plots?")
)),
shinydashboard::tabItem(tabName = "global_distr",
shiny::h2("Global comparison"),
shiny::fluidRow(
#shinydashboard::box(plotly::plotlyOutput('polya_global') %>% shinycssloaders::withSpinner(type = 4),collapsible=TRUE))),
shinydashboard::box(plotly::plotlyOutput('polya_global') %>% shinycssloaders::withSpinner(type = 4),collapsible=TRUE))),
shinydashboard::tabItem(tabName = "diff_polya",
shiny::fluidRow(
shinydashboard::box(DT::dataTableOutput('diff_polya'),collapsible = TRUE,height = '90%'),
shinydashboard::tabBox(title = "per transcript plots",id="tabset1", height="550px",
shiny::tabPanel("boxplot",shiny::plotOutput('polya_boxplot') %>% shinycssloaders::withSpinner(type = 4),
shiny::selectInput("color_by","Color by",choices=grouping_factor_levels),shiny::checkboxInput("add_boxplot",value=FALSE,label = "Add boxplot?")),
shiny::tabPanel("distribution plot",plotly::plotlyOutput('polya_distribution') %>% shinycssloaders::withSpinner(type = 4)
),
shiny::tabPanel("isoforms_boxplot",shiny::plotOutput('isoforms_boxplot') %>% shinycssloaders::withSpinner(type = 4)
),
shiny::tabPanel("virtual gel",shiny::plotOutput('virtual_gel') %>% shinycssloaders::withSpinner(type = 4),
shiny::sliderInput("virtual_gel_bw",min=0.1,max=10,step=0.1,value=1,label = "smoothing bandwitdth to be used ")),
shiny::tabPanel("volcano plot polya",plotly::plotlyOutput('polya_volcano') %>% shinycssloaders::withSpinner(type = 4)))),
shiny::fluidRow(
shinydashboard::box(shiny::actionButton("compute_diff_polya",
shiny::HTML("Compute Polya statistics"),
icon = shiny::icon("spinner")),
shiny::checkboxInput("use_dwell_time_for_statistics",label = "Use dwell time instead of calculated polya_length for statistics",value = FALSE),
shiny::selectInput("polya_stat_test","statistical test to use",choices = c("Wilcoxon","KS"),selected = "Wilcoxon"),
shiny::sliderInput("min_reads_for_polya_stats","Minimal count of reads per transcript",0,100,10,1),collapsible = TRUE,width=12))),
shinydashboard::tabItem(tabName = "diff_exp",
shiny::fluidRow(
shinydashboard::box(DT::dataTableOutput('diff_exp_table')),
shinydashboard::tabBox(title = "plots",id="tabset2",height="500px",
#tabPanel("pca_biplot",plotOutput('pca_biplot') %>% shinycssloaders::withSpinner(type = 4)),
shiny::tabPanel("scatter plot of counts",shiny::actionButton("show_scatter_plot",
shiny::HTML("Show scatter plot"),
icon = shiny::icon("spinner")),plotly::plotlyOutput('counts_plot') %>% shinycssloaders::withSpinner(type = 4)),
shiny::tabPanel("volcano_plot",plotly::plotlyOutput('volcano') %>% shinycssloaders::withSpinner(type = 4)),
shiny::tabPanel("MA_plot",plotly::plotlyOutput('MAplot') %>% shinycssloaders::withSpinner(type = 4)))),
shiny::fluidRow(shiny::actionButton("compute_diff_exp",
shiny::HTML("Compute Differential Expression using Binomial Test"),
icon = shiny::icon("spinner")))),
shinydashboard::tabItem(tabName = "annotation_analysis",
shiny::fluidRow(
shinydashboard::box(DT::dataTableOutput('annotation_table')),
shinydashboard::tabBox(title = "annotation_plots",id="tabset3",height="500px",
#tabPanel("pca_biplot",plotOutput('pca_biplot') %>% shinycssloaders::withSpinner(type = 4)),
shiny::tabPanel("box_plot_of_annotations",plotly::plotlyOutput('annotations_box_plot') %>% shinycssloaders::withSpinner(type = 4),
shiny::uiOutput("select_annotation_factor_levelsUI")
),
shiny::tabPanel("distribution_plot_of_annotations",plotly::plotlyOutput('annotation_distribution_plot') %>% shinycssloaders::withSpinner(type = 4)))),
shiny::fluidRow(shiny::actionButton("get_annotables_annotation",
shiny::HTML("Annotate input dataset"),
icon = shiny::icon("spinner")),
shiny::selectInput("annotables_genome",label="Annotables genome:",choices = c("bdgp6","grch38","grcm38","grch37","rnor6","galgal5","wbcel235","mmul801"),selected="grch38",multiple = FALSE,selectize = TRUE),
shiny::uiOutput("select_annotation_factorUI")
)))))
# call to shiny app and server-siude functions
shiny::shinyApp(ui = nanotail_shiny_ui, server = function(input, output) {
## Reactive values definition ---------------------------------------------------------
values <- reactiveValues()
values$polya_table <- polya_table_passed
values$polya_statistics_summary_table = initial_summary_table
values$polya_table_for_volcano = initial_table_for_volcano
values$diffexp_summary_table = initial_summary_table_counts
values$polya_table_summarized <- polya_summary_table
values$processing_info_global <- processing_info_global
values$processing_info_global_spread <- processing_info_global_spread
values$processing_info_per_sample <- processing_info_per_sample
values$processing_info_per_sample_spread <- processing_info_per_sample_spread
values$diff_exp_grouping_factor = "sample_name"
if(!is.na(precomputed_annotations)) {
values$polya_table_annotables_annotated <- precomputed_annotations
}
else {
values$polya_table_annotables_annotated <- polya_table_passed %>% dplyr::mutate(annotation_group=factor(gsub("^(.).*","\\1",transcript)))
}
# get polyA data for currently selected transcript
data_transcript <- shiny::reactive({
summary_table = values$polya_statistics_summary_table
selected_row <- input$diff_polya_rows_selected
selected_transcript = summary_table[selected_row,]$transcript
data_transcript = subset(polya_table_passed, transcript==selected_transcript)
data_transcript <- data_transcript %>% dplyr::mutate(replicate = gsub(".*(.)$","\\1", sample_name))
data_transcript
})
# get polyA data for currently selected transcript
data_transcript_annot <- shiny::reactive({
summary_table = values$polya_statistics_summary_table
selected_row <- input$diff_polya_rows_selected
selected_transcript = summary_table[selected_row,]$transcript
data_transcript_annot = subset(values$polya_table_annotables_annotated, transcript==selected_transcript)
data_transcript_annot
})
# UI elements rendering ---------------------------------------------------------
## Generate select inputs for conditions used for comparison, based on selected groupingFactor: ---------------------------------------------------------
output$condition1UI <- shiny::renderUI({
group_factor = input$groupingFactor
shiny::selectizeInput("condition1_diff_exp","Condition 1",choices=levels(polya_table_passed[[group_factor]]),selected = levels(polya_table_passed[[group_factor]])[1])
})
output$condition2UI <- shiny::renderUI({
group_factor = input$groupingFactor
shiny::selectizeInput("condition2_diff_exp","Condition 2",choices=levels(polya_table_passed[[group_factor]]),selected = levels(polya_table_passed[[group_factor]])[2])
})
output$Colorblind <- shiny::renderUI({
if(RColorBrewer::brewer.pal.info %>% subset(rownames(.) == input$col_palette) %>% .$colorblind) {
txt_ <- "Yes"
} else {
txt_ <- "No"
}
shiny::HTML(paste("<p>Is colour scale colour blind friendly?<b>", txt_, "</b></p>"))
})
# Output elements rendering section ---------------------------------------------------------
## Use DT::datatable for tables
output$diff_polya = DT::renderDataTable(values$polya_statistics_summary_table %>% dplyr::rename_all(dplyr::funs(stringr::str_replace_all(.,"_"," "))), server = TRUE, selection=list(mode = 'multiple',selected = 1,target = "row"),options = list(dom = 'ftip',scrollX = TRUE))
output$diff_exp_table = DT::renderDataTable(values$diffexp_summary_table %>% dplyr::rename_all(dplyr::funs(stringr::str_replace_all(.,"_"," "))), server = TRUE, selection=list(mode = 'single',selected = 1,target = "row"),options = list(dom = 'ftip',scrollX = TRUE))
output$annotation_table = DT::renderDataTable(data_annotation(), server = TRUE, selection=list(mode = 'multiple',selected = 1,target = "row"),options = list(dom = 'ftip',scrollX = TRUE))
annotation_proxy <- DT::dataTableProxy('annotation_table')
## Show global polyA distribution as density plot
output$polya_global = plotly::renderPlotly({
if (input$plot_only_selected_conditions) {
global_distribution_plot <- plot_polya_distribution(polya_data = polya_table_passed,groupingFactor = input$groupingFactor,scale_x_limit_low = input$scale_limit_low,scale_x_limit_high = input$scale_limit_high,color_palette = input$col_palette,condition1 = input$condition1_diff_exp,condition2 = input$condition2_diff_exp,show_center_values=input$center_values_for_distribution_plot)
}
else {
global_distribution_plot <- plot_polya_distribution(polya_data = polya_table_passed,groupingFactor = input$groupingFactor,scale_x_limit_low = input$scale_limit_low,scale_x_limit_high = input$scale_limit_high,color_palette = input$col_palette,show_center_values=input$center_values_for_distribution_plot)
}
plotly::ggplotly(global_distribution_plot)
})
## Differential adenylation analysis section ---------------------------------------------------------
# Show boxplot of estimated polya lengths for selected transcript
output$polya_boxplot = shiny::renderPlot({
summary_table = values$polya_statistics_summary_table
if (length(input$diff_polya_rows_selected)>0) {
selected_row <- input$diff_polya_rows_selected
if (length(input$diff_polya_rows_selected)>1) {
selected_transcript = paste0(summary_table[selected_row,]$transcript,collapse=",")
}
else {
selected_transcript = summary_table[selected_row,]$transcript
}
cat(selected_transcript)
cat(selected_row)
data_transcript = data_transcript()
if (input$plot_only_selected_conditions) {
polya_boxplot <- plot_polya_violin(polya_data = data_transcript,groupingFactor = input$groupingFactor,scale_y_limit_low = input$scale_limit_low,scale_y_limit_high = input$scale_limit_high,color_palette = input$col_palette,plot_title = selected_transcript,condition1 = input$condition1_diff_exp,condition2 = input$condition2_diff_exp, add_boxplot=input$add_boxplot,fill_by=input$color_by,color_mode='fill')
}
else {
polya_boxplot <- plot_polya_violin(polya_data = data_transcript,groupingFactor = input$groupingFactor,scale_y_limit_low = input$scale_limit_low,scale_y_limit_high = input$scale_limit_high,color_palette = input$col_palette,plot_title = selected_transcript, add_boxplot=input$add_boxplot,fill_by=input$color_by,color_mode='fill')
}
#plotly::ggplotly(polya_boxplot)
print(polya_boxplot)
}
else {
#suppressWarnings(plotly::plotly_empty())
print(no_plot)
}
})
# Show boxplot of estimated polya lengths for selected transcript
output$isoforms_boxplot = shiny::renderPlot({
req(values$polya_table_annotables_annotated)
summary_table = values$polya_statistics_summary_table
if (length(input$diff_polya_rows_selected)>0) {
selected_row <- input$diff_polya_rows_selected
selected_transcript = summary_table[selected_row,]$transcript
data_transcript = data_transcript_annot()
polya_boxplot <- plot_polya_boxplot(polya_data = data_transcript,groupingFactor = "ensembl_transcript_id_short",scale_y_limit_low = input$scale_limit_low,scale_y_limit_high = input$scale_limit_high,color_palette = input$col_palette,plot_title = selected_transcript, violin=FALSE,additional_grouping_factor = input$groupingFactor)
polya_boxplot <- polya_boxplot + ggplot2::theme(axis.text.x = ggplot2::element_text(angle=90))
print(polya_boxplot)
#plotly::ggplotly(polya_boxplot)
}
else {
print("no data")
suppressWarnings(plotly::plotly_empty())
}
})
# Show boxplot of estimated polya lengths for selected transcript
output$virtual_gel = shiny::renderPlot({
summary_table = values$polya_statistics_summary_table
if (length(input$diff_polya_rows_selected)>0) {
selected_row <- input$diff_polya_rows_selected
selected_transcript = summary_table[selected_row,]$transcript
data_transcript = data_transcript()
polya_virtual_gel <- plot_virtual_gel(input_data = data_transcript,groupingFactor = input$groupingFactor,kernel_from =input$scale_limit_low,kernel_to = input$scale_limit_high,valuesColumn="polya_length",density_bw = input$virtual_gel_bw)
polya_virtual_gel$plot <- polya_virtual_gel$plot + ggplot2::theme(axis.text.x = ggplot2::element_text(angle=90))
print(polya_virtual_gel$plot)
#plotly::ggplotly(polya_boxplot)
}
else {
print("no data")
suppressWarnings(plotly::plotly_empty())
}
})
# Show denisty plot of estimated polya lengths for selected transcript
output$polya_distribution = plotly::renderPlotly({
summary_table = values$polya_statistics_summary_table
if (length(input$diff_polya_rows_selected)>0) {
selected_row <- input$diff_polya_rows_selected
if (length(input$diff_polya_rows_selected)>1) {
selected_transcript = paste0(summary_table[selected_row,]$transcript,collapse=",")
}
else {
selected_transcript = summary_table[selected_row,]$transcript
}
data_transcript = data_transcript()
if (input$plot_only_selected_conditions) {
transcript_distribution_plot <- plot_polya_distribution(polya_data = data_transcript,groupingFactor = input$groupingFactor,scale_x_limit_low = input$scale_limit_low,scale_x_limit_high = input$scale_limit_high,color_palette = input$col_palette, plot_title = selected_transcript,condition1 = input$condition1_diff_exp,condition2 = input$condition2_diff_exp,show_center_values=input$center_values_for_distribution_plot)
}
else {
transcript_distribution_plot <- plot_polya_distribution(polya_data = data_transcript,groupingFactor = input$groupingFactor,scale_x_limit_low = input$scale_limit_low,scale_x_limit_high = input$scale_limit_high,color_palette = input$col_palette, plot_title = selected_transcript,show_center_values=input$center_values_for_distribution_plot)
}
plotly::ggplotly(transcript_distribution_plot)
}
else {
suppressWarnings(plotly::plotly_empty())
}
})
# Show volcano plot of differential adenylation analysis
output$polya_volcano <- plotly::renderPlotly({
# show only if differential adenylation analysis was already done
if("fold_change" %in% colnames(values$polya_table_for_volcano)) {
volcanoPlot <- plot_volcano(input_data =values$polya_table_for_volcano,color_palette = input$col_palette,transcript_id_column = "transcript",labels=input$diff_polya_rows_selected)
plotly::ggplotly(volcanoPlot)
}
else {
shiny::showNotification("Launch differential polyadenylation analysis first",closeButton=TRUE,type="warning")
suppressWarnings(plotly::plotly_empty())
}
})
# Calculate differential adenylation if action button was clicked
shiny::observeEvent(input$compute_diff_polya,
{
if(number_of_samples>1) {
shiny::withProgress(message="Computing polya statistics...",
detail = "This step can take a little while",
value = 0.05,min=0,max=1,
{
condition1 = input$condition1_diff_exp
condition2 = input$condition2_diff_exp
shiny::incProgress(1/20)
polya_stats <- calculate_polya_stats(values$polya_table,grouping_factor = input$groupingFactor,condition1 = input$condition1_diff_exp,condition2 = input$condition2_diff_exp,stat_test=input$polya_stat_test,use_dwell_time=input$use_dwell_time_for_statistics,min_reads = input$min_reads_for_polya_stats,add_summary = TRUE,length_summary_to_show = "median")
shiny::incProgress(15/20)
#values$polya_statistics_summary_table <- polya_stats %>% dplyr::select(transcript,!!rlang::sym(paste0(condition1,"_counts")),!!rlang::sym(paste0(condition2,"_counts")),!!rlang::sym(paste0(condition1,"_polya_median")),!!rlang::sym(paste0(condition2,"_polya_median")),p.value,padj)
values$polya_statistics_summary_table <- polya_stats
values$polya_table_for_volcano <- polya_stats %>% dplyr::select(transcript,fold_change,padj,significance)
values$condition1 <- input$condition1_diff_exp
values$condition2 <- input$condition2_diff_exp
shiny::incProgress(3/20)
})
}
else {
shiny::showNotification("Differential analysis is not possible when only single sample is present",closeButton=TRUE,type="error")
}
})
diff_polya_proxy <- DT::dataTableProxy('diff_polya')
# Calculate differential adenylation if action button was clicked
shiny::observeEvent(input$clear_selection,
{
diff_polya_proxy %>% DT::selectRows(NULL)
})
## Differential expression analysis section ---------------------------------------------------------
# Show scatter plot of raw counts
output$counts_plot <- plotly::renderPlotly({
counts_scatter_plot <- plot_counts_scatter(polya_data = values$polya_table_summarized_scatter,groupingFactor = values$groupingFactor_scatter,color_palette = input$col_palette,condition1 = values$condition1_scatter, condition2 = values$condition2_scatter, min_counts = input$counts_scatter_low_value, max_counts = input$counts_scatter_high_value)
plotly::ggplotly(counts_scatter_plot)
})
# Show volcano plot of differential expression analysis
output$volcano <- plotly::renderPlotly({
# show only if differential expression analysis was already done
if("fold_change" %in% colnames(values$diffexp_summary_table)) {
volcanoPlot <- plot_volcano(input_data =values$diffexp_summary_table,color_palette = input$col_palette)
plotly::ggplotly(volcanoPlot)
}
else {
shiny::showNotification("Launch differential expression analysis first",closeButton=TRUE,type="warning")
suppressWarnings(plotly::plotly_empty())
}
})
# Show MA plot of differential expression analysis
output$MAplot <- plotly::renderPlotly({
# show only if differential expression analysis was already done
if("fold_change" %in% colnames(values$diffexp_summary_table)) {
MAplot <- plot_MA(input_data =values$diffexp_summary_table,color_palette = input$col_palette)
plotly::ggplotly(MAplot)
}
else {
shiny::showNotification("Launch differential expression analysis first",closeButton=TRUE,type="warning")
plotly::plotly_empty()
}
})
# Calculate differential expression if action button was clicked
shiny::observeEvent(input$compute_diff_exp,
{
if(number_of_samples>1) {
shiny::withProgress(message="Computing Differential epxression using binomial test...",
detail = "This step can take a little while",
value = 0.05,min=0,max=1,
{
values$polya_table_summarized <- summarize_polya(values$polya_table,summary_factors = c(input$groupingFactor))
shiny::incProgress(4/20)
values$polya_table_summarized_per_sample <- summarize_polya(values$polya_table,summary_factors = "sample_name")
shiny::incProgress(4/20)
values$diffexp_summary_table <- calculate_diff_exp_binom(values$polya_table_summarized,grouping_factor = input$groupingFactor,condition1 = input$condition1_diff_exp,condition2 = input$condition2_diff_exp,summarized_input=TRUE)
shiny::incProgress(10/20)
values$condition1 <- input$condition1_diff_exp
values$condition2 <- input$condition2_diff_exp
values$diff_exp_grouping_factor = input$groupingFactor
})
}
else {
shiny::showNotification("Differential analysis is not possible when only single sample is present",closeButton=TRUE,type="error")
}
})
# show scatter plot of counts if action button was clicked
shiny::observeEvent(input$show_scatter_plot,
{
if(number_of_samples>1) {
shiny::withProgress(message="Computing data for scatter counts plot...",
detail = "This step can take a little while",
value = 0.05,min=0,max=1,
{
values$polya_table_summarized_scatter <- summarize_polya(values$polya_table,summary_factors = c(input$groupingFactor))
shiny::incProgress(18/20)
values$condition1_scatter <- input$condition1_diff_exp
values$condition2_scatter <- input$condition2_diff_exp
values$groupingFactor_scatter = input$groupingFactor
})
}
else {
shiny::showNotification("Scatter plot cannot be shown when only single sample is present",closeButton=TRUE,type="error")
}
})
## Basic QC Info section ---------------------------------------------------------
output$numberOfSamples <- shinydashboard::renderInfoBox({
shinydashboard::infoBox(
"Samples in the analysis: ", number_of_samples, icon = shiny::icon("vials"),
color = "yellow"
)
})
output$numberOfTranscripts <- shinydashboard::renderInfoBox({
shinydashboard::infoBox(
"Transcripts analyzed ", number_of_transcripts, icon = shiny::icon("dna"),
color = "blue"
)
})
output$numberOfAllReads <- shinydashboard::renderInfoBox({
shinydashboard::infoBox(
"Reads processed ", sum(values$processing_info_global$n), icon = shiny::icon("stream"),
color = "black"
)
})
output$numberOfPassReads <- shinydashboard::renderInfoBox({
shinydashboard::infoBox(
"Reads passing filters ", values$processing_info_global_spread$PASS, icon = shiny::icon("check"),
color = "green"
)
})
output$numberOfAdapterReads <- shinydashboard::renderInfoBox({
shinydashboard::infoBox(
"Adapter reads ", values$processing_info_global_spread$ADAPTER, icon = shiny::icon("times"),
color = "red"
)
})
output$numberOfReadLoadFailedReads <- shinydashboard::renderInfoBox({
shinydashboard::infoBox(
"Reads failed to load ", values$processing_info_global_spread$READ_FAILED_LOAD, icon = shiny::icon("times"),
color = "red"
)
})
output$numberOfNoregionReads <- shinydashboard::renderInfoBox({
shinydashboard::infoBox(
"NoRegion reads ", values$processing_info_global_spread$NOREGION, icon = shiny::icon("times"),
color = "red"
)
})
#Show Nanopolish QC summary plot
output$nanopolish_qc_summary_plot <- plotly::renderPlotly({
nanopolish_qc_summary_plot<-plot_nanopolish_qc(values$processing_info_per_sample,color_palette = input$col_palette, frequency=input$show_frequency_plot_nanopolioshQC)
plotly::ggplotly(nanopolish_qc_summary_plot)
})
## Annotation analysis section ---------------------------------------------------------
data_selected_annotation <- shiny::reactive({
#summary_table = values$polya_table_annotated
selected_row <- input$annotation_table_rows_selected
data_annotation = data_annotation()
data_annotation = as.data.frame(data_annotation)
column_name = input$annotation_factor
print(paste0("column name: ",column_name))
selected_annotation = as.character(data_annotation[selected_row,column_name])
print("selected annotation:")
print(selected_annotation)
data_transcript = subset(values$polya_table_annotables_annotated, eval(parse(text=column_name)) %in% c(selected_annotation))
data_transcript <- data_transcript %>% dplyr::group_by(.dots = c("read_id",input$groupingFactor)) %>% dplyr::slice(1)
print("nrow data transc:")
print(nrow(data_transcript))
data_transcript
})
data_annotation <- shiny::reactive({
message(input$annotation_factor)
if (!is.null(input$annotation_factor)) {
data_annotation_summarized <- values$polya_table_annotables_annotated %>% dplyr::group_by(.dots = c(input$annotation_factor,input$groupingFactor)) %>% dplyr::summarize(polya_gm_mean=gm_mean(polya_length)) %>%
tidyr::spread(!!rlang::sym(input$groupingFactor),polya_gm_mean)
data_annotation_summarized
}
else {
shiny::showNotification(ui = "Annotation factor is empty",closeButton = TRUE,type = "message")
}
})
## Generate selectinputs for annotations:
output$select_annotation_factorUI <- shiny::renderUI({
annotation_factor_choices=values$polya_table_annotables_annotated %>% dplyr::select_if(is.factor) %>% dplyr::select(-grouping_factor_levels) %>% colnames
shiny::selectizeInput("annotation_factor","Column used for annotation comparisons",choices=annotation_factor_choices,selected = annotation_factor_choices[1])
})
output$select_annotation_factor_levelsUI <- shiny::renderUI({
annotation_factor_selected = input$annotation_factor
annotation_factor_possible_levels = levels(values$polya_table_annotables_annotated[[annotation_factor_selected]])
shiny::selectInput(inputId = "annotation_factor_levels",label="Which annotations to show on the plot",choices=annotation_factor_possible_levels,selected = annotation_factor_possible_levels,multiple = TRUE)
})
shiny::observeEvent(input$annotation_factor,{
annotation_proxy %>% DT::selectRows(NULL)
})
shiny::observeEvent(input$get_annotables_annotation,
{
shiny::withProgress(message="Annotating input polyA table with annotables...",
detail = "This step can take a little while",
value = 0.05,min=0,max=1,
{
values$polya_table_annotables_annotated <- annotate_with_annotables(values$polya_table,genome = input$annotables_genome)
values$polya_table_annotables_annotated <-data.table::data.table(values$polya_table_annotables_annotated)
shiny::incProgress(18/20)
data.table::setkeyv(values$polya_table_annotables_annotated,c("transcript","biotype","strand","chr"))
shiny::incProgress(1/20)
})
})
# Show denisty plot of estimated polya lengths for selected transcript
output$annotation_distribution_plot = plotly::renderPlotly({
#print(input$annotation_table_rows_selected)
#if (length(input$annotation_table_rows_selected)>0) {
selected_row <- input$annotation_table_rows_selected
print(selected_row)
data_transcript_annotated = data_selected_annotation()
#print(data_transcript_annotated)
if (input$plot_only_selected_conditions) {
transcript_distribution_plot <- plot_polya_distribution(polya_data = data_transcript_annotated,groupingFactor = input$groupingFactor,scale_x_limit_low = input$scale_limit_low,scale_x_limit_high = input$scale_limit_high,color_palette = input$col_palette,condition1 = input$condition1_diff_exp,condition2 = input$condition2_diff_exp,show_center_values=input$center_values_for_distribution_plot)
}
else {
transcript_distribution_plot <- plot_polya_distribution(polya_data = data_transcript_annotated,groupingFactor = input$groupingFactor,scale_x_limit_low = input$scale_limit_low,scale_x_limit_high = input$scale_limit_high,color_palette = input$col_palette,show_center_values=input$center_values_for_distribution_plot)
}
plotly::ggplotly(transcript_distribution_plot)
# }
# else {
# suppressWarnings(plotly::plotly_empty())
# }
})
output$annotations_box_plot = plotly::renderPlotly({
print(paste0("annot_factor: ",input$annotation_factor))
print(input$annotation_factor_levels)
transcript_distribution_plot <- plot_annotations_comparison_boxplot(annotated_polya_data = values$polya_table_annotables_annotated,grouping_factor = input$groupingFactor,annotation_factor = input$annotation_factor, annotation_levels = input$annotation_factor_levels, scale_y_limit_low = input$scale_limit_low,scale_y_limit_high = input$scale_limit_high,color_palette = input$col_palette,violin=FALSE)
plotly::ggplotly(transcript_distribution_plot) %>% plotly::layout(boxmode = "group")
})
})
}
smaegol/nanotail documentation built on June 29, 2024, 12:35 a.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Defines functions nanoTailApp
Documented in nanoTailApp
# TODO ShinyApp: Allow to choose columns shown in the table
# TODO ShinyApp: Allow launching on raw polya data.frame, calculate statistics using selected factor in the app
# TODO ShinyApp: Export report
# TODO allow statistics on more than 2 levels of factor (anova?)
# TODO Add withProgress
#' wrapper for NanoTail Shiny interface
#'
#' @param polya_table polyA predictions table. Can be obtained with \link{read_polya_multiple}
#' @param precomputed_polya_statistics precomputed differential adenylation table (obtained with \link{calculate_polya_stats})
#' @param precomputed_annotations precomputed annotations
#'
#' @export
#'
nanoTailApp <- function(polya_table,precomputed_polya_statistics=NA,precomputed_annotations=NA) {
if ( !requireNamespace('shiny',quietly = TRUE) ) {
stop("NanoTail requires 'shiny'. Please install it using
install.packages('shiny')")
}
if ( !requireNamespace('shiny',quietly = TRUE) ) {
stop("NanoTail requires 'shinydashboard'. Please install it using
install.packages('shinydashboard')")
}
if ( !requireNamespace('shinycssloaders',quietly = TRUE) ) {
stop("NanoTail requires 'shinycssloaders'. Please install it using
install.packages('shinycssloaders')")
}
if ( !requireNamespace('plotly',quietly = TRUE) ) {
stop("NanoTail requires 'plotly'. Please install it using
install.packages('plotly')")
}
#check if parameters are provided
if (missing(polya_table)) {
stop("The polyA lengths table (argument polya_table) is missing",
call. = FALSE)
}
assertthat::assert_that(assertive::has_rows(polya_table),msg = "Empty data frame provided as an input (polya_table). Please provide the correct polya_table")
assertthat::assert_that("polya_length" %in% colnames(polya_table),msg = "Input polya_table should contain polya length predictions. Did you provide valid input?")
assertthat::assert_that("transcript" %in% colnames(polya_table),msg = "Input polya_table should contain 'transcript' column. Did you provide valid input?")
assertthat::assert_that("sample_name" %in% colnames(polya_table),msg = "Input polya_table should contain at least `sample_name` column for grouping. If data were read using read_polya_single() function please use read_polya_single(...,sample_name= SAMPLE_NAME ) or read_polya_multiple() to provide metadata")
assertthat::assert_that(is.factor(polya_table$sample_name),msg = "Sample_name column should be a factor")
# Calculate processing statistics
# TODO Move to reactive values
#remove failed reads from polya_table before further analysis
polya_table_passed <- remove_failed_reads(polya_table)
# make sure transcript names are not factor
polya_table_passed$transcript <- as.character(polya_table_passed$transcript)
# Convert polya_table to data.table (if required)
# TODO check if is already datatable or not
polya_table_passed <- data.table::data.table(polya_table_passed)
data.table::setkey(polya_table_passed,"transcript")
number_of_samples = length(levels(polya_table_passed$sample_name))
number_of_transcripts = length(levels(as.factor(polya_table_passed$transcript)))
### !<- colorpicker start ->! ###
## colorpicker based on Zbyszek Pietras Bar Plot app (https://github.com/zetp/Bar_plot_NC/blob/master/app.R)
### here is code to make color scale picker (palette picker) UI input
### it is taken from shinyWidgets website: https://dreamrs.github.io/shinyWidgets/articles/palette_picker.html
# List of palettes
colors_pal <- list()
colors_pal$ggsci_based_colors <- c("npg", "aaas", "nejm", "lancet", "jama",
"jco", "ucscgb", "d3", "locuszoom",
"igv", "uchicago", "startrek", "tron",
"futurama", "rickandmorty", "simpsons",
"gsea")
colors_pal$RColorBrewer_qualitative = rownames(RColorBrewer::brewer.pal.info[RColorBrewer::brewer.pal.info$category=="qual",])
# Get all colors given a palette name(s)
get_brewer_name <- function(name) {
pals <- RColorBrewer::brewer.pal.info[rownames(RColorBrewer::brewer.pal.info) %in% name, ]
if(nrow(pals)>0) {
res <- lapply(
X = seq_len(nrow(pals)),
FUN = function(i) {
RColorBrewer::brewer.pal(n = pals$maxcolors[i], name = rownames(pals)[i])
}
)
unlist(res)
}
else {
eval(parse(text = paste0("ggsci::pal_",name,"()")))(7)
}
}
background_pals <- sapply(unlist(colors_pal, use.names = FALSE), get_brewer_name)
# Calc linear gradient for CSS
linear_gradient <- function(cols) {
x <- round(seq(from = 0, to = 100, length.out = length(cols)+1))
ind <- c(1, rep(seq_along(x)[-c(1, length(x))], each = 2), length(x))
m <- matrix(data = paste0(x[ind], "%"), ncol = 2, byrow = TRUE)
res <- lapply(
X = seq_len(nrow(m)),
FUN = function(i) {
paste(paste(cols[i], m[i, 1]), paste(cols[i], m[i, 2]), sep = ", ")
}
)
res <- unlist(res)
res <- paste(res, collapse = ", ")
paste0("linear-gradient(to right, ", res, ");")
}
background_pals <- unlist(lapply(X = background_pals, FUN = linear_gradient))
colortext_pals <- rep(c("white", "black"), times = sapply(colors_pal, length))
### !<- colorpicker end ->! ###
processing_info_global <- get_nanopolish_processing_info(polya_table,grouping_factor = NA)
processing_info_global_spread <- processing_info_global %>% tidyr::spread(qc_tag,n)
processing_info_per_sample <- get_nanopolish_processing_info(polya_table,grouping_factor = "sample_name")
processing_info_per_sample_spread <- processing_info_per_sample %>% tidyr::spread(qc_tag,n)
polya_summary_table <- summarize_polya(polya_data = polya_table_passed,summary_factors = "sample_name")
if (!is.na(precomputed_polya_statistics)) {
#assertthat::assert_that(is.list(precomputed_polya_statistics),msg="Please provide the output of calculate_polya_stats() as precomputed_polya_statistics")
#assertthat::assert_that("summary" %in% names(precomputed_polya_statistics),msg="Please provide the output of calculate_polya_stats() as precomputed_polya_statistics")
initial_summary_table = precomputed_polya_statistics %>% dplyr::select(transcript,dplyr::ends_with("_counts"),dplyr::ends_with("_polya_median"),p.value,padj,contains("effect_size"))
initial_table_for_volcano <- precomputed_polya_statistics %>% dplyr::select(transcript,fold_change,padj,significance)
}
else {
if (number_of_samples>1) {
initial_summary_table <- polya_summary_table %>% dplyr::select(transcript,sample_name,polya_median) %>% tidyr::spread(sample_name,polya_median)
}
else {
initial_summary_table <- polya_summary_table
}
initial_table_for_volcano <- initial_summary_table
}
initial_summary_table_counts <- polya_summary_table %>% dplyr::select(transcript,sample_name,counts) %>% tidyr::spread(sample_name,counts)
grouping_factor_levels <- polya_table_passed %>% dplyr::select_if(is.factor) %>% colnames
nanotail_shiny_ui <- shinydashboard::dashboardPage(
skin="blue",
# dashboard header definition -----------------------------------------------------------
shinydashboard::dashboardHeader(
title = paste0("NanoTail - Interactive exploration of polyA lengths estimations ",
"done using Nanopore sequencing - version ",
packageVersion("nanotail")),
titleWidth = 900),
# dashboard sidebar definition -----------------------------------------------------------
shinydashboard::dashboardSidebar(
width = 180,
shinydashboard::sidebarMenu(id="nanotail_menu",
shinydashboard::menuItem("QC info", icon = shiny::icon("info-square"), tabName = "basicInfo"),
shinydashboard::menuItem("Global polyA distribution", icon = shiny::icon("chart-line"), tabName = "global_distr"),
shinydashboard::menuItem("Differential adenylation", icon = shiny::icon("dna"), tabName = "diff_polya"),
shinydashboard::menuItem("Differential expression", icon = shiny::icon("dna"), tabName = "diff_exp"),
shinydashboard::menuItem("Annotation_analysis", icon = shiny::icon("dna"), tabName = "annotation_analysis_inactive"),
shinydashboard::menuItem("Plot settings", icon = shiny::icon("settings"), tabName = "plot_settings"),
shinydashboard::menuItem("About", tabName = "dashboard", icon = shiny::icon("info"))
),
shiny::selectInput("groupingFactor","Group by",choices=grouping_factor_levels),
shiny::uiOutput("condition1UI"),
shiny::uiOutput("condition2UI"),
shiny::checkboxInput("plot_only_selected_conditions","Plot only selected conditions",value=FALSE),
shiny::actionButton("clear_selection",
shiny::HTML("Clear selection"),
icon = shiny::icon("spinner"))
),
# dashboard body definition ---------------------------------------------------------
shinydashboard::dashboardBody(
shinydashboard::tabItems(
shinydashboard::tabItem(tabName = "dashboard",
shiny::fillPage(padding = 0, title = NULL, bootstrap = F, theme = NULL,
shiny::wellPanel(style = "background-color: #ffffff; overflow-y:scroll; max-height: 750px;",
shiny::includeMarkdown(system.file("extdata", "about.md",package = "nanotail")))
)),
shinydashboard::tabItem(tabName = "basicInfo",
shiny::fluidRow(
shinydashboard::infoBoxOutput("numberOfSamples"),
shinydashboard::infoBoxOutput("numberOfTranscripts"),
shinydashboard::infoBoxOutput("numberOfAllReads")),
shiny::fluidRow(
shiny::column(1),
shiny::column(7,shiny::fluidRow(shiny::h4("Per sample plot:")),
shiny::fluidRow(
plotly::plotlyOutput("nanopolish_qc_summary_plot"),
shiny::checkboxInput("show_frequency_plot_nanopolioshQC",label = "Show QC as frequencies?",value = TRUE))),
shiny::column(1),
shiny::column(3,shiny::fluidRow(shiny::h4("Information for all samples:")),
shiny::fluidRow(shinydashboard::infoBoxOutput("numberOfPassReads",width = NULL)),
shiny::fluidRow(shinydashboard::infoBoxOutput("numberOfAdapterReads",width = NULL)),
shiny::fluidRow(shinydashboard::infoBoxOutput("numberOfReadLoadFailedReads",width = NULL)),
shiny::fluidRow(shinydashboard::infoBoxOutput("numberOfNoregionReads",width = NULL))))),
shinydashboard::tabItem(tabName = "plot_settings",
shinydashboard::box(
shinyWidgets::pickerInput(inputId = "col_palette", label = "Colour scale",
choices = colors_pal, selected = "Set1", width = "100%",
choicesOpt = list(
content = sprintf(
"<div style='width:100%%;padding:5px;border-radius:4px;background:%s;color:%s'>%s</div>",
unname(background_pals), colortext_pals, names(background_pals)))),
#shiny::uiOutput("Colorblind"),
shiny::sliderInput("scale_limit_low","Plot scale limit low",0,1024,0,10),
shiny::sliderInput("scale_limit_high","Plot scale limit high",50,1024,200,10),
shiny::sliderInput("counts_scatter_low_value","Counts scatter lower counts limit",0,10024,0,50),
shiny::sliderInput("counts_scatter_high_value","Counts scatter high counts limit",100,10024,1000,50),
shiny::selectInput("center_values_for_distribution_plot",choices=c("none","median","mean","gm_mean","mode"),selected = "median",label = "Show vertical lines with centered values on density plots?")
)),
shinydashboard::tabItem(tabName = "global_distr",
shiny::h2("Global comparison"),
shiny::fluidRow(
#shinydashboard::box(plotly::plotlyOutput('polya_global') %>% shinycssloaders::withSpinner(type = 4),collapsible=TRUE))),
shinydashboard::box(plotly::plotlyOutput('polya_global') %>% shinycssloaders::withSpinner(type = 4),collapsible=TRUE))),
shinydashboard::tabItem(tabName = "diff_polya",
shiny::fluidRow(
shinydashboard::box(DT::dataTableOutput('diff_polya'),collapsible = TRUE,height = '90%'),
shinydashboard::tabBox(title = "per transcript plots",id="tabset1", height="550px",
shiny::tabPanel("boxplot",shiny::plotOutput('polya_boxplot') %>% shinycssloaders::withSpinner(type = 4),
shiny::selectInput("color_by","Color by",choices=grouping_factor_levels),shiny::checkboxInput("add_boxplot",value=FALSE,label = "Add boxplot?")),
shiny::tabPanel("distribution plot",plotly::plotlyOutput('polya_distribution') %>% shinycssloaders::withSpinner(type = 4)
),
shiny::tabPanel("isoforms_boxplot",shiny::plotOutput('isoforms_boxplot') %>% shinycssloaders::withSpinner(type = 4)
),
shiny::tabPanel("virtual gel",shiny::plotOutput('virtual_gel') %>% shinycssloaders::withSpinner(type = 4),
shiny::sliderInput("virtual_gel_bw",min=0.1,max=10,step=0.1,value=1,label = "smoothing bandwitdth to be used ")),
shiny::tabPanel("volcano plot polya",plotly::plotlyOutput('polya_volcano') %>% shinycssloaders::withSpinner(type = 4)))),
shiny::fluidRow(
shinydashboard::box(shiny::actionButton("compute_diff_polya",
shiny::HTML("Compute Polya statistics"),
icon = shiny::icon("spinner")),
shiny::checkboxInput("use_dwell_time_for_statistics",label = "Use dwell time instead of calculated polya_length for statistics",value = FALSE),
shiny::selectInput("polya_stat_test","statistical test to use",choices = c("Wilcoxon","KS"),selected = "Wilcoxon"),
shiny::sliderInput("min_reads_for_polya_stats","Minimal count of reads per transcript",0,100,10,1),collapsible = TRUE,width=12))),
shinydashboard::tabItem(tabName = "diff_exp",
shiny::fluidRow(
shinydashboard::box(DT::dataTableOutput('diff_exp_table')),
shinydashboard::tabBox(title = "plots",id="tabset2",height="500px",
#tabPanel("pca_biplot",plotOutput('pca_biplot') %>% shinycssloaders::withSpinner(type = 4)),
shiny::tabPanel("scatter plot of counts",shiny::actionButton("show_scatter_plot",
shiny::HTML("Show scatter plot"),
icon = shiny::icon("spinner")),plotly::plotlyOutput('counts_plot') %>% shinycssloaders::withSpinner(type = 4)),
shiny::tabPanel("volcano_plot",plotly::plotlyOutput('volcano') %>% shinycssloaders::withSpinner(type = 4)),
shiny::tabPanel("MA_plot",plotly::plotlyOutput('MAplot') %>% shinycssloaders::withSpinner(type = 4)))),
shiny::fluidRow(shiny::actionButton("compute_diff_exp",
shiny::HTML("Compute Differential Expression using Binomial Test"),
icon = shiny::icon("spinner")))),
shinydashboard::tabItem(tabName = "annotation_analysis",
shiny::fluidRow(
shinydashboard::box(DT::dataTableOutput('annotation_table')),
shinydashboard::tabBox(title = "annotation_plots",id="tabset3",height="500px",
#tabPanel("pca_biplot",plotOutput('pca_biplot') %>% shinycssloaders::withSpinner(type = 4)),
shiny::tabPanel("box_plot_of_annotations",plotly::plotlyOutput('annotations_box_plot') %>% shinycssloaders::withSpinner(type = 4),
shiny::uiOutput("select_annotation_factor_levelsUI")
),
shiny::tabPanel("distribution_plot_of_annotations",plotly::plotlyOutput('annotation_distribution_plot') %>% shinycssloaders::withSpinner(type = 4)))),
shiny::fluidRow(shiny::actionButton("get_annotables_annotation",
shiny::HTML("Annotate input dataset"),
icon = shiny::icon("spinner")),
shiny::selectInput("annotables_genome",label="Annotables genome:",choices = c("bdgp6","grch38","grcm38","grch37","rnor6","galgal5","wbcel235","mmul801"),selected="grch38",multiple = FALSE,selectize = TRUE),
shiny::uiOutput("select_annotation_factorUI")
)))))
# call to shiny app and server-siude functions
shiny::shinyApp(ui = nanotail_shiny_ui, server = function(input, output) {
## Reactive values definition ---------------------------------------------------------
values <- reactiveValues()
values$polya_table <- polya_table_passed
values$polya_statistics_summary_table = initial_summary_table
values$polya_table_for_volcano = initial_table_for_volcano
values$diffexp_summary_table = initial_summary_table_counts
values$polya_table_summarized <- polya_summary_table
values$processing_info_global <- processing_info_global
values$processing_info_global_spread <- processing_info_global_spread
values$processing_info_per_sample <- processing_info_per_sample
values$processing_info_per_sample_spread <- processing_info_per_sample_spread
values$diff_exp_grouping_factor = "sample_name"
if(!is.na(precomputed_annotations)) {
values$polya_table_annotables_annotated <- precomputed_annotations
}
else {
values$polya_table_annotables_annotated <- polya_table_passed %>% dplyr::mutate(annotation_group=factor(gsub("^(.).*","\\1",transcript)))
}
# get polyA data for currently selected transcript
data_transcript <- shiny::reactive({
summary_table = values$polya_statistics_summary_table
selected_row <- input$diff_polya_rows_selected
selected_transcript = summary_table[selected_row,]$transcript
data_transcript = subset(polya_table_passed, transcript==selected_transcript)
data_transcript <- data_transcript %>% dplyr::mutate(replicate = gsub(".*(.)$","\\1", sample_name))
data_transcript
})
# get polyA data for currently selected transcript
data_transcript_annot <- shiny::reactive({
summary_table = values$polya_statistics_summary_table
selected_row <- input$diff_polya_rows_selected
selected_transcript = summary_table[selected_row,]$transcript
data_transcript_annot = subset(values$polya_table_annotables_annotated, transcript==selected_transcript)
data_transcript_annot
})
# UI elements rendering ---------------------------------------------------------
## Generate select inputs for conditions used for comparison, based on selected groupingFactor: ---------------------------------------------------------
output$condition1UI <- shiny::renderUI({
group_factor = input$groupingFactor
shiny::selectizeInput("condition1_diff_exp","Condition 1",choices=levels(polya_table_passed[[group_factor]]),selected = levels(polya_table_passed[[group_factor]])[1])
})
output$condition2UI <- shiny::renderUI({
group_factor = input$groupingFactor
shiny::selectizeInput("condition2_diff_exp","Condition 2",choices=levels(polya_table_passed[[group_factor]]),selected = levels(polya_table_passed[[group_factor]])[2])
})
output$Colorblind <- shiny::renderUI({
if(RColorBrewer::brewer.pal.info %>% subset(rownames(.) == input$col_palette) %>% .$colorblind) {
txt_ <- "Yes"
} else {
txt_ <- "No"
}
shiny::HTML(paste("<p>Is colour scale colour blind friendly?<b>", txt_, "</b></p>"))
})
# Output elements rendering section ---------------------------------------------------------
## Use DT::datatable for tables
output$diff_polya = DT::renderDataTable(values$polya_statistics_summary_table %>% dplyr::rename_all(dplyr::funs(stringr::str_replace_all(.,"_"," "))), server = TRUE, selection=list(mode = 'multiple',selected = 1,target = "row"),options = list(dom = 'ftip',scrollX = TRUE))
output$diff_exp_table = DT::renderDataTable(values$diffexp_summary_table %>% dplyr::rename_all(dplyr::funs(stringr::str_replace_all(.,"_"," "))), server = TRUE, selection=list(mode = 'single',selected = 1,target = "row"),options = list(dom = 'ftip',scrollX = TRUE))
output$annotation_table = DT::renderDataTable(data_annotation(), server = TRUE, selection=list(mode = 'multiple',selected = 1,target = "row"),options = list(dom = 'ftip',scrollX = TRUE))
annotation_proxy <- DT::dataTableProxy('annotation_table')
## Show global polyA distribution as density plot
output$polya_global = plotly::renderPlotly({
if (input$plot_only_selected_conditions) {
global_distribution_plot <- plot_polya_distribution(polya_data = polya_table_passed,groupingFactor = input$groupingFactor,scale_x_limit_low = input$scale_limit_low,scale_x_limit_high = input$scale_limit_high,color_palette = input$col_palette,condition1 = input$condition1_diff_exp,condition2 = input$condition2_diff_exp,show_center_values=input$center_values_for_distribution_plot)
}
else {
global_distribution_plot <- plot_polya_distribution(polya_data = polya_table_passed,groupingFactor = input$groupingFactor,scale_x_limit_low = input$scale_limit_low,scale_x_limit_high = input$scale_limit_high,color_palette = input$col_palette,show_center_values=input$center_values_for_distribution_plot)
}
plotly::ggplotly(global_distribution_plot)
})
## Differential adenylation analysis section ---------------------------------------------------------
# Show boxplot of estimated polya lengths for selected transcript
output$polya_boxplot = shiny::renderPlot({
summary_table = values$polya_statistics_summary_table
if (length(input$diff_polya_rows_selected)>0) {
selected_row <- input$diff_polya_rows_selected
if (length(input$diff_polya_rows_selected)>1) {
selected_transcript = paste0(summary_table[selected_row,]$transcript,collapse=",")
}
else {
selected_transcript = summary_table[selected_row,]$transcript
}
cat(selected_transcript)
cat(selected_row)
data_transcript = data_transcript()
if (input$plot_only_selected_conditions) {
polya_boxplot <- plot_polya_violin(polya_data = data_transcript,groupingFactor = input$groupingFactor,scale_y_limit_low = input$scale_limit_low,scale_y_limit_high = input$scale_limit_high,color_palette = input$col_palette,plot_title = selected_transcript,condition1 = input$condition1_diff_exp,condition2 = input$condition2_diff_exp, add_boxplot=input$add_boxplot,fill_by=input$color_by,color_mode='fill')
}
else {
polya_boxplot <- plot_polya_violin(polya_data = data_transcript,groupingFactor = input$groupingFactor,scale_y_limit_low = input$scale_limit_low,scale_y_limit_high = input$scale_limit_high,color_palette = input$col_palette,plot_title = selected_transcript, add_boxplot=input$add_boxplot,fill_by=input$color_by,color_mode='fill')
}
#plotly::ggplotly(polya_boxplot)
print(polya_boxplot)
}
else {
#suppressWarnings(plotly::plotly_empty())
print(no_plot)
}
})
# Show boxplot of estimated polya lengths for selected transcript
output$isoforms_boxplot = shiny::renderPlot({
req(values$polya_table_annotables_annotated)
summary_table = values$polya_statistics_summary_table
if (length(input$diff_polya_rows_selected)>0) {
selected_row <- input$diff_polya_rows_selected
selected_transcript = summary_table[selected_row,]$transcript
data_transcript = data_transcript_annot()
polya_boxplot <- plot_polya_boxplot(polya_data = data_transcript,groupingFactor = "ensembl_transcript_id_short",scale_y_limit_low = input$scale_limit_low,scale_y_limit_high = input$scale_limit_high,color_palette = input$col_palette,plot_title = selected_transcript, violin=FALSE,additional_grouping_factor = input$groupingFactor)
polya_boxplot <- polya_boxplot + ggplot2::theme(axis.text.x = ggplot2::element_text(angle=90))
print(polya_boxplot)
#plotly::ggplotly(polya_boxplot)
}
else {
print("no data")
suppressWarnings(plotly::plotly_empty())
}
})
# Show boxplot of estimated polya lengths for selected transcript
output$virtual_gel = shiny::renderPlot({
summary_table = values$polya_statistics_summary_table
if (length(input$diff_polya_rows_selected)>0) {
selected_row <- input$diff_polya_rows_selected
selected_transcript = summary_table[selected_row,]$transcript
data_transcript = data_transcript()
polya_virtual_gel <- plot_virtual_gel(input_data = data_transcript,groupingFactor = input$groupingFactor,kernel_from =input$scale_limit_low,kernel_to = input$scale_limit_high,valuesColumn="polya_length",density_bw = input$virtual_gel_bw)
polya_virtual_gel$plot <- polya_virtual_gel$plot + ggplot2::theme(axis.text.x = ggplot2::element_text(angle=90))
print(polya_virtual_gel$plot)
#plotly::ggplotly(polya_boxplot)
}
else {
print("no data")
suppressWarnings(plotly::plotly_empty())
}
})
# Show denisty plot of estimated polya lengths for selected transcript
output$polya_distribution = plotly::renderPlotly({
summary_table = values$polya_statistics_summary_table
if (length(input$diff_polya_rows_selected)>0) {
selected_row <- input$diff_polya_rows_selected
if (length(input$diff_polya_rows_selected)>1) {
selected_transcript = paste0(summary_table[selected_row,]$transcript,collapse=",")
}
else {
selected_transcript = summary_table[selected_row,]$transcript
}
data_transcript = data_transcript()
if (input$plot_only_selected_conditions) {
transcript_distribution_plot <- plot_polya_distribution(polya_data = data_transcript,groupingFactor = input$groupingFactor,scale_x_limit_low = input$scale_limit_low,scale_x_limit_high = input$scale_limit_high,color_palette = input$col_palette, plot_title = selected_transcript,condition1 = input$condition1_diff_exp,condition2 = input$condition2_diff_exp,show_center_values=input$center_values_for_distribution_plot)
}
else {
transcript_distribution_plot <- plot_polya_distribution(polya_data = data_transcript,groupingFactor = input$groupingFactor,scale_x_limit_low = input$scale_limit_low,scale_x_limit_high = input$scale_limit_high,color_palette = input$col_palette, plot_title = selected_transcript,show_center_values=input$center_values_for_distribution_plot)
}
plotly::ggplotly(transcript_distribution_plot)
}
else {
suppressWarnings(plotly::plotly_empty())
}
})
# Show volcano plot of differential adenylation analysis
output$polya_volcano <- plotly::renderPlotly({
# show only if differential adenylation analysis was already done
if("fold_change" %in% colnames(values$polya_table_for_volcano)) {
volcanoPlot <- plot_volcano(input_data =values$polya_table_for_volcano,color_palette = input$col_palette,transcript_id_column = "transcript",labels=input$diff_polya_rows_selected)
plotly::ggplotly(volcanoPlot)
}
else {
shiny::showNotification("Launch differential polyadenylation analysis first",closeButton=TRUE,type="warning")
suppressWarnings(plotly::plotly_empty())
}
})
# Calculate differential adenylation if action button was clicked
shiny::observeEvent(input$compute_diff_polya,
{
if(number_of_samples>1) {
shiny::withProgress(message="Computing polya statistics...",
detail = "This step can take a little while",
value = 0.05,min=0,max=1,
{
condition1 = input$condition1_diff_exp
condition2 = input$condition2_diff_exp
shiny::incProgress(1/20)
polya_stats <- calculate_polya_stats(values$polya_table,grouping_factor = input$groupingFactor,condition1 = input$condition1_diff_exp,condition2 = input$condition2_diff_exp,stat_test=input$polya_stat_test,use_dwell_time=input$use_dwell_time_for_statistics,min_reads = input$min_reads_for_polya_stats,add_summary = TRUE,length_summary_to_show = "median")
shiny::incProgress(15/20)
#values$polya_statistics_summary_table <- polya_stats %>% dplyr::select(transcript,!!rlang::sym(paste0(condition1,"_counts")),!!rlang::sym(paste0(condition2,"_counts")),!!rlang::sym(paste0(condition1,"_polya_median")),!!rlang::sym(paste0(condition2,"_polya_median")),p.value,padj)
values$polya_statistics_summary_table <- polya_stats
values$polya_table_for_volcano <- polya_stats %>% dplyr::select(transcript,fold_change,padj,significance)
values$condition1 <- input$condition1_diff_exp
values$condition2 <- input$condition2_diff_exp
shiny::incProgress(3/20)
})
}
else {
shiny::showNotification("Differential analysis is not possible when only single sample is present",closeButton=TRUE,type="error")
}
})
diff_polya_proxy <- DT::dataTableProxy('diff_polya')
# Calculate differential adenylation if action button was clicked
shiny::observeEvent(input$clear_selection,
{
diff_polya_proxy %>% DT::selectRows(NULL)
})
## Differential expression analysis section ---------------------------------------------------------
# Show scatter plot of raw counts
output$counts_plot <- plotly::renderPlotly({
counts_scatter_plot <- plot_counts_scatter(polya_data = values$polya_table_summarized_scatter,groupingFactor = values$groupingFactor_scatter,color_palette = input$col_palette,condition1 = values$condition1_scatter, condition2 = values$condition2_scatter, min_counts = input$counts_scatter_low_value, max_counts = input$counts_scatter_high_value)
plotly::ggplotly(counts_scatter_plot)
})
# Show volcano plot of differential expression analysis
output$volcano <- plotly::renderPlotly({
# show only if differential expression analysis was already done
if("fold_change" %in% colnames(values$diffexp_summary_table)) {
volcanoPlot <- plot_volcano(input_data =values$diffexp_summary_table,color_palette = input$col_palette)
plotly::ggplotly(volcanoPlot)
}
else {
shiny::showNotification("Launch differential expression analysis first",closeButton=TRUE,type="warning")
suppressWarnings(plotly::plotly_empty())
}
})
# Show MA plot of differential expression analysis
output$MAplot <- plotly::renderPlotly({
# show only if differential expression analysis was already done
if("fold_change" %in% colnames(values$diffexp_summary_table)) {
MAplot <- plot_MA(input_data =values$diffexp_summary_table,color_palette = input$col_palette)
plotly::ggplotly(MAplot)
}
else {
shiny::showNotification("Launch differential expression analysis first",closeButton=TRUE,type="warning")
plotly::plotly_empty()
}
})
# Calculate differential expression if action button was clicked
shiny::observeEvent(input$compute_diff_exp,
{
if(number_of_samples>1) {
shiny::withProgress(message="Computing Differential epxression using binomial test...",
detail = "This step can take a little while",
value = 0.05,min=0,max=1,
{
values$polya_table_summarized <- summarize_polya(values$polya_table,summary_factors = c(input$groupingFactor))
shiny::incProgress(4/20)
values$polya_table_summarized_per_sample <- summarize_polya(values$polya_table,summary_factors = "sample_name")
shiny::incProgress(4/20)
values$diffexp_summary_table <- calculate_diff_exp_binom(values$polya_table_summarized,grouping_factor = input$groupingFactor,condition1 = input$condition1_diff_exp,condition2 = input$condition2_diff_exp,summarized_input=TRUE)
shiny::incProgress(10/20)
values$condition1 <- input$condition1_diff_exp
values$condition2 <- input$condition2_diff_exp
values$diff_exp_grouping_factor = input$groupingFactor
})
}
else {
shiny::showNotification("Differential analysis is not possible when only single sample is present",closeButton=TRUE,type="error")
}
})
# show scatter plot of counts if action button was clicked
shiny::observeEvent(input$show_scatter_plot,
{
if(number_of_samples>1) {
shiny::withProgress(message="Computing data for scatter counts plot...",
detail = "This step can take a little while",
value = 0.05,min=0,max=1,
{
values$polya_table_summarized_scatter <- summarize_polya(values$polya_table,summary_factors = c(input$groupingFactor))
shiny::incProgress(18/20)
values$condition1_scatter <- input$condition1_diff_exp
values$condition2_scatter <- input$condition2_diff_exp
values$groupingFactor_scatter = input$groupingFactor
})
}
else {
shiny::showNotification("Scatter plot cannot be shown when only single sample is present",closeButton=TRUE,type="error")
}
})
## Basic QC Info section ---------------------------------------------------------
output$numberOfSamples <- shinydashboard::renderInfoBox({
shinydashboard::infoBox(
"Samples in the analysis: ", number_of_samples, icon = shiny::icon("vials"),
color = "yellow"
)
})
output$numberOfTranscripts <- shinydashboard::renderInfoBox({
shinydashboard::infoBox(
"Transcripts analyzed ", number_of_transcripts, icon = shiny::icon("dna"),
color = "blue"
)
})
output$numberOfAllReads <- shinydashboard::renderInfoBox({
shinydashboard::infoBox(
"Reads processed ", sum(values$processing_info_global$n), icon = shiny::icon("stream"),
color = "black"
)
})
output$numberOfPassReads <- shinydashboard::renderInfoBox({
shinydashboard::infoBox(
"Reads passing filters ", values$processing_info_global_spread$PASS, icon = shiny::icon("check"),
color = "green"
)
})
output$numberOfAdapterReads <- shinydashboard::renderInfoBox({
shinydashboard::infoBox(
"Adapter reads ", values$processing_info_global_spread$ADAPTER, icon = shiny::icon("times"),
color = "red"
)
})
output$numberOfReadLoadFailedReads <- shinydashboard::renderInfoBox({
shinydashboard::infoBox(
"Reads failed to load ", values$processing_info_global_spread$READ_FAILED_LOAD, icon = shiny::icon("times"),
color = "red"
)
})
output$numberOfNoregionReads <- shinydashboard::renderInfoBox({
shinydashboard::infoBox(
"NoRegion reads ", values$processing_info_global_spread$NOREGION, icon = shiny::icon("times"),
color = "red"
)
})
#Show Nanopolish QC summary plot
output$nanopolish_qc_summary_plot <- plotly::renderPlotly({
nanopolish_qc_summary_plot<-plot_nanopolish_qc(values$processing_info_per_sample,color_palette = input$col_palette, frequency=input$show_frequency_plot_nanopolioshQC)
plotly::ggplotly(nanopolish_qc_summary_plot)
})
## Annotation analysis section ---------------------------------------------------------
data_selected_annotation <- shiny::reactive({
#summary_table = values$polya_table_annotated
selected_row <- input$annotation_table_rows_selected
data_annotation = data_annotation()
data_annotation = as.data.frame(data_annotation)
column_name = input$annotation_factor
print(paste0("column name: ",column_name))
selected_annotation = as.character(data_annotation[selected_row,column_name])
print("selected annotation:")
print(selected_annotation)
data_transcript = subset(values$polya_table_annotables_annotated, eval(parse(text=column_name)) %in% c(selected_annotation))
data_transcript <- data_transcript %>% dplyr::group_by(.dots = c("read_id",input$groupingFactor)) %>% dplyr::slice(1)
print("nrow data transc:")
print(nrow(data_transcript))
data_transcript
})
data_annotation <- shiny::reactive({
message(input$annotation_factor)
if (!is.null(input$annotation_factor)) {
data_annotation_summarized <- values$polya_table_annotables_annotated %>% dplyr::group_by(.dots = c(input$annotation_factor,input$groupingFactor)) %>% dplyr::summarize(polya_gm_mean=gm_mean(polya_length)) %>%
tidyr::spread(!!rlang::sym(input$groupingFactor),polya_gm_mean)
data_annotation_summarized
}
else {
shiny::showNotification(ui = "Annotation factor is empty",closeButton = TRUE,type = "message")
}
})
## Generate selectinputs for annotations:
output$select_annotation_factorUI <- shiny::renderUI({
annotation_factor_choices=values$polya_table_annotables_annotated %>% dplyr::select_if(is.factor) %>% dplyr::select(-grouping_factor_levels) %>% colnames
shiny::selectizeInput("annotation_factor","Column used for annotation comparisons",choices=annotation_factor_choices,selected = annotation_factor_choices[1])
})
output$select_annotation_factor_levelsUI <- shiny::renderUI({
annotation_factor_selected = input$annotation_factor
annotation_factor_possible_levels = levels(values$polya_table_annotables_annotated[[annotation_factor_selected]])
shiny::selectInput(inputId = "annotation_factor_levels",label="Which annotations to show on the plot",choices=annotation_factor_possible_levels,selected = annotation_factor_possible_levels,multiple = TRUE)
})
shiny::observeEvent(input$annotation_factor,{
annotation_proxy %>% DT::selectRows(NULL)
})
shiny::observeEvent(input$get_annotables_annotation,
{
shiny::withProgress(message="Annotating input polyA table with annotables...",
detail = "This step can take a little while",
value = 0.05,min=0,max=1,
{
values$polya_table_annotables_annotated <- annotate_with_annotables(values$polya_table,genome = input$annotables_genome)
values$polya_table_annotables_annotated <-data.table::data.table(values$polya_table_annotables_annotated)
shiny::incProgress(18/20)
data.table::setkeyv(values$polya_table_annotables_annotated,c("transcript","biotype","strand","chr"))
shiny::incProgress(1/20)
})
})
# Show denisty plot of estimated polya lengths for selected transcript
output$annotation_distribution_plot = plotly::renderPlotly({
#print(input$annotation_table_rows_selected)
#if (length(input$annotation_table_rows_selected)>0) {
selected_row <- input$annotation_table_rows_selected
print(selected_row)
data_transcript_annotated = data_selected_annotation()
#print(data_transcript_annotated)
if (input$plot_only_selected_conditions) {
transcript_distribution_plot <- plot_polya_distribution(polya_data = data_transcript_annotated,groupingFactor = input$groupingFactor,scale_x_limit_low = input$scale_limit_low,scale_x_limit_high = input$scale_limit_high,color_palette = input$col_palette,condition1 = input$condition1_diff_exp,condition2 = input$condition2_diff_exp,show_center_values=input$center_values_for_distribution_plot)
}
else {
transcript_distribution_plot <- plot_polya_distribution(polya_data = data_transcript_annotated,groupingFactor = input$groupingFactor,scale_x_limit_low = input$scale_limit_low,scale_x_limit_high = input$scale_limit_high,color_palette = input$col_palette,show_center_values=input$center_values_for_distribution_plot)
}
plotly::ggplotly(transcript_distribution_plot)
# }
# else {
# suppressWarnings(plotly::plotly_empty())
# }
})
output$annotations_box_plot = plotly::renderPlotly({
print(paste0("annot_factor: ",input$annotation_factor))
print(input$annotation_factor_levels)
transcript_distribution_plot <- plot_annotations_comparison_boxplot(annotated_polya_data = values$polya_table_annotables_annotated,grouping_factor = input$groupingFactor,annotation_factor = input$annotation_factor, annotation_levels = input$annotation_factor_levels, scale_y_limit_low = input$scale_limit_low,scale_y_limit_high = input$scale_limit_high,color_palette = input$col_palette,violin=FALSE)
plotly::ggplotly(transcript_distribution_plot) %>% plotly::layout(boxmode = "group")
})
})
}
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.