inst/shiny/ui/ui_01_upload.R

In compbiomed/BatchQC: Batch Effects Quality Control Software

accepted <- c("text/csv",
    "text/comma-separated-values",
    "text/plain",
    ".csv")


tabPanel("Upload Data",
    useShinyjs(),
    tags$style(appCSS),
    tags$div(
        class = "jumbotron",
        tags$div(
            class = "container",
            fluidRow(column(7, h1("BatchQC"))),
            tags$p("Batch Effects Quality Control Software"),
            uiOutput("tab")

        )
    ),
    # Application title
    titlePanel("Upload Data"),

    # Place for uploading data
    sidebarLayout(
        sidebarPanel(
            h4("Select the type of input you would like to provide:"),
            radioButtons("uploadChoice", "",
                c("Count File and Metadata File" = "countFile",
                    "Summarized Experiment Object" = "seObject",
                    "Example Data" = "example"
                )),
            #Only show panel if uploading count and metadata files
            conditionalPanel(condition = "input.uploadChoice == 'countFile'",
                h4("Upload counts and metadata table"),
                tags$div(tags$p(
                    'Metadata file must be a table with headers and sample names.'
                )),
                fileInput(
                    "counts",
                    "Counts table",
                    multiple = FALSE,
                    accept = accepted
                ),
                fileInput("md", "Metadata",
                    multiple = FALSE,
                    accept = accepted)),
            conditionalPanel(condition = "input.uploadChoice == 'seObject'",
                fileInput(
                    "se",
                    "Summarized Experiment",
                    multiple = FALSE,
                    accept = c(accepted, ".RDS")
                )),
            conditionalPanel(condition = "input.uploadChoice == 'example'",
                selectInput("exampleData",
                    "Example Data",
                    choices = c("", "proteinData", "signatureData",
                        "bladderData", "TBData", "No Selection"),
                    selected = ""), ),
            withBusyIndicatorUI(actionButton(inputId = 'submit',
                label = 'Upload'))
        ),

        # Show a table of the inputted data
        mainPanel(
            tabsetPanel(
                tabPanel('Input Summary',
                    h4(strong("Usage")),
                    h5("The table displayed here is a preview of the data selected to be uploaded to BatchQC. Please take a moment to verify that this is the intended data to upload, then select the Upload button to complete uploading the data to BatchQC."),
                    conditionalPanel(condition = "input.uploadChoice == 'example'",
                        h4(strong("Example Data Description")),
                        conditionalPanel(condition = "input.exampleData == 'proteinData'",
                            h5("This data set is mass-spectromery data for protein expression in an unpublished data set.")
                        ),
                        conditionalPanel(condition = "input.exampleData == 'signatureData'",
                            h5("This data set is RNA-seq data from signature data captured when activating different growth pathway genes in human mammary epithelial cells (GEO accession: GSE73628)."),
                            h6("Reference: Rahman M, et al. Activity of distinct growth factor receptor network components in breast tumors uncovers two biologically relevant subtypes. Genome Med. 2017 Apr 26;9(1):40. doi: 10.1186/s13073-017-0429-x.")
                        ),
                        conditionalPanel(condition = "input.exampleData == 'bladderData'",
                            h5("This data set is from bladder cancer data and contains Affymetrix gene expression microarray data. This data set is from the
                                    bladderbatch package which must be installed to use this data example set"),
                            h6("Reference: Leek J., et al. Tackling the widespread and critical impact of batch effects in high-throughput data. Nat Rev Genet 11, 733–739 (2010). https://doi.org/10.1038/nrg2825")
                        ),
                        conditionalPanel(condition = "input.exampleData == 'TBData'",
                            h5("This data set is from a TB cohort study from Pondicherry India and contains RNA-seq counts data."),
                            h6("Reference: VanValkenburg A, et al. Malnutrition leads to increased inflammation and expression of tuberculosis risk signatures in recently exposed household contacts of pulmonary tuberculosis. Front Immunol. 2022 Sep 28;13:1011166. doi: 10.3389/fimmu.2022.1011166")
                        )),
                    h4(strong("Data Preview")),
                    DTOutput('counts_header'),
                    textOutput('counts_dimensions'),
                    DTOutput('se_counts'),
                    textOutput('se_dimensions'),
                    br()
                ),
                tabPanel('Full Metadata',
                    h4(strong("Usage")),
                    h5("The table displayed here shows all metadata associated with the uploaded data. Please refer to this table to view sample condition information."),
                    h4(strong("Metadata Table")),
                    DTOutput('metadata_header'),
                    DTOutput('se_meta'),
                    br()
                ),
                tabPanel('Normalization',
                    h4(strong("Usage")),
                    h5("CPM calculates the counts mapped to a feature relative to the total counts mapped to a sample times one million. CPM may be used to adjust expression count biases introduced by sequencing depth. CPM adjusted values are not recommended for differential expression analysis or within sample comparison."),
                    h5("DESeq calculates the counts mapped to a feature divided by sample-specific size factors. Size factors are determined by the median ratio of gene counts relative to the geometric mean per feature. DESeq may be used to adjust expression count biases introduced by sequencing depth and RNA composition. DESeq adjusted values are not recommended for within sample comparison."),
                    h5("edgeR calculates scale factors using a trimmed mean of M-values between each pair of samples and multiplies the scale factors with the original library size to get the normalized library size. This normalization method should be selected when you plan to use edgeR for differential expression analysis. The original read (raw) counts should be the provided input for this normalization method."),
                    selectizeInput('normalization_method',
                        'Choose normalization method',
                        multiple = FALSE,
                        choices = c('CPM', 'DESeq', 'edgeR', 'none'),
                        selected = NULL,
                        options = list(placeholder =
                                'Please select an option below',
                            onInitialize = I(
                                'function() { this.setValue(""); }'
                            ))),
                    selectizeInput('normalization_assay',
                        'Choose the assay on which to do normalization',
                        multiple = FALSE,
                        choices = c(''),
                        selected = NULL),
                    textInput(inputId = 'normalized_assay_name', 'Name for the normalized assay',
                        value = ''),
                    checkboxInput('log', 'log(x+1) transform the results'),
                    withBusyIndicatorUI(actionButton(inputId = 'normalize',
                        label = 'Normalize')),
                    br()
                ),
                tabPanel('Negative Binomial Check',
                    h4(strong("Usage")),
                    h5("This features allows you to check your data to see if it conforms to the required negative binomial assumption needed for various downstream analysis. If The negative binomial assumption is not met, you should normalize your data or perform other preprocessing step and/or use other more appropraite analysis tools."),
                    selectizeInput('nb_test',
                        'Choose the test to perform',
                        multiple = FALSE,
                        choices = c('nb_DESeq2'),
                        selected = NULL,
                        options = list(placeholder =
                                'Please select an option below',
                            onInitialize = I(
                                'function() { this.setValue(""); }'
                            ))),
                    selectizeInput('counts_matrix',
                        'Choose the array data to check',
                        multiple = FALSE,
                        choices = c(''),
                        selected = NULL,
                        options = list(placeholder =
                                'Please select an option below',
                            onInitialize = I(
                                'function() { this.setValue(""); }'
                            ))),
                    selectizeInput('condition_of_interest',
                        'Select the variable you are interested in analyzing',
                        multiple = FALSE,
                        choices = c(''),
                        selected = NULL, options = list(placeholder =
                                'Please select an option below',
                            onInitialize = I(
                                'function() { this.setValue(""); }'
                            ))),
                    selectizeInput('nb_variables',
                        'Select other variables you would like to include in your analysis',
                        multiple = TRUE,
                        choices = c(''),
                        selected = NULL,
                        options = list(placeholder =
                                'Please select an option below',
                            onInitialize = I(
                                'function() { this.setValue(""); }'
                            ))),
                    checkboxInput('nb_advanced_options', 'Advanced Options',
                        value = FALSE),
                    conditionalPanel(condition = "input.nb_advanced_options == 1",
                        numericInput('num_genes',
                            'Number of genes to analyze (downsampling)',
                            value = 500,
                            min = 2)),
                    withBusyIndicatorUI(actionButton(inputId = 'nb_check',
                        label = 'Check Distribution')),
                    textOutput('recommendation'),
                    plotOutput('nb_histogram'),
                    textOutput('reference'),
                    br()
                ),
                tabPanel('Lambda Statistic',
                    h4(strong("Usage")),
                    h5("An experiment with an unbalanced design will always be
                    improved with a batch correction applies. However, when an
                    experiment has a balanced design, anadjusted lambda
                    statistic greater than -2 indicates a need for batch
                    correction."),
                    selectizeInput('lambda_matrix',
                        'Choose the array data to compute lambda',
                        multiple = FALSE,
                        choices = c(''),
                        selected = NULL,
                        options = list(placeholder =
                                'Please select an option below',
                            onInitialize = I(
                                'function() { this.setValue(""); }'
                            ))),
                    selectizeInput('batch_ind',
                        'Select the batch variable',
                        multiple = FALSE,
                        choices = c(''),
                        selected = NULL, options = list(placeholder =
                                'Please select an option below',
                            onInitialize = I(
                                'function() { this.setValue(""); }'
                            ))),
                    selectizeInput('group_ind',
                        'Select the experimental variable',
                        multiple = FALSE,
                        choices = c(''),
                        selected = NULL, options = list(placeholder =
                                'Please select an option below',
                            onInitialize = I(
                                'function() { this.setValue(""); }'
                            ))),
                    withBusyIndicatorUI(actionButton(inputId = 'lambda_stat',
                        label = 'Calculate Lambda')),
                    DTOutput('lambda_table')
                    ),
                tabPanel('Batch Effect Correction',
                    h4(strong("Usage")),
                    h5("ComBat-Seq uses a negative binomial regression to model batch effects. It requires untransformed, raw count data to adjust for batch effect. Please use this option with a counts assay"),
                    h5("ComBat corrects for Batch effect using a parametric empirical Bayes framework and data should be cleaned and normalized. Therefore, please select a normalized assay to run this on."),
                    h5("limma batch correction fits a linear model to the data, including batch and regular treatments, then removes the component due to the batch effects. Please run limma on log expression data."),
                    selectizeInput('correction_method', 'Choose correction method',
                        multiple = FALSE,
                        choices = c('ComBat-Seq', 'ComBat', 'limma', 'sva'),
                        selected = NULL,
                        options = list(placeholder =
                                'Please select an option below',
                            onInitialize = I(
                                'function() { this.setValue(""); }'
                            ))),
                    selectizeInput('correction_assay',
                        'Choose the assay on which to do correction',
                        multiple = FALSE,
                        choices = c(''),
                        selected = NULL,
                        options = list(placeholder =
                                'Please select an option below',
                            onInitialize = I(
                                'function() { this.setValue(""); }')
                        )),
                    selectizeInput('correction_batch',
                        'Select the variable that represents batch (or the experimental variable for sva)',
                        multiple = FALSE,
                        choices = c(''),
                        selected = NULL,
                        options = list(placeholder =
                                'Please select an option below',
                            onInitialize = I(
                                'function() { this.setValue(""); }')
                        )),
                    selectizeInput('correction_covariates',
                        'Choose the covariates you would like to preserve (or for sva, include as adjustment variables)',
                        multiple = TRUE,
                        choices = c(''),
                        selected = NULL,
                        options = list(placeholder =
                                'Please select an option below',
                            onInitialize = I(
                                'function() { this.setValue(""); }')
                        )),
                    textInput(inputId = 'corrected_assay_name',
                        'Name for the corrected assay'),
                    actionButton(inputId = 'correct', label = 'Correct')
                )

            )
        )
    )
)