Subseting object with reactive values will fail - r

Question

List item
I have a problem with reactive values not working as i think it should work.
The small code below describes the problem. The function firstsub2 will basically subset an object into a smaller one by removing samples we don´t want to keep for some reason (this is using the subset_samples function from the phyloseq biocondcutor package).
UI.R
myui <-
fluidPage(
navbarPage("Project",
## foldchanges
tabPanel("Foldchanges",
titlePanel("Permanova: Analysis of variance using distance matrices"),
# Sidebar layout with input and output definitions ----
sidebarLayout(
# Sidebar panel for inputs ----
sidebarPanel(
actionButton("dofoldchanges", "Generate foldchanges")
),
mainPanel(
# Output: Tabset w/ plot, summary, and table ----
tabsetPanel(id="foldchanges",type = "tabs",
tabPanel(title="Summary", value=1, verbatimTextOutput("summary_foldchanges"))
#tabPanel("Table pairwise",value=4, dataTableOutput("tablepermanovapw"))
)
)
)
)
)
)
SERVER.R
#To install phyloseq
#source("https://bioconductor.org/biocLite.R")
#biocLite("phyloseq")
library(shiny)
library(phyloseq)
myserver <- function(input, output, session) {
source("foldchanges.R", local = TRUE)
}
foldchanges.R
# Filter object
firstsub2 <- reactive({
values$rn <- as.character(sample_data(values$physeq)[,"Description"]$Description)))
cat(values$rn)
#The subset_samples function below will not work
filteredtaxo <- subset_samples(values$physeq, Description %in% values$rn)
return(filteredtaxo)
})
values <- reactiveValues()
observeEvent(input$dofoldchanges, {
rich_sparse_biom = system.file("extdata", "rich_sparse_otu_table.biom", package = "phyloseq")
physeq = import_biom(rich_sparse_biom, parseFunction = parse_taxonomy_greengenes)
print(physeq)
values$physeq <- physeq
values$filtered <- firstsub2()
})
This example will basically return the same object that was imported

Answer 1

Your code still has some errors, especially the server file (Some unnecessary commas).
I think your app works as it should, but it will always be the same file as the values$rn will always be the same. I included another line (uncommented right now) below your subset_sample, to test if the subsetting is working. If you change values$rn to "human skin" for example, you will see different results.
But I didn't figure out how to change it. When taking only the first element of values$rn for example, I'm getting an object not found error. But when I include "human skin" like that, it works.
But maybe that helps you already.
library(shiny)
# source("https://bioconductor.org/biocLite.R")
# biocLite("phyloseq")
library(phyloseq)
myui <- {fluidPage(
navbarPage("Project",
## foldchanges
tabPanel("Foldchanges",
titlePanel("Permanova: Analysis of variance using distance matrices"),
# Sidebar layout with input and output definitions ----
sidebarLayout(
# Sidebar panel for inputs ----
sidebarPanel(
actionButton("dofoldchanges", "Generate foldchanges")
),
mainPanel(
# Output: Tabset w/ plot, summary, and table ----
tabsetPanel(id="foldchanges",type = "tabs",
tabPanel(title="Summary", value=1,
verbatimTextOutput("summary_foldchanges"),
verbatimTextOutput("summary_physeq"))
)))))
)}
myserver <- function(input, output, session) {
values <- reactiveValues()
firstsub2 <- reactive({
req(values$physeq)
input$dofoldchanges
values$rn <- as.character(sample_data(values$physeq)[,"Description"]$Description)
cat(values$rn)
#The subset_samples function below will not work
filteredtaxo <- subset_samples(values$physeq, "Description" %in% values$rn)
## Change it to one of the next lines, to see that subsetting works.
# filteredtaxo <- subset_samples(values$physeq, Description %in% "human skin")
# filteredtaxo <- subset_samples(values$physeq, Description %in% "human gut")
return(filteredtaxo)
})
observeEvent(input$dofoldchanges, {
rich_sparse_biom = system.file("extdata", "rich_sparse_otu_table.biom", package = "phyloseq")
physeq = import_biom(rich_sparse_biom, parseFunction = parse_taxonomy_greengenes)
print(physeq)
values$physeq <- physeq
values$filtered <- firstsub2()
})
output$summary_physeq <- renderPrint({
req(values$physeq)
values$physeq
})
output$summary_foldchanges <- renderPrint({
req(values$filtered)
values$filtered
})
}
shinyApp(myui, myserver)