Merge pull request #18 from LUMC/develop

Develop

CHANGELOG.md

@@ -4,8 +4,15 @@ Changelog
 <!--
 Newest changes should be on top.
 -->
+version 1.4.3
+---------------------------
++ Add prior count to CPM calculation
++ Add FDR sort again to markdown files
++ Fix filenames when downloading
++ Add check reload to barcodeplot to update add genes list
++ Change gProfiler correction to gSCS
 
 version 1.4.2
 ---------------------------
 + Remove ordering based on FDR in deTab
-+ Add Changelog file
++ Add Changelog file

DESCRIPTION

@@ -1,7 +1,7 @@
 Package: dgeAnalysis
 Type: Package
 Title: dgeAnalysis
-Version: 1.4.2
+Version: 1.4.3
 Author@R:
   person("Tom", "Kuipers", email = "t.b.kuipers@outlook.com", role = c("aut", "cre"))
 Description:

README.md

@@ -1,6 +1,6 @@
 # dgeAnalysis
 
-This R package contains the R-Shiny application v1.4.2 developed to perform differential gene expression analysis.
+This R package contains the R-Shiny application v1.4.3 developed to perform differential gene expression analysis.
 * dgeAnalysis manual: https://github.com/LUMC/dgeAnalysis/blob/master/MANUAL.pdf
 
 ## Installing

inst/src/markdown/analysisDESeq2.Rmd

@@ -114,7 +114,7 @@ dge <- DGEList(counts = assay(se),
 dge <- dge[rowSums(abs(dge$counts)) > 1,]
 
 tempDge <- dge
-tempDge$counts <- cpm(dge, log = TRUE)
+tempDge$counts <- cpm(dge, log = TRUE, prior.count = 1)
 countDistributionLinePlot(tempDge)
 
 dge <- DGEList(counts = assay(se), samples = colData(se))
@@ -137,7 +137,7 @@ analysis <- DESeqDataSet(se, design = design)
 analysis <- analysis[rowSums(abs(assay(analysis))) > 1,]
 
 # FILTER IF NESSECARY
-counts <- cpm(counts(analysis), log = TRUE)
+counts <- cpm(counts(analysis), log = TRUE, prior.count = 1)
 selectedFeatures <- rownames(analysis)[apply(counts, 1, function(v)
   sum(v >= cpm_value)) >= 1 / 4 * ncol(counts)]
 analysis <- analysis[selectedFeatures, ]
@@ -161,10 +161,8 @@ Extract the normalized data from the analysis results.
 # GET NORMALIZED COUNTS
 
 getSize <- estimateSizeFactors(analysis)
-normCounts <- cpm(data.frame(counts(getSize, normalized = TRUE)))
-normDge <- DGEList(counts = normCounts, samples = dge$samples)
-normDge$counts <- log2(normDge$counts)
-normDge$counts[is.infinite(normDge$counts)] <- 0
+normDge <- DGEList(counts = data.frame(counts(getSize, normalized = TRUE)), samples = dge$samples)
+normDge$counts <- cpm(normDge, log = TRUE, prior.count = 1)
 
 countDistributionLinePlot(normDge)
 samplePca2dPlot(normDge, design_base, "PC1", "PC2")
@@ -217,10 +215,6 @@ if (!is.null(data_annotation)) {
   deTab$Row.names <- NULL
 }
 
-# MAKE BOTH TABLES EQUAL
-deTab <- deTab[intersect(rownames(deTab), rownames(normDge$counts)), ]
-normDge$counts <- normDge$counts[intersect(rownames(deTab), rownames(normDge$counts)), ]
-
 #ORDER deTab TABLE
 deTab <- rename(deTab, "FDR" = "adj.P.Val")
 deOrder <- c("avgLog2CPM", "avgLog2FC", "P.Value", "FDR", "DE")
@@ -264,6 +258,7 @@ Results are saved, so they can be retrieved by the application.
 ```{r save}
 # SAVE ANALYSIS
 
+deTab <- deTab[order(deTab$FDR),]
 save(deTab, normDge, file = "analysis.RData")
 
 ```

inst/src/markdown/analysisEdgeR.Rmd

@@ -122,7 +122,7 @@ row.names(dge$genes) <- row.names(dge$counts)
 dge <- dge[rowSums(abs(dge$counts)) > 1,]
 
 tempDge <- dge
-tempDge$counts <- cpm(dge, log = TRUE)
+tempDge$counts <- cpm(dge, log = TRUE, prior.count = 1)
 countDistributionLinePlot(tempDge)
 
 ```
@@ -134,7 +134,7 @@ The raw data is filtered based on the input values provided.
 # GET SELECTED FEATURES
 
 edger <- calcNormFactors(dge, method = "TMM")
-counts <- cpm(edger, log = TRUE)
+counts <- cpm(edger, log = TRUE, prior.count = 1)
 selectedFeatures <- rownames(edger)[apply(counts, 1, function(v)
   sum(v >= cpm_value)) >= 1 / 4 * ncol(counts)]
 
@@ -157,7 +157,7 @@ The filtered data is normalized using TMM.
 normDge <- calcNormFactors(highExprDge, method = "TMM")
 
 tempDge <- normDge
-tempDge$counts <- cpm(normDge, log = TRUE)
+tempDge$counts <- cpm(normDge, log = TRUE, prior.count = 1)
 countDistributionLinePlot(tempDge)
 samplePca2dPlot(tempDge, design_base, "PC1", "PC2")
 
@@ -255,7 +255,8 @@ Results are saved, so they can be retrieved by the application.
 ```{r save}
 # SAVE ANALYSIS
 
-normDge$counts <- cpm(normDge, log = TRUE)
+deTab <- deTab[order(deTab$FDR),]
+normDge$counts <- cpm(normDge, log = TRUE, prior.count = 1)
 save(deTab, normDge, file = "analysis.RData")
 
 ```

inst/src/markdown/analysisLimma.Rmd

@@ -122,7 +122,7 @@ row.names(dge$genes) <- row.names(dge$counts)
 dge <- dge[rowSums(abs(dge$counts)) > 1,]
 
 tempDge <- dge
-tempDge$counts <- cpm(dge, log = TRUE)
+tempDge$counts <- cpm(dge, log = TRUE, prior.count = 1)
 countDistributionLinePlot(tempDge)
 
 ```
@@ -134,7 +134,7 @@ The raw data is filtered based on the input values provided.
 # GET SELECTED FEATURES
 
 limmaV <- calcNormFactors(dge, method = "TMM")
-counts <- cpm(limmaV, log = TRUE)
+counts <- cpm(limmaV, log = TRUE, prior.count = 1)
 selectedFeatures <- rownames(limmaV)[apply(counts, 1, function(v)
   sum(v >= cpm_value)) >= 1 / 4 * ncol(counts)]
 
@@ -157,7 +157,7 @@ The filtered data is normalized using TMM.
 normDge <- calcNormFactors(highExprDge, method = "TMM")
 
 tempDge <- normDge
-tempDge$counts <- cpm(normDge, log = TRUE)
+tempDge$counts <- cpm(normDge, log = TRUE, prior.count = 1)
 countDistributionLinePlot(tempDge)
 samplePca2dPlot(tempDge, design_base, "PC1", "PC2")
 
@@ -257,7 +257,8 @@ Results are saved, so they can be retrieved by the application.
 ```{r save}
 # SAVE ANALYSIS
 
-normDge$counts <- cpm(normDge, log = TRUE)
+deTab <- deTab[order(deTab$FDR),]
+normDge$counts <- cpm(normDge, log = TRUE, prior.count = 1)
 save(deTab, normDge, file = "analysis.RData")
 
 ```

inst/src/markdown/enrichmentGProfiler2.Rmd

@@ -67,7 +67,7 @@ enrich <- gost(
   ordered_query = FALSE,
   significant = significant,
   user_threshold = alfa,
-  correction_method = "bonferroni",
+  correction_method = "gSCS",
   domain_scope = "annotated",
   sources = database,
   evcodes = TRUE

inst/src/tabs/analysis/svr_analysis.R

@@ -109,6 +109,7 @@ output[["group_analysis_bar"]] <- renderUI({
 ## Add specific gene to barplot
 observe({
   tryCatch({
+    checkReload()
     updateSelectizeInput(
       session = session,
       inputId = 'selected_analysis_bar',

inst/src/tabs/export/svr_export.R

@@ -46,7 +46,7 @@ datasetInput <- reactive({
 ## create filename and save data as CSV
 output$downloadCSV <- downloadHandler(
   filename = function() {
-    paste("shiny_analysis_", gsub(" ", "_", tolower(input$dataset_select)), ".csv", sep = "")
+    paste0("shiny_analysis_", gsub(" ", "_", tolower(input$dataset_select)), ".csv")
   },
   content = function(file) {
     write.table(
@@ -62,7 +62,7 @@ output$downloadCSV <- downloadHandler(
 ## create filename and save data as TSV
 output$downloadTSV <- downloadHandler(
   filename = function() {
-    paste("shiny_analysis_", gsub(" ", "_", tolower(input$dataset_select)), ".tsv", sep = "")
+    paste0("shiny_analysis_", gsub(" ", "_", tolower(input$dataset_select)), ".tsv")
   },
   content = function(file) {
     write.table(
@@ -77,22 +77,19 @@ output$downloadTSV <- downloadHandler(
 
 ## Download markdown HTML report DGE
 output[["downloadDGE_HTML"]] <- downloadHandler(
-  filename = paste(input$analysis_method,
-                   '_',
-                   gsub('-', '', Sys.Date()),
-                   '.html',
-                   sep = ''),
+  filename = function() {
+    paste0(input$analysis_method, '_', gsub('-', '', Sys.Date()),'.html')
+  },
   content <- function(file) {
     file.copy(paste0('markdown/', input$analysis_method, '.html'), file)
   }
 )
 
 ## Download markdown HTML report Enrichment
 output[["downloadENRICH_HTML"]] <- downloadHandler(
-  filename = paste('gprofiler_',
-                   gsub('-', '', Sys.Date()),
-                   '.html',
-                   sep = ''),
+  filename = function() {
+    paste0('gprofiler_', gsub('-', '', Sys.Date()), '.html')
+  },
   content <- function(file) {
     file.copy(paste0('markdown/enrichmentGProfiler2.html'), file)
   }

inst/src/tabs/upload/svr_upload.R

@@ -246,7 +246,7 @@ get_raw_dge <- reactive({
                  samples = colData(se),
                  genes = rowData(se))
   dge <- dge[rowSums(abs(dge$counts)) > 1,]
-  dge$counts <- cpm(dge, log = TRUE)
+  dge$counts <- cpm(dge, log = TRUE, prior.count = 1)
   dge
 })