Merge pull request #25 from yerkes-gencore/dev

Dev

DESCRIPTION

@@ -1,7 +1,7 @@
 Package: gencoreBulk
 Title: NPRC Genomics Core Bulk RNA Analysis Utilities
-Version: 0.2.0
-Date: 2023-04-14
+Version: 0.2.1
+Date: 2024-03-22
 Author: person("Derrik", "Gratz", email = "[email protected]", role =
     c("aut", "cre"), comment = c(ORCID = "0009-0002-5934-576X"))
 Maintainer: Derrik Gratz <[email protected]>

NAMESPACE

@@ -17,7 +17,6 @@ export(mappingBinsPlot)
 export(parseReadPerGeneFiles)
 export(plotFilterByExpr)
 export(plotGeneExpression)
-export(plotModelCoeffs)
 export(plotPCAFromConfig)
 export(plotVoomByGroup)
 export(runfgsea)

R/getTopNGenes.R

@@ -38,7 +38,7 @@ getTopNGenes <- function(result,
     tibble::rownames_to_column('Gene') %>%
     dplyr::arrange(.data$pvalue)
   if (exclude_ENS){
-    filtered_results <- filtered_results[grepl(pattern = ENS_pattern,
+    filtered_results <- filtered_results[!grepl(pattern = ENS_pattern,
                                                filtered_results$Gene),]
   }
   if (direction == 'mixed') {

R/heatmapFromGenelist.R

@@ -125,6 +125,7 @@ heatmapFromGenelist <- function(geneList,
       NULL
     }),
     column_gap = column_gap,
+    column_labels = column_labels,
     col = circlize::colorRamp2(c(scale_min, 0, scale_max), colors),
     ...
   )

R/plotGeneExpression.R

@@ -80,6 +80,7 @@ plotGeneExpression <- function(gene,
       stop('All groups not found in grouping variable of metadata')
     }
     plotdata$grouping <- factor(plotdata$grouping, levels = groups)
+    plotdata <- plotdata[!is.na(plotdata$grouping),]
   }
   ggplot2::ggplot(plotdata, aes(x=grouping, y=.data[['Gene']])) + 
     (if (boxplot) { ggplot2::geom_boxplot(outlier.color = if (jitter) {NA} else {'black'}) }) +
@@ -107,7 +108,7 @@ plotGeneExpression <- function(gene,
 #'  Default FALSE.
 #'
 #' @returns A ggplot2::geom_segment if data_only = FALSE, else a data.frame
-#' @export
+#' 
 #'
 #' @examples 
 #' \dontrun{
@@ -164,15 +165,17 @@ plotModelCoeffs <- function(gene,
 }
 
 .substitute_terms <- function(formula, value_list) {
-  terms <- unlist(strsplit(as.character(formula), "[+-]"))
+  # terms <- unlist(strsplit(as.character(formula), "[+-]"))
   term_dict <- list()
-  for (term in terms) {
-    term <- trimws(term)
-    if (term %in% names(value_list)) {
-      value <- value_list[[term]]
-      formula <- gsub(term, value, formula, fixed = TRUE)
-      term_dict[[term]] <- as.numeric(value)
+  expr <- formula
+  for (term in names(value_list)) {
+    value <- value_list[[term]]
+    # term_dict[[term]] <- as.numeric(value)
+    pattern <- paste0('\\s+',term,'\\s+')
+    if (grepl(pattern, expr)) {
+      term_dict[term] <- as.numeric(value)
     }
+    expr <- gsub(pattern, paste0(' ',value,' '), expr)
   }
-  return(list(expr = formula, dict = term_dict))
+  return(list(expr = expr, dict = term_dict))
 }

R/runfgsea.R

@@ -8,6 +8,8 @@
 #' @param result A DESeqResults result object
 #' @param breakdown_pathway_names Attempt to split pathway names into source and
 #'  pathway for easier reading
+#' @param na_omit Bool, exclude genes with missing data in results? This checks
+#'  all values for NAs so be mindful of modified `result` objects
 #' @inheritParams fgsea::fgseaSimple
 #'
 #' @inherit fgsea::fgseaSimple return
@@ -25,8 +27,11 @@ runfgsea <- function(result,
                      nperm = 1000,
                      minSize = 10,
                      maxSize = 500,
-                     breakdown_pathway_names = FALSE) {
-  # result <- result %>% na.omit()
+                     breakdown_pathway_names = FALSE,
+                     na_omit = TRUE) {
+  if (na_omit) {
+    result <- result %>% na.omit() 
+  }
   fgsea_data <- result$stat
   names(fgsea_data) <- rownames(result)
   fgsea_data <- fgsea_data[!is.na(fgsea_data)]