diff --git a/R/ComponentScores.R b/R/ComponentScores.R
index 4cff378..6954385 100644
--- a/R/ComponentScores.R
+++ b/R/ComponentScores.R
@@ -30,6 +30,30 @@ ScoreUsingSavedComponent <- function(seuratObj, componentOrName, fieldName, assa
   names(geneWeights) <- savedComponent$feature
 
   ad <- Seurat::GetAssayData(seuratObj, assay = assayName, layer = layer)
+  toDrop <- names(geneWeights)[!names(geneWeights) %in% rownames(ad)]
+
+  # Perform the check first for scale.data, since we can add missing genes:
+  if (length(toDrop) > 0 && layer == 'scale.data') {
+    dataLayer <- Seurat::GetAssayData(seuratObj, assay = assayName, layer = 'data')
+    toScale <- toDrop[toDrop %in% rownames(dataLayer)]
+    if (length(toScale) > 0) {
+      print(paste0('The following ', length(toScale), ' genes are missing from scale.data and will be added now: ', paste0(toScale, collapse = ',')))
+      scaled <- Seurat::ScaleData(
+        dataLayer,
+        features = toScale,
+        do.scale = TRUE,
+        do.center = TRUE
+      )
+
+      if (nrow(scaled) != length(toScale)) {
+        stop('The number of rows after scaling did not match the expected number')
+      }
+
+      ad <- rbind(ad, scaled)
+      seuratObj <- suppressWarnings(Seurat::SetAssayData(seuratObj, assay = assayName, layer = layer, new.data = ad))
+    }
+  }
+
   toDrop <- names(geneWeights)[!names(geneWeights) %in% rownames(ad)]
   if (length(toDrop) > 0) {
     print(paste0('The following ', length(toDrop), ' genes were in the component but not the assay, skipping: ', paste0(toDrop, collapse = ',')))
@@ -47,8 +71,6 @@ ScoreUsingSavedComponent <- function(seuratObj, componentOrName, fieldName, assa
 
   if (length(names(seuratObj@reductions)) > 0) {
     suppressMessages(print(FeaturePlot(seuratObj, features = fieldName, order = T) & ggplot2::scale_colour_gradientn(colours = c("navy", "dodgerblue", "gold", "red"))))
-  } else {
-    print('No reductions present, cannot plot')
   }
 
   graphics::hist(seuratObj@meta.data[[fieldName]], breaks = 300, main = fieldName)
diff --git a/tests/testthat/test-classification.R b/tests/testthat/test-classification.R
index 9b3d6af..64944ea 100644
--- a/tests/testthat/test-classification.R
+++ b/tests/testthat/test-classification.R
@@ -5,7 +5,7 @@ testthat::context("Classification")
 
 source('installSeuratData.R')
 
-SimulateSeuratData <- function() {
+SimulateSeuratDataForTcrPrediction <- function() {
   set.seed(GetSeed())
 
   #union features from all activation component files
@@ -52,7 +52,9 @@ SimulateSeuratData <- function() {
   ))
 
   seuratObj <- Seurat::NormalizeData(seuratObj, verbose = FALSE)
-  seuratObj <- Seurat::ScaleData(seuratObj, features = features, verbose = FALSE)
+
+  # NOTE: only scale a subset of genes, so we can test whether PredictTcrActivation automatically adds others:
+  seuratObj <- Seurat::ScaleData(seuratObj, features = features[1:10], verbose = FALSE)
   return(seuratObj)
 }
 
@@ -141,8 +143,15 @@ test_that("Cell type classification works", {
 
 test_that("Predict TCell Activation works ", {
 
-  SimulateSeuratObj <- SimulateSeuratData()
+  SimulateSeuratObj <- SimulateSeuratDataForTcrPrediction()
+  testthat::expect_equal(2995, nrow(Seurat::GetAssayData(SimulateSeuratObj, assay = 'RNA', layer = 'data')))
+  testthat::expect_equal(10, nrow(Seurat::GetAssayData(SimulateSeuratObj, assay = 'RNA', layer = 'scale.data')))
   SimulateSeuratObj <- PredictTcellActivation(SimulateSeuratObj)
+
+  # This tests whether features were added to scale.data:
+  testthat::expect_equal(2995, nrow(Seurat::GetAssayData(SimulateSeuratObj, assay = 'RNA', layer = 'data')))
+  testthat::expect_equal(2874, nrow(Seurat::GetAssayData(SimulateSeuratObj, assay = 'RNA', layer = 'scale.data')))
+
   SimulateSeuratObj <- PredictTcellActivation(SimulateSeuratObj, modelName = 'CD4')
   SimulateSeuratObj <- PredictTcellActivation(SimulateSeuratObj, modelName = 'CD8')
 
@@ -190,7 +199,7 @@ test_that("Predict TCell Activation works ", {
 
 
 test_that("CombineTcellActivationClasses combines classes and writes versioned columns", {
-  seuratObj <- SimulateSeuratData()
+  seuratObj <- SimulateSeuratDataForTcrPrediction()
   seuratObj <- PredictTcellActivation(seuratObj, combineClasses = FALSE)
 
   classMapping <- list(