diff --git a/scripts/tensorize.sh b/scripts/tensorize.sh
index 6491c0766..5f02226f1 100755
--- a/scripts/tensorize.sh
+++ b/scripts/tensorize.sh
@@ -119,25 +119,47 @@ START_TIME=$(date +%s)
 MIN_SAMPLE_ID=$SAMPLE_IDS_START
 MAX_SAMPLE_ID=$SAMPLE_IDS_END
 
-# We want to get the zip folder that was passes to recipes.py - look for the --zip_folder argument and extract the value passed after that
-ZIP_FOLDER=$(echo ${PYTHON_ARGS} | sed 's/--zip_folder \([^ ]*\).*/\1/')
-if [ ! -e $ZIP_FOLDER ]; then
-    echo "ERROR: Zip folder passed was not valid, found $ZIP_FOLDER but expected folder path." 1>&2
-    exit 1
-fi
-
 # create a directory in the /tmp/ folder to store some utilities for use later
 mkdir -p /tmp/ml4h
-# Write out a file with the ids of every sample in the input folder
-echo "Gathering list of input zips to process between $MIN_SAMPLE_ID and $MAX_SAMPLE_ID, this takes several seconds..."
-find $ZIP_FOLDER -name '*.zip' | xargs -I {} basename {} | cut -d '_' -f 1 \
-                               | awk -v min="$MIN_SAMPLE_ID" -v max="$MAX_SAMPLE_ID" '$1 > min && $1 < max' \
-                               | sort | uniq > /tmp/ml4h/sample_ids_trimmed.txt
+
+
+if [[ "$TENSORIZE_MODE" == "tensorize" ]]; then
+  # We want to get the zip folder that was passes to recipes.py - look for the --zip_folder argument and extract the value passed after that
+  ZIP_FOLDER=$(echo ${PYTHON_ARGS} | sed -n 's/.*--zip_folder \([^ ]*\).*/\1/p')
+
+  # Write out a file with the ids of every sample in the input folder
+  if [ -e "$ZIP_FOLDER" ]; then
+    echo "Gathering list of input zips to process between $MIN_SAMPLE_ID and $MAX_SAMPLE_ID, this takes several seconds..."
+    find $ZIP_FOLDER -name '*.zip' | xargs -I {} basename {} | cut -d '_' -f 1 \
+                                   | awk -v min="$MIN_SAMPLE_ID" -v max="$MAX_SAMPLE_ID" '$1 > min && $1 < max' \
+                                   | sort | uniq > /tmp/ml4h/sample_ids_trimmed.txt
+  else
+    echo "ERROR: Invalid zip folder, found zip folder $ZIP_FOLDER"
+    exit 1
+  fi
+
+elif [[ "$TENSORIZE_MODE" == "tensorize_pngs" ]]; then
+  # If tensorizing pngs, we can look for sample_ids as the first column of the manifest file in the --app_csv argument
+  MANIFEST=$(echo ${PYTHON_ARGS} | sed -n 's/.*--app_csv \([^ ]*\).*/\1/p')
+
+  # Write out a file with the ids of every sample in the manifest file
+  if [ -e "$MANIFEST" ]; then
+    echo "Gathering list of sample ids to process between $MIN_SAMPLE_ID and $MAX_SAMPLE_ID, this takes several seconds..."
+    cat $MANIFEST | awk '{print $1}' | grep -v sample \
+                                     | awk -v min="$MIN_SAMPLE_ID" -v max="$MAX_SAMPLE_ID" '$1 > min && $1 < max' \
+                                     | sort | uniq > /tmp/ml4h/sample_ids_trimmed.txt
+  else
+    echo "ERROR: Invalid manifest, found manifest $MANIFEST"
+    exit 1
+  fi
+else
+  echo "ERROR: The tensorize mode $TENSORIZE_MODE is not supported"
+  exit 1
+fi
 
 NUM_SAMPLES_TO_PROCESS=$(cat /tmp/ml4h/sample_ids_trimmed.txt | wc -l)
 echo "Including $NUM_SAMPLES_TO_PROCESS samples in this tensorization job."
 
-
 echo -e "\nLaunching job for sample IDs starting with $MIN_SAMPLE_ID and ending with $MAX_SAMPLE_ID via:"
 
 # we need to run a command using xargs in parallel, and it gets rather complex and messy unless we can just run