diff --git a/notebooks/goea_nbt3102.ipynb b/notebooks/goea_nbt3102.ipynb
index a8f4f31..6f7c956 100755
--- a/notebooks/goea_nbt3102.ipynb
+++ b/notebooks/goea_nbt3102.ipynb
@@ -11,7 +11,7 @@
     "subpopulations of cells\n",
     "](http://www.nature.com/nbt/journal/v33/n2/full/nbt.3102.html#methods)\n",
     "\n",
-    "Note: you must have the Python package, **xlrd**, installed to run this example. \n",
+    "Note: you must have the Python packages, **xlrd**, **pandas** and **openpyxl**, installed to run this example. \n",
     "\n",
     "Note: To create plots, you must have:\n",
     "  * Python packages: **pyparsing**, **pydot**\n",
@@ -160,9 +160,7 @@
    "metadata": {},
    "source": [
     "### 2c. Load Background gene set\n",
-    "In this example, the background is all mouse protein-codinge genes.    \n",
-    "\n",
-    "Follow the instructions in the `background_genes_ncbi` notebook to download a set of background population genes from NCBI."
+    "In this example, the background is all mouse protein-codinge genes.",
    ]
   },
   {
@@ -179,7 +177,7 @@
     }
    ],
    "source": [
-    "from genes_ncbi_10090_proteincoding import GENEID2NT as GeneID2nt_mus\n",
+    "from goatools.test_data.genes_NCBI_10090_ProteinCoding import GENEID2NT as GeneID2nt_mus\n",
     "print(len(GeneID2nt_mus))"
    ]
   },
@@ -251,21 +249,22 @@
     }
    ],
    "source": [
-    "# Data will be stored in this variable\n",
     "import os\n",
-    "geneid2symbol = {}\n",
+    "import goatools\n",
     "# Get xlsx filename where data is stored\n",
-    "ROOT = os.path.dirname(os.getcwd()) # go up 1 level from current working directory\n",
-    "din_xlsx = os.path.join(ROOT, \"goatools/test_data/nbt_3102/nbt.3102-S4_GeneIDs.xlsx\")\n",
+    "ROOT = os.path.dirname(goatools.__file__) # goatool root directory\n",
+    "din_xlsx = os.path.join(ROOT, \"test_data\", \"nbt_3102\", \"nbt.3102-S4_GeneIDs.xlsx\")\n",
     "# Read data\n",
     "if os.path.isfile(din_xlsx):  \n",
-    "    import xlrd\n",
-    "    book = xlrd.open_workbook(din_xlsx)\n",
-    "    pg = book.sheet_by_index(0)\n",
-    "    for r in range(pg.nrows):\n",
-    "        symbol, geneid, pval = [pg.cell_value(r, c) for c in range(pg.ncols)]\n",
-    "        if geneid:\n",
-    "            geneid2symbol[int(geneid)] = symbol\n",
+    "    import pandas as pd\n",
+    "    df = pd.read_excel(\n",
+    "        din_xlsx,\n",
+    "        header=None, \n",
+    "        names=[\"symbol\", \"geneid\", \"pval\"], \n",
+    "        dtype={\"symbol\": str, \"geneid\": int, \"pval\": float},\n",
+    "        index_col=1,\n",
+    "    )  # requires openpyxl\n",
+    "    geneid2symbol = df[\"symbol\"].to_dict()\n",
     "    print('{N} genes READ: {XLSX}'.format(N=len(geneid2symbol), XLSX=din_xlsx))\n",
     "else:\n",
     "    raise RuntimeError('FILE NOT FOUND: {XLSX}'.format(XLSX=din_xlsx))"