fix notebook header levels

Author: tony-kuo

docs/notebooks/1. Querying Attributions.ipynb

@@ -45,7 +45,9 @@
    "id": "f6d70e5b-6e56-45f7-b164-6a23ff565491",
    "metadata": {},
    "source": [
-    "# Lung Example"
+    "# Lung Example\n",
+    "\n",
+    "Compare type I pneumocyte marker genes in that cell type and compare to ionocytes"
    ]
   },
   {
@@ -235,6 +237,24 @@
     "ax.set_xlabel(\"Cell Type\")"
    ]
   },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "eb4b81e5-9202-4f1d-ac99-78cf56ac7a3a",
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  },
+  {
+   "cell_type": "markdown",
+   "id": "eb4479b2-6a54-46ab-b878-de76bd1256c8",
+   "metadata": {},
+   "source": [
+    "# MS1 Example\n",
+    "\n",
+    "Replicate analysis in paper by calculating mean attributions for MS1 gene set and comparing across disease"
+   ]
+  },
   {
    "cell_type": "markdown",
    "id": "22506b56-3e99-4302-9819-a8a03101d8ad",
@@ -356,7 +376,7 @@
    "id": "cbc6bf4d-e73c-432f-8f77-a0d47239ee48",
    "metadata": {},
    "source": [
-    "### Calculate SIGnature Scores (Mean Attribution Scores) for each Cell Type\n",
+    "### Calculate Mean Attribution Scores for each Cell Type\n",
     "\n",
     "Iterate through tiledb of each cell type, calculate score, concatenate the metadata"
    ]
@@ -656,7 +676,7 @@
    "id": "38abb391-5efa-47a1-aa29-47fd56a2f896",
    "metadata": {},
    "source": [
-    "## Analyze Scores\n",
+    "### Analyze Scores\n",
     "\n",
     "Load metadata into \"Meta\" object to perform gene signature analysis"
    ]
@@ -666,7 +686,7 @@
    "id": "ae7c8666-5596-4661-86dd-34b450f9bc73",
    "metadata": {},
    "source": [
-    "### Load and Filter Metadata"
+    "#### Load and Filter Metadata"
    ]
   },
   {
@@ -735,7 +755,7 @@
    "id": "7ae97d49-64b9-4604-afcb-6e92c38e77c4",
    "metadata": {},
    "source": [
-    "### Define \"Hits\" Based on Cutoff\n",
+    "#### Define \"Hits\" Based on Cutoff\n",
     "Show multiple options for how to define hits and add to metadata (e.g. 90th percentile, 0.95 quantile, 2 stdev above mean)"
    ]
   },
@@ -783,7 +803,7 @@
    "id": "4f227a83-d598-4052-acbe-e0b00566133c",
    "metadata": {},
    "source": [
-    "### Calculate hit percentage per sample\n",
+    "#### Calculate hit percentage per sample\n",
     "For all the monocytes/macrophages in each sample, calculate the percentage that are hits\n",
     "\n",
     "Only consider samples with at least 25 monocytes/macrophages and diseases that have at least 3 samples\n",
@@ -808,7 +828,7 @@
    "id": "ee7b6eb8-b186-4a72-9f07-2cbdf1ee4cb1",
    "metadata": {},
    "source": [
-    "### Generate sample boxplot (like figure 4a)"
+    "#### Generate sample boxplot (like figure 4a)"
    ]
   },
   {
@@ -867,7 +887,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.10.18"
+   "version": "3.10.14"
   }
  },
  "nbformat": 4,

docs/notebooks/2. Generating Attributions.ipynb

@@ -5,7 +5,7 @@
    "id": "340b550f-303d-4170-9324-99f74ee2a980",
    "metadata": {},
    "source": [
-    "# Tutorial 2. Generating Attributions\n",
+    "# Tutorial 2: Generating Attributions\n",
     "\n",
     "This first part of this tutorial shows how to generate new attribution scores using integrated gradients and SCimilarity. \n",
     "\n",
@@ -44,15 +44,15 @@
    "id": "d75bd0ec-0c5c-49a2-9071-a48d41f0deac",
    "metadata": {},
    "source": [
-    "## Part 1: Generate Attributions for New Data"
+    "# Part 1: Generate Attributions for New Data"
    ]
   },
   {
    "cell_type": "markdown",
    "id": "df81f956-d8a9-4ca3-9786-c44f8f89c896",
    "metadata": {},
    "source": [
-    "#### Load SCimilarity Model and Attributions TileDB"
+    "### Load SCimilarity Model and Attributions TileDB"
    ]
   },
   {
@@ -75,7 +75,7 @@
    "id": "0b626c3a-eb71-42e4-864a-9127467abf30",
    "metadata": {},
    "source": [
-    "#### Load dataset\n",
+    "### Load dataset\n",
     "\n",
     "Load your custom dataset. This tutorial shows randomly generated cells. 5000 are entirely random and 500 selectively upregulate MS1.\n",
     "\n",
@@ -234,7 +234,7 @@
    "id": "89287600-b5ce-4163-ba76-dd97dc26c2c0",
    "metadata": {},
    "source": [
-    "#### Align Dataset to SCimilarity Genes\n",
+    "### Align Dataset to SCimilarity Genes\n",
     "\n",
     "Attributions collected against SCimilarity model need to be aligned to SCimilarity gene order and then log-normalized"
    ]
@@ -264,7 +264,7 @@
    "id": "37f5d7a8-9aa0-4929-a53f-997c417d3818",
    "metadata": {},
    "source": [
-    "#### Generate attributions\n",
+    "### Generate attributions\n",
     "Buffer size controls how many cells processed at once. Can adjust to fit on GPU"
    ]
   },
@@ -373,15 +373,15 @@
    "id": "f79704ee-95b7-4491-a498-f43ff6b7ba2e",
    "metadata": {},
    "source": [
-    "## Part 2: Integrate new data with pre-computed attributions"
+    "# Part 2: Integrate new data with pre-computed attributions"
    ]
   },
   {
    "cell_type": "markdown",
    "id": "7d29af9b-0b69-4dab-9be6-4e459568837f",
    "metadata": {},
    "source": [
-    "#### Create a tiledb for integrating with known matrices"
+    "### Create a tiledb for integrating with known matrices"
    ]
   },
   {
@@ -431,7 +431,7 @@
    "id": "cb6e801e-511f-4ce1-8511-542078002777",
    "metadata": {},
    "source": [
-    "#### Search existing tiledbs and new tiledb"
+    "### Search existing tiledbs and new tiledb"
    ]
   },
   {
@@ -766,7 +766,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.10.18"
+   "version": "3.10.14"
   }
  },
  "nbformat": 4,