Merge pull request #15 from theislab/sergio_dev

Adding cell scores functions

Author: sergiomarco25

notebooks/spatialdata_tutorials/.ipynb_checkpoints/4_quantify_exRNA-checkpoint.ipynb

@@ -929,7 +929,6 @@
    "source": [
     "import numpy as np\n",
     "\n",
-    "\n",
     "def compute_projection_score(sdata):\n",
     "    \"\"\"\n",
     "    Compute a segmentation score for each cell based on the expression of genes weighted by their intracellular proportion (1 - extracellular proportion).\n",
@@ -994,7 +993,6 @@
     "    # Store the score in the AnnData object under obs\n",
     "    adata.obs[\"projection_score\"] = score\n",
     "\n",
-    "\n",
     "#   return sdata\n",
     "\n",
     "# Example usage:\n",
@@ -2959,6 +2957,13 @@
     "    adata.obs[\"extracellularly_enriched_proportion\"] = proportion"
    ]
   },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  },
   {
    "cell_type": "code",
    "execution_count": 210,
@@ -3031,7 +3036,7 @@
  ],
  "metadata": {
   "kernelspec": {
-   "display_name": "exrna",
+   "display_name": "Python 3 (ipykernel)",
    "language": "python",
    "name": "python3"
   },
@@ -3045,9 +3050,9 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.10.15"
+   "version": "3.10.16"
   }
  },
  "nbformat": 4,
- "nbformat_minor": 2
+ "nbformat_minor": 4
 }

notebooks/spatialdata_tutorials/.ipynb_checkpoints/5_explore_comunication-checkpoint.ipynb

@@ -23,7 +23,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 1,
    "metadata": {},
    "outputs": [
     {
@@ -64,6 +64,27 @@
     "sdata = sd.read_zarr(xenium_path_cropped)"
    ]
   },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  },
   {
    "cell_type": "markdown",
    "metadata": {},
@@ -146,6 +167,142 @@
     "**PLOTTING FUNCTION 2 (troutpy.pl) Sorted scatterplot/barplot**"
    ]
   },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# Compute contribution source score"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import numpy as np\n",
+    "\n",
+    "def compute_projection_score(sdata):\n",
+    "    \"\"\"\n",
+    "    Compute a segmentation score for each cell based on the expression of genes weighted by their intracellular proportion (1 - extracellular proportion).\n",
+    "\n",
+    "    Parameters\n",
+    "    ----------\n",
+    "    sdata : dict\n",
+    "        A spatialdata object with keys 'table' and 'xrna_metadata'.\n",
+    "        - sdata['table'] is an AnnData object containing expression data in layers['raw']\n",
+    "          and cell metadata in .obs.\n",
+    "        - sdata['xrna_metadata'].var is a DataFrame with gene names as the index and\n",
+    "          an 'extracellular proportion' column.\n",
+    "\n",
+    "    Returns\n",
+    "    -------\n",
+    "    sdata : dict\n",
+    "        The same sdata object with a new column 'segmentation_score' in sdata['table'].obs.\n",
+    "    \"\"\"\n",
+    "    # Retrieve the AnnData object with cells in .obs and genes in .var\n",
+    "    adata = sdata[\"table\"]\n",
+    "\n",
+    "    # Retrieve raw expression data; assume shape (n_cells, n_genes)\n",
+    "    raw_expr = adata.layers[\"raw\"]\n",
+    "\n",
+    "    # If raw_expr is a sparse matrix, convert to a dense array\n",
+    "    if hasattr(raw_expr, \"toarray\"):\n",
+    "        raw_expr = raw_expr.toarray()\n",
+    "\n",
+    "    # Get gene names from the AnnData object\n",
+    "    genes = adata.var_names\n",
+    "\n",
+    "    # Retrieve gene metadata containing the extracellular proportions\n",
+    "    gene_meta = sdata[\"xrna_metadata\"].var\n",
+    "\n",
+    "    # Identify the genes common to both the expression data and the metadata\n",
+    "    common_genes = gene_meta.index.intersection(genes)\n",
+    "    if len(common_genes) == 0:\n",
+    "        raise ValueError(\"No common genes found between adata and gene metadata.\")\n",
+    "\n",
+    "    # Subset the expression matrix to only those common genes\n",
+    "    # Here we assume adata.var_names preserves order; get indices corresponding to common genes.\n",
+    "    common_idx = [i for i, gene in enumerate(genes) if gene in common_genes]\n",
+    "    raw_expr = raw_expr[:, common_idx]\n",
+    "\n",
+    "    # Reorder gene_meta so that it matches the ordering in the expression data.\n",
+    "    # This assumes that the order of genes in adata.var_names is the desired order.\n",
+    "    ordered_genes = [gene for gene in genes if gene in common_genes]\n",
+    "    gene_weights = gene_meta.loc[ordered_genes, \"extracellular_proportion\"]\n",
+    "\n",
+    "    # Convert extracellular proportion to intracellular weight (1 - extracellular proportion)\n",
+    "    intracellular_weights = 1 - gene_weights.values  # numpy array\n",
+    "\n",
+    "    # Compute the numerator and denominator for the weighted average per cell.\n",
+    "    # Numerator: dot product of cell expression with intracellular weights.\n",
+    "    # Denominator: total expression (for the common genes) per cell.\n",
+    "    numerator = raw_expr.dot(intracellular_weights)\n",
+    "    denominator = raw_expr.sum(axis=1)\n",
+    "\n",
+    "    # Avoid division by zero (if a cell has zero expression for these genes)\n",
+    "    score = np.divide(numerator, denominator, out=np.full_like(numerator, np.nan), where=denominator != 0)\n",
+    "\n",
+    "    # Store the score in the AnnData object under obs\n",
+    "    adata.obs[\"projection_score\"] = score\n",
+    "\n",
+    "#   return sdata\n",
+    "\n",
+    "# Example usage:\n",
+    "# sdata = compute_segmentation_score(sdata)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  },
   {
    "cell_type": "markdown",
    "metadata": {},
@@ -714,5 +871,5 @@
   }
  },
  "nbformat": 4,
- "nbformat_minor": 2
+ "nbformat_minor": 4
 }