documentation: agea volumes examples

examples/atlas_genomics_load_allen_gene_expression_atlas.py → examples/atlas_genomics_agea_regions_predictor.py

@@ -1,21 +1,25 @@
+"""
+In this example we train a neural network to predict the cosmos regions from the gene expression
+This demonstrates how to align and sample brain regions relative to the gene expression volumes.
+
+This examples requires `sklearn` and `seaborn` to be installed on top of the `iblatlas` requirements.
+"""
+
+import numpy as np
 import matplotlib.pyplot as plt
-from iblatlas.genomics import agea
 
-df_genes, gene_expression_volumes, atlas_agea = agea.load()
+import seaborn as sns
+from sklearn.neural_network import MLPClassifier
+from sklearn.preprocessing import StandardScaler
+from sklearn.model_selection import train_test_split
+import sklearn.metrics
 
-igenes = (0,)
-fig, axs = plt.subplots(3, 2, sharex=True, sharey=True)
-atlas_agea.plot_cslice(0, ax=axs[0, 0])
-atlas_agea.plot_cslice(0, ax=axs[1, 0], volume='annotation')
-atlas_agea.plot_cslice(0, ax=axs[2, 0], volume=gene_expression_volumes[igenes[0]], cmap='viridis')
-atlas_agea.plot_sslice(0, ax=axs[0, 1])
-atlas_agea.plot_sslice(0, ax=axs[1, 1], volume='annotation')
-atlas_agea.plot_sslice(0, ax=axs[2, 1], volume=gene_expression_volumes[igenes[0]], cmap='viridis')
-fig.tight_layout()
+from iblatlas.genomics import agea
 
+# %% Load the agea atlas
+df_genes, gene_expression_volumes, atlas_agea = agea.load()
 
 # %% remap the the agea atlas at the cosmos level parcellation
-import numpy as np
 ne = gene_expression_volumes.shape[0]
 sel = atlas_agea.label.flatten() != 0  # remove void voxels
 # reshape in a big array nexp x nvoxels this takes a little while
@@ -24,10 +28,6 @@
 aids_cosmos = atlas_agea.regions.remap(aids, 'Allen-lr', 'Cosmos')
 
 # %% now we learn to predict the cosmos labels from the gene expression
-from sklearn.neural_network import MLPClassifier
-from sklearn.preprocessing import StandardScaler
-from sklearn.model_selection import train_test_split
-import sklearn.metrics
 X_train, X_test, y_train, y_test = train_test_split(gexps, aids)
 scaler = StandardScaler()
 scaler.fit(gexps)
@@ -40,7 +40,6 @@
 classes_labels = atlas_agea.regions.id2acronym(clf.classes_)
 
 # %% Plot the confusion matrix
-import seaborn as sns
 cm = sklearn.metrics.confusion_matrix(y_test, clf.predict(X_test), normalize='pred')
 sklearn.metrics.ConfusionMatrixDisplay(cm, display_labels=classes_labels).plot(ax=plt.gca(), cmap='magma')
 

examples/atlas_genomics_load_agea.ipynb

@@ -0,0 +1,88 @@
+{
+ "cells": [
+  {
+   "cell_type": "markdown",
+   "source": [
+    "# Loading the gene expression atlas"
+   ],
+   "metadata": {
+    "collapsed": false
+   },
+   "id": "48e25dbdfb3b180c"
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "initial_id",
+   "metadata": {
+    "collapsed": true
+   },
+   "outputs": [],
+   "source": [
+    "import matplotlib.pyplot as plt\n",
+    "from iblatlas.genomics import agea\n",
+    "\n",
+    "df_genes, gene_expression_volumes, atlas_agea = agea.load()"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "source": [
+    "Displaying a coronal and a sagittal slice.\n",
+    "The top row is the diffusion MRI Allen template, the middle row corresponds to the brain regions annotations and the bottom row corresponds to the first gene expression volume."
+   ],
+   "metadata": {
+    "collapsed": false
+   },
+   "id": "62db803f3b14fc9"
+  },
+  {
+   "cell_type": "code",
+   "outputs": [],
+   "source": [
+    "igenes = (0,)\n",
+    "fig, axs = plt.subplots(3, 2, sharex=True, sharey=True)\n",
+    "atlas_agea.plot_cslice(0, ax=axs[0, 0])\n",
+    "atlas_agea.plot_cslice(0, ax=axs[1, 0], volume='annotation')\n",
+    "atlas_agea.plot_cslice(0, ax=axs[2, 0], volume=gene_expression_volumes[igenes[0]], cmap='viridis')\n",
+    "atlas_agea.plot_sslice(0, ax=axs[0, 1])\n",
+    "atlas_agea.plot_sslice(0, ax=axs[1, 1], volume='annotation')\n",
+    "atlas_agea.plot_sslice(0, ax=axs[2, 1], volume=gene_expression_volumes[igenes[0]], cmap='viridis')\n",
+    "fig.tight_layout()"
+   ],
+   "metadata": {
+    "collapsed": false
+   },
+   "id": "d05447d166c56453"
+  },
+  {
+   "cell_type": "markdown",
+   "source": [],
+   "metadata": {
+    "collapsed": false
+   },
+   "id": "9a7afbf944dd385c"
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "Python 3",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 2
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython2",
+   "version": "2.7.6"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}