Merge pull request #47 from volkamerlab/co-author-review

Follow-up manuscript analyses
volkamerlab · Dec 14, 2021 · cf8f1b4 · cf8f1b4
2 parents 3012da2 + 5aec66f
commit cf8f1b4
Show file tree

Hide file tree

Showing 20 changed files with 1,765 additions and 228 deletions.
diff --git a/notebooks/005_comparison/008_alternative_comparison.ipynb b/notebooks/005_comparison/008_alternative_comparison.ipynb
@@ -83,7 +83,17 @@
    "source": [
     "def load_fingerprints(path_fingerprints):\n",
     "    \"\"\"\n",
-    "    TODO\n",
+    "    Load fingerprints from file.\n",
+    "\n",
+    "    Parameters\n",
+    "    ----------\n",
+    "    path_fingerprints : pathlib.Path or str\n",
+    "        Path to fingerprints file.\n",
+    "\n",
+    "    Returns\n",
+    "    -------\n",
+    "    fingerprints_df : pandas.DataFrame\n",
+    "        Fingerprints (index: structures, columns: fingerprint bits)\n",
     "    \"\"\"\n",
     "\n",
     "    kissim_fingerprints = FingerprintGenerator.from_json(path_fingerprints)\n",
@@ -112,7 +122,19 @@
    "source": [
     "def normalize_fingerprints(fingerprints_df, _type):\n",
     "    \"\"\"\n",
-    "    TODO\n",
+    "    Normalize fingerprints with different normalization measures.\n",
+    "\n",
+    "    Parameters\n",
+    "    ----------\n",
+    "    fingerprints_df : pandas.DataFrame\n",
+    "        Fingerprints (index: structures, columns: fingerprint bits)\n",
+    "    _type : str\n",
+    "        Normalization type: min-max-norm or zscore\n",
+    "\n",
+    "    Returns\n",
+    "    -------\n",
+    "    pandas.DataFrame\n",
+    "        Normalized fingerprints (index: structures, columns: fingerprint bits)\n",
     "    \"\"\"\n",
     "\n",
     "    if _type == \"min-max-norm\":\n",
@@ -138,7 +160,23 @@
    "source": [
     "def weight_fingerprints(fingerprints_df, w1, w2, w3):\n",
     "    \"\"\"\n",
-    "    TODO\n",
+    "    Weight fingerprints!\n",
+    "\n",
+    "    Parameters\n",
+    "    ----------\n",
+    "    fingerprints_df : pandas.DataFrame\n",
+    "        Fingerprints (index: structures, columns: fingerprint bits)\n",
+    "    w1 : float\n",
+    "        Weight for the 8 physicochemical features\n",
+    "    w2 : float\n",
+    "        Weight for the 4 spatial distance features\n",
+    "    w3 : float\n",
+    "        Weight for the 3 spatial moment features\n",
+    "\n",
+    "    Returns\n",
+    "    -------\n",
+    "    pandas.DataFrame\n",
+    "        Weighted fingerprints (index: structures, columns: fingerprint bits)\n",
     "    \"\"\"\n",
     "\n",
     "    n1 = 85 * 8\n",
@@ -165,7 +203,17 @@
    "source": [
     "def generate_structure_distance_matrix(fingerprints_df):\n",
     "    \"\"\"\n",
-    "    TODO\n",
+    "    Generate the structure distance matrix based on fingerprints.\n",
+    "\n",
+    "    Parameters\n",
+    "    ----------\n",
+    "    fingerprints_df : pandas.DataFrame\n",
+    "        Fingerprints (index: structures, columns: fingerprint bits)\n",
+    "\n",
+    "    Returns\n",
+    "    -------\n",
+    "    pandas.DataFrame\n",
+    "        Structure distance matrix\n",
     "    \"\"\"\n",
     "\n",
     "    structure_distance_matrix_array = pairwise.nan_euclidean_distances(fingerprints_df.values)\n",
@@ -189,7 +237,18 @@
    "source": [
     "def generate_kinase_distance_matrix(structure_distance_matrix_df):\n",
     "    \"\"\"\n",
-    "    TODO\n",
+    "    Generate kinase distance matrix.\n",
+    "    Based on minimum structure pair distance per kinase pair.\n",
+    "\n",
+    "    Parameters\n",
+    "    ----------\n",
+    "    structure_distance_matrix_df : pandas.DataFrame\n",
+    "        Structure distance matrix\n",
+    "\n",
+    "    Returns\n",
+    "    -------\n",
+    "    pandas.DataFrame\n",
+    "        Kinase distance matrix\n",
     "    \"\"\"\n",
     "\n",
     "    # Copy distance matrix to kinase matrix\n",
@@ -219,7 +278,12 @@
    "source": [
     "def generate_tree(kinase_distance_matrix_df):\n",
     "    \"\"\"\n",
-    "    TODO\n",
+    "    Generate and visualize tree from kinase distance matrix.\n",
+    "\n",
+    "    Parameters\n",
+    "    ----------\n",
+    "    kinase_distance_matrix_df : pandas.DataFrame\n",
+    "        Kinase distance matrix\n",
     "    \"\"\"\n",
     "\n",
     "    tree.from_distance_matrix(kinase_distance_matrix_df, \"kissim.tree\", \"kinase_annotation.csv\")\n",
@@ -259,7 +323,26 @@
    "source": [
     "def pipeline(kissim_fingerprints_df, norm_type, weights):\n",
     "    \"\"\"\n",
-    "    TODO\n",
+    "    Full pipeline:\n",
+    "    - Load fingerprints\n",
+    "    - Normalize fingerprints\n",
+    "    - Weight fingerprints\n",
+    "    - Generate structure distance matrix\n",
+    "    - Generate kinase distance matrix\n",
+    "    - Generate and visualize tree\n",
+    "\n",
+    "    Parameters\n",
+    "    ----------\n",
+    "    kissim_fingerprints_df : pandas.DataFrame\n",
+    "        Fingerprints (index: structures, columns: fingerprint bits)\n",
+    "    norm_type : str\n",
+    "        Normalization type: min-max-norm or zscore\n",
+    "    weights : list of float\n",
+    "        [\n",
+    "            Weight for the 8 physicochemical features,\n",
+    "            Weight for the 4 spatial distance features,\n",
+    "            Weight for the 3 spatial moment features,\n",
+    "        ]\n",
     "    \"\"\"\n",
     "\n",
     "    print(\"Fingerprints \", kissim_fingerprints_df.shape)\n",