export data

fteufel · Aug 19, 2022 · 4caa27e · 4caa27e
1 parent ea062c3
commit 4caa27e
Show file tree

Hide file tree

Showing 3 changed files with 66,770 additions and 3 deletions.
diff --git a/benchmark.ipynb b/benchmark.ipynb
@@ -124,6 +124,24 @@
     "df.to_pickle('tempsave_raw_df.pkl')"
    ]
   },
+  {
+   "cell_type": "code",
+   "execution_count": 26,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "df = pd.read_pickle('tempsave_raw_df.pkl')"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 29,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "df.sort_index().to_csv('results/pdb_benchmark_raw.csv')"
+   ]
+  },
   {
    "cell_type": "code",
    "execution_count": 10,
@@ -158,7 +176,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 11,
+   "execution_count": 39,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -167,7 +185,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 12,
+   "execution_count": 40,
    "metadata": {},
    "outputs": [
     {
@@ -187,7 +205,7 @@
        "dtype: int64"
       ]
      },
-     "execution_count": 12,
+     "execution_count": 40,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -846,6 +864,33 @@
     "plt.legend(loc='lower right')"
    ]
   },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# Export data.\n",
+    "df_median = df.groupby(level=[0,1]).median()\n",
+    "df_median['label'] = 'unknown'\n",
+    "df_median.loc[[x.split('_')[::-1] for x in complexes.values()], 'label'] = 'known'\n",
+    "df_median = df_median.reset_index()\n",
+    "df_median['peptide'] = df_median['peptide'].replace({'P09683:28-54': 'Secretin',\n",
+    " 'P22466:33-62': 'Galanin',\n",
+    " 'P01178:20-28': 'Oxytocin',\n",
+    " 'P06307:96-103': 'Cholecystokinin-8',\n",
+    " 'P61278:103-116': 'Somatostatin-14',\n",
+    " 'P01303:29-64': 'Neuropeptide Y',\n",
+    " 'Q9UBU3:24-50': 'Ghrelin-27',\n",
+    " 'P20366:58-68': 'Substance P',\n",
+    " 'P01286:32-75': 'Somatoliberin',\n",
+    " 'P01350:76-92': 'Gastrin-17'})\n",
+    "\n",
+    "df_median = df_median.set_index(['peptide', 'receptor'])\n",
+    "\n",
+    "df_median.to_csv('results/pdb_benchmark_median.csv')"
+   ]
+  },
   {
    "cell_type": "code",
    "execution_count": 56,