predict_test_tanimoto_2out_v1_mf.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "b5e34019-55b5-49e9-acde-3d36a803dafa",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "\n",
    "df_samples = pd.read_csv(\"/Users/delser/ownCloud/Writing/MetSoc2022_CASMI_results.txt\",index_col= \"Compound Number\", sep=\"\\t\") # mzmine quant table\n",
    "\n",
    "df_pred = pd.read_csv('/Users/delser/mass2smiles/retrain/nist/tcn_seq/09022023_casmi_result_4l_50ep_twilight-durian-197.tsv',sep=\"\\t\") # MS2LDA download\n",
    "\n",
    "df_ref = pd.read_csv('/Users/delser/mass2smiles/retrain/nist/tcn_seq/casmi_func_groups_2201_mf.tsv',sep=\"\\t\") "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "f77a97cc-daea-49d0-beb5-cf5f90c85532",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "\n",
    "#df_samples = pd.read_csv(\"/Users/delser/ownCloud/Writing/MetSoc2022_CASMI_results.txt\",index_col= \"Compound Number\", sep=\"\\t\") # mzmine quant table\n",
    "\n",
    "df_pred = pd.read_csv('/Users/delser/mass2smiles/retrain/nist/tcn_seq/icy-lion-205-50ep_val.tsv',sep=\"\\t\") # MS2LDA download\n",
    "\n",
    "df_ref = pd.read_csv('/Users/delser/mass2smiles/retrain/nist/tcn_seq/all_HRMS_valid_24012023_cddd_refine_s_mf.tsv',sep=\"\\t\") "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "060cdaf2-dcb5-4094-abde-8488e86299d2",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Unnamed: 0</th>\n",
       "      <th>0</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>CC(C)=CC1C(O)CCc2ccc(=O)oc2C1CC(C)c1ccc(O)cc1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>O=C(Nc1cccs1)C(F)Br</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>O=c1ccc(Br)c2c(-c3ccccc3)[nH]cc12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>CCOC(=O)c1nnc(SC)nc1N</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4</td>\n",
       "      <td>COC(=O)C1CC(O)C(C)(O)C=C1C(O)Cc1ccc(O)c(CCCO)c1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>739</th>\n",
       "      <td>739</td>\n",
       "      <td>COc1cc(O)c(C(=O)C=Cc2ccc(O)cc2)c(OC)c1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>740</th>\n",
       "      <td>740</td>\n",
       "      <td>Nc1nc2c([O-])nc[nH]c2cc1=O</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>741</th>\n",
       "      <td>741</td>\n",
       "      <td>CC(=O)CC(=O)N(C)S(C)(=O)=O</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>742</th>\n",
       "      <td>742</td>\n",
       "      <td>CC(CO)c1ccccc1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>743</th>\n",
       "      <td>743</td>\n",
       "      <td>COc1cc(CC(=O)C2CCC(OC)c3ccc(OC)cc32)cc(OC)c1C</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>744 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     Unnamed: 0                                                0\n",
       "0             0    CC(C)=CC1C(O)CCc2ccc(=O)oc2C1CC(C)c1ccc(O)cc1\n",
       "1             1                              O=C(Nc1cccs1)C(F)Br\n",
       "2             2                O=c1ccc(Br)c2c(-c3ccccc3)[nH]cc12\n",
       "3             3                            CCOC(=O)c1nnc(SC)nc1N\n",
       "4             4  COC(=O)C1CC(O)C(C)(O)C=C1C(O)Cc1ccc(O)c(CCCO)c1\n",
       "..          ...                                              ...\n",
       "739         739           COc1cc(O)c(C(=O)C=Cc2ccc(O)cc2)c(OC)c1\n",
       "740         740                       Nc1nc2c([O-])nc[nH]c2cc1=O\n",
       "741         741                       CC(=O)CC(=O)N(C)S(C)(=O)=O\n",
       "742         742                                   CC(CO)c1ccccc1\n",
       "743         743    COc1cc(CC(=O)C2CCC(OC)c3ccc(OC)cc32)cc(OC)c1C\n",
       "\n",
       "[744 rows x 2 columns]"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_pred"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "90cf0f20-a907-4940-b8d1-1827be936e92",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Unnamed: 0.1</th>\n",
       "      <th>Unnamed: 0</th>\n",
       "      <th>spectrum_id</th>\n",
       "      <th>adduct</th>\n",
       "      <th>precursor_mz</th>\n",
       "      <th>mzs</th>\n",
       "      <th>intensities</th>\n",
       "      <th>loss_mzs</th>\n",
       "      <th>loss_intensities</th>\n",
       "      <th>inchikey</th>\n",
       "      <th>...</th>\n",
       "      <th>C</th>\n",
       "      <th>H</th>\n",
       "      <th>O</th>\n",
       "      <th>N</th>\n",
       "      <th>S</th>\n",
       "      <th>I</th>\n",
       "      <th>Br</th>\n",
       "      <th>Cl</th>\n",
       "      <th>F</th>\n",
       "      <th>P</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>17</td>\n",
       "      <td>CCMSLIB00004694894</td>\n",
       "      <td>[M+K]+</td>\n",
       "      <td>509.1060</td>\n",
       "      <td>[55.018578, 55.054958, 56.965504, 57.070599, 5...</td>\n",
       "      <td>[0.01610015, 0.07654480999999999, 0.0660954399...</td>\n",
       "      <td>[121.25959499999999, 123.254682, 130.257276, 1...</td>\n",
       "      <td>[0.07481905, 0.01660272, 0.06307306, 0.0109070...</td>\n",
       "      <td>AOPMSFXOYJXDNJ-IRFSQMTFSA-N</td>\n",
       "      <td>...</td>\n",
       "      <td>21</td>\n",
       "      <td>26</td>\n",
       "      <td>12</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>389</td>\n",
       "      <td>NIST460218</td>\n",
       "      <td>[M+H]+</td>\n",
       "      <td>205.9461</td>\n",
       "      <td>[78.0336, 79.0412, 96.0439, 197.0509, 205.9456...</td>\n",
       "      <td>[0.2914014014014014, 1.0, 0.3114014014014014, ...</td>\n",
       "      <td>[109.90220000000001, 126.9049, 127.9125]</td>\n",
       "      <td>[0.3114014014014014, 1.0, 0.2914014014014014]</td>\n",
       "      <td>CCZWSTFVHJPCEM-UHFFFAOYSA-N</td>\n",
       "      <td>...</td>\n",
       "      <td>5</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>359</td>\n",
       "      <td>NIST404672</td>\n",
       "      <td>[M+H]+</td>\n",
       "      <td>225.0481</td>\n",
       "      <td>[51.0227, 61.007, 62.0148, 63.0226, 64.0179, 6...</td>\n",
       "      <td>[0.058998998998999, 0.013703703703703704, 0.04...</td>\n",
       "      <td>[26.007900000000006, 26.022899999999993, 27.01...</td>\n",
       "      <td>[0.0463963963963964, 0.024304304304304306, 1.0...</td>\n",
       "      <td>XZSIGWOVDPSPMG-UHFFFAOYSA-N</td>\n",
       "      <td>...</td>\n",
       "      <td>13</td>\n",
       "      <td>8</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>500</td>\n",
       "      <td>NIST659328</td>\n",
       "      <td>[M+H]+</td>\n",
       "      <td>220.0209</td>\n",
       "      <td>[55.0288, 57.0319, 57.9743, 58.9822, 58.9947, ...</td>\n",
       "      <td>[1.0, 0.0216016016016016, 0.0308008008008008, ...</td>\n",
       "      <td>[28.031500000000023, 46.04220000000001, 74.037...</td>\n",
       "      <td>[0.0664964964964965, 1.0, 0.0394994994994995, ...</td>\n",
       "      <td>IYLGYUIQSDVGCP-UHFFFAOYSA-N</td>\n",
       "      <td>...</td>\n",
       "      <td>6</td>\n",
       "      <td>9</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4</td>\n",
       "      <td>14</td>\n",
       "      <td>CCMSLIB00004692719</td>\n",
       "      <td>[M+NH4]+</td>\n",
       "      <td>525.3170</td>\n",
       "      <td>[53.039364, 55.01862, 55.054993, 57.034267, 60...</td>\n",
       "      <td>[0.29508745000000003, 0.17668741000000002, 0.2...</td>\n",
       "      <td>[278.184218, 304.969313, 342.200118]</td>\n",
       "      <td>[0.17646052999999998, 0.030979589999999998, 0....</td>\n",
       "      <td>YQDJUXPUIRGKNV-MNOLRWOYSA-N</td>\n",
       "      <td>...</td>\n",
       "      <td>27</td>\n",
       "      <td>41</td>\n",
       "      <td>8</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>739</th>\n",
       "      <td>739</td>\n",
       "      <td>197</td>\n",
       "      <td>NIST103438</td>\n",
       "      <td>[M+H]+</td>\n",
       "      <td>301.1071</td>\n",
       "      <td>[52.0307, 68.9972, 69.0335, 74.0152, 75.0229, ...</td>\n",
       "      <td>[0.013103103103103104, 0.7116016016016016, 0.0...</td>\n",
       "      <td>[17.00630000000001, 17.042700000000025, 18.009...</td>\n",
       "      <td>[0.06, 0.015395395395395396, 1.0, 0.1130030030...</td>\n",
       "      <td>ONIYXBSQSPCBOF-BQYQJAHWSA-N</td>\n",
       "      <td>...</td>\n",
       "      <td>17</td>\n",
       "      <td>16</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>740</th>\n",
       "      <td>740</td>\n",
       "      <td>382</td>\n",
       "      <td>NIST45101</td>\n",
       "      <td>[M+H]+</td>\n",
       "      <td>393.0054</td>\n",
       "      <td>[51.0102, 51.7363, 52.018, 53.0132, 53.0258, 5...</td>\n",
       "      <td>[0.1053953953953954, 0.0381981981981982, 1.0, ...</td>\n",
       "      <td>[131.0383, 132.0251, 132.03070000000002, 132.0...</td>\n",
       "      <td>[0.0424024024024024, 0.013403403403403404, 0.0...</td>\n",
       "      <td>WHSIXKUPQCKWBY-FCKMSMMTSA-N</td>\n",
       "      <td>...</td>\n",
       "      <td>11</td>\n",
       "      <td>13</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>741</th>\n",
       "      <td>741</td>\n",
       "      <td>616</td>\n",
       "      <td>NIST854615</td>\n",
       "      <td>[M+H]+</td>\n",
       "      <td>253.0853</td>\n",
       "      <td>[53.0022, 53.0386, 54.01, 54.0338, 55.0178, 55...</td>\n",
       "      <td>[0.1897997997997998, 0.016796796796796798, 0.0...</td>\n",
       "      <td>[17.007199999999983, 18.010899999999992, 41.02...</td>\n",
       "      <td>[0.066996996996997, 1.0, 0.0207007007007007, 0...</td>\n",
       "      <td>KKQBVQCZKQBNPO-UHFFFAOYSA-N</td>\n",
       "      <td>...</td>\n",
       "      <td>8</td>\n",
       "      <td>16</td>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>742</th>\n",
       "      <td>742</td>\n",
       "      <td>448</td>\n",
       "      <td>NIST539936</td>\n",
       "      <td>[M-H2O+H]+</td>\n",
       "      <td>196.9960</td>\n",
       "      <td>[50.0151, 51.0229, 52.0307, 62.015, 63.0229, 6...</td>\n",
       "      <td>[0.3347047047047047, 0.8486986986986987, 0.042...</td>\n",
       "      <td>[26.0334, 27.028100000000023, 28.0314000000000...</td>\n",
       "      <td>[1.0, 0.058998998998999, 1.0, 0.02090090090090...</td>\n",
       "      <td>QLMFTKHAIDWIDC-UHFFFAOYSA-N</td>\n",
       "      <td>...</td>\n",
       "      <td>9</td>\n",
       "      <td>11</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>743</th>\n",
       "      <td>743</td>\n",
       "      <td>107</td>\n",
       "      <td>CCMSLIB00010105357</td>\n",
       "      <td>[M+H]+</td>\n",
       "      <td>360.2170</td>\n",
       "      <td>[81.070084, 93.069992, 107.08551, 121.100868, ...</td>\n",
       "      <td>[0.013257500375739077, 0.02456798916920971, 0....</td>\n",
       "      <td>[165.15193699999998, 191.13123, 206.154668, 21...</td>\n",
       "      <td>[0.03636381998028929, 0.09352071376761444, 0.1...</td>\n",
       "      <td>PXTMHMPZOUHPQU-UHFFFAOYSA-N</td>\n",
       "      <td>...</td>\n",
       "      <td>21</td>\n",
       "      <td>29</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>744 rows × 82 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     Unnamed: 0.1  Unnamed: 0         spectrum_id      adduct  precursor_mz  \\\n",
       "0               0          17  CCMSLIB00004694894      [M+K]+      509.1060   \n",
       "1               1         389          NIST460218      [M+H]+      205.9461   \n",
       "2               2         359          NIST404672      [M+H]+      225.0481   \n",
       "3               3         500          NIST659328      [M+H]+      220.0209   \n",
       "4               4          14  CCMSLIB00004692719    [M+NH4]+      525.3170   \n",
       "..            ...         ...                 ...         ...           ...   \n",
       "739           739         197          NIST103438      [M+H]+      301.1071   \n",
       "740           740         382           NIST45101      [M+H]+      393.0054   \n",
       "741           741         616          NIST854615      [M+H]+      253.0853   \n",
       "742           742         448          NIST539936  [M-H2O+H]+      196.9960   \n",
       "743           743         107  CCMSLIB00010105357      [M+H]+      360.2170   \n",
       "\n",
       "                                                   mzs  \\\n",
       "0    [55.018578, 55.054958, 56.965504, 57.070599, 5...   \n",
       "1    [78.0336, 79.0412, 96.0439, 197.0509, 205.9456...   \n",
       "2    [51.0227, 61.007, 62.0148, 63.0226, 64.0179, 6...   \n",
       "3    [55.0288, 57.0319, 57.9743, 58.9822, 58.9947, ...   \n",
       "4    [53.039364, 55.01862, 55.054993, 57.034267, 60...   \n",
       "..                                                 ...   \n",
       "739  [52.0307, 68.9972, 69.0335, 74.0152, 75.0229, ...   \n",
       "740  [51.0102, 51.7363, 52.018, 53.0132, 53.0258, 5...   \n",
       "741  [53.0022, 53.0386, 54.01, 54.0338, 55.0178, 55...   \n",
       "742  [50.0151, 51.0229, 52.0307, 62.015, 63.0229, 6...   \n",
       "743  [81.070084, 93.069992, 107.08551, 121.100868, ...   \n",
       "\n",
       "                                           intensities  \\\n",
       "0    [0.01610015, 0.07654480999999999, 0.0660954399...   \n",
       "1    [0.2914014014014014, 1.0, 0.3114014014014014, ...   \n",
       "2    [0.058998998998999, 0.013703703703703704, 0.04...   \n",
       "3    [1.0, 0.0216016016016016, 0.0308008008008008, ...   \n",
       "4    [0.29508745000000003, 0.17668741000000002, 0.2...   \n",
       "..                                                 ...   \n",
       "739  [0.013103103103103104, 0.7116016016016016, 0.0...   \n",
       "740  [0.1053953953953954, 0.0381981981981982, 1.0, ...   \n",
       "741  [0.1897997997997998, 0.016796796796796798, 0.0...   \n",
       "742  [0.3347047047047047, 0.8486986986986987, 0.042...   \n",
       "743  [0.013257500375739077, 0.02456798916920971, 0....   \n",
       "\n",
       "                                              loss_mzs  \\\n",
       "0    [121.25959499999999, 123.254682, 130.257276, 1...   \n",
       "1             [109.90220000000001, 126.9049, 127.9125]   \n",
       "2    [26.007900000000006, 26.022899999999993, 27.01...   \n",
       "3    [28.031500000000023, 46.04220000000001, 74.037...   \n",
       "4                 [278.184218, 304.969313, 342.200118]   \n",
       "..                                                 ...   \n",
       "739  [17.00630000000001, 17.042700000000025, 18.009...   \n",
       "740  [131.0383, 132.0251, 132.03070000000002, 132.0...   \n",
       "741  [17.007199999999983, 18.010899999999992, 41.02...   \n",
       "742  [26.0334, 27.028100000000023, 28.0314000000000...   \n",
       "743  [165.15193699999998, 191.13123, 206.154668, 21...   \n",
       "\n",
       "                                      loss_intensities  \\\n",
       "0    [0.07481905, 0.01660272, 0.06307306, 0.0109070...   \n",
       "1        [0.3114014014014014, 1.0, 0.2914014014014014]   \n",
       "2    [0.0463963963963964, 0.024304304304304306, 1.0...   \n",
       "3    [0.0664964964964965, 1.0, 0.0394994994994995, ...   \n",
       "4    [0.17646052999999998, 0.030979589999999998, 0....   \n",
       "..                                                 ...   \n",
       "739  [0.06, 0.015395395395395396, 1.0, 0.1130030030...   \n",
       "740  [0.0424024024024024, 0.013403403403403404, 0.0...   \n",
       "741  [0.066996996996997, 1.0, 0.0207007007007007, 0...   \n",
       "742  [1.0, 0.058998998998999, 1.0, 0.02090090090090...   \n",
       "743  [0.03636381998028929, 0.09352071376761444, 0.1...   \n",
       "\n",
       "                        inchikey  ...   C   H   O   N  S  I  Br  Cl  F  P  \n",
       "0    AOPMSFXOYJXDNJ-IRFSQMTFSA-N  ...  21  26  12   0  0  0   0   0  0  0  \n",
       "1    CCZWSTFVHJPCEM-UHFFFAOYSA-N  ...   5   4   0   1  0  1   0   0  0  0  \n",
       "2    XZSIGWOVDPSPMG-UHFFFAOYSA-N  ...  13   8   0   2  1  0   0   0  0  0  \n",
       "3    IYLGYUIQSDVGCP-UHFFFAOYSA-N  ...   6   9   2   3  2  0   0   0  0  0  \n",
       "4    YQDJUXPUIRGKNV-MNOLRWOYSA-N  ...  27  41   8   1  0  0   0   0  0  0  \n",
       "..                           ...  ...  ..  ..  ..  .. .. ..  ..  .. .. ..  \n",
       "739  ONIYXBSQSPCBOF-BQYQJAHWSA-N  ...  17  16   5   0  0  0   0   0  0  0  \n",
       "740  WHSIXKUPQCKWBY-FCKMSMMTSA-N  ...  11  13   4   4  0  1   0   0  0  0  \n",
       "741  KKQBVQCZKQBNPO-UHFFFAOYSA-N  ...   8  16   5   2  1  0   0   0  0  0  \n",
       "742  QLMFTKHAIDWIDC-UHFFFAOYSA-N  ...   9  11   1   0  0  0   1   0  0  0  \n",
       "743  PXTMHMPZOUHPQU-UHFFFAOYSA-N  ...  21  29   4   1  0  0   0   0  0  0  \n",
       "\n",
       "[744 rows x 82 columns]"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_ref"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "33bc7690-c77c-4376-a337-14584b7b32e3",
   "metadata": {},
   "source": [
    "df_pred['casmi_ID']=df_ref[\"spectrum_id\"]\n",
    "df_pred"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0ebe37e8-a2c9-464f-a3a3-fd789952887a",
   "metadata": {},
   "source": [
    "df_pred=df_pred.set_index('casmi_ID')\n",
    "df_pred"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "73150d50-5caf-4765-914a-75ca0c7dbf72",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_pred=df_pred.rename(columns={\"0\": \"predicted\"})"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "30252555-f427-473f-8440-ee240f1024b9",
   "metadata": {},
   "source": [
    "df_ref=df_ref.set_index('spectrum_id')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "24d69f4d-f6c4-459d-8fc8-3d8a54000663",
   "metadata": {},
   "source": [
    "df_samples=df_samples.rename(columns={\"SMILES\": \"CASMI_true\"})\n",
    "df_samples"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "4c293d33-b0fc-4fa9-9545-b10cf9b1c862",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>predicted</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>CC(C)=CC1C(O)CCc2ccc(=O)oc2C1CC(C)c1ccc(O)cc1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>O=C(Nc1cccs1)C(F)Br</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>O=c1ccc(Br)c2c(-c3ccccc3)[nH]cc12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>CCOC(=O)c1nnc(SC)nc1N</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>COC(=O)C1CC(O)C(C)(O)C=C1C(O)Cc1ccc(O)c(CCCO)c1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>739</th>\n",
       "      <td>COc1cc(O)c(C(=O)C=Cc2ccc(O)cc2)c(OC)c1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>740</th>\n",
       "      <td>Nc1nc2c([O-])nc[nH]c2cc1=O</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>741</th>\n",
       "      <td>CC(=O)CC(=O)N(C)S(C)(=O)=O</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>742</th>\n",
       "      <td>CC(CO)c1ccccc1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>743</th>\n",
       "      <td>COc1cc(CC(=O)C2CCC(OC)c3ccc(OC)cc32)cc(OC)c1C</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>744 rows × 1 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                           predicted\n",
       "0      CC(C)=CC1C(O)CCc2ccc(=O)oc2C1CC(C)c1ccc(O)cc1\n",
       "1                                O=C(Nc1cccs1)C(F)Br\n",
       "2                  O=c1ccc(Br)c2c(-c3ccccc3)[nH]cc12\n",
       "3                              CCOC(=O)c1nnc(SC)nc1N\n",
       "4    COC(=O)C1CC(O)C(C)(O)C=C1C(O)Cc1ccc(O)c(CCCO)c1\n",
       "..                                               ...\n",
       "739           COc1cc(O)c(C(=O)C=Cc2ccc(O)cc2)c(OC)c1\n",
       "740                       Nc1nc2c([O-])nc[nH]c2cc1=O\n",
       "741                       CC(=O)CC(=O)N(C)S(C)(=O)=O\n",
       "742                                   CC(CO)c1ccccc1\n",
       "743    COc1cc(CC(=O)C2CCC(OC)c3ccc(OC)cc32)cc(OC)c1C\n",
       "\n",
       "[744 rows x 1 columns]"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\n",
    "df_pred=df_pred.drop([\"Unnamed: 0\"], axis=1)\n",
    "df_pred"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "7a880736-b0d6-47ee-a505-3fb301ed7bae",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Unnamed: 0.1</th>\n",
       "      <th>Unnamed: 0</th>\n",
       "      <th>spectrum_id</th>\n",
       "      <th>adduct</th>\n",
       "      <th>precursor_mz</th>\n",
       "      <th>mzs</th>\n",
       "      <th>intensities</th>\n",
       "      <th>loss_mzs</th>\n",
       "      <th>loss_intensities</th>\n",
       "      <th>inchikey</th>\n",
       "      <th>...</th>\n",
       "      <th>H</th>\n",
       "      <th>O</th>\n",
       "      <th>N</th>\n",
       "      <th>S</th>\n",
       "      <th>I</th>\n",
       "      <th>Br</th>\n",
       "      <th>Cl</th>\n",
       "      <th>F</th>\n",
       "      <th>P</th>\n",
       "      <th>predicted</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>17</td>\n",
       "      <td>CCMSLIB00004694894</td>\n",
       "      <td>[M+K]+</td>\n",
       "      <td>509.1060</td>\n",
       "      <td>[55.018578, 55.054958, 56.965504, 57.070599, 5...</td>\n",
       "      <td>[0.01610015, 0.07654480999999999, 0.0660954399...</td>\n",
       "      <td>[121.25959499999999, 123.254682, 130.257276, 1...</td>\n",
       "      <td>[0.07481905, 0.01660272, 0.06307306, 0.0109070...</td>\n",
       "      <td>AOPMSFXOYJXDNJ-IRFSQMTFSA-N</td>\n",
       "      <td>...</td>\n",
       "      <td>26</td>\n",
       "      <td>12</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>CC(C)=CC1C(O)CCc2ccc(=O)oc2C1CC(C)c1ccc(O)cc1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>389</td>\n",
       "      <td>NIST460218</td>\n",
       "      <td>[M+H]+</td>\n",
       "      <td>205.9461</td>\n",
       "      <td>[78.0336, 79.0412, 96.0439, 197.0509, 205.9456...</td>\n",
       "      <td>[0.2914014014014014, 1.0, 0.3114014014014014, ...</td>\n",
       "      <td>[109.90220000000001, 126.9049, 127.9125]</td>\n",
       "      <td>[0.3114014014014014, 1.0, 0.2914014014014014]</td>\n",
       "      <td>CCZWSTFVHJPCEM-UHFFFAOYSA-N</td>\n",
       "      <td>...</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>O=C(Nc1cccs1)C(F)Br</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>359</td>\n",
       "      <td>NIST404672</td>\n",
       "      <td>[M+H]+</td>\n",
       "      <td>225.0481</td>\n",
       "      <td>[51.0227, 61.007, 62.0148, 63.0226, 64.0179, 6...</td>\n",
       "      <td>[0.058998998998999, 0.013703703703703704, 0.04...</td>\n",
       "      <td>[26.007900000000006, 26.022899999999993, 27.01...</td>\n",
       "      <td>[0.0463963963963964, 0.024304304304304306, 1.0...</td>\n",
       "      <td>XZSIGWOVDPSPMG-UHFFFAOYSA-N</td>\n",
       "      <td>...</td>\n",
       "      <td>8</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>O=c1ccc(Br)c2c(-c3ccccc3)[nH]cc12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>500</td>\n",
       "      <td>NIST659328</td>\n",
       "      <td>[M+H]+</td>\n",
       "      <td>220.0209</td>\n",
       "      <td>[55.0288, 57.0319, 57.9743, 58.9822, 58.9947, ...</td>\n",
       "      <td>[1.0, 0.0216016016016016, 0.0308008008008008, ...</td>\n",
       "      <td>[28.031500000000023, 46.04220000000001, 74.037...</td>\n",
       "      <td>[0.0664964964964965, 1.0, 0.0394994994994995, ...</td>\n",
       "      <td>IYLGYUIQSDVGCP-UHFFFAOYSA-N</td>\n",
       "      <td>...</td>\n",
       "      <td>9</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>CCOC(=O)c1nnc(SC)nc1N</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4</td>\n",
       "      <td>14</td>\n",
       "      <td>CCMSLIB00004692719</td>\n",
       "      <td>[M+NH4]+</td>\n",
       "      <td>525.3170</td>\n",
       "      <td>[53.039364, 55.01862, 55.054993, 57.034267, 60...</td>\n",
       "      <td>[0.29508745000000003, 0.17668741000000002, 0.2...</td>\n",
       "      <td>[278.184218, 304.969313, 342.200118]</td>\n",
       "      <td>[0.17646052999999998, 0.030979589999999998, 0....</td>\n",
       "      <td>YQDJUXPUIRGKNV-MNOLRWOYSA-N</td>\n",
       "      <td>...</td>\n",
       "      <td>41</td>\n",
       "      <td>8</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>COC(=O)C1CC(O)C(C)(O)C=C1C(O)Cc1ccc(O)c(CCCO)c1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>739</th>\n",
       "      <td>739</td>\n",
       "      <td>197</td>\n",
       "      <td>NIST103438</td>\n",
       "      <td>[M+H]+</td>\n",
       "      <td>301.1071</td>\n",
       "      <td>[52.0307, 68.9972, 69.0335, 74.0152, 75.0229, ...</td>\n",
       "      <td>[0.013103103103103104, 0.7116016016016016, 0.0...</td>\n",
       "      <td>[17.00630000000001, 17.042700000000025, 18.009...</td>\n",
       "      <td>[0.06, 0.015395395395395396, 1.0, 0.1130030030...</td>\n",
       "      <td>ONIYXBSQSPCBOF-BQYQJAHWSA-N</td>\n",
       "      <td>...</td>\n",
       "      <td>16</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>COc1cc(O)c(C(=O)C=Cc2ccc(O)cc2)c(OC)c1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>740</th>\n",
       "      <td>740</td>\n",
       "      <td>382</td>\n",
       "      <td>NIST45101</td>\n",
       "      <td>[M+H]+</td>\n",
       "      <td>393.0054</td>\n",
       "      <td>[51.0102, 51.7363, 52.018, 53.0132, 53.0258, 5...</td>\n",
       "      <td>[0.1053953953953954, 0.0381981981981982, 1.0, ...</td>\n",
       "      <td>[131.0383, 132.0251, 132.03070000000002, 132.0...</td>\n",
       "      <td>[0.0424024024024024, 0.013403403403403404, 0.0...</td>\n",
       "      <td>WHSIXKUPQCKWBY-FCKMSMMTSA-N</td>\n",
       "      <td>...</td>\n",
       "      <td>13</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>Nc1nc2c([O-])nc[nH]c2cc1=O</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>741</th>\n",
       "      <td>741</td>\n",
       "      <td>616</td>\n",
       "      <td>NIST854615</td>\n",
       "      <td>[M+H]+</td>\n",
       "      <td>253.0853</td>\n",
       "      <td>[53.0022, 53.0386, 54.01, 54.0338, 55.0178, 55...</td>\n",
       "      <td>[0.1897997997997998, 0.016796796796796798, 0.0...</td>\n",
       "      <td>[17.007199999999983, 18.010899999999992, 41.02...</td>\n",
       "      <td>[0.066996996996997, 1.0, 0.0207007007007007, 0...</td>\n",
       "      <td>KKQBVQCZKQBNPO-UHFFFAOYSA-N</td>\n",
       "      <td>...</td>\n",
       "      <td>16</td>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>CC(=O)CC(=O)N(C)S(C)(=O)=O</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>742</th>\n",
       "      <td>742</td>\n",
       "      <td>448</td>\n",
       "      <td>NIST539936</td>\n",
       "      <td>[M-H2O+H]+</td>\n",
       "      <td>196.9960</td>\n",
       "      <td>[50.0151, 51.0229, 52.0307, 62.015, 63.0229, 6...</td>\n",
       "      <td>[0.3347047047047047, 0.8486986986986987, 0.042...</td>\n",
       "      <td>[26.0334, 27.028100000000023, 28.0314000000000...</td>\n",
       "      <td>[1.0, 0.058998998998999, 1.0, 0.02090090090090...</td>\n",
       "      <td>QLMFTKHAIDWIDC-UHFFFAOYSA-N</td>\n",
       "      <td>...</td>\n",
       "      <td>11</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>CC(CO)c1ccccc1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>743</th>\n",
       "      <td>743</td>\n",
       "      <td>107</td>\n",
       "      <td>CCMSLIB00010105357</td>\n",
       "      <td>[M+H]+</td>\n",
       "      <td>360.2170</td>\n",
       "      <td>[81.070084, 93.069992, 107.08551, 121.100868, ...</td>\n",
       "      <td>[0.013257500375739077, 0.02456798916920971, 0....</td>\n",
       "      <td>[165.15193699999998, 191.13123, 206.154668, 21...</td>\n",
       "      <td>[0.03636381998028929, 0.09352071376761444, 0.1...</td>\n",
       "      <td>PXTMHMPZOUHPQU-UHFFFAOYSA-N</td>\n",
       "      <td>...</td>\n",
       "      <td>29</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>COc1cc(CC(=O)C2CCC(OC)c3ccc(OC)cc32)cc(OC)c1C</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>744 rows × 83 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     Unnamed: 0.1  Unnamed: 0         spectrum_id      adduct  precursor_mz  \\\n",
       "0               0          17  CCMSLIB00004694894      [M+K]+      509.1060   \n",
       "1               1         389          NIST460218      [M+H]+      205.9461   \n",
       "2               2         359          NIST404672      [M+H]+      225.0481   \n",
       "3               3         500          NIST659328      [M+H]+      220.0209   \n",
       "4               4          14  CCMSLIB00004692719    [M+NH4]+      525.3170   \n",
       "..            ...         ...                 ...         ...           ...   \n",
       "739           739         197          NIST103438      [M+H]+      301.1071   \n",
       "740           740         382           NIST45101      [M+H]+      393.0054   \n",
       "741           741         616          NIST854615      [M+H]+      253.0853   \n",
       "742           742         448          NIST539936  [M-H2O+H]+      196.9960   \n",
       "743           743         107  CCMSLIB00010105357      [M+H]+      360.2170   \n",
       "\n",
       "                                                   mzs  \\\n",
       "0    [55.018578, 55.054958, 56.965504, 57.070599, 5...   \n",
       "1    [78.0336, 79.0412, 96.0439, 197.0509, 205.9456...   \n",
       "2    [51.0227, 61.007, 62.0148, 63.0226, 64.0179, 6...   \n",
       "3    [55.0288, 57.0319, 57.9743, 58.9822, 58.9947, ...   \n",
       "4    [53.039364, 55.01862, 55.054993, 57.034267, 60...   \n",
       "..                                                 ...   \n",
       "739  [52.0307, 68.9972, 69.0335, 74.0152, 75.0229, ...   \n",
       "740  [51.0102, 51.7363, 52.018, 53.0132, 53.0258, 5...   \n",
       "741  [53.0022, 53.0386, 54.01, 54.0338, 55.0178, 55...   \n",
       "742  [50.0151, 51.0229, 52.0307, 62.015, 63.0229, 6...   \n",
       "743  [81.070084, 93.069992, 107.08551, 121.100868, ...   \n",
       "\n",
       "                                           intensities  \\\n",
       "0    [0.01610015, 0.07654480999999999, 0.0660954399...   \n",
       "1    [0.2914014014014014, 1.0, 0.3114014014014014, ...   \n",
       "2    [0.058998998998999, 0.013703703703703704, 0.04...   \n",
       "3    [1.0, 0.0216016016016016, 0.0308008008008008, ...   \n",
       "4    [0.29508745000000003, 0.17668741000000002, 0.2...   \n",
       "..                                                 ...   \n",
       "739  [0.013103103103103104, 0.7116016016016016, 0.0...   \n",
       "740  [0.1053953953953954, 0.0381981981981982, 1.0, ...   \n",
       "741  [0.1897997997997998, 0.016796796796796798, 0.0...   \n",
       "742  [0.3347047047047047, 0.8486986986986987, 0.042...   \n",
       "743  [0.013257500375739077, 0.02456798916920971, 0....   \n",
       "\n",
       "                                              loss_mzs  \\\n",
       "0    [121.25959499999999, 123.254682, 130.257276, 1...   \n",
       "1             [109.90220000000001, 126.9049, 127.9125]   \n",
       "2    [26.007900000000006, 26.022899999999993, 27.01...   \n",
       "3    [28.031500000000023, 46.04220000000001, 74.037...   \n",
       "4                 [278.184218, 304.969313, 342.200118]   \n",
       "..                                                 ...   \n",
       "739  [17.00630000000001, 17.042700000000025, 18.009...   \n",
       "740  [131.0383, 132.0251, 132.03070000000002, 132.0...   \n",
       "741  [17.007199999999983, 18.010899999999992, 41.02...   \n",
       "742  [26.0334, 27.028100000000023, 28.0314000000000...   \n",
       "743  [165.15193699999998, 191.13123, 206.154668, 21...   \n",
       "\n",
       "                                      loss_intensities  \\\n",
       "0    [0.07481905, 0.01660272, 0.06307306, 0.0109070...   \n",
       "1        [0.3114014014014014, 1.0, 0.2914014014014014]   \n",
       "2    [0.0463963963963964, 0.024304304304304306, 1.0...   \n",
       "3    [0.0664964964964965, 1.0, 0.0394994994994995, ...   \n",
       "4    [0.17646052999999998, 0.030979589999999998, 0....   \n",
       "..                                                 ...   \n",
       "739  [0.06, 0.015395395395395396, 1.0, 0.1130030030...   \n",
       "740  [0.0424024024024024, 0.013403403403403404, 0.0...   \n",
       "741  [0.066996996996997, 1.0, 0.0207007007007007, 0...   \n",
       "742  [1.0, 0.058998998998999, 1.0, 0.02090090090090...   \n",
       "743  [0.03636381998028929, 0.09352071376761444, 0.1...   \n",
       "\n",
       "                        inchikey  ...   H   O   N  S  I  Br  Cl  F  P  \\\n",
       "0    AOPMSFXOYJXDNJ-IRFSQMTFSA-N  ...  26  12   0  0  0   0   0  0  0   \n",
       "1    CCZWSTFVHJPCEM-UHFFFAOYSA-N  ...   4   0   1  0  1   0   0  0  0   \n",
       "2    XZSIGWOVDPSPMG-UHFFFAOYSA-N  ...   8   0   2  1  0   0   0  0  0   \n",
       "3    IYLGYUIQSDVGCP-UHFFFAOYSA-N  ...   9   2   3  2  0   0   0  0  0   \n",
       "4    YQDJUXPUIRGKNV-MNOLRWOYSA-N  ...  41   8   1  0  0   0   0  0  0   \n",
       "..                           ...  ...  ..  ..  .. .. ..  ..  .. .. ..   \n",
       "739  ONIYXBSQSPCBOF-BQYQJAHWSA-N  ...  16   5   0  0  0   0   0  0  0   \n",
       "740  WHSIXKUPQCKWBY-FCKMSMMTSA-N  ...  13   4   4  0  1   0   0  0  0   \n",
       "741  KKQBVQCZKQBNPO-UHFFFAOYSA-N  ...  16   5   2  1  0   0   0  0  0   \n",
       "742  QLMFTKHAIDWIDC-UHFFFAOYSA-N  ...  11   1   0  0  0   1   0  0  0   \n",
       "743  PXTMHMPZOUHPQU-UHFFFAOYSA-N  ...  29   4   1  0  0   0   0  0  0   \n",
       "\n",
       "                                           predicted  \n",
       "0      CC(C)=CC1C(O)CCc2ccc(=O)oc2C1CC(C)c1ccc(O)cc1  \n",
       "1                                O=C(Nc1cccs1)C(F)Br  \n",
       "2                  O=c1ccc(Br)c2c(-c3ccccc3)[nH]cc12  \n",
       "3                              CCOC(=O)c1nnc(SC)nc1N  \n",
       "4    COC(=O)C1CC(O)C(C)(O)C=C1C(O)Cc1ccc(O)c(CCCO)c1  \n",
       "..                                               ...  \n",
       "739           COc1cc(O)c(C(=O)C=Cc2ccc(O)cc2)c(OC)c1  \n",
       "740                       Nc1nc2c([O-])nc[nH]c2cc1=O  \n",
       "741                       CC(=O)CC(=O)N(C)S(C)(=O)=O  \n",
       "742                                   CC(CO)c1ccccc1  \n",
       "743    COc1cc(CC(=O)C2CCC(OC)c3ccc(OC)cc32)cc(OC)c1C  \n",
       "\n",
       "[744 rows x 83 columns]"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_final = df_ref.join(df_pred, how=\"outer\")\n",
    "#df_final=df_final.drop([\"Column 3\"], axis=1)\n",
    "#df_final=df_final.drop([\"Adduct\",\"spec\",\"smiles\"], axis=1)\n",
    "df_final"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "0bb2681e-c18e-4633-9684-0d4ea49361ea",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Unnamed: 0.1</th>\n",
       "      <th>Unnamed: 0</th>\n",
       "      <th>spectrum_id</th>\n",
       "      <th>precursor_mz</th>\n",
       "      <th>mzs</th>\n",
       "      <th>intensities</th>\n",
       "      <th>loss_mzs</th>\n",
       "      <th>loss_intensities</th>\n",
       "      <th>CASMI_true</th>\n",
       "      <th>num_of_sugars</th>\n",
       "      <th>...</th>\n",
       "      <th>H</th>\n",
       "      <th>O</th>\n",
       "      <th>N</th>\n",
       "      <th>S</th>\n",
       "      <th>I</th>\n",
       "      <th>Br</th>\n",
       "      <th>Cl</th>\n",
       "      <th>F</th>\n",
       "      <th>P</th>\n",
       "      <th>predicted</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>719.2538</td>\n",
       "      <td>[53.0387, 55.0179, 55.0545, 57.0338, 59.0492, ...</td>\n",
       "      <td>[0.017272727272727273, 0.01818181818181818, 0....</td>\n",
       "      <td>[314.15709999999996]</td>\n",
       "      <td>[0.01818181818181818]</td>\n",
       "      <td>CC1C(C(C(C(O1)OC2=C(OC3=C(C(=CC(=C3C2=O)O)O)CC...</td>\n",
       "      <td>2.0</td>\n",
       "      <td>...</td>\n",
       "      <td>42</td>\n",
       "      <td>16</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>CC1CC(O)c2ccc(CO)cc2C1N</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>499.2298</td>\n",
       "      <td>[67.0543, 69.0698, 81.0699, 83.0492, 83.0853, ...</td>\n",
       "      <td>[0.010357142857142856, 0.04642857142857143, 0....</td>\n",
       "      <td>[60.02260000000001, 118.02660000000003, 160.03...</td>\n",
       "      <td>[0.060714285714285714, 0.39285714285714285, 0....</td>\n",
       "      <td>CC1CC2(C(C1O)C=C(C(CC3C(C3(C)C)C=C(C2=O)C)OC(=...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>36</td>\n",
       "      <td>8</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>CCCCC(=O)C1CC(C)(O)C=C2C(CCC(=O)O)C3CC(C)CCC3(...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>6</td>\n",
       "      <td>1102.5777</td>\n",
       "      <td>[81.0334, 83.0491, 85.0224, 85.0284, 85.0333, ...</td>\n",
       "      <td>[0.14210526315789473, 0.2236842105263158, 0.01...</td>\n",
       "      <td>[]</td>\n",
       "      <td>[]</td>\n",
       "      <td>CC1C(C(C(C(O1)OC2C(OC(C(C2O)O)OCC3C(C(C(C(O3)O...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>...</td>\n",
       "      <td>84</td>\n",
       "      <td>22</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>CC1CCC(O)c2cc(ccc2O)C(CO)OC1=O</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>10</td>\n",
       "      <td>472.2082</td>\n",
       "      <td>[72.2736, 145.0759, 148.0868, 149.0707, 172.06...</td>\n",
       "      <td>[0.010697674418604652, 0.03488372093023256, 0....</td>\n",
       "      <td>[135.07919999999996, 148.0632, 165.08959999999...</td>\n",
       "      <td>[0.053488372093023255, 0.10465116279069768, 0....</td>\n",
       "      <td>CN1C2=C(C=C(C=C2)C(=O)N(CCC(=O)O)C3=CC=CC=N3)N...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>25</td>\n",
       "      <td>3</td>\n",
       "      <td>7</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>Cc1cc(Nc2ccc(N)cc2C=O)ncn1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>11</td>\n",
       "      <td>657.3116</td>\n",
       "      <td>[55.0542, 55.9818, 57.0334, 60.2456, 69.0335, ...</td>\n",
       "      <td>[0.018260869565217393, 0.013043478260869565, 0...</td>\n",
       "      <td>[158.094, 210.1282, 228.1352, 246.1485, 280.16...</td>\n",
       "      <td>[0.02217391304347826, 0.06086956521739131, 0.1...</td>\n",
       "      <td>CCCC(=O)OCC(C(C(CN1C2=C(C=C(C(=C2)C)C)N=C3C1=N...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>44</td>\n",
       "      <td>10</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>CC1CC(c2ccccc2)Nc2ccccc2C1=O</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>237</th>\n",
       "      <td>237</td>\n",
       "      <td>237</td>\n",
       "      <td>490</td>\n",
       "      <td>268.1541</td>\n",
       "      <td>[86.0599, 109.065, 109.1013, 121.101, 123.1168...</td>\n",
       "      <td>[0.10689655172413794, 0.04482758620689655, 0.0...</td>\n",
       "      <td>[18.01030000000003, 36.02120000000002, 46.0054...</td>\n",
       "      <td>[0.21724137931034482, 0.06206896551724138, 1.0...</td>\n",
       "      <td>CC1CCC2C1C(=O)OC(C2C)NC(=O)C3C(O3)C</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>21</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>O=C(Nc1ccccc1)C1CCc2ccccc21</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>238</th>\n",
       "      <td>238</td>\n",
       "      <td>238</td>\n",
       "      <td>491</td>\n",
       "      <td>411.3254</td>\n",
       "      <td>[55.0544, 57.0701, 67.0544, 69.0699, 81.0699, ...</td>\n",
       "      <td>[0.0215625, 0.053125, 0.015, 0.34375, 0.046875...</td>\n",
       "      <td>[18.01060000000001, 88.08850000000001, 142.135...</td>\n",
       "      <td>[0.153125, 0.02125, 0.01875, 0.02625, 0.028437...</td>\n",
       "      <td>CC(C)C(C)C=CC(C)C1CCC2C1(CCC3=C2C(=O)C=C4C3(CC...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>42</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>CC1=CC(C2CCCCC2)C(C)(O)Cc2ccc(Cl)cc21</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>239</th>\n",
       "      <td>239</td>\n",
       "      <td>239</td>\n",
       "      <td>492</td>\n",
       "      <td>430.2432</td>\n",
       "      <td>[50.4192, 52.9216, 74.1355, 81.0699, 86.4846, ...</td>\n",
       "      <td>[0.05555555555555555, 0.058333333333333334, 0....</td>\n",
       "      <td>[98.63150000000002, 179.0794, 197.0896, 214.55...</td>\n",
       "      <td>[0.06666666666666667, 0.2777777777777778, 1.0,...</td>\n",
       "      <td>CC1CCC(C2(C1=CC(CC2)C(=C)C(=O)O)C)OC3C(C(C(C(O...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>32</td>\n",
       "      <td>8</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>CC(=O)C1CC(c2ccccc2)Oc2ccccc21</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>240</th>\n",
       "      <td>240</td>\n",
       "      <td>240</td>\n",
       "      <td>495</td>\n",
       "      <td>578.2076</td>\n",
       "      <td>[54.5814, 57.0338, 61.0286, 69.0335, 81.0331, ...</td>\n",
       "      <td>[0.01, 0.010769230769230769, 0.038461538461538...</td>\n",
       "      <td>[160.76359999999994]</td>\n",
       "      <td>[0.011923076923076923]</td>\n",
       "      <td>COC1=C2C(=CC(=C1OC3C(C(C(C(O3)CO)O)O)OC4C(C(C(...</td>\n",
       "      <td>2.0</td>\n",
       "      <td>...</td>\n",
       "      <td>32</td>\n",
       "      <td>15</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>COC(=O)c1cc(=O)c2ccccc2o1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>241</th>\n",
       "      <td>241</td>\n",
       "      <td>241</td>\n",
       "      <td>500</td>\n",
       "      <td>243.1014</td>\n",
       "      <td>[71.9563, 104.546, 105.0698, 107.0491, 107.056...</td>\n",
       "      <td>[0.015833333333333335, 0.018333333333333333, 0...</td>\n",
       "      <td>[24.813600000000008, 42.197700000000026, 94.04...</td>\n",
       "      <td>[0.025833333333333333, 0.0175, 0.065, 1.0, 0.6...</td>\n",
       "      <td>C1C(COC2=C1C=CC(=C2)O)C3=CC=C(C=C3)O</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>14</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>Cc1ccc(C2CCc3cc(O)ccc32)cc1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>242 rows × 80 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     Unnamed: 0.1  Unnamed: 0  spectrum_id  precursor_mz  \\\n",
       "0               0           0            3      719.2538   \n",
       "1               1           1            5      499.2298   \n",
       "2               2           2            6     1102.5777   \n",
       "3               3           3           10      472.2082   \n",
       "4               4           4           11      657.3116   \n",
       "..            ...         ...          ...           ...   \n",
       "237           237         237          490      268.1541   \n",
       "238           238         238          491      411.3254   \n",
       "239           239         239          492      430.2432   \n",
       "240           240         240          495      578.2076   \n",
       "241           241         241          500      243.1014   \n",
       "\n",
       "                                                   mzs  \\\n",
       "0    [53.0387, 55.0179, 55.0545, 57.0338, 59.0492, ...   \n",
       "1    [67.0543, 69.0698, 81.0699, 83.0492, 83.0853, ...   \n",
       "2    [81.0334, 83.0491, 85.0224, 85.0284, 85.0333, ...   \n",
       "3    [72.2736, 145.0759, 148.0868, 149.0707, 172.06...   \n",
       "4    [55.0542, 55.9818, 57.0334, 60.2456, 69.0335, ...   \n",
       "..                                                 ...   \n",
       "237  [86.0599, 109.065, 109.1013, 121.101, 123.1168...   \n",
       "238  [55.0544, 57.0701, 67.0544, 69.0699, 81.0699, ...   \n",
       "239  [50.4192, 52.9216, 74.1355, 81.0699, 86.4846, ...   \n",
       "240  [54.5814, 57.0338, 61.0286, 69.0335, 81.0331, ...   \n",
       "241  [71.9563, 104.546, 105.0698, 107.0491, 107.056...   \n",
       "\n",
       "                                           intensities  \\\n",
       "0    [0.017272727272727273, 0.01818181818181818, 0....   \n",
       "1    [0.010357142857142856, 0.04642857142857143, 0....   \n",
       "2    [0.14210526315789473, 0.2236842105263158, 0.01...   \n",
       "3    [0.010697674418604652, 0.03488372093023256, 0....   \n",
       "4    [0.018260869565217393, 0.013043478260869565, 0...   \n",
       "..                                                 ...   \n",
       "237  [0.10689655172413794, 0.04482758620689655, 0.0...   \n",
       "238  [0.0215625, 0.053125, 0.015, 0.34375, 0.046875...   \n",
       "239  [0.05555555555555555, 0.058333333333333334, 0....   \n",
       "240  [0.01, 0.010769230769230769, 0.038461538461538...   \n",
       "241  [0.015833333333333335, 0.018333333333333333, 0...   \n",
       "\n",
       "                                              loss_mzs  \\\n",
       "0                                 [314.15709999999996]   \n",
       "1    [60.02260000000001, 118.02660000000003, 160.03...   \n",
       "2                                                   []   \n",
       "3    [135.07919999999996, 148.0632, 165.08959999999...   \n",
       "4    [158.094, 210.1282, 228.1352, 246.1485, 280.16...   \n",
       "..                                                 ...   \n",
       "237  [18.01030000000003, 36.02120000000002, 46.0054...   \n",
       "238  [18.01060000000001, 88.08850000000001, 142.135...   \n",
       "239  [98.63150000000002, 179.0794, 197.0896, 214.55...   \n",
       "240                               [160.76359999999994]   \n",
       "241  [24.813600000000008, 42.197700000000026, 94.04...   \n",
       "\n",
       "                                      loss_intensities  \\\n",
       "0                                [0.01818181818181818]   \n",
       "1    [0.060714285714285714, 0.39285714285714285, 0....   \n",
       "2                                                   []   \n",
       "3    [0.053488372093023255, 0.10465116279069768, 0....   \n",
       "4    [0.02217391304347826, 0.06086956521739131, 0.1...   \n",
       "..                                                 ...   \n",
       "237  [0.21724137931034482, 0.06206896551724138, 1.0...   \n",
       "238  [0.153125, 0.02125, 0.01875, 0.02625, 0.028437...   \n",
       "239  [0.06666666666666667, 0.2777777777777778, 1.0,...   \n",
       "240                             [0.011923076923076923]   \n",
       "241  [0.025833333333333333, 0.0175, 0.065, 1.0, 0.6...   \n",
       "\n",
       "                                            CASMI_true  num_of_sugars  ...  \\\n",
       "0    CC1C(C(C(C(O1)OC2=C(OC3=C(C(=CC(=C3C2=O)O)O)CC...            2.0  ...   \n",
       "1    CC1CC2(C(C1O)C=C(C(CC3C(C3(C)C)C=C(C2=O)C)OC(=...            0.0  ...   \n",
       "2    CC1C(C(C(C(O1)OC2C(OC(C(C2O)O)OCC3C(C(C(C(O3)O...            4.0  ...   \n",
       "3    CN1C2=C(C=C(C=C2)C(=O)N(CCC(=O)O)C3=CC=CC=N3)N...            0.0  ...   \n",
       "4    CCCC(=O)OCC(C(C(CN1C2=C(C=C(C(=C2)C)C)N=C3C1=N...            1.0  ...   \n",
       "..                                                 ...            ...  ...   \n",
       "237                CC1CCC2C1C(=O)OC(C2C)NC(=O)C3C(O3)C            0.0  ...   \n",
       "238  CC(C)C(C)C=CC(C)C1CCC2C1(CCC3=C2C(=O)C=C4C3(CC...            0.0  ...   \n",
       "239  CC1CCC(C2(C1=CC(CC2)C(=C)C(=O)O)C)OC3C(C(C(C(O...            1.0  ...   \n",
       "240  COC1=C2C(=CC(=C1OC3C(C(C(C(O3)CO)O)O)OC4C(C(C(...            2.0  ...   \n",
       "241               C1C(COC2=C1C=CC(=C2)O)C3=CC=C(C=C3)O            0.0  ...   \n",
       "\n",
       "      H   O   N  S  I  Br  Cl  F  P  \\\n",
       "0    42  16   0  0  0   0   0  0  0   \n",
       "1    36   8   0  0  0   0   0  0  0   \n",
       "2    84  22   0  0  0   0   0  0  0   \n",
       "3    25   3   7  0  0   0   0  0  0   \n",
       "4    44  10   4  0  0   0   0  0  0   \n",
       "..   ..  ..  .. .. ..  ..  .. .. ..   \n",
       "237  21   4   1  0  0   0   0  0  0   \n",
       "238  42   2   0  0  0   0   0  0  0   \n",
       "239  32   8   0  0  0   0   0  0  0   \n",
       "240  32  15   0  0  0   0   0  0  0   \n",
       "241  14   3   0  0  0   0   0  0  0   \n",
       "\n",
       "                                             predicted  \n",
       "0                              CC1CC(O)c2ccc(CO)cc2C1N  \n",
       "1    CCCCC(=O)C1CC(C)(O)C=C2C(CCC(=O)O)C3CC(C)CCC3(...  \n",
       "2                       CC1CCC(O)c2cc(ccc2O)C(CO)OC1=O  \n",
       "3                           Cc1cc(Nc2ccc(N)cc2C=O)ncn1  \n",
       "4                         CC1CC(c2ccccc2)Nc2ccccc2C1=O  \n",
       "..                                                 ...  \n",
       "237                        O=C(Nc1ccccc1)C1CCc2ccccc21  \n",
       "238              CC1=CC(C2CCCCC2)C(C)(O)Cc2ccc(Cl)cc21  \n",
       "239                     CC(=O)C1CC(c2ccccc2)Oc2ccccc21  \n",
       "240                          COC(=O)c1cc(=O)c2ccccc2o1  \n",
       "241                        Cc1ccc(C2CCc3cc(O)ccc32)cc1  \n",
       "\n",
       "[242 rows x 80 columns]"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_final=df_final.rename(columns={\"smiles_preprocessed\": \"CASMI_true\"})\n",
    "df_final"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "12fac9ff-a7a7-4f49-ace5-9f09780654d4",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Unnamed: 0.1</th>\n",
       "      <th>Unnamed: 0</th>\n",
       "      <th>spectrum_id</th>\n",
       "      <th>adduct</th>\n",
       "      <th>precursor_mz</th>\n",
       "      <th>mzs</th>\n",
       "      <th>intensities</th>\n",
       "      <th>loss_mzs</th>\n",
       "      <th>loss_intensities</th>\n",
       "      <th>inchikey</th>\n",
       "      <th>...</th>\n",
       "      <th>H</th>\n",
       "      <th>O</th>\n",
       "      <th>N</th>\n",
       "      <th>S</th>\n",
       "      <th>I</th>\n",
       "      <th>Br</th>\n",
       "      <th>Cl</th>\n",
       "      <th>F</th>\n",
       "      <th>P</th>\n",
       "      <th>predicted</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>17</td>\n",
       "      <td>CCMSLIB00004694894</td>\n",
       "      <td>[M+K]+</td>\n",
       "      <td>509.1060</td>\n",
       "      <td>[55.018578, 55.054958, 56.965504, 57.070599, 5...</td>\n",
       "      <td>[0.01610015, 0.07654480999999999, 0.0660954399...</td>\n",
       "      <td>[121.25959499999999, 123.254682, 130.257276, 1...</td>\n",
       "      <td>[0.07481905, 0.01660272, 0.06307306, 0.0109070...</td>\n",
       "      <td>AOPMSFXOYJXDNJ-IRFSQMTFSA-N</td>\n",
       "      <td>...</td>\n",
       "      <td>26</td>\n",
       "      <td>12</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>CC(C)=CC1C(O)CCc2ccc(=O)oc2C1CC(C)c1ccc(O)cc1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>389</td>\n",
       "      <td>NIST460218</td>\n",
       "      <td>[M+H]+</td>\n",
       "      <td>205.9461</td>\n",
       "      <td>[78.0336, 79.0412, 96.0439, 197.0509, 205.9456...</td>\n",
       "      <td>[0.2914014014014014, 1.0, 0.3114014014014014, ...</td>\n",
       "      <td>[109.90220000000001, 126.9049, 127.9125]</td>\n",
       "      <td>[0.3114014014014014, 1.0, 0.2914014014014014]</td>\n",
       "      <td>CCZWSTFVHJPCEM-UHFFFAOYSA-N</td>\n",
       "      <td>...</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>O=C(Nc1cccs1)C(F)Br</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>359</td>\n",
       "      <td>NIST404672</td>\n",
       "      <td>[M+H]+</td>\n",
       "      <td>225.0481</td>\n",
       "      <td>[51.0227, 61.007, 62.0148, 63.0226, 64.0179, 6...</td>\n",
       "      <td>[0.058998998998999, 0.013703703703703704, 0.04...</td>\n",
       "      <td>[26.007900000000006, 26.022899999999993, 27.01...</td>\n",
       "      <td>[0.0463963963963964, 0.024304304304304306, 1.0...</td>\n",
       "      <td>XZSIGWOVDPSPMG-UHFFFAOYSA-N</td>\n",
       "      <td>...</td>\n",
       "      <td>8</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>O=c1ccc(Br)c2c(-c3ccccc3)[nH]cc12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>500</td>\n",
       "      <td>NIST659328</td>\n",
       "      <td>[M+H]+</td>\n",
       "      <td>220.0209</td>\n",
       "      <td>[55.0288, 57.0319, 57.9743, 58.9822, 58.9947, ...</td>\n",
       "      <td>[1.0, 0.0216016016016016, 0.0308008008008008, ...</td>\n",
       "      <td>[28.031500000000023, 46.04220000000001, 74.037...</td>\n",
       "      <td>[0.0664964964964965, 1.0, 0.0394994994994995, ...</td>\n",
       "      <td>IYLGYUIQSDVGCP-UHFFFAOYSA-N</td>\n",
       "      <td>...</td>\n",
       "      <td>9</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>CCOC(=O)c1nnc(SC)nc1N</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4</td>\n",
       "      <td>14</td>\n",
       "      <td>CCMSLIB00004692719</td>\n",
       "      <td>[M+NH4]+</td>\n",
       "      <td>525.3170</td>\n",
       "      <td>[53.039364, 55.01862, 55.054993, 57.034267, 60...</td>\n",
       "      <td>[0.29508745000000003, 0.17668741000000002, 0.2...</td>\n",
       "      <td>[278.184218, 304.969313, 342.200118]</td>\n",
       "      <td>[0.17646052999999998, 0.030979589999999998, 0....</td>\n",
       "      <td>YQDJUXPUIRGKNV-MNOLRWOYSA-N</td>\n",
       "      <td>...</td>\n",
       "      <td>41</td>\n",
       "      <td>8</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>COC(=O)C1CC(O)C(C)(O)C=C1C(O)Cc1ccc(O)c(CCCO)c1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>739</th>\n",
       "      <td>739</td>\n",
       "      <td>197</td>\n",
       "      <td>NIST103438</td>\n",
       "      <td>[M+H]+</td>\n",
       "      <td>301.1071</td>\n",
       "      <td>[52.0307, 68.9972, 69.0335, 74.0152, 75.0229, ...</td>\n",
       "      <td>[0.013103103103103104, 0.7116016016016016, 0.0...</td>\n",
       "      <td>[17.00630000000001, 17.042700000000025, 18.009...</td>\n",
       "      <td>[0.06, 0.015395395395395396, 1.0, 0.1130030030...</td>\n",
       "      <td>ONIYXBSQSPCBOF-BQYQJAHWSA-N</td>\n",
       "      <td>...</td>\n",
       "      <td>16</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>COc1cc(O)c(C(=O)C=Cc2ccc(O)cc2)c(OC)c1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>740</th>\n",
       "      <td>740</td>\n",
       "      <td>382</td>\n",
       "      <td>NIST45101</td>\n",
       "      <td>[M+H]+</td>\n",
       "      <td>393.0054</td>\n",
       "      <td>[51.0102, 51.7363, 52.018, 53.0132, 53.0258, 5...</td>\n",
       "      <td>[0.1053953953953954, 0.0381981981981982, 1.0, ...</td>\n",
       "      <td>[131.0383, 132.0251, 132.03070000000002, 132.0...</td>\n",
       "      <td>[0.0424024024024024, 0.013403403403403404, 0.0...</td>\n",
       "      <td>WHSIXKUPQCKWBY-FCKMSMMTSA-N</td>\n",
       "      <td>...</td>\n",
       "      <td>13</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>Nc1nc2c([O-])nc[nH]c2cc1=O</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>741</th>\n",
       "      <td>741</td>\n",
       "      <td>616</td>\n",
       "      <td>NIST854615</td>\n",
       "      <td>[M+H]+</td>\n",
       "      <td>253.0853</td>\n",
       "      <td>[53.0022, 53.0386, 54.01, 54.0338, 55.0178, 55...</td>\n",
       "      <td>[0.1897997997997998, 0.016796796796796798, 0.0...</td>\n",
       "      <td>[17.007199999999983, 18.010899999999992, 41.02...</td>\n",
       "      <td>[0.066996996996997, 1.0, 0.0207007007007007, 0...</td>\n",
       "      <td>KKQBVQCZKQBNPO-UHFFFAOYSA-N</td>\n",
       "      <td>...</td>\n",
       "      <td>16</td>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>CC(=O)CC(=O)N(C)S(C)(=O)=O</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>742</th>\n",
       "      <td>742</td>\n",
       "      <td>448</td>\n",
       "      <td>NIST539936</td>\n",
       "      <td>[M-H2O+H]+</td>\n",
       "      <td>196.9960</td>\n",
       "      <td>[50.0151, 51.0229, 52.0307, 62.015, 63.0229, 6...</td>\n",
       "      <td>[0.3347047047047047, 0.8486986986986987, 0.042...</td>\n",
       "      <td>[26.0334, 27.028100000000023, 28.0314000000000...</td>\n",
       "      <td>[1.0, 0.058998998998999, 1.0, 0.02090090090090...</td>\n",
       "      <td>QLMFTKHAIDWIDC-UHFFFAOYSA-N</td>\n",
       "      <td>...</td>\n",
       "      <td>11</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>CC(CO)c1ccccc1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>743</th>\n",
       "      <td>743</td>\n",
       "      <td>107</td>\n",
       "      <td>CCMSLIB00010105357</td>\n",
       "      <td>[M+H]+</td>\n",
       "      <td>360.2170</td>\n",
       "      <td>[81.070084, 93.069992, 107.08551, 121.100868, ...</td>\n",
       "      <td>[0.013257500375739077, 0.02456798916920971, 0....</td>\n",
       "      <td>[165.15193699999998, 191.13123, 206.154668, 21...</td>\n",
       "      <td>[0.03636381998028929, 0.09352071376761444, 0.1...</td>\n",
       "      <td>PXTMHMPZOUHPQU-UHFFFAOYSA-N</td>\n",
       "      <td>...</td>\n",
       "      <td>29</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>COc1cc(CC(=O)C2CCC(OC)c3ccc(OC)cc32)cc(OC)c1C</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>744 rows × 83 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     Unnamed: 0.1  Unnamed: 0         spectrum_id      adduct  precursor_mz  \\\n",
       "0               0          17  CCMSLIB00004694894      [M+K]+      509.1060   \n",
       "1               1         389          NIST460218      [M+H]+      205.9461   \n",
       "2               2         359          NIST404672      [M+H]+      225.0481   \n",
       "3               3         500          NIST659328      [M+H]+      220.0209   \n",
       "4               4          14  CCMSLIB00004692719    [M+NH4]+      525.3170   \n",
       "..            ...         ...                 ...         ...           ...   \n",
       "739           739         197          NIST103438      [M+H]+      301.1071   \n",
       "740           740         382           NIST45101      [M+H]+      393.0054   \n",
       "741           741         616          NIST854615      [M+H]+      253.0853   \n",
       "742           742         448          NIST539936  [M-H2O+H]+      196.9960   \n",
       "743           743         107  CCMSLIB00010105357      [M+H]+      360.2170   \n",
       "\n",
       "                                                   mzs  \\\n",
       "0    [55.018578, 55.054958, 56.965504, 57.070599, 5...   \n",
       "1    [78.0336, 79.0412, 96.0439, 197.0509, 205.9456...   \n",
       "2    [51.0227, 61.007, 62.0148, 63.0226, 64.0179, 6...   \n",
       "3    [55.0288, 57.0319, 57.9743, 58.9822, 58.9947, ...   \n",
       "4    [53.039364, 55.01862, 55.054993, 57.034267, 60...   \n",
       "..                                                 ...   \n",
       "739  [52.0307, 68.9972, 69.0335, 74.0152, 75.0229, ...   \n",
       "740  [51.0102, 51.7363, 52.018, 53.0132, 53.0258, 5...   \n",
       "741  [53.0022, 53.0386, 54.01, 54.0338, 55.0178, 55...   \n",
       "742  [50.0151, 51.0229, 52.0307, 62.015, 63.0229, 6...   \n",
       "743  [81.070084, 93.069992, 107.08551, 121.100868, ...   \n",
       "\n",
       "                                           intensities  \\\n",
       "0    [0.01610015, 0.07654480999999999, 0.0660954399...   \n",
       "1    [0.2914014014014014, 1.0, 0.3114014014014014, ...   \n",
       "2    [0.058998998998999, 0.013703703703703704, 0.04...   \n",
       "3    [1.0, 0.0216016016016016, 0.0308008008008008, ...   \n",
       "4    [0.29508745000000003, 0.17668741000000002, 0.2...   \n",
       "..                                                 ...   \n",
       "739  [0.013103103103103104, 0.7116016016016016, 0.0...   \n",
       "740  [0.1053953953953954, 0.0381981981981982, 1.0, ...   \n",
       "741  [0.1897997997997998, 0.016796796796796798, 0.0...   \n",
       "742  [0.3347047047047047, 0.8486986986986987, 0.042...   \n",
       "743  [0.013257500375739077, 0.02456798916920971, 0....   \n",
       "\n",
       "                                              loss_mzs  \\\n",
       "0    [121.25959499999999, 123.254682, 130.257276, 1...   \n",
       "1             [109.90220000000001, 126.9049, 127.9125]   \n",
       "2    [26.007900000000006, 26.022899999999993, 27.01...   \n",
       "3    [28.031500000000023, 46.04220000000001, 74.037...   \n",
       "4                 [278.184218, 304.969313, 342.200118]   \n",
       "..                                                 ...   \n",
       "739  [17.00630000000001, 17.042700000000025, 18.009...   \n",
       "740  [131.0383, 132.0251, 132.03070000000002, 132.0...   \n",
       "741  [17.007199999999983, 18.010899999999992, 41.02...   \n",
       "742  [26.0334, 27.028100000000023, 28.0314000000000...   \n",
       "743  [165.15193699999998, 191.13123, 206.154668, 21...   \n",
       "\n",
       "                                      loss_intensities  \\\n",
       "0    [0.07481905, 0.01660272, 0.06307306, 0.0109070...   \n",
       "1        [0.3114014014014014, 1.0, 0.2914014014014014]   \n",
       "2    [0.0463963963963964, 0.024304304304304306, 1.0...   \n",
       "3    [0.0664964964964965, 1.0, 0.0394994994994995, ...   \n",
       "4    [0.17646052999999998, 0.030979589999999998, 0....   \n",
       "..                                                 ...   \n",
       "739  [0.06, 0.015395395395395396, 1.0, 0.1130030030...   \n",
       "740  [0.0424024024024024, 0.013403403403403404, 0.0...   \n",
       "741  [0.066996996996997, 1.0, 0.0207007007007007, 0...   \n",
       "742  [1.0, 0.058998998998999, 1.0, 0.02090090090090...   \n",
       "743  [0.03636381998028929, 0.09352071376761444, 0.1...   \n",
       "\n",
       "                        inchikey  ...   H   O   N  S  I  Br  Cl  F  P  \\\n",
       "0    AOPMSFXOYJXDNJ-IRFSQMTFSA-N  ...  26  12   0  0  0   0   0  0  0   \n",
       "1    CCZWSTFVHJPCEM-UHFFFAOYSA-N  ...   4   0   1  0  1   0   0  0  0   \n",
       "2    XZSIGWOVDPSPMG-UHFFFAOYSA-N  ...   8   0   2  1  0   0   0  0  0   \n",
       "3    IYLGYUIQSDVGCP-UHFFFAOYSA-N  ...   9   2   3  2  0   0   0  0  0   \n",
       "4    YQDJUXPUIRGKNV-MNOLRWOYSA-N  ...  41   8   1  0  0   0   0  0  0   \n",
       "..                           ...  ...  ..  ..  .. .. ..  ..  .. .. ..   \n",
       "739  ONIYXBSQSPCBOF-BQYQJAHWSA-N  ...  16   5   0  0  0   0   0  0  0   \n",
       "740  WHSIXKUPQCKWBY-FCKMSMMTSA-N  ...  13   4   4  0  1   0   0  0  0   \n",
       "741  KKQBVQCZKQBNPO-UHFFFAOYSA-N  ...  16   5   2  1  0   0   0  0  0   \n",
       "742  QLMFTKHAIDWIDC-UHFFFAOYSA-N  ...  11   1   0  0  0   1   0  0  0   \n",
       "743  PXTMHMPZOUHPQU-UHFFFAOYSA-N  ...  29   4   1  0  0   0   0  0  0   \n",
       "\n",
       "                                           predicted  \n",
       "0      CC(C)=CC1C(O)CCc2ccc(=O)oc2C1CC(C)c1ccc(O)cc1  \n",
       "1                                O=C(Nc1cccs1)C(F)Br  \n",
       "2                  O=c1ccc(Br)c2c(-c3ccccc3)[nH]cc12  \n",
       "3                              CCOC(=O)c1nnc(SC)nc1N  \n",
       "4    COC(=O)C1CC(O)C(C)(O)C=C1C(O)Cc1ccc(O)c(CCCO)c1  \n",
       "..                                               ...  \n",
       "739           COc1cc(O)c(C(=O)C=Cc2ccc(O)cc2)c(OC)c1  \n",
       "740                       Nc1nc2c([O-])nc[nH]c2cc1=O  \n",
       "741                       CC(=O)CC(=O)N(C)S(C)(=O)=O  \n",
       "742                                   CC(CO)c1ccccc1  \n",
       "743    COc1cc(CC(=O)C2CCC(OC)c3ccc(OC)cc32)cc(OC)c1C  \n",
       "\n",
       "[744 rows x 83 columns]"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_final=df_final.rename(columns={\"smiles_preprocessed\": \"val_true\"})\n",
    "df_final"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "73b62324-0e8c-4995-937d-2c6850cbdf55",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Unnamed: 0.1</th>\n",
       "      <th>spectrum_id</th>\n",
       "      <th>adduct</th>\n",
       "      <th>precursor_mz</th>\n",
       "      <th>mzs</th>\n",
       "      <th>intensities</th>\n",
       "      <th>loss_mzs</th>\n",
       "      <th>loss_intensities</th>\n",
       "      <th>inchikey</th>\n",
       "      <th>val_true</th>\n",
       "      <th>...</th>\n",
       "      <th>H</th>\n",
       "      <th>O</th>\n",
       "      <th>N</th>\n",
       "      <th>S</th>\n",
       "      <th>I</th>\n",
       "      <th>Br</th>\n",
       "      <th>Cl</th>\n",
       "      <th>F</th>\n",
       "      <th>P</th>\n",
       "      <th>predicted</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>CCMSLIB00004694894</td>\n",
       "      <td>[M+K]+</td>\n",
       "      <td>509.1060</td>\n",
       "      <td>[55.018578, 55.054958, 56.965504, 57.070599, 5...</td>\n",
       "      <td>[0.01610015, 0.07654480999999999, 0.0660954399...</td>\n",
       "      <td>[121.25959499999999, 123.254682, 130.257276, 1...</td>\n",
       "      <td>[0.07481905, 0.01660272, 0.06307306, 0.0109070...</td>\n",
       "      <td>AOPMSFXOYJXDNJ-IRFSQMTFSA-N</td>\n",
       "      <td>COC(=O)C1=COC(OC2OC(CO)C(O)C(O)C2O)C2C1C=CC21C...</td>\n",
       "      <td>...</td>\n",
       "      <td>26</td>\n",
       "      <td>12</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>CC(C)=CC1C(O)CCc2ccc(=O)oc2C1CC(C)c1ccc(O)cc1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>NIST460218</td>\n",
       "      <td>[M+H]+</td>\n",
       "      <td>205.9461</td>\n",
       "      <td>[78.0336, 79.0412, 96.0439, 197.0509, 205.9456...</td>\n",
       "      <td>[0.2914014014014014, 1.0, 0.3114014014014014, ...</td>\n",
       "      <td>[109.90220000000001, 126.9049, 127.9125]</td>\n",
       "      <td>[0.3114014014014014, 1.0, 0.2914014014014014]</td>\n",
       "      <td>CCZWSTFVHJPCEM-UHFFFAOYSA-N</td>\n",
       "      <td>Ic1ccccn1</td>\n",
       "      <td>...</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>O=C(Nc1cccs1)C(F)Br</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>NIST404672</td>\n",
       "      <td>[M+H]+</td>\n",
       "      <td>225.0481</td>\n",
       "      <td>[51.0227, 61.007, 62.0148, 63.0226, 64.0179, 6...</td>\n",
       "      <td>[0.058998998998999, 0.013703703703703704, 0.04...</td>\n",
       "      <td>[26.007900000000006, 26.022899999999993, 27.01...</td>\n",
       "      <td>[0.0463963963963964, 0.024304304304304306, 1.0...</td>\n",
       "      <td>XZSIGWOVDPSPMG-UHFFFAOYSA-N</td>\n",
       "      <td>N#Cc1ccc2c(c1)Nc1ccccc1S2</td>\n",
       "      <td>...</td>\n",
       "      <td>8</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>O=c1ccc(Br)c2c(-c3ccccc3)[nH]cc12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>NIST659328</td>\n",
       "      <td>[M+H]+</td>\n",
       "      <td>220.0209</td>\n",
       "      <td>[55.0288, 57.0319, 57.9743, 58.9822, 58.9947, ...</td>\n",
       "      <td>[1.0, 0.0216016016016016, 0.0308008008008008, ...</td>\n",
       "      <td>[28.031500000000023, 46.04220000000001, 74.037...</td>\n",
       "      <td>[0.0664964964964965, 1.0, 0.0394994994994995, ...</td>\n",
       "      <td>IYLGYUIQSDVGCP-UHFFFAOYSA-N</td>\n",
       "      <td>CCOC(=O)CSc1nnc(N)s1</td>\n",
       "      <td>...</td>\n",
       "      <td>9</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>CCOC(=O)c1nnc(SC)nc1N</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4</td>\n",
       "      <td>CCMSLIB00004692719</td>\n",
       "      <td>[M+NH4]+</td>\n",
       "      <td>525.3170</td>\n",
       "      <td>[53.039364, 55.01862, 55.054993, 57.034267, 60...</td>\n",
       "      <td>[0.29508745000000003, 0.17668741000000002, 0.2...</td>\n",
       "      <td>[278.184218, 304.969313, 342.200118]</td>\n",
       "      <td>[0.17646052999999998, 0.030979589999999998, 0....</td>\n",
       "      <td>YQDJUXPUIRGKNV-MNOLRWOYSA-N</td>\n",
       "      <td>COC(C=CCCC=CC(=O)O)C(O)C(C)C(O)C(C)=CC(C)C(=O)...</td>\n",
       "      <td>...</td>\n",
       "      <td>41</td>\n",
       "      <td>8</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>COC(=O)C1CC(O)C(C)(O)C=C1C(O)Cc1ccc(O)c(CCCO)c1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>739</th>\n",
       "      <td>739</td>\n",
       "      <td>NIST103438</td>\n",
       "      <td>[M+H]+</td>\n",
       "      <td>301.1071</td>\n",
       "      <td>[52.0307, 68.9972, 69.0335, 74.0152, 75.0229, ...</td>\n",
       "      <td>[0.013103103103103104, 0.7116016016016016, 0.0...</td>\n",
       "      <td>[17.00630000000001, 17.042700000000025, 18.009...</td>\n",
       "      <td>[0.06, 0.015395395395395396, 1.0, 0.1130030030...</td>\n",
       "      <td>ONIYXBSQSPCBOF-BQYQJAHWSA-N</td>\n",
       "      <td>COc1cc(O)c(C(=O)C=Cc2ccccc2O)c(OC)c1</td>\n",
       "      <td>...</td>\n",
       "      <td>16</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>COc1cc(O)c(C(=O)C=Cc2ccc(O)cc2)c(OC)c1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>740</th>\n",
       "      <td>740</td>\n",
       "      <td>NIST45101</td>\n",
       "      <td>[M+H]+</td>\n",
       "      <td>393.0054</td>\n",
       "      <td>[51.0102, 51.7363, 52.018, 53.0132, 53.0258, 5...</td>\n",
       "      <td>[0.1053953953953954, 0.0381981981981982, 1.0, ...</td>\n",
       "      <td>[131.0383, 132.0251, 132.03070000000002, 132.0...</td>\n",
       "      <td>[0.0424024024024024, 0.013403403403403404, 0.0...</td>\n",
       "      <td>WHSIXKUPQCKWBY-FCKMSMMTSA-N</td>\n",
       "      <td>Nc1ncnc2c1c(I)cn2C1OC(CO)C(O)C1O</td>\n",
       "      <td>...</td>\n",
       "      <td>13</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>Nc1nc2c([O-])nc[nH]c2cc1=O</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>741</th>\n",
       "      <td>741</td>\n",
       "      <td>NIST854615</td>\n",
       "      <td>[M+H]+</td>\n",
       "      <td>253.0853</td>\n",
       "      <td>[53.0022, 53.0386, 54.01, 54.0338, 55.0178, 55...</td>\n",
       "      <td>[0.1897997997997998, 0.016796796796796798, 0.0...</td>\n",
       "      <td>[17.007199999999983, 18.010899999999992, 41.02...</td>\n",
       "      <td>[0.066996996996997, 1.0, 0.0207007007007007, 0...</td>\n",
       "      <td>KKQBVQCZKQBNPO-UHFFFAOYSA-N</td>\n",
       "      <td>CNC(=O)CN(C)S(=O)(=O)CCCC(=O)O</td>\n",
       "      <td>...</td>\n",
       "      <td>16</td>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>CC(=O)CC(=O)N(C)S(C)(=O)=O</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>742</th>\n",
       "      <td>742</td>\n",
       "      <td>NIST539936</td>\n",
       "      <td>[M-H2O+H]+</td>\n",
       "      <td>196.9960</td>\n",
       "      <td>[50.0151, 51.0229, 52.0307, 62.015, 63.0229, 6...</td>\n",
       "      <td>[0.3347047047047047, 0.8486986986986987, 0.042...</td>\n",
       "      <td>[26.0334, 27.028100000000023, 28.0314000000000...</td>\n",
       "      <td>[1.0, 0.058998998998999, 1.0, 0.02090090090090...</td>\n",
       "      <td>QLMFTKHAIDWIDC-UHFFFAOYSA-N</td>\n",
       "      <td>CCC(O)c1ccc(Br)cc1</td>\n",
       "      <td>...</td>\n",
       "      <td>11</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>CC(CO)c1ccccc1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>743</th>\n",
       "      <td>743</td>\n",
       "      <td>CCMSLIB00010105357</td>\n",
       "      <td>[M+H]+</td>\n",
       "      <td>360.2170</td>\n",
       "      <td>[81.070084, 93.069992, 107.08551, 121.100868, ...</td>\n",
       "      <td>[0.013257500375739077, 0.02456798916920971, 0....</td>\n",
       "      <td>[165.15193699999998, 191.13123, 206.154668, 21...</td>\n",
       "      <td>[0.03636381998028929, 0.09352071376761444, 0.1...</td>\n",
       "      <td>PXTMHMPZOUHPQU-UHFFFAOYSA-N</td>\n",
       "      <td>COc1cc(C(O)=NCC23CC4CC(CC(C4)C2)C3)cc(OC)c1OC</td>\n",
       "      <td>...</td>\n",
       "      <td>29</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>COc1cc(CC(=O)C2CCC(OC)c3ccc(OC)cc32)cc(OC)c1C</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>744 rows × 82 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     Unnamed: 0.1         spectrum_id      adduct  precursor_mz  \\\n",
       "0               0  CCMSLIB00004694894      [M+K]+      509.1060   \n",
       "1               1          NIST460218      [M+H]+      205.9461   \n",
       "2               2          NIST404672      [M+H]+      225.0481   \n",
       "3               3          NIST659328      [M+H]+      220.0209   \n",
       "4               4  CCMSLIB00004692719    [M+NH4]+      525.3170   \n",
       "..            ...                 ...         ...           ...   \n",
       "739           739          NIST103438      [M+H]+      301.1071   \n",
       "740           740           NIST45101      [M+H]+      393.0054   \n",
       "741           741          NIST854615      [M+H]+      253.0853   \n",
       "742           742          NIST539936  [M-H2O+H]+      196.9960   \n",
       "743           743  CCMSLIB00010105357      [M+H]+      360.2170   \n",
       "\n",
       "                                                   mzs  \\\n",
       "0    [55.018578, 55.054958, 56.965504, 57.070599, 5...   \n",
       "1    [78.0336, 79.0412, 96.0439, 197.0509, 205.9456...   \n",
       "2    [51.0227, 61.007, 62.0148, 63.0226, 64.0179, 6...   \n",
       "3    [55.0288, 57.0319, 57.9743, 58.9822, 58.9947, ...   \n",
       "4    [53.039364, 55.01862, 55.054993, 57.034267, 60...   \n",
       "..                                                 ...   \n",
       "739  [52.0307, 68.9972, 69.0335, 74.0152, 75.0229, ...   \n",
       "740  [51.0102, 51.7363, 52.018, 53.0132, 53.0258, 5...   \n",
       "741  [53.0022, 53.0386, 54.01, 54.0338, 55.0178, 55...   \n",
       "742  [50.0151, 51.0229, 52.0307, 62.015, 63.0229, 6...   \n",
       "743  [81.070084, 93.069992, 107.08551, 121.100868, ...   \n",
       "\n",
       "                                           intensities  \\\n",
       "0    [0.01610015, 0.07654480999999999, 0.0660954399...   \n",
       "1    [0.2914014014014014, 1.0, 0.3114014014014014, ...   \n",
       "2    [0.058998998998999, 0.013703703703703704, 0.04...   \n",
       "3    [1.0, 0.0216016016016016, 0.0308008008008008, ...   \n",
       "4    [0.29508745000000003, 0.17668741000000002, 0.2...   \n",
       "..                                                 ...   \n",
       "739  [0.013103103103103104, 0.7116016016016016, 0.0...   \n",
       "740  [0.1053953953953954, 0.0381981981981982, 1.0, ...   \n",
       "741  [0.1897997997997998, 0.016796796796796798, 0.0...   \n",
       "742  [0.3347047047047047, 0.8486986986986987, 0.042...   \n",
       "743  [0.013257500375739077, 0.02456798916920971, 0....   \n",
       "\n",
       "                                              loss_mzs  \\\n",
       "0    [121.25959499999999, 123.254682, 130.257276, 1...   \n",
       "1             [109.90220000000001, 126.9049, 127.9125]   \n",
       "2    [26.007900000000006, 26.022899999999993, 27.01...   \n",
       "3    [28.031500000000023, 46.04220000000001, 74.037...   \n",
       "4                 [278.184218, 304.969313, 342.200118]   \n",
       "..                                                 ...   \n",
       "739  [17.00630000000001, 17.042700000000025, 18.009...   \n",
       "740  [131.0383, 132.0251, 132.03070000000002, 132.0...   \n",
       "741  [17.007199999999983, 18.010899999999992, 41.02...   \n",
       "742  [26.0334, 27.028100000000023, 28.0314000000000...   \n",
       "743  [165.15193699999998, 191.13123, 206.154668, 21...   \n",
       "\n",
       "                                      loss_intensities  \\\n",
       "0    [0.07481905, 0.01660272, 0.06307306, 0.0109070...   \n",
       "1        [0.3114014014014014, 1.0, 0.2914014014014014]   \n",
       "2    [0.0463963963963964, 0.024304304304304306, 1.0...   \n",
       "3    [0.0664964964964965, 1.0, 0.0394994994994995, ...   \n",
       "4    [0.17646052999999998, 0.030979589999999998, 0....   \n",
       "..                                                 ...   \n",
       "739  [0.06, 0.015395395395395396, 1.0, 0.1130030030...   \n",
       "740  [0.0424024024024024, 0.013403403403403404, 0.0...   \n",
       "741  [0.066996996996997, 1.0, 0.0207007007007007, 0...   \n",
       "742  [1.0, 0.058998998998999, 1.0, 0.02090090090090...   \n",
       "743  [0.03636381998028929, 0.09352071376761444, 0.1...   \n",
       "\n",
       "                        inchikey  \\\n",
       "0    AOPMSFXOYJXDNJ-IRFSQMTFSA-N   \n",
       "1    CCZWSTFVHJPCEM-UHFFFAOYSA-N   \n",
       "2    XZSIGWOVDPSPMG-UHFFFAOYSA-N   \n",
       "3    IYLGYUIQSDVGCP-UHFFFAOYSA-N   \n",
       "4    YQDJUXPUIRGKNV-MNOLRWOYSA-N   \n",
       "..                           ...   \n",
       "739  ONIYXBSQSPCBOF-BQYQJAHWSA-N   \n",
       "740  WHSIXKUPQCKWBY-FCKMSMMTSA-N   \n",
       "741  KKQBVQCZKQBNPO-UHFFFAOYSA-N   \n",
       "742  QLMFTKHAIDWIDC-UHFFFAOYSA-N   \n",
       "743  PXTMHMPZOUHPQU-UHFFFAOYSA-N   \n",
       "\n",
       "                                              val_true  ...   H   O   N  S  I  \\\n",
       "0    COC(=O)C1=COC(OC2OC(CO)C(O)C(O)C2O)C2C1C=CC21C...  ...  26  12   0  0  0   \n",
       "1                                            Ic1ccccn1  ...   4   0   1  0  1   \n",
       "2                            N#Cc1ccc2c(c1)Nc1ccccc1S2  ...   8   0   2  1  0   \n",
       "3                                 CCOC(=O)CSc1nnc(N)s1  ...   9   2   3  2  0   \n",
       "4    COC(C=CCCC=CC(=O)O)C(O)C(C)C(O)C(C)=CC(C)C(=O)...  ...  41   8   1  0  0   \n",
       "..                                                 ...  ...  ..  ..  .. .. ..   \n",
       "739               COc1cc(O)c(C(=O)C=Cc2ccccc2O)c(OC)c1  ...  16   5   0  0  0   \n",
       "740                   Nc1ncnc2c1c(I)cn2C1OC(CO)C(O)C1O  ...  13   4   4  0  1   \n",
       "741                     CNC(=O)CN(C)S(=O)(=O)CCCC(=O)O  ...  16   5   2  1  0   \n",
       "742                                 CCC(O)c1ccc(Br)cc1  ...  11   1   0  0  0   \n",
       "743      COc1cc(C(O)=NCC23CC4CC(CC(C4)C2)C3)cc(OC)c1OC  ...  29   4   1  0  0   \n",
       "\n",
       "     Br  Cl  F  P                                        predicted  \n",
       "0     0   0  0  0    CC(C)=CC1C(O)CCc2ccc(=O)oc2C1CC(C)c1ccc(O)cc1  \n",
       "1     0   0  0  0                              O=C(Nc1cccs1)C(F)Br  \n",
       "2     0   0  0  0                O=c1ccc(Br)c2c(-c3ccccc3)[nH]cc12  \n",
       "3     0   0  0  0                            CCOC(=O)c1nnc(SC)nc1N  \n",
       "4     0   0  0  0  COC(=O)C1CC(O)C(C)(O)C=C1C(O)Cc1ccc(O)c(CCCO)c1  \n",
       "..   ..  .. .. ..                                              ...  \n",
       "739   0   0  0  0           COc1cc(O)c(C(=O)C=Cc2ccc(O)cc2)c(OC)c1  \n",
       "740   0   0  0  0                       Nc1nc2c([O-])nc[nH]c2cc1=O  \n",
       "741   0   0  0  0                       CC(=O)CC(=O)N(C)S(C)(=O)=O  \n",
       "742   1   0  0  0                                   CC(CO)c1ccccc1  \n",
       "743   0   0  0  0    COc1cc(CC(=O)C2CCC(OC)c3ccc(OC)cc32)cc(OC)c1C  \n",
       "\n",
       "[744 rows x 82 columns]"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\n",
    "df_final=df_final.drop([\"Unnamed: 0\"], axis=1)\n",
    "df_final"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "c640f099-2b01-4ce3-98a3-68b6591aa4a7",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>spectrum_id</th>\n",
       "      <th>adduct</th>\n",
       "      <th>precursor_mz</th>\n",
       "      <th>mzs</th>\n",
       "      <th>intensities</th>\n",
       "      <th>loss_mzs</th>\n",
       "      <th>loss_intensities</th>\n",
       "      <th>inchikey</th>\n",
       "      <th>val_true</th>\n",
       "      <th>num_of_sugars</th>\n",
       "      <th>...</th>\n",
       "      <th>H</th>\n",
       "      <th>O</th>\n",
       "      <th>N</th>\n",
       "      <th>S</th>\n",
       "      <th>I</th>\n",
       "      <th>Br</th>\n",
       "      <th>Cl</th>\n",
       "      <th>F</th>\n",
       "      <th>P</th>\n",
       "      <th>predicted</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>CCMSLIB00004694894</td>\n",
       "      <td>[M+K]+</td>\n",
       "      <td>509.1060</td>\n",
       "      <td>[55.018578, 55.054958, 56.965504, 57.070599, 5...</td>\n",
       "      <td>[0.01610015, 0.07654480999999999, 0.0660954399...</td>\n",
       "      <td>[121.25959499999999, 123.254682, 130.257276, 1...</td>\n",
       "      <td>[0.07481905, 0.01660272, 0.06307306, 0.0109070...</td>\n",
       "      <td>AOPMSFXOYJXDNJ-IRFSQMTFSA-N</td>\n",
       "      <td>COC(=O)C1=COC(OC2OC(CO)C(O)C(O)C2O)C2C1C=CC21C...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>26</td>\n",
       "      <td>12</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>CC(C)=CC1C(O)CCc2ccc(=O)oc2C1CC(C)c1ccc(O)cc1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>NIST460218</td>\n",
       "      <td>[M+H]+</td>\n",
       "      <td>205.9461</td>\n",
       "      <td>[78.0336, 79.0412, 96.0439, 197.0509, 205.9456...</td>\n",
       "      <td>[0.2914014014014014, 1.0, 0.3114014014014014, ...</td>\n",
       "      <td>[109.90220000000001, 126.9049, 127.9125]</td>\n",
       "      <td>[0.3114014014014014, 1.0, 0.2914014014014014]</td>\n",
       "      <td>CCZWSTFVHJPCEM-UHFFFAOYSA-N</td>\n",
       "      <td>Ic1ccccn1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>O=C(Nc1cccs1)C(F)Br</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>NIST404672</td>\n",
       "      <td>[M+H]+</td>\n",
       "      <td>225.0481</td>\n",
       "      <td>[51.0227, 61.007, 62.0148, 63.0226, 64.0179, 6...</td>\n",
       "      <td>[0.058998998998999, 0.013703703703703704, 0.04...</td>\n",
       "      <td>[26.007900000000006, 26.022899999999993, 27.01...</td>\n",
       "      <td>[0.0463963963963964, 0.024304304304304306, 1.0...</td>\n",
       "      <td>XZSIGWOVDPSPMG-UHFFFAOYSA-N</td>\n",
       "      <td>N#Cc1ccc2c(c1)Nc1ccccc1S2</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>8</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>O=c1ccc(Br)c2c(-c3ccccc3)[nH]cc12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>NIST659328</td>\n",
       "      <td>[M+H]+</td>\n",
       "      <td>220.0209</td>\n",
       "      <td>[55.0288, 57.0319, 57.9743, 58.9822, 58.9947, ...</td>\n",
       "      <td>[1.0, 0.0216016016016016, 0.0308008008008008, ...</td>\n",
       "      <td>[28.031500000000023, 46.04220000000001, 74.037...</td>\n",
       "      <td>[0.0664964964964965, 1.0, 0.0394994994994995, ...</td>\n",
       "      <td>IYLGYUIQSDVGCP-UHFFFAOYSA-N</td>\n",
       "      <td>CCOC(=O)CSc1nnc(N)s1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>9</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>CCOC(=O)c1nnc(SC)nc1N</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>CCMSLIB00004692719</td>\n",
       "      <td>[M+NH4]+</td>\n",
       "      <td>525.3170</td>\n",
       "      <td>[53.039364, 55.01862, 55.054993, 57.034267, 60...</td>\n",
       "      <td>[0.29508745000000003, 0.17668741000000002, 0.2...</td>\n",
       "      <td>[278.184218, 304.969313, 342.200118]</td>\n",
       "      <td>[0.17646052999999998, 0.030979589999999998, 0....</td>\n",
       "      <td>YQDJUXPUIRGKNV-MNOLRWOYSA-N</td>\n",
       "      <td>COC(C=CCCC=CC(=O)O)C(O)C(C)C(O)C(C)=CC(C)C(=O)...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>41</td>\n",
       "      <td>8</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>COC(=O)C1CC(O)C(C)(O)C=C1C(O)Cc1ccc(O)c(CCCO)c1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>739</th>\n",
       "      <td>NIST103438</td>\n",
       "      <td>[M+H]+</td>\n",
       "      <td>301.1071</td>\n",
       "      <td>[52.0307, 68.9972, 69.0335, 74.0152, 75.0229, ...</td>\n",
       "      <td>[0.013103103103103104, 0.7116016016016016, 0.0...</td>\n",
       "      <td>[17.00630000000001, 17.042700000000025, 18.009...</td>\n",
       "      <td>[0.06, 0.015395395395395396, 1.0, 0.1130030030...</td>\n",
       "      <td>ONIYXBSQSPCBOF-BQYQJAHWSA-N</td>\n",
       "      <td>COc1cc(O)c(C(=O)C=Cc2ccccc2O)c(OC)c1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>16</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>COc1cc(O)c(C(=O)C=Cc2ccc(O)cc2)c(OC)c1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>740</th>\n",
       "      <td>NIST45101</td>\n",
       "      <td>[M+H]+</td>\n",
       "      <td>393.0054</td>\n",
       "      <td>[51.0102, 51.7363, 52.018, 53.0132, 53.0258, 5...</td>\n",
       "      <td>[0.1053953953953954, 0.0381981981981982, 1.0, ...</td>\n",
       "      <td>[131.0383, 132.0251, 132.03070000000002, 132.0...</td>\n",
       "      <td>[0.0424024024024024, 0.013403403403403404, 0.0...</td>\n",
       "      <td>WHSIXKUPQCKWBY-FCKMSMMTSA-N</td>\n",
       "      <td>Nc1ncnc2c1c(I)cn2C1OC(CO)C(O)C1O</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>13</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>Nc1nc2c([O-])nc[nH]c2cc1=O</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>741</th>\n",
       "      <td>NIST854615</td>\n",
       "      <td>[M+H]+</td>\n",
       "      <td>253.0853</td>\n",
       "      <td>[53.0022, 53.0386, 54.01, 54.0338, 55.0178, 55...</td>\n",
       "      <td>[0.1897997997997998, 0.016796796796796798, 0.0...</td>\n",
       "      <td>[17.007199999999983, 18.010899999999992, 41.02...</td>\n",
       "      <td>[0.066996996996997, 1.0, 0.0207007007007007, 0...</td>\n",
       "      <td>KKQBVQCZKQBNPO-UHFFFAOYSA-N</td>\n",
       "      <td>CNC(=O)CN(C)S(=O)(=O)CCCC(=O)O</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>16</td>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>CC(=O)CC(=O)N(C)S(C)(=O)=O</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>742</th>\n",
       "      <td>NIST539936</td>\n",
       "      <td>[M-H2O+H]+</td>\n",
       "      <td>196.9960</td>\n",
       "      <td>[50.0151, 51.0229, 52.0307, 62.015, 63.0229, 6...</td>\n",
       "      <td>[0.3347047047047047, 0.8486986986986987, 0.042...</td>\n",
       "      <td>[26.0334, 27.028100000000023, 28.0314000000000...</td>\n",
       "      <td>[1.0, 0.058998998998999, 1.0, 0.02090090090090...</td>\n",
       "      <td>QLMFTKHAIDWIDC-UHFFFAOYSA-N</td>\n",
       "      <td>CCC(O)c1ccc(Br)cc1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>11</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>CC(CO)c1ccccc1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>743</th>\n",
       "      <td>CCMSLIB00010105357</td>\n",
       "      <td>[M+H]+</td>\n",
       "      <td>360.2170</td>\n",
       "      <td>[81.070084, 93.069992, 107.08551, 121.100868, ...</td>\n",
       "      <td>[0.013257500375739077, 0.02456798916920971, 0....</td>\n",
       "      <td>[165.15193699999998, 191.13123, 206.154668, 21...</td>\n",
       "      <td>[0.03636381998028929, 0.09352071376761444, 0.1...</td>\n",
       "      <td>PXTMHMPZOUHPQU-UHFFFAOYSA-N</td>\n",
       "      <td>COc1cc(C(O)=NCC23CC4CC(CC(C4)C2)C3)cc(OC)c1OC</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>29</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>COc1cc(CC(=O)C2CCC(OC)c3ccc(OC)cc32)cc(OC)c1C</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>744 rows × 81 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "            spectrum_id      adduct  precursor_mz  \\\n",
       "0    CCMSLIB00004694894      [M+K]+      509.1060   \n",
       "1            NIST460218      [M+H]+      205.9461   \n",
       "2            NIST404672      [M+H]+      225.0481   \n",
       "3            NIST659328      [M+H]+      220.0209   \n",
       "4    CCMSLIB00004692719    [M+NH4]+      525.3170   \n",
       "..                  ...         ...           ...   \n",
       "739          NIST103438      [M+H]+      301.1071   \n",
       "740           NIST45101      [M+H]+      393.0054   \n",
       "741          NIST854615      [M+H]+      253.0853   \n",
       "742          NIST539936  [M-H2O+H]+      196.9960   \n",
       "743  CCMSLIB00010105357      [M+H]+      360.2170   \n",
       "\n",
       "                                                   mzs  \\\n",
       "0    [55.018578, 55.054958, 56.965504, 57.070599, 5...   \n",
       "1    [78.0336, 79.0412, 96.0439, 197.0509, 205.9456...   \n",
       "2    [51.0227, 61.007, 62.0148, 63.0226, 64.0179, 6...   \n",
       "3    [55.0288, 57.0319, 57.9743, 58.9822, 58.9947, ...   \n",
       "4    [53.039364, 55.01862, 55.054993, 57.034267, 60...   \n",
       "..                                                 ...   \n",
       "739  [52.0307, 68.9972, 69.0335, 74.0152, 75.0229, ...   \n",
       "740  [51.0102, 51.7363, 52.018, 53.0132, 53.0258, 5...   \n",
       "741  [53.0022, 53.0386, 54.01, 54.0338, 55.0178, 55...   \n",
       "742  [50.0151, 51.0229, 52.0307, 62.015, 63.0229, 6...   \n",
       "743  [81.070084, 93.069992, 107.08551, 121.100868, ...   \n",
       "\n",
       "                                           intensities  \\\n",
       "0    [0.01610015, 0.07654480999999999, 0.0660954399...   \n",
       "1    [0.2914014014014014, 1.0, 0.3114014014014014, ...   \n",
       "2    [0.058998998998999, 0.013703703703703704, 0.04...   \n",
       "3    [1.0, 0.0216016016016016, 0.0308008008008008, ...   \n",
       "4    [0.29508745000000003, 0.17668741000000002, 0.2...   \n",
       "..                                                 ...   \n",
       "739  [0.013103103103103104, 0.7116016016016016, 0.0...   \n",
       "740  [0.1053953953953954, 0.0381981981981982, 1.0, ...   \n",
       "741  [0.1897997997997998, 0.016796796796796798, 0.0...   \n",
       "742  [0.3347047047047047, 0.8486986986986987, 0.042...   \n",
       "743  [0.013257500375739077, 0.02456798916920971, 0....   \n",
       "\n",
       "                                              loss_mzs  \\\n",
       "0    [121.25959499999999, 123.254682, 130.257276, 1...   \n",
       "1             [109.90220000000001, 126.9049, 127.9125]   \n",
       "2    [26.007900000000006, 26.022899999999993, 27.01...   \n",
       "3    [28.031500000000023, 46.04220000000001, 74.037...   \n",
       "4                 [278.184218, 304.969313, 342.200118]   \n",
       "..                                                 ...   \n",
       "739  [17.00630000000001, 17.042700000000025, 18.009...   \n",
       "740  [131.0383, 132.0251, 132.03070000000002, 132.0...   \n",
       "741  [17.007199999999983, 18.010899999999992, 41.02...   \n",
       "742  [26.0334, 27.028100000000023, 28.0314000000000...   \n",
       "743  [165.15193699999998, 191.13123, 206.154668, 21...   \n",
       "\n",
       "                                      loss_intensities  \\\n",
       "0    [0.07481905, 0.01660272, 0.06307306, 0.0109070...   \n",
       "1        [0.3114014014014014, 1.0, 0.2914014014014014]   \n",
       "2    [0.0463963963963964, 0.024304304304304306, 1.0...   \n",
       "3    [0.0664964964964965, 1.0, 0.0394994994994995, ...   \n",
       "4    [0.17646052999999998, 0.030979589999999998, 0....   \n",
       "..                                                 ...   \n",
       "739  [0.06, 0.015395395395395396, 1.0, 0.1130030030...   \n",
       "740  [0.0424024024024024, 0.013403403403403404, 0.0...   \n",
       "741  [0.066996996996997, 1.0, 0.0207007007007007, 0...   \n",
       "742  [1.0, 0.058998998998999, 1.0, 0.02090090090090...   \n",
       "743  [0.03636381998028929, 0.09352071376761444, 0.1...   \n",
       "\n",
       "                        inchikey  \\\n",
       "0    AOPMSFXOYJXDNJ-IRFSQMTFSA-N   \n",
       "1    CCZWSTFVHJPCEM-UHFFFAOYSA-N   \n",
       "2    XZSIGWOVDPSPMG-UHFFFAOYSA-N   \n",
       "3    IYLGYUIQSDVGCP-UHFFFAOYSA-N   \n",
       "4    YQDJUXPUIRGKNV-MNOLRWOYSA-N   \n",
       "..                           ...   \n",
       "739  ONIYXBSQSPCBOF-BQYQJAHWSA-N   \n",
       "740  WHSIXKUPQCKWBY-FCKMSMMTSA-N   \n",
       "741  KKQBVQCZKQBNPO-UHFFFAOYSA-N   \n",
       "742  QLMFTKHAIDWIDC-UHFFFAOYSA-N   \n",
       "743  PXTMHMPZOUHPQU-UHFFFAOYSA-N   \n",
       "\n",
       "                                              val_true  num_of_sugars  ...  \\\n",
       "0    COC(=O)C1=COC(OC2OC(CO)C(O)C(O)C2O)C2C1C=CC21C...            1.0  ...   \n",
       "1                                            Ic1ccccn1            0.0  ...   \n",
       "2                            N#Cc1ccc2c(c1)Nc1ccccc1S2            0.0  ...   \n",
       "3                                 CCOC(=O)CSc1nnc(N)s1            0.0  ...   \n",
       "4    COC(C=CCCC=CC(=O)O)C(O)C(C)C(O)C(C)=CC(C)C(=O)...            0.0  ...   \n",
       "..                                                 ...            ...  ...   \n",
       "739               COc1cc(O)c(C(=O)C=Cc2ccccc2O)c(OC)c1            0.0  ...   \n",
       "740                   Nc1ncnc2c1c(I)cn2C1OC(CO)C(O)C1O            0.0  ...   \n",
       "741                     CNC(=O)CN(C)S(=O)(=O)CCCC(=O)O            0.0  ...   \n",
       "742                                 CCC(O)c1ccc(Br)cc1            0.0  ...   \n",
       "743      COc1cc(C(O)=NCC23CC4CC(CC(C4)C2)C3)cc(OC)c1OC            0.0  ...   \n",
       "\n",
       "      H   O   N  S  I  Br  Cl  F  P  \\\n",
       "0    26  12   0  0  0   0   0  0  0   \n",
       "1     4   0   1  0  1   0   0  0  0   \n",
       "2     8   0   2  1  0   0   0  0  0   \n",
       "3     9   2   3  2  0   0   0  0  0   \n",
       "4    41   8   1  0  0   0   0  0  0   \n",
       "..   ..  ..  .. .. ..  ..  .. .. ..   \n",
       "739  16   5   0  0  0   0   0  0  0   \n",
       "740  13   4   4  0  1   0   0  0  0   \n",
       "741  16   5   2  1  0   0   0  0  0   \n",
       "742  11   1   0  0  0   1   0  0  0   \n",
       "743  29   4   1  0  0   0   0  0  0   \n",
       "\n",
       "                                           predicted  \n",
       "0      CC(C)=CC1C(O)CCc2ccc(=O)oc2C1CC(C)c1ccc(O)cc1  \n",
       "1                                O=C(Nc1cccs1)C(F)Br  \n",
       "2                  O=c1ccc(Br)c2c(-c3ccccc3)[nH]cc12  \n",
       "3                              CCOC(=O)c1nnc(SC)nc1N  \n",
       "4    COC(=O)C1CC(O)C(C)(O)C=C1C(O)Cc1ccc(O)c(CCCO)c1  \n",
       "..                                               ...  \n",
       "739           COc1cc(O)c(C(=O)C=Cc2ccc(O)cc2)c(OC)c1  \n",
       "740                       Nc1nc2c([O-])nc[nH]c2cc1=O  \n",
       "741                       CC(=O)CC(=O)N(C)S(C)(=O)=O  \n",
       "742                                   CC(CO)c1ccccc1  \n",
       "743    COc1cc(CC(=O)C2CCC(OC)c3ccc(OC)cc32)cc(OC)c1C  \n",
       "\n",
       "[744 rows x 81 columns]"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\n",
    "df_final=df_final.drop([\"Unnamed: 0.1\"], axis=1)\n",
    "df_final"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1bce8634-8bc8-49c1-a512-d2a34405f3ea",
   "metadata": {},
   "source": [
    "df_final=df_final.drop([\"Unnamed: 0.1\"], axis=1)\n",
    "df_final"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "de1f2420-c803-4d07-9d87-8fd0af606d5d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'CC1CC2(C(C1O)C=C(C(CC3C(C3(C)C)C=C(C2=O)C)OC(=O)CCC(=O)O)C)OC(=O)C'"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_final[\"CASMI_true\"][1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "ad931806-e173-4ac3-8f91-cfd6de36721f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'CC(C=CC1(O)CCC(=O)C=C1C)=CC=CC(C)C1OC(CC(=O)O)CCC1C'"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_final[\"predicted\"][1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "df831f84-be49-4a6c-a895-df3dd747aabe",
   "metadata": {},
   "outputs": [],
   "source": [
    "from rdkit import DataStructs\n",
    "from rdkit import Chem\n",
    "#m_true=Chem.MolFromSmiles(df_final[\"decoded_test\"][1])\n",
    "#m_predict=Chem.MolFromSmiles(df_final[\"predicted\"][1])\n",
    "#ms=[m_true,m_predict]\n",
    "#fps = [Chem.RDKFingerprint(x) for x in ms]\n",
    "#DataStructs.FingerprintSimilarity(fps[0],fps[1])\n",
    "\n",
    "\n",
    "def get_tani(smiles1,smiles2):\n",
    "    try:\n",
    "        m_true=Chem.MolFromSmiles(smiles1,sanitize=False)\n",
    "        m_predict=Chem.MolFromSmiles(smiles2,sanitize=False)\n",
    "        ms=[m_true,m_predict]\n",
    "        fps = [Chem.RDKFingerprint(x) for x in ms]\n",
    "        return DataStructs.FingerprintSimilarity(fps[0],fps[1])\n",
    "    except:\n",
    "        return \"no_prediction_or_error_with_parsing_by_rdkit\"\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "4ecf2903-9aa4-4035-ba92-8b95598ef2f8",
   "metadata": {},
   "outputs": [],
   "source": [
    "from rdkit import DataStructs\n",
    "from rdkit import Chem\n",
    "#m_true=Chem.MolFromSmiles(df_final[\"decoded_test\"][1])\n",
    "#m_predict=Chem.MolFromSmiles(df_final[\"predicted\"][1])\n",
    "#ms=[m_true,m_predict]\n",
    "#fps = [Chem.RDKFingerprint(x) for x in ms]\n",
    "#DataStructs.FingerprintSimilarity(fps[0],fps[1])\n",
    "\n",
    "\n",
    "def get_tani(smiles1,smiles2):\n",
    "    try:\n",
    "        m_true=Chem.MolFromSmiles(smiles1,)\n",
    "        m_predict=Chem.MolFromSmiles(smiles2)\n",
    "        ms=[m_true,m_predict]\n",
    "        fps = [Chem.RDKFingerprint(x) for x in ms]\n",
    "        return DataStructs.FingerprintSimilarity(fps[0],fps[1])\n",
    "    except:\n",
    "        return \"no_prediction_or_error_with_parsing_by_rdkit\"\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "98789ca8-b4ad-4a61-ad05-654ee819c34b",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "RDKit ERROR: [14:15:45] Can't kekulize mol.  Unkekulized atoms: 2 3 4 6 7 15 16\n",
      "[14:15:45] Can't kekulize mol.  Unkekulized atoms: 2 3 4 6 7 15 16\n",
      "\n",
      "RDKit ERROR: \n",
      "RDKit ERROR: [14:15:46] Can't kekulize mol.  Unkekulized atoms: 3 4 5 6 7 9 10 12 14\n",
      "[14:15:46] Can't kekulize mol.  Unkekulized atoms: 3 4 5 6 7 9 10 12 14\n",
      "\n",
      "RDKit ERROR: \n",
      "RDKit ERROR: [14:15:46] Can't kekulize mol.  Unkekulized atoms: 2 3 4 6 7 8 13 14 15 16 19\n",
      "[14:15:46] Can't kekulize mol.  Unkekulized atoms: 2 3 4 6 7 8 13 14 15 16 19\n",
      "\n",
      "RDKit ERROR: \n",
      "RDKit ERROR: [14:15:46] Can't kekulize mol.  Unkekulized atoms: 2 3 4 5 7 8 11 12 13 17 18\n",
      "[14:15:46] Can't kekulize mol.  Unkekulized atoms: 2 3 4 5 7 8 11 12 13 17 18\n",
      "\n",
      "RDKit ERROR: \n",
      "RDKit ERROR: [14:15:46] Can't kekulize mol.  Unkekulized atoms: 2 3 4 6 7 8 13\n",
      "[14:15:46] Can't kekulize mol.  Unkekulized atoms: 2 3 4 6 7 8 13\n",
      "\n",
      "RDKit ERROR: \n",
      "RDKit ERROR: [14:15:47] Can't kekulize mol.  Unkekulized atoms: 15 16 17 18 19\n",
      "RDKit ERROR: \n",
      "[14:15:47] Can't kekulize mol.  Unkekulized atoms: 15 16 17 18 19\n",
      "\n",
      "RDKit ERROR: [14:15:47] Can't kekulize mol.  Unkekulized atoms: 2 3 4 7 8 9 22 26 28\n",
      "[14:15:47] Can't kekulize mol.  Unkekulized atoms: 2 3 4 7 8 9 22 26 28\n",
      "\n",
      "RDKit ERROR: \n",
      "RDKit ERROR: [14:15:47] Can't kekulize mol.  Unkekulized atoms: 3 4 5 6 7 8 11 12 13 14 15\n",
      "[14:15:47] Can't kekulize mol.  Unkekulized atoms: 3 4 5 6 7 8 11 12 13 14 15\n",
      "\n",
      "RDKit ERROR: \n",
      "RDKit ERROR: [14:15:48] Can't kekulize mol.  Unkekulized atoms: 2 3 4 5 8 9 16 19 20\n",
      "[14:15:48] Can't kekulize mol.  Unkekulized atoms: 2 3 4 5 8 9 16 19 20\n",
      "\n",
      "RDKit ERROR: \n",
      "RDKit ERROR: [14:15:48] Can't kekulize mol.  Unkekulized atoms: 3 4 6 7 8 10 11 12 13\n",
      "[14:15:48] Can't kekulize mol.  Unkekulized atoms: 3 4 6 7 8 10 11 12 13\n",
      "\n",
      "RDKit ERROR: \n",
      "RDKit ERROR: [14:15:48] Can't kekulize mol.  Unkekulized atoms: 2 3 4 5 6 7 8 9 10\n",
      "[14:15:48] Can't kekulize mol.  Unkekulized atoms: 2 3 4 5 6 7 8 9 10\n",
      "\n",
      "RDKit ERROR: \n",
      "[14:15:48] Can't kekulize mol.  Unkekulized atoms: 5 6 9 10 13 14 16 27 29\n",
      "\n",
      "RDKit ERROR: [14:15:48] Can't kekulize mol.  Unkekulized atoms: 5 6 9 10 13 14 16 27 29\n",
      "RDKit ERROR: \n"
     ]
    }
   ],
   "source": [
    "#df_final['tanimoto'] = df_final.apply(lambda x: get_tani(x.CASMI_true, x.predicted), axis=1)\n",
    "df_final['tanimoto'] = df_final.apply(lambda x: get_tani(x.val_true, x.predicted), axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "65f5b522-e8b4-4ea9-ab95-bffc427e9753",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>spectrum_id</th>\n",
       "      <th>adduct</th>\n",
       "      <th>precursor_mz</th>\n",
       "      <th>mzs</th>\n",
       "      <th>intensities</th>\n",
       "      <th>loss_mzs</th>\n",
       "      <th>loss_intensities</th>\n",
       "      <th>inchikey</th>\n",
       "      <th>val_true</th>\n",
       "      <th>num_of_sugars</th>\n",
       "      <th>...</th>\n",
       "      <th>O</th>\n",
       "      <th>N</th>\n",
       "      <th>S</th>\n",
       "      <th>I</th>\n",
       "      <th>Br</th>\n",
       "      <th>Cl</th>\n",
       "      <th>F</th>\n",
       "      <th>P</th>\n",
       "      <th>predicted</th>\n",
       "      <th>tanimoto</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>CCMSLIB00004694894</td>\n",
       "      <td>[M+K]+</td>\n",
       "      <td>509.1060</td>\n",
       "      <td>[55.018578, 55.054958, 56.965504, 57.070599, 5...</td>\n",
       "      <td>[0.01610015, 0.07654480999999999, 0.0660954399...</td>\n",
       "      <td>[121.25959499999999, 123.254682, 130.257276, 1...</td>\n",
       "      <td>[0.07481905, 0.01660272, 0.06307306, 0.0109070...</td>\n",
       "      <td>AOPMSFXOYJXDNJ-IRFSQMTFSA-N</td>\n",
       "      <td>COC(=O)C1=COC(OC2OC(CO)C(O)C(O)C2O)C2C1C=CC21C...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>12</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>CC(C)=CC1C(O)CCc2ccc(=O)oc2C1CC(C)c1ccc(O)cc1</td>\n",
       "      <td>0.383088</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>NIST460218</td>\n",
       "      <td>[M+H]+</td>\n",
       "      <td>205.9461</td>\n",
       "      <td>[78.0336, 79.0412, 96.0439, 197.0509, 205.9456...</td>\n",
       "      <td>[0.2914014014014014, 1.0, 0.3114014014014014, ...</td>\n",
       "      <td>[109.90220000000001, 126.9049, 127.9125]</td>\n",
       "      <td>[0.3114014014014014, 1.0, 0.2914014014014014]</td>\n",
       "      <td>CCZWSTFVHJPCEM-UHFFFAOYSA-N</td>\n",
       "      <td>Ic1ccccn1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>O=C(Nc1cccs1)C(F)Br</td>\n",
       "      <td>0.048338</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>NIST404672</td>\n",
       "      <td>[M+H]+</td>\n",
       "      <td>225.0481</td>\n",
       "      <td>[51.0227, 61.007, 62.0148, 63.0226, 64.0179, 6...</td>\n",
       "      <td>[0.058998998998999, 0.013703703703703704, 0.04...</td>\n",
       "      <td>[26.007900000000006, 26.022899999999993, 27.01...</td>\n",
       "      <td>[0.0463963963963964, 0.024304304304304306, 1.0...</td>\n",
       "      <td>XZSIGWOVDPSPMG-UHFFFAOYSA-N</td>\n",
       "      <td>N#Cc1ccc2c(c1)Nc1ccccc1S2</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>O=c1ccc(Br)c2c(-c3ccccc3)[nH]cc12</td>\n",
       "      <td>no_prediction_or_error_with_parsing_by_rdkit</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>NIST659328</td>\n",
       "      <td>[M+H]+</td>\n",
       "      <td>220.0209</td>\n",
       "      <td>[55.0288, 57.0319, 57.9743, 58.9822, 58.9947, ...</td>\n",
       "      <td>[1.0, 0.0216016016016016, 0.0308008008008008, ...</td>\n",
       "      <td>[28.031500000000023, 46.04220000000001, 74.037...</td>\n",
       "      <td>[0.0664964964964965, 1.0, 0.0394994994994995, ...</td>\n",
       "      <td>IYLGYUIQSDVGCP-UHFFFAOYSA-N</td>\n",
       "      <td>CCOC(=O)CSc1nnc(N)s1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>CCOC(=O)c1nnc(SC)nc1N</td>\n",
       "      <td>0.155352</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>CCMSLIB00004692719</td>\n",
       "      <td>[M+NH4]+</td>\n",
       "      <td>525.3170</td>\n",
       "      <td>[53.039364, 55.01862, 55.054993, 57.034267, 60...</td>\n",
       "      <td>[0.29508745000000003, 0.17668741000000002, 0.2...</td>\n",
       "      <td>[278.184218, 304.969313, 342.200118]</td>\n",
       "      <td>[0.17646052999999998, 0.030979589999999998, 0....</td>\n",
       "      <td>YQDJUXPUIRGKNV-MNOLRWOYSA-N</td>\n",
       "      <td>COC(C=CCCC=CC(=O)O)C(O)C(C)C(O)C(C)=CC(C)C(=O)...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>8</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>COC(=O)C1CC(O)C(C)(O)C=C1C(O)Cc1ccc(O)c(CCCO)c1</td>\n",
       "      <td>0.385116</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>739</th>\n",
       "      <td>NIST103438</td>\n",
       "      <td>[M+H]+</td>\n",
       "      <td>301.1071</td>\n",
       "      <td>[52.0307, 68.9972, 69.0335, 74.0152, 75.0229, ...</td>\n",
       "      <td>[0.013103103103103104, 0.7116016016016016, 0.0...</td>\n",
       "      <td>[17.00630000000001, 17.042700000000025, 18.009...</td>\n",
       "      <td>[0.06, 0.015395395395395396, 1.0, 0.1130030030...</td>\n",
       "      <td>ONIYXBSQSPCBOF-BQYQJAHWSA-N</td>\n",
       "      <td>COc1cc(O)c(C(=O)C=Cc2ccccc2O)c(OC)c1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>COc1cc(O)c(C(=O)C=Cc2ccc(O)cc2)c(OC)c1</td>\n",
       "      <td>0.937381</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>740</th>\n",
       "      <td>NIST45101</td>\n",
       "      <td>[M+H]+</td>\n",
       "      <td>393.0054</td>\n",
       "      <td>[51.0102, 51.7363, 52.018, 53.0132, 53.0258, 5...</td>\n",
       "      <td>[0.1053953953953954, 0.0381981981981982, 1.0, ...</td>\n",
       "      <td>[131.0383, 132.0251, 132.03070000000002, 132.0...</td>\n",
       "      <td>[0.0424024024024024, 0.013403403403403404, 0.0...</td>\n",
       "      <td>WHSIXKUPQCKWBY-FCKMSMMTSA-N</td>\n",
       "      <td>Nc1ncnc2c1c(I)cn2C1OC(CO)C(O)C1O</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>Nc1nc2c([O-])nc[nH]c2cc1=O</td>\n",
       "      <td>0.30091</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>741</th>\n",
       "      <td>NIST854615</td>\n",
       "      <td>[M+H]+</td>\n",
       "      <td>253.0853</td>\n",
       "      <td>[53.0022, 53.0386, 54.01, 54.0338, 55.0178, 55...</td>\n",
       "      <td>[0.1897997997997998, 0.016796796796796798, 0.0...</td>\n",
       "      <td>[17.007199999999983, 18.010899999999992, 41.02...</td>\n",
       "      <td>[0.066996996996997, 1.0, 0.0207007007007007, 0...</td>\n",
       "      <td>KKQBVQCZKQBNPO-UHFFFAOYSA-N</td>\n",
       "      <td>CNC(=O)CN(C)S(=O)(=O)CCCC(=O)O</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>CC(=O)CC(=O)N(C)S(C)(=O)=O</td>\n",
       "      <td>0.259657</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>742</th>\n",
       "      <td>NIST539936</td>\n",
       "      <td>[M-H2O+H]+</td>\n",
       "      <td>196.9960</td>\n",
       "      <td>[50.0151, 51.0229, 52.0307, 62.015, 63.0229, 6...</td>\n",
       "      <td>[0.3347047047047047, 0.8486986986986987, 0.042...</td>\n",
       "      <td>[26.0334, 27.028100000000023, 28.0314000000000...</td>\n",
       "      <td>[1.0, 0.058998998998999, 1.0, 0.02090090090090...</td>\n",
       "      <td>QLMFTKHAIDWIDC-UHFFFAOYSA-N</td>\n",
       "      <td>CCC(O)c1ccc(Br)cc1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>CC(CO)c1ccccc1</td>\n",
       "      <td>0.306383</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>743</th>\n",
       "      <td>CCMSLIB00010105357</td>\n",
       "      <td>[M+H]+</td>\n",
       "      <td>360.2170</td>\n",
       "      <td>[81.070084, 93.069992, 107.08551, 121.100868, ...</td>\n",
       "      <td>[0.013257500375739077, 0.02456798916920971, 0....</td>\n",
       "      <td>[165.15193699999998, 191.13123, 206.154668, 21...</td>\n",
       "      <td>[0.03636381998028929, 0.09352071376761444, 0.1...</td>\n",
       "      <td>PXTMHMPZOUHPQU-UHFFFAOYSA-N</td>\n",
       "      <td>COc1cc(C(O)=NCC23CC4CC(CC(C4)C2)C3)cc(OC)c1OC</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>COc1cc(CC(=O)C2CCC(OC)c3ccc(OC)cc32)cc(OC)c1C</td>\n",
       "      <td>0.303761</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>744 rows × 82 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "            spectrum_id      adduct  precursor_mz  \\\n",
       "0    CCMSLIB00004694894      [M+K]+      509.1060   \n",
       "1            NIST460218      [M+H]+      205.9461   \n",
       "2            NIST404672      [M+H]+      225.0481   \n",
       "3            NIST659328      [M+H]+      220.0209   \n",
       "4    CCMSLIB00004692719    [M+NH4]+      525.3170   \n",
       "..                  ...         ...           ...   \n",
       "739          NIST103438      [M+H]+      301.1071   \n",
       "740           NIST45101      [M+H]+      393.0054   \n",
       "741          NIST854615      [M+H]+      253.0853   \n",
       "742          NIST539936  [M-H2O+H]+      196.9960   \n",
       "743  CCMSLIB00010105357      [M+H]+      360.2170   \n",
       "\n",
       "                                                   mzs  \\\n",
       "0    [55.018578, 55.054958, 56.965504, 57.070599, 5...   \n",
       "1    [78.0336, 79.0412, 96.0439, 197.0509, 205.9456...   \n",
       "2    [51.0227, 61.007, 62.0148, 63.0226, 64.0179, 6...   \n",
       "3    [55.0288, 57.0319, 57.9743, 58.9822, 58.9947, ...   \n",
       "4    [53.039364, 55.01862, 55.054993, 57.034267, 60...   \n",
       "..                                                 ...   \n",
       "739  [52.0307, 68.9972, 69.0335, 74.0152, 75.0229, ...   \n",
       "740  [51.0102, 51.7363, 52.018, 53.0132, 53.0258, 5...   \n",
       "741  [53.0022, 53.0386, 54.01, 54.0338, 55.0178, 55...   \n",
       "742  [50.0151, 51.0229, 52.0307, 62.015, 63.0229, 6...   \n",
       "743  [81.070084, 93.069992, 107.08551, 121.100868, ...   \n",
       "\n",
       "                                           intensities  \\\n",
       "0    [0.01610015, 0.07654480999999999, 0.0660954399...   \n",
       "1    [0.2914014014014014, 1.0, 0.3114014014014014, ...   \n",
       "2    [0.058998998998999, 0.013703703703703704, 0.04...   \n",
       "3    [1.0, 0.0216016016016016, 0.0308008008008008, ...   \n",
       "4    [0.29508745000000003, 0.17668741000000002, 0.2...   \n",
       "..                                                 ...   \n",
       "739  [0.013103103103103104, 0.7116016016016016, 0.0...   \n",
       "740  [0.1053953953953954, 0.0381981981981982, 1.0, ...   \n",
       "741  [0.1897997997997998, 0.016796796796796798, 0.0...   \n",
       "742  [0.3347047047047047, 0.8486986986986987, 0.042...   \n",
       "743  [0.013257500375739077, 0.02456798916920971, 0....   \n",
       "\n",
       "                                              loss_mzs  \\\n",
       "0    [121.25959499999999, 123.254682, 130.257276, 1...   \n",
       "1             [109.90220000000001, 126.9049, 127.9125]   \n",
       "2    [26.007900000000006, 26.022899999999993, 27.01...   \n",
       "3    [28.031500000000023, 46.04220000000001, 74.037...   \n",
       "4                 [278.184218, 304.969313, 342.200118]   \n",
       "..                                                 ...   \n",
       "739  [17.00630000000001, 17.042700000000025, 18.009...   \n",
       "740  [131.0383, 132.0251, 132.03070000000002, 132.0...   \n",
       "741  [17.007199999999983, 18.010899999999992, 41.02...   \n",
       "742  [26.0334, 27.028100000000023, 28.0314000000000...   \n",
       "743  [165.15193699999998, 191.13123, 206.154668, 21...   \n",
       "\n",
       "                                      loss_intensities  \\\n",
       "0    [0.07481905, 0.01660272, 0.06307306, 0.0109070...   \n",
       "1        [0.3114014014014014, 1.0, 0.2914014014014014]   \n",
       "2    [0.0463963963963964, 0.024304304304304306, 1.0...   \n",
       "3    [0.0664964964964965, 1.0, 0.0394994994994995, ...   \n",
       "4    [0.17646052999999998, 0.030979589999999998, 0....   \n",
       "..                                                 ...   \n",
       "739  [0.06, 0.015395395395395396, 1.0, 0.1130030030...   \n",
       "740  [0.0424024024024024, 0.013403403403403404, 0.0...   \n",
       "741  [0.066996996996997, 1.0, 0.0207007007007007, 0...   \n",
       "742  [1.0, 0.058998998998999, 1.0, 0.02090090090090...   \n",
       "743  [0.03636381998028929, 0.09352071376761444, 0.1...   \n",
       "\n",
       "                        inchikey  \\\n",
       "0    AOPMSFXOYJXDNJ-IRFSQMTFSA-N   \n",
       "1    CCZWSTFVHJPCEM-UHFFFAOYSA-N   \n",
       "2    XZSIGWOVDPSPMG-UHFFFAOYSA-N   \n",
       "3    IYLGYUIQSDVGCP-UHFFFAOYSA-N   \n",
       "4    YQDJUXPUIRGKNV-MNOLRWOYSA-N   \n",
       "..                           ...   \n",
       "739  ONIYXBSQSPCBOF-BQYQJAHWSA-N   \n",
       "740  WHSIXKUPQCKWBY-FCKMSMMTSA-N   \n",
       "741  KKQBVQCZKQBNPO-UHFFFAOYSA-N   \n",
       "742  QLMFTKHAIDWIDC-UHFFFAOYSA-N   \n",
       "743  PXTMHMPZOUHPQU-UHFFFAOYSA-N   \n",
       "\n",
       "                                              val_true  num_of_sugars  ...  \\\n",
       "0    COC(=O)C1=COC(OC2OC(CO)C(O)C(O)C2O)C2C1C=CC21C...            1.0  ...   \n",
       "1                                            Ic1ccccn1            0.0  ...   \n",
       "2                            N#Cc1ccc2c(c1)Nc1ccccc1S2            0.0  ...   \n",
       "3                                 CCOC(=O)CSc1nnc(N)s1            0.0  ...   \n",
       "4    COC(C=CCCC=CC(=O)O)C(O)C(C)C(O)C(C)=CC(C)C(=O)...            0.0  ...   \n",
       "..                                                 ...            ...  ...   \n",
       "739               COc1cc(O)c(C(=O)C=Cc2ccccc2O)c(OC)c1            0.0  ...   \n",
       "740                   Nc1ncnc2c1c(I)cn2C1OC(CO)C(O)C1O            0.0  ...   \n",
       "741                     CNC(=O)CN(C)S(=O)(=O)CCCC(=O)O            0.0  ...   \n",
       "742                                 CCC(O)c1ccc(Br)cc1            0.0  ...   \n",
       "743      COc1cc(C(O)=NCC23CC4CC(CC(C4)C2)C3)cc(OC)c1OC            0.0  ...   \n",
       "\n",
       "      O   N  S  I  Br  Cl  F  P  \\\n",
       "0    12   0  0  0   0   0  0  0   \n",
       "1     0   1  0  1   0   0  0  0   \n",
       "2     0   2  1  0   0   0  0  0   \n",
       "3     2   3  2  0   0   0  0  0   \n",
       "4     8   1  0  0   0   0  0  0   \n",
       "..   ..  .. .. ..  ..  .. .. ..   \n",
       "739   5   0  0  0   0   0  0  0   \n",
       "740   4   4  0  1   0   0  0  0   \n",
       "741   5   2  1  0   0   0  0  0   \n",
       "742   1   0  0  0   1   0  0  0   \n",
       "743   4   1  0  0   0   0  0  0   \n",
       "\n",
       "                                           predicted  \\\n",
       "0      CC(C)=CC1C(O)CCc2ccc(=O)oc2C1CC(C)c1ccc(O)cc1   \n",
       "1                                O=C(Nc1cccs1)C(F)Br   \n",
       "2                  O=c1ccc(Br)c2c(-c3ccccc3)[nH]cc12   \n",
       "3                              CCOC(=O)c1nnc(SC)nc1N   \n",
       "4    COC(=O)C1CC(O)C(C)(O)C=C1C(O)Cc1ccc(O)c(CCCO)c1   \n",
       "..                                               ...   \n",
       "739           COc1cc(O)c(C(=O)C=Cc2ccc(O)cc2)c(OC)c1   \n",
       "740                       Nc1nc2c([O-])nc[nH]c2cc1=O   \n",
       "741                       CC(=O)CC(=O)N(C)S(C)(=O)=O   \n",
       "742                                   CC(CO)c1ccccc1   \n",
       "743    COc1cc(CC(=O)C2CCC(OC)c3ccc(OC)cc32)cc(OC)c1C   \n",
       "\n",
       "                                         tanimoto  \n",
       "0                                        0.383088  \n",
       "1                                        0.048338  \n",
       "2    no_prediction_or_error_with_parsing_by_rdkit  \n",
       "3                                        0.155352  \n",
       "4                                        0.385116  \n",
       "..                                            ...  \n",
       "739                                      0.937381  \n",
       "740                                       0.30091  \n",
       "741                                      0.259657  \n",
       "742                                      0.306383  \n",
       "743                                      0.303761  \n",
       "\n",
       "[744 rows x 82 columns]"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_final"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "002f3da7-8d0a-4c0a-b7bb-dcdad33c4e32",
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "#x=np.load('allclean050122_predicted.npy')\n",
    "#x=np.load('/Users/delser/mass2smiles/retrain/nist/tcn_seq/casmi_predict1.npy')\n",
    "x=np.load('/Users/delser/mass2smiles/retrain/nist/tcn_seq/val_predict1.npy')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "eee988c0-1f21-4cdf-b625-cadd68931e73",
   "metadata": {},
   "outputs": [],
   "source": [
    "def float_oupt_to_class(oupt, k):\n",
    "  end_pts = np.zeros(k+1, dtype=np.float32) \n",
    "  delta = 1.0 / k\n",
    "  for i in range(k):\n",
    "    end_pts[i] = i * delta\n",
    "  end_pts[k] = 1.0\n",
    "  # if k=4, [0.0, 0.25, 0.50, 0.75, 1.0] \n",
    "\n",
    "  for i in range(k):\n",
    "    if oupt >= end_pts[i] and oupt <= end_pts[i+1]:\n",
    "      return i\n",
    "  return -1  # fatal error \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "ae696725-eba5-4bbc-8e27-119671446dfc",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "funct=['#num_of_sugars',\"#Number of aliphatic carboxylic acids\",\n",
    " \"#Number of aliphatic hydroxyl groups\",\n",
    " \"#Number of aliphatic hydroxyl groups excluding tert-OH\",\n",
    " \"#Number of N functional groups attached to aromatics\",\n",
    " \"#Number of Aromatic carboxylic acides\",\n",
    " \"#Number of aromatic nitrogens\",\n",
    " \"#Number of aromatic amines\",\n",
    " \"#Number of aromatic hydroxyl groups\",\n",
    " \"#Number of carboxylic acids\",\n",
    " \"#Number of carboxylic acids\",\n",
    " \"#Number of carbonyl O\",\n",
    " \"#Number of carbonyl O, excluding COOH\",\n",
    " \"#Number of thiocarbonyl\",\n",
    " \"#Number of Imines\",\n",
    " \"#Number of Tertiary amines\",\n",
    " \"#Number of Secondary amines\",\n",
    " \"#Number of Primary amines\",\n",
    " \"#Number of hydroxylamine groups\",\n",
    " \"#Number of tert-alicyclic amines (no heteroatoms, not quinine-like bridged N)\",\n",
    " \"#Number of H-pyrrole nitrogens\",\n",
    " \"#Number of thiol groups\",\n",
    " \"#Number of aldehydes\",\n",
    " \"#Number of alkyl carbamates (subject to hydrolysis)\",\n",
    " \"#Number of alkyl halides\",\n",
    " \"#Number of allylic oxidation sites excluding steroid dienone\",\n",
    " \"#Number of amides\",\n",
    " \"#Number of anilines\",\n",
    " \"#Number of aryl methyl sites for hydroxylation\",\n",
    " \"#Number of azo groups\",\n",
    " \"#Number of benzene rings\",\n",
    " \"#Bicyclic\",\n",
    " \"#Number of dihydropyridines\",\n",
    " \"#Number of epoxide rings\",\n",
    " \"#Number of esters\",\n",
    " \"#Number of ether oxygens (including phenoxy)\",\n",
    " \"#Number of furan rings\",\n",
    " \"#Number of guanidine groups\",\n",
    " \"#Number of halogens\",\n",
    " \"#Number of imidazole rings\",\n",
    " \"#Number of isothiocyanates\",\n",
    " \"#Number of ketones\",\n",
    " \"#Number of ketones excluding diaryl, a,b-unsat. dienones, heteroatom on Calpha\",\n",
    " \"#Number of beta lactams\",\n",
    " \"#Number of cyclic esters (lactones)\",\n",
    " \"#Number of methoxy groups -OCH3\",\n",
    " \"#Number of nitriles\",\n",
    " \"#Number of nitro groups\",\n",
    " \"#Number of oxazole rings\",\n",
    " \"#Number of para-hydroxylation sites\",\n",
    " \"#Number of phenols\",\n",
    " \"#Number of phosphoric acid groups\",\n",
    " \"#Number of phosphoric ester groups\",\n",
    " \"#Number of piperdine rings\",\n",
    " \"#Number of primary amides\",\n",
    " \"#Number of pyridine rings\",\n",
    " \"#Number of quaternary nitrogens\",\n",
    " \"#Number of thioether\",\n",
    " \"#Number of thiazole rings\",\n",
    " \"#Number of unbranched alkanes of at least 4 members (excludes halogenated alkanes)\",\n",
    "      \"#adduct_enc\",\"#C\",\"#H\", \"#O\",\"#N\", \"#S\", \"#I\", \"#Br\",\"#Cl\", \"#F\",\"#P\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "3206a1a6-d3cb-4626-988a-b673e9f99a52",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.024481766"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x[0][0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "488b4b30-1912-4c73-9aa4-8636ab11bf3c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "float_oupt_to_class(x[0][2], 24)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "d168756e-50a3-4898-9f33-4b16129ca6a0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(744, 71)"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "all_arr=[]\n",
    "for sample in x:\n",
    "    cache=[]\n",
    "    for i in sample:\n",
    "        cache.append(float_oupt_to_class(i,65))\n",
    "    x=np.array(cache)\n",
    "    all_arr.append(x)\n",
    "    \n",
    "result=np.stack(all_arr)   \n",
    "result.shape  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "b9fd00fa-c5de-4c13-8746-37d51d6568a0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>#num_of_sugars</th>\n",
       "      <th>#Number of aliphatic carboxylic acids</th>\n",
       "      <th>#Number of aliphatic hydroxyl groups</th>\n",
       "      <th>#Number of aliphatic hydroxyl groups excluding tert-OH</th>\n",
       "      <th>#Number of N functional groups attached to aromatics</th>\n",
       "      <th>#Number of Aromatic carboxylic acides</th>\n",
       "      <th>#Number of aromatic nitrogens</th>\n",
       "      <th>#Number of aromatic amines</th>\n",
       "      <th>#Number of aromatic hydroxyl groups</th>\n",
       "      <th>#Number of carboxylic acids</th>\n",
       "      <th>...</th>\n",
       "      <th>#C</th>\n",
       "      <th>#H</th>\n",
       "      <th>#O</th>\n",
       "      <th>#N</th>\n",
       "      <th>#S</th>\n",
       "      <th>#I</th>\n",
       "      <th>#Br</th>\n",
       "      <th>#Cl</th>\n",
       "      <th>#F</th>\n",
       "      <th>#P</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>25</td>\n",
       "      <td>24</td>\n",
       "      <td>8</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>7</td>\n",
       "      <td>6</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>14</td>\n",
       "      <td>8</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>7</td>\n",
       "      <td>9</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>29</td>\n",
       "      <td>45</td>\n",
       "      <td>9</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>739</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>17</td>\n",
       "      <td>16</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>740</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>9</td>\n",
       "      <td>10</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>741</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>8</td>\n",
       "      <td>15</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>742</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>9</td>\n",
       "      <td>10</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>743</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>21</td>\n",
       "      <td>28</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>744 rows × 71 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     #num_of_sugars  #Number of aliphatic carboxylic acids  \\\n",
       "0                 0                                      0   \n",
       "1                 0                                      0   \n",
       "2                 0                                      0   \n",
       "3                 0                                      0   \n",
       "4                 0                                      1   \n",
       "..              ...                                    ...   \n",
       "739               0                                      0   \n",
       "740               0                                      0   \n",
       "741               0                                      0   \n",
       "742               0                                      0   \n",
       "743               0                                      0   \n",
       "\n",
       "     #Number of aliphatic hydroxyl groups  \\\n",
       "0                                       3   \n",
       "1                                       0   \n",
       "2                                       0   \n",
       "3                                       0   \n",
       "4                                       4   \n",
       "..                                    ...   \n",
       "739                                     0   \n",
       "740                                     0   \n",
       "741                                     0   \n",
       "742                                     0   \n",
       "743                                     0   \n",
       "\n",
       "     #Number of aliphatic hydroxyl groups excluding tert-OH  \\\n",
       "0                                                    2        \n",
       "1                                                    0        \n",
       "2                                                    0        \n",
       "3                                                    0        \n",
       "4                                                    3        \n",
       "..                                                 ...        \n",
       "739                                                  0        \n",
       "740                                                  0        \n",
       "741                                                  0        \n",
       "742                                                  0        \n",
       "743                                                  0        \n",
       "\n",
       "     #Number of N functional groups attached to aromatics  \\\n",
       "0                                                    0      \n",
       "1                                                    0      \n",
       "2                                                    0      \n",
       "3                                                    0      \n",
       "4                                                    0      \n",
       "..                                                 ...      \n",
       "739                                                  0      \n",
       "740                                                  0      \n",
       "741                                                  0      \n",
       "742                                                  0      \n",
       "743                                                  0      \n",
       "\n",
       "     #Number of Aromatic carboxylic acides  #Number of aromatic nitrogens  \\\n",
       "0                                        0                              0   \n",
       "1                                        0                              2   \n",
       "2                                        0                              2   \n",
       "3                                        0                              2   \n",
       "4                                        0                              0   \n",
       "..                                     ...                            ...   \n",
       "739                                      0                              0   \n",
       "740                                      0                              4   \n",
       "741                                      0                              0   \n",
       "742                                      0                              0   \n",
       "743                                      0                              0   \n",
       "\n",
       "     #Number of aromatic amines  #Number of aromatic hydroxyl groups  \\\n",
       "0                             0                                    2   \n",
       "1                             0                                    0   \n",
       "2                             1                                    0   \n",
       "3                             0                                    0   \n",
       "4                             0                                    0   \n",
       "..                          ...                                  ...   \n",
       "739                           0                                    1   \n",
       "740                           0                                    0   \n",
       "741                           0                                    0   \n",
       "742                           0                                    0   \n",
       "743                           0                                    0   \n",
       "\n",
       "     #Number of carboxylic acids  ...  #C  #H  #O  #N  #S  #I  #Br  #Cl  #F  \\\n",
       "0                              0  ...  25  24   8   0   0   0    0    0   0   \n",
       "1                              0  ...   7   6   1   2   0   0    0    0   0   \n",
       "2                              0  ...  14   8   1   3   0   0    0    0   0   \n",
       "3                              0  ...   7   9   2   3   2   0    0    0   0   \n",
       "4                              1  ...  29  45   9   1   0   0    0    0   0   \n",
       "..                           ...  ...  ..  ..  ..  ..  ..  ..  ...  ...  ..   \n",
       "739                            0  ...  17  16   5   0   0   0    0    0   0   \n",
       "740                            0  ...   9  10   3   4   0   0    0    0   0   \n",
       "741                            0  ...   8  15   4   2   1   0    0    0   0   \n",
       "742                            0  ...   9  10   1   0   0   0    0    0   0   \n",
       "743                            0  ...  21  28   5   0   0   0    0    0   0   \n",
       "\n",
       "     #P  \n",
       "0     0  \n",
       "1     0  \n",
       "2     0  \n",
       "3     0  \n",
       "4     0  \n",
       "..   ..  \n",
       "739   0  \n",
       "740   0  \n",
       "741   0  \n",
       "742   0  \n",
       "743   0  \n",
       "\n",
       "[744 rows x 71 columns]"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.DataFrame(result, columns = funct)\n",
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "51e0fed6-0be5-40f1-9f19-63d0cf40e060",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>spectrum_id</th>\n",
       "      <th>adduct</th>\n",
       "      <th>precursor_mz</th>\n",
       "      <th>mzs</th>\n",
       "      <th>intensities</th>\n",
       "      <th>loss_mzs</th>\n",
       "      <th>loss_intensities</th>\n",
       "      <th>inchikey</th>\n",
       "      <th>val_true</th>\n",
       "      <th>num_of_sugars</th>\n",
       "      <th>...</th>\n",
       "      <th>#C</th>\n",
       "      <th>#H</th>\n",
       "      <th>#O</th>\n",
       "      <th>#N</th>\n",
       "      <th>#S</th>\n",
       "      <th>#I</th>\n",
       "      <th>#Br</th>\n",
       "      <th>#Cl</th>\n",
       "      <th>#F</th>\n",
       "      <th>#P</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>CCMSLIB00004694894</td>\n",
       "      <td>[M+K]+</td>\n",
       "      <td>509.1060</td>\n",
       "      <td>[55.018578, 55.054958, 56.965504, 57.070599, 5...</td>\n",
       "      <td>[0.01610015, 0.07654480999999999, 0.0660954399...</td>\n",
       "      <td>[121.25959499999999, 123.254682, 130.257276, 1...</td>\n",
       "      <td>[0.07481905, 0.01660272, 0.06307306, 0.0109070...</td>\n",
       "      <td>AOPMSFXOYJXDNJ-IRFSQMTFSA-N</td>\n",
       "      <td>COC(=O)C1=COC(OC2OC(CO)C(O)C(O)C2O)C2C1C=CC21C...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>25</td>\n",
       "      <td>24</td>\n",
       "      <td>8</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>NIST460218</td>\n",
       "      <td>[M+H]+</td>\n",
       "      <td>205.9461</td>\n",
       "      <td>[78.0336, 79.0412, 96.0439, 197.0509, 205.9456...</td>\n",
       "      <td>[0.2914014014014014, 1.0, 0.3114014014014014, ...</td>\n",
       "      <td>[109.90220000000001, 126.9049, 127.9125]</td>\n",
       "      <td>[0.3114014014014014, 1.0, 0.2914014014014014]</td>\n",
       "      <td>CCZWSTFVHJPCEM-UHFFFAOYSA-N</td>\n",
       "      <td>Ic1ccccn1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>7</td>\n",
       "      <td>6</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>NIST404672</td>\n",
       "      <td>[M+H]+</td>\n",
       "      <td>225.0481</td>\n",
       "      <td>[51.0227, 61.007, 62.0148, 63.0226, 64.0179, 6...</td>\n",
       "      <td>[0.058998998998999, 0.013703703703703704, 0.04...</td>\n",
       "      <td>[26.007900000000006, 26.022899999999993, 27.01...</td>\n",
       "      <td>[0.0463963963963964, 0.024304304304304306, 1.0...</td>\n",
       "      <td>XZSIGWOVDPSPMG-UHFFFAOYSA-N</td>\n",
       "      <td>N#Cc1ccc2c(c1)Nc1ccccc1S2</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>14</td>\n",
       "      <td>8</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>NIST659328</td>\n",
       "      <td>[M+H]+</td>\n",
       "      <td>220.0209</td>\n",
       "      <td>[55.0288, 57.0319, 57.9743, 58.9822, 58.9947, ...</td>\n",
       "      <td>[1.0, 0.0216016016016016, 0.0308008008008008, ...</td>\n",
       "      <td>[28.031500000000023, 46.04220000000001, 74.037...</td>\n",
       "      <td>[0.0664964964964965, 1.0, 0.0394994994994995, ...</td>\n",
       "      <td>IYLGYUIQSDVGCP-UHFFFAOYSA-N</td>\n",
       "      <td>CCOC(=O)CSc1nnc(N)s1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>7</td>\n",
       "      <td>9</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>CCMSLIB00004692719</td>\n",
       "      <td>[M+NH4]+</td>\n",
       "      <td>525.3170</td>\n",
       "      <td>[53.039364, 55.01862, 55.054993, 57.034267, 60...</td>\n",
       "      <td>[0.29508745000000003, 0.17668741000000002, 0.2...</td>\n",
       "      <td>[278.184218, 304.969313, 342.200118]</td>\n",
       "      <td>[0.17646052999999998, 0.030979589999999998, 0....</td>\n",
       "      <td>YQDJUXPUIRGKNV-MNOLRWOYSA-N</td>\n",
       "      <td>COC(C=CCCC=CC(=O)O)C(O)C(C)C(O)C(C)=CC(C)C(=O)...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>29</td>\n",
       "      <td>45</td>\n",
       "      <td>9</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>739</th>\n",
       "      <td>NIST103438</td>\n",
       "      <td>[M+H]+</td>\n",
       "      <td>301.1071</td>\n",
       "      <td>[52.0307, 68.9972, 69.0335, 74.0152, 75.0229, ...</td>\n",
       "      <td>[0.013103103103103104, 0.7116016016016016, 0.0...</td>\n",
       "      <td>[17.00630000000001, 17.042700000000025, 18.009...</td>\n",
       "      <td>[0.06, 0.015395395395395396, 1.0, 0.1130030030...</td>\n",
       "      <td>ONIYXBSQSPCBOF-BQYQJAHWSA-N</td>\n",
       "      <td>COc1cc(O)c(C(=O)C=Cc2ccccc2O)c(OC)c1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>17</td>\n",
       "      <td>16</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>740</th>\n",
       "      <td>NIST45101</td>\n",
       "      <td>[M+H]+</td>\n",
       "      <td>393.0054</td>\n",
       "      <td>[51.0102, 51.7363, 52.018, 53.0132, 53.0258, 5...</td>\n",
       "      <td>[0.1053953953953954, 0.0381981981981982, 1.0, ...</td>\n",
       "      <td>[131.0383, 132.0251, 132.03070000000002, 132.0...</td>\n",
       "      <td>[0.0424024024024024, 0.013403403403403404, 0.0...</td>\n",
       "      <td>WHSIXKUPQCKWBY-FCKMSMMTSA-N</td>\n",
       "      <td>Nc1ncnc2c1c(I)cn2C1OC(CO)C(O)C1O</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>9</td>\n",
       "      <td>10</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>741</th>\n",
       "      <td>NIST854615</td>\n",
       "      <td>[M+H]+</td>\n",
       "      <td>253.0853</td>\n",
       "      <td>[53.0022, 53.0386, 54.01, 54.0338, 55.0178, 55...</td>\n",
       "      <td>[0.1897997997997998, 0.016796796796796798, 0.0...</td>\n",
       "      <td>[17.007199999999983, 18.010899999999992, 41.02...</td>\n",
       "      <td>[0.066996996996997, 1.0, 0.0207007007007007, 0...</td>\n",
       "      <td>KKQBVQCZKQBNPO-UHFFFAOYSA-N</td>\n",
       "      <td>CNC(=O)CN(C)S(=O)(=O)CCCC(=O)O</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>8</td>\n",
       "      <td>15</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>742</th>\n",
       "      <td>NIST539936</td>\n",
       "      <td>[M-H2O+H]+</td>\n",
       "      <td>196.9960</td>\n",
       "      <td>[50.0151, 51.0229, 52.0307, 62.015, 63.0229, 6...</td>\n",
       "      <td>[0.3347047047047047, 0.8486986986986987, 0.042...</td>\n",
       "      <td>[26.0334, 27.028100000000023, 28.0314000000000...</td>\n",
       "      <td>[1.0, 0.058998998998999, 1.0, 0.02090090090090...</td>\n",
       "      <td>QLMFTKHAIDWIDC-UHFFFAOYSA-N</td>\n",
       "      <td>CCC(O)c1ccc(Br)cc1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>9</td>\n",
       "      <td>10</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>743</th>\n",
       "      <td>CCMSLIB00010105357</td>\n",
       "      <td>[M+H]+</td>\n",
       "      <td>360.2170</td>\n",
       "      <td>[81.070084, 93.069992, 107.08551, 121.100868, ...</td>\n",
       "      <td>[0.013257500375739077, 0.02456798916920971, 0....</td>\n",
       "      <td>[165.15193699999998, 191.13123, 206.154668, 21...</td>\n",
       "      <td>[0.03636381998028929, 0.09352071376761444, 0.1...</td>\n",
       "      <td>PXTMHMPZOUHPQU-UHFFFAOYSA-N</td>\n",
       "      <td>COc1cc(C(O)=NCC23CC4CC(CC(C4)C2)C3)cc(OC)c1OC</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>21</td>\n",
       "      <td>28</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>744 rows × 153 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "            spectrum_id      adduct  precursor_mz  \\\n",
       "0    CCMSLIB00004694894      [M+K]+      509.1060   \n",
       "1            NIST460218      [M+H]+      205.9461   \n",
       "2            NIST404672      [M+H]+      225.0481   \n",
       "3            NIST659328      [M+H]+      220.0209   \n",
       "4    CCMSLIB00004692719    [M+NH4]+      525.3170   \n",
       "..                  ...         ...           ...   \n",
       "739          NIST103438      [M+H]+      301.1071   \n",
       "740           NIST45101      [M+H]+      393.0054   \n",
       "741          NIST854615      [M+H]+      253.0853   \n",
       "742          NIST539936  [M-H2O+H]+      196.9960   \n",
       "743  CCMSLIB00010105357      [M+H]+      360.2170   \n",
       "\n",
       "                                                   mzs  \\\n",
       "0    [55.018578, 55.054958, 56.965504, 57.070599, 5...   \n",
       "1    [78.0336, 79.0412, 96.0439, 197.0509, 205.9456...   \n",
       "2    [51.0227, 61.007, 62.0148, 63.0226, 64.0179, 6...   \n",
       "3    [55.0288, 57.0319, 57.9743, 58.9822, 58.9947, ...   \n",
       "4    [53.039364, 55.01862, 55.054993, 57.034267, 60...   \n",
       "..                                                 ...   \n",
       "739  [52.0307, 68.9972, 69.0335, 74.0152, 75.0229, ...   \n",
       "740  [51.0102, 51.7363, 52.018, 53.0132, 53.0258, 5...   \n",
       "741  [53.0022, 53.0386, 54.01, 54.0338, 55.0178, 55...   \n",
       "742  [50.0151, 51.0229, 52.0307, 62.015, 63.0229, 6...   \n",
       "743  [81.070084, 93.069992, 107.08551, 121.100868, ...   \n",
       "\n",
       "                                           intensities  \\\n",
       "0    [0.01610015, 0.07654480999999999, 0.0660954399...   \n",
       "1    [0.2914014014014014, 1.0, 0.3114014014014014, ...   \n",
       "2    [0.058998998998999, 0.013703703703703704, 0.04...   \n",
       "3    [1.0, 0.0216016016016016, 0.0308008008008008, ...   \n",
       "4    [0.29508745000000003, 0.17668741000000002, 0.2...   \n",
       "..                                                 ...   \n",
       "739  [0.013103103103103104, 0.7116016016016016, 0.0...   \n",
       "740  [0.1053953953953954, 0.0381981981981982, 1.0, ...   \n",
       "741  [0.1897997997997998, 0.016796796796796798, 0.0...   \n",
       "742  [0.3347047047047047, 0.8486986986986987, 0.042...   \n",
       "743  [0.013257500375739077, 0.02456798916920971, 0....   \n",
       "\n",
       "                                              loss_mzs  \\\n",
       "0    [121.25959499999999, 123.254682, 130.257276, 1...   \n",
       "1             [109.90220000000001, 126.9049, 127.9125]   \n",
       "2    [26.007900000000006, 26.022899999999993, 27.01...   \n",
       "3    [28.031500000000023, 46.04220000000001, 74.037...   \n",
       "4                 [278.184218, 304.969313, 342.200118]   \n",
       "..                                                 ...   \n",
       "739  [17.00630000000001, 17.042700000000025, 18.009...   \n",
       "740  [131.0383, 132.0251, 132.03070000000002, 132.0...   \n",
       "741  [17.007199999999983, 18.010899999999992, 41.02...   \n",
       "742  [26.0334, 27.028100000000023, 28.0314000000000...   \n",
       "743  [165.15193699999998, 191.13123, 206.154668, 21...   \n",
       "\n",
       "                                      loss_intensities  \\\n",
       "0    [0.07481905, 0.01660272, 0.06307306, 0.0109070...   \n",
       "1        [0.3114014014014014, 1.0, 0.2914014014014014]   \n",
       "2    [0.0463963963963964, 0.024304304304304306, 1.0...   \n",
       "3    [0.0664964964964965, 1.0, 0.0394994994994995, ...   \n",
       "4    [0.17646052999999998, 0.030979589999999998, 0....   \n",
       "..                                                 ...   \n",
       "739  [0.06, 0.015395395395395396, 1.0, 0.1130030030...   \n",
       "740  [0.0424024024024024, 0.013403403403403404, 0.0...   \n",
       "741  [0.066996996996997, 1.0, 0.0207007007007007, 0...   \n",
       "742  [1.0, 0.058998998998999, 1.0, 0.02090090090090...   \n",
       "743  [0.03636381998028929, 0.09352071376761444, 0.1...   \n",
       "\n",
       "                        inchikey  \\\n",
       "0    AOPMSFXOYJXDNJ-IRFSQMTFSA-N   \n",
       "1    CCZWSTFVHJPCEM-UHFFFAOYSA-N   \n",
       "2    XZSIGWOVDPSPMG-UHFFFAOYSA-N   \n",
       "3    IYLGYUIQSDVGCP-UHFFFAOYSA-N   \n",
       "4    YQDJUXPUIRGKNV-MNOLRWOYSA-N   \n",
       "..                           ...   \n",
       "739  ONIYXBSQSPCBOF-BQYQJAHWSA-N   \n",
       "740  WHSIXKUPQCKWBY-FCKMSMMTSA-N   \n",
       "741  KKQBVQCZKQBNPO-UHFFFAOYSA-N   \n",
       "742  QLMFTKHAIDWIDC-UHFFFAOYSA-N   \n",
       "743  PXTMHMPZOUHPQU-UHFFFAOYSA-N   \n",
       "\n",
       "                                              val_true  num_of_sugars  ...  \\\n",
       "0    COC(=O)C1=COC(OC2OC(CO)C(O)C(O)C2O)C2C1C=CC21C...            1.0  ...   \n",
       "1                                            Ic1ccccn1            0.0  ...   \n",
       "2                            N#Cc1ccc2c(c1)Nc1ccccc1S2            0.0  ...   \n",
       "3                                 CCOC(=O)CSc1nnc(N)s1            0.0  ...   \n",
       "4    COC(C=CCCC=CC(=O)O)C(O)C(C)C(O)C(C)=CC(C)C(=O)...            0.0  ...   \n",
       "..                                                 ...            ...  ...   \n",
       "739               COc1cc(O)c(C(=O)C=Cc2ccccc2O)c(OC)c1            0.0  ...   \n",
       "740                   Nc1ncnc2c1c(I)cn2C1OC(CO)C(O)C1O            0.0  ...   \n",
       "741                     CNC(=O)CN(C)S(=O)(=O)CCCC(=O)O            0.0  ...   \n",
       "742                                 CCC(O)c1ccc(Br)cc1            0.0  ...   \n",
       "743      COc1cc(C(O)=NCC23CC4CC(CC(C4)C2)C3)cc(OC)c1OC            0.0  ...   \n",
       "\n",
       "     #C  #H  #O  #N  #S  #I  #Br  #Cl  #F  #P  \n",
       "0    25  24   8   0   0   0    0    0   0   0  \n",
       "1     7   6   1   2   0   0    0    0   0   0  \n",
       "2    14   8   1   3   0   0    0    0   0   0  \n",
       "3     7   9   2   3   2   0    0    0   0   0  \n",
       "4    29  45   9   1   0   0    0    0   0   0  \n",
       "..   ..  ..  ..  ..  ..  ..  ...  ...  ..  ..  \n",
       "739  17  16   5   0   0   0    0    0   0   0  \n",
       "740   9  10   3   4   0   0    0    0   0   0  \n",
       "741   8  15   4   2   1   0    0    0   0   0  \n",
       "742   9  10   1   0   0   0    0    0   0   0  \n",
       "743  21  28   5   0   0   0    0    0   0   0  \n",
       "\n",
       "[744 rows x 153 columns]"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_final = df_final.join(df, how=\"outer\")\n",
    "#df_final=df_final.drop([\"Column 3\"], axis=1)\n",
    "#df_final=df_final.drop([\"Adduct\",\"spec\",\"smiles\"], axis=1)\n",
    "df_final"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "892b3a38-d101-4a64-88b7-535e4942178a",
   "metadata": {},
   "outputs": [],
   "source": [
    "#df_final.to_csv(\"first-predictions-validation_050122.tsv\",sep='\\t')\n",
    "#df_final.to_csv(\"/Users/delser/ownCloud/Writing/casmi_tanimoto_0707022.tsv\",sep='\\t')\n",
    "df_final.to_csv(\"/Users/delser/mass2smiles/retrain/predicted_results_casmi16082022_tanimoto.tsv\",sep='\\t')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "799eb9fd-e047-453e-bdb3-c4d622483fa1",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "f62db435-8e6e-4163-8b55-472656363292",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_final.to_csv(\"/Users/delser/ownCloud/mass2smiles_paper/predicted_results_casmi28102022_tanimoto.tsv\",sep='\\t')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "cdc3ce7e-60de-420b-8758-5e9ca3aed8e3",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_final.to_csv(\"/Users/delser/mass2smiles/retrain/nist/spconv/casmi_result_tanimoto_14112022.tsv\",sep='\\t')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "614633c8-a667-4ccd-8a20-92f452efcba2",
   "metadata": {},
   "outputs": [],
   "source": [
    "'/Users/delser/mass2smiles/retrain/nist/tcn_seq/12122022_casmi_result.tsv'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "47ffffff-ff19-4cf6-bdb8-70bc857c31c9",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_final.to_csv('/Users/delser/mass2smiles/retrain/nist/tcn_seq/val_result_5l_50ep_icy_lion_205_tanimoto_mf.tsv',sep='\\t')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "bf170144-73ef-40e6-8de9-87894989cdee",
   "metadata": {},
   "outputs": [],
   "source": [
    "09022023_val_result_4l_50ep_twilight-durian-197.tsv"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}