tutorials/delta_vs_direct.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# DelftaCalculator basics and $\\Delta$-learning vs. direct learning "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import glob\n",
    "from openbabel.pybel import readfile, readstring\n",
    "import numpy as np\n",
    "from delfta.utils import TESTS_PATH, DATA_PATH"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Next, we read a few example molecules from SDFs (structure data files) from the QMugs dataset. For more details on how to input your own molecules, see the documentation of the Pybel API [here](https://open-babel.readthedocs.io/en/latest/UseTheLibrary/Python_PybelAPI.html)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "if not os.path.exists(os.path.join(TESTS_PATH)): \n",
    "    from delfta.download import TESTS_REMOTE, download\n",
    "    import tarfile \n",
    "    \n",
    "    download(TESTS_REMOTE, os.path.join(DATA_PATH, \"test_data.tar.gz\"))\n",
    "    with tarfile.open(os.path.join(DATA_PATH, \"test_data.tar.gz\")) as handle:\n",
    "        handle.extractall(DATA_PATH)\n",
    "\n",
    "mol_files = glob.glob(os.path.join(TESTS_PATH, \"mols_working\", \"CHEMBL*.sdf\"))\n",
    "mols = [next(readfile(\"sdf\", mol_file)) for mol_file in mol_files]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We'll initialize a calculator by stating which tasks we want to perform. Options are `all` (the default), `E_form`, `E_homo`, `E_lumo`, `E_gap`, `dipole`, and `charges`. You can also choose between `delta=True` (run a GNF2-xTB calculation and predict a delta correction to obtain an approximation of the DFT-value ($\\omega$B97X-D/def2-SVP)) or `delta=False` (predict the DFT-value directly from the molecular structure)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2021/12/08 06:17:05 PM | DelFTa | INFO: Now running xTB...\n",
      "100%|██████████| 100/100 [00:28<00:00,  3.53it/s]\n",
      "2021/12/08 06:17:33 PM | DelFTa | INFO: Now running network for model single_energy_delta...\n",
      "100%|██████████| 100/100 [00:00<00:00, 131.78it/s]\n",
      "2021/12/08 06:17:35 PM | DelFTa | INFO: Now running network for model multitask_delta...\n",
      "100%|██████████| 100/100 [00:00<00:00, 157.75it/s]\n",
      "2021/12/08 06:17:36 PM | DelFTa | INFO: Now running network for model charges_delta...\n",
      "100%|██████████| 100/100 [00:00<00:00, 158.15it/s]\n",
      "2021/12/08 06:17:37 PM | DelFTa | INFO: Now running network for model charges_direct...\n",
      "100%|██████████| 100/100 [00:00<00:00, 160.94it/s]\n",
      "2021/12/08 06:17:37 PM | DelFTa | INFO: Now running network for model single_energy_direct...\n",
      "100%|██████████| 100/100 [00:00<00:00, 156.61it/s]\n",
      "2021/12/08 06:17:38 PM | DelFTa | INFO: Now running network for model multitask_direct...\n",
      "100%|██████████| 100/100 [00:00<00:00, 142.11it/s]\n"
     ]
    }
   ],
   "source": [
    "from delfta.calculator import DelftaCalculator\n",
    "\n",
    "calc_delta = DelftaCalculator(tasks=[\"E_form\", \"E_homo\", \"E_lumo\", \"E_gap\", \"dipole\", \"charges\"], delta=True) \n",
    "# We'll skip Wiberg bond orders in this tutorial as this requires some more verbose handling of atom/bond ids later on \n",
    "predictions_delta = calc_delta.predict(mols, batch_size=32) # lower the batch size if you run into memory issues\n",
    "calc_direct = DelftaCalculator(tasks=[\"E_form\", \"E_homo\", \"E_lumo\", \"E_gap\", \"dipole\", \"charges\"], delta=False) \n",
    "predictions_direct = calc_direct.predict(mols)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's put the output in a dataframe to get an overview: "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "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>structure</th>\n",
       "      <th>E_form [Ha]</th>\n",
       "      <th>E_homo [Ha]</th>\n",
       "      <th>E_lumo [Ha]</th>\n",
       "      <th>E_gap [Ha]</th>\n",
       "      <th>dipole [D]</th>\n",
       "      <th>charges [e]</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>CHEMBL225201_conf_02.sdf</td>\n",
       "      <td>-8.813721</td>\n",
       "      <td>-0.274999</td>\n",
       "      <td>0.005881</td>\n",
       "      <td>0.280759</td>\n",
       "      <td>7.328785</td>\n",
       "      <td>[-0.030875619599075324, 0.005195767134704592, ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>CHEMBL3917059_conf_01.sdf</td>\n",
       "      <td>-10.935448</td>\n",
       "      <td>-0.273494</td>\n",
       "      <td>0.009785</td>\n",
       "      <td>0.283435</td>\n",
       "      <td>10.684330</td>\n",
       "      <td>[0.054004443623275764, -0.1176872677693367, -0...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>CHEMBL498248_conf_02.sdf</td>\n",
       "      <td>-7.160987</td>\n",
       "      <td>-0.310790</td>\n",
       "      <td>-0.055976</td>\n",
       "      <td>0.254909</td>\n",
       "      <td>2.619457</td>\n",
       "      <td>[0.16150189902526857, -0.4162774630255127, 0.1...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>CHEMBL2178189_conf_01.sdf</td>\n",
       "      <td>-10.881596</td>\n",
       "      <td>-0.279899</td>\n",
       "      <td>0.031184</td>\n",
       "      <td>0.311134</td>\n",
       "      <td>8.439722</td>\n",
       "      <td>[0.13710493664514542, -0.3518399607843399, 0.2...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>CHEMBL1702692_conf_00.sdf</td>\n",
       "      <td>-7.063443</td>\n",
       "      <td>-0.278748</td>\n",
       "      <td>-0.029215</td>\n",
       "      <td>0.249319</td>\n",
       "      <td>6.440857</td>\n",
       "      <td>[0.13054176561258315, -0.33679879455505374, 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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>95</th>\n",
       "      <td>CHEMBL549536_conf_00.sdf</td>\n",
       "      <td>-6.982763</td>\n",
       "      <td>-0.284059</td>\n",
       "      <td>-0.000596</td>\n",
       "      <td>0.283686</td>\n",
       "      <td>7.779823</td>\n",
       "      <td>[0.0775152101145935, -0.0325183975387001, -0.1...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>96</th>\n",
       "      <td>CHEMBL3911894_conf_02.sdf</td>\n",
       "      <td>-8.601434</td>\n",
       "      <td>-0.297836</td>\n",
       "      <td>0.008123</td>\n",
       "      <td>0.305994</td>\n",
       "      <td>1.267288</td>\n",
       "      <td>[-0.13680183943115232, 0.20445370790084838, -0...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>97</th>\n",
       "      <td>CHEMBL3310704_conf_00.sdf</td>\n",
       "      <td>-9.649964</td>\n",
       "      <td>-0.272366</td>\n",
       "      <td>0.040883</td>\n",
       "      <td>0.313243</td>\n",
       "      <td>4.099394</td>\n",
       "      <td>[0.14251945604553223, -0.3266561704756737, 0.2...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>98</th>\n",
       "      <td>CHEMBL1731586_conf_00.sdf</td>\n",
       "      <td>-9.637108</td>\n",
       "      <td>-0.251576</td>\n",
       "      <td>0.027726</td>\n",
       "      <td>0.279228</td>\n",
       "      <td>2.892029</td>\n",
       "      <td>[0.09452940721343994, -0.38360732514808654, 0....</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99</th>\n",
       "      <td>CHEMBL3679736_conf_00.sdf</td>\n",
       "      <td>-11.973742</td>\n",
       "      <td>-0.279266</td>\n",
       "      <td>-0.008116</td>\n",
       "      <td>0.271179</td>\n",
       "      <td>3.807105</td>\n",
       "      <td>[-0.08393573112254142, 0.12980520748504637, -0...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>100 rows × 7 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                    structure  E_form [Ha]  E_homo [Ha]  E_lumo [Ha]  \\\n",
       "0    CHEMBL225201_conf_02.sdf    -8.813721    -0.274999     0.005881   \n",
       "1   CHEMBL3917059_conf_01.sdf   -10.935448    -0.273494     0.009785   \n",
       "2    CHEMBL498248_conf_02.sdf    -7.160987    -0.310790    -0.055976   \n",
       "3   CHEMBL2178189_conf_01.sdf   -10.881596    -0.279899     0.031184   \n",
       "4   CHEMBL1702692_conf_00.sdf    -7.063443    -0.278748    -0.029215   \n",
       "..                        ...          ...          ...          ...   \n",
       "95   CHEMBL549536_conf_00.sdf    -6.982763    -0.284059    -0.000596   \n",
       "96  CHEMBL3911894_conf_02.sdf    -8.601434    -0.297836     0.008123   \n",
       "97  CHEMBL3310704_conf_00.sdf    -9.649964    -0.272366     0.040883   \n",
       "98  CHEMBL1731586_conf_00.sdf    -9.637108    -0.251576     0.027726   \n",
       "99  CHEMBL3679736_conf_00.sdf   -11.973742    -0.279266    -0.008116   \n",
       "\n",
       "    E_gap [Ha]  dipole [D]                                        charges [e]  \n",
       "0     0.280759    7.328785  [-0.030875619599075324, 0.005195767134704592, ...  \n",
       "1     0.283435   10.684330  [0.054004443623275764, -0.1176872677693367, -0...  \n",
       "2     0.254909    2.619457  [0.16150189902526857, -0.4162774630255127, 0.1...  \n",
       "3     0.311134    8.439722  [0.13710493664514542, -0.3518399607843399, 0.2...  \n",
       "4     0.249319    6.440857  [0.13054176561258315, -0.33679879455505374, 0....  \n",
       "..         ...         ...                                                ...  \n",
       "95    0.283686    7.779823  [0.0775152101145935, -0.0325183975387001, -0.1...  \n",
       "96    0.305994    1.267288  [-0.13680183943115232, 0.20445370790084838, -0...  \n",
       "97    0.313243    4.099394  [0.14251945604553223, -0.3266561704756737, 0.2...  \n",
       "98    0.279228    2.892029  [0.09452940721343994, -0.38360732514808654, 0....  \n",
       "99    0.271179    3.807105  [-0.08393573112254142, 0.12980520748504637, -0...  \n",
       "\n",
       "[100 rows x 7 columns]"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import pandas as pd\n",
    "\n",
    "df = pd.DataFrame(predictions_delta)\n",
    "df.insert(0, \"structure\", [os.path.basename(mol_file) for mol_file in mol_files])\n",
    "df.rename(columns={\"E_form\": \"E_form [Ha]\", \"E_homo\": \"E_homo [Ha]\", \"E_lumo\": \"E_lumo [Ha]\", \"E_gap\": \"E_gap [Ha]\", \"dipole\": \"dipole [D]\", \"charges\": \"charges [e]\"})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Get the DFT-calculated values from the QMugs SDFs and prepare everything for plotting:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "predictions_delta[\"charges\"] = np.concatenate(predictions_delta[\"charges\"])\n",
    "predictions_direct[\"charges\"] = np.concatenate(predictions_direct[\"charges\"])\n",
    "dft_keys = [\"DFT:FORMATION_ENERGY\", \"DFT:HOMO_ENERGY\", \"DFT:LUMO_ENERGY\", \"DFT:HOMO_LUMO_GAP\", \"DFT:DIPOLE\", \"DFT:MULLIKEN_CHARGES\"]\n",
    "dft_values = {}\n",
    "for dft_key in dft_keys:\n",
    "    if dft_key == \"DFT:DIPOLE\": \n",
    "        dft_values[dft_key] =  [float(mol.data[dft_key].split(\"|\")[-1]) for mol in mols]\n",
    "    elif dft_key == \"DFT:MULLIKEN_CHARGES\": \n",
    "        dft_values[dft_key] = [float(elem) for mol in mols for elem in mol.data[dft_key].split(\"|\")]\n",
    "    else:\n",
    "        dft_values[dft_key] = [float(mol.data[dft_key]) for mol in mols]\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2021/12/08 06:17:39 PM | matplotlib.font_manager | INFO: Generating new fontManager, this may take some time...\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7fddbd0898d0>"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1440x864 with 6 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "from matplotlib.gridspec import GridSpec\n",
    "from matplotlib.lines import Line2D\n",
    "\n",
    "def make_plot(ax, dft_vals, pred_vals, color): \n",
    "    ax.scatter(dft_vals, pred_vals, color=color, s=2)\n",
    "    ax.set_xlabel(\"DFT calculated\")\n",
    "    ax.set_ylabel(\"Predicted\")\n",
    "    min_val = [min(ax.get_ylim()[0], ax.get_xlim()[0])]\n",
    "    max_val = [max(ax.get_ylim()[1], ax.get_xlim()[1])]\n",
    "    ax.plot([min_val, max_val], [min_val, max_val], '--', color=\"black\", zorder=-1)\n",
    "    ax.set_aspect(\"equal\", \"box\")\n",
    "\n",
    "fig = plt.figure(figsize=(20, 12))\n",
    "gs = GridSpec(nrows=2, ncols=3)\n",
    "gs.update(wspace=0.3, hspace=0.2)\n",
    "pred_keys = [\"E_form\", \"E_homo\", \"E_lumo\", \"E_gap\", \"dipole\", \"charges\"]\n",
    "all_dft_vals = [dft_values[dft_key] for dft_key in dft_keys]\n",
    "all_pred_vals_delta = [predictions_delta[pred_key] for pred_key in pred_keys]\n",
    "all_pred_vals_direct = [predictions_direct[pred_key] for pred_key in pred_keys]\n",
    "units = [\"Ha\", \"Ha\", \"Ha\", \"Ha\", \"D\", \"e\"]\n",
    "axes = [fig.add_subplot(gs[i,j]) for i in range(2) for j in range(3)]\n",
    "for ax, pred_key, dft_key, unit in zip(axes, pred_keys, dft_keys, units): \n",
    "    make_plot(ax, dft_values[dft_key], predictions_direct[pred_key], \"orange\")\n",
    "    make_plot(ax, dft_values[dft_key], predictions_delta[pred_key], \"blue\")\n",
    "    ax.title.set_text(f\"{pred_key} / {unit}\")\n",
    "    ax.title.set_fontsize(16)\n",
    "\n",
    "handles = [Line2D([0], [0], marker='o', color=\"white\", markerfacecolor=\"blue\"), Line2D([0], [0], marker='o', color=\"white\", markerfacecolor=\"orange\")]\n",
    "fig.legend(handles, [\"delta\", \"direct\"], fontsize=16, loc=\"lower center\", ncol=2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As you can see, both $\\Delta$- and direct-learning approximate the $\\omega$B97X-D/def2-SVP ground truth well. Check our paper for a more detailed discussion."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "You can also create molecules directly from SMILES. In that case, you need to keep `force3D=True` (the default), so that the Pybel generates a three-dimensional structure for the molecule using the MMFF94 force field. Alternatively, you can create three-dimensional coordinates in some other way and feed the molecule to the calculator afterwards."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2021/12/08 06:17:41 PM | DelFTa | INFO: Assigned MMFF94 coordinates and added hydrogens to molecules at position(s) [0]\n",
      "2021/12/08 06:17:41 PM | DelFTa | INFO: Now running xTB...\n",
      "100%|██████████| 1/1 [00:00<00:00, 11.33it/s]\n",
      "2021/12/08 06:17:41 PM | DelFTa | INFO: Now running network for model single_energy_delta...\n",
      "100%|██████████| 1/1 [00:00<00:00, 122.50it/s]\n",
      "2021/12/08 06:17:41 PM | DelFTa | INFO: Now running network for model multitask_delta...\n",
      "100%|██████████| 1/1 [00:00<00:00, 128.79it/s]\n",
      "2021/12/08 06:17:41 PM | DelFTa | INFO: Now running network for model charges_delta...\n",
      "100%|██████████| 1/1 [00:00<00:00, 65.51it/s]\n",
      "2021/12/08 06:17:41 PM | DelFTa | INFO: Now running network for model wbo_delta...\n",
      "100%|██████████| 1/1 [00:00<00:00, 60.47it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'E_form': array([-3.8762815], dtype=float32), 'E_homo': array([-0.34509787], dtype=float32), 'E_lumo': array([-0.01114559], dtype=float32), 'E_gap': array([0.33391201], dtype=float32), 'dipole': array([3.698949], dtype=float32), 'charges': [array([-0.19695041,  0.18014771, -0.09179275, -0.2915747 ,  0.20729421,\n",
      "       -0.04023171,  0.01246643, -0.00152081, -0.01100436, -0.10213734,\n",
      "        0.19624388, -0.2294366 , -0.19212152,  0.06875062,  0.0843718 ,\n",
      "        0.07212227,  0.03232232,  0.03542026,  0.02956192,  0.0426783 ,\n",
      "        0.21608485])], 'wbo': [{'0-1': 2.2922393119399898, '0-3': 0.15052526246164533, '1-2': 1.0618275224391063, '1-3': 1.2581694726330652, '2-13': 0.9273212582222881, '2-14': 0.9442758618455975, '2-15': 0.9486681854011312, '3-4': 1.17273141301408, '4-5': 1.41910056634316, '4-9': 1.365871508336803, '5-6': 1.470211942518288, '5-8': 0.11313759159555414, '5-16': 0.926460847769334, '6-7': 1.4510870471804158, '6-17': 0.9309598911554917, '7-8': 1.4754809913379463, '7-18': 0.931736700349708, '8-9': 1.3943825856676701, '8-19': 0.9220687038701757, '9-10': 1.0401675536132569, '10-11': 2.191835577121868, '10-12': 1.3611355531073726, '11-12': 0.16731167503427397, '12-20': 1.0677662324114374}]}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "mols = [readstring(\"smi\", \"O=C(C)Oc1ccccc1C(=O)O\")] # aspirin\n",
    "calc = DelftaCalculator(delta=True, force3d=True)\n",
    "preds_delta = calc.predict(mols)\n",
    "print(preds_delta)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "interpreter": {
   "hash": "d1a309c350f9312609ebcb7a37fefc7494367ccc4e0da436af3d332477491a2a"
  },
  "kernelspec": {
   "display_name": "Python 3.8.11 64-bit ('delfta': conda)",
   "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": "3.7.4"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}