ipyparallel_example.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "3cae514e",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[21:10:28] Initializing Normalizer\n",
      "[21:10:28] Explicit valence for atom # 8 Sb, 9, is greater than permitted\n",
      "[21:10:28] ERROR: Could not sanitize molecule on line 1825\n",
      "[21:10:28] ERROR: Explicit valence for atom # 8 Sb, 9, is greater than permitted\n",
      "[21:10:28] Explicit valence for atom # 9 N, 5, is greater than permitted\n",
      "[21:10:28] ERROR: Could not sanitize molecule on line 2097\n",
      "[21:10:28] ERROR: Explicit valence for atom # 9 N, 5, is greater than permitted\n",
      "[21:10:28] Explicit valence for atom # 4 Al, 9, is greater than permitted\n",
      "[21:10:28] ERROR: Could not sanitize molecule on line 3226\n",
      "[21:10:28] ERROR: Explicit valence for atom # 4 Al, 9, is greater than permitted\n",
      "[21:10:28] Explicit valence for atom # 4 C, 5, is greater than permitted\n",
      "[21:10:28] ERROR: Could not sanitize molecule on line 3399\n",
      "[21:10:28] ERROR: Explicit valence for atom # 4 C, 5, is greater than permitted\n",
      "[21:10:28] Explicit valence for atom # 2 O, 4, is greater than permitted\n",
      "[21:10:28] ERROR: Could not sanitize molecule on line 4508\n",
      "[21:10:28] ERROR: Explicit valence for atom # 2 O, 4, is greater than permitted\n",
      "[21:10:28] Explicit valence for atom # 3 Be, 4, is greater than permitted\n",
      "[21:10:28] ERROR: Could not sanitize molecule on line 4596\n",
      "[21:10:28] ERROR: Explicit valence for atom # 3 Be, 4, is greater than permitted\n"
     ]
    }
   ],
   "source": [
    "from rdkit import Chem\n",
    "from rdkit.Chem import AllChem\n",
    "from rdkit.Chem.Draw import IPythonConsole\n",
    "from chembl_structure_pipeline import standardizer\n",
    "from rdkit import RDLogger\n",
    "import time\n",
    "RDLogger.DisableLog('rdApp.info')\n",
    "suppl = Chem.SmilesMolSupplier('./first_5K.smi', smilesColumn=0, delimiter='\\t')\n",
    "mols = [m for m in suppl if m!=None]\n",
    "mols = mols[:1000]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "74b598d5",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "def stdmol(mol):\n",
    "    stdmol = standardizer.standardize_mol(mol)\n",
    "    pmol, _ = standardizer.get_parent_mol(stdmol)\n",
    "    return pmol\n",
    "mols = [stdmol(m) for m in mols]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "f750f1e4",
   "metadata": {},
   "outputs": [],
   "source": [
    "def confgen(tpl, mols=mols):\n",
    "    from rdkit import Chem\n",
    "    from rdkit.Chem import AllChem     \n",
    "    nengine, clsidx = tpl\n",
    "    nwrite = 0\n",
    "    for idx, mol in enumerate(mols):\n",
    "        if idx % nengine == clsidx:\n",
    "            try:\n",
    "                hmol = Chem.AddHs(mol)\n",
    "                res = AllChem.EmbedMolecule(hmol)\n",
    "                if res == 0:\n",
    "                    AllChem.MMFFOptimizeMolecule(hmol)\n",
    "                    nwrite += 1\n",
    "            except:\n",
    "                continue\n",
    "    return nwrite"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "525afae9",
   "metadata": {},
   "outputs": [],
   "source": [
    "import ipyparallel as ipp"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "470611a5",
   "metadata": {},
   "outputs": [],
   "source": [
    "def runtask(nengines=1):\n",
    "    start = time.time()\n",
    "    with ipp.Cluster(n=nengines) as rc:\n",
    "        view = rc.load_balanced_view()\n",
    "        asyncresult = view.map_async(confgen, [(nengines, idx) for idx in range(nengines)])\n",
    "        asyncresult.wait_interactive()\n",
    "        # retrieve actual results\n",
    "        result = asyncresult.get()\n",
    "        print(result)\n",
    "    finish = time.time()\n",
    "    print(finish - start)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "11e016c5",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Starting 1 engines with <class 'ipyparallel.cluster.launcher.LocalEngineSetLauncher'>\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "53bc35b5a1364998b4749afac3a3a0d1",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?engine/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "24386b2a4ea047be88cae75313cf5a8e",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "confgen:   0%|          | 0/1 [00:00<?, ?tasks/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[992]\n",
      "Stopping engine(s): 1680783042\n",
      "engine set stopped 1680783042: {'engines': {'0': {'exit_code': 0, 'pid': 429592, 'identifier': '0'}}, 'exit_code': 0}\n",
      "Stopping controller\n",
      "Controller stopped: {'exit_code': 0, 'pid': 429563, 'identifier': 'ipcontroller-1680783041-w6r9-429542'}\n",
      "32.62002611160278\n"
     ]
    }
   ],
   "source": [
    "runtask(1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "679451a7",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Starting 16 engines with <class 'ipyparallel.cluster.launcher.LocalEngineSetLauncher'>\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a5058565a70948b3a453522b496044b3",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/16 [00:00<?, ?engine/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "c7599b5a98ad43dc8e3cf4d187689fef",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "confgen:   0%|          | 0/16 [00:00<?, ?tasks/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[63, 61, 63, 63, 63, 62, 63, 63, 61, 61, 61, 62, 62, 62, 62, 61]\n",
      "Stopping engine(s): 1680783075\n",
      "engine set stopped 1680783075: {'engines': {'0': {'exit_code': 0, 'pid': 429664, 'identifier': '0'}, '1': {'exit_code': 0, 'pid': 429666, 'identifier': '1'}, '2': {'exit_code': 0, 'pid': 429668, 'identifier': '2'}, '3': {'exit_code': 0, 'pid': 429670, 'identifier': '3'}, '4': {'exit_code': 0, 'pid': 429676, 'identifier': '4'}, '5': {'exit_code': 0, 'pid': 429682, 'identifier': '5'}, '6': {'exit_code': 0, 'pid': 429688, 'identifier': '6'}, '7': {'exit_code': 0, 'pid': 429700, 'identifier': '7'}, '8': {'exit_code': 0, 'pid': 429713, 'identifier': '8'}, '9': {'exit_code': 0, 'pid': 429728, 'identifier': '9'}, '10': {'exit_code': 0, 'pid': 429738, 'identifier': '10'}, '11': {'exit_code': 0, 'pid': 429754, 'identifier': '11'}, '12': {'exit_code': 0, 'pid': 429770, 'identifier': '12'}, '13': {'exit_code': 0, 'pid': 429786, 'identifier': '13'}, '14': {'exit_code': 0, 'pid': 429802, 'identifier': '14'}, '15': {'exit_code': 0, 'pid': 429818, 'identifier': '15'}}, 'exit_code': 0}\n",
      "Stopping controller\n",
      "Controller stopped: {'exit_code': 0, 'pid': 429633, 'identifier': 'ipcontroller-1680783074-y6mr-429542'}\n",
      "12.983936071395874\n"
     ]
    }
   ],
   "source": [
    "runtask(16)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "55eeb63e",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.9.16"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}