notebooks/protein_ligand_interactions/binding_site.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "7cfe619e-c8e0-416b-99cd-338f6eb65ee3",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "577ab528dac94d73a536d41b3adcff0a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": []
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import pl_interactions as pli\n",
    "\n",
    "import mdtraj as mdt\n",
    "import nglview as nv\n",
    "import numpy as np\n",
    "from matplotlib.colors import to_rgb\n",
    "from rdkit import Chem"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "16b43004-921a-4d11-b47a-dc9774e7e841",
   "metadata": {},
   "source": [
    "# Visualize the binding site of protein-ligand complex"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "cdd524d1-6527-4507-b87a-5b9dae0b0f27",
   "metadata": {},
   "outputs": [],
   "source": [
    "def extract_chain(traj, chains):\n",
    "    \"\"\" Extract chains from a trajectory\n",
    "\n",
    "        Parameters\n",
    "        ----------\n",
    "        traj: mdtraj.trajectory\n",
    "\n",
    "        chains: list[int]\n",
    "\n",
    "        Returns\n",
    "        -------\n",
    "        new_traj: mdtraj.Trajectory\n",
    "\n",
    "    \"\"\"\n",
    "    topology = traj.topology\n",
    "    return traj.atom_slice(\n",
    "        [atom.index for atom in topology.atoms if (atom.residue.chain.index in chains)]\n",
    "    )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "e8472f6a-ac09-4c6e-b293-71397a045029",
   "metadata": {},
   "outputs": [],
   "source": [
    "def add_sphere_to_view(view, center, radius):\n",
    "    \"\"\" Add a sphere to a view. \"\"\"\n",
    "    color = '#B03A2E'\n",
    "    n_components = len(view._ngl_component_ids)\n",
    "    view.shape.add_sphere(center.tolist(), to_rgb(color), radius, \"sphere\")\n",
    "    view.update_representation(component=n_components, repr_index=0, opacity=0.3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "90885657-0d0b-4bbe-8efc-802520035f5b",
   "metadata": {},
   "outputs": [],
   "source": [
    "traj_all_chains = mdt.load(\"../../datasets/ligand-receptor/test_cases/eralpha/1qku/1qku.pdb\")\n",
    "traj = extract_chain(traj_all_chains, [0, 3])\n",
    "view = nv.show_mdtraj(traj)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "d378115a-dd8f-4948-8cbc-be123ea8c69c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "0b2aaa938a074143940d0806d00c2e47",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "NGLWidget()"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "ligand = extract_chain(traj_all_chains, [3])\n",
    "nv.show_mdtraj(ligand)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "b3b9d112-9e62-41ed-8889-cf945fa4ef93",
   "metadata": {},
   "outputs": [],
   "source": [
    "traj.save_pdb(\"./data/er_alpha/er_alpha_A_chain.pdb\")\n",
    "ligand.save_pdb(\"./data/er_alpha/estradiol.pdb\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "f09675df-4910-44e6-968c-d1a2458cea7e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['EST:B']\n"
     ]
    }
   ],
   "source": [
    "print(pli.find_ligands_in_traj(traj))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "140a82ee-aa23-43a7-9b02-8d4eca8582e7",
   "metadata": {},
   "outputs": [],
   "source": [
    "ligand_id = \"EST:B\"\n",
    "centroid = pli.ligand_centroid(traj, ligand_id) * 10 # convert to angstroms\n",
    "extent = pli.ligand_maximum_extent(traj, ligand_id) * 10"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "de2ac114-954f-4f2f-9b07-298e2e4f2202",
   "metadata": {},
   "outputs": [],
   "source": [
    "bs_site = extent + 8.5"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "5bdeeb6c-41b8-4953-9e27-e1d553cdc829",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "42b7672cbf404c308b8bbc3ba8bbcd5d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "NGLWidget()"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "add_sphere_to_view(view, centroid, bs_site)\n",
    "view.representations = [\n",
    "     {\"type\": \"cartoon\", \n",
    "      \"params\": {\n",
    "          \"sele\": \"protein\",\n",
    "          \"color\": \"residueindex\"\n",
    "      }},\n",
    "    {\"type\": \"ball+stick\", \n",
    "     \"params\": {\n",
    "         \"sele\": \"( not polymer or hetero ) and not ( water or ion )\"\n",
    "    }}\n",
    "]\n",
    "view"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a3f192ac-5619-4b0d-aec2-e25c75adcf53",
   "metadata": {},
   "source": [
    "## Extract binding site\n",
    "Extract binding site to a new trajectory for visualization purposes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "fa26007d-7b31-4ab4-8cb4-e60b29cc58e7",
   "metadata": {},
   "outputs": [],
   "source": [
    "BS_DIST = 5.0  # in angstroms\n",
    "\n",
    "\n",
    "def get_binding_site_atoms_indices(lig_center, lig_max_extent, coords):\n",
    "    \"\"\" Get the indices of all the atoms that belong to the binding site.\n",
    "\n",
    "        Parameters\n",
    "        ----------\n",
    "        lig_center : np.array of shape (3,)\n",
    "        lig_max_extent : float\n",
    "        coords: np.array of shape(n_atoms, 3)\n",
    "\n",
    "        Returns\n",
    "        -------\n",
    "        bs_indices: list[int]\n",
    "    \"\"\"\n",
    "    bs_cutoff = lig_max_extent + BS_DIST\n",
    "    distance = np.sqrt(np.sum(np.power(coords - lig_center, 2), axis=1))\n",
    "    return np.where(distance <= bs_cutoff)[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "0bf39658-a54d-408e-a127-7f863fcac42e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(287,)"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bs_indices = get_binding_site_atoms_indices(centroid, extent, traj.xyz[0] * 10)\n",
    "bs_indices.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "f9b98f34-d086-4269-a4f0-6529b09bda70",
   "metadata": {},
   "outputs": [],
   "source": [
    "bs_traj = traj.atom_slice(bs_indices)\n",
    "\n",
    "# Some of the rsidue"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "a6af649a-d87b-42ee-8465-c491b90b03d8",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f3bc8b7e51f04675a1c407e27239f8fc",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "NGLWidget()"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "bs_view = nv.show_mdtraj(bs_traj)\n",
    "bs_view.representations = [\n",
    "     {\"type\": \"ball+stick\", \n",
    "      \"params\": {\n",
    "          \"sele\": \"protein\",\n",
    "      }},\n",
    "    {\"type\": \"ball+stick\", \n",
    "     \"params\": {\n",
    "         \"sele\": \"( not polymer or hetero ) and not ( water or ion )\",\n",
    "         \"color\": \"blue\"\n",
    "    }}\n",
    "]\n",
    "bs_view"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "09e648cb-6f55-46e7-84d1-a71377c24669",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Num chains 2\n",
      "Num residues 53\n",
      "Num atoms 287\n"
     ]
    }
   ],
   "source": [
    "print(f\"Num chains {bs_traj.n_chains}\")\n",
    "print(f\"Num residues {bs_traj.n_residues}\")\n",
    "print(f\"Num atoms {bs_traj.n_atoms}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7ad9ae58-a2ff-44f4-9c2a-79b3d7492b65",
   "metadata": {},
   "source": [
    "### Removing incomplete residues from the binding site\n",
    "\n",
    "When we extracted the binding site, some of the residues lost some of their atoms and they are \n",
    "incomplete. We will remove this residues."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "9f684227-69be-407f-a919-2dddd0e9a567",
   "metadata": {},
   "outputs": [],
   "source": [
    "bs_df, _ = bs_traj.topology.to_dataframe()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3147b799-4573-4dd5-85e4-32094c6f34c2",
   "metadata": {},
   "source": [
    "LEU 540 is an example of a residue that lost part of its atoms"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "6b1f2140-3e40-4bb6-8434-d13464caff8f",
   "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>serial</th>\n",
       "      <th>name</th>\n",
       "      <th>element</th>\n",
       "      <th>resSeq</th>\n",
       "      <th>resName</th>\n",
       "      <th>chainID</th>\n",
       "      <th>segmentID</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>259</th>\n",
       "      <td>1898</td>\n",
       "      <td>CA</td>\n",
       "      <td>C</td>\n",
       "      <td>540</td>\n",
       "      <td>LEU</td>\n",
       "      <td>0</td>\n",
       "      <td></td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>260</th>\n",
       "      <td>1900</td>\n",
       "      <td>O</td>\n",
       "      <td>O</td>\n",
       "      <td>540</td>\n",
       "      <td>LEU</td>\n",
       "      <td>0</td>\n",
       "      <td></td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>261</th>\n",
       "      <td>1901</td>\n",
       "      <td>CB</td>\n",
       "      <td>C</td>\n",
       "      <td>540</td>\n",
       "      <td>LEU</td>\n",
       "      <td>0</td>\n",
       "      <td></td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>262</th>\n",
       "      <td>1902</td>\n",
       "      <td>CG</td>\n",
       "      <td>C</td>\n",
       "      <td>540</td>\n",
       "      <td>LEU</td>\n",
       "      <td>0</td>\n",
       "      <td></td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>263</th>\n",
       "      <td>1903</td>\n",
       "      <td>CD1</td>\n",
       "      <td>C</td>\n",
       "      <td>540</td>\n",
       "      <td>LEU</td>\n",
       "      <td>0</td>\n",
       "      <td></td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>264</th>\n",
       "      <td>1904</td>\n",
       "      <td>CD2</td>\n",
       "      <td>C</td>\n",
       "      <td>540</td>\n",
       "      <td>LEU</td>\n",
       "      <td>0</td>\n",
       "      <td></td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     serial name element  resSeq resName  chainID segmentID\n",
       "259    1898   CA       C     540     LEU        0          \n",
       "260    1900    O       O     540     LEU        0          \n",
       "261    1901   CB       C     540     LEU        0          \n",
       "262    1902   CG       C     540     LEU        0          \n",
       "263    1903  CD1       C     540     LEU        0          \n",
       "264    1904  CD2       C     540     LEU        0          "
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bs_df[bs_df[\"resSeq\"] == 540]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "d1e10c67-c0b7-4023-99fc-4796920daf19",
   "metadata": {},
   "outputs": [],
   "source": [
    "traj_df, _ =  traj.topology.to_dataframe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "833e47af-c6d5-4c38-8371-3d9a72923841",
   "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>serial</th>\n",
       "      <th>name</th>\n",
       "      <th>element</th>\n",
       "      <th>resSeq</th>\n",
       "      <th>resName</th>\n",
       "      <th>chainID</th>\n",
       "      <th>segmentID</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1896</th>\n",
       "      <td>1897</td>\n",
       "      <td>N</td>\n",
       "      <td>N</td>\n",
       "      <td>540</td>\n",
       "      <td>LEU</td>\n",
       "      <td>0</td>\n",
       "      <td></td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1897</th>\n",
       "      <td>1898</td>\n",
       "      <td>CA</td>\n",
       "      <td>C</td>\n",
       "      <td>540</td>\n",
       "      <td>LEU</td>\n",
       "      <td>0</td>\n",
       "      <td></td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1898</th>\n",
       "      <td>1899</td>\n",
       "      <td>C</td>\n",
       "      <td>C</td>\n",
       "      <td>540</td>\n",
       "      <td>LEU</td>\n",
       "      <td>0</td>\n",
       "      <td></td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1899</th>\n",
       "      <td>1900</td>\n",
       "      <td>O</td>\n",
       "      <td>O</td>\n",
       "      <td>540</td>\n",
       "      <td>LEU</td>\n",
       "      <td>0</td>\n",
       "      <td></td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1900</th>\n",
       "      <td>1901</td>\n",
       "      <td>CB</td>\n",
       "      <td>C</td>\n",
       "      <td>540</td>\n",
       "      <td>LEU</td>\n",
       "      <td>0</td>\n",
       "      <td></td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1901</th>\n",
       "      <td>1902</td>\n",
       "      <td>CG</td>\n",
       "      <td>C</td>\n",
       "      <td>540</td>\n",
       "      <td>LEU</td>\n",
       "      <td>0</td>\n",
       "      <td></td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1902</th>\n",
       "      <td>1903</td>\n",
       "      <td>CD1</td>\n",
       "      <td>C</td>\n",
       "      <td>540</td>\n",
       "      <td>LEU</td>\n",
       "      <td>0</td>\n",
       "      <td></td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1903</th>\n",
       "      <td>1904</td>\n",
       "      <td>CD2</td>\n",
       "      <td>C</td>\n",
       "      <td>540</td>\n",
       "      <td>LEU</td>\n",
       "      <td>0</td>\n",
       "      <td></td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      serial name element  resSeq resName  chainID segmentID\n",
       "1896    1897    N       N     540     LEU        0          \n",
       "1897    1898   CA       C     540     LEU        0          \n",
       "1898    1899    C       C     540     LEU        0          \n",
       "1899    1900    O       O     540     LEU        0          \n",
       "1900    1901   CB       C     540     LEU        0          \n",
       "1901    1902   CG       C     540     LEU        0          \n",
       "1902    1903  CD1       C     540     LEU        0          \n",
       "1903    1904  CD2       C     540     LEU        0          "
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "traj_df[traj_df[\"resSeq\"] == 540]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "98be5d62-384f-44a7-9eab-eaaf15e40696",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([342, 343, 344, 345, 346, 347, 348, 349, 350, 351, 352, 353, 354,\n",
       "       357, 382, 383, 384, 385, 386, 387, 388, 389, 390, 391, 392, 394,\n",
       "       402, 403, 404, 405, 408, 410, 418, 420, 421, 424, 425, 428, 517,\n",
       "       518, 519, 520, 521, 522, 523, 524, 525, 526, 528, 534, 540, 544,\n",
       "       600])"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bs_res_seq = bs_df[\"resSeq\"].unique()\n",
    "bs_res_seq"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "97a891bc-f84b-46fc-857f-3ca6f7e13455",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "33 incomplete residues\n"
     ]
    }
   ],
   "source": [
    "incomplete_residues = []\n",
    "for res in bs_res_seq:\n",
    "    res_bs = bs_df[bs_df[\"resSeq\"] == res]\n",
    "    res_traj = traj_df[traj_df[\"resSeq\"] == res]\n",
    "    if len(res_bs) != len(res_traj):\n",
    "        incomplete_residues.append(res)\n",
    "\n",
    "print(f\"{len(incomplete_residues)} incomplete residues\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "b1d4dd78-6ae5-4ec0-9902-ab5ce3c77f70",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[342, 345, 348, 351, 352, 353, 354, 357, 382, 383, 385, 386, 389, 390, 392, 394, 402, 403, 405, 408, 410, 418, 420, 428, 517, 519, 520, 523, 526, 528, 534, 540, 544]\n"
     ]
    }
   ],
   "source": [
    "print(incomplete_residues)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "9797968e-653a-4959-b868-db8b7a15d441",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<mdtraj.Trajectory with 1 frames, 166 atoms, 20 residues, and unitcells at 0x7ff91df1b610>"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "atoms_bs = []\n",
    "for atom in bs_traj.topology.atoms:\n",
    "    \n",
    "    if atom.residue.resSeq not in incomplete_residues:\n",
    "        atoms_bs.append(atom.index)\n",
    "\n",
    "bs_traj = bs_traj.atom_slice(atoms_bs)\n",
    "bs_traj"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "3d20d806-a993-4cc6-949b-102d8fee9646",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "4b8da2552f154b829788d9b5f13bea15",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "NGLWidget()"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "bs_view = nv.show_mdtraj(bs_traj)\n",
    "bs_view.representations = [\n",
    "     {\"type\": \"ball+stick\", \n",
    "      \"params\": {\n",
    "          \"sele\": \"protein\",\n",
    "      }},\n",
    "    {\"type\": \"ball+stick\", \n",
    "     \"params\": {\n",
    "         \"sele\": \"( not polymer or hetero ) and not ( water or ion )\",\n",
    "         \"color\": \"blue\"\n",
    "    }}\n",
    "]\n",
    "bs_view"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "57f4e694-a862-4bdb-b5fe-c6e33d42ff69",
   "metadata": {},
   "outputs": [],
   "source": [
    "bs_traj.save_pdb(\"./data/er_alpha/binding_site.pdb\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "dbd913ce-2648-494e-8bd3-de6dc48d4e39",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "166"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bs_mol = Chem.MolFromPDBFile(\"./data/er_alpha/binding_site.pdb\")\n",
    "bs_mol.GetNumAtoms()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "d0bdc439-32ac-4900-ac8b-1911edba6cbb",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<rdkit.Chem.rdchem.Mol at 0x7ff90122f440>"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bs_mol"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "f4b61b72-57cd-48ac-a1dc-5c0418baf8b6",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "343"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bs_traj.topology.atom(0).residue.resSeq"
   ]
  }
 ],
 "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": "3.7.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}