{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "63b89129-eb15-489c-acae-c6255df4d9f0",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ubuntu/.local/lib/python3.8/site-packages/pandas/core/computation/expressions.py:20: UserWarning: Pandas requires version '2.7.3' or newer of 'numexpr' (version '2.7.1' currently installed).\n",
      "  from pandas.core.computation.check import NUMEXPR_INSTALLED\n",
      "/home/ubuntu/.local/lib/python3.8/site-packages/Bio/pairwise2.py:278: BiopythonDeprecationWarning: Bio.pairwise2 has been deprecated, and we intend to remove it in a future release of Biopython. As an alternative, please consider using Bio.Align.PairwiseAligner as a replacement, and contact the Biopython developers if you still need the Bio.pairwise2 module.\n",
      "  warnings.warn(\n"
     ]
    }
   ],
   "source": [
    "# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved.\n",
    "# SPDX-License-Identifier: MIT-0\n",
    "\n",
    "\"\"\"\n",
    "Parse PDB files to extract the coordinates of the 4 key atoms from AAs to\n",
    "generate json records compatible to the LM-GVP model.\n",
    "\n",
    "This script is intended for Fluorescence and Protease datasets from TAPE.\n",
    "\"\"\"\n",
    "\n",
    "import json\n",
    "import os\n",
    "import argparse\n",
    "from collections import defaultdict\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "\n",
    "from tqdm import tqdm\n",
    "from joblib import Parallel, delayed\n",
    "from Bio.PDB import PDBParser\n",
    "from Bio.PDB.Polypeptide import three_to_one\n",
    "\n",
    "import xpdb\n",
    "from contact_map_utils import parse_pdb_structure\n",
    "\n",
    "\n",
    "def parse_args():\n",
    "    \"\"\"Prepare argument parser.\n",
    "\n",
    "    Args:\n",
    "\n",
    "    Return:\n",
    "\n",
    "    \"\"\"\n",
    "    parser = argparse.ArgumentParser(\n",
    "        description=\"Generate GVP ProteinGraph datasets representing protein\"\n",
    "        + \"structures.\"\n",
    "    )\n",
    "    parser.add_argument(\n",
    "        \"--data-file\",\n",
    "        help=\"Path to protein data frame, including sequences, paths to\"\n",
    "        + \" structure files and labels\",\n",
    "        required=True,\n",
    "    )\n",
    "    parser.add_argument(\n",
    "        \"-t\",\n",
    "        \"--target-variable\",\n",
    "        help=\"target variable in the protein data frame\",\n",
    "        required=True,\n",
    "    )\n",
    "    parser.add_argument(\"-o\", \"--output\", help=\"output dir for graphs\")\n",
    "\n",
    "    args = parser.parse_args()\n",
    "    return args\n",
    "\n",
    "\n",
    "def get_atom_coords(residue, target_atoms=[\"N\", \"CA\", \"C\", \"O\"]):\n",
    "    \"\"\"Extract the coordinates of the target_atoms from an AA residue.\n",
    "\n",
    "    Args:\n",
    "        residue: a Bio.PDB.Residue object representing the residue.\n",
    "        target_atoms: Target atoms which residues will be returned.\n",
    "\n",
    "    Retruns:\n",
    "        Array of residue's target atoms (in the same order as target atoms).\n",
    "    \"\"\"\n",
    "    return np.asarray([residue[atom].coord for atom in target_atoms])\n",
    "\n",
    "\n",
    "def structure_to_coords(struct, target_atoms=[\"N\", \"CA\", \"C\", \"O\"], name=\"\"):\n",
    "    \"\"\"Convert a PDB structure in to coordinates of target atoms from all AAs\n",
    "\n",
    "    Args:\n",
    "        struct: a Bio.PDB.Structure object representing the protein structure\n",
    "        target_atoms: Target atoms which residues will be returned.\n",
    "        name: String. Name of the structure\n",
    "\n",
    "    Return:\n",
    "        Dictionary with the pdb sequence, atom 3D coordinates and name.\n",
    "    \"\"\"\n",
    "    output = {}\n",
    "    # get AA sequence in the pdb structure\n",
    "    pdb_seq = \"\".join(\n",
    "        [three_to_one(res.get_resname()) for res in struct.get_residues()]\n",
    "    )\n",
    "    output[\"seq\"] = pdb_seq\n",
    "    # get the atom coords\n",
    "    coords = np.asarray(\n",
    "        [\n",
    "            get_atom_coords(res, target_atoms=target_atoms)\n",
    "            for res in struct.get_residues()\n",
    "        ]\n",
    "    )\n",
    "    output[\"coords\"] = coords.tolist()\n",
    "    output[\"name\"] = name\n",
    "    return output\n",
    "\n",
    "\n",
    "def parse_pdb_gz_to_json_record(parser, sequence, pdb_file_path, name=\"\"):\n",
    "    \"\"\"\n",
    "    Reads and reformats a pdb strcuture into a dictionary.\n",
    "\n",
    "    Args:\n",
    "        parser: a Bio.PDB.PDBParser or Bio.PDB.MMCIFParser instance.\n",
    "        sequence: String. Sequence of the structure.\n",
    "        pdb_file_path: String. Path to the pdb file.\n",
    "        name: String. Name of the protein.\n",
    "\n",
    "    Return:\n",
    "        Dictionary with the pdb sequence, atom 3D coordinates and name.\n",
    "    \"\"\"\n",
    "    struct = parse_pdb_structure(parser, sequence, pdb_file_path)\n",
    "    record = structure_to_coords(struct, name=name)\n",
    "    return record"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "402ad072-edd1-4c04-a4c0-93390862770b",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "72469cb0-878a-478b-849d-ee88774f3708",
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "  0%|          | 0/23197 [00:00<?, ?it/s]\u001b[A\n",
      "  0%|          | 30/23197 [00:00<05:33, 69.52it/s]\u001b[A\n",
      "  0%|          | 60/23197 [00:00<03:52, 99.48it/s]\u001b[A\n",
      "  0%|          | 90/23197 [00:01<04:16, 90.06it/s]\u001b[A\n",
      "  1%|          | 150/23197 [00:01<02:19, 164.87it/s]\u001b[A\n",
      "  1%|          | 210/23197 [00:01<01:34, 243.85it/s]\u001b[A\n",
      "  0%|          | 0/2320 [08:17<?, ?it/s] 436.75it/s]\u001b[A\n",
      "\n",
      "  2%|▏         | 393/23197 [00:02<02:12, 171.72it/s]\u001b[A\n",
      "  2%|▏         | 510/23197 [00:02<01:23, 272.13it/s]\u001b[A\n",
      "  3%|▎         | 630/23197 [00:02<01:06, 339.34it/s]\u001b[A\n",
      "  3%|▎         | 750/23197 [00:02<00:56, 399.22it/s]\u001b[A\n",
      "  4%|▍         | 870/23197 [00:03<00:50, 439.06it/s]\u001b[A\n",
      "  4%|▍         | 990/23197 [00:03<00:47, 464.22it/s]\u001b[A\n",
      "  5%|▍         | 1110/23197 [00:04<01:36, 227.71it/s]\u001b[A\n",
      "  5%|▌         | 1230/23197 [00:04<01:16, 288.41it/s]\u001b[A\n",
      "  6%|▌         | 1350/23197 [00:04<01:05, 332.80it/s]\u001b[A\n",
      "  6%|▋         | 1470/23197 [00:04<00:57, 380.43it/s]\u001b[A\n",
      "  7%|▋         | 1590/23197 [00:05<00:52, 412.55it/s]\u001b[A\n",
      "  7%|▋         | 1710/23197 [00:05<00:46, 457.87it/s]\u001b[A\n",
      "  8%|▊         | 1830/23197 [00:05<00:45, 471.04it/s]\u001b[A\n",
      "  8%|▊         | 1950/23197 [00:06<01:42, 207.42it/s]\u001b[A\n",
      "  9%|▉         | 2070/23197 [00:07<01:19, 265.95it/s]\u001b[A\n",
      "  9%|▉         | 2190/23197 [00:07<01:05, 321.43it/s]\u001b[A\n",
      " 10%|▉         | 2310/23197 [00:07<00:54, 381.64it/s]\u001b[A\n",
      " 10%|█         | 2430/23197 [00:07<00:49, 418.37it/s]\u001b[A\n",
      " 11%|█         | 2550/23197 [00:07<00:45, 454.37it/s]\u001b[A\n",
      " 12%|█▏        | 2670/23197 [00:08<00:43, 474.03it/s]\u001b[A\n",
      " 12%|█▏        | 2790/23197 [00:08<00:40, 498.03it/s]\u001b[A\n",
      " 13%|█▎        | 2910/23197 [00:08<00:36, 562.53it/s]\u001b[A\n",
      " 13%|█▎        | 3030/23197 [00:08<00:39, 511.14it/s]\u001b[A\n",
      " 14%|█▎        | 3150/23197 [00:10<01:47, 185.83it/s]\u001b[A\n",
      " 14%|█▍        | 3270/23197 [00:10<01:23, 238.75it/s]\u001b[A\n",
      " 15%|█▍        | 3390/23197 [00:10<01:07, 295.04it/s]\u001b[A\n",
      " 15%|█▌        | 3510/23197 [00:10<00:59, 331.53it/s]\u001b[A\n",
      " 16%|█▌        | 3630/23197 [00:11<00:51, 379.40it/s]\u001b[A\n",
      " 16%|█▌        | 3750/23197 [00:11<00:45, 432.09it/s]\u001b[A\n",
      " 17%|█▋        | 3870/23197 [00:11<00:41, 468.00it/s]\u001b[A\n",
      " 17%|█▋        | 3990/23197 [00:11<00:40, 470.86it/s]\u001b[A\n",
      " 18%|█▊        | 4110/23197 [00:12<00:39, 481.31it/s]\u001b[A\n",
      " 18%|█▊        | 4230/23197 [00:12<00:39, 485.39it/s]\u001b[A\n",
      " 19%|█▉        | 4350/23197 [00:14<01:58, 158.62it/s]\u001b[A\n",
      " 19%|█▉        | 4470/23197 [00:14<01:31, 204.84it/s]\u001b[A\n",
      " 19%|█▉        | 4517/23197 [00:14<01:32, 201.93it/s]\u001b[A\n",
      " 20%|█▉        | 4555/23197 [00:14<01:27, 212.57it/s]\u001b[A\n",
      " 20%|█▉        | 4620/23197 [00:14<01:14, 250.91it/s]\u001b[A\n",
      " 21%|██        | 4779/23197 [00:15<00:44, 417.69it/s]\u001b[A\n",
      " 21%|██        | 4881/23197 [00:15<00:36, 508.67it/s]\u001b[A\n",
      " 21%|██▏       | 4965/23197 [00:15<00:36, 497.82it/s]\u001b[A\n",
      " 22%|██▏       | 5040/23197 [00:15<00:40, 450.07it/s]\u001b[A\n",
      " 22%|██▏       | 5160/23197 [00:15<00:38, 468.90it/s]\u001b[A\n",
      " 23%|██▎       | 5280/23197 [00:16<00:36, 497.42it/s]\u001b[A\n",
      " 23%|██▎       | 5400/23197 [00:16<00:32, 543.94it/s]\u001b[A\n",
      " 24%|██▍       | 5520/23197 [00:16<00:32, 545.80it/s]\u001b[A\n",
      " 24%|██▍       | 5640/23197 [00:16<00:31, 555.94it/s]\u001b[A\n",
      " 25%|██▍       | 5760/23197 [00:16<00:31, 545.56it/s]\u001b[A\n",
      " 25%|██▌       | 5880/23197 [00:17<00:30, 561.35it/s]\u001b[A\n",
      " 26%|██▌       | 6000/23197 [00:19<02:05, 136.68it/s]\u001b[A\n",
      " 26%|██▌       | 6042/23197 [00:19<01:57, 145.48it/s]\u001b[A\n",
      " 26%|██▌       | 6079/23197 [00:19<01:49, 156.87it/s]\u001b[A\n",
      " 26%|██▋       | 6120/23197 [00:19<01:36, 176.68it/s]\u001b[A\n",
      " 27%|██▋       | 6234/23197 [00:19<01:01, 274.74it/s]\u001b[A\n",
      " 27%|██▋       | 6300/23197 [00:20<00:52, 321.50it/s]\u001b[A\n",
      " 28%|██▊       | 6420/23197 [00:20<00:40, 413.74it/s]\u001b[A\n",
      " 28%|██▊       | 6540/23197 [00:20<00:36, 453.33it/s]\u001b[A\n",
      " 29%|██▊       | 6660/23197 [00:20<00:34, 478.46it/s]\u001b[A\n",
      " 29%|██▉       | 6780/23197 [00:20<00:30, 532.01it/s]\u001b[A\n",
      " 30%|██▉       | 6900/23197 [00:20<00:28, 573.82it/s]\u001b[A\n",
      " 30%|███       | 7020/23197 [00:21<00:29, 554.08it/s]\u001b[A\n",
      " 31%|███       | 7140/23197 [00:21<00:28, 562.69it/s]\u001b[A\n",
      " 31%|███▏      | 7260/23197 [00:21<00:28, 555.14it/s]\u001b[A\n",
      " 32%|███▏      | 7380/23197 [00:21<00:28, 551.73it/s]\u001b[A\n",
      " 32%|███▏      | 7500/23197 [00:22<00:28, 556.21it/s]\u001b[A\n",
      " 33%|███▎      | 7620/23197 [00:22<00:27, 574.53it/s]\u001b[A\n",
      " 33%|███▎      | 7740/23197 [00:22<00:27, 572.23it/s]\u001b[A\n",
      " 34%|███▍      | 7860/23197 [00:22<00:26, 580.83it/s]\u001b[A\n",
      " 34%|███▍      | 7980/23197 [00:25<02:11, 116.07it/s]\u001b[A\n",
      " 35%|███▍      | 8100/23197 [00:25<01:35, 158.64it/s]\u001b[A\n",
      " 35%|███▌      | 8163/23197 [00:26<01:34, 159.20it/s]\u001b[A\n",
      " 36%|███▌      | 8250/23197 [00:26<01:16, 195.64it/s]\u001b[A\n",
      " 36%|███▌      | 8351/23197 [00:26<00:57, 259.76it/s]\u001b[A\n",
      " 36%|███▋      | 8430/23197 [00:26<00:49, 298.77it/s]\u001b[A\n",
      " 37%|███▋      | 8550/23197 [00:26<00:38, 376.68it/s]\u001b[A\n",
      " 37%|███▋      | 8670/23197 [00:26<00:35, 413.07it/s]\u001b[A\n",
      " 38%|███▊      | 8790/23197 [00:27<00:31, 460.38it/s]\u001b[A\n",
      " 38%|███▊      | 8910/23197 [00:27<00:31, 453.23it/s]\u001b[A\n",
      " 39%|███▉      | 9030/23197 [00:27<00:26, 531.89it/s]\u001b[A\n",
      " 39%|███▉      | 9150/23197 [00:27<00:24, 573.90it/s]\u001b[A\n",
      " 40%|███▉      | 9270/23197 [00:27<00:24, 557.27it/s]\u001b[A\n",
      " 40%|████      | 9390/23197 [00:28<00:23, 595.00it/s]\u001b[A\n",
      " 41%|████      | 9510/23197 [00:28<00:24, 561.83it/s]\u001b[A\n",
      " 42%|████▏     | 9630/23197 [00:28<00:23, 585.56it/s]\u001b[A\n",
      " 42%|████▏     | 9750/23197 [00:28<00:24, 559.54it/s]\u001b[A\n",
      " 43%|████▎     | 9870/23197 [00:29<00:24, 544.38it/s]\u001b[A\n",
      " 43%|████▎     | 9990/23197 [00:29<00:26, 491.72it/s]\u001b[A\n",
      " 44%|████▎     | 10110/23197 [00:29<00:25, 511.47it/s]\u001b[A\n",
      " 44%|████▍     | 10230/23197 [00:29<00:24, 530.12it/s]\u001b[A\n",
      " 45%|████▍     | 10350/23197 [00:29<00:24, 525.07it/s]\u001b[A\n",
      " 45%|████▌     | 10470/23197 [00:33<02:11, 96.94it/s] \u001b[A\n",
      " 45%|████▌     | 10509/23197 [00:33<02:01, 104.64it/s]\u001b[A\n",
      " 46%|████▌     | 10576/23197 [00:33<01:37, 129.95it/s]\u001b[A\n",
      " 46%|████▌     | 10650/23197 [00:34<01:17, 161.65it/s]\u001b[A\n",
      " 46%|████▋     | 10772/23197 [00:34<00:51, 243.02it/s]\u001b[A\n",
      " 47%|████▋     | 10890/23197 [00:34<00:38, 317.92it/s]\u001b[A\n",
      " 47%|████▋     | 11010/23197 [00:34<00:33, 368.02it/s]\u001b[A\n",
      " 48%|████▊     | 11130/23197 [00:34<00:30, 400.26it/s]\u001b[A\n",
      " 48%|████▊     | 11250/23197 [00:34<00:26, 447.80it/s]\u001b[A\n",
      " 49%|████▉     | 11370/23197 [00:35<00:24, 484.29it/s]\u001b[A\n",
      " 50%|████▉     | 11490/23197 [00:35<00:21, 541.63it/s]\u001b[A\n",
      " 50%|█████     | 11610/23197 [00:35<00:20, 562.89it/s]\u001b[A\n",
      " 51%|█████     | 11730/23197 [00:35<00:19, 573.70it/s]\u001b[A\n",
      " 51%|█████     | 11850/23197 [00:35<00:19, 588.09it/s]\u001b[A\n",
      " 52%|█████▏    | 11970/23197 [00:36<00:19, 586.69it/s]\u001b[A\n",
      " 52%|█████▏    | 12090/23197 [00:36<00:20, 538.58it/s]\u001b[A\n",
      " 53%|█████▎    | 12210/23197 [00:36<00:20, 546.82it/s]\u001b[A\n",
      " 53%|█████▎    | 12330/23197 [00:36<00:18, 572.09it/s]\u001b[A\n",
      " 54%|█████▎    | 12450/23197 [00:36<00:18, 588.52it/s]\u001b[A\n",
      " 54%|█████▍    | 12570/23197 [00:37<00:19, 551.50it/s]\u001b[A\n",
      " 55%|█████▍    | 12690/23197 [00:37<00:19, 545.71it/s]\u001b[A\n",
      " 55%|█████▌    | 12810/23197 [00:37<00:17, 589.86it/s]\u001b[A\n",
      " 56%|█████▌    | 12930/23197 [00:37<00:18, 552.70it/s]\u001b[A\n",
      " 56%|█████▋    | 13050/23197 [00:38<00:17, 565.60it/s]\u001b[A\n",
      " 57%|█████▋    | 13170/23197 [00:38<00:19, 504.44it/s]\u001b[A\n",
      " 57%|█████▋    | 13290/23197 [00:38<00:18, 521.65it/s]\u001b[A\n",
      " 58%|█████▊    | 13410/23197 [00:38<00:17, 569.59it/s]\u001b[A\n",
      " 58%|█████▊    | 13530/23197 [00:38<00:16, 591.79it/s]\u001b[A\n",
      " 59%|█████▉    | 13650/23197 [00:39<00:15, 608.51it/s]\u001b[A\n",
      " 59%|█████▉    | 13770/23197 [00:43<01:54, 82.57it/s] \u001b[A\n",
      " 60%|█████▉    | 13815/23197 [00:43<01:43, 90.83it/s]\u001b[A\n",
      " 60%|█████▉    | 13854/23197 [00:43<01:32, 100.74it/s]\u001b[A\n",
      " 60%|█████▉    | 13890/23197 [00:44<01:27, 106.30it/s]\u001b[A\n",
      " 60%|██████    | 13920/23197 [00:44<01:18, 118.00it/s]\u001b[A\n",
      " 60%|██████    | 14010/23197 [00:44<00:50, 180.32it/s]\u001b[A\n",
      " 61%|██████    | 14159/23197 [00:44<00:28, 317.51it/s]\u001b[A\n",
      " 61%|██████▏   | 14234/23197 [00:44<00:24, 370.77it/s]\u001b[A\n",
      " 62%|██████▏   | 14310/23197 [00:44<00:22, 395.36it/s]\u001b[A\n",
      " 62%|██████▏   | 14430/23197 [00:44<00:19, 457.24it/s]\u001b[A\n",
      " 63%|██████▎   | 14550/23197 [00:44<00:16, 530.91it/s]\u001b[A\n",
      " 63%|██████▎   | 14670/23197 [00:45<00:15, 558.90it/s]\u001b[A\n",
      " 64%|██████▍   | 14790/23197 [00:45<00:13, 601.73it/s]\u001b[A\n",
      " 64%|██████▍   | 14910/23197 [00:45<00:12, 659.58it/s]\u001b[A\n",
      " 65%|██████▍   | 15030/23197 [00:45<00:12, 677.29it/s]\u001b[A\n",
      " 65%|██████▌   | 15150/23197 [00:45<00:12, 650.98it/s]\u001b[A\n",
      " 66%|██████▌   | 15270/23197 [00:45<00:11, 692.94it/s]\u001b[A\n",
      " 66%|██████▋   | 15390/23197 [00:46<00:11, 666.52it/s]\u001b[A\n",
      " 67%|██████▋   | 15510/23197 [00:46<00:11, 683.07it/s]\u001b[A\n",
      " 67%|██████▋   | 15630/23197 [00:46<00:11, 687.59it/s]\u001b[A\n",
      " 68%|██████▊   | 15750/23197 [00:46<00:11, 676.43it/s]\u001b[A\n",
      " 68%|██████▊   | 15870/23197 [00:46<00:10, 672.69it/s]\u001b[A\n",
      " 69%|██████▉   | 15990/23197 [00:47<00:11, 641.36it/s]\u001b[A\n",
      " 69%|██████▉   | 16110/23197 [00:47<00:10, 650.64it/s]\u001b[A\n",
      " 70%|██████▉   | 16230/23197 [00:47<00:11, 599.16it/s]\u001b[A\n",
      " 70%|███████   | 16350/23197 [00:47<00:11, 603.14it/s]\u001b[A\n",
      " 71%|███████   | 16470/23197 [00:47<00:10, 649.41it/s]\u001b[A\n",
      " 72%|███████▏  | 16590/23197 [00:48<00:09, 698.82it/s]\u001b[A\n",
      " 72%|███████▏  | 16710/23197 [00:48<00:09, 658.98it/s]\u001b[A\n",
      " 73%|███████▎  | 16830/23197 [00:48<00:10, 610.98it/s]\u001b[A\n",
      " 73%|███████▎  | 16950/23197 [00:48<00:10, 582.14it/s]\u001b[A\n",
      " 74%|███████▎  | 17070/23197 [00:48<00:10, 557.19it/s]\u001b[A\n",
      " 74%|███████▍  | 17190/23197 [00:49<00:10, 559.83it/s]\u001b[A\n",
      " 75%|███████▍  | 17310/23197 [00:49<00:11, 532.98it/s]\u001b[A\n",
      " 75%|███████▌  | 17430/23197 [00:49<00:09, 578.49it/s]\u001b[A\n",
      " 76%|███████▌  | 17550/23197 [00:49<00:09, 611.19it/s]\u001b[A\n",
      " 76%|███████▌  | 17670/23197 [00:49<00:09, 574.97it/s]\u001b[A\n",
      " 77%|███████▋  | 17790/23197 [00:50<00:09, 546.17it/s]\u001b[A\n",
      " 77%|███████▋  | 17910/23197 [00:55<01:18, 67.25it/s] \u001b[A\n",
      " 77%|███████▋  | 17950/23197 [00:55<01:11, 73.56it/s]\u001b[A\n",
      " 78%|███████▊  | 18030/23197 [00:55<00:54, 95.36it/s]\u001b[A\n",
      " 78%|███████▊  | 18178/23197 [00:56<00:32, 154.75it/s]\u001b[A\n",
      " 79%|███████▉  | 18323/23197 [00:56<00:21, 229.51it/s]\u001b[A\n",
      " 79%|███████▉  | 18418/23197 [00:56<00:17, 267.38it/s]\u001b[A\n",
      " 80%|███████▉  | 18500/23197 [00:56<00:15, 298.46it/s]\u001b[A\n",
      " 80%|████████  | 18572/23197 [00:56<00:14, 310.68it/s]\u001b[A\n",
      " 81%|████████  | 18690/23197 [00:56<00:12, 370.83it/s]\u001b[A\n",
      " 81%|████████  | 18810/23197 [00:57<00:10, 408.45it/s]\u001b[A\n",
      " 82%|████████▏ | 18930/23197 [00:57<00:08, 483.96it/s]\u001b[A\n",
      " 82%|████████▏ | 19050/23197 [00:57<00:08, 491.89it/s]\u001b[A\n",
      " 83%|████████▎ | 19170/23197 [00:57<00:07, 517.65it/s]\u001b[A\n",
      " 83%|████████▎ | 19290/23197 [00:57<00:07, 529.76it/s]\u001b[A\n",
      " 84%|████████▎ | 19410/23197 [00:58<00:06, 591.14it/s]\u001b[A\n",
      " 84%|████████▍ | 19530/23197 [00:58<00:06, 562.74it/s]\u001b[A\n",
      " 85%|████████▍ | 19650/23197 [00:58<00:06, 563.73it/s]\u001b[A\n",
      " 85%|████████▌ | 19770/23197 [00:58<00:05, 576.34it/s]\u001b[A\n",
      " 86%|████████▌ | 19890/23197 [00:59<00:06, 498.76it/s]\u001b[A\n",
      " 86%|████████▋ | 20010/23197 [00:59<00:06, 519.95it/s]\u001b[A\n",
      " 87%|████████▋ | 20130/23197 [00:59<00:05, 514.41it/s]\u001b[A\n",
      " 87%|████████▋ | 20250/23197 [00:59<00:05, 516.80it/s]\u001b[A\n",
      " 88%|████████▊ | 20370/23197 [01:00<00:05, 474.45it/s]\u001b[A\n",
      " 88%|████████▊ | 20490/23197 [01:00<00:05, 472.73it/s]\u001b[A\n",
      " 89%|████████▉ | 20610/23197 [01:00<00:05, 486.93it/s]\u001b[A\n",
      " 89%|████████▉ | 20730/23197 [01:00<00:05, 463.81it/s]\u001b[A\n",
      " 90%|████████▉ | 20850/23197 [01:01<00:05, 450.18it/s]\u001b[A\n",
      " 90%|█████████ | 20970/23197 [01:01<00:05, 436.39it/s]\u001b[A\n",
      " 91%|█████████ | 21090/23197 [01:01<00:05, 396.26it/s]\u001b[A\n",
      " 91%|█████████▏| 21210/23197 [01:02<00:04, 422.49it/s]\u001b[A\n",
      " 92%|█████████▏| 21330/23197 [01:02<00:04, 460.05it/s]\u001b[A\n",
      " 92%|█████████▏| 21450/23197 [01:02<00:03, 481.24it/s]\u001b[A\n",
      " 93%|█████████▎| 21570/23197 [01:02<00:03, 492.61it/s]\u001b[A\n",
      " 94%|█████████▎| 21690/23197 [01:02<00:02, 518.25it/s]\u001b[A\n",
      " 94%|█████████▍| 21810/23197 [01:03<00:02, 539.52it/s]\u001b[A\n",
      " 95%|█████████▍| 21930/23197 [01:03<00:02, 558.66it/s]\u001b[A\n",
      " 95%|█████████▌| 22050/23197 [01:03<00:02, 477.28it/s]\u001b[A\n",
      " 96%|█████████▌| 22170/23197 [01:03<00:01, 540.81it/s]\u001b[A\n",
      " 96%|█████████▌| 22290/23197 [01:04<00:01, 497.30it/s]\u001b[A\n",
      " 97%|█████████▋| 22410/23197 [01:10<00:14, 54.55it/s] \u001b[A\n",
      " 97%|█████████▋| 22449/23197 [01:11<00:12, 59.72it/s]\u001b[A\n",
      " 97%|█████████▋| 22482/23197 [01:11<00:10, 65.92it/s]\u001b[A\n",
      " 97%|█████████▋| 22512/23197 [01:11<00:09, 73.94it/s]\u001b[A\n",
      " 97%|█████████▋| 22590/23197 [01:11<00:05, 107.42it/s]\u001b[A\n",
      " 98%|█████████▊| 22701/23197 [01:11<00:02, 171.55it/s]\u001b[A\n",
      " 98%|█████████▊| 22770/23197 [01:11<00:02, 213.01it/s]\u001b[A\n",
      " 99%|█████████▊| 22890/23197 [01:11<00:01, 288.99it/s]\u001b[A\n",
      " 99%|█████████▉| 23010/23197 [01:12<00:00, 349.43it/s]\u001b[A\n",
      "100%|██████████| 23197/23197 [01:12<00:00, 321.23it/s]\u001b[A\n"
     ]
    }
   ],
   "source": [
    "df = pd.read_csv('../../../data/processed/merged/kcat_max_wt_singleSeqs_wpdbs.csv')\n",
    "\n",
    "# PDB parser\n",
    "sloppyparser = PDBParser(\n",
    "    QUIET=True,\n",
    "    PERMISSIVE=True,\n",
    "    structure_builder=xpdb.SloppyStructureBuilder(),\n",
    ")\n",
    "\n",
    "additional_path_pdb = '../../../'\n",
    "# 2. Parallel parsing structures and converting to protein records\n",
    "records = Parallel(n_jobs=-1)(\n",
    "    delayed(parse_pdb_gz_to_json_record)(\n",
    "        sloppyparser,\n",
    "        df.iloc[i][\"sequence\"],\n",
    "        additional_path_pdb+df.iloc[i][\"pdbpath\"],\n",
    "        df.iloc[i][\"pdbpath\"].split(\"/\")[-1],\n",
    "    )\n",
    "    for i in tqdm(range(df.shape[0]))\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "0add270e-bf61-4513-acb6-3ace461fdaa3",
   "metadata": {},
   "outputs": [],
   "source": [
    "target_col = 'log10kcat_max'\n",
    "smiles_col = 'reactant_smiles'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "50e0b9ac-dcfd-4a43-88b4-d7ae36338644",
   "metadata": {},
   "outputs": [],
   "source": [
    "outprefix = '../../../data/processed/merged/kcat_max_wt_singleSeqs_wpdbs'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "bb37b182-aa17-41d5-a7ba-5066248804ef",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "number of records: 23197\n"
     ]
    }
   ],
   "source": [
    "# 3. Segregate records by splits\n",
    "for i, rec in enumerate(records):\n",
    "    row = df.iloc[i]\n",
    "    target = row[target_col]\n",
    "    smiles = row[smiles_col]\n",
    "    rec[\"target\"] = target\n",
    "    rec[\"ec\"] = row[\"ec\"]\n",
    "    rec[\"taxonomy_id\"] = row[\"taxonomy_id\"]\n",
    "    \n",
    "# 4. write to disk\n",
    "print(\"number of records:\", len(records))\n",
    "outfile = os.path.join(f\"{outprefix}_pdbrecords.json\")\n",
    "json.dump(records, open(outfile, \"w\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "62b95f57-9214-4649-879b-c63f60e72726",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "dict_keys(['seq', 'coords', 'name', 'target', 'ec', 'taxonomy_id'])"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "records[0].keys()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "5cf2060a-b444-4e7b-88d5-5987d9553507",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'AF-U3KRF2-F1-model_v4.pdb'"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "records[0]['name']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "4857a4ed-7ae7-4603-8baf-df113f2b7fa4",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.8.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}