07_full_length_all_aa_different_probabilites_reduced_indel_prob.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "695ff11e",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<frozen importlib._bootstrap>:228: RuntimeWarning: scipy._lib.messagestream.MessageStream size changed, may indicate binary incompatibility. Expected 56 from C header, got 64 from PyObject\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO: Pandarallel will run on 48 workers.\n",
      "INFO: Pandarallel will use Memory file system to transfer data between the main process and workers.\n"
     ]
    }
   ],
   "source": [
    "import shutil\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "from multiprocessing import Pool, cpu_count\n",
    "from functools import partial\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "sns.set_context(\"paper\", font_scale=2.0)\n",
    "sns.set_style('whitegrid')\n",
    "\n",
    "# reduced indel probabilites in this file\n",
    "from functions2 import *\n",
    "\n",
    "from pandarallel import pandarallel\n",
    "\n",
    "pandarallel.initialize(progress_bar=False)\n",
    "\n",
    "with SequenceFile(\"data/uniprot-9606.fasta\", digital=True, alphabet=alphabet) as seq_file:\n",
    "    sequences = list(seq_file)\n",
    "    \n",
    "\n",
    "uniprot = fasta_reader('data/uniprot-9606.fasta')\n",
    "\n",
    "aa_known = 'ACDEFGHIKLMNPQRSTVWY'\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "b881ff8a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "95 %%5 %95 %97 %95 %78 %87 %86 %87 %87 %88 %88 %90 %89 %89 %90 %91 %91 %91 %91 %94 %94 %93 %92 %94 %94 %94 %95 %94 %96 %97 %97 %31 %82 %89 %91 %92 %93 %93 %94 %94 %95 %98 %59 %84 %94 %95 %95 %95 %96 %96 %63 %94 %94 %95 %95 %95 %95 %95 %95 %95 %98 %77 %91 %93 %93 %94 %94 %94 %96 %88 %95 %95 %94 %94 %94 %95 %95 %95 %85 %90 %87 %87 %88 %88 %88 %90 %89 %89 %91 %91 %92 %94 %94 %94 %94 %94 %95 %95 %58 %83 %93 %93 %94 %94 %94 %94 %94 %94 %94 %95 %95 %55 %75 %80 %79 %93 %41 %23 %93 %95 %78 %87 %93 %89 %94 %95 %95 %96 %95 %95 %95 %95 %95 %95 %95 %95 %95 %96 %92 %79 %79 %94 %84 %54 %55 %80 %94 %95 %93 %94 %94 %98 %94 %95 %95 %95 %95 %95 %91 %93 %92 %96 %97 %91 %86 %95 %46 %95 %75 %92 %76 %94 %95 %95 %95 %96 %88 %80 %88 %\r"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "100 %1 %89 %94 %74 %10 %90 %78 %74 %\r"
     ]
    }
   ],
   "source": [
    "prob_range = [0.9, 0.8, 0.6, 0.4, 0.3, 0.2, 0.1, 0.09, 0.08, 0.07, 0.06, 0.05]\n",
    "\n",
    "\n",
    "\n",
    "for prob in prob_range:\n",
    "    temp_df = uniprot.copy()\n",
    "\n",
    "    temp_df['temp'] = temp_df.values.tolist() \n",
    "    temp_df['posteriors'] =  temp_df.temp.parallel_apply(\\\n",
    "                         lambda x: generate_reads(seq=x[1], aa_known=aa_known, max_prob=prob))\n",
    "\n",
    "\n",
    "    temp_df['temp1'] = temp_df[['Accession', 'posteriors']].values.tolist() \n",
    "    temp_df['hmm'] = temp_df.temp1.parallel_apply(\\\n",
    "             lambda x: hmm_build(x[1], x[0], f'full_length_{x[0]}_{aa_known}_max_prob_{prob}'))\n",
    "\n",
    "\n",
    "    hmms = np.array_split(temp_df.hmm, 10)\n",
    "    hmms = [i.tolist() for i in hmms]\n",
    "\n",
    "    results = []\n",
    "    for i, v in enumerate(hmms):\n",
    "        results.append(score(v, sequences, background))\n",
    "    #             print(f'\\t\\t\\t\\tdone: {i}', end='\\r')\n",
    "    fname = f'full_length_{aa_known}_max_prob_{prob}_del_reduced_indel_prob.pkl.gz'\n",
    "\n",
    "    res_df = pd.concat(results)\n",
    "    res_df.to_pickle('results/scan_results/' + fname)\n",
    "    del temp_df, res_df, hmms\n",
    "    shutil.rmtree('temp/', ignore_errors=True)\n",
    "    try:\n",
    "        os.makedirs('temp/')\n",
    "    except FileExistsError:\n",
    "        pass"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "743d7ea1",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "39c56c2e",
   "metadata": {},
   "source": [
    "#### plotting "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "d3959ed2",
   "metadata": {},
   "outputs": [],
   "source": [
    "aa_known = 'ACDEFGHIKLMNPQRSTVWY'\n",
    "prob_range = [0.9, 0.8, 0.6, 0.4, 0.2, 0.3, 0.1, 0.09, 0.08, 0.07, 0.06, 0.05]\n",
    "\n",
    "all_files = ['results/scan_results/' + f'full_length_{aa_known}_max_prob_{prob}_del_reduced_indel_prob.pkl.gz' \\\n",
    "             for prob in prob_range]\n",
    "result_dfs = [pd.read_pickle(i) for i in all_files]\n",
    "# total number of HMMS\n",
    "hmms = list(range(uniprot.shape[0]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "f69a2849",
   "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>known aa</th>\n",
       "      <th>max_prob</th>\n",
       "      <th>total queries</th>\n",
       "      <th>hits</th>\n",
       "      <th>identified</th>\n",
       "      <th>identified_frac</th>\n",
       "      <th>Type</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>ACDEFGHIKLMNPQRSTVWY</td>\n",
       "      <td>0.90</td>\n",
       "      <td>20181</td>\n",
       "      <td>20179</td>\n",
       "      <td>19969</td>\n",
       "      <td>0.989495</td>\n",
       "      <td>low indel prob.</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>ACDEFGHIKLMNPQRSTVWY</td>\n",
       "      <td>0.80</td>\n",
       "      <td>20181</td>\n",
       "      <td>20179</td>\n",
       "      <td>19965</td>\n",
       "      <td>0.989297</td>\n",
       "      <td>low indel prob.</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>ACDEFGHIKLMNPQRSTVWY</td>\n",
       "      <td>0.60</td>\n",
       "      <td>20181</td>\n",
       "      <td>20179</td>\n",
       "      <td>19888</td>\n",
       "      <td>0.985481</td>\n",
       "      <td>low indel prob.</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>ACDEFGHIKLMNPQRSTVWY</td>\n",
       "      <td>0.40</td>\n",
       "      <td>20181</td>\n",
       "      <td>20179</td>\n",
       "      <td>19802</td>\n",
       "      <td>0.98122</td>\n",
       "      <td>low indel prob.</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>ACDEFGHIKLMNPQRSTVWY</td>\n",
       "      <td>0.20</td>\n",
       "      <td>20181</td>\n",
       "      <td>20180</td>\n",
       "      <td>19732</td>\n",
       "      <td>0.977751</td>\n",
       "      <td>low indel prob.</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>ACDEFGHIKLMNPQRSTVWY</td>\n",
       "      <td>0.30</td>\n",
       "      <td>20181</td>\n",
       "      <td>20178</td>\n",
       "      <td>19760</td>\n",
       "      <td>0.979139</td>\n",
       "      <td>low indel prob.</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>ACDEFGHIKLMNPQRSTVWY</td>\n",
       "      <td>0.10</td>\n",
       "      <td>20181</td>\n",
       "      <td>20144</td>\n",
       "      <td>19791</td>\n",
       "      <td>0.980675</td>\n",
       "      <td>low indel prob.</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>ACDEFGHIKLMNPQRSTVWY</td>\n",
       "      <td>0.09</td>\n",
       "      <td>20181</td>\n",
       "      <td>20146</td>\n",
       "      <td>19804</td>\n",
       "      <td>0.981319</td>\n",
       "      <td>low indel prob.</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>ACDEFGHIKLMNPQRSTVWY</td>\n",
       "      <td>0.08</td>\n",
       "      <td>20181</td>\n",
       "      <td>20141</td>\n",
       "      <td>19805</td>\n",
       "      <td>0.981369</td>\n",
       "      <td>low indel prob.</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>ACDEFGHIKLMNPQRSTVWY</td>\n",
       "      <td>0.07</td>\n",
       "      <td>20181</td>\n",
       "      <td>20148</td>\n",
       "      <td>19607</td>\n",
       "      <td>0.971557</td>\n",
       "      <td>low indel prob.</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>ACDEFGHIKLMNPQRSTVWY</td>\n",
       "      <td>0.06</td>\n",
       "      <td>20181</td>\n",
       "      <td>20158</td>\n",
       "      <td>16268</td>\n",
       "      <td>0.806105</td>\n",
       "      <td>low indel prob.</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>ACDEFGHIKLMNPQRSTVWY</td>\n",
       "      <td>0.05</td>\n",
       "      <td>20181</td>\n",
       "      <td>20161</td>\n",
       "      <td>2</td>\n",
       "      <td>0.000099</td>\n",
       "      <td>low indel prob.</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                known aa  max_prob total queries   hits identified  \\\n",
       "0   ACDEFGHIKLMNPQRSTVWY      0.90         20181  20179      19969   \n",
       "1   ACDEFGHIKLMNPQRSTVWY      0.80         20181  20179      19965   \n",
       "2   ACDEFGHIKLMNPQRSTVWY      0.60         20181  20179      19888   \n",
       "3   ACDEFGHIKLMNPQRSTVWY      0.40         20181  20179      19802   \n",
       "4   ACDEFGHIKLMNPQRSTVWY      0.20         20181  20180      19732   \n",
       "5   ACDEFGHIKLMNPQRSTVWY      0.30         20181  20178      19760   \n",
       "6   ACDEFGHIKLMNPQRSTVWY      0.10         20181  20144      19791   \n",
       "7   ACDEFGHIKLMNPQRSTVWY      0.09         20181  20146      19804   \n",
       "8   ACDEFGHIKLMNPQRSTVWY      0.08         20181  20141      19805   \n",
       "9   ACDEFGHIKLMNPQRSTVWY      0.07         20181  20148      19607   \n",
       "10  ACDEFGHIKLMNPQRSTVWY      0.06         20181  20158      16268   \n",
       "11  ACDEFGHIKLMNPQRSTVWY      0.05         20181  20161          2   \n",
       "\n",
       "   identified_frac             Type  \n",
       "0         0.989495  low indel prob.  \n",
       "1         0.989297  low indel prob.  \n",
       "2         0.985481  low indel prob.  \n",
       "3          0.98122  low indel prob.  \n",
       "4         0.977751  low indel prob.  \n",
       "5         0.979139  low indel prob.  \n",
       "6         0.980675  low indel prob.  \n",
       "7         0.981319  low indel prob.  \n",
       "8         0.981369  low indel prob.  \n",
       "9         0.971557  low indel prob.  \n",
       "10        0.806105  low indel prob.  \n",
       "11        0.000099  low indel prob.  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "plot_df = pd.DataFrame(columns=['known aa', 'max_prob', \\\n",
    "                                'total queries', 'hits',  'identified', ])\n",
    "\n",
    "for i, v in enumerate(result_dfs):\n",
    "    stats = gen_stats(hmms, v)\n",
    "    plot_df.loc[i] = [aa_known, prob_range[i], stats['Total queries'],\\\n",
    "                       stats['Hits'], stats['Identified'],]\n",
    "plot_df['identified_frac'] = plot_df['identified']/plot_df['total queries']\n",
    "plot_df['Type'] = 'low indel prob.'\n",
    "plot_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "30be1fec",
   "metadata": {},
   "outputs": [],
   "source": [
    "plot_df.to_pickle('results/full_length_all_p_max_reduced_indel_prob.pkl.gz')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "9abbf8c0",
   "metadata": {},
   "outputs": [],
   "source": [
    "plot_df = pd.read_pickle('results/full_length_all_p_max_reduced_indel_prob.pkl.gz')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "cce760e8",
   "metadata": {},
   "outputs": [],
   "source": [
    "plot_df1 = pd.read_pickle('results/full_length_all_p_max.pkl.gz')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "d1d291d8",
   "metadata": {},
   "outputs": [],
   "source": [
    "new_df = pd.concat([plot_df1, plot_df, ])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "5a1f5d4b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.set_context(\"paper\", font_scale=2)\n",
    "sns.set_style('ticks')\n",
    "from matplotlib.ticker import ScalarFormatter\n",
    "plt.figure(figsize=(8, 4))\n",
    "ax = sns.lineplot(data=new_df, \\\n",
    "                     x='max_prob', y='identified_frac', hue='Type', \\\n",
    "                  palette=\"Set1\", marker='o', markersize=8, lw=2,)\n",
    "ax.set_xscale('log')\n",
    "ax.set_xticks(plot_df['max_prob'].value_counts().index, )\n",
    "ax.tick_params(axis='x', rotation=90)\n",
    "ax.get_xaxis().set_major_formatter(ScalarFormatter())\n",
    "ax.set_xlabel(r'${p_{max}}$')\n",
    "ax.set_ylabel('Accuracy')\n",
    "plt.savefig('figs/fig1-logscale-reduced_indels.pdf');"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f8bbab77",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "003c0a60",
   "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": "3.9.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}