scripts/2.plot_model_summary.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 🧮 2. Visualize model metrics"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "You are now familiar with how CarveMe works, next we will inspect the output GEMs. This notebook loads the models and plots the number of genes, reactions, and metabolites for the different communities. You may follow along passively on GitHub or launch the jupyter notebook from your terminal to interactively run through each code chunk. To do so, launch jupyter from your terminal:\n",
    "```bash\n",
    "$ cd $ROOT\n",
    "$ jupyter notebook --browser firefox\n",
    "```\n",
    "This should launch a browser window where you can navigate to the `SymbNET/scripts/` folder and click on the appropriate file to launch this notebook interactively.\n",
    "\n",
    "Note: this script is forked from Daniel's [carveme_paper repo](https://github.com/cdanielmachado/carveme_paper/blob/master/notebooks/Model%20summary.ipynb)."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Load libraries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "from framed import load_cbmodel\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "import pandas as pd\n",
    "sns.set_style('white')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 🍎 Healthy gut community"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-3-4daafaf93186>:38: UserWarning: You have mixed positional and keyword arguments, some input may be discarded.\n",
      "  ax1.legend(legend_text, bbox_to_anchor=(1.1, 1), ncol=3, handles=legend_handles)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x324 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "### Gut community members\n",
    "organisms = ['B. uniformis', 'R. bromii', 'B. wexlerae', 'E. rectale', 'F. saccharivorans']\n",
    "\n",
    "### Load gut normal models\n",
    "gut_normal = {\n",
    "    'B. uniformis': ('../models/gut_normal/ERR260255_bin.24.s.xml', 'cobra'),\n",
    "    'R. bromii': ('../models/gut_normal/ERR260255_bin.7.p.xml', 'cobra'),\n",
    "    'B. wexlerae': ('../models/gut_normal/ERR260255_bin.9.s.xml', 'cobra'),\n",
    "    'E. rectale': ('../models/gut_normal/ERR260255_bin.19.p.xml', 'cobra'),\n",
    "    'F. saccharivorans': ('../models/gut_normal/ERR260255_bin.14.p.xml', 'cobra')\n",
    "}\n",
    "\n",
    "for label, (model_file, flavor) in gut_normal.items():\n",
    "    gut_normal[label] =  load_cbmodel(model_file, flavor=flavor)\n",
    "\n",
    "data = []\n",
    "spontaneous = {'G_spontaneous', 'G_Unknown', 'G_s0001', 'G_NoAssignment'} \n",
    "\n",
    "for organism in organisms:\n",
    "    \n",
    "    model = gut_normal[organism]\n",
    "    n_genes = len(set(model.genes) - spontaneous)\n",
    "    #n_rxns = len([x for x in model.reactions if not x.startswith('R_EX_')])\n",
    "    n_enzrxns = len([x for x, rxn in model.reactions.items()\n",
    "                     if rxn.gpr is not None and str(rxn.gpr) not in spontaneous])\n",
    "    n_mets = len({x[2:-2] for x in model.metabolites})\n",
    "    data.append((organism, 'gut_normal', n_genes, n_enzrxns, n_mets))\n",
    "    \n",
    "df = pd.DataFrame(data, columns=['Organism', 'Source', 'Genes', 'Reactions', 'Metabolites'])\n",
    "\n",
    "### Plot\n",
    "\n",
    "f, (ax1, ax2, ax3) = plt.subplots(3, sharex=True, figsize=(5, 4.5))\n",
    "\n",
    "sns.barplot(x=\"Genes\", y=\"Organism\", hue=\"Source\", data=df, palette='deep', ax=ax1, alpha=0.8)\n",
    "ax1.set_ylabel('')\n",
    "legend_handles, legend_text = ax1.get_legend_handles_labels()\n",
    "ax1.legend(legend_text, bbox_to_anchor=(1.1, 1), ncol=3, handles=legend_handles)\n",
    "\n",
    "sns.barplot(x=\"Reactions\", y=\"Organism\", hue=\"Source\", data=df, palette='deep', ax=ax2, alpha=0.8)\n",
    "ax2.get_legend().set_title('')\n",
    "#ax2.set_xlabel('')\n",
    "ax2.legend([])\n",
    "\n",
    "res = sns.barplot(x=\"Metabolites\", y=\"Organism\", hue=\"Source\", data=df, palette='deep', ax=ax3, alpha=0.8)\n",
    "#ax3.set_xticklabels(ax3.get_xticklabels(), rotation=90)\n",
    "ax3.set_ylabel('')\n",
    "ax3.legend([])\n",
    "\n",
    "plt.savefig('../plots/model_summary/gut_normal.png', dpi=300, bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 🍰 Impaired gut community"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-4-431150b4456b>:38: UserWarning: You have mixed positional and keyword arguments, some input may be discarded.\n",
      "  ax1.legend(legend_text, bbox_to_anchor=(1.1, 1), ncol=3, handles=legend_handles)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x324 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "### Gut community members\n",
    "organisms = ['B. uniformis', 'R. bromii', 'B. wexlerae', 'E. rectale', 'F. saccharivorans']\n",
    "\n",
    "### Load gut impaired models\n",
    "gut_impaired = {\n",
    "    'B. uniformis': ('../models/gut_impaired/ERR260172_bin.10.p.xml', 'cobra'),\n",
    "    'R. bromii': ('../models/gut_impaired/ERR260172_bin.44.p.xml', 'cobra'),\n",
    "    'B. wexlerae': ('../models/gut_impaired/ERR260172_bin.7.s.xml', 'cobra'),\n",
    "    'E. rectale': ('../models/gut_impaired/ERR260172_bin.31.s.xml', 'cobra'),\n",
    "    'F. saccharivorans': ('../models/gut_impaired/ERR260172_bin.34.p.xml', 'cobra')\n",
    "}\n",
    "\n",
    "for label, (model_file, flavor) in gut_impaired.items():\n",
    "    gut_impaired[label] =  load_cbmodel(model_file, flavor=flavor)\n",
    "    \n",
    "data = []\n",
    "spontaneous = {'G_spontaneous', 'G_Unknown', 'G_s0001', 'G_NoAssignment'} \n",
    "\n",
    "for organism in organisms:\n",
    "    \n",
    "    model = gut_impaired[organism]\n",
    "    n_genes = len(set(model.genes) - spontaneous)\n",
    "    #n_rxns = len([x for x in model.reactions if not x.startswith('R_EX_')])\n",
    "    n_enzrxns = len([x for x, rxn in model.reactions.items()\n",
    "                     if rxn.gpr is not None and str(rxn.gpr) not in spontaneous])\n",
    "    n_mets = len({x[2:-2] for x in model.metabolites})\n",
    "    data.append((organism, 'gut_impaired', n_genes, n_enzrxns, n_mets))\n",
    "    \n",
    "df = pd.DataFrame(data, columns=['Organism', 'Source', 'Genes', 'Reactions', 'Metabolites'])\n",
    "\n",
    "### Plot\n",
    "\n",
    "f, (ax1, ax2, ax3) = plt.subplots(3, sharex=True, figsize=(5, 4.5))\n",
    "\n",
    "sns.barplot(x=\"Genes\", y=\"Organism\", hue=\"Source\", data=df, palette='deep', ax=ax1, alpha=0.8)\n",
    "ax1.set_ylabel('')\n",
    "legend_handles, legend_text = ax1.get_legend_handles_labels()\n",
    "ax1.legend(legend_text, bbox_to_anchor=(1.1, 1), ncol=3, handles=legend_handles)\n",
    "\n",
    "sns.barplot(x=\"Reactions\", y=\"Organism\", hue=\"Source\", data=df, palette='deep', ax=ax2, alpha=0.8)\n",
    "ax2.get_legend().set_title('')\n",
    "#ax2.set_xlabel('')\n",
    "ax2.legend([])\n",
    "\n",
    "res = sns.barplot(x=\"Metabolites\", y=\"Organism\", hue=\"Source\", data=df, palette='deep', ax=ax3, alpha=0.8)\n",
    "#ax3.set_xticklabels(ax3.get_xticklabels(), rotation=90)\n",
    "ax3.set_ylabel('')\n",
    "ax3.legend([])\n",
    "\n",
    "plt.savefig('../plots/model_summary/gut_impaired.png', dpi=300, bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 🏥 T2D gut community"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-5-03e9c6147ef2>:38: UserWarning: You have mixed positional and keyword arguments, some input may be discarded.\n",
      "  ax1.legend(legend_text, bbox_to_anchor=(1.1, 1), ncol=3, handles=legend_handles)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x324 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "### Gut community members\n",
    "organisms = ['B. uniformis', 'R. bromii', 'B. wexlerae', 'E. rectale', 'F. saccharivorans']\n",
    "\n",
    "### Load gut t2d models\n",
    "gut_t2d = {\n",
    "    'B. uniformis': ('../models/gut_t2d/ERR260140_bin.3.s.xml', 'cobra'),\n",
    "    'R. bromii': ('../models/gut_t2d/ERR260140_bin.19.p.xml', 'cobra'),\n",
    "    'B. wexlerae': ('../models/gut_t2d/ERR260140_bin.13.s.xml', 'cobra'),\n",
    "    'E. rectale': ('../models/gut_t2d/ERR260140_bin.11.s.xml', 'cobra'),\n",
    "    'F. saccharivorans': ('../models/gut_t2d/ERR260140_bin.10.p.xml', 'cobra')\n",
    "}\n",
    "\n",
    "for label, (model_file, flavor) in gut_t2d.items():\n",
    "    gut_t2d[label] =  load_cbmodel(model_file, flavor=flavor)\n",
    "    \n",
    "data = []\n",
    "spontaneous = {'G_spontaneous', 'G_Unknown', 'G_s0001', 'G_NoAssignment'} \n",
    "\n",
    "for organism in organisms:\n",
    "    \n",
    "    model = gut_t2d[organism]\n",
    "    n_genes = len(set(model.genes) - spontaneous)\n",
    "    #n_rxns = len([x for x in model.reactions if not x.startswith('R_EX_')])\n",
    "    n_enzrxns = len([x for x, rxn in model.reactions.items()\n",
    "                     if rxn.gpr is not None and str(rxn.gpr) not in spontaneous])\n",
    "    n_mets = len({x[2:-2] for x in model.metabolites})\n",
    "    data.append((organism, 'gut_t2d', n_genes, n_enzrxns, n_mets))\n",
    "    \n",
    "df = pd.DataFrame(data, columns=['Organism', 'Source', 'Genes', 'Reactions', 'Metabolites'])\n",
    "\n",
    "### Plot\n",
    "\n",
    "f, (ax1, ax2, ax3) = plt.subplots(3, sharex=True, figsize=(5, 4.5))\n",
    "\n",
    "sns.barplot(x=\"Genes\", y=\"Organism\", hue=\"Source\", data=df, palette='deep', ax=ax1, alpha=0.8)\n",
    "ax1.set_ylabel('')\n",
    "legend_handles, legend_text = ax1.get_legend_handles_labels()\n",
    "ax1.legend(legend_text, bbox_to_anchor=(1.1, 1), ncol=3, handles=legend_handles)\n",
    "\n",
    "sns.barplot(x=\"Reactions\", y=\"Organism\", hue=\"Source\", data=df, palette='deep', ax=ax2, alpha=0.8)\n",
    "ax2.get_legend().set_title('')\n",
    "#ax2.set_xlabel('')\n",
    "ax2.legend([])\n",
    "\n",
    "res = sns.barplot(x=\"Metabolites\", y=\"Organism\", hue=\"Source\", data=df, palette='deep', ax=ax3, alpha=0.8)\n",
    "#ax3.set_xticklabels(ax3.get_xticklabels(), rotation=90)\n",
    "ax3.set_ylabel('')\n",
    "ax3.legend([])\n",
    "\n",
    "plt.savefig('../plots/model_summary/gut_t2d.png', dpi=300, bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 🤖 RefSeq genome based community"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-6-5ff0471d6fb7>:35: UserWarning: You have mixed positional and keyword arguments, some input may be discarded.\n",
      "  ax1.legend(legend_text, bbox_to_anchor=(1.1, 1), ncol=3, handles=legend_handles)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x324 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "### Load gut refseq models\n",
    "gut_refseq = {\n",
    "    'B. uniformis': ('../models/gut_refseq/GCF_018292165.1_ASM1829216v1_protein.xml', 'cobra'),\n",
    "    'R. bromii': ('../models/gut_refseq/GCF_002834225.1_ASM283422v1_protein.xml', 'cobra'),\n",
    "    'B. wexlerae': ('../models/gut_refseq/GCF_018785545.1_ASM1878554v1_protein.xml', 'cobra'),\n",
    "    'E. rectale': ('../models/gut_refseq/GCF_000020605.1_ASM2060v1_protein.xml', 'cobra'),\n",
    "    'F. saccharivorans': ('../models/gut_refseq/GCF_015557635.1_ASM1555763v1_protein.xml', 'cobra')\n",
    "}\n",
    "\n",
    "for label, (model_file, flavor) in gut_refseq.items():\n",
    "    gut_refseq[label] =  load_cbmodel(model_file, flavor=flavor)\n",
    "\n",
    "data = []\n",
    "spontaneous = {'G_spontaneous', 'G_Unknown', 'G_s0001', 'G_NoAssignment'} \n",
    "\n",
    "for organism in organisms:\n",
    "    \n",
    "    model = gut_refseq[organism]\n",
    "    n_genes = len(set(model.genes) - spontaneous)\n",
    "    #n_rxns = len([x for x in model.reactions if not x.startswith('R_EX_')])\n",
    "    n_enzrxns = len([x for x, rxn in model.reactions.items()\n",
    "                     if rxn.gpr is not None and str(rxn.gpr) not in spontaneous])\n",
    "    n_mets = len({x[2:-2] for x in model.metabolites})\n",
    "    data.append((organism, 'gut_refseq', n_genes, n_enzrxns, n_mets))\n",
    "    \n",
    "df = pd.DataFrame(data, columns=['Organism', 'Source', 'Genes', 'Reactions', 'Metabolites'])\n",
    "\n",
    "### Plot\n",
    "\n",
    "f, (ax1, ax2, ax3) = plt.subplots(3, sharex=True, figsize=(5, 4.5))\n",
    "\n",
    "sns.barplot(x=\"Genes\", y=\"Organism\", hue=\"Source\", data=df, palette='deep', ax=ax1, alpha=0.8)\n",
    "ax1.set_ylabel('')\n",
    "legend_handles, legend_text = ax1.get_legend_handles_labels()\n",
    "ax1.legend(legend_text, bbox_to_anchor=(1.1, 1), ncol=3, handles=legend_handles)\n",
    "\n",
    "sns.barplot(x=\"Reactions\", y=\"Organism\", hue=\"Source\", data=df, palette='deep', ax=ax2, alpha=0.8)\n",
    "ax2.get_legend().set_title('')\n",
    "#ax2.set_xlabel('')\n",
    "ax2.legend([])\n",
    "\n",
    "res = sns.barplot(x=\"Metabolites\", y=\"Organism\", hue=\"Source\", data=df, palette='deep', ax=ax3, alpha=0.8)\n",
    "#ax3.set_xticklabels(ax3.get_xticklabels(), rotation=90)\n",
    "ax3.set_ylabel('')\n",
    "ax3.legend([])\n",
    "\n",
    "plt.savefig('../plots/model_summary/gut_refseq.png', dpi=300, bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 🥛 Kefir community"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-7-3d5255ea0f66>:38: UserWarning: You have mixed positional and keyword arguments, some input may be discarded.\n",
      "  ax1.legend(legend_text, bbox_to_anchor=(1.1, 1), ncol=3, handles=legend_handles)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x324 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "### Kefir community members\n",
    "organisms = ['L. mesenteroides', 'L. lactis', 'A. fabarum', 'L. kefiranofaciens', 'L. kefiri']\n",
    "\n",
    "### Load kefir refseq models\n",
    "kefir = {\n",
    "    'L. mesenteroides': ('../models/kefir/230a_03182016.xml','cobra'),\n",
    "    'L. lactis': ('../models/kefir/261_03182016.xml','cobra'),\n",
    "    'A. fabarum': ('../models/kefir/290_03182016.xml','cobra'),\n",
    "    'L. kefiranofaciens': ('../models/kefir/322b_03182016.xml','cobra'),\n",
    "    'L. kefiri': ('../models/kefir/376_03182016.xml','cobra')\n",
    "}\n",
    "\n",
    "for label, (model_file, flavor) in kefir.items():\n",
    "    kefir[label] =  load_cbmodel(model_file, flavor=flavor)\n",
    "\n",
    "data = []\n",
    "spontaneous = {'G_spontaneous', 'G_Unknown', 'G_s0001', 'G_NoAssignment'} \n",
    "\n",
    "for organism in organisms:\n",
    "    \n",
    "    model = kefir[organism]\n",
    "    n_genes = len(set(model.genes) - spontaneous)\n",
    "    #n_rxns = len([x for x in model.reactions if not x.startswith('R_EX_')])\n",
    "    n_enzrxns = len([x for x, rxn in model.reactions.items()\n",
    "                     if rxn.gpr is not None and str(rxn.gpr) not in spontaneous])\n",
    "    n_mets = len({x[2:-2] for x in model.metabolites})\n",
    "    data.append((organism, 'kefir', n_genes, n_enzrxns, n_mets))\n",
    "    \n",
    "df = pd.DataFrame(data, columns=['Organism', 'Source', 'Genes', 'Reactions', 'Metabolites'])\n",
    "\n",
    "### Plot\n",
    "\n",
    "f, (ax1, ax2, ax3) = plt.subplots(3, sharex=True, figsize=(5, 4.5))\n",
    "\n",
    "sns.barplot(x=\"Genes\", y=\"Organism\", hue=\"Source\", data=df, palette='deep', ax=ax1, alpha=0.8)\n",
    "ax1.set_ylabel('')\n",
    "legend_handles, legend_text = ax1.get_legend_handles_labels()\n",
    "ax1.legend(legend_text, bbox_to_anchor=(1.1, 1), ncol=3, handles=legend_handles)\n",
    "\n",
    "sns.barplot(x=\"Reactions\", y=\"Organism\", hue=\"Source\", data=df, palette='deep', ax=ax2, alpha=0.8)\n",
    "ax2.get_legend().set_title('')\n",
    "#ax2.set_xlabel('')\n",
    "ax2.legend([])\n",
    "\n",
    "res = sns.barplot(x=\"Metabolites\", y=\"Organism\", hue=\"Source\", data=df, palette='deep', ax=ax3, alpha=0.8)\n",
    "#ax3.set_xticklabels(ax3.get_xticklabels(), rotation=90)\n",
    "ax3.set_ylabel('')\n",
    "ax3.legend([])\n",
    "\n",
    "plt.savefig('../plots/model_summary/kefir.png', dpi=300, bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 🪱 Soil community"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-8-0577bdab9749>:38: UserWarning: You have mixed positional and keyword arguments, some input may be discarded.\n",
      "  ax1.legend(legend_text, bbox_to_anchor=(1.1, 1), ncol=3, handles=legend_handles)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x324 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "### Soil community members\n",
    "organisms = ['f__Thermoleophilaceae;g__', 'f__Herpetosiphonaceae;g__', 'f__Phormidiaceae;g__Microcoleus', 'f__Geodermatophilaceae;g__Geodermatophilus', 'f__Rubrobacteraceae;g__Rubrobacter_A']\n",
    "\n",
    "## soil models\n",
    "soil = {\n",
    "    'f__Thermoleophilaceae;g__': ('../ensembles/ERR671933_bin.1.o.xml','cobra'),\n",
    "    'f__Herpetosiphonaceae;g__': ('../ensembles/ERR671933_bin.2.o.xml','cobra'),\n",
    "    'f__Phormidiaceae;g__Microcoleus': ('../ensembles/ERR671933_bin.3.s.xml','cobra'),\n",
    "    'f__Geodermatophilaceae;g__Geodermatophilus': ('../ensembles/ERR671933_bin.4.o.xml','cobra'),\n",
    "    'f__Rubrobacteraceae;g__Rubrobacter_A': ('../ensembles/ERR671933_bin.5.s.xml','cobra')\n",
    "}\n",
    "\n",
    "for label, (model_file, flavor) in soil.items():\n",
    "    soil[label] =  load_cbmodel(model_file, flavor=flavor)\n",
    "    \n",
    "data = []\n",
    "spontaneous = {'G_spontaneous', 'G_Unknown', 'G_s0001', 'G_NoAssignment'} \n",
    "\n",
    "for organism in organisms:\n",
    "    \n",
    "    model = soil[organism]\n",
    "    n_genes = len(set(model.genes) - spontaneous)\n",
    "    #n_rxns = len([x for x in model.reactions if not x.startswith('R_EX_')])\n",
    "    n_enzrxns = len([x for x, rxn in model.reactions.items()\n",
    "                     if rxn.gpr is not None and str(rxn.gpr) not in spontaneous])\n",
    "    n_mets = len({x[2:-2] for x in model.metabolites})\n",
    "    data.append((organism, 'soil', n_genes, n_enzrxns, n_mets))\n",
    "    \n",
    "df = pd.DataFrame(data, columns=['Organism', 'Source', 'Genes', 'Reactions', 'Metabolites'])\n",
    "\n",
    "### Plot\n",
    "\n",
    "f, (ax1, ax2, ax3) = plt.subplots(3, sharex=True, figsize=(5, 4.5))\n",
    "\n",
    "sns.barplot(x=\"Genes\", y=\"Organism\", hue=\"Source\", data=df, palette='deep', ax=ax1, alpha=0.8)\n",
    "ax1.set_ylabel('')\n",
    "legend_handles, legend_text = ax1.get_legend_handles_labels()\n",
    "ax1.legend(legend_text, bbox_to_anchor=(1.1, 1), ncol=3, handles=legend_handles)\n",
    "\n",
    "sns.barplot(x=\"Reactions\", y=\"Organism\", hue=\"Source\", data=df, palette='deep', ax=ax2, alpha=0.8)\n",
    "ax2.get_legend().set_title('')\n",
    "#ax2.set_xlabel('')\n",
    "ax2.legend([])\n",
    "\n",
    "res = sns.barplot(x=\"Metabolites\", y=\"Organism\", hue=\"Source\", data=df, palette='deep', ax=ax3, alpha=0.8)\n",
    "#ax3.set_xticklabels(ax3.get_xticklabels(), rotation=90)\n",
    "ax3.set_ylabel('')\n",
    "ax3.legend([])\n",
    "\n",
    "plt.savefig('../plots/model_summary/soil.png', dpi=300, bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 🧠 Compare gut models across datasets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-9-c0728b392f06>:49: UserWarning: You have mixed positional and keyword arguments, some input may be discarded.\n",
      "  ax1.legend(legend_text, bbox_to_anchor=(1.1, 1), ncol=3, handles=legend_handles)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x324 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "organisms = ['B. uniformis', 'R. bromii', 'B. wexlerae', 'E. rectale', 'F. saccharivorans']\n",
    "\n",
    "data = []\n",
    "spontaneous = {'G_spontaneous', 'G_Unknown', 'G_s0001', 'G_NoAssignment'} \n",
    "\n",
    "for organism in organisms:\n",
    "    \n",
    "    model = gut_normal[organism]\n",
    "    n_genes = len(set(model.genes) - spontaneous)\n",
    "    #n_rxns = len([x for x in model.reactions if not x.startswith('R_EX_')])\n",
    "    n_enzrxns = len([x for x, rxn in model.reactions.items()\n",
    "                     if rxn.gpr is not None and str(rxn.gpr) not in spontaneous])\n",
    "    n_mets = len({x[2:-2] for x in model.metabolites})\n",
    "    data.append((organism, 'gut_normal', n_genes, n_enzrxns, n_mets))\n",
    "    \n",
    "    model = gut_impaired[organism]\n",
    "    n_genes = len(set(model.genes) - spontaneous)\n",
    "    #n_rxns = len([x for x in model.reactions if not x.startswith('R_EX_')])\n",
    "    n_enzrxns = len([x for x, rxn in model.reactions.items()\n",
    "                     if rxn.gpr is not None and str(rxn.gpr) not in spontaneous])\n",
    "    n_mets = len({x[2:-2] for x in model.metabolites})\n",
    "    data.append((organism, 'gut_impaired', n_genes, n_enzrxns, n_mets))\n",
    "    \n",
    "    model = gut_t2d[organism]\n",
    "    n_genes = len(set(model.genes) - spontaneous)\n",
    "    #n_rxns = len([x for x in model.reactions if not x.startswith('R_EX_')])\n",
    "    n_enzrxns = len([x for x, rxn in model.reactions.items()\n",
    "                     if rxn.gpr is not None and str(rxn.gpr) not in spontaneous])\n",
    "    n_mets = len({x[2:-2] for x in model.metabolites})\n",
    "    data.append((organism, 'gut_t2d', n_genes, n_enzrxns, n_mets))\n",
    "    \n",
    "    model = gut_refseq[organism]\n",
    "    n_genes = len(set(model.genes) - spontaneous)\n",
    "    #n_rxns = len([x for x in model.reactions if not x.startswith('EX_')])\n",
    "    n_enzrxns = len([x for x, rxn in model.reactions.items()\n",
    "                     if rxn.gpr is not None and str(rxn.gpr) not in spontaneous])\n",
    "    n_mets = len({x[2:-2] for x in model.metabolites})\n",
    "    data.append((organism, 'gut_refseq', n_genes, n_enzrxns, n_mets))\n",
    "    \n",
    "df = pd.DataFrame(data, columns=['Organism', 'Source', 'Genes', 'Reactions', 'Metabolites'])\n",
    "\n",
    "### Plot\n",
    "\n",
    "f, (ax1, ax2, ax3) = plt.subplots(3, sharex=True, figsize=(5, 4.5))\n",
    "\n",
    "sns.barplot(x=\"Organism\", y=\"Genes\", hue=\"Source\", data=df, palette='deep', ax=ax1, alpha=0.8)\n",
    "ax1.set_xlabel('')\n",
    "legend_handles, legend_text = ax1.get_legend_handles_labels()\n",
    "ax1.legend(legend_text, bbox_to_anchor=(1.1, 1), ncol=3, handles=legend_handles)\n",
    "\n",
    "\n",
    "sns.barplot(x=\"Organism\", y=\"Reactions\", hue=\"Source\", data=df, palette='deep', ax=ax2, alpha=0.8)\n",
    "ax2.get_legend().set_title('')\n",
    "ax2.set_xlabel('')\n",
    "ax2.legend([])\n",
    "\n",
    "res = sns.barplot(x=\"Organism\", y=\"Metabolites\", hue=\"Source\", data=df, palette='deep', ax=ax3, alpha=0.8)\n",
    "ax3.set_xticklabels(ax3.get_xticklabels(), rotation=45)\n",
    "ax3.set_xlabel('')\n",
    "ax3.legend([])\n",
    "\n",
    "plt.savefig('../plots/model_summary/gut_all.png', dpi=300, bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 🔬 Discussion questions\n",
    "\n",
    " - Approximately how many genes do you expect to find in a typical prokaryotic genome? What percentage do you expect to be metabolism encoding?\n",
    " - Are the number of metabolites, reactions, and genes in your community's models comparable to other metabolic reconstructions?\n",
    " - Is there variation in the metabolism of conspecific models reconstructed from different metagenomes?"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Move on to [exercise 3](https://github.com/franciscozorrilla/SymbNET/blob/main/scripts/3.run_smetana_detailed_interactions.md)"
   ]
  }
 ],
 "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": "3.8.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}