demo/DIVERS_analysis.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## DIVERS demo\n",
    "### overview\n",
    "this is a quick guide demonstrating simple analysis of data generated by DIVERS. we will analyze fecal samples spike-in sequencing data (Fig. 1-3 in the manuscript)."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### input\n",
    "* **list of sample IDs:** [../test_data/test.sample_list.txt]\n",
    "* **matrix of OTU reads counts (exclude spike-in strain):** [../test_data/test.OTU_readsCount.csv]\n",
    "* **table of sample weights:** [../test_data/test.sample_weight.tsv]\n",
    "* **table of spike-in strain reads counts:** [../test_data/test.spikein_readsCount.tsv]\n",
    "* **table of OTU taxonomy annotation:** [../test_data/test.taxonomy.tsv]\n",
    "* **configure file for DIVERS:** [../test_data/test.sample_info.variance_3.config]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### absolute abundances estimation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Write total bacterial density to ../test_output/test.total_density.tsv\n",
      "Write relative abundance to ../test_output/test.relative_abundance.csv\n",
      "Write absolute abundance to ../test_output/test.absolute_abundance.csv\n",
      "Complete!\n"
     ]
    }
   ],
   "source": [
    "! mkdir -p ../test_output\n",
    "! chmod +x ../script/calculate_absolute_abundance.R\n",
    "\n",
    "! ../script/calculate_absolute_abundance.R -s ../test_data/test.sample_list.txt \\\n",
    "                                        -i ../test_data/test.OTU_readsCount.csv \\\n",
    "                                        -w ../test_data/test.sample_weight.tsv \\\n",
    "                                        -p ../test_data/test.spikein_readsCount.tsv \\\n",
    "                                        -o ../test_output/test \\\n",
    "                                        -r"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### variance and covariance decomposition of absolute bacterial abundances using DIVERS"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "DIVERS will perform variance decomposition for sample X, Y and Z in configure file\n",
      "\n",
      "0.Check configure file...\n",
      "Complete!\n",
      "\n",
      "1.Calculate marginal means and variances of each OTU...\n",
      "  processing [============================] 500/500 eta:  0s\n",
      "Complete!\n",
      "\n",
      "2.Variance decomposition of OTU abundances...\n",
      "  processing [============================] 500/500 eta:  0s\n",
      "  write variance decomposition result to ../test_output/test.variance_3.variance_decomposition.tsv\n",
      "Complete!\n",
      "\n",
      "3.Calculate total covariances...\n",
      "  processing [============================] 500/500 eta:  0s\n",
      "Complete!\n",
      "\n",
      "4.Covariance decomposition of OTU abundances...\n",
      "  processing [============================] 500/500 eta:  0s\n",
      "  write correlation decomposition results to ../test_output/test.variance_3.correlation_decomposition_*.csv\n",
      "Complete!\n",
      "\n",
      "DIVERS complete!\n",
      "\n"
     ]
    }
   ],
   "source": [
    "! chmod +x ../script/DIVERS.R\n",
    "\n",
    "# For temporal, spatial and technical variance decomposition\n",
    "! ../script/DIVERS.R -i ../test_output/test.absolute_abundance.csv \\\n",
    "                    -c ../test_data/test.sample_info.variance_3.config \\\n",
    "                    -o ../test_output/test.variance_3 \\\n",
    "                    -v 3 \\\n",
    "                    -t 1e-4 \\\n",
    "                    -n 500"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### load libraries\n",
    "load the utility functions as well as a few python libraries for analysis, visualization, etc."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# load libraries\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt # Plotting is done using matplotlib.pyplot\n",
    "from matplotlib.colors import LinearSegmentedColormap\n",
    "from matplotlib.ticker import StrMethodFormatter\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### set data directory\n",
    "set the directory from which the input files will be drawn from, including the prefix of DIVERS results and the taxonomy annotation of these OTUs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "DIVERS_output_prefix = \"../test_output/test.variance_3\"\n",
    "taxonomy_annotation_path = \"../test_data/test.taxonomy.tsv\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### load taxonomy annotation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>OTU</th>\n",
       "      <th>phlum</th>\n",
       "      <th>class</th>\n",
       "      <th>order</th>\n",
       "      <th>family</th>\n",
       "      <th>genus</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>otu_1</td>\n",
       "      <td>Firmicutes</td>\n",
       "      <td>Clostridia</td>\n",
       "      <td>Clostridiales</td>\n",
       "      <td>Lachnospiraceae</td>\n",
       "      <td>Roseburia</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>otu_2</td>\n",
       "      <td>Firmicutes</td>\n",
       "      <td>Clostridia</td>\n",
       "      <td>Clostridiales</td>\n",
       "      <td>Lachnospiraceae</td>\n",
       "      <td>Blautia</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>otu_3</td>\n",
       "      <td>Firmicutes</td>\n",
       "      <td>Clostridia</td>\n",
       "      <td>Clostridiales</td>\n",
       "      <td>Lachnospiraceae</td>\n",
       "      <td>Blautia</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>otu_4</td>\n",
       "      <td>Firmicutes</td>\n",
       "      <td>Clostridia</td>\n",
       "      <td>Clostridiales</td>\n",
       "      <td>Lachnospiraceae</td>\n",
       "      <td>Blautia</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>otu_5</td>\n",
       "      <td>Firmicutes</td>\n",
       "      <td>Clostridia</td>\n",
       "      <td>Clostridiales</td>\n",
       "      <td>Ruminococcaceae</td>\n",
       "      <td>Faecalibacterium</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     OTU       phlum       class          order           family  \\\n",
       "0  otu_1  Firmicutes  Clostridia  Clostridiales  Lachnospiraceae   \n",
       "1  otu_2  Firmicutes  Clostridia  Clostridiales  Lachnospiraceae   \n",
       "2  otu_3  Firmicutes  Clostridia  Clostridiales  Lachnospiraceae   \n",
       "3  otu_4  Firmicutes  Clostridia  Clostridiales  Lachnospiraceae   \n",
       "4  otu_5  Firmicutes  Clostridia  Clostridiales  Ruminococcaceae   \n",
       "\n",
       "              genus  \n",
       "0         Roseburia  \n",
       "1           Blautia  \n",
       "2           Blautia  \n",
       "3           Blautia  \n",
       "4  Faecalibacterium  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def removeQuote(string):\n",
    "    if string[0] == \"\\\"\":\n",
    "        return string[1:-1]\n",
    "    else:\n",
    "        return string\n",
    "    \n",
    "with open(taxonomy_annotation_path) as f:\n",
    "    data_taxonomy = pd.read_table(f, sep = '\\t', lineterminator = '\\n')\n",
    "taxonomy_Pool = []\n",
    "for i in range(data_taxonomy.shape[0]):\n",
    "    tmpList = data_taxonomy[\"taxonomy\"][i].split(\"; \")\n",
    "    if len(tmpList) == 14:\n",
    "        tmpOut = [removeQuote(tmpList[4]), removeQuote(tmpList[6]), removeQuote(tmpList[8]), removeQuote(tmpList[10]), removeQuote(tmpList[12])]\n",
    "    elif len(tmpList) == 16:\n",
    "        tmpOut = [removeQuote(tmpList[4]), removeQuote(tmpList[6]), removeQuote(tmpList[10]), removeQuote(tmpList[12]), removeQuote(tmpList[14])]\n",
    "    elif len(tmpList) == 18:\n",
    "        tmpOut = [removeQuote(tmpList[4]), removeQuote(tmpList[6]), removeQuote(tmpList[10]), removeQuote(tmpList[14]), removeQuote(tmpList[16])]\n",
    "    else:\n",
    "        tmpOut = [\"\",] * 5\n",
    "    taxonomy_Pool.append([data_taxonomy[\"OTU\"][i],] + tmpOut)\n",
    "df_taxonomy = pd.DataFrame(taxonomy_Pool)\n",
    "df_taxonomy.columns = [\"OTU\",\"phlum\",\"class\",\"order\",\"family\",\"genus\"]\n",
    "df_taxonomy.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### load variance decompostion result"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>OTU_ID</th>\n",
       "      <th>Average_abundances</th>\n",
       "      <th>Total_variances</th>\n",
       "      <th>Temporal_variances</th>\n",
       "      <th>Spatial_variances</th>\n",
       "      <th>Technical_variances</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>otu_1</td>\n",
       "      <td>0.145453</td>\n",
       "      <td>0.007708</td>\n",
       "      <td>0.005961</td>\n",
       "      <td>0.001012</td>\n",
       "      <td>0.000869</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>otu_2</td>\n",
       "      <td>0.097650</td>\n",
       "      <td>0.001766</td>\n",
       "      <td>0.001197</td>\n",
       "      <td>0.000515</td>\n",
       "      <td>0.000203</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>otu_3</td>\n",
       "      <td>0.058688</td>\n",
       "      <td>0.000357</td>\n",
       "      <td>0.000216</td>\n",
       "      <td>0.000030</td>\n",
       "      <td>0.000075</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>otu_4</td>\n",
       "      <td>0.040297</td>\n",
       "      <td>0.000488</td>\n",
       "      <td>0.000346</td>\n",
       "      <td>0.000086</td>\n",
       "      <td>0.000026</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>otu_5</td>\n",
       "      <td>0.091128</td>\n",
       "      <td>0.005634</td>\n",
       "      <td>0.004873</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000152</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  OTU_ID  Average_abundances  Total_variances  Temporal_variances  \\\n",
       "0  otu_1            0.145453         0.007708            0.005961   \n",
       "1  otu_2            0.097650         0.001766            0.001197   \n",
       "2  otu_3            0.058688         0.000357            0.000216   \n",
       "3  otu_4            0.040297         0.000488            0.000346   \n",
       "4  otu_5            0.091128         0.005634            0.004873   \n",
       "\n",
       "   Spatial_variances  Technical_variances  \n",
       "0           0.001012             0.000869  \n",
       "1           0.000515             0.000203  \n",
       "2           0.000030             0.000075  \n",
       "3           0.000086             0.000026  \n",
       "4           0.000000             0.000152  "
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "with open(DIVERS_output_prefix + \".variance_decomposition.tsv\") as f:\n",
    "    data_variance_decomposition = pd.read_table(f, sep = '\\t', lineterminator = '\\n')\n",
    "\n",
    "data_variance_decomposition.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### bin OTUs according to average OTU abundance\n",
    "we will bin OTUs according to their average abundance aross different samples and calculate decomposed variances for each bin"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "b_w = 1\n",
    "bins = [i for i in range(-7, -1, b_w)]\n",
    "means_var_fraction_T = []\n",
    "means_var_fraction_S = []\n",
    "means_var_fraction_N = []\n",
    "sem_var_fraction_T = []\n",
    "sem_var_fraction_S = []\n",
    "sem_var_fraction_N = []\n",
    "for i in range(len(bins)):\n",
    "    if i < len(bins) - 1:\n",
    "        tmpDF = data_variance_decomposition[(data_variance_decomposition[\"Average_abundances\"] > 10 ** bins[i]) &\n",
    "                                            (data_variance_decomposition[\"Average_abundances\"] <= 10 ** bins[i + 1])]\n",
    "    else:\n",
    "        tmpDF = data_variance_decomposition[(data_variance_decomposition[\"Average_abundances\"] > 10 ** bins[i])]\n",
    "\n",
    "    means_var_fraction_T.append(np.nanmean(tmpDF[\"Temporal_variances\"] / (tmpDF[\"Temporal_variances\"] + tmpDF[\"Spatial_variances\"] + tmpDF[\"Technical_variances\"])))\n",
    "    means_var_fraction_S.append(np.nanmean(tmpDF[\"Spatial_variances\"] / (tmpDF[\"Temporal_variances\"] + tmpDF[\"Spatial_variances\"] + tmpDF[\"Technical_variances\"])))\n",
    "    means_var_fraction_N.append(np.nanmean(tmpDF[\"Technical_variances\"] / (tmpDF[\"Temporal_variances\"] + tmpDF[\"Spatial_variances\"] + tmpDF[\"Technical_variances\"])))\n",
    "    sem_var_fraction_T.append(np.nanstd(tmpDF[\"Temporal_variances\"] / (tmpDF[\"Temporal_variances\"] + tmpDF[\"Spatial_variances\"] + tmpDF[\"Technical_variances\"])) / np.sqrt(tmpDF.shape[0]))\n",
    "    sem_var_fraction_S.append(np.nanstd(tmpDF[\"Spatial_variances\"] / (tmpDF[\"Temporal_variances\"] + tmpDF[\"Spatial_variances\"] + tmpDF[\"Technical_variances\"])) / np.sqrt(tmpDF.shape[0]))\n",
    "    sem_var_fraction_N.append(np.nanstd(tmpDF[\"Technical_variances\"] / (tmpDF[\"Temporal_variances\"] + tmpDF[\"Spatial_variances\"] + tmpDF[\"Technical_variances\"])) / np.sqrt(tmpDF.shape[0]))\n",
    "\n",
    "array_means_var_fraction_T = np.array(means_var_fraction_T)\n",
    "array_means_var_fraction_S = np.array(means_var_fraction_S)\n",
    "array_means_var_fraction_N = np.array(means_var_fraction_N)\n",
    "array_sem_var_fraction_T = np.array(sem_var_fraction_T)\n",
    "array_sem_var_fraction_S = np.array(sem_var_fraction_S)\n",
    "array_sem_var_fraction_N = np.array(sem_var_fraction_N)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### plot variance fraction decomposition as a function of average abundance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAikAAAFECAYAAADiEa4HAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAMBIIgUAGEmkAAAjiRQAYKQXPmQIIvkJQuQAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x101ec7f60>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig1, ax1 = plt.subplots(figsize=(6,5))\n",
    "ax1.spines['top'].set_color('none')\n",
    "ax1.spines['right'].set_color('none')\n",
    "ax1.tick_params(axis='both', which='major', labelsize = 11)\n",
    "plt.rcParams[\"errorbar.capsize\"] = 2\n",
    "ax1.set(xlim = (-7.2, -0.8))\n",
    "width = 0.5\n",
    "\n",
    "x_pos = [e + b_w / 2.0  for e in bins]\n",
    "\n",
    "p1 = plt.bar(x_pos, array_means_var_fraction_N, width, \n",
    "             yerr = array_sem_var_fraction_N, color = \"#8b579b\", \n",
    "             linewidth = [1.2,] * len(bins), edgecolor = [\"black\",] * len(bins))\n",
    "\n",
    "p2 = plt.bar(x_pos, array_means_var_fraction_S, width,\n",
    "             bottom = array_means_var_fraction_N, yerr = array_sem_var_fraction_S, color = \"#1eb3ec\",\n",
    "             linewidth = [1.2,] * len(bins), edgecolor = [\"black\",] * len(bins))\n",
    "\n",
    "p3 = plt.bar(x_pos, array_means_var_fraction_T, width,\n",
    "             bottom = array_means_var_fraction_N + array_means_var_fraction_S, \n",
    "             yerr = array_sem_var_fraction_T, color = \"#e82d35\",\n",
    "             linewidth = [1.2,] * len(bins), edgecolor = [\"black\",] * len(bins))\n",
    "\n",
    "plt.ylabel('Variance Fraction', fontsize = 15, fontname=\"Arial\")\n",
    "plt.xlabel('log10(Mean absolute abundance (a.u.))', fontsize = 15, fontname=\"Arial\")\n",
    "plt.legend((p3[0], p2[0], p1[0]), ('Temporal','Spatial sampling','Techinical'), \n",
    "           loc = \"center left\", bbox_to_anchor=(1, 0.5), fontsize = 13)\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### load covariance/correlation decompostion result"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>otu_1</th>\n",
       "      <th>otu_2</th>\n",
       "      <th>otu_3</th>\n",
       "      <th>otu_4</th>\n",
       "      <th>otu_5</th>\n",
       "      <th>otu_6</th>\n",
       "      <th>otu_7</th>\n",
       "      <th>otu_8</th>\n",
       "      <th>otu_9</th>\n",
       "      <th>otu_10</th>\n",
       "      <th>...</th>\n",
       "      <th>otu_284</th>\n",
       "      <th>otu_304</th>\n",
       "      <th>otu_315</th>\n",
       "      <th>otu_352</th>\n",
       "      <th>otu_370</th>\n",
       "      <th>otu_383</th>\n",
       "      <th>otu_388</th>\n",
       "      <th>otu_394</th>\n",
       "      <th>otu_416</th>\n",
       "      <th>otu_429</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>otu_1</th>\n",
       "      <td>1.010177</td>\n",
       "      <td>0.396150</td>\n",
       "      <td>0.655665</td>\n",
       "      <td>-0.255742</td>\n",
       "      <td>0.673559</td>\n",
       "      <td>0.472517</td>\n",
       "      <td>0.666926</td>\n",
       "      <td>0.583428</td>\n",
       "      <td>0.306945</td>\n",
       "      <td>0.686890</td>\n",
       "      <td>...</td>\n",
       "      <td>0.503124</td>\n",
       "      <td>0.761594</td>\n",
       "      <td>-0.112431</td>\n",
       "      <td>0.643203</td>\n",
       "      <td>0.256222</td>\n",
       "      <td>0.606065</td>\n",
       "      <td>-0.067260</td>\n",
       "      <td>0.372274</td>\n",
       "      <td>0.128460</td>\n",
       "      <td>-0.024871</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>otu_2</th>\n",
       "      <td>0.396150</td>\n",
       "      <td>0.998092</td>\n",
       "      <td>0.728573</td>\n",
       "      <td>0.377353</td>\n",
       "      <td>0.248537</td>\n",
       "      <td>0.844729</td>\n",
       "      <td>0.699585</td>\n",
       "      <td>0.102890</td>\n",
       "      <td>0.737745</td>\n",
       "      <td>-0.079987</td>\n",
       "      <td>...</td>\n",
       "      <td>0.431139</td>\n",
       "      <td>0.236563</td>\n",
       "      <td>0.478209</td>\n",
       "      <td>0.299753</td>\n",
       "      <td>-0.202513</td>\n",
       "      <td>0.259806</td>\n",
       "      <td>-0.245788</td>\n",
       "      <td>0.568209</td>\n",
       "      <td>0.503665</td>\n",
       "      <td>0.303136</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>otu_3</th>\n",
       "      <td>0.655665</td>\n",
       "      <td>0.728573</td>\n",
       "      <td>1.007463</td>\n",
       "      <td>0.145758</td>\n",
       "      <td>0.407829</td>\n",
       "      <td>0.810984</td>\n",
       "      <td>0.906859</td>\n",
       "      <td>0.397608</td>\n",
       "      <td>0.577209</td>\n",
       "      <td>0.391374</td>\n",
       "      <td>...</td>\n",
       "      <td>0.201446</td>\n",
       "      <td>0.551413</td>\n",
       "      <td>0.463480</td>\n",
       "      <td>0.405504</td>\n",
       "      <td>0.048430</td>\n",
       "      <td>0.328564</td>\n",
       "      <td>-0.206732</td>\n",
       "      <td>0.516860</td>\n",
       "      <td>0.245647</td>\n",
       "      <td>0.205115</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>otu_4</th>\n",
       "      <td>-0.255742</td>\n",
       "      <td>0.377353</td>\n",
       "      <td>0.145758</td>\n",
       "      <td>1.009353</td>\n",
       "      <td>-0.202443</td>\n",
       "      <td>0.329569</td>\n",
       "      <td>0.099897</td>\n",
       "      <td>-0.364880</td>\n",
       "      <td>0.388706</td>\n",
       "      <td>-0.443963</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.154311</td>\n",
       "      <td>-0.266471</td>\n",
       "      <td>0.487658</td>\n",
       "      <td>-0.194044</td>\n",
       "      <td>-0.287663</td>\n",
       "      <td>-0.069719</td>\n",
       "      <td>-0.015127</td>\n",
       "      <td>0.051577</td>\n",
       "      <td>0.402829</td>\n",
       "      <td>0.313857</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>otu_5</th>\n",
       "      <td>0.673559</td>\n",
       "      <td>0.248537</td>\n",
       "      <td>0.407829</td>\n",
       "      <td>-0.202443</td>\n",
       "      <td>1.007391</td>\n",
       "      <td>0.245542</td>\n",
       "      <td>0.447812</td>\n",
       "      <td>0.450035</td>\n",
       "      <td>0.386151</td>\n",
       "      <td>0.683891</td>\n",
       "      <td>...</td>\n",
       "      <td>0.497728</td>\n",
       "      <td>0.885943</td>\n",
       "      <td>-0.185406</td>\n",
       "      <td>0.966599</td>\n",
       "      <td>0.478691</td>\n",
       "      <td>0.721345</td>\n",
       "      <td>-0.291836</td>\n",
       "      <td>0.381167</td>\n",
       "      <td>0.220358</td>\n",
       "      <td>-0.216007</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 136 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          otu_1     otu_2     otu_3     otu_4     otu_5     otu_6     otu_7  \\\n",
       "otu_1  1.010177  0.396150  0.655665 -0.255742  0.673559  0.472517  0.666926   \n",
       "otu_2  0.396150  0.998092  0.728573  0.377353  0.248537  0.844729  0.699585   \n",
       "otu_3  0.655665  0.728573  1.007463  0.145758  0.407829  0.810984  0.906859   \n",
       "otu_4 -0.255742  0.377353  0.145758  1.009353 -0.202443  0.329569  0.099897   \n",
       "otu_5  0.673559  0.248537  0.407829 -0.202443  1.007391  0.245542  0.447812   \n",
       "\n",
       "          otu_8     otu_9    otu_10    ...      otu_284   otu_304   otu_315  \\\n",
       "otu_1  0.583428  0.306945  0.686890    ...     0.503124  0.761594 -0.112431   \n",
       "otu_2  0.102890  0.737745 -0.079987    ...     0.431139  0.236563  0.478209   \n",
       "otu_3  0.397608  0.577209  0.391374    ...     0.201446  0.551413  0.463480   \n",
       "otu_4 -0.364880  0.388706 -0.443963    ...    -0.154311 -0.266471  0.487658   \n",
       "otu_5  0.450035  0.386151  0.683891    ...     0.497728  0.885943 -0.185406   \n",
       "\n",
       "        otu_352   otu_370   otu_383   otu_388   otu_394   otu_416   otu_429  \n",
       "otu_1  0.643203  0.256222  0.606065 -0.067260  0.372274  0.128460 -0.024871  \n",
       "otu_2  0.299753 -0.202513  0.259806 -0.245788  0.568209  0.503665  0.303136  \n",
       "otu_3  0.405504  0.048430  0.328564 -0.206732  0.516860  0.245647  0.205115  \n",
       "otu_4 -0.194044 -0.287663 -0.069719 -0.015127  0.051577  0.402829  0.313857  \n",
       "otu_5  0.966599  0.478691  0.721345 -0.291836  0.381167  0.220358 -0.216007  \n",
       "\n",
       "[5 rows x 136 columns]"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "with open(DIVERS_output_prefix + \".correlation_total.csv\") as f:\n",
    "    data_correlation_total = pd.read_csv(f, index_col=0, lineterminator = '\\n')\n",
    "with open(DIVERS_output_prefix + \".correlation_decomposition_temporal.csv\") as f:\n",
    "    data_correlation_decomposition_T = pd.read_csv(f, index_col=0, lineterminator = '\\n')\n",
    "with open(DIVERS_output_prefix + \".correlation_decomposition_spatial.csv\") as f:\n",
    "    data_correlation_decomposition_S = pd.read_csv(f, index_col=0, lineterminator = '\\n')\n",
    "with open(DIVERS_output_prefix + \".correlation_decomposition_technical.csv\") as f:\n",
    "    data_correlation_decomposition_N = pd.read_csv(f, index_col=0, lineterminator = '\\n')\n",
    "\n",
    "data_correlation_total.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### find OTUs with high abundance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "if_abundance_10e4 = data_variance_decomposition[\"Average_abundances\"] > 10 ** -4\n",
    "index_abundance_10e4 = [i for i in range(len(if_abundance_10e4)) if if_abundance_10e4[i] == True]\n",
    "otu_abundance_10e4 = data_variance_decomposition[\"OTU_ID\"][index_abundance_10e4].values\n",
    "\n",
    "data_variance_decomposition_highAbundance = data_variance_decomposition.loc[index_abundance_10e4,:]\n",
    "data_correlation_total_highAbundance = data_correlation_total.loc[otu_abundance_10e4,otu_abundance_10e4]\n",
    "data_correlation_decomposition_T_highAbundance = data_correlation_decomposition_T.loc[otu_abundance_10e4,otu_abundance_10e4]\n",
    "data_correlation_decomposition_S_highAbundance = data_correlation_decomposition_S.loc[otu_abundance_10e4,otu_abundance_10e4]\n",
    "data_correlation_decomposition_N_highAbundance = data_correlation_decomposition_N.loc[otu_abundance_10e4,otu_abundance_10e4]\n",
    "df_taxonomy_highAbundance = df_taxonomy.loc[index_abundance_10e4,:]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### get pairwise correlations (total, temporal, spatial and technical)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "triu_index_pair = []\n",
    "triu_otu_pair = []\n",
    "for i in range(len(otu_abundance_10e4)):\n",
    "    for j in range(i + 1,len(otu_abundance_10e4)):\n",
    "        triu_index_pair.append([i,j])\n",
    "        triu_otu_pair.append([otu_abundance_10e4[i],otu_abundance_10e4[j]])\n",
    "\n",
    "cors_total = [data_correlation_total_highAbundance.loc[e[0],e[1]] for e in triu_otu_pair]\n",
    "cors_T = [data_correlation_decomposition_T_highAbundance.loc[e[0],e[1]] for e in triu_otu_pair]\n",
    "cors_S = [data_correlation_decomposition_S_highAbundance.loc[e[0],e[1]] for e in triu_otu_pair]\n",
    "cors_N = [data_correlation_decomposition_N_highAbundance.loc[e[0],e[1]] for e in triu_otu_pair]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### boxplot of pairwise correlations (total, temporal, spatial and technical)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAUoAAAGoCAYAAAAgpayMAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4wLCBo\ndHRwOi8vbWF0cGxvdGxpYi5vcmcvpW3flQAAIABJREFUeJztnXmYXFW1t99FNyQhCcgQiIJBpiBj\nQtI4QRBBUfRDUJxB0KvpNonCRURBuZfJEcXxMiRRBBHUC6KoCCIKVwYHkkiAgIR5NBKmmBCmdNb3\nx9pFTorqruqkqs6pqt/7POfpqr1PnVr7VPWv1t577bXN3RFCCDEw6+RtgBBCFB0JpRBCVEFCKYQQ\nVZBQCiFEFSSUQghRBQmlEEJUQUIphBBVKKRQmtknzWyOmT1nZudWOfdoM1tkZkvM7BwzG5ape5WZ\nXW1my83sH2b25oYbL4RoOwoplMAjwBeBcwY7yczeChwH7Ae8CtgGODlzyk+AvwObAF8ALjazMQ2w\nVwjRxliRV+aY2ReBLd39IwPUXwjc5+6fT8/3Ay5w97FmNh64BdjU3Zem+mtT/dlNaYAQoi0oqkdZ\nKzsD8zPP5wObm9kmqe6ekkhm6nduon1CiDagO28D1pJRwJLM89Lj0RXqSvVbVLqQmfUCvQA77bTT\n5AULFtTX0g7HzOpynSL3gERLUtMXs9U9ymXABpnnpcdLK9SV6pdSAXef5e497t4zYsSIuhsqhGhd\nWl0oFwATMs8nAP9y98dT3TZmNrqsPjdX0czqcrQi7j7oMdTz2okZM2bQ3d3NjBkz8jZFDEAhhdLM\nus1sONAFdJnZcDOrNEzwI+BjZraTmW0EnACcC+DuC4GbgBPT698F7Ab8vCmNEKJGZs6cSX9/PzNn\nzszbFDEAhRRKQvCeIUJ/DkuPTzCzcWa2zMzGAbj7FcBpwNXA/ek4MXOdDwA9wJPAV4H3uPviprWi\njGreUq2HaC/6+vro6uqir68vb1PEABRyMsfdTwJOGqB6VNm53wS+OcB17gP2qZ9lolFMnTo1bxOE\nGJBCx1HmRU9Pj8+ZM6fp79vb2wvArFmzmv7eIj+6u7vp7++nq6uLFStW5G1Op9ERs95txezZs5k9\ne3beZogmo6538ZFHWYG8PMrSjHYnfiZz584FYPLkyTlbIjqMmjzKQo5Ris6jp6cH6MwfCVF81PUW\nQogqSCiFEKIKEkohhKiCxigLxKRJk/I2QQhRAQllgSjN/AohioW63kIIUQV5lKIQ5BG3KkStSCgL\nRCcHnCvQXBQZdb2FEKIKEkpRCHp7e19MCiJE0dBa7wporXfz6eS2i1xR9iAhhKgHEkohckZ75hQf\ndb0roK538+nktitxb64ozVqroc2lOpO+vj5mzpypxL0FRh5lBfLyKDuZTvYoRa7IoxStgxKCiCKj\nyZwCMWvWrI7dWGzu3LkdmxREkznFR13vCmgyRzQTTebkiuIohWgFtAtj8ZFHWQF5lM2nk9suckUe\npRCtgMYoi488ygrIo2w+anvQie3PGXmUQrQCJaHMCqYoFhJKIXJm5513Xu2vKB4KOC8Q6nZ1Jrff\nfvtqf0XxkEcpRM4oPKj4FHIyx8w2Bn4A7A88Bhzv7hdWOO9yYEqmaD3gDnffNdXfB2wO9Kf6G9x9\n/2rvr7Xezae0IklZzkWTaem13mcAzxMiNxG4zMzmu/uC7EnufkD2uZldA/yx7FoHuvtVDbS1bpQ2\n2OrEpXwSSFFkCieUZjYSOATYxd2XAdeZ2a+ADwPHDfK6VxHe5UebYGZDmDdvXt4mCCEqUMQxyvFA\nv7svzJTNB6pNCR4OXOvu95aVX2Bmi83sSjObUE9DRf3o5IQgovgUUShHAUvKypYAo6u87nDg3LKy\nQ4FXAVsBVwO/M7OXVXqxmfWa2Rwzm7N48eKh2izWkr6+vo6dzNDKnOJTuMkcM9sduN7d18+UHQPs\n4+4HDvCavYArgLGpuz7Qtf8BHOvuvx7MBq3MaT6d3HZlD8qVll2ZsxDoNrPtM2UTgAUDnA9wBHDJ\nYCKZcGq8MUI0C4UHFZ/CeZQAZvZTQtQ+Tsx6/xZ4Q/msdzp3BPBP4N3u/sdM+TjglcCNxA/Cp4DP\nAq9298cHe395lM2nk9sucqWlw4OmA+cAjwKPA9PcfYGZTQEud/dRmXMPJsYwry67xmjgLGBb4Fng\nJuCAaiKZJ1OnTs3bBCFEBQrpUeaNAs6bjzxKkRMtO0YphBCFQkJZIDp5gy1371hvUuFBxUdd7wpo\nMkc0EyXuzRV1vYVoBZS4t/hIKEUhmDx58otJQTqNadOm0dXVxbRp0/I2RQxAUcODRIehhCCiyGiM\nsgIao2w+ndx2LWHMFY1RCtEKaAlj8ZFHWQF5lM2nk9sucqWllzB2JFoNJEQxkVAWiE6d9W136hH2\nI087XySUohAoIYgoMhLKAlHaYKsTt0Ro5zbLG2x9NJlTAU3mCNExKDxItA6dnBCkt7dX2/UWHHmU\nFZBH2XzU9s5sewGQRymEEPVAQimEEFWQUAohRBUUHlQgJk2alLcJQogKSCgLRKfO+gpRdCSUQuSM\nehLFR0IpCkEnJwRRT6L4SCgLRCfH0ykhiCgymvUWQogqSChFIejkZXxmph0YC46WMFZASxibj9re\nmW0vAFrCKIQQ9UBCKYQQVZBQCiFEFRQeVCBmzpyZtwlCiAoU0qM0s43N7Bdm9rSZ3W9mHxrgvJPM\n7AUzW5Y5tsnUTzSzuWa2PP2d2LxWDJ1OnvkVosgU1aM8A3ge2ByYCFxmZvPdfUGFc3/m7oeVF5rZ\nesClwLeBM4E+4FIz297dn2+c6WJN6ORlfOpJFJ/ChQeZ2UjgSWAXd1+Yys4HHnb348rOPQnYbgCh\n3B/4IbClp0aa2QNAr7tfMZgNeYUHlTbYklcpRNNo2fCg8UB/SSQT84GdBzj/QDN7wswWmNm0TPnO\nwM2++i/BzQNdx8x6zWyOmc1ZvHjx2ti/xvT19dHX15fLewshBqaIQjkKWFJWtgQYXeHc/wV2BMYA\nU4H/NrMPrsF1cPdZ7t7j7j1jxoxZU9uFGDKzZs1q6+1624EiCuUyYIOysg2ApeUnuvtt7v6Iu/e7\n+w3Ad4D3DPU6In86eRmfehLFp4hCuRDoNrPtM2UTgEoTOeU4q8YcFgC72er/fbvVeB0hhHiRwgml\nuz8NXAKcYmYjzWxP4CDg/PJzzewgM9vIgtcARxIz3QDXAP3AkWY2zMw+mcr/2PBGCCHaisIJZWI6\nMAJ4FPgJMM3dF5jZFDNbljnvA8BdRHf6R8DX3P08gBQCdDBwOPAU8B/AwQoNEkIMlcKFBxWBNQ0P\nKsoYWyt+pp2cQaeT214AWjY8SAghCkVRV+a0NN874+xc3vdTMz6Ry/sK0e5IKEUh6ORlfOpyFx8J\npSgEWrYpiozGKIUQogoSSlEIOnkZ3+TJk7Vdb8FR11sUgtISvk7sgs+bNy9vE0QV5FEKIUQVJJRC\nCFEFCaUQQlRBQimEEFWQUAohRBVqmvVO+9gcBbweWI+yheTuvn/9TROtRL0SgqzNdVp1hcvUqVPz\nNkFUodbwoLOJzOFXAI81zhwhOo9OjR9tJWoVyv2BD7v7xY00RrQ+i3bdo+nvOfaWG5v+nqKzqHWM\nsovYCVEIUWfmzp3L3Llz8zZDDEKtHuUFxJYKR3qrDgQJUVB6enqA1h1j7QRqFcqRwGHAu8zsbuC5\nbKUmc4QQ7UytQtlF7F0jhBAdR01C6e4fbbQhQojOYsaMGcycOZO+vj7OOOOMvM0ZlJoDzs1sDzP7\nmZktMLObzOyCtEWsEEIMmZkzZ9Lf398S2e1rEkoz2xe4HhgHXAb8AdgWuM7M3tg484QQ7UpfXx9d\nXV0vptgrMrWOUX4ZONPd/zNbaGbfBL4ITKm3YUKI9uaMM84ofJe7RK1COQE4vEL5TKDzMq0KUUfW\nZA950VxqFcpFRLd7YVn5OGBZXS0SosPQNhDFp9bJnJ8BZ5vZfmY2wszWN7O3AGcBWtYohBgyM2bM\noLu7mxkzZuRtSlVqFcqTgduA3xMe5FIiQcbfgM82xjQhOoPe3t6O3Cuo7Wa93f0Zd38nsAvwAeBg\nYHt3/5C7L2+kgUK0O7Nnz2b27Nl5m9F02mLW28xe4e6PlB6n4qeIMCGy5aXzhBCiHRnMo3zQzDZL\njx8CHqxwlMqFEGJInHnmmfT393PmmWfmbUpVBpv13hd4Ij1+UxNsEUKIQjKgR+nu/+fuK9LTNwI3\nprIXD+DvwEH1NsrMNjazX5jZ02Z2v5l9aIDzjjWzW81sqZnda2bHltXfZ2bPmNmydFxZb1uFEO3P\nYGOUmwLrp6cnAr8xs/JtICYB04BP19muM4Dngc2BicBlZjbf3ReUm0kEwt9MLKm80swedPefZs45\n0N2vqrN9QogOYrCu9wHAeUApm+hA+fYvqadBaSOzQ4Bd3H0ZsZ78V8CHgeOy57r7aZmnd5jZpcCe\nQFYohSg0kyZNytsEUYUBhdLdz09JetcB/kR0sZ/InkLEU95WZ5vGA/3unl0FNJ/o/g+IxfZ9U4hl\nlVkuMLN1iGGCY9294pYWZtZLWo45bty4NTRdiKGjbSCKz6BLGN39BgAz2xp4oEnbQIwClpSVLQFG\nV3ndSYSo/zBTdigwj+iiHwX8zsxe7e5Plb/Y3WcBswB6enrWqp2fmvGJtXm5EKJg1LrW+1/Ap8xs\nVyLbOYT4DAN63H18HW1aBmxQVrYB4b1WxMw+SYxVTnH3F7epcPfrM6d9xcyOILzOX9fPXCFEu1Or\nUJ4BfJBYsrgX0RXfFtgSOL3ONi0Eus1se3e/M5VNAMoncgAws/8gxi73dveHqlzbCYFvKN874+xG\nv0VF5Mm2JjFq1Hmbi02fPv3FDOdFp9a13gcCR7j7PsA9wAxgG+DnRFe5brj708QE0SlmNtLM9iTG\nR88vP9fMDiVyZb7F3e8pqxtnZnua2XpmNjyFDm1KZmWREELUQq1CuSHw1/R4ATDZ3fuBrwBvb4Bd\n04ERwKPEpmbT3H2BmU0xs2xaty8CmwA3ZmIlS+7caCK70ZPAw8DbgAPc/fEG2CuEGCLtsjInyz+B\nLYAHiK7xbql8CTCm3ka5+xNE4o3y8mvJeLDuvvUg11jAKjuFEGKNqVUoLwHONbOPAFcBPzaz6wkx\nu7tBtgkhRCGoVSiPB9YFtnb3C1MA+M+JGer3Nso4IYQoArXu6/0c8KnM86lmdjTwbGY9uBBCtCWD\nrfWumIiiwnnu7j+pn0lCdBatkOG70xnMo/xxjddwYmZaCMbeMlBKADEQnbgNRKsx2FrvWkOHhBCi\nral1MgcAM9sc2JGIqRzt7o82xCrRsizadY+mv2ere7GzZs0C5FkWmZq8RjMbZmbnEPGUVwEvB2aa\n2R/MbMNGGihEu9PX19cSy/g6mVq71ycCPcQ672dT2WnAK9NfIYRoW2oVyvcBR6a0aw7g7n8GpgLv\nbJBtQghRCGoVylcA91coX0SsAxdCiLalVqH8O6uvvS7lg+olso8LIUTbUuus93HAFWb2OmIp43Fm\ntiPwWuAdjTJOCCGKQE0eZcra8wZiZ8S7gD2Irvhk7XAoOh0zW6ujHtcpIkOxuehtq8mjNLNzgS+7\n+4cba44QQhSPWrveBxObdwkhBmDcr8u3vW88Dxy4adPfs1aqbW2R9RaLvg1GrZM5FxBbM2xnZkNa\nzSOEEINRdJGE2j3K/Yj9tg8F3MxWZivdfb16GyaEEEWhVqH8SkOtEEJ0HK2UXq5WoXwTMZmzsJHG\nCCE6h1ZKAlLrGOXBRGiQEEJ0HJrMEULkwqxZs15MMVd0NJkjhMiFUmq5VuiCazJHCCGqUOsujOeV\nHpvZGOAFd3+qYVYJIUSBqHlfHDM72swWEanVHjezh83s040zTQghikGta71nAF8Cvgv8CegCpgBf\nNLNl7t4aI7JCCLEG1DpG+Z/AUe4+O1P2azO7M9VJKIUQbUutQrkF8McK5X8kvEwhOp4iJ6hoFPVI\ng7a212jGWvFaxyjvBvauUL4P8FDdrBFCiAJSq0f5LeB/zGwb4IZUthdwFPD5ehtlZhsDPwD2Bx4D\njnf3CyucZ8BXgY+noh8An/P0E2NmE1PZjsDtwMfc/aZ62ysEdHaata+/f3b1k+rMsT+b2rT3qjU8\n6BwzexlwLPCFVPwQ8OkGTeScQSyZ3ByYCFxmZvPdfUHZeb3E8soJxD4+vwfuAc42s/WAS4FvA2cC\nfcClZra9uzd0OeanZnyikZcXonA0U7TyoObwIHf/JrA1IV4bAns1QiTNbCRwCPBf7r7M3a8DfgVU\nyq5+BHC6uz/k7g8DpwMfSXX7ED8E33b359z9u4AB+9bbZiFEe1NreNBY4CLgWnf/fCq728wWAO91\n93r2OcYD/WWZiuYDb6xw7s6svgvk/FRWqrvZVx/pvTmVXzGYAcuXL+emm1bvoY8ZM4YtttiC/v5+\nbrnllpe8ZuzYsbg7L7zwAgsWlDu+8IpXvILNNtuMZ599ln/84x8vqd9yyy0ZM2YM66+/Ptdff/1L\n6rfaais22mgjli1bxl133fWS+q233poNN9yQJUuWvMR2gO22245Ro0bx5JNPcv/9L915ePz48ay/\n/vo89thjPPTQS4edX/3qVzN8+HAeffRRHnnkkZfUF4V77rmHf//736uVDRs2jB133BGAu+66i2XL\nlq1WP2LECHbYYQcA7rjjDp555pnV6keNGsV2220HwO23385zzz3XKPPXirX57m266aYsX76chQtf\nmiCs2ncvb1544QXWXXddFi1axKJFi15Sv+uuu9LV1cXDDz/M4sWLV6ubOHFiTe9Rq0f5XaJr+8NM\n2d5EPOU3a7xGrYwClpSVLQFG13DuEmBUGrscynUws14zm2Nmc5YuXbpGhgsh2hOrZWrdzJ4AppSP\nEZrZBOAqdx9TN4PMdgeud/f1M2XHAPu4+4Fl5y4B3uLuf0vPJwPXuPtoMzs61b09c/6vU/3pg9nQ\n09Pjc+bMqVeTaqYUJtEKqfHLKdm+aNc9mv7eY2+5EcjvvpXanudkTit/Z3K2vabYpFo9SgfWr1De\nBdQ7c9BCoNvMts+UTQBe2qeIsgkDnLcA2M1WD9LabYDrCCHEgNQqlFcA3zazcaUCM3slMXlyZT0N\ncvengUuI/JcjzWxP4CDg/Aqn/wj4tJltYWavAI4Bzk111wD9wJFmNszMPpnKKwXOCyHEgNQaR3k0\nEXpzr5mVRkPHAPOIHJX1ZjpwDvAo8Dgwzd0XmNkU4HJ3H5XOmwlsA5RmV76fynD3583s4FT2VSKO\n8uBGhwZ1OqVusBDtRK1xlI+a2STgzcAuwAuE8FzlDRhgcPcniPjI8vJriUma0nMHPpuOStf5OzC5\n3vY1ilbabEmITqLmbR3cvR/4XTpEA2iFTM8Dsba/lwUZ2BeiItr/RgiRC63Ug6opPKjTyCs8qLTR\nUit7lmtKK3uU9cigs7a04n0rCDV9eBLKCiiOsvm0ctsllC1NTR/ekLveZtbt7iuGbo8Q7YnGZ9eM\nVupB1exRmtmHicxBWxNpy44FHnH3UxtnXj7Io2w+anvntb0g7a6fR2lmhxPpyr4BnJCK5wOnm9lz\n7n7aGpkoRKKVBvZF51HrWu/5wNfc/UIzWwpMcPd7zOxDwCnuvl2jDW0m8ihFM+nUz70g7a7rGOX2\nwJ8rlP+F2E9HCCFWo9ZJrmrnFeEHpNa13g8QmcbL2S/VCbFWzJo168XBfSGKRq0e5enAWSmB7zrA\nG83sI8CngeMbZFvHUYRfzrzo6+sDWmMGtN606/hste9z1pMs+nd/KLPe04iNxEpd7YeBL7p7233K\neY1RdjIFGa8STaQgQlnfOEp3P4vwKjcFnnf3f1d7jRBCtAM1by5mZnuZ2aZpf5y3m9lvzOwEM6v5\nGmJwJk+ezOTJLZPsSNQJjc8Wn1rDg6YT++a8GXgSmEMk7J0E/MDdTxjk5S2HwoOaj9reeW1vpa53\nrd7gUUCvu19DJOq92d3fARxG5W1khRCibahVKLciMpwDvBX4bXp8J7BZvY0SQogiUatQPgxsa2bb\nAruyKnnvXsCDjTBMdBbu3nFdT9E61DrrPQu4GHgOuNXdr0vjltm130II0ZbUumfO18zsNmBb4IJU\n/BjQ5+6VdkcUQoi2YShxlL8ue/6/9Tens5k6dWreJuRGKSxq7ty5OVsixEsZMDzIzBYCr3P3J8zs\nTmDAASR3H98g+3JBK3OaT6eGyHQyrRQeNJhHeQHwbHr847U2RwghWpQBhdLdT848/YW739wEezqa\nUrdTq3OEKBa1hgf93czuNbPvmtl+ZqZtbhtAT08PPT09eZshmkynLl2dPn06XV1dTJ8+PW9TqlKr\n4I0FDgDeBvwM6DazK4BfAb9196caZJ8Qbc+8efPyNkFUYcjb1aYkGK8F+ojljCvdfVgDbMsNrfVu\nPmp757W9u7ub/v5+urq6WLEit41d65tmzcxeDrwR2Ccd44E7gD8M3TYhVqeTQ6M6lb6+PmbOnPli\n0uYiU2v2oIVEsPndwHXA1cAf3P2RxpqXD/IoRTPR554rdc0etAh4gVjCuBT4N/D0mtklhBCtRU1C\n6e57AxsDxwLPAycCj5rZHDP7WgPtEx3C3LlztSpHFJahLGFcDlxhZlcCrwfeD0wFdgc+Vw9jzGxj\n4AfA/sRa8uPd/cIBzj0WOIJIAfcYcKa7fz1Tfx+wOdCfim5w9/3rYWej6OTVQKWwqE7sfmp8tvjU\nJJRmtgPwlnS8EegC/kgk9P1NHe05g/BYNye2x73MzOa7+4JKZgGHAzcT46dXmtmD7v7TzDkHuvtV\ndbSvoXRiLJ1A20C0ALV6lLcT+3f/FvgQMZHzXD0NMbORwCHALu6+DLjOzH5FZFA/rvx8dz8t8/QO\nM7sU2BP4afm5QgixNtQ6mfNL4K3uPt3df1tvkUyMB/rdfWGmbD6wc7UXWkwbTgHKPc8LzGyxmV1p\nZhPqZ2pj6O3t7ch9rTsdjc8Wn1rDg54CJrj7/Q0zxGwKcJG7j82UTQUOdfd9qrz2ZOBg4DUlETez\nPYF5RBf9qHS8eqBVRGbWC/QCjBs3bvL99zesqQPSyWEiantntr0A1DU86ALgVDPbbk3XeZvZNWbm\nAxzXAcuADcpetgERjjTYdT9JjFW+I+vpuvv17v6Muy93968ATxFeZ0XcfZa797h7z5gxY9akiUKI\nNqVW0duP6BofCriZrcxWuvt61S5Qg1c4klhDvr2735mKJ/DS7nT2Nf9BjF/u7e4PVTOBGn89hBAi\nS61C+ZWGWgG4+9Nmdglwipl9nJj1Pgh4Q6XzzexQ4MvAm9z9nrK6ccArgRsJr/lTwKbA9Y1rgVgb\nOjk0ShSfWvfMOa/RhiSmA+cAjwKPA9NKoUFpDPNydx+Vzv0isAlwYyZT8o/d/RPAaOAsImzoWeAm\n4AB3f7xJ7RBDRKFRosgMKJRmNgv4tLsvS48Hwt29Lqva3f0JYlKmUt21wKjM860Huc4CYLd62CSE\nEIN5lNtn6rdvgi0dz6RJk/I2ITdKYVEKvhZFZMj5KDsBbS7WfDo5REZbgORK3fNRjiFmvrsybzAM\n2MPdvzRk84QQgASyFah1rfeHgVmEMJbCbEo//XcBEkohRNtSa8D5F4AfEbPITwGTgbcT678bHjrU\nKZjZansdi85AS1eLT61LGJ8DdnX3hWb2B+Cb7n6Zmb0TONndd2+0oc1EGc6bj9remW0vAHVdwrgc\nKK3GuRPYJT2+CdhuaHYJIURrUatQ3gAca2bDgb8DB6by1xJrtIVYKyZNmtTR4VGi2NQ66/154HfE\nxM3ZwOfN7DEiacW3G2Sb6CCUZkwUmVqXMM43s22Ake6+1MxeC7wbWOzuFzXUQiGEyJlBhTJl9NmX\n2H3xBndfDODui4AzG2+eEELkz2BrvXcjutubp6JHzOxd7n5jUyzrQGbOnJm3CbnRyTO/GpstPgOG\nB5nZb4kxyM8QOxl+FdjA3fdonnn5oCWMzaeThVLkylovYXw9kevxJnhxW4aFZjbS3Z+ug4FCCNES\nDBYeNBpYVHqSkuOuIHJAigYwa9YsZc8RooAM1vVeCYx190czZUuJTcbuqfiiNkErc5qP2t6ZbS8A\ndV2ZI4QQHUu1OMqjzCw7HtkNTDezJ7InufuX626ZEEIUhMGE8gHgQ2Vli4BDysqc2ORLiDWmk0Oj\nRPFRhvMKaIxSNBN97rmiMUohhKgHEkpRCBQaJYpMzXvmiMbTyV2vvr7Y8bgTM31rfLb4SCiFyJlO\n/HFoNdT1FkKIKkgoC8TkyZO1dWkHovHZ4qPwoAooPKj5qO2d2fYCoPAgIYSoBxJKIYSogma9RSFQ\nt1MUGXmUQghRhcIJpZltbGa/MLOnzex+MytPzJE99yQze8HMlmWObTL1E81srpktT38nNqcVQoh2\noohd7zOA54lNzSYCl5nZfHdfMMD5P3P3w8oLzWw94FJi3/EzgT7gUjPb3t2fb4zpa8fUqVPzNiE3\nSmFR2t9bFJFChQel7XGfBHZx94Wp7HzgYXc/rsL5JwHbDSCU+wM/BLb01EgzewDodfcrBrNDm4s1\nH4XIiJxoyfCg8UB/SSQT84GdB3nNgWb2hJktMLNpmfKdgZt99f+8mwe6lpn1mtkcM5uzePHiNbVf\nCNGGFE0oRwFLysqWEBudVeJ/gR2BMcBU4L/N7INrci13n+XuPe7eM2bMmDWxfa2ZO3euup5CFJCm\nCqWZXWNmPsBxHbCM2Es8ywbA0krXc/fb3P0Rd+939xuA7wDvSdVDulYR6OnpoaenJ28zRJPR0tXi\n09TJHHffZ7D6NEbZnSZc7kzFE4CBJnJe8hasGnNYABxjZpbpfu9GTBYJURjmzZuXtwmiCoXqerv7\n08AlwClmNtLM9gQOAs6vdL6ZHWRmG1nwGuBIYqYb4BqgHzjSzIaZ2SdT+R8b2ohBMLNBj6GeJ4Ro\nDkUMD5oOnAM8CjwOTCuFBpnZFOBydx+Vzv1AOncY8BDwNXc/D8Ddnzezg4HvA18FbgcOLmpoUKfT\nyaFRovgUKjyoKCg8SDQThUblSkuGBwkhROGQUIpCoNAoUWSKOEYpOpBSWFQndj81Plt8JJRCNJha\nIxVmz549YF0n/oAUCXW9hRC5/l4lAAAgAElEQVSiCvIohWgw8gZbH3mUQghRBQmlEEJUQUIphBBV\n0BilKARaCSWKjDxKUQg6OdXYjBkz6O7uZsaMGXmbIgZAa70roLXeopl0d3fT399PV1cXK1asyNuc\nTkNrvUXr0NvbS29vb95m5EJfXx9dXV309fXlbYoYAAmlKASzZ88edGWKEHmirncF1PWuP/VKONyO\n39fsvWnH9hUcdb2FEKIeSChFU3D3uhztyC677LLaX1E8FEcpRM7ccssteZsgqiCPUgghqiChFEKI\nKkgohRCiChJKIYSogoRSCCGqIKEUQogqSCiFEKIKEkohhKiChFIIIaogoRRCiCpIKIUQogoSSiGE\nqIKEUoic0Z45xadQQmlmG5vZL8zsaTO738w+NMi5l5vZsszxvJndkqm/z8yeydRf2ZxWCDE0Zs6c\nSX9/PzNnzszbFDEAhRJK4AzgeWBz4FDgLDPbudKJ7n6Au48qHcANwEVlpx2YOWf/hlouxBqiPXOK\nT2G2gjCzkcCTwC7uvjCVnQ887O7HVXntq4C7ge3c/d5Udh/wcXe/aqi2aCsIITqGltsKYjzQXxLJ\nxHygokdZxuHAtSWRzHCBmS02syvNbMJgFzCzXjObY2ZzFi9ePDTLhRBtTZGEchSwpKxsCTC6htce\nDpxbVnYo8CpgK+Bq4Hdm9rKBLuDus9y9x917xowZU6vNQogOoGlCaWbXmJkPcFwHLAM2KHvZBsDS\nKtfdCxgLXJwtd/fr3f0Zd1/u7l8BngKm1K9FQohOoWl75rj7PoPVpzHKbjPb3t3vTMUTgAVVLn0E\ncIm7L6tmAjWORwghRJbCdL3d/WngEuAUMxtpZnsCBwHnD/QaMxsBvJeybreZjTOzPc1sPTMbbmbH\nApsC1zesAUKItqUwQpmYDowAHgV+Akxz9wUAZjbFzMq9xoOJccyry8pHA2cRs+gPA28DDnD3xxto\nuxCiTSlMeFCRUHiQEB1Dy4UHCSFEIZFQCiFEFSSUQghRBQmlEEJUQUIphBBVkFAKIUQVJJRCCFEF\nCaUQQlRBQimEEFWQUAohRBUklEIIUQUJpRBCVEFCKYQQVZBQCiFEFSSUQghRBQmlEEJUQUIphBBV\nkFAKIUQVJJRCCFEFCaUQQlRBQimEEFWQUAohRBUklEIIUQUJpRBCVEFCKYQQVZBQCiFEFSSUQghR\nBQmlEEJUQUIphBBVkFAKIUQVCiWUZvZJM5tjZs+Z2bk1nH+0mS0ysyVmdo6ZDcvUvcrMrjaz5Wb2\nDzN7c0ONF0K0LYUSSuAR4IvAOdVONLO3AscB+wGvArYBTs6c8hPg78AmwBeAi81sTJ3tFUJ0AIUS\nSne/xN1/CTxew+lHAD9w9wXu/iRwKvARADMbD0wCTnT3Z9z958AtwCGNsVwI0c4USiiHyM7A/Mzz\n+cDmZrZJqrvH3ZeW1e/cRPuEEG1Cd94GrAWjgCWZ56XHoyvUleq3GOhiZtYL9Kany8zsjjrZOVQ2\nBR7L6b3zRm3vPPJu9xXu/rZqJzVNKM3sGuCNA1Rf7+57DfGSy4ANMs9Lj5dWqCvVL2UA3H0WMGuI\nNtQdM5vj7j1525EHanvntb1V2t20rre77+PuNsAxVJEEWABMyDyfAPzL3R9PdduY2eiy+gVr3gIh\nRKdSqDFKM+s2s+FAF9BlZsPNbCCv90fAx8xsJzPbCDgBOBfA3RcCNwEnpmu8C9gN+HnDGyGEaDsK\nJZSE2D1DhP0clh6fAGBm48xsmZmNA3D3K4DTgKuB+9NxYuZaHwB6gCeBrwLvcffFTWrH2pB79z9H\n1PbOoyXabe6etw1CCFFoiuZRCiFE4ZBQCiFEFSSUDcTMLG8b8sLM1svbBtFcBpl4bXkklA3CzNYH\nzjGzNQl9amnMbCLwczPbLm9b1hYz29LMXpG3HUUnJaS5yMyOzduWRiChbBybE+vRv2Rmr8nbmCYz\nHDgA+I6ZbZO3MWuKmW0JXAd8wcwGXNUlABgHbAgcYWbT8jam3kgoG4CZrePu9wI7peObnSSW7v4X\nYArwGuCsVhVLd38IuAx4J3CUxHJg3P1O4CjgH8S9mp6zSXVFQtkA3H1lEst/AHsDr6bzxPLPwEFE\nFqeWE0sz6wJw9xnAT4HDkVgOirvfApxCrIA7sp3EUkLZIJJYdrv77YR31RFiWZrAMjNz9xsIsdyd\nFhNLd+83s3XS42OB85FYVsXdbybywraVWEoo60jJCynh7ivS37YXy4wH5mV/bwAOpoU8y5LYpx+7\nddPjY4EfI7F8kfLve4kkll+kjcRSK3PqhJl1uXt/evw+YDNgJPAdd382le8E/IkYx/m0u/8tL3vr\nSfKcV6SZ/o8CY4HngK+5+wvpnD2BXwLzgGnufk9uBg9C6XNMQycrzWx0Nq+pmZ0OHErkGviOuz+c\nm7E5UvZ9n0ZM5owg8in81d2fN7PdiSXIOwPfdfczczN4LZFQ1oHSP1V6fCEwmVinPgJYDkwH5qYv\nT0ksbwW+4O7X52R2XcgKCjFD/BzR9o2BF4D3uftd6dw9gUsIsTwqJS8pDBmRHAWcQWwvsjXwDeDS\nNEHX8WJZ9n3/CbAnse3KVkTqxguAb7n7sxmxHA+c6+6n52T22uHuOup0AGcC9wJ7pudfAlYSHuQU\nYL1UvmMqvxwYnrfddWj3COAa4I/A2FR2ZWrjQuDVmXPfkMq/lbfdZW1YJ/0dRfyIXQ0cCRxDCP53\ngZ0y538DeBiYCYzJ2/6c7tm3gbuA16Xnn8t85v8NDEvlE4HfA38DNsrb7jVqa94GtPJB8sjT473T\nF+Ft6fmxwPPAtFS+MJ1T+vLsAIzPuw11avshwLXADun5RcCDRAanW4ixqh0y5+8KdOXdhgptWhf4\nGfA7YNNU9iMiEfRK4IdlYvl94NfZe9EpB9Gd/iXw7vT88+kH5aPAb4isXV8oOQJEmsMt87Z7jdub\ntwGtdgCWjq6y8g2SOG4IfBD4N3BYqntrEs0biF0j1827HWt5D7rS39HEWOx2mbZ+LYnkZCKv6Iwk\nMjcCu1S6Tk5tmAL8V1nZ1kkkSz92PwUeAl4OTE3tOAvYNft9yP5t14PkcWeejya87ZcBBwKLgf9I\ndZsQ3vZthOfd+r2mvA1opSMJ5KGlL0QquzLzBSl5iz8FzgNGpecbAbenf7RbgRF5t2Vt7kH6OxJ4\nlOhibZhEcRThPX8+c952wB3A08AledufbOpOdp9Roe41xMqij6Z/9n2I6JARwF+IMeeLgK0yr1mn\nGXbneL+6Mo/fCmxbuo/p77eJYZfNSucTibMfTM7Bpnm3YW2Ptl3E3kDWB2aa2cuIPYB2I7oYuPtz\nKTHANsCj7r4svWYrouv9bmCpuz/TfLPXnuyMMLA/sbPlz9x9SaofQXiYG3v6jyG62QuJeMpCTN54\nzNCf7u5Pm9lI4FCPPZPwFIlgZpOAu4H/c3c3s37Ca7qK2KTuwcz1Vja9EU0iTdyUZrd/THS5rzWz\nzwHPpRChHYgf/0fTy8YSY5efAO5395bfNE1COQTSP//sNMP7NaJ7/Q53vxHiS5VOvQo40MyOBv5M\ndNteRYhnLXuWF5IkksOJ7udWxNhjVvyeAeYCe5nZUcCdwGeAB4A7kuC8GFaSF8mGp9PTI4D/MbPN\n3f3UzGnPEpNuw4l2jSc85w95LG1cbfa3wvVzbWO98FWz2+cQY+wfA+Znf+zNbCZwsZl9lYhoeDvw\nWuBub41dBaoioVwzRhDdi5cB+5rZXHd/IX2pVprZD4k13v9FdLeXAwe2skhm2Ivonm4OXJvEbx2i\nq73MzD5LTHqcSLT9duDwdJ7lJSBpX6V13f1RXxUC9HLgB8SWqScn4Ts5veRyopt5U9pB9A2EYP4z\nXc8GEEnL/KBs5+63NrxxDSC1w9PjSYRIHg9clektlPhTqjuFcB6WAf+vXUQS0BhlLQdlA/bEYPWO\nRJe7P/0dVvaazYh/rnfTwrN9A9yPdxPd7iXAmzLlpTGrTYn9ivZmVdhNd472rkNMsN1ITNisS4w/\nnpfqxwAnEcJ+UuZ17wX+N73u/Ez7Kk5CsWqSq4sQ4AXApLw/ryHcpw2IhRDl5fsRk5GvK7+vZc+3\nIoai2i5cKncDin6w+kD28PIvATEpUBLL0j/SCGBK3rbXue1W9vzdxMTN30ptLT+n0nVybMurgf8D\nFhEz2b/L/oCVieUpWduJLnfpR7Ki4Gc++/WBtxGTG4uJsKnd825/jffo6GRv+Y/+3uk7vn/mnmTD\nw3qBHfO2v6H3Jm8DinyUCcPpwF/Tl/83xPrlUozYfwMr0j/a64Gz0z/cGFo0bCTjHY0Cvg78KnlV\nn8mc8x5iVvOvwF5521xDm3rS5/I08NYK9VmxPLFC/aCfZbpXd6Xvx4XAT4ju+p+APfJufw33ZyNW\nLYo4OFO+IREwfiuwfdlrtkjfjf+kjWf/czegFY70hb8vCcbxRJfq/iSQI9I5pVUJ9yWPpWW6XBXa\nW/KeRhGrim5O/wx/JSY5fsuqUKj3A9cnwdwvb9urtGc/4FIizOefwMRsfXo8Jn2uK4GPDvF9vkVM\nbr0yUzaVGKb4E63jWR7ESz3rw1I7/pp+cIYTM+DnEMMY2+Rtd0PvSd4GFOkgxmhmAKMzZVOI7tqB\nmbJhSTzvBT6S+Ud8HbFK5ZXNtLtB98KA7xBd6+1S2agkjIuByzLnHkLMcM/O2+6yNpSPoRkxLLIz\nseRyUSXxAtZLAjekcVViFc+1xBjoOpnyXmKM77pW+AEFXkF41v3AlzLlHyeiOF4gIhnuJsKkJuZt\nc8PvSd4GFOUgVhosIsaWRmbK30Mketg8PS95Uuslz+SvtGj3uoZ7cjnwq7Ky9Yilic8C/5kpfyMF\nGIvM2FMaM1yPWGv8mozgG7BLRix3TeVj0w/gtuXXqfE9/we4N3uvMo8vAh4jurC7DLU9DbxPA01M\nvYLIK7kS+HKmfBvgw6nuCDKB9+185G5AEY4kkvcAV1A2Q00ETK8gsuCUykpiuV/6Ir22ncSSmCVe\nl1jL++tU1p2pHwXMIbpd5V5b7mLJ6kss/0gspXuIWCF0cKozwrO8mliXfBoxu31XNXEcRFx2Tx7W\neWX3ch3gYmLI4mbgq+n9c/3OsPoY/AeIXJvZKIYtMmL5pWbbV6Sj4xP3pni6a4lB96megokzPJ7q\nDzezPSBW4KS6MUQ39FFP36xWpELC4ZUeeSR/DbzDzA70WM1SSs67jGj3aC+LJfQCBFp7xDGOJD63\nLkIADiaE8xIzOyJ9XrcRAdSXESuN7iNmb1eU35MSKfdmv5mNMLMPmtn7MrtNLiQySE0xs/NTfOl6\nRNjMRsBXCGE+nOi15Pqd8VUrbi4mliHOBn5gZt9P9Q8Ds4BTgePN7KTSa0vJjTuGvJU6z4MYk7yb\n+MW8A9iXVXF/2QH+dxID1hcBB6SyrYlYuZto4bWsrOqiDiM8or3S45JXdgEx9PDOzGt2JCZ5Ts3b\n/gHatA6xcuoaVg2ZXERMwP0ifd6Hlb3m5VQJAcqcO5qY2HicCK5eyipPdWPgOMKzfJAY430AuDXV\nH0UIcm7fGVb3JN9D5JLcl4iB/DIx9n5J5pwtWDXB9fm8P99c7lneBuT4ZdmAmIC4nFhFsyD98785\nIxLlX6i/EisP7kvnLgYm5N2WOtyL0cTSs8dYlbjjRMIb2ozYAqGUP/O3SSRuqiYoObZnPWLi4aPp\n+blE13sXIp7y1tSe91d4bcXucOY7YUSEw+XE9hZvIiZxnsu833BiyObbxI/pqaz6QbqC8CpH1rPN\na3ifPkps2fDdTPs2BD6dxP0XmXPHpR+AnfKwNe8jdwNy+oJ0JcG7hjQmSXSNbksCuN8AYrkL8K70\nxTqSzKB/qx1l//g/J8byDiImPn6S7sWFrIqr+wixSuWXxHhe6R8/d7GsJG5Ed3ddYvz4TqLrXfIY\nTyRmbleSiWaocI11y56PSGJxEfCJTPmmRJf7BeAjA1xrx3Q/n6TJkznEmPJPWT2aY4fU/pVElvby\n80tieVH5d6YTj9wNyK3hMAHYIj0u/QNtXE0s2+kgVpG8nwiQf1umfBQRL3obsV69dB+6y15fBJEs\nCXZpAqrcxkOIrvFrMmXfJMbjBgwBIlajnFomLscQ3e1HgH1K75v57pxJeJYfLrvWbsSM+Dxgtxzu\nUWkLjpFl5a9Nwn1P9v5kvgNHEUML56eytpmwHPI9zNuAohxlX/hKYtlWXxLCkzw1eRTPAG9I5aUZ\n/VGEB3kLybNi9djA3O9H5jMbTQwP/IUYg/xs5py3E3kzpxNe5jbEUsajM+e8RCwZOF/ll4nQqN+T\nxj8zdRsnQVxJGsvO1O1MyteYx33K3KvPklmGS+QjWE5MaO1a9roN0n3bvlm2FvXI3YAiHdXEst0O\nYI/kWa0ATs6UD8vUrwRen7etFWwv9QKGE5MRNxFjhZcmIbuUVYlkzyMSeNybjprHV4kExTOAV2XK\nvkKEEZ1F2aQM0Q0/hlWebu4/KBnbdiSCyK8BNsmUT0k/lpXEsjD253rv8jagaEcFsbyVSHLQ0mI5\nkP3EEMSF6R/oE2V17yUmeJreXazSlpJIdhFhPVeTJhmScL4V+BcpBjSVf5LwoP+bAbIAEdtXnERm\nbJJIPruSSCGWTaLxDSJS4mxW7a9jZdfLfWii3J70w/9ABbHcK4nlpbTIUsum3ru8DSjiUSaWi4gQ\nj/Xztmst2lMaPlg//eN/hMxSuiSW5yexPJWYyT2Q6MreQAGSHbD6KhcjZrb/lkTyNxXOfxvhWR43\n2D0pu+aRGVHMTuKVQmO+xOrruL9B9DrOIKdu9QBtq5jFKdUNS2L5UAWx3DO182fZ+61DQjnwjVkl\nlhvRwrPbmfaMJDzk+5MgziezuRarPMsVxLrks4jZ71KGpDw3AtsR+B5lM8rEpMzK5CGVZ7UZToQy\nXZw8qapdSGJW+6h0f75UJs4nDSCWpwFPkcmqlON96qrwA/DpZOOxrBqKGEwsX0dme2Ed6b7kbUCR\njyJ4UnVoQ6mb2kfE/u1AxI3+JHlDp2XOnUjE/f0TmJYpH9ZMm8vsfz2xKOA64OwK9ccnATuZsj2j\nUxuvZWjrtYcncakmltlu+Cfz/CHJ2P1LYvKl9JlfnD7Lu5MoLiSNtZaJ5VW0YbLdut7fvA3Q0aAP\ntmwygVjL+5VM/RbE+NqdZWLZQ0yKPEZmt8mc2jAZeIKYSR6fKS/3mr6UhO0UVu0tvhuxDcU5Nb5X\ndiXW+jWI5alkJngq2dXkezUy/fDdRgyt7EkMS7ye6BW9g1jL/i9W7aI4jFiR8wyxXLXlHYOG3d+8\nDdDRgA919bCZ7xO5JG8Evlh23tgklncAX82U70rse7OS2O8mjzYMI2arL6SG5X7ETPRKYqnp5YQH\n+mdWBcwPuuKmQvkIYva6kliWxix78/6sK3ze1xLxmt8mEgiXIhgsieffysRyOJH5qeNDgAa9x3kb\noKPOH+jqYTNziDCWy4lxx3+RCSxP540lAqWXANMz5bsT45Tjm2V7mV3DiIiD08vKNyf2Vp9NeJD7\nZupOSAJ2CZkM5pStsMmUlya5Ribx+zIRZ7hx6V4Su0hWEsuPUZBZ7WRnSSw3IBIpryQ2fys/tySW\nD+X12bbikbsBOur4Ya4eNvMGItRj+1S2e/K2rgbeXPa6LYjB/vIubW4zn0T39y9JEDdMYvB6Igby\nhSTszxBDBx/LvO7k9KNwDGkiqsq9Gpl+TO5Nfx9Pf9+ffmwsXesFortdvp9MrmLJ6rPzY9Lf0URA\n/PPAtPLPMX03FqZjXRQrWf0+522Ajjp/oDHDOy8dl5TV9RCz3tcAbxng9YWJFyUCoZ8HriQ2A3s2\nCdopRLKO3YgxuTmsnmy5tHrmv4BRA7WRWLHSl669NTGWtxmx7v2fwHvTeSOIPWEK090u/6wIj/dH\nQE96XkofeAcxZlm+bv21wNZ5t6FVjtwN0NGADzW6j88TXddtUlnJg+ohsh/9gdh7OXd7q7RlLyJr\n033EOGRPWf2BScB2Lyv/DjEhtfEA110/Xe83wP9UqL+a8FY3Ss9HJi+zEN3tMlsvItZrH8Pq2dlH\nE2O1FcVSxxDucd4G6FjLD7DC+uvkKU1PAnIasGFZfQ+RvOElAlHEgwguf1mF8i5idvpWKkz4MEjI\nC9GNfzbdh+9l3yv9fTkx435MhdcWRixZld9ybzKRDpnHJbG8lVhsUBjbW+no+AznrUzKtr3SzNYz\ns22IcBoA3P1MYtzxM8BnzWxDd3czM3efQwSYH5WP5UPD3Z9396cg2pz+GtFdPhiYS4wtlr9ucelx\neUZud/8zsfzxUeD9ZvaO0nulU1YQqfhGV7juirVvVd3YnRiOmFOyy0MhV6bHS4EDCPH8GOEZiyHS\nnbcBYs0wsy6PLQtGE8lgXwGMM7N5wPfN7AJ3P93MVhJ7kruZnebu/wZw939krpP79g21ktq8EdEl\n/xzxj/+x9COwjpdtTQEv/qCsSGLZ7bHNBe7+JzM7gkjse0z6EflNetlGxP/Hs01o1pBJ20wYsB2w\n2N2XZz/L0n0wsx3c/Q4zex2xAmdJfla3LqWumGgBykXNzIYTA/ZPEaE8dxNe4luIEJnPuPsLZnYU\nsSf5TGLt89NNN75OmNkGRLagJcTE1HuTCFYU/JJ4ph+UM4FNiLHH7xE7Jvab2f5EvOlIIrVcFzCe\n2BNpQhE8yOyPQBJ0T4+PJ0Kb3uTufyn9KKS6/YjJqs+5+7152d4OqOvdIpjZFOBYMxuRKX4D8Y9/\nEpEpZz4xo70FsCDjOX2HCG2ZSOQebFmSR/xuYguDQ5JIdg/kFSeRHEZMXk0kPMTDCEHcN732SmKy\n4xmie/oyYifFnX2QjcaaRfoRKInkJsS2DCUuIXKGftfMJmdEcgzwISLu9N9NNrn9yHuQVEf1g+hm\nrgS+xerbxk4lPKtSYt0PE8HRx6fnowkxXW3DNNoobo4Blt2x+iTXOGK7i+2I7upYIlbyDmL8rjTx\n8SYiwcb/kjKY532/ytoxixDFpUTy4cMI7/cdRIjU44R3+XUihvZJCpYir1UPjVEWHDN7AxHX93Xg\nFF+9G3gb0SvY3cy2IJb8fcHdv5K8oBlEqribgWXuL07mtM14iw8+JrkeIYpvABx4KLV9kZn1EOLy\nLeBoM/u9u19tZr2EII00s/Xc/co875ev8iTPI5YafpMQ8w8Sovhad/+UmS0lNsDrIzzIfwB7ufuC\nXAxvN/JWah0DH0SG8acJ72B0hfptidUrc4lZ2s9k6nYgxi+/kXc7crp3o4n4y4eJsdtbWLVKqeRB\nvoxYnbKA2DSuFIi+H7FXzMU0OQ8psQRxBqvv1bNLasN7WRW+tAvRy/gmq3udmxJe84i8P4N2OjRG\nWVDMbG9ize4wosv4mjTT+WKoi7vfTXia2xL/7PNT/b5EUothxK6BHbFhfSl0KHEeIXbfILrd2wIn\nmNkIXzWu+RQRU/pK4H2E14m7/4Hozh7v7k0b000TTguJDdGynvLmRFznPHd/3sx2IgLiLwJO8BiH\nnWhm67r7Yx6KWcjZ+pYlb6XW8dKDCB5eTizF24JV+/cMtOf4IYRn+QgRF3g7MXmxbvm57X4QM9fv\nIn5A9k5lw4nx2+XAOSQvkVWe5UhWX9bY9DFJwgO+hwj12rKsbsdk+5uIMLAniCzko1L9YcQQwibN\ntrtTDoUHFQwz25AYhP8e4S0sNbONidUV6xDdsms8wlpeDIkxs22JcJbtCa/kRg9P48VwkXYnec0n\nE1mEniI2RbsjlXcRSxBnE3tcz3D3Z8ruYS4xpWY2ivh81yWyOz1YVj+WSEJsxMz95USu0GfMbFNi\nuea6RDzp0qYa3ynkrdQ6XnoQXe2S11NtG90BvUU6MBErsTrpbGL2/9Nldd1EyMy/iW1tc8vcnrFp\nA2L8cSUxC78vZVEK6fG7iR/Q+4G3p7KdifjPf6HtGxp6yKNsATJB0wN5lm01k10rg6zE2Y1IDPI+\nYmfJczJ13UTM5AeJDEoveX2zSMHzc4lQpWOIMccuYmuJq/2lvYb3EN7yhkTSk2eJLEGHuPtNOTSh\nY5BQtgjVxDJf65pPJgRoGDHcMBa4093vT/UTiIQZBwNHl4llVnwqim0T7O8iPMR5wGHu/lBamnk9\ngw+x7ApsBbyGENl5XtZVF/VHQtlClInltan4WOCKPD2jZlMSjsw699HEhMd1wO/d/cvpvInA0cBB\nwH+6+7k5mVyRJOaPufvDpV5BLePRovlIKFuMMrFcSAjDB/O2q9mY2frEj8USQgyXETP9I4kNxT6X\nzptAJN09AjjQ3S/Lx+LqaIiluEgoW5DMP9QGwNOd5mmkWewTiQzohyeP7EIifOY6YiXOOe7+X+n8\nPYC3EhuoFToCQEMsxURC2aKUZZNp+25Z2TidEXvB/Nvdf2xms4G3E0L5DLHr5G7E9rwnlF2n8OFS\nFcRyJZFX9Pft/jkXFQmlKDyZiZsRxP7kNxBLE18AdiJW3nwO+EU672PEvjqjiFjU7+Vk+hpTJpa3\nEeu79/EmrhQSq1BSDFFobPUExf9HjEmOAM5Kkx9bEkl2H/BVyXnHE+u8ryFyULYcSSTXcfcnzGxH\nYu8fiWROSChFoUmTGOsTIvkY8AXg5sykxoPEBM4bzexhYgb8dcDPPLbDaNmhiYxYPkmEEomcUNdb\nFB4z+wgRQP4ud78jle1KJIpYRsRR/pBY5w6x1eweRR+LFK2DPErRCmxKpJG7y8xeToxTnkisTNmM\nyPDeA7yTWOP9Pz7I9hBCDBV5lKLwpFjIvxF75XQDrwaOT2UHEHGU27j745nXSCRF3ZBHKQqPu883\nszcBHyelkHP3eQBm9lpirXR/2WskkqJuyKMULYmZrUsk451NhAp9UKtWRKOQUIqWI+1E+DHgQCJW\nco80JplLggvR/mgrCNGKTCImbu5jlUh2SyRFo5BHKVqOFFT+SuDBFHSuiRvRUCSUoqVRd1s0Awml\nEEJUQWOUQghRBQmlEHUunSkAAAAlSURBVEJUQUIphBBVkFAKIUQVJJRCCFEFCaUQQlRBQimEEFX4\n/7jrTfn5v++mAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1107f0c50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig2, ax2 = plt.subplots(figsize=(4.5,6))\n",
    "ax2.spines['top'].set_color('none')\n",
    "ax2.spines['right'].set_color('none')\n",
    "ax2.tick_params(axis='both', which='major', labelsize = 12)\n",
    "ax2.set(ylim = (-1, 1))\n",
    "for label in ax2.get_xmajorticklabels():\n",
    "    label.set_rotation(45)\n",
    "    label.set_horizontalalignment(\"right\")\n",
    "    label.set_fontsize(15)\n",
    "    \n",
    "color_list = [\"#939598\",\"#e82d35\",\"#1eb3ec\",\"#8b579b\"]\n",
    "flierprops = dict(marker='o', markerfacecolor='black', markersize = 2,\n",
    "                  linestyle='none')\n",
    "bp0 = plt.boxplot([cors_total, cors_T, cors_S, cors_N], flierprops = flierprops, patch_artist = True,\n",
    "                  labels = ['Total','Temporal','Spatial\\nsampling','Techinical'],\n",
    "                  widths = 0.5, whis = 3)\n",
    "for item in ['boxes', 'medians', 'caps']:\n",
    "        plt.setp(bp0[item], color = \"black\", linewidth = 2, linestyle = \"-\")  \n",
    "for item in ['whiskers']:\n",
    "        plt.setp(bp0[item], color = \"black\", linewidth = 2, linestyle = \"--\")\n",
    "for patch, color in zip(bp0['boxes'], color_list):\n",
    "        patch.set_facecolor(color)\n",
    "\n",
    "plt.axhline(y = 0, color = 'grey', linestyle='--', alpha = 0.5, linewidth = 1.5)\n",
    "plt.ylabel('Pairwise correlation', fontsize = 15, fontname=\"Arial\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Covariance decomposition by phyla\n",
    "we will get indices for four dominant phyla and get correlations (total, temporal, spatial and technical) within and between phyla"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "phyla_inds = [\"Actinobacteria\", \"Bacteroidetes\", \"Firmicutes\", \"Proteobacteria\"]\n",
    "otu_phyla = []\n",
    "for each in phyla_inds:\n",
    "    tmp_if = [e for e in df_taxonomy_highAbundance[\"phlum\"] == each]\n",
    "    tmp_index = [i for i in range(len(tmp_if)) if tmp_if[i] == True]\n",
    "    tmp_otu = np.array(data_variance_decomposition_highAbundance[\"OTU_ID\"])[tmp_index]\n",
    "    otu_phyla.append(tmp_otu)\n",
    "\n",
    "cors_phyla_total = np.zeros((len(phyla_inds), len(phyla_inds)))\n",
    "cors_phyla_T = np.zeros((len(phyla_inds), len(phyla_inds)))\n",
    "cors_phyla_S = np.zeros((len(phyla_inds), len(phyla_inds)))\n",
    "cors_phyla_N = np.zeros((len(phyla_inds), len(phyla_inds)))\n",
    "\n",
    "for i in range(len(phyla_inds)):\n",
    "    for j in range(i,len(phyla_inds)):\n",
    "        tmp_phyla_1 = phyla_inds[i]\n",
    "        tmp_phyla_2 = phyla_inds[j]\n",
    "        tmp_cors_total = data_correlation_total.loc[otu_phyla[i],otu_phyla[j]]\n",
    "        tmp_cors_T = data_correlation_decomposition_T.loc[otu_phyla[i],otu_phyla[j]]\n",
    "        tmp_cors_S = data_correlation_decomposition_S.loc[otu_phyla[i],otu_phyla[j]]\n",
    "        tmp_cors_N = data_correlation_decomposition_N.loc[otu_phyla[i],otu_phyla[j]]\n",
    "        \n",
    "        if i == j:\n",
    "            triu_otu_pair = []\n",
    "            for p in range(len(otu_phyla[i])):\n",
    "                for q in range(p + 1,len(otu_phyla[i])):\n",
    "                    triu_otu_pair.append([otu_phyla[i][p],otu_phyla[i][q]])\n",
    "\n",
    "            cors_phyla_total[i,j] = np.mean([tmp_cors_total.loc[e[0],e[1]] for e in triu_otu_pair])\n",
    "            cors_phyla_T[i,j] = np.mean([tmp_cors_T.loc[e[0],e[1]] for e in triu_otu_pair])\n",
    "            cors_phyla_S[i,j] = np.mean([tmp_cors_S.loc[e[0],e[1]] for e in triu_otu_pair])\n",
    "            cors_phyla_N[i,j] = np.mean([tmp_cors_N.loc[e[0],e[1]] for e in triu_otu_pair])\n",
    "        else:\n",
    "            cors_phyla_total[i,j] = np.mean(np.mean(tmp_cors_total))\n",
    "            cors_phyla_total[j,i] = np.mean(np.mean(tmp_cors_total))\n",
    "            cors_phyla_T[i,j] = np.mean(np.mean(tmp_cors_T))\n",
    "            cors_phyla_T[j,i] = np.mean(np.mean(tmp_cors_T))\n",
    "            cors_phyla_S[i,j] = np.mean(np.mean(tmp_cors_S))\n",
    "            cors_phyla_S[j,i] = np.mean(np.mean(tmp_cors_S))\n",
    "            cors_phyla_N[i,j] = np.mean(np.mean(tmp_cors_N))\n",
    "            cors_phyla_N[j,i] = np.mean(np.mean(tmp_cors_N))\n",
    "            "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Heatmap of average total correlations within and between phyla"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAWMAAAEYCAYAAACJJ5fjAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1108a6710>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "data_to_draw = cors_phyla_total\n",
    "\n",
    "color_list = np.array(((68, 191, 236),  (255 ,255 ,255), (240, 135, 49))) / 255\n",
    "cm = LinearSegmentedColormap.from_list(\"heatmapColor\", color_list, N=100)\n",
    "valfmt = StrMethodFormatter(\"{x:.3f}\")\n",
    "\n",
    "fig3, ax3 = plt.subplots()  \n",
    "im = ax3.imshow(data_to_draw, interpolation = 'nearest', cmap = cm, vmin = -1, vmax = 1)\n",
    "ax3.set_xticks(np.arange(len(phyla_inds)))\n",
    "ax3.set_yticks(np.arange(len(phyla_inds)))\n",
    "ax3.set_xticklabels(phyla_inds, fontsize = 13)\n",
    "ax3.set_yticklabels(phyla_inds, fontsize = 13)\n",
    "plt.setp(ax3.get_xticklabels(), rotation = 45, ha=\"right\",\n",
    "        rotation_mode = \"anchor\")\n",
    "for i in range(len(phyla_inds)):\n",
    "    for j in range(len(phyla_inds)):\n",
    "        text = ax3.text(j, i, valfmt(data_to_draw[i, j]),\n",
    "                        ha=\"center\", va=\"center\", color=\"black\", fontsize = 10)        \n",
    "for edge, spine in ax3.spines.items():\n",
    "        spine.set_visible(True)\n",
    "cbar = ax3.figure.colorbar(im, ax = ax3, ticks=[-1, -0.5, 0, 0.5, 1])\n",
    "cbar.ax.set_ylabel(\"Pairwise correlation\", rotation = -90, va=\"bottom\", fontsize = 13)\n",
    "ax3.set_title(\"Total correlations\", fontsize = 15)\n",
    "fig3.tight_layout()    \n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Heatmap of average decomposed correlations (temporal, spatial and technical) within and between phyla"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA9UAAAD7CAYAAABkBLfEAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nu2QguQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1108265f8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "data_to_draw = [cors_phyla_T, cors_phyla_S, cors_phyla_N]\n",
    "label_list = [\"Temporal correlations\", \"Spatial sampling correlations\",\"Technical correlations\"]\n",
    "\n",
    "color_list = np.array(((68, 191, 236),  (255 ,255 ,255), (240, 135, 49))) / 255\n",
    "cm = LinearSegmentedColormap.from_list(\"heatmapColor\", color_list, N=100)\n",
    "valfmt = StrMethodFormatter(\"{x:.3f}\")\n",
    "\n",
    "fig4, ax4 = plt.subplots(1, 3, figsize=(14, 3.5))\n",
    "\n",
    "for tmp_i, tmp_ax in enumerate(ax4):\n",
    "    im = tmp_ax.imshow(data_to_draw[tmp_i], interpolation = 'nearest', cmap = cm, vmin = -1, vmax = 1)\n",
    "    tmp_ax.set_xticks(np.arange(len(phyla_inds)))\n",
    "    tmp_ax.set_yticks(np.arange(len(phyla_inds)))\n",
    "    tmp_ax.set_xticklabels(phyla_inds, fontsize = 13)\n",
    "    tmp_ax.set_yticklabels(phyla_inds, fontsize = 13)\n",
    "    plt.setp(tmp_ax.get_xticklabels(), rotation = 45, ha=\"right\",\n",
    "             rotation_mode = \"anchor\")\n",
    "    for i in range(len(phyla_inds)):\n",
    "        for j in range(len(phyla_inds)):\n",
    "            text = tmp_ax.text(j, i, valfmt(data_to_draw[tmp_i][i, j]),\n",
    "                              ha=\"center\", va=\"center\", color=\"black\", fontsize = 8)        \n",
    "    for edge, spine in tmp_ax.spines.items():\n",
    "            spine.set_visible(True)\n",
    "    cbar = tmp_ax.figure.colorbar(im, ax = ax4[tmp_i], ticks=[-1, -0.5, 0, 0.5, 1])\n",
    "    cbar.ax.set_ylabel(\"Pairwise correlation\", rotation = -90, va=\"bottom\", fontsize = 13)\n",
    "    tmp_ax.set_title(label_list[tmp_i], fontsize = 15)\n",
    "\n",
    "fig4.tight_layout()    \n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "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": "3.6.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}