BH_KOs_clustermap.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## This script takes relative abundances of all genes and finds those with the biggest difference between 60m and Deep metagenomes (calculated as the absolute value of difference in means). Then a heat map of these genes is generated using the Seaborn clustermap function."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os as os\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Navigate to directory containing read counts file\n",
    "os.chdir(\"/Users/nastassia.patin/Desktop/Projects/BlueHole/ISMEJ_revisions/\")\n",
    "path = \"60Ms_vs_deeps_KOs_readcounts.csv\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "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>Group</th>\n",
       "      <th>Subgroup</th>\n",
       "      <th>Subgroup2</th>\n",
       "      <th>KO</th>\n",
       "      <th>Function</th>\n",
       "      <th>ReadCounts</th>\n",
       "      <th>Sample</th>\n",
       "      <th>GE</th>\n",
       "      <th>ReadCounts_Norm</th>\n",
       "      <th>dataset</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Metabolism</td>\n",
       "      <td>Carbohydrate metabolism</td>\n",
       "      <td>Glycolysis / Gluconeogenesis [PATH:ko00010]</td>\n",
       "      <td>K00845</td>\n",
       "      <td>glk; glucokinase [EC:2.7.1.2]</td>\n",
       "      <td>629</td>\n",
       "      <td>BH091960M</td>\n",
       "      <td>2401.2</td>\n",
       "      <td>0.261952</td>\n",
       "      <td>60Ms</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Metabolism</td>\n",
       "      <td>Carbohydrate metabolism</td>\n",
       "      <td>Galactose metabolism [PATH:ko00052]</td>\n",
       "      <td>K00845</td>\n",
       "      <td>glk; glucokinase [EC:2.7.1.2]</td>\n",
       "      <td>629</td>\n",
       "      <td>BH091960M</td>\n",
       "      <td>2401.2</td>\n",
       "      <td>0.261952</td>\n",
       "      <td>60Ms</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Metabolism</td>\n",
       "      <td>Carbohydrate metabolism</td>\n",
       "      <td>Starch and sucrose metabolism [PATH:ko00500]</td>\n",
       "      <td>K00845</td>\n",
       "      <td>glk; glucokinase [EC:2.7.1.2]</td>\n",
       "      <td>629</td>\n",
       "      <td>BH091960M</td>\n",
       "      <td>2401.2</td>\n",
       "      <td>0.261952</td>\n",
       "      <td>60Ms</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Metabolism</td>\n",
       "      <td>Carbohydrate metabolism</td>\n",
       "      <td>Amino sugar and nucleotide sugar metabolism [P...</td>\n",
       "      <td>K00845</td>\n",
       "      <td>glk; glucokinase [EC:2.7.1.2]</td>\n",
       "      <td>629</td>\n",
       "      <td>BH091960M</td>\n",
       "      <td>2401.2</td>\n",
       "      <td>0.261952</td>\n",
       "      <td>60Ms</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Metabolism</td>\n",
       "      <td>Biosynthesis of other secondary metabolites</td>\n",
       "      <td>Streptomycin biosynthesis [PATH:ko00521]</td>\n",
       "      <td>K00845</td>\n",
       "      <td>glk; glucokinase [EC:2.7.1.2]</td>\n",
       "      <td>629</td>\n",
       "      <td>BH091960M</td>\n",
       "      <td>2401.2</td>\n",
       "      <td>0.261952</td>\n",
       "      <td>60Ms</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        Group                                     Subgroup  \\\n",
       "0  Metabolism                      Carbohydrate metabolism   \n",
       "1  Metabolism                      Carbohydrate metabolism   \n",
       "2  Metabolism                      Carbohydrate metabolism   \n",
       "3  Metabolism                      Carbohydrate metabolism   \n",
       "4  Metabolism  Biosynthesis of other secondary metabolites   \n",
       "\n",
       "                                           Subgroup2      KO  \\\n",
       "0        Glycolysis / Gluconeogenesis [PATH:ko00010]  K00845   \n",
       "1                Galactose metabolism [PATH:ko00052]  K00845   \n",
       "2       Starch and sucrose metabolism [PATH:ko00500]  K00845   \n",
       "3  Amino sugar and nucleotide sugar metabolism [P...  K00845   \n",
       "4           Streptomycin biosynthesis [PATH:ko00521]  K00845   \n",
       "\n",
       "                         Function  ReadCounts     Sample      GE  \\\n",
       "0   glk; glucokinase [EC:2.7.1.2]         629  BH091960M  2401.2   \n",
       "1   glk; glucokinase [EC:2.7.1.2]         629  BH091960M  2401.2   \n",
       "2   glk; glucokinase [EC:2.7.1.2]         629  BH091960M  2401.2   \n",
       "3   glk; glucokinase [EC:2.7.1.2]         629  BH091960M  2401.2   \n",
       "4   glk; glucokinase [EC:2.7.1.2]         629  BH091960M  2401.2   \n",
       "\n",
       "   ReadCounts_Norm dataset  \n",
       "0         0.261952    60Ms  \n",
       "1         0.261952    60Ms  \n",
       "2         0.261952    60Ms  \n",
       "3         0.261952    60Ms  \n",
       "4         0.261952    60Ms  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.read_csv(path)\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Exclude irrelevant KEGG categories\n",
    "df = df.loc[(df['Group'] != 'Human Diseases') & \n",
    "               (df['Group'] != 'Brite Hierarchies') & \n",
    "               (df['Group'] != 'Organismal Systems') &\n",
    "               (df['Subgroup'] != 'Cellular community - eukaryotes')]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### For KOs, drop duplicates\n",
    "#### Some genes will belong to multiple categories, but we don't want to count them twice"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "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>Group</th>\n",
       "      <th>Subgroup</th>\n",
       "      <th>Subgroup2</th>\n",
       "      <th>KO</th>\n",
       "      <th>Function</th>\n",
       "      <th>ReadCounts</th>\n",
       "      <th>Sample</th>\n",
       "      <th>GE</th>\n",
       "      <th>ReadCounts_Norm</th>\n",
       "      <th>dataset</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Metabolism</td>\n",
       "      <td>Carbohydrate metabolism</td>\n",
       "      <td>Pyruvate metabolism</td>\n",
       "      <td>K01573</td>\n",
       "      <td>oadG; oxaloacetate decarboxylase (Na+ extrudi...</td>\n",
       "      <td>1</td>\n",
       "      <td>BH091995M</td>\n",
       "      <td>1283.9</td>\n",
       "      <td>0.000779</td>\n",
       "      <td>deeps</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Metabolism</td>\n",
       "      <td>Amino acid metabolism</td>\n",
       "      <td>Glycine, serine and threonine metabolism</td>\n",
       "      <td>K06720</td>\n",
       "      <td>ectC; L-ectoine synthase [EC:4.2.1.108]</td>\n",
       "      <td>1</td>\n",
       "      <td>BH091995M</td>\n",
       "      <td>1283.9</td>\n",
       "      <td>0.000779</td>\n",
       "      <td>deeps</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Not Included in Pathway or Brite</td>\n",
       "      <td>Poorly characterized</td>\n",
       "      <td>Function unknown</td>\n",
       "      <td>K09991</td>\n",
       "      <td>K09991; uncharacterized protein</td>\n",
       "      <td>2</td>\n",
       "      <td>BH091960M</td>\n",
       "      <td>2401.2</td>\n",
       "      <td>0.000833</td>\n",
       "      <td>60Ms</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Environmental Information Processing</td>\n",
       "      <td>Signal transduction</td>\n",
       "      <td>Two-component system</td>\n",
       "      <td>K13588</td>\n",
       "      <td>chpT; histidine phosphotransferase ChpT</td>\n",
       "      <td>2</td>\n",
       "      <td>BH091960M</td>\n",
       "      <td>2401.2</td>\n",
       "      <td>0.000833</td>\n",
       "      <td>60Ms</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Not Included in Pathway or Brite</td>\n",
       "      <td>Unclassified: genetic information processing</td>\n",
       "      <td>Protein processing</td>\n",
       "      <td>K19411</td>\n",
       "      <td>mcsA; protein arginine kinase activator</td>\n",
       "      <td>2</td>\n",
       "      <td>BH091960M</td>\n",
       "      <td>2401.2</td>\n",
       "      <td>0.000833</td>\n",
       "      <td>60Ms</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                  Group  \\\n",
       "0                            Metabolism   \n",
       "1                            Metabolism   \n",
       "2      Not Included in Pathway or Brite   \n",
       "3  Environmental Information Processing   \n",
       "4      Not Included in Pathway or Brite   \n",
       "\n",
       "                                       Subgroup  \\\n",
       "0                       Carbohydrate metabolism   \n",
       "1                         Amino acid metabolism   \n",
       "2                          Poorly characterized   \n",
       "3                           Signal transduction   \n",
       "4  Unclassified: genetic information processing   \n",
       "\n",
       "                                   Subgroup2      KO  \\\n",
       "0                       Pyruvate metabolism   K01573   \n",
       "1  Glycine, serine and threonine metabolism   K06720   \n",
       "2                           Function unknown  K09991   \n",
       "3                      Two-component system   K13588   \n",
       "4                         Protein processing  K19411   \n",
       "\n",
       "                                            Function  ReadCounts     Sample  \\\n",
       "0   oadG; oxaloacetate decarboxylase (Na+ extrudi...           1  BH091995M   \n",
       "1            ectC; L-ectoine synthase [EC:4.2.1.108]           1  BH091995M   \n",
       "2                    K09991; uncharacterized protein           2  BH091960M   \n",
       "3            chpT; histidine phosphotransferase ChpT           2  BH091960M   \n",
       "4            mcsA; protein arginine kinase activator           2  BH091960M   \n",
       "\n",
       "       GE  ReadCounts_Norm dataset  \n",
       "0  1283.9         0.000779   deeps  \n",
       "1  1283.9         0.000779   deeps  \n",
       "2  2401.2         0.000833    60Ms  \n",
       "3  2401.2         0.000833    60Ms  \n",
       "4  2401.2         0.000833    60Ms  "
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df2 = df.sort_values('ReadCounts_Norm').drop_duplicates(subset=['Sample', 'KO'], keep='last')\n",
    "df2 = df2.reset_index(drop=True)\n",
    "df2.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Get a distribution of read counts\n",
    "ax = df2['ReadCounts_Norm'].plot.hist(bins=50, alpha=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Split the data frame by dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_G1 = df2.loc[(df2['dataset'] == '60Ms')]\n",
    "df_G2 = df2.loc[(df2['dataset'] == 'deeps')]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_G1_sub = df_G1[['KO','Sample','ReadCounts_Norm']]\n",
    "df_G2_sub = df_G2[['KO','Sample','ReadCounts_Norm']]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Concatenate the two data sets while maintaining the information as to which row is from which data set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "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>KO</th>\n",
       "      <th>Sample</th>\n",
       "      <th>ReadCounts_Norm</th>\n",
       "      <th>dataset</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>K09991</td>\n",
       "      <td>BH091960M</td>\n",
       "      <td>0.000833</td>\n",
       "      <td>60Ms</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>K13588</td>\n",
       "      <td>BH091960M</td>\n",
       "      <td>0.000833</td>\n",
       "      <td>60Ms</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>K19411</td>\n",
       "      <td>BH091960M</td>\n",
       "      <td>0.000833</td>\n",
       "      <td>60Ms</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>K15761</td>\n",
       "      <td>BH091960M</td>\n",
       "      <td>0.001249</td>\n",
       "      <td>60Ms</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>K08973</td>\n",
       "      <td>BH37</td>\n",
       "      <td>0.001520</td>\n",
       "      <td>60Ms</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       KO     Sample  ReadCounts_Norm dataset\n",
       "2  K09991  BH091960M         0.000833    60Ms\n",
       "3  K13588  BH091960M         0.000833    60Ms\n",
       "4  K19411  BH091960M         0.000833    60Ms\n",
       "5  K15761  BH091960M         0.001249    60Ms\n",
       "6  K08973       BH37         0.001520    60Ms"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "concat = pd.concat([df_G1_sub.assign(dataset='60Ms'), df_G2_sub.assign(dataset='Deeps')])\n",
    "concat.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "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>KO</th>\n",
       "      <th>K00003</th>\n",
       "      <th>K00005</th>\n",
       "      <th>K00009</th>\n",
       "      <th>K00010</th>\n",
       "      <th>K00012</th>\n",
       "      <th>K00013</th>\n",
       "      <th>K00014</th>\n",
       "      <th>K00015</th>\n",
       "      <th>K00016</th>\n",
       "      <th>K00018</th>\n",
       "      <th>...</th>\n",
       "      <th>K23492</th>\n",
       "      <th>K23508</th>\n",
       "      <th>K23514</th>\n",
       "      <th>K23518</th>\n",
       "      <th>K23535</th>\n",
       "      <th>K23536</th>\n",
       "      <th>K23537</th>\n",
       "      <th>K23548</th>\n",
       "      <th>K23549</th>\n",
       "      <th>K23573</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Sample</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>BH37</th>\n",
       "      <td>1.358456</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.308464</td>\n",
       "      <td>0.512080</td>\n",
       "      <td>0.743048</td>\n",
       "      <td>0.761282</td>\n",
       "      <td>0.057742</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.217292</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.945145</td>\n",
       "      <td>0.791673</td>\n",
       "      <td>1.186750</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BH091960M</th>\n",
       "      <td>0.592620</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.084541</td>\n",
       "      <td>0.534733</td>\n",
       "      <td>0.415209</td>\n",
       "      <td>0.906630</td>\n",
       "      <td>0.034150</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.285274</td>\n",
       "      <td>...</td>\n",
       "      <td>0.016242</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.032067</td>\n",
       "      <td>0.052057</td>\n",
       "      <td>0.853323</td>\n",
       "      <td>1.465517</td>\n",
       "      <td>0.294436</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.018741</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BH51</th>\n",
       "      <td>1.137606</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.072032</td>\n",
       "      <td>0.444610</td>\n",
       "      <td>1.043219</td>\n",
       "      <td>1.140089</td>\n",
       "      <td>0.094386</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.057129</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.074516</td>\n",
       "      <td>0.074516</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.146547</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.176354</td>\n",
       "      <td>1.266766</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BH091995M</th>\n",
       "      <td>0.840408</td>\n",
       "      <td>0.053743</td>\n",
       "      <td>0.031155</td>\n",
       "      <td>0.031155</td>\n",
       "      <td>1.314744</td>\n",
       "      <td>0.498481</td>\n",
       "      <td>0.774204</td>\n",
       "      <td>0.056858</td>\n",
       "      <td>0.014799</td>\n",
       "      <td>0.071657</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.022587</td>\n",
       "      <td>0.116832</td>\n",
       "      <td>0.038165</td>\n",
       "      <td>0.147987</td>\n",
       "      <td>0.153439</td>\n",
       "      <td>0.232884</td>\n",
       "      <td>0.102033</td>\n",
       "      <td>0.131630</td>\n",
       "      <td>2.562505</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>4 rows × 3374 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "KO           K00003    K00005    K00009    K00010    K00012    K00013  \\\n",
       "Sample                                                                  \n",
       "BH37       1.358456  0.000000  0.000000  0.308464  0.512080  0.743048   \n",
       "BH091960M  0.592620  0.000000  0.000000  0.084541  0.534733  0.415209   \n",
       "BH51       1.137606  0.000000  0.000000  0.072032  0.444610  1.043219   \n",
       "BH091995M  0.840408  0.053743  0.031155  0.031155  1.314744  0.498481   \n",
       "\n",
       "KO           K00014    K00015    K00016    K00018  ...    K23492    K23508  \\\n",
       "Sample                                             ...                       \n",
       "BH37       0.761282  0.057742  0.000000  0.217292  ...  0.000000  0.000000   \n",
       "BH091960M  0.906630  0.034150  0.000000  0.285274  ...  0.016242  0.000000   \n",
       "BH51       1.140089  0.094386  0.000000  0.057129  ...  0.000000  0.000000   \n",
       "BH091995M  0.774204  0.056858  0.014799  0.071657  ...  0.000000  0.022587   \n",
       "\n",
       "KO           K23514    K23518    K23535    K23536    K23537    K23548  \\\n",
       "Sample                                                                  \n",
       "BH37       0.000000  0.000000  0.945145  0.791673  1.186750  0.000000   \n",
       "BH091960M  0.032067  0.052057  0.853323  1.465517  0.294436  0.000000   \n",
       "BH51       0.074516  0.074516  0.000000  0.000000  0.146547  0.000000   \n",
       "BH091995M  0.116832  0.038165  0.147987  0.153439  0.232884  0.102033   \n",
       "\n",
       "KO           K23549    K23573  \n",
       "Sample                         \n",
       "BH37       0.000000  0.000000  \n",
       "BH091960M  0.000000  0.018741  \n",
       "BH51       0.176354  1.266766  \n",
       "BH091995M  0.131630  2.562505  \n",
       "\n",
       "[4 rows x 3374 columns]"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Reformat data frame for heat map\n",
    "df4 = pd.pivot_table(data=concat, index='Sample', values='ReadCounts_Norm', columns='KO')\n",
    "df4 = df4.fillna(0).reindex([\"BH37\",\"BH091960M\",\"BH51\",\"BH091995M\"])\n",
    "df4.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Calculate the fold difference between sample pairs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "def fold_difference(a, b, c, d):\n",
    "    diff = ((a + b)/2) / ((c + d)/2)\n",
    "    return(diff)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-14-00ed9ada64a9>:2: DeprecationWarning: The default dtype for empty Series will be 'object' instead of 'float64' in a future version. Specify a dtype explicitly to silence this warning.\n",
      "  df4 = df4.append(pd.Series(name='Average'))\n"
     ]
    },
    {
     "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>KO</th>\n",
       "      <th>K00003</th>\n",
       "      <th>K00005</th>\n",
       "      <th>K00009</th>\n",
       "      <th>K00010</th>\n",
       "      <th>K00012</th>\n",
       "      <th>K00013</th>\n",
       "      <th>K00014</th>\n",
       "      <th>K00015</th>\n",
       "      <th>K00016</th>\n",
       "      <th>K00018</th>\n",
       "      <th>...</th>\n",
       "      <th>K23492</th>\n",
       "      <th>K23508</th>\n",
       "      <th>K23514</th>\n",
       "      <th>K23518</th>\n",
       "      <th>K23535</th>\n",
       "      <th>K23536</th>\n",
       "      <th>K23537</th>\n",
       "      <th>K23548</th>\n",
       "      <th>K23549</th>\n",
       "      <th>K23573</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Sample</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>BH37</th>\n",
       "      <td>1.358456</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>0.308464</td>\n",
       "      <td>0.512080</td>\n",
       "      <td>0.743048</td>\n",
       "      <td>0.761282</td>\n",
       "      <td>0.057742</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>0.217292</td>\n",
       "      <td>...</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>9.451451e-01</td>\n",
       "      <td>7.916730e-01</td>\n",
       "      <td>1.186750</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>1.000000e-09</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BH091960M</th>\n",
       "      <td>0.592620</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>0.084541</td>\n",
       "      <td>0.534733</td>\n",
       "      <td>0.415209</td>\n",
       "      <td>0.906630</td>\n",
       "      <td>0.034150</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>0.285274</td>\n",
       "      <td>...</td>\n",
       "      <td>1.624188e-02</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>3.206730e-02</td>\n",
       "      <td>5.205730e-02</td>\n",
       "      <td>8.533233e-01</td>\n",
       "      <td>1.465517e+00</td>\n",
       "      <td>0.294436</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>1.874063e-02</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BH51</th>\n",
       "      <td>1.137606</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>0.072032</td>\n",
       "      <td>0.444610</td>\n",
       "      <td>1.043219</td>\n",
       "      <td>1.140089</td>\n",
       "      <td>0.094386</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>0.057129</td>\n",
       "      <td>...</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>7.451565e-02</td>\n",
       "      <td>7.451565e-02</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>0.146547</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>1.763537e-01</td>\n",
       "      <td>1.266766e+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BH091995M</th>\n",
       "      <td>0.840408</td>\n",
       "      <td>5.374250e-02</td>\n",
       "      <td>3.115507e-02</td>\n",
       "      <td>0.031155</td>\n",
       "      <td>1.314744</td>\n",
       "      <td>0.498481</td>\n",
       "      <td>0.774204</td>\n",
       "      <td>0.056858</td>\n",
       "      <td>1.479866e-02</td>\n",
       "      <td>0.071657</td>\n",
       "      <td>...</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>2.258743e-02</td>\n",
       "      <td>1.168315e-01</td>\n",
       "      <td>3.816497e-02</td>\n",
       "      <td>1.479866e-01</td>\n",
       "      <td>1.534387e-01</td>\n",
       "      <td>0.232884</td>\n",
       "      <td>1.020329e-01</td>\n",
       "      <td>1.316302e-01</td>\n",
       "      <td>2.562505e+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Average</th>\n",
       "      <td>1.029561</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>0.155012</td>\n",
       "      <td>0.497141</td>\n",
       "      <td>0.733825</td>\n",
       "      <td>0.936001</td>\n",
       "      <td>0.062093</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>0.186565</td>\n",
       "      <td>...</td>\n",
       "      <td>5.413960e-03</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>3.552765e-02</td>\n",
       "      <td>4.219098e-02</td>\n",
       "      <td>5.994895e-01</td>\n",
       "      <td>7.523967e-01</td>\n",
       "      <td>0.542578</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>5.878457e-02</td>\n",
       "      <td>4.285022e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Fold difference</th>\n",
       "      <td>0.986382</td>\n",
       "      <td>3.721449e-08</td>\n",
       "      <td>6.419500e-08</td>\n",
       "      <td>3.808671</td>\n",
       "      <td>0.594998</td>\n",
       "      <td>0.751286</td>\n",
       "      <td>0.871294</td>\n",
       "      <td>0.607570</td>\n",
       "      <td>1.351474e-07</td>\n",
       "      <td>3.902356</td>\n",
       "      <td>...</td>\n",
       "      <td>8.120940e+06</td>\n",
       "      <td>8.854482e-08</td>\n",
       "      <td>1.675870e-01</td>\n",
       "      <td>4.619899e-01</td>\n",
       "      <td>1.215291e+01</td>\n",
       "      <td>1.471069e+01</td>\n",
       "      <td>3.903696</td>\n",
       "      <td>1.960153e-08</td>\n",
       "      <td>6.493846e-09</td>\n",
       "      <td>4.894047e-03</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>6 rows × 3374 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "KO                 K00003        K00005        K00009    K00010    K00012  \\\n",
       "Sample                                                                      \n",
       "BH37             1.358456  1.000000e-09  1.000000e-09  0.308464  0.512080   \n",
       "BH091960M        0.592620  1.000000e-09  1.000000e-09  0.084541  0.534733   \n",
       "BH51             1.137606  1.000000e-09  1.000000e-09  0.072032  0.444610   \n",
       "BH091995M        0.840408  5.374250e-02  3.115507e-02  0.031155  1.314744   \n",
       "Average          1.029561  1.000000e-09  1.000000e-09  0.155012  0.497141   \n",
       "Fold difference  0.986382  3.721449e-08  6.419500e-08  3.808671  0.594998   \n",
       "\n",
       "KO                 K00013    K00014    K00015        K00016    K00018  ...  \\\n",
       "Sample                                                                 ...   \n",
       "BH37             0.743048  0.761282  0.057742  1.000000e-09  0.217292  ...   \n",
       "BH091960M        0.415209  0.906630  0.034150  1.000000e-09  0.285274  ...   \n",
       "BH51             1.043219  1.140089  0.094386  1.000000e-09  0.057129  ...   \n",
       "BH091995M        0.498481  0.774204  0.056858  1.479866e-02  0.071657  ...   \n",
       "Average          0.733825  0.936001  0.062093  1.000000e-09  0.186565  ...   \n",
       "Fold difference  0.751286  0.871294  0.607570  1.351474e-07  3.902356  ...   \n",
       "\n",
       "KO                     K23492        K23508        K23514        K23518  \\\n",
       "Sample                                                                    \n",
       "BH37             1.000000e-09  1.000000e-09  1.000000e-09  1.000000e-09   \n",
       "BH091960M        1.624188e-02  1.000000e-09  3.206730e-02  5.205730e-02   \n",
       "BH51             1.000000e-09  1.000000e-09  7.451565e-02  7.451565e-02   \n",
       "BH091995M        1.000000e-09  2.258743e-02  1.168315e-01  3.816497e-02   \n",
       "Average          5.413960e-03  1.000000e-09  3.552765e-02  4.219098e-02   \n",
       "Fold difference  8.120940e+06  8.854482e-08  1.675870e-01  4.619899e-01   \n",
       "\n",
       "KO                     K23535        K23536    K23537        K23548  \\\n",
       "Sample                                                                \n",
       "BH37             9.451451e-01  7.916730e-01  1.186750  1.000000e-09   \n",
       "BH091960M        8.533233e-01  1.465517e+00  0.294436  1.000000e-09   \n",
       "BH51             1.000000e-09  1.000000e-09  0.146547  1.000000e-09   \n",
       "BH091995M        1.479866e-01  1.534387e-01  0.232884  1.020329e-01   \n",
       "Average          5.994895e-01  7.523967e-01  0.542578  1.000000e-09   \n",
       "Fold difference  1.215291e+01  1.471069e+01  3.903696  1.960153e-08   \n",
       "\n",
       "KO                     K23549        K23573  \n",
       "Sample                                       \n",
       "BH37             1.000000e-09  1.000000e-09  \n",
       "BH091960M        1.000000e-09  1.874063e-02  \n",
       "BH51             1.763537e-01  1.266766e+00  \n",
       "BH091995M        1.316302e-01  2.562505e+00  \n",
       "Average          5.878457e-02  4.285022e-01  \n",
       "Fold difference  6.493846e-09  4.894047e-03  \n",
       "\n",
       "[6 rows x 3374 columns]"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df4 = df4.replace(0, 0.000000001)\n",
    "df4 = df4.append(pd.Series(name='Average'))\n",
    "df4 = df4.append(pd.Series(name='Fold difference', dtype='float64'))\n",
    "df4.loc['Average'] = np.array(np.mean(df4.iloc[:3, :], axis=0))\n",
    "df4.loc['Fold difference'] = df4.apply(lambda x: \n",
    "                                           fold_difference(x.loc['BH37'], x.loc['BH091960M'], \n",
    "                                                           x.loc['BH51'], x.loc['BH091995M']), axis=0)\n",
    "df4.head(6)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Get the distribution of average normalized read counts across all samples"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "ax = df4.loc['Average'].plot.hist(bins=50, alpha=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Select only the columns (KOs) with an average relative abundnace above 0.2 and a twenty-fold or greater difference"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "df5 = df4.loc[:, ((df4.loc['Average'] > 0.2) & \n",
    "                  ((df4.loc['Fold difference'] < 0.05) | (df4.loc['Fold difference'] > 20)))]\n",
    "df5.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "df5 = df5.sort_values(by='Fold difference', axis=1)\n",
    "df5.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Sample\n",
       "BH37               1.000000e-09\n",
       "BH091960M          1.874063e-02\n",
       "BH51               1.266766e+00\n",
       "BH091995M          2.562505e+00\n",
       "Average            4.285022e-01\n",
       "Fold difference    4.894047e-03\n",
       "Name: K23573, dtype: float64"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Get average and fold difference values for K25373\n",
    "df5['K23573']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "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>KO</th>\n",
       "      <th>K05349</th>\n",
       "      <th>K02567</th>\n",
       "      <th>K07716</th>\n",
       "      <th>K11942</th>\n",
       "      <th>K06959</th>\n",
       "      <th>K13040</th>\n",
       "      <th>K05367</th>\n",
       "      <th>K07641</th>\n",
       "      <th>K04561</th>\n",
       "      <th>K02319</th>\n",
       "      <th>...</th>\n",
       "      <th>K02182</th>\n",
       "      <th>K09472</th>\n",
       "      <th>K18593</th>\n",
       "      <th>K03338</th>\n",
       "      <th>K07577</th>\n",
       "      <th>K06954</th>\n",
       "      <th>K14260</th>\n",
       "      <th>K22187</th>\n",
       "      <th>K23464</th>\n",
       "      <th>K01637</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Sample</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>BH37</th>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>...</td>\n",
       "      <td>6.154080e-01</td>\n",
       "      <td>6.321228e-01</td>\n",
       "      <td>1.066707e+00</td>\n",
       "      <td>1.098617e+00</td>\n",
       "      <td>1.021121e+00</td>\n",
       "      <td>7.582434e-01</td>\n",
       "      <td>7.308920e-01</td>\n",
       "      <td>1.189789e+00</td>\n",
       "      <td>1.083422e+00</td>\n",
       "      <td>1.340222e+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BH091960M</th>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>...</td>\n",
       "      <td>3.452440e-01</td>\n",
       "      <td>3.960520e-01</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>1.415959e-02</td>\n",
       "      <td>1.465934e-01</td>\n",
       "      <td>4.193736e-01</td>\n",
       "      <td>5.234883e-01</td>\n",
       "      <td>1.961519e-01</td>\n",
       "      <td>5.805431e-01</td>\n",
       "      <td>1.033233e+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BH51</th>\n",
       "      <td>6.756085e-01</td>\n",
       "      <td>3.060109e+00</td>\n",
       "      <td>2.071535e+00</td>\n",
       "      <td>7.973174e-01</td>\n",
       "      <td>6.209637e-01</td>\n",
       "      <td>1.671634e+00</td>\n",
       "      <td>6.135122e-01</td>\n",
       "      <td>1.403378e+00</td>\n",
       "      <td>1.239444e+00</td>\n",
       "      <td>9.612519e-01</td>\n",
       "      <td>...</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>1.000000e-09</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BH091995M</th>\n",
       "      <td>4.618740e+00</td>\n",
       "      <td>9.253057e-01</td>\n",
       "      <td>9.790482e-01</td>\n",
       "      <td>1.671470e+00</td>\n",
       "      <td>1.336553e+00</td>\n",
       "      <td>8.801309e-02</td>\n",
       "      <td>1.084197e+00</td>\n",
       "      <td>2.102968e-02</td>\n",
       "      <td>8.411870e-02</td>\n",
       "      <td>1.386401e-01</td>\n",
       "      <td>...</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>1.000000e-09</td>\n",
       "      <td>1.000000e-09</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>4 rows × 101 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "KO               K05349        K02567        K07716        K11942  \\\n",
       "Sample                                                              \n",
       "BH37       1.000000e-09  1.000000e-09  1.000000e-09  1.000000e-09   \n",
       "BH091960M  1.000000e-09  1.000000e-09  1.000000e-09  1.000000e-09   \n",
       "BH51       6.756085e-01  3.060109e+00  2.071535e+00  7.973174e-01   \n",
       "BH091995M  4.618740e+00  9.253057e-01  9.790482e-01  1.671470e+00   \n",
       "\n",
       "KO               K06959        K13040        K05367        K07641  \\\n",
       "Sample                                                              \n",
       "BH37       1.000000e-09  1.000000e-09  1.000000e-09  1.000000e-09   \n",
       "BH091960M  1.000000e-09  1.000000e-09  1.000000e-09  1.000000e-09   \n",
       "BH51       6.209637e-01  1.671634e+00  6.135122e-01  1.403378e+00   \n",
       "BH091995M  1.336553e+00  8.801309e-02  1.084197e+00  2.102968e-02   \n",
       "\n",
       "KO               K04561        K02319  ...        K02182        K09472  \\\n",
       "Sample                                 ...                               \n",
       "BH37       1.000000e-09  1.000000e-09  ...  6.154080e-01  6.321228e-01   \n",
       "BH091960M  1.000000e-09  1.000000e-09  ...  3.452440e-01  3.960520e-01   \n",
       "BH51       1.239444e+00  9.612519e-01  ...  1.000000e-09  1.000000e-09   \n",
       "BH091995M  8.411870e-02  1.386401e-01  ...  1.000000e-09  1.000000e-09   \n",
       "\n",
       "KO               K18593        K03338        K07577        K06954  \\\n",
       "Sample                                                              \n",
       "BH37       1.066707e+00  1.098617e+00  1.021121e+00  7.582434e-01   \n",
       "BH091960M  1.000000e-09  1.415959e-02  1.465934e-01  4.193736e-01   \n",
       "BH51       1.000000e-09  1.000000e-09  1.000000e-09  1.000000e-09   \n",
       "BH091995M  1.000000e-09  1.000000e-09  1.000000e-09  1.000000e-09   \n",
       "\n",
       "KO               K14260        K22187        K23464        K01637  \n",
       "Sample                                                             \n",
       "BH37       7.308920e-01  1.189789e+00  1.083422e+00  1.340222e+00  \n",
       "BH091960M  5.234883e-01  1.961519e-01  5.805431e-01  1.033233e+00  \n",
       "BH51       1.000000e-09  1.000000e-09  1.000000e-09  1.000000e-09  \n",
       "BH091995M  1.000000e-09  1.000000e-09  1.000000e-09  1.000000e-09  \n",
       "\n",
       "[4 rows x 101 columns]"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Just get the sample rows\n",
    "df6 = df5.iloc[0:4, :]\n",
    "df6.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Match KOs with KEGG Groups"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "kos = list(df6.columns.values)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(106, 2)"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df6_func = df2.loc[df2['KO'].str.fullmatch('|'.join(kos))]\n",
    "df6_func = df6_func[['Group','KO']]\n",
    "df6_func = df6_func.drop_duplicates()\n",
    "df6_func.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Transform values as needed for a nice visual"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "df6_log = np.log(df6)\n",
    "df6_sqrt = np.sqrt(df6)\n",
    "df6_frthrt = np.sqrt(df6_sqrt)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Make color bar for heat map"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'Metabolism': (0.6509803921568628, 0.807843137254902, 0.8901960784313725),\n",
       " 'Not Included in Pathway or Brite': (0.12156862745098039,\n",
       "  0.47058823529411764,\n",
       "  0.7058823529411765),\n",
       " 'Environmental Information Processing': (0.6980392156862745,\n",
       "  0.8745098039215686,\n",
       "  0.5411764705882353),\n",
       " 'Genetic Information Processing': (0.2,\n",
       "  0.6274509803921569,\n",
       "  0.17254901960784313),\n",
       " 'Cellular Processes': (0.984313725490196, 0.6039215686274509, 0.6)}"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lut = dict(zip(df6_func['Group'].unique(), \n",
    "                sns.color_palette(\"Paired\")))\n",
    "lut"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(106, 3)"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df6_func['Color'] = df6_func['Group'].map(lut)\n",
    "df6_func.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "lut3 = dict(zip(df6_func['KO'], df6_func['Color']))\n",
    "column_colors = pd.Series(lut3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'K02568': (0.6509803921568628, 0.807843137254902, 0.8901960784313725),\n",
       " 'K07182': (0.12156862745098039, 0.47058823529411764, 0.7058823529411765),\n",
       " 'K02305': (0.6509803921568628, 0.807843137254902, 0.8901960784313725),\n",
       " 'K03281': (0.12156862745098039, 0.47058823529411764, 0.7058823529411765),\n",
       " 'K01434': (0.6509803921568628, 0.807843137254902, 0.8901960784313725),\n",
       " 'K05815': (0.6980392156862745, 0.8745098039215686, 0.5411764705882353),\n",
       " 'K03338': (0.6509803921568628, 0.807843137254902, 0.8901960784313725),\n",
       " 'K06863': (0.6509803921568628, 0.807843137254902, 0.8901960784313725),\n",
       " 'K08299': (0.12156862745098039, 0.47058823529411764, 0.7058823529411765),\n",
       " 'K03851': (0.6509803921568628, 0.807843137254902, 0.8901960784313725),\n",
       " 'K17241': (0.6980392156862745, 0.8745098039215686, 0.5411764705882353),\n",
       " 'K08975': (0.12156862745098039, 0.47058823529411764, 0.7058823529411765),\n",
       " 'K23573': (0.6980392156862745, 0.8745098039215686, 0.5411764705882353),\n",
       " 'K18277': (0.6509803921568628, 0.807843137254902, 0.8901960784313725),\n",
       " 'K18682': (0.2, 0.6274509803921569, 0.17254901960784313),\n",
       " 'K07641': (0.6980392156862745, 0.8745098039215686, 0.5411764705882353),\n",
       " 'K17315': (0.6980392156862745, 0.8745098039215686, 0.5411764705882353),\n",
       " 'K01623': (0.6509803921568628, 0.807843137254902, 0.8901960784313725),\n",
       " 'K12256': (0.6509803921568628, 0.807843137254902, 0.8901960784313725),\n",
       " 'K00425': (0.6509803921568628, 0.807843137254902, 0.8901960784313725),\n",
       " 'K09703': (0.12156862745098039, 0.47058823529411764, 0.7058823529411765),\n",
       " 'K11710': (0.6980392156862745, 0.8745098039215686, 0.5411764705882353),\n",
       " 'K11074': (0.6980392156862745, 0.8745098039215686, 0.5411764705882353),\n",
       " 'K19519': (0.12156862745098039, 0.47058823529411764, 0.7058823529411765),\n",
       " 'K00860': (0.6509803921568628, 0.807843137254902, 0.8901960784313725),\n",
       " 'K00453': (0.6509803921568628, 0.807843137254902, 0.8901960784313725),\n",
       " 'K22082': (0.6509803921568628, 0.807843137254902, 0.8901960784313725),\n",
       " 'K00479': (0.12156862745098039, 0.47058823529411764, 0.7058823529411765),\n",
       " 'K00130': (0.6509803921568628, 0.807843137254902, 0.8901960784313725),\n",
       " 'K00128': (0.6509803921568628, 0.807843137254902, 0.8901960784313725),\n",
       " 'K20023': (0.6509803921568628, 0.807843137254902, 0.8901960784313725),\n",
       " 'K07487': (0.12156862745098039, 0.47058823529411764, 0.7058823529411765),\n",
       " 'K07086': (0.12156862745098039, 0.47058823529411764, 0.7058823529411765),\n",
       " 'K15580': (0.6980392156862745, 0.8745098039215686, 0.5411764705882353),\n",
       " 'K09722': (0.6509803921568628, 0.807843137254902, 0.8901960784313725),\n",
       " 'K03337': (0.6509803921568628, 0.807843137254902, 0.8901960784313725),\n",
       " 'K00499': (0.6509803921568628, 0.807843137254902, 0.8901960784313725),\n",
       " 'K00042': (0.6509803921568628, 0.807843137254902, 0.8901960784313725),\n",
       " 'K01697': (0.6509803921568628, 0.807843137254902, 0.8901960784313725),\n",
       " 'K03841': (0.6509803921568628, 0.807843137254902, 0.8901960784313725),\n",
       " 'K00376': (0.6509803921568628, 0.807843137254902, 0.8901960784313725),\n",
       " 'K09967': (0.12156862745098039, 0.47058823529411764, 0.7058823529411765),\n",
       " 'K05829': (0.6509803921568628, 0.807843137254902, 0.8901960784313725),\n",
       " 'K06909': (0.12156862745098039, 0.47058823529411764, 0.7058823529411765),\n",
       " 'K07151': (0.2, 0.6274509803921569, 0.17254901960784313),\n",
       " 'K00758': (0.6509803921568628, 0.807843137254902, 0.8901960784313725),\n",
       " 'K09667': (0.6509803921568628, 0.807843137254902, 0.8901960784313725),\n",
       " 'K00150': (0.6509803921568628, 0.807843137254902, 0.8901960784313725),\n",
       " 'K03406': (0.6980392156862745, 0.8745098039215686, 0.5411764705882353),\n",
       " 'K07485': (0.12156862745098039, 0.47058823529411764, 0.7058823529411765),\n",
       " 'K00549': (0.6509803921568628, 0.807843137254902, 0.8901960784313725),\n",
       " 'K19200': (0.6509803921568628, 0.807843137254902, 0.8901960784313725),\n",
       " 'K08997': (0.12156862745098039, 0.47058823529411764, 0.7058823529411765),\n",
       " 'K07216': (0.12156862745098039, 0.47058823529411764, 0.7058823529411765),\n",
       " 'K00317': (0.6509803921568628, 0.807843137254902, 0.8901960784313725),\n",
       " 'K04561': (0.6509803921568628, 0.807843137254902, 0.8901960784313725),\n",
       " 'K13040': (0.6980392156862745, 0.8745098039215686, 0.5411764705882353),\n",
       " 'K07139': (0.12156862745098039, 0.47058823529411764, 0.7058823529411765),\n",
       " 'K09471': (0.6509803921568628, 0.807843137254902, 0.8901960784313725),\n",
       " 'K01023': (0.12156862745098039, 0.47058823529411764, 0.7058823529411765),\n",
       " 'K07495': (0.12156862745098039, 0.47058823529411764, 0.7058823529411765),\n",
       " 'K02044': (0.6980392156862745, 0.8745098039215686, 0.5411764705882353),\n",
       " 'K05962': (0.12156862745098039, 0.47058823529411764, 0.7058823529411765),\n",
       " 'K02319': (0.2, 0.6274509803921569, 0.17254901960784313),\n",
       " 'K07577': (0.12156862745098039, 0.47058823529411764, 0.7058823529411765),\n",
       " 'K14067': (0.6509803921568628, 0.807843137254902, 0.8901960784313725),\n",
       " 'K22187': (0.6509803921568628, 0.807843137254902, 0.8901960784313725),\n",
       " 'K06330': (0.12156862745098039, 0.47058823529411764, 0.7058823529411765),\n",
       " 'K07046': (0.6509803921568628, 0.807843137254902, 0.8901960784313725),\n",
       " 'K17225': (0.6509803921568628, 0.807843137254902, 0.8901960784313725),\n",
       " 'K17316': (0.6980392156862745, 0.8745098039215686, 0.5411764705882353),\n",
       " 'K01446': (0.12156862745098039, 0.47058823529411764, 0.7058823529411765),\n",
       " 'K17317': (0.6980392156862745, 0.8745098039215686, 0.5411764705882353),\n",
       " 'K01477': (0.6509803921568628, 0.807843137254902, 0.8901960784313725),\n",
       " 'K18456': (0.12156862745098039, 0.47058823529411764, 0.7058823529411765),\n",
       " 'K06137': (0.12156862745098039, 0.47058823529411764, 0.7058823529411765),\n",
       " 'K09735': (0.12156862745098039, 0.47058823529411764, 0.7058823529411765),\n",
       " 'K09696': (0.6980392156862745, 0.8745098039215686, 0.5411764705882353),\n",
       " 'K00166': (0.6509803921568628, 0.807843137254902, 0.8901960784313725),\n",
       " 'K11073': (0.6980392156862745, 0.8745098039215686, 0.5411764705882353),\n",
       " 'K02182': (0.12156862745098039, 0.47058823529411764, 0.7058823529411765),\n",
       " 'K07149': (0.12156862745098039, 0.47058823529411764, 0.7058823529411765),\n",
       " 'K09472': (0.6509803921568628, 0.807843137254902, 0.8901960784313725),\n",
       " 'K06954': (0.12156862745098039, 0.47058823529411764, 0.7058823529411765),\n",
       " 'K00496': (0.6509803921568628, 0.807843137254902, 0.8901960784313725),\n",
       " 'K14260': (0.6509803921568628, 0.807843137254902, 0.8901960784313725),\n",
       " 'K23464': (0.12156862745098039, 0.47058823529411764, 0.7058823529411765),\n",
       " 'K05367': (0.6509803921568628, 0.807843137254902, 0.8901960784313725),\n",
       " 'K06959': (0.12156862745098039, 0.47058823529411764, 0.7058823529411765),\n",
       " 'K07516': (0.6509803921568628, 0.807843137254902, 0.8901960784313725),\n",
       " 'K01139': (0.6509803921568628, 0.807843137254902, 0.8901960784313725),\n",
       " 'K05349': (0.6509803921568628, 0.807843137254902, 0.8901960784313725),\n",
       " 'K17760': (0.12156862745098039, 0.47058823529411764, 0.7058823529411765),\n",
       " 'K00154': (0.12156862745098039, 0.47058823529411764, 0.7058823529411765),\n",
       " 'K11942': (0.12156862745098039, 0.47058823529411764, 0.7058823529411765),\n",
       " 'K23463': (0.12156862745098039, 0.47058823529411764, 0.7058823529411765),\n",
       " 'K12308': (0.6509803921568628, 0.807843137254902, 0.8901960784313725),\n",
       " 'K02567': (0.6509803921568628, 0.807843137254902, 0.8901960784313725),\n",
       " 'K07716': (0.6980392156862745, 0.8745098039215686, 0.5411764705882353),\n",
       " 'K01637': (0.6509803921568628, 0.807843137254902, 0.8901960784313725),\n",
       " 'K18593': (0.6509803921568628, 0.807843137254902, 0.8901960784313725)}"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lut3"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "K02568    (0.6509803921568628, 0.807843137254902, 0.8901...\n",
       "K07182    (0.12156862745098039, 0.47058823529411764, 0.7...\n",
       "K02305    (0.6509803921568628, 0.807843137254902, 0.8901...\n",
       "K03281    (0.12156862745098039, 0.47058823529411764, 0.7...\n",
       "K01434    (0.6509803921568628, 0.807843137254902, 0.8901...\n",
       "                                ...                        \n",
       "K12308    (0.6509803921568628, 0.807843137254902, 0.8901...\n",
       "K02567    (0.6509803921568628, 0.807843137254902, 0.8901...\n",
       "K07716    (0.6980392156862745, 0.8745098039215686, 0.541...\n",
       "K01637    (0.6509803921568628, 0.807843137254902, 0.8901...\n",
       "K18593    (0.6509803921568628, 0.807843137254902, 0.8901...\n",
       "Length: 101, dtype: object"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "column_colors"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Set palette for color bar"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "my_palette = dict(zip(concat.dataset.unique(), [\"purple\",\"orange\"]))\n",
    "row_colors = concat.dataset.map(my_palette)\n",
    "sampz = concat.Sample"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "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>dataset</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Sample</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>BH091960M</th>\n",
       "      <td>purple</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BH37</th>\n",
       "      <td>purple</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BH091995M</th>\n",
       "      <td>orange</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BH51</th>\n",
       "      <td>orange</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          dataset\n",
       "Sample           \n",
       "BH091960M  purple\n",
       "BH37       purple\n",
       "BH091995M  orange\n",
       "BH51       orange"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sample_colors = sampz.to_frame().join(row_colors)\n",
    "sample_colors = sample_colors.drop_duplicates(subset=['Sample'])\n",
    "sample_colors = sample_colors.set_index('Sample')\n",
    "sample_colors.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Make the clustermap"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x720 with 6 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "p = sns.clustermap(df6_frthrt, cmap='YlGnBu', figsize=(20,10), xticklabels=True, method='ward',\n",
    "                   metric='euclidean', row_colors = sample_colors, col_colors=column_colors)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 6 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "p = sns.clustermap(df6_frthrt, cmap='YlGnBu', method='ward', metric='euclidean', xticklabels=True, row_cluster=False,\n",
    "                   col_cluster=False, yticklabels=False, row_colors = sample_colors, col_colors = column_colors)\n",
    "hm = p.ax_heatmap.get_position()\n",
    "#col = p.ax_col_dendrogram.get_position()\n",
    "#p.ax_heatmap.invert_yaxis()\n",
    "#plt.setp(p.ax_heatmap.yaxis.get_majorticklabels(), fontsize=18)\n",
    "#plt.setp(p.ax_heatmap.xaxis.get_majorticklabels(), fontsize=22)\n",
    "#plt.setp(p.ax_heatmap.yaxis.invert())\n",
    "p.ax_heatmap.set_position([hm.x0, hm.y0, hm.width*3, hm.height])\n",
    "p.ax_col_colors.set_position([col.x0, col.y0-0.04, col.width*3, col.height*0.25])\n",
    "row = p.ax_row_dendrogram.get_position()\n",
    "#p.ax_col_dendrogram.set_position([col.x0, col.y0, col.width*3, col.height*0.5])\n",
    "p.ax_row_dendrogram.set_position([row.x0+0.07, row.y0, row.width*0.75, row.height])\n",
    "p.ax_heatmap.set_xticklabels(p.ax_heatmap.get_xmajorticklabels(), fontsize = 16)\n",
    "p.ax_col_dendrogram.set_visible(False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "p.savefig(\"FigS6B_new2_KOs.png\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Make a separate legend for the KEGG category color bar"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAd0AAAFDCAYAAAB/UdRdAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nO3deVhV1f7H8fcSUJEDMimCIoNS2mCaWHpz+pV5c2wyp9TKtPyladlgaSmaaXYzrcws9Wqaptb9XcuBbLa6mmmFpZaGAyInB1AQHDBg//5AzxUEORhsQj+v5znPw957rb2/e3EePqx9NhtjWRYiIiJS/qpUdAEiIiKXCoWuiIiITRS6IiIiNlHoioiI2EShKyIiYhOFroiIiE1KDF1jzD+NMQeNMVuK2W6MMa8aYxKNMT8ZY64t+zJFREQqP3dmuvOBW86zvRMQc/r1APDGny9LRETk4lNi6FqW9RVw+DxNbgUWWPm+BfyNMaFlVaCIiMjFoiw+060LJJ+1vO/0OhERETmLZxnswxSxrshnSxpjHiD/EjQ+Pj7NGzVqVAaHFxERsc/333+fallWrQvpWxahuw8IP2u5HuAsqqFlWW8BbwHExsZamzZtKoPDi4iI2McYk3Shfcvi8vKHwIDTdzG3BDIsy/q9DPYrIiJyUSlxpmuMeRdoDwQbY/YB4wAvAMuyZgGrgc5AInAcuK+8ihUREanMSgxdy7L6lLDdAoaWWUUiIiIXKT2RSkRExCYKXREREZsodEVERGyi0BUREbGJQldERMQmCl0RERGbKHRFRERsotAVERGxiUJXRETEJgpdERERmyh0RUREbKLQFRERsYlCV0RExCYKXREREZsodEVERGyi0BUREbGJQldERMQmCl0RERGbKHRFRERsotAVERGxiUJXRETEJgpdERERmyh0RUREbKLQFRERsYlCV0RExCYKXREREZsodEVERGyi0BUREbGJQldERMQmCl0RERGbKHRFRERsotAVERGxiUJXRETEJgpdERERmyh0RUREbKLQFRERsYlCV0RExCYKXREREZsodEVERGyi0BUREbGJQldERMQmCl0RERGbKHRFRERsotAVERGxiUJXRETEJgpdERERmyh0RUREbKLQFRERsYlCV0RExCZuha4x5hZjzHZjTKIx5qkittc0xqwwxmw2xmw1xtxX9qWKiIhUbiWGrjHGA3gd6ARcAfQxxlxRqNlQYJtlWdcA7YGpxpiqZVyriIhIpebOTPc6INGyrF2WZZ0ClgC3FmpjAb7GGAM4gMNATplWKiIiUsm5E7p1geSzlvedXne2GUBjwAn8DIywLCuv8I6MMQ8YYzYZYzYdOnToAksWERGpnNwJXVPEOqvQ8t+BBCAMaArMMMb4ndPJst6yLCvWsqzYWrVqlbpYERGRysyd0N0HhJ+1XI/8Ge3Z7gP+z8qXCOwGGpVNiSIiIhcHd0J3IxBjjIk6fXNUb+DDQm32AjcBGGNCgMuBXWVZqIiISGXnWVIDy7JyjDHDgDWAB/BPy7K2GmOGnN4+C3gOmG+M+Zn8y9GjLMtKLce6RUREKp0SQxfAsqzVwOpC62ad9bUT6Fi2pYmIiFxc9EQqERERmyh0RUREbKLQFRERsYlCV0RExCYKXREREZsodEVERGyi0BUREbGJQldERMQmCl0RERGbKHRFRERsotAVERGxiUJXRETEJgpdERERmyh0RUREbKLQFRERsYlCV0RExCYKXREREZsodEVERGyi0BUREbGJQldERMQmCl0RERGbKHRFRERsotAVERGxiUJXRETEJgpdERERmyh0RUREbKLQFRERsYlCV0RExCYKXREREZsodEVERGyi0BUREbGJQldERMQmCl0RERGbKHRFRERsotAVERGxiUJXRETEJgpdERERmyh0RUREbKLQFRERsYlCV0RExCYKXREREZsodEVERGyi0BUREbGJQldERMQmCl0RERGbKHRFRERsotAVERGxiUJXRETEJm6FrjHmFmPMdmNMojHmqWLatDfGJBhjthpj1pZtmSIiIpWfZ0kNjDEewOvAzcA+YKMx5kPLsrad1cYfmAncYlnWXmNM7fIqWEREpLJyZ6Z7HZBoWdYuy7JOAUuAWwu16Qv8n2VZewEsyzpYtmWKiIhUfu6Ebl0g+azlfafXne0yIMAY86Ux5ntjzICyKlBERORiUeLlZcAUsc4qYj/NgZsAb2C9MeZby7J2FNiRMQ8ADwDUr1+/9NVWkJzZb1V0CaXiOfiBii5BRESK4M5Mdx8QftZyPcBZRJuPLMs6ZllWKvAVcE3hHVmW9ZZlWbGWZcXWqlXrQmsWERGplNwJ3Y1AjDEmyhhTFegNfFiozQdAG2OMpzGmBnA98EvZlioiIlK5lXh52bKsHGPMMGAN4AH807KsrcaYIae3z7Is6xdjzEfAT0AeMMeyrC3lWbiIiEhl485nuliWtRpYXWjdrELL/wD+UXaliYiIXFz0RCoRERGbKHRFRERsotAVERGxiUJXRETEJgpdERERmyh0RUREbKLQlVLbs2cPxhhycnIAaN++PXPmzHGrb2naiohcbBS6l7DFixcTGxuLw+EgNDSUTp068c0331R0WaVijMHHxweHw0HdunUZOXIkubm5FV2WiEiRFLqXqJdffplHHnmE0aNHc+DAAfbu3ctDDz3EBx98UNGlFenMrLoomzdvJisri88++4zFixcze/bsUvUXEbGLQvcSlJGRwdixY3n99de544478PHxwcvLi27duvGPf+Q/VCwvL48XXniBBg0aEBQURM+ePTl8+HCJ+46Li6Nfv36u5cKXos+2c+dObrzxRoKCgggODubuu+8mPT3dtT0yMpIpU6bQpEkTfHx8SgzORo0a0aZNG7Zs2eI67ty5c6lfvz433ngjeXl5TJw4kYiICGrXrs2AAQPIyMhw9f/mm2/429/+hr+/P+Hh4cyfPx+A7OxsHn/8cerXr09ISAhDhgzhxIkTAKSmptK1a1f8/f0JDAykTZs25OXlATBlyhTq1q2Lr68vl19+OZ999lmJY3vy5En69etHUFAQ/v7+tGjRggMHDpQ47iJSOSh0L0Hr16/n5MmT3H777cW2efXVV1m+fDlr167F6XQSEBDA0KFDy7QOy7J4+umncTqd/PLLLyQnJxMXF1egzbvvvsuqVatIT0/H0/P8Ty3dtm0bX3/9Nc2aNXOtW7t2Lb/88gtr1qxh/vz5zJ8/ny+++IJdu3aRlZXFsGHDANi7dy+dOnXi4Ycf5tChQyQkJNC0aVMARo0axY4dO0hISCAxMZGUlBQmTJgAwNSpU6lXrx6HDh3iwIEDTJo0CWMM27dvZ8aMGWzcuJHMzEzWrFlDZGQkcP6xffvtt8nIyCA5OZm0tDRmzZqFt7d3WQy3iPwFKHQvQWlpaQQHB583xN58802ef/556tWrR7Vq1YiLi+P9998v08u0DRs25Oabb6ZatWrUqlWLkSNHsnbt2gJthg8fTnh4+HmD59prryUgIIBu3boxaNAg7rvvPte2uLg4fHx88Pb2ZtGiRYwcOZLo6GgcDgeTJ09myZIl5OTksGjRIjp06ECfPn3w8vIiKCiIpk2bYlkWs2fPZtq0aQQGBuLr68vo0aNZsmQJAF5eXvz+++8kJSXh5eVFmzZtMMbg4eFBdnY227Zt448//iAyMpIGDRqUOLZeXl6kpaWRmJiIh4cHzZs3x8/Pr8zGXEQqllv/8EAuLkFBQaSmppKTk1Ns8CYlJXH77bdTpcp/fy/z8PAo00udBw8eZPjw4Xz99ddkZmaSl5dHQEBAgTbh4eHF9P6vH374gYYNGxa57ez+TqeTiIgI13JERAQ5OTkcOHCA5ORkVyie7dChQxw/fpzmzZu71lmW5bpZ64knniAuLo6OHTsC8MADD/DUU0/RsGFDpk+fTlxcHFu3buXvf/87L7/8MmFhYecd2/79+5OcnEzv3r1JT0+nX79+PP/883h5eZU4DiLy16eZ7iWoVatWVK9eneXLlxfbJjw8nPj4eNLT012vkydPUrdu3fPu28fHh+PHj7uW9+/fX2zbp59+GmMMP/30E0ePHuWdd97BsqwCbYwxbp5V0c7ufybwzti7dy+enp6EhIQQHh7Ozp07z+kfHByMt7c3W7dudY1DRkYGWVlZAPj6+jJ16lR27drFihUrePnll12f3fbt25dvvvmGpKQkjDGMGjUKOP/Yenl5MW7cOLZt28a6detYuXIlCxYs+FNjICJ/HQrdS1DNmjWZMGECQ4cOZfny5Rw/fpw//viD+Ph4nnzySQCGDBnCmDFjXCF16NAht+5sbtq0KV999RV79+4lIyODyZMnF9s2MzMTh8OBv78/KSkprpu4ykufPn2YNm0au3fvJisri9GjR9OrVy88PT25++67+fTTT1m2bBk5OTmkpaWRkJBAlSpVGDx4MI8++igHDx4EICUlhTVr1gCwcuVKEhMTsSwLPz8/PDw88PDwYPv27Xz++edkZ2dTvXp1vL298fDwAM4/tl988QU///wzubm5+Pn54eXl5eonIpWfQvcSNXLkSF5++WUmTpxIrVq1CA8PZ8aMGdx2220AjBgxgu7du9OxY0d8fX1p2bIlGzZsKHG/N998M7169aJJkyY0b96crl27Ftt23Lhx/PDDD9SsWZMuXbpwxx13lNn5FWXgwIH079+ftm3bEhUVRfXq1XnttdcAqF+/PqtXr2bq1KkEBgbStGlTNm/eDOTfhdywYUNatmyJn58fHTp0YPv27QD89ttvdOjQAYfDQatWrXjooYdo37492dnZPPXUUwQHB1OnTh0OHjzIpEmTgPOP7f79++nRowd+fn40btyYdu3aFbgbXEQqN1P4cp5dYmNjrU2bNlXIsUsrZ/ZbFV1CqXgOfqCiSxARuWgZY763LCv2QvpqpisiImITha6IiIhNFLoiIiI2UeiKiIjYRKErIiJiE4WuiIiITRS6IiIiNtGzl93QsuprFV1CqWxCf6crIvJXpJmuiIiITRS6IiIiNlHoioiI2EShKyIiYhOFroiIiE0UuiIiIjZR6IqIiNhEoSsiImITha6IiIhNFLrylzBp0iQGDRpUrsc4ceIE3bp1o2bNmtx1113leqwLsWjRIjp27FjRZfxldOrUibfffruiyxApUwrdS9iSJUu4/vrr8fHxoXbt2lx//fXMnDkTy7LK9bhffvkl9erVK7Bu9OjRzJkzp9T7at++vdv93n//fQ4cOEBaWhrvvfdeqY9Vlvbs2YMxhpycHNe6u+++m48//rjMj/Xll19SpUoVHA4Hvr6+XH755cybN6/Mj1PW4uPjueeeeyq6DJEypdC9RE2dOpURI0bwxBNPsH//fg4cOMCsWbP4z3/+w6lTpyq6vHKRlJTEZZddhqdn6R85fnY4VkZhYWFkZWVx9OhRpkyZwuDBg9m2bds57Sr7eYr81Sl0L0EZGRmMHTuWmTNn0qNHD3x9fTHG0KxZMxYtWkS1atUAyM7O5vHHH6d+/fqEhIQwZMgQTpw4Afx3tjp16lRq165NaGhogdlTcX2PHTtGp06dcDqdOBwOHA4HTqeTuLg4+vXr5+r/zTff8Le//Q1/f3/Cw8OZP39+ied1vprGjRvHhAkTWLp0KQ6Hg7lz55KXl8fEiROJiIigdu3aDBgwgIyMDOC/M9G5c+dSv359brzxRubPn88NN9zAo48+ir+/P9HR0axbt4758+cTHh5O7dq1C1wOXbVqFc2aNcPPz4/w8HDi4uJc29q2bQuAv78/DoeD9evXM3/+fFq3bu1qs27dOlq0aEHNmjVp0aIF69atc21r3749zz77LDfccAO+vr507NiR1NTUEsfIGMNtt91GQEAA27ZtK3BOgYGBxMXFkZGRwYABA6hVqxYRERFMnDiRvLw81z5mz55N48aN8fX15YorruCHH34AwOl0cuedd1KrVi2ioqJ49dVXXX2+++47YmNj8fPzIyQkhJEjRwJw8uRJ+vXrR1BQEP7+/rRo0YIDBw64zvHMVYwzY/P4448TEBBAVFQU8fHxrv3v3r2btm3b4uvrS4cOHRg6dGiB95PIX4VC9xK0fv16srOzufXWW8/bbtSoUezYsYOEhAQSExNJSUlhwoQJru379+8nIyODlJQU5s6dy9ChQzly5Mh5+/r4+BAfH++aeWVlZREWFlbguHv37qVTp048/PDDHDp0iISEBJo2berWuRVX0/jx4xk9ejS9evUiKyuL+++/n/nz5zN//ny++OILdu3aRVZWFsOGDSuwv7Vr1/LLL7+wZs0aADZs2ECTJk1IS0ujb9++9O7dm40bN5KYmMg777zDsGHDyMrKAsDHx4cFCxaQnp7OqlWreOONN1i+fDkAX331FQDp6elkZWXRqlWrAsc9fPgwXbp0Yfjw4aSlpTFy5Ei6dOlCWlqaq83ixYuZN28eBw8e5NSpU7z00ksljk9eXh7//ve/SU9P5+qrr3adU3R0NAcPHmTMmDE8/PDDZGRksGvXLtauXcuCBQtcv7y89957xMXFsWDBAo4ePcqHH35IUFAQeXl5dOvWjWuuuYaUlBQ+++wzpk+f7hq3ESNGMGLECI4ePcrOnTvp2bMnAG+//TYZGRkkJyeTlpbGrFmz8Pb2LrL2DRs2cPnll5OamsqTTz7J/fff7/oopG/fvlx33XWkpaURFxfHwoULSxwLkYqg0L0EpaamEhwcXOAy65lZpbe3N1999RWWZTF79mymTZtGYGAgvr6+jB49miVLlrj6eHl5MXbsWLy8vOjcuTMOh4Pt27e71fd8Fi1aRIcOHejTpw9eXl4EBQW5HbrF1VTccUaOHEl0dDQOh4PJkyezZMmSApdY4+Li8PHxcQVBVFQU9913Hx4eHvTq1Yvk5GTGjh1LtWrV6NixI1WrViUxMRHIn6ldffXVVKlShSZNmtCnTx/Wrl3r1nmsWrWKmJgY+vfvj6enJ3369KFRo0asWLHC1ea+++7jsssuw9vbm549e5KQkFDs/pxOJ/7+/gQHBzN+/HgWLlzI5ZdfDuRfen744Yfx9PSkatWqLF26lMmTJ+Pr60tkZCSPPfaYK8TmzJnDk08+SYsWLTDG0LBhQyIiIti4cSOHDh1i7NixVK1alejoaAYPHuz6nnt5eZGYmEhqaioOh4OWLVu61qelpZGYmIiHhwfNmzfHz8+vyHOIiIhg8ODBeHh4cM899/D7779z4MAB9u7dy8aNG5kwYQJVq1aldevWdO/e3a1xFrGb/p/uJSgoKIjU1FRycnJcwXvm0mW9evXIy8vj0KFDHD9+nObNm7v6WZZFbm5ugf2cHdw1atQgKyvLrb7nk5ycTIMGDS743IqqqShOp5OIiAjXckREBDk5Oa7LmwDh4eEF+oSEhLi+PhPEhdedOd6GDRt46qmn2LJlC6dOnSI7O9vtu6YL13amvpSUFNdynTp13DpPyA/Wffv2Fbnt7HNMTU3l1KlT54zLmeMW971JSkpyBfsZubm5tGnTBoC5c+cyduxYGjVqRFRUFOPGjaNr167079+f5ORkevfuTXp6Ov369eP555/Hy8vrnGMUPl+ArKwsUlNTCQwMdK07c07JycnFjodIRdFM9xLUqlUrqlWrxgcffFBsm+DgYLy9vdm6dSvp6emkp6eTkZFx3h/s7vY1xpy3f3h4ODt37izdSV2AsLAwkpKSXMt79+7F09OzQIiWVOv59O3bl+7du5OcnExGRgZDhgxxXQ4tab+FaztTX926dS+4nuKcXUtwcDBeXl7njMuZ4xb3vQkPDycqKsr1/U5PTyczM5PVq1cDEBMTw7vvvsvBgwcZNWoUPXr04NixY3h5eTFu3Di2bdvGunXrWLlyJQsWLChV/aGhoRw+fJjjx4+71ilw5a9KoXsJ8vf3Z9y4cTz00EO8//77ZGVlkZeXR0JCAseOHQOgSpUqDB48mEcffZSDBw8CkJKS4vqM7nxK6hsSEkJaWprrpqXC7r77bj799FOWLVtGTk4OaWlp5710eqH69OnDtGnT2L17N1lZWa7PfC/k7uaiZGZmEhgYSPXq1fnuu+9YvHixa1utWrWoUqUKu3btKrJv586d2bFjB4sXLyYnJ4elS5eybds2unbtWia1FcfDw4OePXsyZswYMjMzSUpK4uWXX3bdlDRo0CBeeuklvv/+eyzLIjExkaSkJK677jr8/PyYMmUKJ06cIDc3ly1btrBx40YA3nnnHQ4dOkSVKlVcs2EPDw+++OILfv75Z3Jzc/Hz88PLywsPD49S1RwREUFsbCxxcXGcOnWK9evXF7gML/JXotC9RD355JO8/PLLvPjii9SuXZuQkBAefPBBpkyZwt/+9jcApkyZQsOGDWnZsiV+fn506NCh2M9HCztf30aNGtGnTx+io6Px9/fH6XQW6Fu/fn1Wr17N1KlTCQwMpGnTpmzevLlsBwAYOHAg/fv3p23btkRFRVG9enVee+21Mtv/zJkzGTt2LL6+vkyYMMF18xDkXx4dM2YMN9xwA/7+/nz77bcF+gYFBbFy5UqmTp1KUFAQL774IitXriQ4OLjM6ivOa6+9ho+PD9HR0bRu3Zq+ffsycOBAAO666y7GjBlD37598fX15bbbbuPw4cN4eHiwYsUKEhISiIqKIjg4mEGDBrl+sfroo4+48sorcTgcjBgxgiVLllC9enX2799Pjx498PPzo3HjxrRr1+6C7jpetGgR69evJygoiGeeeYZevXq57sIX+Ssx5f0ghOLExsZamzZtqpBjl1bs21dXdAmlsumenyu6BJEK1atXLxo1asT48eMruhS5CBljvrcsK/ZC+mqmKyKV3saNG9m5cyd5eXl89NFHfPDBB9x2220VXZbIOXT3sohUevv37+eOO+4gLS2NevXq8cYbb9CsWbOKLkvkHApdEan0unXrRrdu3Sq6DJES6fKyiIiITdwKXWPMLcaY7caYRGPMU+dp18IYk2uM6VF2JYqIiFwcSgxdY4wH8DrQCbgC6GOMuaKYdlOAkv+QU0RE5BLkzkz3OiDRsqxdlmWdApYART0p/2HgX8DBMqxPRETkouFO6NYFzn6m2r7T61yMMXWB24FZZVeaiIjIxcWdu5eLekhs4SdqTAdGWZaVe75nyhpjHgAegPynDlUW49oNqOgSRETkIuBO6O4Dzv5XK/UAZ6E2scCS04EbDHQ2xuRYlrX87EaWZb0FvAX5T6S60KJFREQqI3dCdyMQY4yJAlKA3kDfsxtYlhV15mtjzHxgZeHAFRERudSVGLqWZeUYY4aRf1eyB/BPy7K2GmOGnN6uz3FFRETc4NYTqSzLWg2sLrSuyLC1LOveP1+WiIjIxUdPpBIREbGJQldERMQmCl0RERGbKHRFRERsotAVERGxiUJXytSQIUN47rnnKroM2917770888wzZbKvN954g5CQEBwOB2lpaWWyz7LkcDjYtWtXRZfxl7Bo0SI6duxY0WVIJaLQvURFRkbi7e2Nw+FwvYYNG/an9ztr1iyeffbZMqiwYkVGRvLpp5+Wyb7mz59P69at3Wr7xx9/MHLkSD7++GOysrIICgoqkxouVPv27ZkzZ06BdVlZWURHR5f5sc5+T4aEhHDfffeRlZVV5scpS3fffTcff/xxRZchlYhC9xK2YsUKsrKyXK8ZM2aU6/FycnLKdf8XgwMHDnDy5EmuvPLKUve1LIu8vLxyqMo+Z96TP/zwAxs3bmTixInntNH7SCozha6c48zM7PHHHycgIICoqCji4+MBWLJkCbGxsQXaT5s2je7duwMFL7N++eWX1KtXjylTplCnTh3uu+8+srOzeeSRRwgLCyMsLIxHHnmE7OzsAu2nTp1K7dq1CQ0NZd68ea7j3HvvvTz00EN06tQJh8PBDTfcwP79+3nkkUcICAigUaNG/Pjjj672TqeTO++8k1q1ahEVFcWrr77q2hYXF0fPnj0ZMGAAvr6+XHnllWzatAmA/v37s3fvXrp164bD4eDFF18E4K677qJOnTrUrFmTtm3bsnXr1gsa38jISF566SWaNGlCzZo16dWrFydPnmTHjh1cfvnlAPj7+3PjjTcCsG7dOlq0aEHNmjVp0aIF69atc+2rffv2jBkzhhtuuIEaNWqwa9cujDHMnDmTmJgYfH19efbZZ9m5cyetWrXCz8+Pnj17curUKQCOHDlC165dqVWrFgEBAXTt2pV9+/YBMGbMGL7++muGDRtW4EqIMYbExEQAMjIyGDBgALVq1SIiIoKJEye6gv9876OS1K1bl06dOrFlyxbXMV9//XViYmKIiYkBYPbs2TRs2JDAwEC6d++O0/nfR8Jv3bqVm2++mcDAQEJCQpg0aRIAeXl5vPDCCzRo0ICgoCB69uzJ4cOHATh58iT9+vUjKCgIf39/WrRowYEDB1znEh0dja+vL1FRUSxatKjAOZ5hjGHWrFnExMQQEBDA0KFDsaz8x8zn5uby2GOPERwcTFRUFDNmzMAYo18iLjEKXSnShg0buPzyy0lNTeXJJ5/k/vvvx7Isunfvzvbt2/ntt99cbRcvXkzfvn2L3M/+/fs5fPgwSUlJvPXWWzz//PN8++23JCQksHnzZr777rsCs5n9+/eTkZFBSkoKc+fOZejQoRw5csS1fdmyZUycOJHU1FSqVatGq1atuPbaa0lNTaVHjx6MHDkSyP/h2q1bN6655hpSUlL47LPPmD59OmvWrHHt68MPP6R3796kp6fTvXt3V6gsXLiQ+vXru2ZdTz75JACdOnXit99+4+DBg1x77bXcfffdFzy+y5Yt46OPPmL37t389NNPzJ8/n8suu8wV5Onp6Xz++eccPnyYLl26MHz4cNLS0hg5ciRdunQp8FnvwoULeeutt8jMzCQiIgKAjz76iO+//55vv/2WF198kQceeIBFixaRnJzMli1bePfdd13jdN9995GUlMTevXvx9vZ2jcPzzz9PmzZtmDFjRrFXQh5++GEyMjLYtWsXa9euZcGCBQV+USrufVSS5ORkVq9eTbNmzVzrli9fzoYNG9i2bRuff/45Tz/9NMuWLeP3338nIiKC3r17A5CZmUmHDh245WzS5t4AAB4cSURBVJZbcDqdJCYmctNNNwHw6quvsnz5ctauXYvT6XQFI8Dbb79NRkYGycnJpKWlMWvWLLy9vTl27BjDhw8nPj6ezMxM1q1bR9OmTYutfeXKlWzcuJHNmzezbNky13tu9uzZxMfHk5CQwA8//MDy5Xo8/aVIoXsJu+222/D393e9Zs+e7doWERHB4MGD8fDw4J577uH333/nwIED1KhRg1tvvdX1Q/u3337j119/dc10C6tSpQrjx4+nWrVqeHt7s2jRIsaOHUvt2rWpVasW48aNY+HCha72Xl5ejB07Fi8vLzp37ozD4WD79u2u7bfffjvNmzenevXq3H777VSvXp0BAwbg4eFBr169XDPdjRs3cujQIcaOHUvVqlWJjo5m8ODBLFmyxLWv1q1b07lzZzw8POjfvz+bN28+73gNHDgQX19fqlWrRlxcHJs3byYjI6P0Aw8MHz6csLAwAgMD6datGwkJCUW2W7VqFTExMfTv3x9PT0/69OlDo0aNWLFihavNvffey5VXXomnpydeXl4AjBo1Cj8/P6688kquuuoqOnbsSHR0NDVr1qRTp06ucQoKCuLOO++kRo0a+Pr6MmbMGNauXevWOeTm5rJ06VImT56Mr68vkZGRPPbYYwW+n8W9j4pz5j3ZunVr2rVrx+jRo13bnn76aQIDA13vo4EDB3LttddSrVo1Jk+ezPr169mzZw8rV66kTp06PPbYY1SvXh1fX1+uv/56AN58802ef/556tWr5/o+vv/+++Tk5ODl5UVaWhqJiYl4eHjQvHlz/Pz8gPz38ZYtWzhx4gShoaHnvfz/1FNP4e/vT/369fmf//kf1/d22bJljBgxgnr16hEQEMBTTz3l1jjLxUWhewlbvnw56enprtfgwYNd2+rUqeP6ukaNGgCum1r69u3rCt3Fixdz2223udoUVqtWLapXr+5adjqdrtkY5P9QPvuyYFBQEJ6e/30keI0aNQrcTBMSEuL62tvb+5zlM22TkpJwOp0FfqmYNGlSgR/4hc/x5MmTxV7qy83N5amnnqJBgwb4+fkRGRkJQGpqapHtS1L42MXdMFR4vCB/zFJSUlzL4eHhhbu5PU7Hjx/nwQcfJCIiAj8/P9q2bUt6ejq5ubklnkNqaiqnTp065/t5dm3nex8V5cx7MikpiZkzZ+Lt7V3keRYeF4fDQVBQECkpKSQnJ9OgQYMi95+UlMTtt9/uek80btwYDw8PDhw4QP/+/fn73/9O7969CQsL48knn+SPP/7Ax8eHpUuXMmvWLEJDQ+nSpQu//vprsedQ3PfW6XQWOIeivm9y8VPoSql17NiR1NRUEhISePfdd4u9tAz5n3GdLSwsjKSkJNfy3r17CQsLK/Maw8PDiYqKKvBLRWZmJqtXry65M+fWvXjxYj744AM+/fRTMjIy2LNnD4Bbl0r/jMLjBfljVrdu3WJrLY2pU6eyfft2NmzYwNGjR/nqq6+A/57X+fYdHByMl5fXOd/Ps2srS2fXUnhcjh07RlpaGnXr1iU8PJydO3cWuY/w8HDi4+MLvC9OnjxJ3bp18fLyYty4cWzbto1169axcuVKFixYAMDf//53PvnkE37//XcaNWpU4BdUd4WGhro+L4f8S+hy6VHoSql5enrSo0cPnnjiCQ4fPszNN9/sdt8+ffowceJEDh06RGpqKhMmTKBfv35lXuN1112Hn58fU6ZM4cSJE+Tm5rJlyxY2btzoVv+QkJACf4uamZlJtWrVCAoK4vjx4wUue5anzp07s2PHDhYvXkxOTg5Lly5l27ZtdO3atUz2n5mZibe3N/7+/hw+fJjx48cX2F54HM7m4eFBz549GTNmDJmZmSQlJfHyyy+Xy/ezsL59+zJv3jwSEhLIzs5m9OjRXH/99URGRtK1a1f279/P9OnTyc7OJjMzkw0bNgD5f0c+ZswYV2AfOnSIDz74AIAvvviCn3/+mdzcXPz8/PDy8nLNgj/88EOOHTtGtWrVcDgceHh4lLrmnj178sorr5CSkkJ6ejpTpkwpuwGRSkOhewk7c3fumdftt9/udt++ffvy6aefctdddxW4HFySZ555htjYWJo0acLVV1/NtddeW2YPlTibh4cHK1asICEhgaioKIKDgxk0aJDbn8E+/fTTTJw4EX9/f1566SUGDBhAREQEdevW5YorrqBly5ZlXnNRgoKCWLlyJVOnTiUoKIgXX3yRlStXEhwcXCb7f+SRRzhx4gTBwcG0bNmSW265pcD2ESNG8P777xMQEMDw4cPP6f/aa6/h4+NDdHQ0rVu3pm/fvgwcOLBMajufm266ieeee44777yT0NBQdu7c6fq83tfXl08++YQVK1ZQp04dYmJi+OKLL1zn0717dzp27Iivry8tW7Z0BfL+/fvp0aMHfn5+NG7cmHbt2tGvXz/y8vKYOnWq6zP4tWvXMnPmzFLXPHjwYDp27EiTJk1o1qwZnTt3xtPT84ICXCovU96Xx4oTGxtrnfkTjb+6FXv+UdEllEq3yCcqugQRKUF8fDxDhgw55+MD+eszxnxvWVZsyS3PpZmuiIgNTpw4werVq8nJySElJYXx48eX6uqSXBwUuiIiNrAsi3HjxhEQEECzZs1o3LgxEyZMqOiyxGbufxgnIiIXrEaNGm7fyCcXL810RUREbKLQFRERsYlCV0RExCYKXREREZvoRio3PPzmFRVdQql0m1zRFYiISFE00xUREbGJQldERMQmCl0RERGbKHRFRERsotAVERGxiUJXRETEJgpdERERmyh0RUREbKLQFRERsYlCV0RExCYKXflLi4uLo1+/fhfU98svv6RevXrl0tfhcLBr164L2ndZ+TPndzG58sor+fLLLyu6DBG3KHQvUZGRkYSEhHDs2DHXujlz5tC+fXu3+rdv3545c+YUu33Pnj0YY8jJyfmzpf4lZWVlER0dXep+Z8bF4XDgcDiIjIzkhRdecKuvMYbExMRSH7OyKDw2ISEhPPTQQ/zxxx/n7bd161bX+/bP/JImYgeF7iUsJyeHV155paLLuCSlp6eTlZXFu+++y4QJE/joo48quiRbne+XsTNj8/PPP7N+/Xpef/31Uu9D5K9KoXsJe+KJJ3jppZdIT08vcvu6deto0aIFNWvWpEWLFqxbtw6AMWPG8PXXXzNs2DAcDgfDhg0r8Vj33nsvQ4cOpUuXLvj6+nL99dezc+dO1/atW7dy8803ExgYSEhICJMmTTpnH0VdTo2MjOTTTz8F4MSJE9x7770EBARwxRVXsHHjxgJtnU4nd955J7Vq1SIqKopXX33Vta2kvoWdPess6dzOp1WrVlx55ZVs2bKF7777jlatWuHv709oaCjDhg3j1KlTALRt2xaAa665BofDwdKlS137mDp1KrVr1yY0NJR58+YBsHv3bvz9/cnLywNg0KBB1K5d29WnX79+TJ8+HYB58+bRuHFjfH19iY6O5s0333S1u+qqq1ixYoVr+Y8//iA4OJiEhIQiz2f27Nk0bNiQwMBAunfvjtPpLDBmr7/+OjExMcTExJQ4NrVr1+bmm29m27ZtrnWRkZFMmTKFJk2a4OPjQ05Ojus98NFHHzFp0iSWLl2Kw+HgmmuuASAjI4P777+f0NBQ6tatyzPPPENubm6JxxcpDwrdS1hsbCzt27fnpZdeOmfb4cOH6dKlC8OHDyctLY2RI0fSpUsX0tLSeP7552nTpg0zZswgKyuLGTNmuHW8d999l3HjxnHkyBEaNmzImDFjAMjMzKRDhw7ccsstOJ1OEhMTuemmm0p9PuPHj2fnzp3s3LmTNWvW8Pbbb7u25eXl0a1bN6655hpSUlL47LPPmD59OmvWrCmx7585t/OxLIv//Oc/bN26lWbNmuHh4cG0adNITU1l/fr1fPbZZ8ycOROAr776CoDNmzeTlZVFr169ANi/fz8ZGRmkpKQwd+5chg4dypEjR4iKisLPz48ff/wRgK+//hqHw8Evv/zi2l+7du2A/HBbuXIlR48eZd68eTz66KP88MMPAAwYMIB33nnHVfPq1asJDQ2ladOm55zP559/ztNPP82yZcv4/fffiYiIoHfv3gXaLF++nA0bNhQI0uI4nU7WrFlDy5YtzxnrVatWkZ6ejqfnf/876S233MLo0aPp1asXWVlZbN68GYB77rkHT09PEhMT+fHHH/n444/P+9GISHlS6F7iJkyYwGuvvcahQ4cKrF+1ahUxMTH0798fT09P+vTpQ6NGjQrMekrrjjvu4LrrrsPT05O7777bNVtauXIlderU4bHHHqN69equ2WJpLVu2jDFjxhAYGEh4eDjDhw93bdu4cSOHDh1i7NixVK1alejoaAYPHsySJUtK7Ptnzq04wcHBBAYGMmjQIF544QVuuukmmjdvTsuWLfH09CQyMpIHH3yQtWvXnnc/Xl5ejB07Fi8vLzp37ozD4WD79u0AtGvXjrVr17J//34AevTowdq1a9m9ezdHjx51zQS7dOlCgwYNMMbQrl07OnbsyNdffw3kz4hXr17N0aNHAVi4cCH9+/cvspZFixYxcOBArr32WqpVq8bkyZNZv349e/bscbV5+umnCQwMxNvb+7xj4+/vT926dfHx8aFHjx4Ftg8fPpzw8PDz7uOMAwcOEB8fz/Tp0/Hx8aF27do8+uijru+7iN30T+wvcVdddRVdu3blhRdeoHHjxq71TqeTiIiIAm0jIiJISUm54GPVqVPH9XWNGjXIysoCIDk5mQYNGlzwfs9wOp2Eh4e7ls+uPykpCafTib+/v2tdbm4ubdq0KbGvO4o7t+KkpqYWmKUB7Nixg5EjR7Jp0yaOHz9OTk4OzZs3P+9+goKCCuzn7GO3a9eODz/8kHr16tG2bVvat2/PwoULqV69Om3atKFKlfzfuePj4xk/fjw7duwgLy+P48ePc/XVVwMQFhbGDTfcwL/+9S9uv/124uPji70PwOl0cu2117qWHQ4HQUFBpKSkEBkZCVBgjEsamxMnTjB27FhuueUW10cb7u7jjKSkJP744w9CQ0Nd6/Ly8kq1D5GypJmuMH78eGbPnl0gUMPCwkhKSirQbu/evdStWxfI/3yurISHh7v1GaiPjw/Hjx93Lefm5haYoYeGhpKcnFyg3rOPERUVRXp6uuuVmZnJ6tWrS+xrl//93/+lUaNG/Pbbbxw9epRJkyZhWdYF769du3Z8/fXXfPnll7Rr147WrVvzn//8h7Vr17ouLWdnZ3PnnXfy+OOPc+DAAdLT0+ncuXOB495zzz288847vPfee7Rq1cr1Hiis8Hvm2LFjpKWlFWhfmveNt7c39957L+vXryc1NdWtfRTeFh4eTrVq1UhNTXV9348ePcrWrVvdrkOkLCl0hYYNG9KrV68CNxZ17tyZHTt2sHjxYnJycli6dCnbtm2ja9euAISEhJTZ36l27dqV/fv3M336dLKzs8nMzGTDhg3ntLvssss4efIkq1at4o8//mDixIlkZ2e7tvfs2ZPJkydz5MgR9u3bx2uvvebadt111+Hn58eUKVM4ceIEubm5bNmyxXXD1Pn62iUzMxM/Pz8cDge//vorb7zxRoHtpR3zmJgYvL29eeedd2jbti1+fn6EhITwr3/9yxW6p06dIjs7m1q1auHp6Ul8fDwff/xxgf3cdttt/PDDD7zyyisMGDCg2OP17duXefPmkZCQQHZ2NqNHj+b66693zXJLKzs7m4ULF1KnTh2CgoLc6hMSEsKePXtcN5CFhobSsWNHHnvsMY4ePUpeXh47d+4s8bK9SHlR6AoAY8eOLfA3u0FBQaxcuZKpU6cSFBTEiy++yMqVKwkODgZgxIgRvP/++wQEBJT688/CfH19+eSTT1ixYgV16tQhJiaGL7744px2NWvWZObMmQwaNMj1ed/ZdzOPGzeOiIgIoqKi6NixY4HPHj08PFixYgUJCQlERUURHBzMoEGDyMjIKLGvXV566SUWL16Mr68vgwcPdt0sdUZcXBz33HMP/v7+LFu2zK19tmvXjqCgIOrXr+9atiyLZs2aAflj/+qrr9KzZ08CAgJYvHgx3bt3L7APb29v7rzzTnbv3s0dd9xR7LFuuukmnnvuOe68805CQ0PZuXPnBX126u/v7/o73fXr1/Phhx+6PUO+6667gPz375lL3QsWLODUqVNcccUVBAQE0KNHD37//fdS1yVSFsyfuXz1Z8TGxlqbNm2qkGOXVuTTqyq6hFLZM7lLRZcgF5kJEyawY8eOAncyi1yqjDHfW5YVeyF9dSOViJzX4cOHmTt3LgsXLqzoUkQqPV1eFpFizZ49m/DwcDp16uR6QIeIXDjNdEWkWIMHD2bw4MEVXYbIRUMzXREREZu4FbrGmFuMMduNMYnGmKeK2H63Mean0691xphryr5UERGRyq3E0DXGeACvA52AK4A+xpgrCjXbDbSzLKsJ8BzwVlkXKiIiUtm5M9O9Dki0LGuXZVmngCXArWc3sCxrnWVZR04vfgvoP2uLiIgU4s6NVHWB5LOW9wHnexr9/UB8URuMMQ8ADwCuP9avDCb31NVyERH589yZ6Rb1KJgin6hhjPkf8kN3VFHbLct6y7KsWMuyYmvVquV+lSIiIhcBd2a6+4Cz/yVHPcBZuJExpgkwB+hkWVZa2ZQnIiJy8XBnprsRiDHGRBljqgK9gQ/PbmCMqQ/8H9DfsqwdZV+miIhI5VfiTNeyrBxjzDBgDeAB/NOyrK3GmCGnt88CxgJBwMzTDybPudDnUoqIiFys3HoilWVZq4HVhdbNOuvrQcCgsi1NRETk4qInUomIiNhEoSsiImITha6IiIhNFLoiIiI2UeiKiIjYRKErtjHGkJiYWOp+e/bswRhDTk4OAJ06deLtt98u6/JERMqdQvcSFRkZSdWqVUlNTS2wvmnTphhj2LNnz3n7f/nll9SrVzH/1yI+Pp577rmnQo4tIvJnKHQvYVFRUbz77ruu5Z9//pkTJ05UYEUiIhc3he4lrH///ixYsMC1/PbbbzNgwADXcnZ2No8//jj169cnJCSEIUOGcOLECY4dO0anTp1wOp04HA4cDgdOp5PvvvuOVq1a4e/vT2hoKMOGDePUqVMFjrl69Wqio6MJDg7miSeeIC8vD4C8vDwmTpxIREQEtWvXZsCAAWRkZBRZd/v27ZkzZw4AiYmJtGvXjpo1axIcHEyvXr1c7YwxzJw5k5iYGHx9fXn22WfZuXMnrVq1ws/Pj549e55Tn4hIeVLoXsJatmzJ0aNH+eWXX8jNzWXp0qX069fPtX3UqFHs2LGDhIQEEhMTSUlJYcKECfj4+BAfH09YWBhZWVlkZWURFhaGh4cH06ZNIzU1lfXr1/PZZ58xc+bMAsf897//zaZNm/jhhx/44IMP+Oc//wnA/PnzmT9/Pl988QW7du0iKyuLYcOGlXgOzz77LB07duTIkSPs27ePhx9+uMD2jz76iO+//55vv/2WF198kQceeIBFixaRnJzMli1bCsz0RUTKm0L3EndmtvvJJ5/QqFEj6tatC4BlWcyePZtp06YRGBiIr68vo0ePZsmSJcXuq3nz5rRs2RJPT08iIyN58MEHWbt2bYE2o0aNIjAwkPr16/PII4+4Qm/RokWMHDmS6OhoHA4HkydPZsmSJa6bp4rj5eVFUlISTqeT6tWr07p163OO5+fnx5VXXslVV11Fx44diY6OpmbNmnTq1Ikff/zxQoZNROSCuPXsZbl49e/fn7Zt27J79+4Cl5YPHTrE8ePHad68uWudZVnk5uYWu68dO3YwcuRINm3axPHjx8nJySnQHyA8/L//JTIiIgKnM/+/RDqdTiIiIgpsy8nJ4cCBA+et/8UXX+TZZ5/luuuuIyAggMcee4yBAwe6toeEhLi+9vb2Pmd5//79592/iEhZ0kz3EhcREUFUVBSrV6/mjjvucK0PDg7G29ubrVu3kp6eTnp6OhkZGWRlZQH5n5cW9r//+780atSI3377jaNHjzJp0iQsyyrQJjk52fX13r17CQsLAyAsLIykpKQC2zw9PQuEZFHq1KnD7NmzcTqdvPnmmzz00EMX9GdJIiJ2UOgKc+fO5fPPP8fHx8e1rkqVKgwePJhHH32UgwcPApCSksKaNWuA/BlkWlpagZudMjMz8fPzw+Fw8Ouvv/LGG2+cc6x//OMfHDlyhOTkZF555RXXjU99+vRh2rRp7N69m6ysLEaPHk2vXr3w9Dz/xZj33nuPffv2ARAQEIAxBg8Pjz83ICIi5UShKzRo0IDY2HP//fGUKVNo2LAhLVu2xM/Pjw4dOrB9+3YAGjVqRJ8+fYiOjsbf3x+n08lLL73E4sWL8fX1ZfDgwQXuJD7j1ltvpXnz5jRt2pQuXbpw//33AzBw4EDXpe6oqCiqV6/Oa6+9VmLtGzdu5Prrr8fhcNC9e3deeeUVoqKi/uSIiIiUD1P48p9dYmNjrU2bNlXIsUvr3R/3VXQJpdKnWcU8tEJE5FJgjPnesqxzZypu0ExXRETEJgpdERERmyh0RUREbKLQFRERsYlCV0RExCYKXREREZsodEVERGyiZy+7QX/3KiIiZUEzXREREZsodEVERGyi0BUREbGJQldERMQmCl0RERGbKHRFRERsotAVERGxiUJXRETEJgpdERERmyh0RUREbKLQFRERsYlCV0RExCYKXREREZsodEVERGyi0BUREbGJQldERMQmCl0RERGbKHRFRERsotAVERGxiUJXRETEJgpdERERmyh0RUREbKLQFRERsYlCV0RExCZuha4x5hZjzHZjTKIx5qkithtjzKunt/9kjLm27EsVERGp3EoMXWOMB/A60Am4AuhjjLmiULNOQMzp1wPAG2Vcp4iISKXnzkz3OiDRsqxdlmWdApYAtxZqcyuwwMr3LeBvjAkt41pFREQqNXdCty6QfNbyvtPrSttGRETkkubpRhtTxDrrAtpgjHmA/MvPANnGmC1uHF9KLxhIregiLkIa1/KhcS0/GtvycfmFdnQndPcB4Wct1wOcF9AGy7LeAt4CMMZssiwrtlTVils0tuVD41o+NK7lR2NbPowxmy60rzuXlzcCMcaYKGNMVaA38GGhNh8CA07fxdwSyLAs6/cLLUpERORiVOJM17KsHGPMMGAN4AH807KsrcaYIae3zwJWA52BROA4cF/5lSwiIlI5uXN5GcuyVpMfrGevm3XW1xYwtJTHfquU7cV9GtvyoXEtHxrX8qOxLR8XPK4mPy9FRESkvOkxkCIiIjYp99DVIyTLhxvjevfp8fzJGLPOGHNNRdRZGZU0tme1a2GMyTXG9LCzvsrKnXE1xrQ3xiQYY7YaY9baXWNl5MbPgprGmBXGmM2nx1X33LjBGPNPY8zB4v609YKzy7KscnuRf+PVTiAaqApsBq4o1KYzEE/+3/q2BDaUZ00Xw8vNcf0bEHD6604a17Ib27PafU7+vQ49Krruv/rLzfesP7ANqH96uXZF1/1Xf7k5rqOBKae/rgUcBqpWdO1/9RfQFrgW2FLM9gvKrvKe6eoRkuWjxHG1LGudZVlHTi9+S/7fTkvJ3HnPAjwM/As4aGdxlZg749oX+D/LsvYCWJalsS2ZO+NqAb7GGAM4yA/dHHvLrHwsy/qK/LEqzgVlV3mHrh4hWT5KO2b3k/8bmZSsxLE1xtQFbgdmIe5y5z17GRBgjPnSGPO9MWaAbdVVXu6M6wygMfkPLPoZGGFZVp495V3ULii73PqToT+hzB4hKQW4PWbGmP8hP3Rbl2tFFw93xnY6MMqyrNz8yYO4wZ1x9QSaAzcB3sB6Y8y3lmXtKO/iKjF3xvXvQAJwI9AA+MQY87VlWUfLu7iL3AVlV3mHbpk9QlIKcGvMjDFNgDlAJ8uy0myqrbJzZ2xjgSWnAzcY6GyMybEsa7k9JVZK7v4sSLUs6xhwzBjzFXANoNAtnjvjeh/wgpX/QWSiMWY30Aj4zp4SL1oXlF3lfXlZj5AsHyWOqzGmPvB/QH/NFEqlxLG1LCvKsqxIy7IigfeBhxS4JXLnZ8EHQBtjjKcxpgZwPfCLzXVWNu6M617yrx5gjAkh/2H9u2yt8uJ0QdlVrjNdS4+QLBdujutYIAiYeXpGlmPpweclcnNspZTcGVfLsn4xxnwE/ATkAXMsy9J/IjsPN9+vzwHzjTE/k39JdJRlWfrPQyUwxrwLtAeCjTH7gHGAF/y57NITqURERGyiJ1KJiIjYRKErIiJiE4WuiIiITRS6IiIiNlHoioiI2EShKyIiYhOFroiIiE0UuiIiIjb5f8G2Jq7UExNzAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "x, y = (0.1, 0)\n",
    "fig = plt.figure()\n",
    "ax = fig.add_axes([0,0,1,1])\n",
    "for func in df6_func['Group'].unique():\n",
    "    y = y + 0.1\n",
    "    color = lut[func]\n",
    "    square = plt.Rectangle((x, y), 0.1, 0.1, fc=color)\n",
    "    plt.gca().add_patch(square)\n",
    "    plt.text(x + 0.12, y + 0.03, func, fontsize=12)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "fig.savefig('FigS6B_legend.png')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}