orthoplot.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "46e73412-61fb-4e6b-9fa0-89fed2186600",
   "metadata": {},
   "source": [
    "# 0. Preparing the environment\n",
    "\n",
    "## This will only load the required libraries, in R, the equivalent is to call `library(\"tidiverse\")`\n",
    "\n",
    "## We will also declare a function we'll use later"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "166c5c47-e542-4f0f-8192-7aa98d4cc24c",
   "metadata": {},
   "outputs": [],
   "source": [
    "import ipywidgets as     widgets\n",
    "import numpy      as     np\n",
    "import pandas     as     pd\n",
    "import seaborn    as     sns\n",
    "from   matplotlib import pyplot as plt\n",
    "from   pyDOE2     import ff2n\n",
    "from   upsetplot  import UpSet,plot"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "08f2ae10-9856-4154-b7ba-ac3fa8a23793",
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_og_class(og_tuple):\n",
    "    og_class = binary_dict[og_tuple]\n",
    "    return og_class"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "28275dbf-0986-44e0-b2d2-394147ad7ec8",
   "metadata": {},
   "source": [
    "# 1. Preparing the files"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ede47cfb-1e31-4a35-b1d5-e06504a9cf48",
   "metadata": {},
   "source": [
    "## Two files are required for the analysis:\n",
    "\n",
    "- ### `Orthogroups.GeneCount.tsv` that, as the name implies, contains the number of genes per orthogroup per species\n",
    "- ### `Orthogroups.GeneCount.tsv` that, as the name implies, contains the number of exclusive genes per species\n",
    "\n",
    "### We will load the files into separate pandas dataframes.\n",
    "### In addition, from the dataframe header we will extract a list of the species"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "96c51f16-3da2-42b0-973c-a5dfb2a938d8",
   "metadata": {},
   "outputs": [],
   "source": [
    "assigned_df   = pd.read_csv(\"Orthogroups.GeneCount.tsv\",sep=\"\\t\")\n",
    "unassigned_df = pd.read_csv(\"Orthogroups_UnassignedGenes.tsv\",sep=\"\\t\",dtype=str).fillna(\"\")\n",
    "org_list      = list(assigned_df.columns.values)[1:-1]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c2586135-5195-47a7-8da6-ea7a8393bc27",
   "metadata": {},
   "source": [
    "### Because the list of unassigned genes contains an \"orthogroup\" per gene per species, the resulting dataframe is excessively long, witht the following lines we can collapse the dataframe into a symmetrical matrix containing a single row per species indicating how many exclusive genes per species we have"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "8f79d856-a466-4d0e-be5a-2ec8d2ddd19c",
   "metadata": {},
   "outputs": [],
   "source": [
    "missing_list = []\n",
    "for org_name in org_list:\n",
    "    org_index = str(org_list.index(org_name))\n",
    "    missed_og_name = \"MISSING_OG\"+org_index\n",
    "    sub_total = len(unassigned_df[unassigned_df[org_name]!= \"\"])\n",
    "    missed_og_dict = {}\n",
    "    missed_og_dict[\"Orthogroup\"] = missed_og_name\n",
    "    for sub_org in org_list:\n",
    "        if (sub_org==org_name):\n",
    "            sub_total_value = sub_total\n",
    "        else:\n",
    "            sub_total_value = 0\n",
    "        missed_og_dict[sub_org] = sub_total_value\n",
    "    missing_list.append(missed_og_dict)\n",
    "missing_df = pd.DataFrame.from_records(missing_list)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4d4ac5c7-b45a-411f-b563-135dc001cfa0",
   "metadata": {},
   "source": [
    "### And then we can put the two dataframes together"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "f00e2ab2-6296-4335-83e0-a79f3cdd6c3a",
   "metadata": {},
   "outputs": [],
   "source": [
    "ortho_df = pd.concat([assigned_df, missing_df], sort=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a0aca47e-4623-4225-ba53-defa66ea33d2",
   "metadata": {},
   "outputs": [],
   "source": [
    "o"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "03735ea7-5d6f-4635-97a0-40b9484601b1",
   "metadata": {},
   "source": [
    "# 2. Selecting the species"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "abe8deea-54ce-4734-ba03-31af7beafb51",
   "metadata": {},
   "source": [
    "## The following code is just to make things interactive, here you can select as many species as you want for your figure\n",
    "\n",
    "> ## Note that if you modify the list of selected species, you need to run all cells from here onwards"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "060be0e2-819c-4424-ab8d-ecb9b9676ee9",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "daf9ad863d0f4bb5afba0d024afc8a8a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "VBox(children=(Checkbox(value=False, description='Acanthamoeba_castellani.NEFF', indent=False), Checkbox(value…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "checkboxes = [widgets.Checkbox(value=False, description=label,indent=False) for label in org_list]\n",
    "output = widgets.VBox(children=checkboxes,)\n",
    "display(output)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "1ab9e41e-4808-4ebf-8141-035b140c2e59",
   "metadata": {},
   "outputs": [],
   "source": [
    "sel_org_list = []\n",
    "for i in range(0, len(checkboxes)):\n",
    "    if checkboxes[i].value == True:\n",
    "        sel_org_list = sel_org_list + [checkboxes[i].description]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ce25ed7e-a211-471e-887d-4aa7591a6a92",
   "metadata": {},
   "source": [
    "### Based on the selected organisms, we'll build a binary matrix that includes all the possibilities for an orthogroup to be shared across the number of species\n",
    "\n",
    "### As an example, a list of 3 organisms will produce 8 different orthogroup classes\n",
    "\n",
    "|Species 1|Species 2|Species 3|Orthogroup class|\n",
    "|:-------:|:-------:|:-------:|:--------------:|\n",
    "|Present  |Present  |Present  |        0       |\n",
    "|Present  |Present  |Absent   |        1       |\n",
    "|Present  |Absent   |Present  |        2       |\n",
    "|Present  |Absent   |Absent   |        3       |\n",
    "|Absent   |Present  |Present  |        4       |\n",
    "|Absent   |Present  |Absent   |        5       |\n",
    "|Absent   |Absent   |Present  |        6       |\n",
    "|Absent   |Absent   |Absent   |        7       |\n",
    "\n",
    "### The number of orthogroup classes grows exponentially with respect to the number of organisms, this is also the number of bars in your figure, thus, it is important to keep the number sensible not to produce an overly large and massively unreadable figure, also, not to consume too much memory"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "203f4523-5610-402b-8199-cb2307a17e98",
   "metadata": {},
   "outputs": [],
   "source": [
    "num_sel_orgs  = len(sel_org_list)\n",
    "binary_matrix = ff2n(num_sel_orgs)\n",
    "binary_list   = binary_matrix.tolist()\n",
    "binary_tuple  = [tuple(row) for row in binary_list]\n",
    "class_list    = list(range(0,len(binary_tuple)))\n",
    "binary_dict   = dict(zip(binary_tuple,class_list))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cab87129-9358-4480-af98-9ab0b9d8fd93",
   "metadata": {},
   "source": [
    "# 3. Data transformation"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "add85a4c-620e-4156-ac0e-be8261dd9cf1",
   "metadata": {},
   "source": [
    "## We will produce two additional columns, one containing the number of orthogroups, and another containing the number of genes, that will make sense later"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "31b5d2e8-d473-4986-9d77-3144478b3fde",
   "metadata": {},
   "outputs": [],
   "source": [
    "sub_cols  = [\"Orthogroup\"] + sel_org_list\n",
    "sub_df    = ortho_df.copy()\n",
    "sub_df    = sub_df[sub_cols]\n",
    "sub_df[\"gene_total\"] = sub_df.loc[:,sub_df.columns != \"Orthogroup\"].apply(np.sum, axis=1)\n",
    "sub_df    = sub_df[sub_df[\"gene_total\"]!=0]\n",
    "for org_name in sel_org_list:\n",
    "    sub_df[org_name] = sub_df[org_name].apply(lambda x: 1 if x > 0 else -1)\n",
    "sub_df[\"og_total\"] = sub_df.apply(lambda x: x[\"gene_total\"] if \"MISS\" in x[\"Orthogroup\"] else 1,axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "76c4f8d1-0767-4b01-98d4-0a69b7c20fe3",
   "metadata": {},
   "source": [
    "## We then encode the orthogroup dataframe into a binary matrix so that we can decode the orthogroup classes based on their presence or absence across the selected species"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "5de7d130-45d4-46e0-bdaf-f6aa6a4eb379",
   "metadata": {},
   "outputs": [],
   "source": [
    "sub_df[\"og_class\"] = sub_df[sel_org_list].apply(lambda x: get_og_class(tuple(x.values.flatten().tolist())),axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "dbd1c120-6a67-483d-877e-07b8c7ad95d8",
   "metadata": {},
   "source": [
    "## Then we calculate the total number of orthogroups and the total number of genes per class"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "7af99091-fb74-483c-96ae-4e26306cf9f2",
   "metadata": {},
   "outputs": [],
   "source": [
    "tuple_list        = []\n",
    "class_og_totals   = []\n",
    "class_gene_totals = []\n",
    "for value in binary_dict.values():\n",
    "    presence_tuple = [k for k,v in binary_dict.items() if v == value][0]\n",
    "    presence_list  = list(presence_tuple)\n",
    "    index_list = []\n",
    "    for dummy,(matrix_value,organism) in enumerate(zip(presence_list,sel_org_list)):\n",
    "        if matrix_value == 1:\n",
    "            index_list.append(True)\n",
    "        else:\n",
    "            index_list.append(False)\n",
    "    class_og_total   = np.sum(sub_df[sub_df[\"og_class\"]==value][\"og_total\"])\n",
    "    class_gene_total = np.sum(sub_df[sub_df[\"og_class\"]==value][\"gene_total\"])\n",
    "    class_og_totals.append(class_og_total)\n",
    "    class_gene_totals.append(class_gene_total)\n",
    "    index_tuple = tuple(index_list)\n",
    "    tuple_list.append(index_tuple)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e2be3563-fb94-4087-a3de-18ef5e4dbe92",
   "metadata": {},
   "source": [
    "## Based on the number of species and the number of classes we'll build two dataframes, one containing the total number of orthogroups per class and one containing the total number of genes per class"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "6e18af57-4843-4cb6-a199-fea91646ccb8",
   "metadata": {},
   "outputs": [],
   "source": [
    "multi_index = pd.MultiIndex.from_tuples(tuple_list,names=sel_org_list)\n",
    "upset_og_df   = pd.Series(class_og_totals,index=multi_index)\n",
    "upset_gene_df = pd.Series(class_gene_totals,index=multi_index)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "442b98b9-6210-4ae0-8ffe-379842ffed9a",
   "metadata": {},
   "source": [
    "# 4. Plotting your results"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3aca2842-c1ea-463c-bc18-39d186b8ab4e",
   "metadata": {},
   "source": [
    "## Once you have the last two dataframes, you can plot them with the following code, the options are self explanatory and you can save your figure to a file, the extension dictates the file format"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "921adbcd-e2b6-43ec-b807-10bc2d522b49",
   "metadata": {},
   "source": [
    "### Code to plot the orthogroup totals"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "b0544ab1-5bb1-4b22-a401-f13b398f93ef",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x720 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=(20,10))\n",
    "upset = plot(upset_og_df, fig=fig, element_size=None,show_counts=True)\n",
    "plt.suptitle(\"Your legend goes here\")\n",
    "plt.savefig(\"og_upset_plot.svg\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "242bab5f-085a-4da3-baa0-5c4fd5aa1165",
   "metadata": {},
   "source": [
    "### Code to plot the gene totals"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "b04fa6b7-0be9-407b-961c-ca643b3b6502",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABIwAAAKHCAYAAADjQdNWAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjQuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/MnkTPAAAACXBIWXMAAAsTAAALEwEAmpwYAABrPklEQVR4nO3deZhcVZ3/8feXhAAhhBAxCRAQEGSHAJFlUIwybC4giAjjSBQUnMEFFxBnfjMobuioiKIoigouLIpK3FAEAm4sYd8hsphECELYt5Dw/f1xbid1m+6ku9PV1dX9fj1PPV116t6qb9Xtqrr1qXPOjcxEkiRJkiRJ6rBSqwuQJEmSJEnS4GJgJEmSJEmSpBoDI0mSJEmSJNUYGEmSJEmSJKnGwEiSJEmSJEk1BkaSJEmSJEmqMTCSJEkDLiLeGRF/anUdnUXEzIh4d6vr6E5EZERs0uo6JEnS0GdgJEnSMBYRP4yI73Vqe01EPBwR67SqLkmSJLWWgZEkScPbB4F9I2JPgIhYFfg28JHMvL8/7iAiRvbH7aj/uE0kSdLyGBhJkjSMZebDwPuB0yNideAE4G+Z+f2I2C8ibomIR6uhWlt0rNd5aFREfD8iPl2dnxYRcyPiYxHxAPA9liMiNo+IiyJiQUTcEREHN1z3koj4ZUQ8HhFXR8SnG4ezLWfd70fE1yPi1xHxRERcGREvb7h+z4i4PSIei4hTgVhGjatFxJkR8UhE3BYRx0XE3Ibrt6iep0er522/hutWiYgvRsTfI2J+RHwzIlarrls7In5VrbcgIv4YEcvaR/vXiLirWv7rEbGk5og4vKrtkYj4XUS8rOG6jIijI+Iu4K6q7Y0RcX11W3+JiG2XuaEkSdKwYWAkSdIwl5k/Aa4FzgaOBI6MiFdUl48BXgr8BvhlRIzq4c1OAsYDL6tus1tVUHUR8GNgAnAI8I2I2LJa5OvAU9VtTq9OPV2Xqu2TwFrAbOAz1bprAz8D/h+wNvA3YLdllHoCsCGwMbAn8O8NdawM/BL4fVXH+4EfRcRm1SInAa8ApgCbAOsB/1td9xFgLuV5ngj8F5DLqOONwCuBbYGDgb2rGvav1j2wuq0/UrZhozcDOwNbRsT2wHeBo4CXAN8CZkTEKsu4b0mSNEwYGEmSJID/BF4HnJiZc4C3Ab/OzIsy83ngi8BqwL/08PZeAE7IzOcy85nlLPtG4N7M/F5mLsrM64DzgbdGxAjgLdVtPZ2ZtwJn9mTdhmV+nplXZeYi4EeU0Abg9cAtmfnT6jF+BXhgGXUeDHw2Mx/JzLnAVxuu2wUYA5yUmQsz8xLgV8ChVQ+gI4EPZeaCzHwC+CwlyAJ4HlgHeFlmPp+Zf8zMZQVGJ2Xmo5n5d+DShsfzXuBzmXlb9Vg/C0xp7GVUXb+g2iZHAt/KzCszc3Fmngk8Vz0WSZI0zBkYSZIkMnM+8BBwS9W0LnBfw/UvAHMoPWN64p+Z+WwPl30ZsHM1LOrRiHgUeDulR9FLgZHVfXeY08N1OzSGQE9Tgh0oj3HJbVUhTeNtd7Yu3dexLjCnep463Ed5vl4KjAauaajxwqod4P8oPZ9+HxF3R8Txy6hhWY/nZcApDfexgDLErnGbdX7uPtLpuVu/eiySJGmYc8JDSZLUlX8A23RcqHrJrA/Mq5qepoQgHSZRhlV1WFYPmc7mAJdl5p6dr6h6GC0CJgN3Vs3r92TdHri/8bYaHuOylp8M3NpFHf8A1o+IlRpCow2qmh8CngG2ysx5dFL1OPoIJbzZGrgkIq7OzIt7+XjmAJ/JzB8tY5nG7dKx/Gd6eT+SJGkYsIeRJEnqynnAGyJij2p+no9Qhiv9pbr+euDfImJEROwDvGYF7utXwCsi4h0RsXJ1emVEbJGZiynzDH0iIkZHxObAYT1Ztwf3+2tgq4g4MMpRwz5AvWdSZ+cBH4+ItSJiPeB9DdddSQnRjqtqmAa8CTinCpC+DZwcERMAImK9iOiYe+iNEbFJFVg9BiymDOnrrW9W9W1V3e6aEfHWZSz/beC9EbFzFKtHxBsiYo0+3LckSRpiDIwkSdKLZOYdlEmdv0bpIfMm4E2ZubBa5INV26OUIWC/WIH7egLYizKnzz8oQ64+D3RMvvw+YM2q/QeUiZyf6+G6y7rfhyhzHZ0EPAxsCvx5GaucSOlFdQ/wB+CnDXUspDwf+1Ker28Ah2Xm7dW6H6MMO7siIh6v1u+YEHvT6vKTwF+Bb2Tmpcurv4vH83PKYz+nuo+bq3q6W34W8B7gVOCRqr539vZ+JUnS0BTLnlNRkiRpcImIzwOTMnP6chdubh3/ARySmSvSu0qSJGlQsoeRJEka1CJi84jYtho2tRNwBPDzFtSxTkTsFhErRcRmlGF6A16HJEnSQHDSa0mSNNitQRmGti4wH/gScEEL6hgFfAvYiDIU7xzK0DNJkqQhxyFpkiRJkiRJqnFImiRJkiRJkmoMjCRJkiRJklRjYCRJkiRJkqQaAyNJkiRJkiTVGBhJkiRJkiSpxsBIkiRJkiRJNQZGkiRJkiRJqjEwkiRJkiRJUo2BkSRJkiRJkmoMjCRJkiRJklRjYCRJkiRJkqQaAyNJkiRJkiTVGBhJkiRJkiSpxsBIkiRJkiRJNQZGkiRJkiRJqjEwkiRJkiRJUo2BkSRJkiRJkmoMjCRJkiRJklRjYCRJkiRJkqQaAyNJkiRJkiTVGBhJkiRJkiSpxsBIkiRJkiRJNQZGkiRJkiRJqjEwkiRJkiRJUo2BkSRJkiRJkmoMjCRJkiRJklRjYCRJkiRJkqQaAyNJkiRJkiTVGBhJkiRJkiSpxsBIkiRJkiRJNQZGkiRJkiRJqjEwkiRJkiRJUo2BkSRJkiRJkmoMjCRJkiRJklRjYCRJkiRJkqQaAyNJkiRJkiTVGBhJkiRJkiSpxsBIkiRJkiRJNQZGkiRJkiRJqjEwkiRJkiRJUo2BkSRJkiRJkmoMjCRJkiRJklRjYCRJkiRJkqQaAyNJkiRJkiTVGBhJkiRJkiSpxsBIkiRJkiRJNQZGkiRJkiRJqjEwkiRJkiRJUo2BkSRJkiRJkmoMjCRJkiRJklRjYCRJkiRJkqQaAyNJkiRJkiTVGBhJkiRJkiSpxsBIkiRJkiRJNQZGkiRJkiRJqjEwkiRJkiRJUo2BkSRJkiRJkmoMjCRJkiRJklQzstUFSNJws88+++SFF17Y6jIkSZIkKbq7wh5GkjTAHnrooVaXIEmSJEnLZGAkaViKiA9FxC0RcXNEnB0Rq0bERhFxZUTMjohzI2JUtewq1eXZ1fUbNtzOx6v2OyJi75Y9IEmSJEnqRwZGkoadiFgP+AAwNTO3BkYAhwCfB07OzE2AR4AjqlWOAB6p2k+uliMitqzW2wrYB/hGRIwYyMciSZIkacXdcccdTJkyZclp7NixfOUrX+Ftb3vbkrYNN9yQKVOm1Nb7+9//zpgxY/jiF7+4pO3CCy9ks802Y5NNNuGkk04a4EfSf5zDSNJwNRJYLSKeB0YD9wOvA/6tuv5M4BPAacD+1XmAnwKnRkRU7edk5nPAPRExG9gJ+OsAPQZJkiRJ/WCzzTbj+uuvB2Dx4sWst956HHDAARxzzDFLlvnIRz7CmmuuWVvvwx/+MPvuu++Sy4sXL+boo4/moosuYvLkybzyla9kv/32Y8sttxyIh9Gv7GEkadjJzHnAF4G/U4Kix4BrgEczc1G12Fxgver8esCcat1F1fIvaWzvYh1JkiRJbejiiy/m5S9/OS972cuWtGUm5513HoceeuiStl/84hdstNFGbLXVVkvarrrqKjbZZBM23nhjRo0axSGHHMIFF1wwoPX3F3sYSRp2ImItSu+gjYBHgZ9QhpQ18z6PBI4EmDhxIjNnzmzm3UmSJEnqo5NPPpmpU6fW9tlvuOEGRo8ezbx585g3bx7PPPMM//3f/80Xv/hFzj33XFZbbTVmzpzJZZddxsiRI5es+8QTT3DbbbcN2v3/adOmdXudgZGk4ehfgXsy858AEfEzYDdgXESMrHoRTQbmVcvPA9YH5kbESGBN4OGG9g6N69Rk5unA6QBTp07NZb0xS5IkSWqNhQsXcvXVV/P973+fiRMnLmk/99xzOfLII5cELB/96Ec54YQT2HfffbnyyisZM2YM06ZN46GHHmLOnDlLlpszZw6PP/74MoOZwcrASNJw9Hdgl4gYDTwD7AHMAi4FDgLOAaYDHX1HZ1SX/1pdf0lmZkTMAH4cEV8G1gU2Ba4ayAciSZIkqf/89re/ZYcddqiFRYsWLeJnP/sZ11xzzZK2K6+8kp/+9Kccd9xxPProo6y00kqsuuqq7LjjjsyZs3TWirlz57Leeu05a4WBkaRhJzOvjIifAtcCi4DrKL1/fg2cExGfrtrOqFY5A/hBNan1AsqR0cjMWyLiPODW6naOzszFA/pgJEmSJPWbs88+uzZPEcAf/vAHNt98cyZPnryk7Y9//OOS85/4xCcYM2YM73vf+1i0aBF33XUX99xzD+uttx7nnHMOP/7xjwes/v5kYCRpWMrME4ATOjXfTTnKWedlnwXe2s3tfAb4TL8XKEmSJGlAPfXUU1x00UV861vfqrWfc845LwqRujNy5EhOPfVU9t57bxYvXszhhx9emxS7nURmtroGSRpWpk6dmrNmzWp1GZIkSZIU3V2x0kBWIUmSJEmSpMHPwEiSJEmSJEk1BkaSJEmSJEmqMTCSJEmSJElSjYGRJEmSJEmSaka2ugBJkiRJkqT+MGnSJObPn9/qMvrFxIkTeeCBB1p2//YwkiRJkiRJQ8JQCYug9Y/FwEiSJEmSJEk1BkaSJEmSJEmqMTCSJEmSJElSjYGRJEmSJEmSagyMJEmSJEmSVGNgJEmSJEmSpBoDI0mSJEmSJNUYGEmSJEmSJKnGwEiSJEmSJEk1BkaSJEmSJEmqMTCSJEmSJElSjYGRJEmSJEmSagyMJEmSJEmSVGNgJEmSJEmSpBoDI0mSJEmSJNUYGEmSJEmSJKnGwEiSJEmSJEk1BkaSJEmSJEmqMTCSJEmSJElSjYGRJEmSJEmSagyMJEmSJEmSVGNgJEmSJEmSpBoDI0mSJEmSJNUYGEmSJEmSJKnGwEiSJEmSJEk1BkaSJEmSJEmqMTCSNCxFxGYRcX3D6fGIOCYixkfERRFxV/V3rWr5iIivRsTsiLgxInZouK3p1fJ3RcT01j0qSZIkSeofBkaShqXMvCMzp2TmFGBH4Gng58DxwMWZuSlwcXUZYF9g0+p0JHAaQESMB04AdgZ2Ak7oCJkkSZIkqV0ZGEkS7AH8LTPvA/YHzqzazwTeXJ3fHzgriyuAcRGxDrA3cFFmLsjMR4CLgH0GtHpJkiRJ6mcGRpIEhwBnV+cnZub91fkHgInV+fWAOQ3rzK3aumuXJEmSpLY1stUFSFIrRcQoYD/g452vy8yMiOyn+zmSMpSNiRMnMnPmzP64WUmSJElDWLO/N0ybNq3b6wyMJA13+wLXZub86vL8iFgnM++vhpw9WLXPA9ZvWG9y1TYPmNapfWbnO8nM04HTAaZOnZrLemOWJEmSJFh2oNNsDkmTNNwdytLhaAAzgI4jnU0HLmhoP6w6WtouwGPV0LXfAXtFxFrVZNd7VW2SJEmS1LbsYSRp2IqI1YE9gaMamk8CzouII4D7gIOr9t8ArwdmU46o9i6AzFwQEZ8Crq6WOzEzFwxA+ZIkSZLUNJHZL9NzSJJ6aOrUqTlr1qxWlyFJkiQNORHR6hL61QBkNt0+YQ5JkyRJkiRJUo2BkSRJkiRJkmoMjCRJkiRJklRjYCRJkiRJkqQaAyNJkiRJkiTVGBhJkiRJkiSpxsBIkiRJkiRJNQZGkiRJkiRJqjEwkiRJkiRJUo2BkSRJkiRJkmoMjCRJkiRJklRjYCRJkiRJkqQaAyNJkiRJkiTVGBhJkiRJkiSpxsBIkiRJkiRJNQZGkiRJkiRJqjEwkiRJkiRJUo2BkSRJkiRJkmoMjCRJkiRJklRjYCRJkiRJkqQaAyNJkiRJkiTVGBhJkiRJkiSpxsBIkiRJkiRJNQZGkiRJkiRJqjEwkiRJkiRJUo2BkSRJkiRJkmoMjCRJkiRJklRjYCRJkiRJkqQaAyNJkiRJkiTVGBhJkiRJkiSpxsBIkiRJkiRJNQZGkiRJkiRJqjEwkiRJkiRJUo2BkSRJkiRJkmoMjCQNSxExLiJ+GhG3R8RtEbFrRIyPiIsi4q7q71rVshERX42I2RFxY0Ts0HA706vl74qI6a17RJIkSZLUfwyMJA1XpwAXZubmwHbAbcDxwMWZuSlwcXUZYF9g0+p0JHAaQESMB04AdgZ2Ak7oCJkkSZIkqZ0ZGEkadiJiTWB34AyAzFyYmY8C+wNnVoudCby5Or8/cFYWVwDjImIdYG/gosxckJmPABcB+wzYA5EkSZKkJjEwkjQcbQT8E/heRFwXEd+JiNWBiZl5f7XMA8DE6vx6wJyG9edWbd21S5IkSVJbG9nqAiSpBUYCOwDvz8wrI+IUlg4/AyAzMyKyv+4wIo6kDGdj4sSJzJw5s79uWpIkSdIQ1ezvDdOmTev2OgMjScPRXGBuZl5ZXf4pJTCaHxHrZOb91ZCzB6vr5wHrN6w/uWqbB0zr1D6zqzvMzNOB0wGmTp2ay3pjliRJkiRYdqDTbA5JkzTsZOYDwJyI2Kxq2gO4FZgBdBzpbDpwQXV+BnBYdbS0XYDHqqFrvwP2ioi1qsmu96raJEmSJKmt2cNI0nD1fuBHETEKuBt4FyVEPy8ijgDuAw6ulv0N8HpgNvB0tSyZuSAiPgVcXS13YmYuGLiHIEmSJEnNEZn9NkWHJKkHpk6dmrNmzWp1GZIkSdKQExGtLqFfDUBm0+0T5pA0SZIkSZIk1RgYSZIkSZIkqcbASJIkSZIkSTUGRpIkSZIkSaoxMJIkSZIkSVKNgZEkSZIkSZJqDIwkSZIkSZJUY2AkSZIkSZKkGgMjSZIkSZIk1RgYSZIkSZIkqcbASJIkSZIkSTUGRpIkSZIkSaoxMJIkSZIkSVKNgZEkSZIkSZJqDIwkSZIkSZJUY2AkSZIkSZKkGgMjSZIkSZIk1RgYSZIkSZIkqcbASJIkSZIkSTUGRpIkSZIkSaoxMJIkSZIkSVKNgZEkSZIkSZJqDIwkSZIkSZJUY2AkSZIkSZKkGgMjSZIkSZIk1RgYSZIkSZIkqcbASJIkSZIkSTUGRpIkSZIkSaoxMJIkSZIkSVKNgZEkSZIkSZJqDIwkSZIkSZJUY2AkSZIkSZKkGgMjSZIkSZIk1RgYSZIkSZIkqcbASJIkSZIkSTUGRpKGrYi4NyJuiojrI2JW1TY+Ii6KiLuqv2tV7RERX42I2RFxY0Ts0HA706vl74qI6a16PJIkSZLUXwyMJA13r83MKZk5tbp8PHBxZm4KXFxdBtgX2LQ6HQmcBiVgAk4AdgZ2Ak7oCJkkSZIkqV0ZGElS3f7AmdX5M4E3N7SflcUVwLiIWAfYG7goMxdk5iPARcA+A1yzJEmSJPUrAyNJw1kCv4+IayLiyKptYmbeX51/AJhYnV8PmNOw7tyqrbt2SZIkSWpbI1tdgCStqIh4GbBpZv4hIlYDRmbmEz1Y9VWZOS8iJgAXRcTtjVdmZkZE9lONR1KGsjFx4kRmzpzZHzcrSZIkaQhr9veGadOmdXudgZGkthYR76EEMeOBlwOTgW8Ceyxv3cycV/19MCJ+TpmDaH5ErJOZ91dDzh6sFp8HrN+w+uSqbR4wrVP7zC7u63TgdICpU6fmst6YJUmSJAmWHeg0m0PSJLW7o4HdgMcBMvMuYMLyVoqI1SNijY7zwF7AzcAMoONIZ9OBC6rzM4DDqqOl7QI8Vg1d+x2wV0SsVU12vVfVJkmSJEltyx5Gktrdc5m5MCIAiIiRlLmJlmci8PNqvZHAjzPzwoi4GjgvIo4A7gMOrpb/DfB6YDbwNPAugMxcEBGfAq6uljsxMxf0yyOTJEmSpBYxMJLU7i6LiP8CVouIPYH/BH65vJUy825guy7aH6aL4WyZmZTeTF3d1neB7/aybkmSJEkatBySJqndHQ/8E7gJOAr4TWb+d2tLkiRJkqT2Zg8jSe3u7cA5mfntjoaIeGNm/qqFNUmSJElSW7OHkaR29zXgjxGxRUPbia0qRpIkSZKGAgMjSe3uHuBw4KcR8daqLVpYjyRJkiS1PYekSWp3mZnXRsRrgLMjYmdgRKuLkiRJkqR2Zg8jSe3ufoDMfAjYG0hg65ZWJEmSJEltzsBIUlvLzDc0nH8hM4/NTN/bNKQsXryY7bffnje+8Y219g984AOMGTNmyeUvf/nLbLnllmy77bbsscce3HfffUuu22effRg3btyLbkOSJEnqil+qJLWliPhK9feXETGj86nF5Un96pRTTmGLLbaotc2aNYtHHnmk1rb99tsza9YsbrzxRg466CCOO+64Jdcde+yx/OAHPxiQeiVJktT+DIwktauOb75fBL7UxUkaEubOncuvf/1r3v3udy9pW7x4Mcceeyxf+MIXasu+9rWvZfTo0QDssssuzJ07d8l1e+yxB2usscbAFC1JkqS256TXktpSZl5T/b2soy0i1gLWz8wbW1aY1M+OOeYYvvCFL/DEE08saTv11FPZb7/9WGeddbpd74wzzmDfffcdiBIlSZI0BBkYSWprETET2I/yfnYN8GBE/DkzP9zSwqR+8Ktf/YoJEyaw4447MnPmTAD+8Y9/8JOf/GTJ5a788Ic/ZNasWVx22WXdLiNJkiQti4GRpHa3ZmY+HhHvBs7KzBMiwh5GGhL+/Oc/M2PGDH7zm9/w7LPP8vjjj7PVVluxyiqrsMkmmwDw9NNPs8kmmzB79mwA/vCHP/CZz3yGyy67jFVWWaWV5UuSJKmNOYeRpHY3MiLWAQ4GftXqYqT+9LnPfY65c+dy7733cs455/C6172ORx55hAceeIB7772Xe++9l9GjRy8Ji6677jqOOuooZsyYwYQJE1pcvSRJktqZPYwktbsTgd8Bf8rMqyNiY+CuFtcktcSxxx7Lk08+yVvf+lYANthgA2bMKAcNfPWrX83tt9/Ok08+yeTJkznjjDPYe++9W1muJEmSBrHIzFbXIEnDytSpU3PWrFmtLkOSJEkaciKi1SX0qwHIbLp9whySJkmSJEmSpBoDI0mSJEmSJNUYGEmSJEmSJKnGSa8ltbWIWAV4C7AhDe9pmXliq2qSJEmSpHZnYCSp3V0APAZcAzzX4lokSZIkaUgwMJLU7iZn5j6tLkJankmTJjF//vxWl7HCJk6cyAMPPNDqMiRJktRkzmEkqd39JSK2aXUR0vIMhbAIhs7jkCRJ0rLZw0hSu3sV8M6IuIcyJC2AzMxtW1uWJEmSJLUvAyNJ7W7fVhcgSZIkSUONQ9IktbXMvA8YB7ypOo2r2iRJkiRJfWRgJKmtRcQHgR8BE6rTDyPi/a2tSpIkSZLam0PSJLW7I4CdM/MpgIj4PPBX4GstrUqSJEmS2pg9jCS1uwAWN1xeXLVJkiRJkvrIHkaS2t33gCsj4ufV5TcDZ7SuHEmSJElqfwZGktpaZn45ImYCr6qa3pWZ17WwJEmSJElqewZGktpSRIzNzMcjYjxwb3XquG58Zi5oVW2SJEmS1O4MjCS1qx8DbwSuAbKhParLG7eiKEmSJEkaCgyMJLWlzHxj9XejVtciSZIkSUONR0mT1NYi4uKetEmSJEmSes7ASFJbiohVq/mL1o6ItSJifHXaEFivh7cxIiKui4hfVZc3iogrI2J2RJwbEaOq9lWqy7Or6zdsuI2PV+13RMTe/f9IJUmSJGngGRhJaldHUeYv2rz623G6ADi1h7fxQeC2hsufB07OzE2AR4AjqvYjgEeq9pOr5YiILYFDgK2AfYBvRMSIFXhMkiRJkjQoGBhJakuZeUo1f9FHM3PjzNyoOm2XmcsNjCJiMvAG4DvV5QBeB/y0WuRM4M3V+f2ry1TX71Etvz9wTmY+l5n3ALOBnfrnEUqSJElS6xgYSWp3L0TEuI4L1fC0/+zBel8BjgNeqC6/BHg0MxdVl+eydGjbesAcgOr6x6rll7R3sY4kSZIktS2Pkiap3b0nM7/ecSEzH4mI9wDf6G6FiHgj8GBmXhMR05pfIkTEkcCRABMnTmTmzJkDcbdSU/j/K0mSNDCavd81bdq0bq8zMJLU7kZERGRmQpnIGhi1nHV2A/aLiNcDqwJjgVOAcRExsupFNBmYVy0/D1gfmBsRI4E1gYcb2js0rlOTmacDpwNMnTo1l/XGLA12/v9KkiQNjFbudzkkTVK7uxA4NyL2iIg9gLOrtm5l5sczc3JmbkiZtPqSzHw7cClwULXYdMoE2gAzqstU119SBVQzgEOqo6htBGwKXNV/D02SJEmSWsMeRpLa3ccoR0z7j+ryRVQTWffxts6JiE8D1wFnVO1nAD+IiNnAAkrIRGbeEhHnAbcCi4CjM3NxH+9bkiRJkgaNqEZxSFLbiojVgA0y845W19ITU6dOzVmzZrW6DA2wcmC9ocF9B0mSNFgNpX0uGJD9rm6fMIekSWprEbEfcD3VMLSImBIRM1palCRJkiS1OQMjSe3uBGAn4FGAzLwe2KiF9UiSJElS2zMwktTuns/Mxzq1OV5GkiRJklaAk15Lane3RMS/ASMiYlPgA8BfWlyTJEmSJLU1exhJanfvB7YCngPOBh4HjmllQZIkSZLU7uxhJKmtZebTwH8D/x0RI4DVM/PZFpclSZIkSW3NHkaS2lpE/DgixkbE6sBNwK0RcWyr65IkSZKkdmZgJKndbZmZjwNvBn5LOULaO1pakSRJkiS1OQMjSe1u5YhYmRIYzcjM5/EoaZIkSZK0QgyMJLW7bwL3AqsDl0fEyygTX0uSJEmS+shJryW1rYhYCZifmes1tP0deG3rqpIkSZKk9mcPI0ltKzNfAI7r1JaZuahFJUmSJEnSkGBgJKnd/SEiPhoR60fE+I5Tq4uSJEmSpHbmkDRJ7e5t1d+jG9oS2LgFtUiSJEnSkGBgJKmtZeZGra5BkiRJkoYah6RJamsRMToi/l9EnF5d3jQi3tjquiRJkiSpnRkYSWp33wMWAv9SXZ4HfLp15UiSJElS+zMwktTuXp6ZXwCeB8jMp4FobUmSJEmS1N4MjCS1u4URsRplomsi4uXAc60tSZIkSZLam5NeS2p3nwAuBNaPiB8BuwHvamlFkiRJktTmDIwktbXM/H1EXAPsQhmK9sHMfKjFZUmSJElSW3NImqS2FhEXZ+bDmfnrzPxVZj4UERe3ui5JkiRJamf2MJLUliJiVWA0sHZErMXSia7HAuu1rDBJkiRJGgIMjCS1q6OAY4B1gWtYGhg9DpzaopokSZIkaUgwMJLUljLzFOCUiHh/Zn6t1fVIkiRJ0lBiYCSprWXm1yLiX4ANaXhPy8yzWlaUJEmSJLU5AyNJbS0ifgC8HLgeWFw1J2BgJEmSJEl9ZGAkqd1NBbbMzGx1IZIkSZI0VKzU6gIkaQXdDExqdRGSJEmSNJTYw0hSu1sbuDUirgKe62jMzP1aV5IkSZIktTcDI0nt7hOtLkCSJEmShhoDI0ltLTMva3UNkiRJkjTUGBhJaksR8QTlaGgvugrIzBw7wCVJkiRJ0pBhYCSpLWXmGq2uQZIkSZKGKo+SJmlYiohVI+KqiLghIm6JiE9W7RtFxJURMTsizo2IUVX7KtXl2dX1Gzbc1ser9jsiYu8WPSRJkiRJ6jcGRpKGq+eA12XmdsAUYJ+I2AX4PHByZm4CPAIcUS1/BPBI1X5ytRwRsSVwCLAVsA/wjYgYMZAPRJIkSZL6m4GRpGEpiyeriytXpwReB/y0aj8TeHN1fv/qMtX1e0REVO3nZOZzmXkPMBvYqfmPQJIkSZKax8BI0rAVESMi4nrgQeAi4G/Ao5m5qFpkLrBedX49YA5Adf1jwEsa27tYR5IkSZLakpNeSxq2MnMxMCUixgE/BzZv1n1FxJHAkQATJ05k5syZzborqen8/5UkSRoYzd7vmjZtWrfXGRhJGvYy89GIuBTYFRgXESOrXkSTgXnVYvOA9YG5ETESWBN4uKG9Q+M6jfdxOnA6wNSpU3NZb8zSYOf/ryRJ0sBo5X6XQ9IkDUsR8dKqZxERsRqwJ3AbcClwULXYdOCC6vyM6jLV9ZdkZlbth1RHUdsI2BS4akAehCRJkiQ1iT2MJA1X6wBnVkc0Wwk4LzN/FRG3AudExKeB64AzquXPAH4QEbOBBZQjo5GZt0TEecCtwCLg6GqomyRJkiS1rSg/kEuSBsrUqVNz1qxZrS5DA6wcVG9ocN9BkiQNVkNpnwsGZL+r2yfMIWmSJEmSJEmqMTCSJEmSJElSjYGRJEmSJEmSagyMJEmSJEmSVGNgJEmSJEmSpBoDI0mSJEmSJNUYGEmSJEmSJKnGwEiSJEmSJEk1BkaSJEmSJEmqMTCSJEmSJElSjYGRJEmSJEmSagyMJEmSJEmSVGNgJEmSJEmSpBoDI0mSJEmSJNUYGEmSJEmSJKnGwEiSJEmSJEk1BkaSJEmSJEmqMTCSJEmSJElSjYGRJEmSJEmSagyMJEmSJEmSVGNgJEmSJEmSpBoDI0mSJEmSJNUYGEmSJEmSJKnGwEiSJEmSJEk1BkaSJEmSJEmqMTCSJEmSJElSjYGRJEmSJEmSagyMJEmSJEmSVGNgJEmSJEmSpBoDI0mSpB46/PDDmTBhAltvvfWStuuvv55ddtmFKVOmMHXqVK666ioAHnnkEQ444AC23XZbdtppJ26++eYl6zz66KMcdNBBbL755myxxRb89a9/HfDHIkmStCwGRpIkST30zne+kwsvvLDWdtxxx3HCCSdw/fXXc+KJJ3LccccB8NnPfpYpU6Zw4403ctZZZ/HBD35wyTof/OAH2Weffbj99tu54YYb2GKLLQb0cUiSJC2PgZEkSVIP7b777owfP77WFhE8/vjjADz22GOsu+66ANx666287nWvA2DzzTfn3nvvZf78+Tz22GNcfvnlHHHEEQCMGjWKcePGDdyDkCRJ6oGRrS5AkiSpnX3lK19h77335qMf/SgvvPACf/nLXwDYbrvt+NnPfsarX/1qrrrqKu677z7mzp3LiBEjeOlLX8q73vUubrjhBnbccUdOOeUUVl999RY/EkmSpKXsYSRJkrQCTjvtNE4++WTmzJnDySefvKTn0PHHH8+jjz7KlClT+NrXvsb222/PiBEjWLRoEddeey3/8R//wXXXXcfqq6/OSSed1OJHIUmSVGdgJGnYiYj1I+LSiLg1Im6JiA9W7eMj4qKIuKv6u1bVHhHx1YiYHRE3RsQODbc1vVr+roiY3qrHJKl1zjzzTA488EAA3vrWty6Z9Hrs2LF873vf4/rrr+ess87in//8JxtvvDGTJ09m8uTJ7LzzzgAcdNBBXHvttS2rX5IkqSsGRpKGo0XARzJzS2AX4OiI2BI4Hrg4MzcFLq4uA+wLbFqdjgROgxIwAScAOwM7ASd0hEySho91112Xyy67DIBLLrmETTfdFChHQlu4cCEA3/nOd9h9990ZO3YskyZNYv311+eOO+4A4OKLL2bLLbdsTfGSJEndcA4jScNOZt4P3F+dfyIibgPWA/YHplWLnQnMBD5WtZ+VmQlcERHjImKdatmLMnMBQERcBOwDnD1gD0bSgDr00EOZOXMmDz30EJMnT+aTn/wk3/72t/ngBz/IokWLWHXVVTn99NMBuO2225g+fToRwVZbbcUZZ5yx5Ha+9rWv8fa3v52FCxey8cYb873vfa9VD0mSJKlLBkaShrWI2BDYHrgSmFiFSQAPABOr8+sBcxpWm1u1ddcuaYg6++yu8+BrrrnmRW277rord955Z5fLT5kyhVmzZvVrbZIkSf3JwEjSsBURY4DzgWMy8/GIWHJdZmZEZD/e15GU4WxMnDiRmTNn9tdNSwPO/19JkqSB0ez9rmnTpnV7nYGRpGEpIlamhEU/ysyfVc3zI2KdzLy/GnL2YNU+D1i/YfXJVds8lg5h62if2dX9ZebpwOkAU6dOzWW9MUuDnf+/kiRJA6OV+11Oei1p2InSlegM4LbM/HLDVTOAjiOdTQcuaGg/rDpa2i7AY9XQtd8Be0XEWtVk13tVbZIkSZLU1gyMJA1HuwHvAF4XEddXp9cDJwF7RsRdwL9WlwF+A9wNzAa+DfwnQDXZ9aeAq6vTiR0TYEuSNFwdfvjhTJgwga233rrW/rWvfY3NN9+crbbaiuOOOw6Aiy66iB133JFtttmGHXfckUsuuQSAJ554gilTpiw5rb322hxzzDED/VAkaViLctAfSdJAmTp1ajrZ7fDTOEdWu3PfQdKyXH755YwZM4bDDjuMm2++GYBLL72Uz3zmM/z6179mlVVW4cEHH2TChAlcd911TJw4kXXXXZebb76Zvffem3nz5r3oNnfccUdOPvlkdt9994F+OJLazFDa54IB2e/q9glzDiNJkjQsTZo0ifnz57e6jBU2ceJEHnjggVaXIS2x++67c++999baTjvtNI4//nhWWWUVACZMmADA9ttvv2SZrbbaimeeeYbnnntuyXIAd955Jw8++CCvfvWrm1+8JGkJh6RJkqRhaSiERTB0HoeGtjvvvJM//vGP7LzzzrzmNa/h6quvftEy559/PjvssEMtLAI455xzeNvb3jbkeg1I0mBnDyNJkiRJTbVo0SIWLFjAFVdcwdVXX83BBx/M3XffvSQEuuWWW/jYxz7G73//+xete8455/CDH/xgoEuWpGHPHkaSJEmSmmry5MkceOCBRAQ77bQTK620Eg899BAAc+fO5YADDuCss87i5S9/eW29G264gUWLFrHjjju2omxJGtYMjCRJkiQ11Zvf/GYuvfRSoAxPW7hwIWuvvTaPPvoob3jDGzjppJPYbbfdXrTe2WefzaGHHjrQ5UqSMDCSJEmS1I8OPfRQdt11V+644w4mT57MGWecweGHH87dd9/N1ltvzSGHHMKZZ55JRHDqqacye/ZsTjzxRKZMmcKUKVN48MEHl9zWeeedZ2AkSS0SHhpXkgbW1KlTc9asWa0uQwNsKE3WOlT2HdwmkiQNPUPp8x0G5DO+2yfMHkaSJEmSJEmqMTCSJEmSJElSjYGRJEmSJEmSagyMJEmSJEmSVGNgJEmSJEmSpJqRrS5AkiRJ0uA0adIk5s+f3+oyVtjEiRN54IEHWl2GJLUVexhJkiRJ6tJQCItg6DwOSRpIBkaSJEmSJEmqMTCSJEmSJElSjYGRJEmSJEmSagyMJEmSJEmSVGNgJEmSJEmSpBoDI0mSJEmSJNUYGEmSJEmSJKnGwEiSJEmSJEk1BkaSJEmSJEmqMTCSJEmSJElSjYGRJEmS2tbhhx/OhAkT2HrrrZe0LViwgD333JNNN92UPffck0ceeQSACy64gG233ZYpU6YwdepU/vSnPwFw6aWXMmXKlCWnVVddlV/84heteDiSJA0akZmtrkGShpWpU6fmrFmzWl2GBlhEtLqEfjNU9h3cJkPD5ZdfzpgxYzjssMO4+eabATjuuOMYP348xx9/PCeddBKPPPIIn//853nyySdZffXViQhuvPFGDj74YG6//fba7S1YsIBNNtmEuXPnMnr06FY8pEHF14mkdjOU3rdgQN67un3C7GEkSZKktrX77rszfvz4WtsFF1zA9OnTAZg+ffqS3kJjxoxZ8kXiqaee6vJLxU9/+lP23XdfwyJJ0rBnYCRJkqQhZf78+ayzzjoATJo0ifnz5y+57uc//zmbb745b3jDG/jud7/7onXPOeccDj300AGrVZKkwcrASJIkSUNWRNR6Eh1wwAHcfvvt/OIXv+B//ud/asvef//93HTTTey9994DXaYkSYOOgZEkSZKGlIkTJ3L//fcDJQSaMGHCi5bZfffdufvuu3nooYeWtJ133nkccMABrLzyygNWqyRJg5WBkSRJkoaU/fbbjzPPPBOAM888k/333x+A2bNnL5k89Nprr+W5557jJS95yZL1zj77bIejSZJUGdnqAiRJkqS+OvTQQ5k5cyYPPfQQkydP5pOf/CTHH388Bx98MGeccQYve9nLOO+88wA4//zzOeuss1h55ZVZbbXVOPfcc5cMV7v33nuZM2cOr3nNa1r5cCRJGjTCw0tK0sCaOnVqzpo1q9VlaIANpUO8DpV9B7eJtHy+TiS1m6H0vgUD8t7V7RPmkDRJkiRJkiTVGBhJkiRJkiSpxsBI0rAUEd+NiAcj4uaGtvERcVFE3FX9Xatqj4j4akTMjogbI2KHhnWmV8vfFRHTW/FYJEmSJKm/GRhJGq6+D+zTqe144OLM3BS4uLoMsC+waXU6EjgNSsAEnADsDOwEnNARMkmSJElSOzMwkjQsZeblwIJOzfsDZ1bnzwTe3NB+VhZXAOMiYh1gb+CizFyQmY8AF/HiEEqS1EOTJk0iItr+NGnSpFY/lZIkrTADI0laamJm3l+dfwCYWJ1fD5jTsNzcqq27dklSH8yfP7/VJfSLofI4JEnD28hWFyBJg1FmZkT02zEsI+JIynA2Jk6cyMyZM/vrpqUB5//v4OM2GXzcJoOP20RSO2r2e9e0adO6vS4y++37kCS1lYjYEPhVZm5dXb4DmJaZ91dDzmZm5mYR8a3q/NmNy3WcMvOoqr22XHemTp2as2bNatKj0mAVEa0uod8MlX0Ht8ng4zYZfNwmktrNUHrfggF57+r2CXNImiQtNQPoONLZdOCChvbDqqOl7QI8Vg1d+x2wV0SsVU12vVfVJkmSJEltzSFpkoaliDib0kNo7YiYSzna2UnAeRFxBHAfcHC1+G+A1wOzgaeBdwFk5oKI+BRwdbXciZnZeSJtSZIkSWo7DkmTpAHmkLThaSh1jx4q+w5uk8HHbTL4uE0ktZuh9L4FDkmTJEmSJEnSIGJgJEmSJEmSpBoDI0mSJEmSJNUYGEmSJEmSJKnGwEiSJEmSJEk1BkaSJEmSJEmqMTCSJEmSJElSjYGRJEmSJEmSagyMJEmSJEmSVGNgJEmSJEmSpBoDI0mSJEkaJp599ll22mkntttuO7baaitOOOEEAI444gi22247tt12Ww466CCefPJJAL785S+z5ZZbsu2227LHHntw3333tbJ8SQPIwEiSJEmSholVVlmFSy65hBtuuIHrr7+eCy+8kCuuuIKTTz6ZG264gRtvvJENNtiAU089FYDtt9+eWbNmceONN3LQQQdx3HHHtfgRSBooBkaSJEmSNExEBGPGjAHg+eef5/nnnyciGDt2LACZyTPPPENEAPDa176W0aNHA7DLLrswd+7c1hQ+hM2ZM4fXvva1bLnllmy11VaccsopANxwww3suuuubLPNNrzpTW/i8ccfB+Cqq65iypQpTJkyhe22246f//znrSxfQ5iBkSRJkiQNI4sXL2bKlClMmDCBPffck5133hmAd73rXUyaNInbb7+d97///S9a74wzzmDfffcd6HKHvJEjR/KlL32JW2+9lSuuuIKvf/3r3Hrrrbz73e/mpJNO4qabbuKAAw7g//7v/wDYeuutmTVr1pIeYkcddRSLFi1q8aPQUGRgJEmSJEnDyIgRI7j++uuZO3cuV111FTfffDMA3/ve9/jHP/7BFltswbnnnltb54c//CGzZs3i2GOPbUXJQ9o666zDDjvsAMAaa6zBFltswbx587jzzjvZfffdAdhzzz05//zzARg9ejQjR44EypxUHb3BpP5mYCRJkiRJw9C4ceN47Wtfy4UXXrikbcSIERxyyCFLwgmAP/zhD3zmM59hxowZrLLKKq0oddi49957ue6669h5553ZaqutuOCCCwD4yU9+wpw5c5Ysd+WVV7LVVluxzTbb8M1vfnNJgCT1JwMjSZIkSRom/vnPf/Loo48C8Mwzz3DRRRex2WabMXv2bKDMYTRjxgw233xzAK677jqOOuooZsyYwYQJE1pV9rDw5JNP8pa3vIWvfOUrjB07lu9+97t84xvfYMcdd+SJJ55g1KhRS5bdeeedueWWW7j66qv53Oc+x7PPPtvCyjVUGUNKkiRJ0jBx//33M336dBYvXswLL7zAwQcfzBve8AZe/epX8/jjj5OZbLfddpx22mkAHHvssTz55JO89a1vBWCDDTZgxowZrXwIQ9Lzzz/PW97yFt7+9rdz4IEHArD55pvz+9//HoA777yTX//61y9ab4sttmDMmDHcfPPNTJ06dUBr1tAXmdnqGiRpWJk6dWrOmjWr1WVogA2l+QWGyr6D22TwcZsMPm4Tqfkyk+nTpzN+/Hi+8pWvLGl/8MEHmTBhAi+88ALvfOc7mTZtGocffjj33HMP66+/PiNHjuS+++5j11135cYbb2Tttddu3YMYRIbS+xYMyHtXt0+YPYwkSZIkSWqRP//5z/zgBz9gm222YcqUKQB89rOf5a677uLrX/86AAceeCDvete7APjTn/7ESSedxMorr8xKK63EN77xDcMiNYU9jCRpgNnDaHgaSr92DZV9B7fJ4OM2GXzcJpLazVB634LW9jBy0mtJkiRJkiTVGBhJkiRJkiSpxsBIkiRJkiRJNQZGkqReu/DCC9lss83YZJNNOOmkk1pdjiRJw8KkSZOIiCFxmjRpUqufTknLYWAkSeqVxYsXc/TRR/Pb3/6WW2+9lbPPPptbb7211WVJkjTkzZ8/v9Ul9Juh8lgM8TSUGRhJknrlqquuYpNNNmHjjTdm1KhRHHLIIVxwwQWtLkuSJGnADZXgC4bWY1H/MDCSJPXKvHnzWH/99Zdcnjx5MvPmzWthRZIkSZL6m4GRJEmSJEmSagyMJEm9st566zFnzpwll+fOnct6663XwookSZIk9TcDI0lSr7zyla/krrvu4p577mHhwoWcc8457Lfffq0uS5IkSVI/GtnqAiRJ7WXkyJGceuqp7L333ixevJjDDz+crbbaqtVlSZIkSepHkZmtrkGShpWpU6fmrFmzWl2GBlhEtLqEfjNU9h3cJoOP22TwcZsMLkNpe4DbZDBymww+A7BNun3CHJImSZIkSZKkGgMjSVpBEbFPRNwREbMj4vhW1yNJkiRJK8rASJJWQESMAL4O7AtsCRwaEVu2tipJkiRJWjEGRpK0YnYCZmfm3Zm5EDgH2L/FNUmSJEnSCjEwkqQVsx4wp+Hy3KpNkiRJktrWyFYXIEnDQUQcCRxZXbwF2LqF5agFhsJRR4Yat8ng4zYZfNwmg4vbY/Bxmww+bpP+Y2AkSStmHrB+w+XJVVtNZp4OnD5QRUmSJEnSinBImiStmKuBTSNio4gYBRwCzGhxTZIkSZK0QuxhJEkrIDMXRcT7gN8BI4DvZuYtLS5LkiRJklZIOL5PkiRJkiRJjRySJkmSJEmSpBoDI0mSJEmSJNUYGEmSJEmSJKnGwEiSJEmSJEk1BkaSJEmSJEmqMTCSJEmSJElSjYGRJEmSJEmSagyMJEmSJEmSVGNgJEmSJEmSpBoDI0mSJEmSJNUYGEmSJEmSJKnGwEiSJEmSJEk1BkaSJEmSJEmqMTCSJEmSJElSjYGRJEmSJEmSagyMJEmSJEmSVGNgJEmSJEmSpBoDI0mSJEmSJNUYGEmSJEmSJKnGwEiSJEmSJEk1BkaSJEmSJEmqMTCSJEmSJElSjYGRJEmSJEmSagyMJEmSJEmSVGNgJEmSJEmSpBoDI0mSJEmSJNUYGEmSJEmSJKlm2AZGF1xwwYWtrgEGRx2DoQYYHHUMhhokSZIkSWq1YRsYAWu3uoDKYKhjMNQAg6OOwVCDJEmSJEktNZwDI0mSJEmSJHXBwEiSJEmSJEk1BkaSJEmSJEmqMTCSJEmSJElSjYGRJEmSJEmSaka2uoCeiIg1+vs2zz///JWacbvtWMdgqGGw1DEYapAkSZIkaSBk5hPdXWcPI0mSJEmSJNUYGEmSJEmSJKnGwEiSJEmSJEk1BkaSJEmSJEmqMTCSJEmSJElSjYGRJEmSJEmSagyMJEmSJEmSVDOy1QVouR5v9h3cc889A3I/7VDHYKihMrbVBUiSJEmShi97GEmSJEmSJKnGwEiSJEmSJEk1BkaSJEmSJEmqMTCSJEmSJElSjZNeS5L6YqXqtBjIFteiYkT1d3FLq1Ajt8ngsRlwMLA+sAi4EzgbmN/KogSU7yMvVCe1TgDTgL2BlwJPAlcB5wPPtq6sYW0k8HrgNcA44FHgMuA3lPcxDbxVgQOBnYExwD+B3wEzGaL7wwZGkqSeWglYCxgPjKrakrJTuaD6q4E1krJN1gJWrtoSeIyyTZ5pUV3D2SjKa2QcSwOjFyg7+guA51pS1fD1L8DngM0pr5GOfd9ngf8GLgE+DtzdkuqGr9Upr5M1KEEFwELgkepkyDqw3gUcT9keq7N0m7wV+CJwFvBJ4OmWVDf8jAA+AryvOr9Gw3WHUF4fpwJfwtfKQBkN/C8wnbKfNaZqT+Awyj7wScB3W1JdExkYSZJ6YnXKL/MjOrUHZUdmDcqH5d8Zor+wDEJrAuuxdMe+Q1DCinGUkGLeQBY1zL0EmNRF+0qUL8fjKb9GPjiQRQ1jBwNfA1br4rpVq797A7sBbwKuG6C6hrMAJgNju7huFDCR0rtlLvDEANY1nH2D0mNidBfXdXwpPhx4HbAXJdBT84wCfgrsRNfbpCM8+jCwO2XbLRyY0oatccDvgQ1Z+tnRISivkzHAZ4FXAv8xgLU1nXMY9d1/AFdSumr+Z6fr3g88Ttkx7PAF4Hrgr8B2Vds2H/nIR15R3cZfKS94SRpsVgNexovDos7GVMup+cZSvnR1Dos6G0cJldR83YVFnb20Oqm5dqf7sKjRSpTX0y8pryk11/p0HRY1WqlabvXmlzPs/S/dh0WNVgE2An6B3x+b7duU4U7L2yajKeHEd5pe0fC2EuX/fmNeHBZ1Nho4ADihyTUNKF/wfbMF8E7KON9dgX0o/0RQdsxfB8xpWH4v4OXAFOADwMlV+zMf+MAH7qMkyAcAn6f8YixJg0lPgokOHcMM1DwBrNuL5cex9FdiNcdISs+InprA0mGdao4vsPywqNFqlCEgap5x1IfWLEtg2N1s4ylDnpYXTHQYBWxK6ZWn5tiC8r2yp+9dq1G2x5ZNq0h7Aa+g55/Zo4GjGUL7wgZGfbMZMIsyN8Ri4E/AftV1JwH/Q31IxhsoEysCXE0JhSYCs1/2spd1zGXwAKWb+tpNrVzDzaPAn4G/AH+k/GKxLBtQes6tqNcDH+rDeh+nhKoranvKl4WeuITyHN0K3FOd/zPldf4T4BpKL8BPdqrzjmq5ayghcE8DlXazOr3/YjtkPiQHqXEsv7dXZ26T5hpP798D3CbNsy1l6EBvrAwcSs+/PKv3evs/vzI9D5jUe/9O74eQjwGO6f9SVDmapfMR9tTK1XpqjmPo/Y9uCbyj/0tpDQOjvrmV0rNoPCXZ3YvyK8jrgX8AN3dafh3KWOwO/+DFvw7vSPlS5qSL6k/PUOZm+BfgE9VpIPyGpT3pWuE64LgeLvs6ynP0GcqRQHarTn8Hvkp5be4G7ALs2bDe16v2qcBWwKv6o/BBaFwf1lmF3v2yr94Z14d11qD3IZN6btwAraOe+XfK+1BvLab8uq/+N4q+fS6s1d+FaIkj6Ns22QF/4G6Wg+n9HMMjq/XU/8ZT9vN7azXg3f1cS8sYGPXNnZQvw78Afg7cRNkx+SjlS2dvTQROp8yL5GSxapY1KD2OoPQa+SWl19EVlLCzw0jKeOhZwA9YujMxBfgtcDnl/75j+MV7KT3n/gp8r2p7O+WoGgD7Unrx/AmYwfLn7tga+AMl9Jne0P4pSu+nK1g639f3qHeNPg3YnxLenFe17cbSXkN/oue/EjxDeX4AnqfMQdbVMKBRlNf/o9XlbYCLKc/Hj2n/L4V9HTbT21/I1HN9fW7dJs3Tl+d2BO6HNUtP5lzryih6Ng+Ves/PksGnr6HPQsqwWvWv0fT9/30EzvnVDBPp+5FNh0yo6o5K3/2AMqHiPpQvirdRuj//hdLDaD3Kl9MJwP3UJ1Jcl9LLiCeffHIlykz4J1K+dEv9aTWWDps6lTJPFpTDCf8b8GpKWPTZhnU2pQRGUymTt7+HEiL9H+VX290p////Wy3/YUoosytdd1P+K6UXz6so/+tdLdNoa+CNwB6UQ7xOogz53La6jzcBn6a8iZ9Pmf8LyofsNOB3nW7vA5R5KXajhEt9Ocz4mpTg67KGtqMpz+1dwGxKcAwl/P3fqtZbqsfQzgyxJbWbF1q0rrrnZ8ngsyLbxNdJ/1uR5zRWcH11zc8SDIxWREdqOJnyZfbHlImvt65O8yhfkB+kDM85tFr+lZQv4fOBlT/zmc9sTJnf6IIBq1zDSceQtB0pvXJOr9qDMoP/Xyk9jdZl6a9Fcym9eADOpQQfm1Im1JtBCUmOY+lklDcDZwBvAxZ1UcO6lN54VwAfpEzotyy/pgRaCyi9mXasavgp5c33n5QwdkfgIkqANYoyXOzP1bqNrqAEYu+lBD+Ll3P/nY0Avgt8E7i3ob1jSNrGlF913kI58suaVR1Q3hd26+X9DTZ9PVRrX3+R0fL1ZZtkH9dTz/TluV3EENqhHGRmU3qG9tZC6lMIqP/4WTL4PNDH9UZRvseofz1L337UXNF11b0H6Nvw5o51hwQDo777EaVH0HmUHhaPLWPZ31G+aN5AOcTrh6v2A2fPnr0GZfhOx5CZbZpUr3QV5bDPa1PCnbUpPYx2owSbHYeK7PyLU1ICpttYOr/PLsCbq+sPogRR21F64HQeBvDF6vpdKIHR8g5J2dX9d+c5yrCxf6UENj/rYpmTKUcBWZUSMG26nPvv7KvA34BvdHP9oup22z0Y6s4jfVjnGdzJb6a+bJPHMZxopr5sk76so545i97/ONDh9/1ZiJZ4HniqD+v5Omme0+j9NknKfpfbpTnOovfh6sJqPfW/xyjfbXrbG+8pyutrSDAw6ru9Kb2F/oX6MJUOW1N6SHT4COUL9S6UuVkAzv3JT35yHUu/hO/G0mEtUn/blBLmLKD0hPknJex4NeXoaB3WB3aqzh9M6YV0FyVg6mgfCWxOCZImU3Ye/re63c5zBI2lGoJJGQa3PG+gpPnjq9qupQz1PJDynvUSymtlVrX8zyhD5f6FEtx0thFlovqvVLf1ih7U0OF/KD2GPrac5XahHGHtccoQ1V2r9kMovaHa2TO8uNfW8ixY/iJaAY/RdW++ZXGbNNcj9G6HMvELVzPdxYsPQLI8Cym9Se2J1zy9fR9aSN9CJvXMufT+u+DTwClNqEXFt+j9jzsvUHrBqzlOofzf98YIyutrSDAwkoa2jjmM/gycCRxF+WA5l3Lo+SsowyXvbFjnLsq8RbMoEzZ/h/LL4Dsoc239pTrtTHlD/E51O3+mpOmde9t9lvLLx+XAwz2o+WbKsLSLKXMuPUAZNnczJbz6NSXIebBa/mJKgHQpXQ9B+E/KZNl/ra7vCJX+3MWyjdYFjqUEY3+qlj+s4fqOOYyuojwP367aj6LMsfRXyrxLn6f9zaXnv9Z3hGZqrrn0PKB4iN7v7Kh3FrM0GO+JB+jbkCn13Ifp+f/9C5T3ra80qxgBvft8eAGY07xSBDxJ+bGvp6+TZyn7e139UK7+cQ9lf72n2+Rpyryi9zStIl1O2afv6Y+nT1Om/XiiaRUNsN4etk9SexnXTfsCyjCuruzYTftNdH244b26aPtRdYIyh9dvurnNzj63jOv+pzp1tohyRJxGf2Jpz55ju7m9zkPIGmuG8uVvbDfrfo7ua72JMmH3UPIcZWfkZSz7CB6PUuZvU/M9BdxH6RG4rKNB/ZOl4aqa69Hq77qU3pddSUpYZI+v5rue0qv1x5Qhyd39SPo8pbfX3vTsRw2tmHmUMGj8MpZZBPyd3vduVe99i9KD/P2Uo3R152nKCIme9BTXijmO0rt9P5a/TX5F9/u56j9vpxxoZ0eWHj26K09TDjI0ZIajgT2MJEk98xylJ9pcygdiR++WRZQvv7MxLBpoTwF3UI7E+SxLt8nzlF5Fd2JYNNAeBW6nhEILKdukY8Lx+ZTtZVg0cC6hHD3zV5T3sI73rhcovSuepgxD25UyV50Gxv2U3swPs3R4bVKGQM+jvE6cwHfgfAZ4JyUQ6jwH4ZOU97NPUY5S63ZpvgSOBD5EeV96inqP1KeAuym9KN+N8xMOhGco//8nUl4PTzZc91x1/fXAuyijDIYUexhJaoV/B/6jU9sVlLm+NLg9xrIn+dfASkoAYQgxeLxA+SJsb5XB4XbKZ87alB3+iSztwfJL/ALcKgspX7yGzJGE2tyF1WlzSi/pjl7U/04Z8t/bSX+14s6uTq+khNodQcT+lOkQNLAWU46Q/A3KDxEdRzj/JPAHymfNkGRgJKkVflidJEkaCA8B32t1EdIgd3t16giMLmlhLSqurk4dgZFhUWslJUTtcGqrChkoDkmTJEmSJElSjYGRJEmSJEmSagyMJEmSJEmSVGNgJEmSJEmSpBonvR78xjb7DjbaaKPLgNc0+37aoY7BUIMkSZIkSa1mDyNJkiRJkiTVGBhJkiRJkiSpxsBIkiRJkiRJNQZGkiRJkiRJqjEwkiRJkiRJUo2BkSRJkiRJkmoMjCRJkiRJklQzstUF9ERmPtHft3nBBRe80Izbbcc6BkMNg6WOwVCDJEmS1CoRATTnO5j6xm0yuAyn7WEPI0mSJEmSJNUYGEmSJEmSJKnGwEiSJEmSJEk1BkaSJEmSJEmqMTCSJEmSJElSjYGRJEmSJEmSagyMJEmSJEmSVGNgJEmSJEmSpBoDI0mSJEmSJNUYGEmSJEmSJKnGwEiSJEmSJEk1BkaSJEmSJEmqMTCSJEmSJElSjYGRJEmSJEmSagyMJEmSJEmSVGNgJEmSJEmSpBoDI0mSJEmSJNUYGPVCRGR/n2bPnr1jM2633WoYLHUMhhoiIlv9vy5JkiRJGt4MjCRJkiRJklRjYCRJkiRJkqQaAyNJkiRJkiTVGBhJkiRJkiSpZmSrC5AktZeICGAU5UeHRZn5fItLGvYiYiXKNoGyTRa1sp7hLiJ2AN4DbAoEcA/wHeDKzPTABi0SESNZuu+7MDNfaGU9w1lEjADeAPwbMAl4FrgW+FZm3tfK2oa7iBjVcD58z2qdiFgD+Hdg34a2/wR+mJmPt6ywYSwiXgoc0XD5Z8DPgJ9m5rMtK6yJDIwkST1S7US+BBgPjGhofxp4GHjUHcuBFRGrAmsD42joNRwRTwIPuUM5sCLi1cA3gQ2BVVj6OnkBOAT4R0QcnZkXtabC4SkixlLeu9ZoaH4hIh6jvE6eaU1lw0/1g8N7gRMpr5HGbTIN+FBE/Bl4T2beM/AVDk/Vjw5rUV4nqzZctUVELAAe9sehgRMRqwFfBqZTPj9Wb7j6C8AXI+Is4EO+fw2MiHgJ5fP9TZRt0uEA4F+Bb0TEKcAnMnNxC0psGgMjSdJyRcSawAaU3hKdja5OL4mIe+3dMjAiYm1g3W6uHgOMiYjHgb/bk6L5IuIA4EfAal1cvRJlh39T4IKIOCozfzCQ9Q1HVTixPiVQ7azjC/JaEfFAZj44kLUNR9X2+BrwTupfgDusUv2dBlwbEa/OzJsHprrhKyJWBjaiHhR1GAlMANaOiPsy84kBLW4YiogxwB+Bzej686TjtXMYsEv1OnG7NFFErAtcSXktjOpikY7g+8PA1Ih401DaF26bOYwi4rsR8WBE3NzQ9taIuCUiXoiIqQ3tG0bEMxFxfXX6ZsN1MyPijqOPPnrL6roJDdcdHBG3Vrf544F7dJI0eFVdorsLixqNBjaqfqlUE1W/dHUXFjUaC2xQfVFTk1T7ID+k6537zlYDvhkRuze3KtF9WNTZpGqYgZrrGLoPixqNANYELq3e69Qk1dDAjek6LGq0ErBhRIxuflXD3s+BzVn+58lqlFDpF80uaDiretdfAkyk67Co0Whgd+Abza5rILXTTv33gX06td0MHAhc3sXyf8vMKdXpvZ2ue/vXv/71W6vrHgSIiE2BjwO7ZeZWlA81SRKsx/LDog6rUbq0q0mqHfx1erHK2Oqk5vkMPQuLOoymDCtQk1S/0o/rxSqTqjmO1AQRsQrwSZYfFi1ZpVr2qKYVJSg9JlZZ7lJFUPYH1CQR8UpgV5Yf4HVYFdi5Wk/N8WbK//3KPVx+NHBY1StpSGibwCgzLwcWdGq7LTPv6Ke7eA/w9cx8pLptuyarZSLi5oj4SccvOdV8JK2u6fsRcVB1/jsRsWUT7+utEXFbRFy6jGU2bOxx2M0yTa1zOKjm/ljeLyqdGRg111r0/vPbbdIkEbE+5RfF3vbi2jYiNmtCSSrW7uXyQZmfTc3x1j6ssxplTqMRy11SvVb1PF2rl6utZi+jpvoIPQ+LOqwGfLQJtaj4GGWYf28kZa62IaFtAqM+2CgirouIy6pJKBt9rxqS9j8N3fRfAbwiIv4cEVdEROfeTNKAycytgYUM0jebzHx3Zt7axLs4gjLh5WtX5Eb6q85h/qvzuD6sMyoievorsnpvXB/WGVPNU6H+d2Af11sZeFt/FqKiChjWWO6CL9bbL8/quSPo2zZZBdi5n2tRsQZ9m8/W10kTVN9JD6DhoCI9tBJwgEPP+19ETAS26sOqqwLv6udyWmaoBkb3Axtk5vaUyad+XP1KDvD2zNzm85///O3Aq4F3VO0jKZNRTgMOBb4dEeMGtGqp7o/AJo0NETEmIi6OiGsj4qaI2L9qXz0ifh0RN1S9k95Wtd8bEZ+r5uuaFRE7RMTvIuJvEfHeapmIiP+r1rupYd2IiFMj4o6I+AOl23JHHTM75g1r7P0UEQdFxPer89+PiNOqAPbuiJgWZS6y2zqW6UpE/C/wKuCMqq4NI+KP1WO+NiL+pYt1RkTEF6vHcGNEvL+rOiPi5ChzlF3cMV9FRLwnIq6unrvzG3p1fT8ivhkRVwJfiIhPRMRHG+7z5qq2DSPi9mr5OyPiRxHxr1X4fFdE7NTD7T1Y9TVkGM4hW7O5TQaXdej9L8JQtsfkfq5FxUh63+OrYz01R2+G0TZ6gYb9D/UrP0sGl9VZse/mve0Fo+WbADzXx3WHTLA6JAOjzHwuMx+uzl8D/I3Sg4jMnAcwZsyYF4AfAx1f5uYCMzLz+eownndSAiRpwFU9WvYFbup01bPAAZm5A/Ba4EvVLwr7AP/IzO2q3kkXNqzz98ycQgmgvg8cBOxCmUsAyq/jU4DtKIeF/L+IWIfyK8dmwJaUIzG8KKjpgbUoY7E/BMwATqYk9dtExJSuVsjME4FZlHD3WOBBYM/qMb8N+GoXqx1JOYz1lMzclnKkos5WB2ZVc5RdBpxQtf8sM1+ZmdsBt1F+Be0wGfiXzPzwch7nJsCXKJMUbg78GyX0+ijwX8tZd7DLAV5Py+c2GVz6ujO5ouuqe/6vDz4rckh2D+feHH6WDC6L6Pt385Wq9dW/nqdvPz7AENoeQzIwioiXdox3joiNKcHP3RExMsphiHn++ecDeCNl4mwoM8xPq9ZZmxIw3T2wlUtLzAL+DpzRqT2Az0bEjcAfKJOwTaQES3tGxOejHF7zsYZ1ZlR/bwKuzMwnMvOfwHNVL7pXAWdn5uLMnE8JU15JmZOjo/0flCME9NYvMzOr+56fmTdVh/e+hRLw9MTKlB5/NwE/oQRYnf0r8K2OQ1hm5oIulnkBOLc6/0PK4wbYuurBdBPwdupdT3+SmYt7UOM9nR7bxQ2Pe8MerD+Y9fULrV+Em6cvz+0L+KWrWf4G9GWeuaeB/pqHUXXPAz157+7s2f4uREvcTt+ChpVxf7xZ/HwfRDLzWeCx5S7YtScy85n+rEcAzKP383h2+Ht/FtJKbRMYRcTZwF+BzSJibkQcEREHRMRcSg+GX0fE76rFdwdujIjrgZ8C762+QK4C/C4ibnz/+9+/JeWf4NvVOr8DHo6IW4FLgWM7eilJA606gt/7M3Nhp6veDrwU2LHqNTQfWDUz7wR2oAQUn66GdXXo+GB/gfqH/Av0T7fixh3AzsMy+uO+P0R5nNsBU+n7G3dnHXV/H3hfZm5D6XXV+Bieajjf+ZefxuU6P7bGx93uXbf78j74VGa6Q9k8fdkmj/Uw/FTvnU/f9qdWovR0Vj+rAvtH+rBqVz82qH98jfpnak/9LTNv6+9iBJn5FL0PSRNfJ830DXofyD0LnNaEWoa9zHwC+CVlf743ngS+3P8VtUbbBEaZeWhmrpOZK2fm5Mw8IzN/Xp1fJTMnZube1bLnZ+ZW1ZfuHTLzl1X7U5m5Y2Zu+81vfvOWzPxgxw50Fh/OzC0zc5vMPKeVj1fqxprAg5n5fES8FngZQJRDNz6dmT8E/o8SHvXUH4G3VfMAvZQSuF4FXN7Qvg5lCFxX5kfEFhGxEmUYW39bE7i/6r3zDrqeDPAi4KhqKB8R0dWRblaiDMeDMmTsT9X5NYD7o0wI/PZl1HEv1fMaETsAG/XuYbSn6hev3u7kP9SMWrTEE5RJ8XvDbdIk1Zeus+hdD67FwC+66Q2p/vEwvevRsgh4tDmliLKv0dv3oaeAzzehFi3V2x8gHs9Me6s2z2n0rSfeN/q7EC3xJaC3vbeCpaMa2l7bBEaSgDI3z9Rq+NRhlC7eANsAV1W96k4APt2L2/w5cCNwA2XY2XGZ+UDVfhdwK+XL0F+7Wf944FfAXygTzve3bwDTI+IGyvxAXYUX36F0/byxWu7fuljmKWCniLgZeB1wYtX+P8CVwJ9Z+nx25XxgfETcAryPMs/ZcDGHnn8ZXtBpSKT6WdV74j56/ovX/XZVb7oTKF+8erJNkhJMHNvMgoa7qpfjP3q6OHBf9dpSE1TP7WH0/IvXc8B1wHlNK0pUoyl6+pm9kDI6Q01SzbX7WXr+Q91TwOcyc27zqhr2rgR+RhlG3hNPA0cNpf2udh8qIQ0LmTmm+vsQZQhmZ/dShlV2Xm/DhvPfpwy/etF1lC8utS8v1c7d+7qpZ1rD+Z9Shn52XuadDefvBbbu6roe3P5dwLYNV3+s821Wcxd9uDp1eTvV5RdNXp2Zp9FFV97ONVZv/Ht1U3KXj63z425XmbkwIv5GmY+pu6NBJfBQZjYjNFQnmflMtU1eRvfDNF8AHqjeN9REmflgRLyK0jNzPN2/Tp6j9BB7jTv4zZeZD0dEAuvS/Y+kz1MODtGX4VLqhcz8Y0QcTPnlfRW6P3z405QfsV5vb5YB8XfKa+Qly1jmaUqoOmQm8h3EPk054tnRlAO2dOcpyv7rpwaiqOEqMzMijqB8rr+e7rdJUgLxYzOzq4PvtC17GEmSliszF1ZzZd1N+TVyIWUIx7OUOaZuNywaWFWIeQclMH6C8sV3EWWH5R/AbYZFAycz/0YJt/+PMn/O45QvWU9X5x8HvgJsnZm3tqjMYaca9ncbpWfE05TXyfOUOSb+TnnvMiwaIJn5K8qBNX5M+fx4ouHqJ4F7KEcYnVbNH6Imq6blmEfpZf0g9Tl0HgVmZ+Zsw7uBUW2PjwFvoQzlfJby2nim+vts1X5QZh5rz8jmq/733wa8hzIqo7G30ROUbfJr4F8zc8gND7SHkaSWiYgrKb8yNnpHZt7U3/fV0UtLKyYzn6RvR4RSE1Q7ih1hhFqsGt7xvxHxKWAPYIPqqn8Av88XH8hAA6Car/Jh+jZhvPpZFZgeFhEfoPTc7ZjrYy/gCr8At0b1/vQA8EBEdLQNmSM9tZvM/B3lYE0vB/4FGEsJJ/5c/UChAVS9L50NnB0R21J6QQK8F7ikms5jSDIwktQymblzq2uQpP5W/Rp5YavrkAazzHwUOC8izq0udzdXojRsVeGQAdEgkpk3NoSqQ/5opw5JkyRJkiRJUo2BkSRJkiRJkmoMjCRJkiRJklRjYCRJkiRJkqQaA6NeyMzo79Mmm2xyTTNut91qGCx1DIYaMjNa/b8uSZIkSRreDIwkSZIkSZJUY2AkSZIkSZKkGgMjSZIkSZIk1RgYSZIkSZIkqcbASJIkSZIkSTUGRpIkSZIkSaoxMJIkSZIkSVKNgZEkSZIkSZJqDIwkSZIkSZJUY2AkSZIkSZKkGgMjSZIkSZIk1RgYSZIkSZIkqcbASJIkSZIkSTUGRpIkSZIkSaoxMJIkSZIkSVKNgZEkSZIkSZJqDIwkSZIkSZJUM7LVBfRERKzR37d5/vnnr9SM223HOgZDDYOljsFQgyRJ0nDj/tfg4zaRlm2ovEYy84nurrOHkSRJkiRJkmoMjCRJkiRJklRjYCRJkiRJkqQaAyNJkiRJkiTVGBhJkiRJkiSpxsBIkiRJkiRJNQZGkiRJkiRJqhnZ6gJ66PH+vsF77rmnKbfbjnUMhhoGSx2DoYbK2FYXIEmSJEkavuxhJEmSJEmSpBoDI0mSJEmSJNUYGEmSJEmSJKnGwEiSJEmSJEk1BkaSpN5aCVgdGAOs2uJaBLsA3wf+CswCfgO8DVilhTUJVqa8TlYHRrW4FhWrUN63Vqd9DvwyVI0GpgO/b2j7FjClJdUIYC3gA8AlDW1fAjZpTTlqMJry3jW61YWITYAvNly+lPK6Was15TSfH5aSpJ5aFXgJsCYQDe0LgQXAI8ALLahruNqF8gVrArAaS38EegXlS9fJwOeBU1pR3DA2FhhPCSUaPU15nTw24BVpLco26RxwPwk8XP3VwAjgf4D/BJL66+RgYH/gPuDdwE0DXt3wNAr4MuX5T8rnSYd3Au8ArgcOB+YOcG3D2cqU9621gBEN7Ysp+1sLgOdbUNdwNRk4A9ie+vbYEdgS+H/AT4APA88NeHVNZGAkSeqJ8cAk6kFRh1HVdeOBe3EHZiDsCfyQ+o59ozHV348DGwHHDEBNgvWAcd1cN7o6jQXmDFRBw1wAL+PF4V2HMdVpAXD/QBU1jAVwFuX9q6ueEiOq9s2Biyjh0ZUDVt3wtCrwW8oX3q56DK9cnV4J/Bl4HfC3Aatu+BoNbEA9mOgwAlibEiTNAZ4awLqGq5dTet6Npett0rEvdhCwFbAP8OzAlNZ8yx2SFhHfjYgHI+LmLq77SERkRKxdXd48Iv4aEc9FxEcblls/Ii6NiFsj4paI+GDDdeMj4qKIuKv6O2S7c0lSm1oTWIeuw6JGo4ANcbhzs20K/IDuw6JGo4FDgKOaWpEA1qX7sKjRWMovlWq+Deg+LGo0ntJTT831/+g+LGoU1TI/o/wYoeb5JiUsWt7nyQjKvsBvcSh6s61CCbq7CiYajaC8xzn8vLlWpQz1X5Plb5PVgC2A05td1EDqyU799ykpWU1ErA/sBfy9oXkBZQzfFzstvgj4SGZuSelCf3REbFlddzxwcWZuClxcXZYkDR4Te7HsKMqwNTXPMfRuB3E08F8sf0dHfbcKvZu/YE16Fvip79ZgaU+7nlgbe94302jgaHo3B8vKGHY30wbAG+j5e9FKlNfVW5pWkaDsc/X0h7eV6N0+mnrvQMoPPT3dJqtRspMNmlbRAFvuA8/MyylBUGcnA8dRxrp2LPtgZl5Np+EImXl/Zl5bnX8CuI3SbRtKd9Mzq/NnAm/u3UOQ1EvjKd2K/wzMBu5ouLxyE+5vA4ZHl/KbKc8tlK70UH4Z/UFryuk3Y+n9/4U9RZtnDUqX596GPyvTxY8/6jfjl7/IixisNldvt0n0YR313Ftp+M7QQ6sC76E5+yYqz21vrQ58qL8L0RIrUz7ne2MNfI0004foWU/VRisBRzahlpbo07CBiNgfmJeZN/Rh3Q0pk0V1fIGcmJkd48YfwJRUarYFwG7V6Qzg6w2XnXumf+xZ/X2AMllkO1uzD+usjEfyaJZplF67vTUGOLR/S1GDvrxOxvZ7FeqwEr3rXdShL9tRPfN2ev+lq8Mr+7MQLXEwfRvOtAFlCK76X18/F/w8aY5JlKkWemsU5fU1JPQ6MIqIjq7t/9uHdccA5wPHZObjna/PzKT3vz5IWnGvAf4EXEEJkEYBO1AO070K5cv/VZRxuWsBZ1fXXUyZ3A3K5LqnA38ArqMcLrezDYDfAX+sTjtX7ROBCym9nK4Edq3a7wc+Vd33DMqRCH4D3Ai8fjm32ZWgHAnkGuAC4KeUXo7dPQdQeg79V3XbV1Dmj4HyS/QvqtpOpT6/T0cIPhR6V/V1iIa/djXHePo+tMw5WpqnL9skcL6vZunr+5ZD0ppn7RVY155fzdHXkOF53CbN4nvX4PISypGA+2LIhHh92VF5OeWIKzdExL2UiRuvjYhlTkoXEStTwqIfZebPGq6aHxHrVMusAzzYh5ok9d0qlEkP30mZY2wk5XC211LCmf+hhDbnUIaT/hdwAyXU+ST1id22Bt4I7EGZj6zz+8I/gf2AV1f394Wq/WBK0LRbdbsdh9JdHbgc2Ily2OP/qdb/N+C/l3ObXdmPEuJMpXTF3mk5z0GHh6vb/w7QMWn/8ZTQbCfgl8D6y7jfdtbXEP+Ffq1CHRbS920yZI7YMYT4I1lz+L41+PT1MNNJ37+wadn62qs8GGKHDR9E+vre5WdJcyxk+Qd86c6QGbXR68AoM2/KzAmZuWFmbgjMBXbIzAe6WycigjL05bbM/HKnq2ewtCfCdMqv/pIGzgjgPsp8RgA/pgQ3ACdRDqG6PfCVqm1XSngEJcwZz9Lx1r+mfCldUF23Y6f7Whn4GqWnzlmUQ+dC6fHz75ReSltRwiEob9Qd8wHdQukBtKg63zGZXHe32ZVdKb2CkhJO/7Fq33QZzwGU9ymA6xvudzfg3Or874BHlnG/7eyZPq5nONEcd/ZxvYWUoFfN0Zf/9xUJ/7Rsz9O3oZt+CW6em4DFfVhvFeCufq5Fxd/6uN5I4B/9WYiW6Ou+k/tczTGPvvfe6uvra9BZbmAUER1DTzaLiLkRccQylp0UEXOBDwP/r1p+LOWL1TuA10XE9dWpYzjJScCeEXEX8K/VZUmDw3hKL5816NlhVDt/+el8+WhKULMrsDtLh339hTIh7z8oPX065lppTOdfYOmvjMnSN/DubrM/dXyJWMzw6/bb1UEPludJhtAvK4PMNcD8Pqz3AuWHGzVHX14nfVlHPdeXEN9t0jyn0bdA7ibgnn6uRcUpLP2BrqcWAz8Dnur/cgQ8Tu/3nxZV66n/PU0ZIdXbHyCepLy+hoSeHCXt0MxcJzNXzszJmXlGp+s3zMyHqvMPVMuMzcxx1fnHM/NPmRmZuW1mTqlOv6nWeTgz98jMTTPzXzPTD2tpYC2m9JrZuLp8CKUnD5Q3u09TetKcWLX9haUTub2KMlzrieryGyi/Bo6nDOG6ttN9jaV82U1KKNQx78f6lNDnzOq0XS/q7+42u3IFZVhaAC+t6ofy62V3z0F3/kw56guUSa6H6pHBnmfp9u2ph5tRiJY4mbIT01MvUIKme5tSjQAepXc7lC8wdHslDhaP0LseXAvxS1czXQfMoXfb5CnK+52a41f0vtfXQso8j2qe3n42+N25ub5O70O8xZTX15DgZIuSngP+gzKc6wqW9kQ4lPIF6CeUiaJ3oPTg+RxliNpfKXMYHdVwWzdThqVdDHyecpSwRt+hzD/0F+AVLP2F6tXV7f0JeAvll8ie6u42u3IBpRfTrGq9GyhfELp7DpblJErvyasoIdScXtTcbubR8zkk/knvf7FU7/yIMvF7T36tT8r2OLqpFSkp7wE9mQOnN8uq756nvHf1xGLg702sRcXh9Dzsfga4jLJPoeZYBLyLng89fxr4LkvnmVRz9GY/6slqeTXPzZTvDL157zqCIdTTfrgNrZBU97mG86/qdN3Z1QnKF5vXNVzX3eG5b6EeIEHZCe84ctnfWHoENFh6tMUfV6fO1umm1sbrurvNriRlsuynKL2gLq1qhrJj2vk5gDKRd4frWHp0tgXAm7u5n47aGh97O1sM3E3pCdbdYZEXU3qJ+UtX8y0CDqQc5W87ut8mz1J2XN5I2X5qrqcpvbgm0/3Q2IWU0NrhHAPjMcrn1zp0f+TG5ygBnvMXNd9NwAGUIR6j6P6Q7k9R5kE8DOf5arY/UA4Ccjplm3T13TApnyU/pBz4RM13H+V9ay26nnQ5KT1bnUtqYPw/ymfIYcBqdL1NFlE+448Cfj9wpTWfgZGk4eYnwJqUHaMv4JEZe2ox5ctwx5DD1Si9VDvGzvd2+IdWzJOUIaAHAsdQjmDaMSQqKL9snUb5VcwhggPnGcoQ1zGUHf2O4GghZee+t8M7teKeqE5jgXEsDY6eowTcvRneqRV3BeVIpUdRjkYalM+OpLxebqYcZONX+JkyUGYAtwLvowzJ7/gsScr8lX+iDA28vCXVDV/3U3oPrUWZS3QlSgD+BGWfqy8T+6tvEjiO8r70IcoPzM+yNDgaSZm+41SG4CT9BkaS+kvnHkCttCXw7U5tz1F6Sb3+xYurF56j7MSo9RZTAtCfUIZjbkz5MvxP4Gr6dkQi9Y8ncWjmYPM4zlE0WNwPfAL4DLAT5UeIhZQvWvaGbI3ZlB8f/gt4JeWHtWcoAZ6f+a2ziPKZ7rCzweHy6rQOZQTCapSerLMYwj2HDYwkDUW3UuYXkoaLO6uTJLWL5ykHkNDg8TRliL6k7t3PMApSnfRakiRJkiRJNQZGkiRJkiRJqjEwkiRJkiRJUo2BkSRJkiRJkmraZdLrsf19gxtttNFlwGv6+3bbsY7BUMNgqWMw1CBJkiRJUqvZw0iSJEmSJEk1BkaSJEmSJEmqMTCSJEmSJElSjYGRJEmSJEmSagyMJEmSJEmSVGNgJEmSJEmSpBoDI0mSJEmSJNWMbHUBPZGZT/T3bV5wwQUvNON227GOwVDDYKljMNQgSZIkSVKr2cNIkiRJkiRJNQZGkiRJkiRJqjEwkiRJkiRJUo2BkSRJkiRJkmoMjCRJkiRJklRjYCRJkiRJkqSayMxW17BcEXEhsHY/3+zawEP9fJt9MRjqGAw1wOCoYzDUALBqZm7d6iIkSZIkScNTWwRGzRARszJzqnUMjhoGSx2DoYbBVIckSZIkaXhySJokSZIkSZJqDIwkSZIkSZJUM5wDo9NbXUBlMNQxGGqAwVHHYKgBBk8dkiRJkqRhaNjOYSRJkiRJkqSuDeceRpIkSZIkSepC2wdGEfHdiHgwIm5uaPu/iLg9Im6MiJ9HxLiqfcOIeCYirq9O32xYZ8eIuCkiZkfEVyMiqvbxEXFRRNxV/V2rhzV0uV4UX63u58aI2KFhnenV8ndFxPTl1bac52Wzhsd5fUQ8HhHHRMQnImJeQ/vrG9b5eHUfd0TE3g3t+1RtsyPi+B5tmKXr3lvVfn1EzOrv56YH979+RFwaEbdGxC0R8cGqfUCfh17U2/T7kCRJkiRpedp+SFpE7A48CZyVmVtXbXsBl2Tmooj4PEBmfiwiNgR+1bFcp9u5CvgAcCXwG+CrmfnbiPgCsCAzT6q+wK+VmR/rQQ1drlcFE+8HXg/sDJySmTtHxHhgFjAVSOAaYMfMfKS72nrxHI0A5lX39y7gycz8YqdltgTOBnYC1gX+ALyiuvpOYE9gLnA1cGhm3trD+74XmJqZDzW09dtz04P7XwdYJzOvjYg1qnXfDBw8kM9DT1Tbqan3IUmSJElST7R9D6PMvBxY0Knt95m5qLp4BTB5WbdRhQpjM/OKLAnaWZRQAWB/4Mzq/JkN7cusYRnr7U8JljIzrwDGVfe/N3BRZi6ogpCLgH2WU1tP7QH8LTPvW8Yy+wPnZOZzmXkPMJsSmuwEzM7MuzNzIXBOteyK6Jfnpid3lJn3Z+a11fkngNuA9ZZT20A9D50NxH1IkiRJkrRcbR8Y9cDhQGNvnI0i4rqIuCwiXl21rUfp0dFhLktDhYmZeX91/gFgYg/vt7v11gPmdHFfy2rvrraeOoTSa6bD+6ohX99tGGLX27p6KoHfR8Q1EXFk1dZfz02vVD3Mtqf01IKBfR56YiDuQ5IkSZKk5RrSgVFE/DewCPhR1XQ/sEFmbg98GPhxRIzt6e1VPXx6PYavr+v1h4gYBewH/KRqOg14OTCF8nx8qcklvCozdwD2BY6uhu8tMVDPTUSMAc4HjsnMxxn450GSJEmSpLYxZAOjiHgn8Ebg7VUoQTXM6OHq/DXA3yjz08yjPmxtctUGML8aFtUxdO3BHpbQ3XrzgPW7uK9ltXdXW0/sC1ybmfMBMnN+Zi7OzBeAb1OGQfWlrh7JzHnV3weBn1f311/PTY9ExMqUsOhHmfmzqp4BfR56aCDuQ5IkSZKk5RqSgVFE7AMcB+yXmU83tL+0mliYiNgY2BS4uxoe9XhE7FIdgeww4IJqtRlAx1G5pje0L093680ADquOCLYL8Fh1/78D9oqItarhUXsBv1tObT1xKA3D0TqCmsoBQMeR3WYAh0TEKhGxEeW5uYoy8fKmEbFR1VvpkGrZ5YqI1auJpomI1avHdDP99Nz0sIYAzgBuy8wvt+J56IWBuA9JkiRJkpZrZKsLWFERcTYwDVg7IuYCJwAfB1YBLip5AVdk5nuB3YETI+J54AXgvZnZMVn1fwLfB1ajzHnUMe/RScB5EXEEcB/l6Fo9qaG79X5DOQrYbOBpylHLyMwFEfEpSmgAcGIPalvec7M65YhbRzU0fyEiplCGgd3bcV1m3hIR5wG3UobxHZ2Zi6vbeR8loBkBfDczb+nJ/VPmJvp5tQ1GAj/OzAsj4mr677lZnt2AdwA3RcT1Vdt/AYcO4PPQI9VR/Zp6H5IkSZIk9URUo7UkSZIkSZIkYIgOSZMkSZIkSVLfGRhJkiRJkiSpxsBIkiRJkiRJNQZGkiRJkiRJqjEwkiRJkiRJUo2BkSRJkiRJkmoMjCRJkiRJklRjYCRJkiRJkqSa/w9x7/i5+9SSMQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 1440x720 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=(20,10))\n",
    "upset = plot(upset_gene_df, fig=fig, element_size=None,show_counts=True)\n",
    "plt.suptitle(\"Your legend goes here\")\n",
    "plt.savefig(\"gene_upset_plot.svg\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4b511efb-ad58-44c2-aaac-04e5a314ba32",
   "metadata": {},
   "source": [
    "# 5. Notes"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "40d0048f-e479-4dad-8c91-575b24883226",
   "metadata": {},
   "source": [
    "## As opposed to plotting the missing values across different channels in TMT tagging, plotting the results of orthogroups presents specific challenges:\n",
    "\n",
    "- ### The number of genes per orthogroup might be different across species\n",
    "- ### The identifiers of shared genes are not identical across species\n",
    "- ### The number of genes per species might differ quite significantly\n",
    "\n",
    "## Let's consider the following toy example:\n",
    "\n",
    "|Orthogroup|Species 1|Species 2|Species 3|OG Class|\n",
    "|:--------:|:-------:|:-------:|:-------:|:------:|\n",
    "|OG0001    |  1000   |     1   |   10    |    0   |\n",
    "|OG0002    |     1   |  1000   |   10    |    0   |\n",
    "|OG0003    |    10   |     0   |   10    |    2   |\n",
    "|OG0004    |     0   |    10   |   10    |    4   |\n",
    "\n",
    "The resulting plot will display results that might be somewhat deceiving, 2 Orthogroups shared across the 3 species, 1 orthogroup shared across 2 species, and 1 orthogroup shared in 2 different species. But the gene counts tell a different story: while orthogroups 1 and 2 are conserved across all three species, they were massively expanded in species 1 and 2, thus, plotting the actual number of genes is helpful (to some extent).\n",
    "\n",
    "There is a caveat to this approach though, because UpsetPlot assume a symetric distribution of the intersected data, the number of genes is inflated in the second plot, thus, the side bars should be removed from the plot, in addition, a pie chart on top of each category could also help to see what species present more genes in each category but so far that has not been implemented"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}