09_benchmark.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "9be010e3",
   "metadata": {},
   "source": [
    "# Benchmarking pipeline"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "88056c0f",
   "metadata": {},
   "source": [
    "Biological activities can be estimated by coupling a statistical method with a prior knowledge network. However, it is not clear which method or network performs better. When perturbation experiments are available, one can benchmark these methods by evaluating how good are they at recovering perturbed regulators. In our original `decoupler` [publication](https://doi.org/10.1093/bioadv/vbac016), we observed that `wsum`, `mlm`, `ulm` and `consensus` where the top performer methods. \n",
    "\n",
    "In this notebook we showcase how to use `decoupler` for the benchmarking of methods and prior knowledge networks. The data used here consist of single-gene perturbation experiments from the [KnockTF2](http://www.licpathway.net/KnockTFv2/index.php) database, an extensive dataset of literature curated perturbation experiments. If you use these data for your resarch, please cite the original publication of the resource:\n",
    "\n",
    ">Chenchen Feng, Chao Song, Yuejuan Liu, Fengcui Qian, Yu Gao, Ziyu Ning, Qiuyu Wang, Yong Jiang, Yanyu Li, Meng Li, Jiaxin Chen, Jian Zhang, Chunquan Li, KnockTF: a comprehensive human gene expression profile database with knockdown/knockout of transcription factors, Nucleic Acids Research, Volume 48, Issue D1, 08 January 2020, Pages D93–D100, https://doi.org/10.1093/nar/gkz881"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b1a45580-cff8-4467-aa07-ec0fbdae2008",
   "metadata": {},
   "source": [
    "生物学的活性は、統計的手法と事前知識ネットワークを組み合わせることで推定できる。しかし、どの手法やネットワークが優れているかは明らかではない。摂動実験が可能であれば、摂動された制御因子を回復させるのにどの程度優れているかを評価することで、これらの手法をベンチマークすることができる。私たちが最初に発表したdecouplerでは、wsum、mlm、ulm、consensusが上位を占めました。\n",
    "\n",
    "このノートブックでは、手法と事前知識ネットワークのベンチマークにdecouplerを使用する方法を紹介する。ここで使用したデータはKnockTF2データベースの単一遺伝子摂動実験です。これらのデータを研究に使用する場合は、そのリソースの原著を引用してください："
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4a2fe17a",
   "metadata": {},
   "source": [
    "## Loading packages\n",
    "\n",
    "First, we need to load the relevant packages."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4ee66a7c-9523-44ec-bd71-7eee954faffc",
   "metadata": {},
   "source": [
    "まず、関連するパッケージをロードする必要がある。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "5a195373-b827-4cee-848d-33b39657a36a",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "f7cfc5d4-3aa7-4748-9623-94a843e3921d",
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "1373e60c-168d-4a1e-85b4-9d742107d2eb",
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "8c610d39",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "import decoupler as dc"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1552661f",
   "metadata": {},
   "source": [
    "## Loading the data\n",
    "\n",
    "We can download the data easily from Zenodo:"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e1c24c02-fe3a-4abd-b6c9-4e5ff755b766",
   "metadata": {},
   "source": [
    "データはZenodoから簡単にダウンロードできる："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "7654e58f",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "--2024-08-08 23:16:58--  https://zenodo.org/record/7035528/files/knockTF_expr.csv?download=1\n",
      "zenodo.org (zenodo.org) をDNSに問いあわせています... 188.184.98.238, 188.185.79.172, 188.184.103.159\n",
      "zenodo.org (zenodo.org)|188.184.98.238|:443 に接続しています... 接続しました。\n",
      "HTTP による接続要求を送信しました、応答を待っています... 301 MOVED PERMANENTLY\n",
      "場所: /records/7035528/files/knockTF_expr.csv [続く]\n",
      "--2024-08-08 23:17:00--  https://zenodo.org/records/7035528/files/knockTF_expr.csv\n",
      "zenodo.org:443 への接続を再利用します。\n",
      "HTTP による接続要求を送信しました、応答を待っています... 200 OK\n",
      "長さ: 146086808 (139M) [text/plain]\n",
      "`knockTF_expr.csv' に保存中\n",
      "\n",
      "knockTF_expr.csv    100%[===================>] 139.32M  2.98MB/s 時間 50s        \n",
      "\n",
      "2024-08-08 23:17:50 (2.81 MB/s) - `knockTF_expr.csv' へ保存完了 [146086808/146086808]\n",
      "\n",
      "--2024-08-08 23:17:51--  https://zenodo.org/record/7035528/files/knockTF_meta.csv?download=1\n",
      "zenodo.org (zenodo.org) をDNSに問いあわせています... 188.185.79.172, 188.184.103.159, 188.184.98.238\n",
      "zenodo.org (zenodo.org)|188.185.79.172|:443 に接続しています... 接続しました。\n",
      "HTTP による接続要求を送信しました、応答を待っています... 301 MOVED PERMANENTLY\n",
      "場所: /records/7035528/files/knockTF_meta.csv [続く]\n",
      "--2024-08-08 23:17:52--  https://zenodo.org/records/7035528/files/knockTF_meta.csv\n",
      "zenodo.org:443 への接続を再利用します。\n",
      "HTTP による接続要求を送信しました、応答を待っています... 200 OK\n",
      "長さ: 144861 (141K) [text/plain]\n",
      "`knockTF_meta.csv' に保存中\n",
      "\n",
      "knockTF_meta.csv    100%[===================>] 141.47K   148KB/s 時間 1.0s       \n",
      "\n",
      "2024-08-08 23:17:54 (148 KB/s) - `knockTF_meta.csv' へ保存完了 [144861/144861]\n",
      "\n"
     ]
    }
   ],
   "source": [
    "!wget 'https://zenodo.org/record/7035528/files/knockTF_expr.csv?download=1' -O knockTF_expr.csv\n",
    "!wget 'https://zenodo.org/record/7035528/files/knockTF_meta.csv?download=1' -O knockTF_meta.csv"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d2e11366",
   "metadata": {},
   "source": [
    "We can then read it using `pandas`:"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d0d814b6-1bae-49f7-b114-cc567dfb43b4",
   "metadata": {},
   "source": [
    "そしてpandasを使ってそれを読むことができる："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "25c778f4",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "# Read data\n",
    "mat = pd.read_csv('knockTF_expr.csv', index_col=0)\n",
    "obs = pd.read_csv('knockTF_meta.csv', index_col=0)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "70634b2b",
   "metadata": {},
   "source": [
    "`mat` consist of a matrix of logFCs between the perturbed and the control samples. Positive values mean that these genes were\n",
    "overexpressed after the perturbation, negative values mean the opposite."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "84812ca3-2465-499a-861a-6d10bcd04982",
   "metadata": {},
   "source": [
    "matは、摂動サンプルとコントロールサンプルの間のlogFCのマトリックスからなる。正の値は、これらの遺伝子が摂動後に過剰発現したことを意味し、負の値はその逆を意味する。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "8fc3a459",
   "metadata": {
    "tags": []
   },
   "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>A1BG</th>\n",
       "      <th>A1BG-AS1</th>\n",
       "      <th>A1CF</th>\n",
       "      <th>A2LD1</th>\n",
       "      <th>A2M</th>\n",
       "      <th>A2ML1</th>\n",
       "      <th>A4GALT</th>\n",
       "      <th>A4GNT</th>\n",
       "      <th>AA06</th>\n",
       "      <th>AAA1</th>\n",
       "      <th>...</th>\n",
       "      <th>ZWINT</th>\n",
       "      <th>ZXDA</th>\n",
       "      <th>ZXDB</th>\n",
       "      <th>ZXDC</th>\n",
       "      <th>ZYG11A</th>\n",
       "      <th>ZYG11B</th>\n",
       "      <th>ZYX</th>\n",
       "      <th>ZZEF1</th>\n",
       "      <th>ZZZ3</th>\n",
       "      <th>psiTPTE22</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>DataSet_01_001</th>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>-0.03322</td>\n",
       "      <td>-0.04690</td>\n",
       "      <td>-0.14383</td>\n",
       "      <td>0.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DataSet_01_002</th>\n",
       "      <td>-0.17375</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>-0.53779</td>\n",
       "      <td>-0.4424</td>\n",
       "      <td>-1.59664</td>\n",
       "      <td>-0.00010</td>\n",
       "      <td>-0.08260</td>\n",
       "      <td>-3.64590</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.39579</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.08611</td>\n",
       "      <td>0.28574</td>\n",
       "      <td>-0.13021</td>\n",
       "      <td>-0.27664</td>\n",
       "      <td>-0.88158</td>\n",
       "      <td>0.81295</td>\n",
       "      <td>0.47896</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DataSet_01_003</th>\n",
       "      <td>-0.21613</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>-0.22027</td>\n",
       "      <td>-0.0008</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.15580</td>\n",
       "      <td>-0.35802</td>\n",
       "      <td>-0.32025</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.82964</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.06447</td>\n",
       "      <td>0.59323</td>\n",
       "      <td>0.31395</td>\n",
       "      <td>-0.23260</td>\n",
       "      <td>0.06551</td>\n",
       "      <td>-0.14714</td>\n",
       "      <td>0.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DataSet_01_004</th>\n",
       "      <td>-0.25568</td>\n",
       "      <td>0.11105</td>\n",
       "      <td>-0.28527</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>-0.03586</td>\n",
       "      <td>-0.46970</td>\n",
       "      <td>0.18559</td>\n",
       "      <td>-0.25601</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.39888</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.10444</td>\n",
       "      <td>-0.16434</td>\n",
       "      <td>0.19146</td>\n",
       "      <td>0.41561</td>\n",
       "      <td>0.39384</td>\n",
       "      <td>0.12790</td>\n",
       "      <td>0.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DataSet_01_005</th>\n",
       "      <td>0.47850</td>\n",
       "      <td>-0.37571</td>\n",
       "      <td>-0.84718</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>3.35445</td>\n",
       "      <td>0.17104</td>\n",
       "      <td>-0.34852</td>\n",
       "      <td>-0.95517</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.24849</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.28692</td>\n",
       "      <td>-0.01815</td>\n",
       "      <td>0.11941</td>\n",
       "      <td>0.07785</td>\n",
       "      <td>0.23474</td>\n",
       "      <td>0.22869</td>\n",
       "      <td>0.00000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 21985 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                   A1BG  A1BG-AS1     A1CF   A2LD1      A2M    A2ML1   A4GALT  \\\n",
       "DataSet_01_001  0.00000   0.00000  0.00000  0.0000  0.00000  0.00000  0.00000   \n",
       "DataSet_01_002 -0.17375   0.00000 -0.53779 -0.4424 -1.59664 -0.00010 -0.08260   \n",
       "DataSet_01_003 -0.21613   0.00000 -0.22027 -0.0008  0.00000  0.15580 -0.35802   \n",
       "DataSet_01_004 -0.25568   0.11105 -0.28527  0.0000 -0.03586 -0.46970  0.18559   \n",
       "DataSet_01_005  0.47850  -0.37571 -0.84718  0.0000  3.35445  0.17104 -0.34852   \n",
       "\n",
       "                  A4GNT  AA06  AAA1  ...  ZWINT     ZXDA  ZXDB     ZXDC  \\\n",
       "DataSet_01_001  0.00000   0.0   0.0  ...    0.0  0.00000   0.0  0.00000   \n",
       "DataSet_01_002 -3.64590   0.0   0.0  ...    0.0  0.39579   0.0  0.08611   \n",
       "DataSet_01_003 -0.32025   0.0   0.0  ...    0.0 -0.82964   0.0  0.06447   \n",
       "DataSet_01_004 -0.25601   0.0   0.0  ...    0.0 -0.39888   0.0  0.10444   \n",
       "DataSet_01_005 -0.95517   0.0   0.0  ...    0.0  0.24849   0.0 -0.28692   \n",
       "\n",
       "                 ZYG11A   ZYG11B      ZYX    ZZEF1     ZZZ3  psiTPTE22  \n",
       "DataSet_01_001  0.00000  0.00000 -0.03322 -0.04690 -0.14383    0.00000  \n",
       "DataSet_01_002  0.28574 -0.13021 -0.27664 -0.88158  0.81295    0.47896  \n",
       "DataSet_01_003  0.59323  0.31395 -0.23260  0.06551 -0.14714    0.00000  \n",
       "DataSet_01_004 -0.16434  0.19146  0.41561  0.39384  0.12790    0.00000  \n",
       "DataSet_01_005 -0.01815  0.11941  0.07785  0.23474  0.22869    0.00000  \n",
       "\n",
       "[5 rows x 21985 columns]"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mat.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c420680b",
   "metadata": {},
   "source": [
    "On the other hand, `obs` contains the meta-data information of each perturbation experiment.\n",
    "Here one can find which TF was perturbed (`TF` column), and many other useful information."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0555fc1f-da6d-429e-a9b1-81171556c6cd",
   "metadata": {},
   "source": [
    "一方、obsには各摂動実験のメタデータ情報が含まれている。ここでは、どのTFが摂動されたか（TFカラム）、その他多くの有用な情報を得ることができる。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "fa5396dc",
   "metadata": {
    "tags": []
   },
   "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>TF</th>\n",
       "      <th>Species</th>\n",
       "      <th>Knock.Method</th>\n",
       "      <th>Biosample.Name</th>\n",
       "      <th>Profile.ID</th>\n",
       "      <th>Platform</th>\n",
       "      <th>TF.Class</th>\n",
       "      <th>TF.Superclass</th>\n",
       "      <th>Tissue.Type</th>\n",
       "      <th>Biosample.Type</th>\n",
       "      <th>Data.Source</th>\n",
       "      <th>Pubmed.ID</th>\n",
       "      <th>logFC</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>DataSet_01_001</th>\n",
       "      <td>ESR1</td>\n",
       "      <td>Homo sapiens</td>\n",
       "      <td>siRNA</td>\n",
       "      <td>MCF7</td>\n",
       "      <td>GSE10061</td>\n",
       "      <td>GPL3921</td>\n",
       "      <td>Nuclear receptors with C4 zinc fingers</td>\n",
       "      <td>Zinc-coordinating DNA-binding domains</td>\n",
       "      <td>Mammary_gland</td>\n",
       "      <td>Cell line</td>\n",
       "      <td>GEO</td>\n",
       "      <td>18631401</td>\n",
       "      <td>-0.71385</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DataSet_01_002</th>\n",
       "      <td>HNF1A</td>\n",
       "      <td>Homo sapiens</td>\n",
       "      <td>shRNA</td>\n",
       "      <td>HuH7</td>\n",
       "      <td>GSE103128</td>\n",
       "      <td>GPL18180</td>\n",
       "      <td>Homeo domain factors</td>\n",
       "      <td>Helix-turn-helix domains</td>\n",
       "      <td>Liver</td>\n",
       "      <td>Cell line</td>\n",
       "      <td>GEO</td>\n",
       "      <td>29466992</td>\n",
       "      <td>0.16428</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DataSet_01_003</th>\n",
       "      <td>MLXIP</td>\n",
       "      <td>Homo sapiens</td>\n",
       "      <td>shRNA</td>\n",
       "      <td>HA1ER</td>\n",
       "      <td>GSE11242</td>\n",
       "      <td>GPL4133</td>\n",
       "      <td>Basic helix-loop-helix factors (bHLH)</td>\n",
       "      <td>Basic domains</td>\n",
       "      <td>Embryo_kidney</td>\n",
       "      <td>Stem cell</td>\n",
       "      <td>GEO</td>\n",
       "      <td>18458340</td>\n",
       "      <td>0.26215</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DataSet_01_004</th>\n",
       "      <td>CREB1</td>\n",
       "      <td>Homo sapiens</td>\n",
       "      <td>shRNA</td>\n",
       "      <td>K562</td>\n",
       "      <td>GSE12056</td>\n",
       "      <td>GPL570</td>\n",
       "      <td>Basic leucine zipper factors (bZIP)</td>\n",
       "      <td>Basic domains</td>\n",
       "      <td>Haematopoietic_and_lymphoid_tissue</td>\n",
       "      <td>Cell line</td>\n",
       "      <td>GEO</td>\n",
       "      <td>18801183</td>\n",
       "      <td>-0.95018</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DataSet_01_005</th>\n",
       "      <td>POU5F1</td>\n",
       "      <td>Homo sapiens</td>\n",
       "      <td>siRNA</td>\n",
       "      <td>GBS6</td>\n",
       "      <td>GSE12320</td>\n",
       "      <td>GPL570</td>\n",
       "      <td>Homeo domain factors</td>\n",
       "      <td>Helix-turn-helix domains</td>\n",
       "      <td>Bone_marrow</td>\n",
       "      <td>Cell line</td>\n",
       "      <td>GEO</td>\n",
       "      <td>20203285</td>\n",
       "      <td>0.00000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                    TF       Species Knock.Method Biosample.Name Profile.ID  \\\n",
       "DataSet_01_001    ESR1  Homo sapiens        siRNA           MCF7   GSE10061   \n",
       "DataSet_01_002   HNF1A  Homo sapiens        shRNA           HuH7  GSE103128   \n",
       "DataSet_01_003   MLXIP  Homo sapiens        shRNA          HA1ER   GSE11242   \n",
       "DataSet_01_004   CREB1  Homo sapiens        shRNA           K562   GSE12056   \n",
       "DataSet_01_005  POU5F1  Homo sapiens        siRNA           GBS6   GSE12320   \n",
       "\n",
       "                Platform                                TF.Class  \\\n",
       "DataSet_01_001   GPL3921  Nuclear receptors with C4 zinc fingers   \n",
       "DataSet_01_002  GPL18180                    Homeo domain factors   \n",
       "DataSet_01_003   GPL4133   Basic helix-loop-helix factors (bHLH)   \n",
       "DataSet_01_004    GPL570     Basic leucine zipper factors (bZIP)   \n",
       "DataSet_01_005    GPL570                    Homeo domain factors   \n",
       "\n",
       "                                        TF.Superclass  \\\n",
       "DataSet_01_001  Zinc-coordinating DNA-binding domains   \n",
       "DataSet_01_002               Helix-turn-helix domains   \n",
       "DataSet_01_003                          Basic domains   \n",
       "DataSet_01_004                          Basic domains   \n",
       "DataSet_01_005               Helix-turn-helix domains   \n",
       "\n",
       "                                       Tissue.Type Biosample.Type Data.Source  \\\n",
       "DataSet_01_001                       Mammary_gland      Cell line         GEO   \n",
       "DataSet_01_002                               Liver      Cell line         GEO   \n",
       "DataSet_01_003                       Embryo_kidney      Stem cell         GEO   \n",
       "DataSet_01_004  Haematopoietic_and_lymphoid_tissue      Cell line         GEO   \n",
       "DataSet_01_005                         Bone_marrow      Cell line         GEO   \n",
       "\n",
       "               Pubmed.ID    logFC  \n",
       "DataSet_01_001  18631401 -0.71385  \n",
       "DataSet_01_002  29466992  0.16428  \n",
       "DataSet_01_003  18458340  0.26215  \n",
       "DataSet_01_004  18801183 -0.95018  \n",
       "DataSet_01_005  20203285  0.00000  "
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "obs.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f6d02df0",
   "metadata": {},
   "source": [
    "## Filtering\n",
    "\n",
    "Since some of the experiments collected inside KnockTF are of low quality, it is best to remove them. We can do this by filtering\n",
    "perturbation experiments where the knocked TF does not show a strong negative logFC (meaning that the knock method did not work).\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "64e76186-a011-4f54-8b3e-cf99e6154f8c",
   "metadata": {},
   "source": [
    "KnockTFの中で収集された実験のいくつかは質が低いので、それらを除去するのが最善である。これは、ノックされたTFが強い負のlogFCを示さない（ノック法が機能しなかったことを意味する）摂動実験をフィルタリングすることによって行うことができる。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "b5450177",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "# Filter out low quality experiments (positive logFCs in knockout experiments)\n",
    "msk = obs['logFC'] < -1\n",
    "mat = mat[msk]\n",
    "obs = obs[msk]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "1751080c-c648-4b69-bf8c-e66c0b9fb135",
   "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>A1BG</th>\n",
       "      <th>A1BG-AS1</th>\n",
       "      <th>A1CF</th>\n",
       "      <th>A2LD1</th>\n",
       "      <th>A2M</th>\n",
       "      <th>A2ML1</th>\n",
       "      <th>A4GALT</th>\n",
       "      <th>A4GNT</th>\n",
       "      <th>AA06</th>\n",
       "      <th>AAA1</th>\n",
       "      <th>...</th>\n",
       "      <th>ZWINT</th>\n",
       "      <th>ZXDA</th>\n",
       "      <th>ZXDB</th>\n",
       "      <th>ZXDC</th>\n",
       "      <th>ZYG11A</th>\n",
       "      <th>ZYG11B</th>\n",
       "      <th>ZYX</th>\n",
       "      <th>ZZEF1</th>\n",
       "      <th>ZZZ3</th>\n",
       "      <th>psiTPTE22</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>DataSet_01_006</th>\n",
       "      <td>-0.41056</td>\n",
       "      <td>3.13675</td>\n",
       "      <td>-0.49427</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.23968</td>\n",
       "      <td>1.37572</td>\n",
       "      <td>-0.03733</td>\n",
       "      <td>0.56758</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.39262</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.24752</td>\n",
       "      <td>-0.03830</td>\n",
       "      <td>-0.05072</td>\n",
       "      <td>0.17918</td>\n",
       "      <td>-0.17708</td>\n",
       "      <td>0.08701</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DataSet_01_007</th>\n",
       "      <td>-0.50900</td>\n",
       "      <td>2.12951</td>\n",
       "      <td>-0.16245</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.02671</td>\n",
       "      <td>-0.61539</td>\n",
       "      <td>0.06548</td>\n",
       "      <td>0.87137</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.22490</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.17279</td>\n",
       "      <td>-0.54982</td>\n",
       "      <td>-0.22164</td>\n",
       "      <td>-0.13836</td>\n",
       "      <td>0.23519</td>\n",
       "      <td>0.13498</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DataSet_01_008</th>\n",
       "      <td>-0.25537</td>\n",
       "      <td>0.06701</td>\n",
       "      <td>0.31111</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>-0.02576</td>\n",
       "      <td>-0.15574</td>\n",
       "      <td>0.00037</td>\n",
       "      <td>-0.14773</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.35065</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.15874</td>\n",
       "      <td>0.04277</td>\n",
       "      <td>0.34284</td>\n",
       "      <td>-0.09468</td>\n",
       "      <td>0.16314</td>\n",
       "      <td>0.02952</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DataSet_01_010</th>\n",
       "      <td>0.05667</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.81121</td>\n",
       "      <td>0.46618</td>\n",
       "      <td>-0.45087</td>\n",
       "      <td>0.05971</td>\n",
       "      <td>0.00786</td>\n",
       "      <td>-0.30456</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.63275</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.40010</td>\n",
       "      <td>1.81472</td>\n",
       "      <td>0.32373</td>\n",
       "      <td>-0.30996</td>\n",
       "      <td>0.39868</td>\n",
       "      <td>0.77080</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DataSet_01_011</th>\n",
       "      <td>-0.03059</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>-1.20615</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.57765</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.08458</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 21985 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                   A1BG  A1BG-AS1     A1CF    A2LD1      A2M    A2ML1  \\\n",
       "DataSet_01_006 -0.41056   3.13675 -0.49427  0.00000  0.23968  1.37572   \n",
       "DataSet_01_007 -0.50900   2.12951 -0.16245  0.00000  0.02671 -0.61539   \n",
       "DataSet_01_008 -0.25537   0.06701  0.31111  0.00000 -0.02576 -0.15574   \n",
       "DataSet_01_010  0.05667   0.00000  0.81121  0.46618 -0.45087  0.05971   \n",
       "DataSet_01_011 -0.03059   0.00000  0.00000  0.00000 -1.20615  0.00000   \n",
       "\n",
       "                 A4GALT    A4GNT  AA06  AAA1  ...  ZWINT     ZXDA  ZXDB  \\\n",
       "DataSet_01_006 -0.03733  0.56758   0.0   0.0  ...    0.0 -0.39262   0.0   \n",
       "DataSet_01_007  0.06548  0.87137   0.0   0.0  ...    0.0 -0.22490   0.0   \n",
       "DataSet_01_008  0.00037 -0.14773   0.0   0.0  ...    0.0  0.35065   0.0   \n",
       "DataSet_01_010  0.00786 -0.30456   0.0   0.0  ...    0.0  1.63275   0.0   \n",
       "DataSet_01_011  0.57765  0.00000   0.0   0.0  ...    0.0  0.00000   0.0   \n",
       "\n",
       "                   ZXDC   ZYG11A   ZYG11B      ZYX    ZZEF1     ZZZ3  \\\n",
       "DataSet_01_006 -0.24752 -0.03830 -0.05072  0.17918 -0.17708  0.08701   \n",
       "DataSet_01_007 -0.17279 -0.54982 -0.22164 -0.13836  0.23519  0.13498   \n",
       "DataSet_01_008  0.15874  0.04277  0.34284 -0.09468  0.16314  0.02952   \n",
       "DataSet_01_010  0.40010  1.81472  0.32373 -0.30996  0.39868  0.77080   \n",
       "DataSet_01_011  0.00000  0.00000  0.00000  0.08458  0.00000  0.00000   \n",
       "\n",
       "                psiTPTE22  \n",
       "DataSet_01_006        0.0  \n",
       "DataSet_01_007        0.0  \n",
       "DataSet_01_008        0.0  \n",
       "DataSet_01_010        0.0  \n",
       "DataSet_01_011        0.0  \n",
       "\n",
       "[5 rows x 21985 columns]"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mat.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "cf1f1929-9126-4f33-8ddc-a4c6041a655b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(388, 21985)"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mat.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "1875b73c-8913-477e-8e7b-cb663c80e232",
   "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>TF</th>\n",
       "      <th>Species</th>\n",
       "      <th>Knock.Method</th>\n",
       "      <th>Biosample.Name</th>\n",
       "      <th>Profile.ID</th>\n",
       "      <th>Platform</th>\n",
       "      <th>TF.Class</th>\n",
       "      <th>TF.Superclass</th>\n",
       "      <th>Tissue.Type</th>\n",
       "      <th>Biosample.Type</th>\n",
       "      <th>Data.Source</th>\n",
       "      <th>Pubmed.ID</th>\n",
       "      <th>logFC</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>DataSet_01_006</th>\n",
       "      <td>PROX1</td>\n",
       "      <td>Homo sapiens</td>\n",
       "      <td>siRNA</td>\n",
       "      <td>Primary lymphatic endothelial cells</td>\n",
       "      <td>GSE12846</td>\n",
       "      <td>GPL570</td>\n",
       "      <td>Homeo domain factors</td>\n",
       "      <td>Helix-turn-helix domains</td>\n",
       "      <td>Foreskin</td>\n",
       "      <td>Primary cell</td>\n",
       "      <td>GEO</td>\n",
       "      <td>18815287</td>\n",
       "      <td>-1.13977</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DataSet_01_007</th>\n",
       "      <td>NR2F2</td>\n",
       "      <td>Homo sapiens</td>\n",
       "      <td>siRNA</td>\n",
       "      <td>Primary lymphatic endothelial cells</td>\n",
       "      <td>GSE12846</td>\n",
       "      <td>GPL570</td>\n",
       "      <td>Nuclear receptors with C4 zinc fingers</td>\n",
       "      <td>Zinc-coordinating DNA-binding domains</td>\n",
       "      <td>Foreskin</td>\n",
       "      <td>Primary cell</td>\n",
       "      <td>GEO</td>\n",
       "      <td>18815287</td>\n",
       "      <td>-1.34696</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DataSet_01_008</th>\n",
       "      <td>WT1</td>\n",
       "      <td>Homo sapiens</td>\n",
       "      <td>siRNA</td>\n",
       "      <td>PLC/PRF/5</td>\n",
       "      <td>GSE12886</td>\n",
       "      <td>GPL570</td>\n",
       "      <td>C2H2 zinc finger factors</td>\n",
       "      <td>Zinc-coordinating DNA-binding domains</td>\n",
       "      <td>Liver</td>\n",
       "      <td>Cell line</td>\n",
       "      <td>GEO</td>\n",
       "      <td>19190340</td>\n",
       "      <td>-1.11474</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DataSet_01_010</th>\n",
       "      <td>TP53</td>\n",
       "      <td>Homo sapiens</td>\n",
       "      <td>shRNA</td>\n",
       "      <td>iPS cells</td>\n",
       "      <td>GSE13334</td>\n",
       "      <td>GPL4133</td>\n",
       "      <td>p53 domain factors</td>\n",
       "      <td>Immunoglobulin fold</td>\n",
       "      <td>Dermal_fibroblasts</td>\n",
       "      <td>Induced pluripotent stem cell</td>\n",
       "      <td>GEO</td>\n",
       "      <td>19668191</td>\n",
       "      <td>-3.13745</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DataSet_01_011</th>\n",
       "      <td>TP53</td>\n",
       "      <td>Homo sapiens</td>\n",
       "      <td>siRNA</td>\n",
       "      <td>U251</td>\n",
       "      <td>GSE13991</td>\n",
       "      <td>GPL1426</td>\n",
       "      <td>p53 domain factors</td>\n",
       "      <td>Immunoglobulin fold</td>\n",
       "      <td>Central_nervous_system</td>\n",
       "      <td>Cell line</td>\n",
       "      <td>GEO</td>\n",
       "      <td>19139068</td>\n",
       "      <td>-2.43895</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   TF       Species Knock.Method  \\\n",
       "DataSet_01_006  PROX1  Homo sapiens        siRNA   \n",
       "DataSet_01_007  NR2F2  Homo sapiens        siRNA   \n",
       "DataSet_01_008    WT1  Homo sapiens        siRNA   \n",
       "DataSet_01_010   TP53  Homo sapiens        shRNA   \n",
       "DataSet_01_011   TP53  Homo sapiens        siRNA   \n",
       "\n",
       "                                     Biosample.Name Profile.ID Platform  \\\n",
       "DataSet_01_006  Primary lymphatic endothelial cells   GSE12846   GPL570   \n",
       "DataSet_01_007  Primary lymphatic endothelial cells   GSE12846   GPL570   \n",
       "DataSet_01_008                            PLC/PRF/5   GSE12886   GPL570   \n",
       "DataSet_01_010                            iPS cells   GSE13334  GPL4133   \n",
       "DataSet_01_011                                 U251   GSE13991  GPL1426   \n",
       "\n",
       "                                              TF.Class  \\\n",
       "DataSet_01_006                    Homeo domain factors   \n",
       "DataSet_01_007  Nuclear receptors with C4 zinc fingers   \n",
       "DataSet_01_008                C2H2 zinc finger factors   \n",
       "DataSet_01_010                      p53 domain factors   \n",
       "DataSet_01_011                      p53 domain factors   \n",
       "\n",
       "                                        TF.Superclass             Tissue.Type  \\\n",
       "DataSet_01_006               Helix-turn-helix domains                Foreskin   \n",
       "DataSet_01_007  Zinc-coordinating DNA-binding domains                Foreskin   \n",
       "DataSet_01_008  Zinc-coordinating DNA-binding domains                   Liver   \n",
       "DataSet_01_010                    Immunoglobulin fold      Dermal_fibroblasts   \n",
       "DataSet_01_011                    Immunoglobulin fold  Central_nervous_system   \n",
       "\n",
       "                               Biosample.Type Data.Source Pubmed.ID    logFC  \n",
       "DataSet_01_006                   Primary cell         GEO  18815287 -1.13977  \n",
       "DataSet_01_007                   Primary cell         GEO  18815287 -1.34696  \n",
       "DataSet_01_008                      Cell line         GEO  19190340 -1.11474  \n",
       "DataSet_01_010  Induced pluripotent stem cell         GEO  19668191 -3.13745  \n",
       "DataSet_01_011                      Cell line         GEO  19139068 -2.43895  "
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "obs.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "0925faf1-8e6f-4692-80bf-af51f33cf171",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(388, 13)"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "obs.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "af388fa0",
   "metadata": {},
   "source": [
    "## Evaluation\n",
    "\n",
    "### Single network\n",
    "\n",
    "As an example, we will evaluate the gene regulatory network\n",
    "[CollecTRI](https://github.com/saezlab/CollecTRI), for more information you can visit this\n",
    "[other vignette](https://decoupler-py.readthedocs.io/en/latest/notebooks/dorothea.html).\n",
    "\n",
    "We can retrieve this network by running:"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5b4c0fac-aa58-466e-a120-0d9cf93fc282",
   "metadata": {},
   "source": [
    "例として、遺伝子制御ネットワークCollecTRIを評価する。\n",
    "\n",
    "このネットワークは"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "e66d6631-575b-4f1e-ae33-89e0f78e2d1d",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "# Get collectri\n",
    "collectri = dc.get_collectri(organism='human', split_complexes=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "43e40621-ce06-4c6c-9bf5-e2c4047e7e57",
   "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>source</th>\n",
       "      <th>target</th>\n",
       "      <th>weight</th>\n",
       "      <th>PMID</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>MYC</td>\n",
       "      <td>TERT</td>\n",
       "      <td>1</td>\n",
       "      <td>10022128;10491298;10606235;10637317;10723141;1...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>SPI1</td>\n",
       "      <td>BGLAP</td>\n",
       "      <td>1</td>\n",
       "      <td>10022617</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>SMAD3</td>\n",
       "      <td>JUN</td>\n",
       "      <td>1</td>\n",
       "      <td>10022869;12374795</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>SMAD4</td>\n",
       "      <td>JUN</td>\n",
       "      <td>1</td>\n",
       "      <td>10022869;12374795</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>STAT5A</td>\n",
       "      <td>IL2</td>\n",
       "      <td>1</td>\n",
       "      <td>10022878;11435608;17182565;17911616;22854263;2...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   source target  weight                                               PMID\n",
       "0     MYC   TERT       1  10022128;10491298;10606235;10637317;10723141;1...\n",
       "1    SPI1  BGLAP       1                                           10022617\n",
       "2   SMAD3    JUN       1                                  10022869;12374795\n",
       "3   SMAD4    JUN       1                                  10022869;12374795\n",
       "4  STAT5A    IL2       1  10022878;11435608;17182565;17911616;22854263;2..."
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "collectri.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3c906e41",
   "metadata": {},
   "source": [
    "We will run the benchmark pipeline using the top performer methods from `decoupler`. If we would like, we could use any other,\n",
    "and we can optionaly change their parametters too (for more information check the `decouple` function).\n",
    "\n",
    "To run the benchmark pipeline, we need input molecular data (`mat`), its associated metadata (`obs`) with the name of the\n",
    "column where the perturbed regulators are (`perturb`), and which direction the pertrubations are (`sign`, positive or negative).\n",
    "    \n",
    "Note that if you want to ignore missing values in the input `mat` (denoted as zeros), you can add the parameter `'dense': True` to `decouple_kws`."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "490898a7-ac57-4e8b-a4d3-1a477120d033",
   "metadata": {},
   "source": [
    "decouplerのトップパフォーマーのメソッドを使ってベンチマークパイプラインを実行する。必要であれば、他のメソッドも使用することができ、オプションでパラメータを変更することもできます（詳細はdecouple関数を参照してください）。\n",
    "\n",
    "ベンチマークパイプラインを実行するには、分子データ(mat)、それに関連するメタデータ(obs)、摂動が加えられたカラムの名前(perturb)、摂動の方向(sign, positive or negative)を入力する必要がある。\n",
    "\n",
    "入力マットの欠損値（ゼロ）を無視したい場合は、'dense' パラメータを追加してください： を指定します。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "947b03a7-f674-41de-91a7-eb241fe8efac",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Example on how to set up decouple arguments\n",
    "decouple_kws={\n",
    "    'methods' : ['wsum', 'ulm', 'mlm'],\n",
    "    'consensus': True,\n",
    "    'args' : {\n",
    "        'wsum' : {'times': 100}\n",
    "    }\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "737759f4",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Extracting inputs...\n",
      "Formating net...\n",
      "Removed 109 experiments without sources in net.\n",
      "Running 279 experiments for 155 unique sources.\n",
      "Running methods...\n",
      "52 features of mat are empty, they will be removed.\n",
      "Running wsum on mat with 279 samples and 21933 targets for 766 sources.\n",
      "52 features of mat are empty, they will be removed.\n",
      "Running ulm on mat with 279 samples and 21933 targets for 766 sources.\n",
      "52 features of mat are empty, they will be removed.\n",
      "Running mlm on mat with 279 samples and 21933 targets for 766 sources.\n",
      "Calculating metrics...\n",
      "Computing metrics...\n",
      "Done.\n"
     ]
    }
   ],
   "source": [
    "# Run benchmark pipeline\n",
    "df = dc.benchmark(\n",
    "    mat, obs, collectri,\n",
    "    perturb='TF', sign=-1, verbose=True,\n",
    "    decouple_kws=decouple_kws\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4d0c141a",
   "metadata": {},
   "source": [
    "The result of running the pipeline is a long-format dataframe containing different performance metrics across methods:\n",
    "\n",
    "- `auroc`: Area Under the Receiver Operating characteristic Curve (AUROC)\n",
    "- `auprc`: Area Under the Precision-Recall Curve (AUPRC), which by default uses a calibrated version where 0.5 indicates a random classifier (adapted from [this publication](https://doi.org/10.1007/978-3-030-44584-3_36)).\n",
    "- `mauroc`: Monte-Carlo Area Under the Precision-Recall Curve (AUPRC)\n",
    "- `mauprc`: Monte-Carlo Area Under the Receiver Operating characteristic Curve (AUROC)\n",
    "- `rank`: Rank of the perturbed source per experiment.\n",
    "- `nrank`: Normalized rank of the perturbed source per experiment (close to 0 is better).\n",
    "\n",
    "The Monte-Carlo metrics perform random permutations where each one samples the same number of true positives and true negatives to have balanced sets, returning one value per iteration done (10k by default)."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7795f124-437f-4cda-af74-b949f234f1a4",
   "metadata": {},
   "source": [
    "パイプラインの実行結果は、メソッド間で異なるパフォーマンスメトリクスを含むロングフォーマットのデータフレームです：\n",
    "\n",
    "- auroc： AUROC: 動作特性曲線下面積 (AUROC)\n",
    "- auprc： Area Under the Precision-Recall Curve (AUPRC)。デフォルトでは、0.5がランダム分類器を示す較正バージョンを使用します（この出版物から引用）。\n",
    "- mauroc： モンテカルロ精度再現曲線下面積 (AUPRC)\n",
    "- mauprc： レシーバ動作特性曲線下モンテカルロ面積 (AUROC)\n",
    "- ランク 実験ごとの摂動ソースのランク。\n",
    "- nrank: 実験ごとの摂動ソースの正規化ランク（0に近いほど良い）。\n",
    "\n",
    "Monte-Carlo メトリクスは、バランスのとれた集合になるように、各々が同じ数の真陽性と真陰性をサンプリングするランダムな並べ替えを実行し、実行された反復ごとに1つの値を返す（デフォルトでは10k）。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "1c98ebbd-db48-47e2-8cb2-1721692ccbeb",
   "metadata": {
    "tags": []
   },
   "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>groupby</th>\n",
       "      <th>group</th>\n",
       "      <th>source</th>\n",
       "      <th>method</th>\n",
       "      <th>metric</th>\n",
       "      <th>score</th>\n",
       "      <th>ci</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>wsum_estimate</td>\n",
       "      <td>auroc</td>\n",
       "      <td>0.713571</td>\n",
       "      <td>0.001305</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>wsum_estimate</td>\n",
       "      <td>auprc</td>\n",
       "      <td>0.738849</td>\n",
       "      <td>0.001305</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>wsum_estimate</td>\n",
       "      <td>mcauroc</td>\n",
       "      <td>0.717372</td>\n",
       "      <td>0.001305</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>wsum_estimate</td>\n",
       "      <td>mcauroc</td>\n",
       "      <td>0.731504</td>\n",
       "      <td>0.001305</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>wsum_estimate</td>\n",
       "      <td>mcauroc</td>\n",
       "      <td>0.713069</td>\n",
       "      <td>0.001305</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  groupby group source         method   metric     score        ci\n",
       "0    None  None   None  wsum_estimate    auroc  0.713571  0.001305\n",
       "1    None  None   None  wsum_estimate    auprc  0.738849  0.001305\n",
       "2    None  None   None  wsum_estimate  mcauroc  0.717372  0.001305\n",
       "3    None  None   None  wsum_estimate  mcauroc  0.731504  0.001305\n",
       "4    None  None   None  wsum_estimate  mcauroc  0.713069  0.001305"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "14098ba7",
   "metadata": {},
   "source": [
    "We can easily visualize the results in a scatterplot: "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b81a3456-81ce-4edd-843b-a7371139f2e5",
   "metadata": {},
   "source": [
    "結果は散布図で簡単に視覚化できる： "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "d8ad87dd-892a-4787-9cf5-5acd6b8077fb",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 500x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "dc.plot_metrics_scatter(df, x='auroc', y='auprc')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a9c647ca",
   "metadata": {},
   "source": [
    "It looks like `mlm` this time had worse performance than other methods like `ulm` at recovering\n",
    "perturbed regulators, this might be caused because there are many TFs that share too many targets.\n",
    "Whenever there are multiple sources (TFs) and they share a lot of targets we recommend to use `ulm`.\n",
    "On the other hand, when there are few sources and they are orthogonal (they share few targets), we recommend to use `mlm`.\n",
    "Since `consensus` remains one of the top performer methods we also recommend its use but at a higher computational cost and less interpretability.\n",
    "\n",
    "We can also visualise the distirbutions of the obtained metrics, for example `mcauroc`, `mcauprc`, `rank` and `nrank`:"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d056c7c7-6bcf-4fbc-b4e3-330df24e6348",
   "metadata": {},
   "source": [
    "今回のmlmは、ulmのような他の方法よりも摂動制御因子を回復させるパフォーマンスが悪かったようだ。これは、ターゲットを共有するTFが多すぎるためかもしれない。複数のソース（TF）があり、それらが多くのターゲットを共有している場合は、ulmを使うことを推奨する。一方、ソースが少なく、それらが直交している（ターゲットを共有しているものが少ない）場合は、mlmを使うことを推奨する。コンセンサスは依然としてトップ・パフォーマーの手法の1つであるため、その使用も推奨するが、計算コストが高くなり、解釈可能性が低くなる。\n",
    "\n",
    "mcauroc、mcauprc、rank、nrankなど、得られたメトリクスの分布を可視化することもできます："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "73690cc0-3a6f-4b66-a6d5-1338bc4ff394",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 600x500 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axes = plt.subplots(\n",
    "    2, 2, figsize=(6, 5),\n",
    "    sharex=True, tight_layout=True\n",
    ")\n",
    "\n",
    "dc.plot_metrics_boxplot(df, metric='mcauroc', ax=axes[0, 0])\n",
    "\n",
    "dc.plot_metrics_boxplot(df, metric='mcauprc', ax=axes[0, 1])\n",
    "\n",
    "dc.plot_metrics_boxplot(df, metric='rank', ax=axes[1, 0])\n",
    "\n",
    "dc.plot_metrics_boxplot(df, metric='nrank', ax=axes[1, 1])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "11d7c52e",
   "metadata": {},
   "source": [
    "### Multiple networks\n",
    "\n",
    "This pipeline can evaluate multiple networks at the same time. Now we will evaluate CollecTRI and\n",
    "[DoRothEA](https://saezlab.github.io/dorothea/), a literature-based network that we developed in the past.\n",
    "We will also add a randomized version as control."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "51b4c1a8-9d88-4df5-b5da-f43afe121351",
   "metadata": {},
   "source": [
    "このパイプラインは複数のネットワークを同時に評価することができます。ここでは、CollecTRIと、過去に開発した文献ベースのネットワークであるDoRothEAを評価する。また、コントロールとしてランダム化バージョンも追加する。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "ff43435f-e746-4b78-a99e-3f7bcb037933",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "8cc6944c8cf9421692f36d33f67069c1",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "0.00B [00:00, ?B/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Get dorothea\n",
    "dorothea = dc.get_dorothea(organism='human')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "a34abe69",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "# Randomize nets\n",
    "ctri_rand = dc.shuffle_net(\n",
    "    collectri, target='target', weight='weight'\n",
    ").drop_duplicates(['source', 'target'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "8cc87364-558e-46dd-9373-d968a376bf7f",
   "metadata": {},
   "outputs": [],
   "source": [
    "doro_rand = dc.shuffle_net(\n",
    "    dorothea, target='target', weight='weight'\n",
    ").drop_duplicates(['source', 'target'])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "65089089",
   "metadata": {},
   "source": [
    "We can run the same pipeline but now using a dictionary of networks (and of extra arguments if needed)."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "691abca6-6e97-47ec-8d82-415257fff3e6",
   "metadata": {},
   "source": [
    "同じパイプラインを実行できるが、今度はネットワークの辞書を使う（必要であれば追加の引数も）。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "ee231442-8dd9-49cf-86e6-fc9a173151b4",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Build dictionary of networks to test\n",
    "nets = {\n",
    "    'r_collectri': ctri_rand,\n",
    "    'r_dorothea': doro_rand,\n",
    "    'collectri': collectri,\n",
    "    'dorothea': dorothea,\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "d572c1df-3127-48f3-8ae9-c47c8f0ba5a9",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Example extra arguments\n",
    "decouple_kws = {\n",
    "    'r_collectri': {\n",
    "        'methods' : ['ulm', 'mlm'],\n",
    "        'consensus': True\n",
    "    },\n",
    "    'r_dorothea': {\n",
    "        'methods' : ['ulm', 'mlm'],\n",
    "        'consensus': True\n",
    "    },\n",
    "    'collectri': {\n",
    "        'methods' : ['ulm', 'mlm'],\n",
    "        'consensus': True\n",
    "    },\n",
    "    'dorothea': {\n",
    "        'methods' : ['ulm', 'mlm'],\n",
    "        'consensus': True\n",
    "    }\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "029d1161",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Using r_collectri network...\n",
      "Extracting inputs...\n",
      "Formating net...\n",
      "Removed 109 experiments without sources in net.\n",
      "Running 279 experiments for 155 unique sources.\n",
      "Running methods...\n",
      "52 features of mat are empty, they will be removed.\n",
      "Running ulm on mat with 279 samples and 21933 targets for 762 sources.\n",
      "52 features of mat are empty, they will be removed.\n",
      "Running mlm on mat with 279 samples and 21933 targets for 762 sources.\n",
      "Calculating metrics...\n",
      "Computing metrics...\n",
      "Done.\n",
      "Using r_dorothea network...\n",
      "Extracting inputs...\n",
      "Formating net...\n",
      "Removed 175 experiments without sources in net.\n",
      "Running 213 experiments for 104 unique sources.\n",
      "Running methods...\n",
      "55 features of mat are empty, they will be removed.\n",
      "Running ulm on mat with 213 samples and 21930 targets for 295 sources.\n",
      "55 features of mat are empty, they will be removed.\n",
      "Running mlm on mat with 213 samples and 21930 targets for 295 sources.\n",
      "Calculating metrics...\n",
      "Computing metrics...\n",
      "Done.\n",
      "Using collectri network...\n",
      "Extracting inputs...\n",
      "Formating net...\n",
      "Removed 109 experiments without sources in net.\n",
      "Running 279 experiments for 155 unique sources.\n",
      "Running methods...\n",
      "52 features of mat are empty, they will be removed.\n",
      "Running ulm on mat with 279 samples and 21933 targets for 766 sources.\n",
      "52 features of mat are empty, they will be removed.\n",
      "Running mlm on mat with 279 samples and 21933 targets for 766 sources.\n",
      "Calculating metrics...\n",
      "Computing metrics...\n",
      "Done.\n",
      "Using dorothea network...\n",
      "Extracting inputs...\n",
      "Formating net...\n",
      "Removed 174 experiments without sources in net.\n",
      "Running 214 experiments for 105 unique sources.\n",
      "Running methods...\n",
      "55 features of mat are empty, they will be removed.\n",
      "Running ulm on mat with 214 samples and 21930 targets for 297 sources.\n",
      "55 features of mat are empty, they will be removed.\n",
      "Running mlm on mat with 214 samples and 21930 targets for 297 sources.\n",
      "Calculating metrics...\n",
      "Computing metrics...\n",
      "Done.\n"
     ]
    }
   ],
   "source": [
    "# Run benchmark pipeline\n",
    "df = dc.benchmark(\n",
    "    mat, obs, nets, perturb='TF', sign=-1, verbose=True,\n",
    "    decouple_kws=decouple_kws\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1c37c44e",
   "metadata": {},
   "source": [
    "Like before we can try to visualize the results in a scatterplot, this time grouping by the network used:"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6d97aed9-c646-4c71-af0d-d5f960ef1c4b",
   "metadata": {},
   "source": [
    "前回同様、散布図で結果を視覚化してみよう。今回は使用したネットワークでグループ化した："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "adb64864-b4b8-49a3-8f78-d170cf6ae050",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAlgAAAHFCAYAAAAudofcAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjkuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/TGe4hAAAACXBIWXMAAA9hAAAPYQGoP6dpAAC0NUlEQVR4nOzdeVhV1frA8e8BZB5FBFQEFUFRFBwT5xHEsVLJnMfMHHBIMyfUyvCGoqX+yhJsUsspMzRxwFlDzCkMJ5RUEHGAcGLavz+47OsRECgQwffzPOe5nrXX3utdG6+8rbX22hpFURSEEEIIIUSx0SntAIQQQgghyhtJsIQQQgghipkkWEIIIYQQxUwSLCGEEEKIYiYJlhBCCCFEMZMESwghhBCimEmCJYQQQghRzCTBEkIIIYQoZpJgCSGEEEIUM0mwhBBCCCGKWaknWCtWrKBGjRoYGhrSuHFjDhw48Mz63333HQ0bNsTY2Bh7e3uGDRvG7du3teps3LgRNzc3DAwMcHNzY/PmzSXZBSGEEEIILaWaYK1fvx5/f39mzpzJ77//TuvWrenatStxcXF51j948CCDBw9mxIgR/PHHH/z4449ERkYycuRItc6RI0fw8/Nj0KBBnDp1ikGDBtGvXz+OHTv2vLolhBBCiJecpjRf9ty8eXMaNWrEypUr1bK6devSu3dvFi5cmKv+J598wsqVK7l06ZJa9umnn7Jo0SL++usvAPz8/EhJSWH79u1qHR8fH6ysrFi7dm0J9kYIIYQQIpteaTWclpZGVFQU7733nlZ5ly5dOHz4cJ7neHl5MXPmTMLCwujatSuJiYls2LCBbt26qXWOHDnCpEmTtM7z9vYmODg431geP37M48eP1e9ZWVncuXMHa2trNBrNP+idEEKI8kBRFP7++2+qVKmCjk6pr6oRZUipJVhJSUlkZmZia2urVW5ra0tCQkKe53h5efHdd9/h5+fHo0ePyMjIoGfPnnz66adqnYSEhCJdE2DhwoXMmzfvX/RGCCFEefbXX39RrVq10g5DlCGllmDleHqESFGUfEeNoqOjmTBhAnPmzMHb25v4+HjeffddxowZw1dfffWPrgkwY8YMJk+erH5PTk6mevXqnD9/nooVK/6Tbr3Q0tPT2bt3L+3bt6dChQqlHU6xkr6VXeW5f9K3suvOnTu4uLhgZmZW2qGIMqbUEqxKlSqhq6uba2QpMTEx1whUjoULF9KyZUveffddABo0aICJiQmtW7fmgw8+wN7eHjs7uyJdE8DAwAADA4Nc5RUrVsTa2rqoXXvhpaenY2xsjLW1dbn7B1H6VnaV5/5J38o+WS4iiqrUJpT19fVp3Lgx4eHhWuXh4eF4eXnlec6DBw9yzYHr6uoC2aNUAC1atMh1zZ07d+Z7TSGEEEKI4laqU4STJ09m0KBBNGnShBYtWvDFF18QFxfHmDFjgOypu+vXr/P1118D0KNHD0aNGsXKlSvVKUJ/f3+aNWtGlSpVAJg4cSJt2rQhMDCQXr168dNPP7Fr1y4OHjxYav0UQgghxMulVBMsPz8/bt++zfz584mPj6d+/fqEhYXh6OgIQHx8vNaeWEOHDuXvv//ms88+Y8qUKVhaWtKhQwcCAwPVOl5eXqxbt45Zs2Yxe/ZsatWqxfr162nevPlz758QQoiChYaG4u/vz71790o7FCGKTakvch87dixjx47N81hoaGiusvHjxzN+/PhnXrNPnz706dOnOMITQghRjJycnPD398ff318t8/Pzw9fXt8TblkROPE+lnmAJIYR4uRkZGWFkZFTaYQhRrGTXNCHESykrK4vAwECcnZ0xMDCgevXqfPjhhwCcOXOGDh06YGRkhLW1NaNHjyY1NVU9d+jQofTu3ZtPPvkEe3t7rK2teeedd0hPT1frrFixgtq1a2NoaEi1atW0ljIoisKiRYuoWbMmRkZGNGzYkA0bNqjHIyIi0Gg07N69myZNmmBsbIyXlxcxMTFqnVOnTtG+fXvMzMwwNzencePGHD9+HICAgAA8PDy0+hscHIyTk5NWG82aNcPExARLS0tatmzJ1atXC3Xvfv75Zxo3boyhoSGurq6sW7eOjIwM9XhAQADVq1fHwMCAKlWqMGHCBADatWvH1atXmTRpEhqNRn0yLzQ0FEtLS63zPTw8WL16NdWrV8fU1JS3336bzMxMFi1ahJ2dHZUrV1Z/XjkWL16Mu7s7JiYmODg4MHbsWPXnFhERwbBhw0hOTlbbDggIALI3vp42bRpVq1bFxMSE5s2bExERUah7IUR+JMESQryUZsyYQWBgILNnzyY6Oprvv/8eW1tbHjx4oL5eKzIykh9//JFdu3Yxbtw4rfP37t3LpUuX2Lt3L2vWrCE0NFRd1nD8+HEmTJjA/PnziYmJ4eeff6ZevXrqubNmzSIkJISVK1fyxx9/MGnSJAYOHMi+ffu02pg5cyZBQUEcP34cPT09hg8frh4bMGAA1apVIzIyUn0rRmG3ScjIyKB37960bduW06dPc+TIEUaPHl2orQh+/fVXBg4cyIQJE4iOjmb58uXs2bNHfb3Zhg0bWLJkCZ9//jkXLlxgy5YtuLu7A7Bp0yaqVaumrruNj4/Pt51Lly6xfft2duzYwdq1a1m9ejXdunXj2rVr7Nu3j8DAQGbNmsXRo0fVc3R0dFi2bBlnz55lzZo17Nmzh2nTpgHZ63ODg4MxNzdX2546dSoAw4YN49ChQ6xbt47Tp0/Tt29ffHx8uHDhQqHupxB5UkQuycnJCqAkJSWVdiglIi0tTdmyZYuSlpZW2qEUO+lb2fU8+5eSkqIYGBgoq1atynXsiy++UKysrJTU1FS17JdfflF0dHSUhIQERVEUZciQIYqjo6OSkZGh1unbt6/i5+enKIqibNy4UTE3N1dSUlIURdHuW2pqqmJoaKgcPnxYq90RI0Yo/fv3VxRFUfbu3asAyq5du7RiAJSHDx8qiqIoZmZmSmhoaJ79mzt3rtKwYUOtsiVLliiOjo6KoijK7du3FUCJiIgo8F49rXXr1spHH32kfk9LS1P8/f0Ve3t7RVEUJSgoSHFxccn35+jo6KgsWbJEqywkJESxsLDQit/Y2Fi9f4qiKN7e3oqTk5OSmZmplrm6uioLFy7MN9YffvhBsba2zrcdRVGUixcvKhqNRrl+/bpWeceOHZUZM2YoSUlJCqAkJyfn244QeZE1WEKIl865c+d4/PgxHTt2zPNYw4YNMTExUctatmxJVlYWMTEx6qbF9erVU/fhA7C3t+fMmTMAdO7cGUdHR2rWrImPjw+dOnVS1xhFR0fz6NEjOnfurNVuWloanp6eWmUNGjTQuj5kb5xcvXp1Jk+ezMiRI/nmm2/o1KkTffv2pVatWoXqf8WKFRk6dCje3t507tyZTp060a9fP7WNZ4mKiiIyMlJrei49PZ20tDQePHhA3759CQ4OVvvu6+tLjx490NMr2q8bJycnrd3TbW1t0dXV1doL0dbWlsTERPX73r17+eijj4iOjiYlJYWMjAwePXrE/fv3tX6eTzpx4gSKouDi4qJV/vjx43K50bR4fmSKUAjx0nnWgmrlGa/WerL86ek4jUZDVlYWAGZmZpw4cYK1a9dib2/P/Pnz1afXcur88ssvnDx5Uv1ER0drrcN6uo2ctnPODwgI4I8//qBbt27s2bMHNzc3Nm/eDGRPlSn/3Xw5x5PrwwBCQkI4cuQIXl5erF+/HhcXF63ptvxkZWUxb948Ne7IyEiWLl1KdHQ0hoaGODg4EBMTw/LlyzEyMmLs2LG0adMmV/sFyev+PuueX716FV9fX+rXr8/GjRuJiopi+fLlefb96f7o6uoSFRWl9fM4d+4cS5cuLVLMQjxJEiwhxEundu3aGBkZsXv37lzH3NzcOHnyJPfv31fLDh06hI6OTq5RjmfR09OjU6dOLFq0iKioKBITE9m7dy9ubm4YGBgQFxeHs7Oz1sfBwaFI/XBxcWHSpEns3LmT1157jZCQEABsbGxISEjQSrJOnjyZ63xPT09mzJjB4cOHqV+/Pt9//32BbTZq1IiYmBituO3t7XF2dlZHl4yMjOjZsyfLli0jIiKCI0eOqKN7+vr6ZGZmFqmfhXH8+HEyMjIICgrilVdewcXFhRs3bmjVyattT09PMjMzSUxMzPXzsLOzK/Y4xctDpgiFEC8dQ0NDpk+fzrRp09DX16dly5bcunWLP/74gwEDBjB37lyGDBlCQEAAt27dYvz48QwaNOiZ7zR90rZt27h8+TJt2rTBysqKn3/+WZ2GMjMzY+rUqUyaNImsrCxatWpFSkoKhw8fxtTUlCFDhhR4/YcPH/Luu+/Sp08fatSowbVr14iMjOT1118Hsp/Wu3XrFosWLaJPnz7s2LGD7du3Y25uDkBsbCxffPEFPXv2pEqVKsTExHD+/HkGDx5cYNtz5syhe/fuODg40LdvXzIzMzl48CBHjx5l4cKFhIaGkpmZSfPmzTE2Nuabb77ByMhI3UDaycmJ/fv388Ybb2BgYEClSpUKdU8LUqtWLTIyMvj000/p0aMHhw4d4v/+7/+06jg5OZGamsru3btp2LAhxsbGuLi4MGDAAAYPHkxQUBCenp4kJSWxZ88e3N3dZZNq8Y/JCJYQ4qU0e/ZspkyZwpw5c6hbty5+fn4kJiZibGzMr7/+yp07d2jatCl9+vShY8eOfPbZZ4W+tqWlJZs2baJDhw7UrVuXL774gsmTJ6tPEi5YsIA5c+awcOFC6tati7e3Nz///DM1atQo1PV1dXW5ffs2gwcPxsXFhX79+tG1a1fmzZsHQN26dVmxYgXLly+nYcOG/Pbbb+oTcwDGxsb8+eefvP7667i4uDB69GjGjRvHW2+9VWDb3t7ebNu2jfDwcJo2bUrr1q356aef1ATK0tKSVatW0bJlSxo0aMDu3bv5+eef1fVM8+fP58qVK9SqVQsbG5tC39OCeHh4sHjxYgIDA6lfvz7fffed+mRjDi8vL8aMGYOfnx82NjYsWrQIyJ4uHTx4MFOmTMHV1ZWePXty7NixIo8oCvEkjfL0RL0gJSUFCwsLkpKSyuUix/T0dMLCwvD19S30Y91lhfSt7CrP/ZO+lV23b9+mUqVKJCcnqyOAQhSGjGAJIYQQQhQzSbCEEEKo6tWrh6mpaZ6f7777rrTDE6LMkEXuQgghVGFhYflua1DYRf5CCEmwhBDipXHlyhWuX79Oy5Yt862Ts1hdCPHvyBShEEK8BO7fv8+PP/7IzZs3SzsUIV4KkmAJIUQ5pygK27ZtA7K3WRBClDxJsIQQopw7ffo0f/75J927d8/3nXxCiOIlCZYQQpRjycnJbN++nQYNGlC3bt3SDkeIl4YkWEIIUU4pisLWrVvR19ena9eupR2OEC8VSbCEEKKcOn78OJcvX6ZXr14YGhqWdjhCvFQkwRJCiHJEo9GwZcsW7ty5Q3h4OE2aNKFWrVqlHdYzBQQE4OHhUdphCFGsJMESQohyJisriy1btmBqakrnzp1LOxwtOQngk6ZOncru3btLvG1J5MTzJBuNCiFEORMTE0NaWhrDhg1DX1+/tMMpUM6reIQoT2QESwghyhAnJyeCg4O1yjw8PAgICFC/nzlzBi8vL7KystBoNPzwww+0bt0aIyMjmjZtyvnz54mMjKRJkyaYmpri4+PDrVu3Ch1DSEgIdevWxdDQkPr16xMWFqYeS0tLY9y4cdjb22NoaIiTkxMLFy5UYwd49dVX0Wg06venR5aGDh1K7969+eijj7C1tcXS0pJ58+aRkZHBu+++S8WKFalWrRqrV6/Wimv69Om4uLhgbGxMzZo1mT17tvran9DQUObNm8epU6fQaDRoNBpCQ0OB7CctR48eTeXKlTE3N6dDhw6cOnWq0PdDiLzICJYQQpQTmZmZAJiZmdG+fXuuXbsGwNy5cwkODqZ69eoMHz6c/v37Y25uztKlSzE2NqZfv37MmTOHlStXFtjGqlWrmDt3Lp999hmenp5ERkYycuRImjVrxogRI1i2bBlbt27lhx9+oHr16vz111/89ddfAERGRlK5cmVCQkLw8fFBV1c333b27NlDtWrV2L9/P4cOHWLEiBEcOXKENm3acOzYMdavX8+YMWPo3LkzDg4Oar9DQ0OpUqUKZ86cYdSoUZiZmTFt2jT8/Pw4e/YsO3bsYNeuXQBYWFigKArdunWjYsWKhIWFYWFhweeff07Hjh05f/78v/p5iJebJFhCCFFO7N+/H4DmzZujp/e/f96nTp2q7uA+ceJE+vfvz+7du9V3Eo4YMUIdzSnIggULCAoK4rXXXgOgWrVq/PTTT3z55ZeMGDGCuLg4ateuTatWrdBoNFrvNrSxsQHA0tISOzu7Z7ZTsWJFli1bho6ODq6urixatIgHDx7w/vvvAzBjxgw+/vhjDh06xBtvvAHArFmz1POdnJyYMmUK69evZ9q0aRgZGWFqaoqenp5W23v27OHMmTMkJiZiYGAAwCeffMKWLVvYsGEDr7/+eqHuixBPkwRLCCHKgZSUFA4cOABkJydPatCggfpnW1tbANzd3bXKEhMTC2zj1q1b/PXXX4wYMYJRo0ap5WlpaVhZWQHZ03udO3fG1dUVHx8funfvTpcuXYrcn3r16qGj879VLLa2ttSvX1/9rquri7W1tVbcGzZsIDg4mIsXL5KamkpGRgbm5ubPbCcqKorU1FSsra21yh8+fMilS5eKHLcQOSTBEkKIMkRHRwdFUcjMUvgt9g6Jfz8i+f5D/vzzT3r37p3nORUqVFD/rNFo8izLysoqsO2cOqtWraJ58+YApKens2/fPjp06ABAo0aNiI2NZfv27ezatYt+/frRqVMnNmzYUKR+PhlfTox5leXEdPToUd544w3mzZuHt7c3FhYWrFu3jqCgoAL7ZG9vT0RERK5jlpaWRYpZiCdJgiWEEGWIjY0Nh89c4IfAPcQnPyLr8QNuXImletUqVKrfqkTbtrW1pWrVqly+fJkBAwYA2QnW+fPnqVGjhlrP3NwcPz8//Pz86NOnDz4+Pty5c4eKFStSoUIFda1YcTp06BCOjo7MnDlTLbt69apWHX19/VxtN2rUiISEBPT09NRF90+6fft2sccqXg6SYAkhRBni5N6Mjeu+o1IvR3QMTbl34Fs0Ojpcy7RgytbYEm8/ICCACRMmYG5uTteuXUlNTWX37t2cP3+ed999lyVLlmBvb4+Hhwc6Ojr8+OOP2NnZqaNBTk5O6vovAwMDdWrx33J2diYuLo5169bRtGlTfvnlFzZv3qxVx8nJidjYWE6ePEm1atUwMzOjU6dOtGjRgt69exMYGIirqys3btwgLCyM3r17ayWOQhSFbNMghBBlRGaWwuWqnTFwqE/ihvkk/jgP49ot0FhW4bZijPLfeufiU8jMUp55rX9q5MiRfPnll4SGhuLu7k6nTp3Ys2ePOvpjampKYGAgTZo0oWnTply5coWwsDB1PVVQUBDh4eE4ODjg6elZbHH16tWLSZMmMW7cODw8PDh8+DCzZ8/WqvP666/j4+ND+/btsbGxYe3atWg0GsLCwmjTpg3Dhw/HxcWFN954gytXrqjr1YT4JzSKopTM/wvLsJSUFCwsLEhKSsq18LE8SE9PJywsDF9f31xrGso66VvZVZ77V1x9O3LpNv1XHS1UXXsLQ+b2cMOnvv0/bq8wyvPPDbKnCCtVqkRycnKBC+aFeJKMYAkhRBmR+PejQtdNSH7E29+eYMfZ+BKMSAiRH0mwhBCijKhsZljImgpOund4pcJVdm76jujoc4VuI+e1NXl9craBEEIUTBa5CyFEGdGsRkXsLQxJSH7Es9Z2WGse0KbCZVIUQxLSTYlPM8CtkG2cPHky32NVq1YtSrhCvNQkwRJCiDJCV0fD3B5uvP3tCTSQb5J1WzHh60eNUP47SeGHQaHbcHZ2/veBCiFkilAIIcoSn/r2rBzYCDuLZ08XKk/88174qUUhRHGRBEsIIcoYn/r2HJzege9GNMfSKP8n9zRkP03YrEbFfOsIIUqGJFhCCFEG6epoaFm7Eh+/7o6G7GTqSTnf5/ZwQ1fn6aNCiJImCZYQQpRh+U0Z2lkYsnJgoxLfB0sIkTdZ5C6EEGWcT317OrvZqS9/rmyWPS0oI1dClB5JsIQQohzQ1dHQolb5e/OEEGWVTBEKIYQQQhQzSbCEEEIIIYpZqSdYK1asoEaNGhgaGtK4ceNnvoph6NChaDSaXJ969eqpdUJDQ/Os8+hR4d/hJYQQQgjxb5RqgrV+/Xr8/f2ZOXMmv//+O61bt6Zr167ExcXlWX/p0qXEx8ern7/++ouKFSvSt29frXrm5uZa9eLj4zE0lI32hBAvnoiICDQaDffu3SvtUIqkXbt2+Pv7l3YYQrywSjXBWrx4MSNGjGDkyJHUrVuX4OBgHBwcWLlyZZ71LSwssLOzUz/Hjx/n7t27DBs2TKueRqPRqmdnZ/c8uiOEEOVOfgngpk2bWLBgQYm3L4mcKKtK7SnCtLQ0oqKieO+997TKu3TpwuHDhwt1ja+++opOnTrh6OioVZ6amoqjoyOZmZl4eHiwYMECPD09873O48ePefz4sfo9JSUFgPT0dNLT0wvbpTIjp0/St7KlPPcNynf/ntW3jIwM9diL2Pf84jMzM1PLn/zf4qYoCllZWaV2b17En4koG0otwUpKSiIzMxNbW1utcltbWxISEgo8Pz4+nu3bt/P9999rldepU4fQ0FDc3d1JSUlh6dKltGzZklOnTlG7du08r7Vw4ULmzZuXq3zv3r0YGxsXoVdlS3h4eGmHUGKkb2VXWe/fzJkzcXR0REdHh71796Knp8ebb75J27Zt6dGjB4cPH8bCwoLRo0fTuHFjzpw5A8DOnTsxNTVl9+7dfPXVV0yaNImQkBCSkpJo3LgxEydO5PDhw6xdu5YHDx7Qtm1bRowYga6uboExpaen891337F//37u379P9erVGTx4MO7u7gAkJibyxRdfcO7cOTIyMqhcuTJDhgzBwcGBt956C4DKlSsD0L59eyZOnMjMmTOpUaMGI0eOBMDJyYnOnTtz48YNjh49ipmZGSNHjqROnTosX76c06dPU7lyZSZMmKC+UDolJYVVq1YRHR3N33//jZ2dHX369KFNmzZA9rKQ/fv3s3//fj799FMAPv/8c2xtbfnrr78ICQkhOjoaAwMDPDw8GDFiBObm5sX404QHDx4U6/XEy6PU98HSaLQ3wlMUJVdZXkJDQ7G0tKR3795a5a+88gqvvPKK+r1ly5Y0atSITz/9lGXLluV5rRkzZjB58mT1e0pKCg4ODrRv3x5r6/K3r0x6ejrh4eF07tyZChXyf49ZWSR9K7vKS/8WL17MgQMHmDJlCoGBgfz444/Mnz+fY8eOMXz4cPXfohUrVnDp0iVMTEyA7NF7S0tLkpKSSE9P59ixY2zevJnU1FT69evHV199hYWFBeHh4cTGxuLn58ebb75Jv379Coxp8ODB3Lx5kw0bNmBvb89PP/3E3LlzOXHiBLVr16Z3795YWlqyf/9+TExMOHfuHObm5nh5eWFpaYmfnx9nz57F3NwcIyMjLCwsWLx4MTVq1KBz586Eh4djZGTEjh07WLBgAcuXL2fZsmV89tlneHl54e/vT4MGDXj//ff58ssvOXXqFBqNhuvXr3Pjxg2CgoIwNzdn+/btTJ06lddee41mzZrRsmVLevToQb169Zg7dy4ANjY2JCYmMnLkSIYPH05oaCgPHz5k5syZrF69mp07dxbrz/P27dvFej3xElFKyePHjxVdXV1l06ZNWuUTJkxQ2rRp88xzs7KyFGdnZ8Xf379QbY0cOVLx8fEpdGzJyckKoCQlJRX6nLIkLS1N2bJli5KWllbaoRQ76VvZVV7617ZtW6VVq1bq94yMDMXExERp166d2rf4+HgFUI4cOaLs3btXAZS7d+8qiqIoISEhCqBcvHhRvcZbb72lGBsbK3///bda5u3trbz11lsFxnPx4kVFo9Eo169f1yrv2LGjMmPGDEVRFMXd3V0JCAjI8/yn43uynxMnTlR/bo6OjsrAgQPV4zl9nD17tlp25MgRBVDi4+PzjdfX11eZMmVKrnaeNHv2bKVLly5aZX/99ZcCKDExMfle+59ISkpSACU5OblYryvKv1Jb5K6vr0/jxo1zTQeEh4fj5eX1zHP37dvHxYsXGTFiRIHtKIrCyZMnsbeX93EJIZ6PBg0aqH/W1dXF2tpaa61oztKIxMTEPM83NjamVq1aWvWdnJwwNTXVKsvv/CedOHECRVFwcXHB1NRU/ezbt49Lly4BMGHCBD744ANatmzJ3LlzOX36dNE6/F9P9junjznTkE+W5cSdmZnJhx9+SIMGDbC2tsbU1JSdO3fm+yR5jqioKPbu3avVnzp16gCofRKitJXqFOHkyZMZNGgQTZo0oUWLFnzxxRfExcUxZswYIHvq7vr163z99dda53311Vc0b96c+vXr57rmvHnzeOWVV6hduzYpKSksW7aMkydPsnz58ufSJyGEeHqKU6PRaK2VylkGkZWVVejz8yrL7/wnZWVloaurS1RUVK71WjkJ28iRI/H29uaXX35h586dLFy4kKCgIMaPH1/g9fOLO6ePeZXlxB0UFMSSJUsIDg7G3d0dExMT/P39SUtLK7BPPXr0IDAwMNcx+Y9p8aIo1QTLz8+P27dvM3/+fOLj46lfvz5hYWHqf+nFx8fn+i+Z5ORkNm7cyNKlS/O85r179xg9ejQJCQlYWFjg6enJ/v37adasWYn3RwghXjSenp5kZmaSmJhI69at863n4ODAmDFjGDNmDDNmzGDVqlWMHz8efX19IHu0qbgdOHCAXr16MXDgQCA7cbpw4QJ169ZV6+jr6+dqu1GjRmzcuBEnJyf09Ep9KbEQeSr1v5ljx45l7NixeR4LDQ3NVWZhYfHMpzqWLFnCkiVLiis8IYQo01xcXBgwYACDBw8mKCgIT09PkpKS2LNnD+7u7vj6+uLv70/Xrl1xcXHh7t277NmzR01yHB0d0Wg0bNu2DV9fX4yMjLSmKv8NZ2dnNm7cyOHDh7GysmLx4sUkJCRoJVhOTk4cO3aMK1euYGpqSsWKFXnnnXdYtWoV/fv3591336VSpUpcvHiRdevWsWrVqkI9WSlESSv1V+UIIYQoWSEhIQwePJgpU6bg6upKz549OXbsGA4ODkD26NQ777xD3bp18fHxwdXVlRUrVgBQtWpV5s2bx3vvvYetrS3jxo0rtrhmz55No0aN8Pb2pl27dtjZ2eV6Mnzq1Kno6uri5uaGjY0NcXFxVKlShUOHDpGZmYm3tzf169dn4sSJWFhYoKMjv9bEi0GjKIpS2kG8aFJSUrCwsCApKancbtMQFhaGr69vmX4cPi/St7KrPPdP+lZ23b59m0qVKpGcnFzse2yJ8k1SfSGEEEKIYiYJlhBClAGpqamkpqbmKj9w4IDWdgVPf4QQpaPUF7kLIYTIW1JSEjExMfz5559cu3YNZ2dnBgwYoFWnSZMmnDx5snQCFELkSxIsIYR4QSiKwvXr1/nzzz+JiYkhKSkJPT09nJ2d6dWrF66urrnOMTIyUt/tJ4R4cUiCJYQQpSgjI4PY2Fj+/PNPzp8/T2pqKsbGxri4uNCpUydq1qxZLhePC1HeSYIlhBDP2cOHD7lw4QIxMTFcvHiRtLQ0rKyscHd3p06dOlSrVk22GxCijJMESwghiiAzS+G32Dsk/v2IymaGNKtREV0dTYHnpaWlERkZycWLF7l69SpZWVlUrVqVVq1a4erqio2NjfoqGSFE2ScJlhBCFNKOs/HM+zma+ORHapm9hSFze7jhU1/7HXiKonDz5k1iYmI4d+4cN2/e5M8//6RGjRrqZp6yr5IQ5ZckWEIIUQg7zsbz9rcneHpn5oTkR7z97QlWDmxEFzdb4uLi1EXq9+7dw8DAgFq1amFoaMjrr7+OmZlZqcQvhHi+JMESQogCZGYpzPs5OldyBaBLJlV1Uti0eQunw/7m4cOHmJmZ4erqSp06dXByciIrK4uwsDAMDQ2fe+xCiNIhCZYQQhTgt9g7WtOCT+pqEEMlnQfczTDCzq0enVp4Ym9vr7WeKisr63mFKoR4QUiCJYQQBUj8O+/kCuBQmhPp6PC3Ykj7Wh5UqVLlOUYmhHhRSYIlhBAFqGyW/9TeHcW4UPWEEC8X2WhFCCEK0KxGRewtDMlvEwUN2U8TNqtR8XmGVW6EhoZiaWlZ2mEIUawkwRJCiALo6miY28MNIFeSlfN9bg+3Qu2H9bJzcnIiODhYq8zPz4/z58+XeNuSyInnSRIsIYQoBJ/69qwc2Ag7C+1pQDsLQ1YObJRrHyxReEZGRlSuXLm0wxCiWEmCJYR4KWVlZREYGIizszMGBgZUr16dDz/8EIAzZ87QoUMHjIyMsLa2ZvTo0aSmpuJT356D0ztQ78K32B77lB6a41xZNpABbevzzjvvkJ6erl5/xYoV1K5dG0NDQ6pVq0ZgYKB6TFEUFi1aRM2aNTEyMqJhw4Zs2LBBPR4REYFGo2H37t00adIEY2NjvLy8iImJUeucOnWK9u3bY2Zmhrm5OY0bN+b48eMABAQE4OHhodXf4OBgnJyctNpo1qwZJiYmWFpa0rJlS65evVqoe/fzzz/TuHFjDA0NcXV1Zd26dWRkZKjHAwICqF69OgYGBlSpUoUJEyYA0K5dO65evcqkSZPQaDTqk5ZPjyzlxL969WqqV6+Oqakpb7/9NpmZmSxatAg7OzsqV66s/rxyLF68GHd3d0xMTHBwcGDs2LGkpqaq/R02bBjJyclq2wEBAUD2LvvTpk2jatWqmJiY0Lx5cyIiIgp1L4TIjyRYQoiX0owZMwgMDGT27NlER0fz/fffY2try4MHD/Dx8cHKyorIyEh+/PFHdu3axbhx44Ds6UIbMwP+PHGEjHsJ7N27lzVr1hAaGkpoaCgAx48fZ8KECcyfP5+YmBh+/vln6tWrp7Y9a9YsQkJCWLlyJX/88QeTJk1i4MCB7Nu3TyvGmTNnEhQUxPHjx9HT02P48OHqsQEDBlCtWjUiIyOJiorivffeK/RLoTMyMujduzdt27bl9OnTHDlyhNGjRxfqVT2//vorAwcOZMKECURHR7N8+XL27NnDwoULAdiwYQNLlizh888/58KFC2zZsgV3d3cANm3aRLVq1Zg/fz7x8fHEx8fn286lS5fYvn07O3bsYO3ataxevZpu3bpx7do19u3bR2BgILNmzeLo0aPqOTo6OixbtoyzZ8+yZs0a9uzZw7Rp0wDw8vIiODgYc3Nzte2pU6cCMGzYMA4dOsS6des4ffo0ffv2xcfHhwsXLhTqfgqRJ0XkkpycrABKUlJSaYdSItLS0pQtW7YoaWlppR1KsZO+lV3/uH+ZGYpyeb+inP4x+38zMwo8JSUlRTEwMFBWrVqV69gXX3yhWFlZKampqWrZL7/8oujo6CgJCQmKoijKkCFDFEdHRyUj439t9e3bV/Hz81MURVE2btyomJubKykpKbn6lpqaqhgaGiqHDx/WanfEiBFK//79FUVRlL179yqAsmvXLq0YAOXhw4eKoiiKmZmZEhoammf/5s6dqzRs2FCrbMmSJYqjo6OiKIpy+/ZtBVAiIiIKvFdPa926tfLRRx+p39PS0hR/f3/F3t5eURRFCQoKUlxcXPL9OTo6OipLlizRKgsJCVEsLCy04jc2Nlbvn6Ioire3t+Lk5KRkZmaqZa6ursrChQvzjfWHH35QrK2t821HURTl4sWLikajUa5fv65V3rFjR2XGjBlKUlKSAijJycn5tiNEXmSbBiFE2RW9FXZMh5Qb/yszrwI+geDWM9/Tzp07x+PHj+nYsWOexxo2bIiJiYla1rJlS7KysoiJicHW1haAevXqoaurq9axt7fnzJkzAHTu3BlHR0dq1qyJj48PnTp1wsjIKDvk6GgePXpE586dtdpNS0vD09NTq6xBgwZa1wdITEykevXqTJ48mZEjR/LNN9/QqVMn+vbtS61atZ55u3JUrFiRoUOH4u3tTefOnenUqRP9+vVT23iWqKgoIiMjtabn0tPTSUtL48GDB/Tt25fg4GC1776+vvTo0QM9vaL9unFyctJ6rZCtrS26urro6OholSUmJqrf9+7dy0cffUR0dDQpKSlkZGTw6NEj7t+/r/XzfNKJEydQFAUXFxet8sePH2NtbV2kmIV4kkwRCiHKpuit8MNg7eQKICU+uzx6a76n5iQ7eVEUJd+psifLn56O02g06o7tZmZmnDhxgrVr12Jvb8/8+fPx9/fn3r17ap1ffvmFkydPqp/o6GitdVhPt5HTds75AQEB/PHHH3Tr1o09e/bg5ubG5s2bgeypMkXRfrHPk+vDAEJCQjhy5AheXl6sX78eFxcXrem2/GRlZTFv3jw17sjISJYuXUp0dDSGhoY4ODgQExPD8uXLMTIyYuzYsbRp0yZX+wXJ6/4+655fvXoVX19f6tevz8aNG4mKimL58uV59v3p/ujq6hIVFaX18zh37hxLly4tUsxCPEkSLCFE2ZOVmT1ylefbAbPLlO3vkf44e/QiZ6Fzjtq1a2NkZMTu3btzne3m5sbJkye5f/++Wnbo0CF0dHRyjXI8i56eHp06dWLRokVERUWRmJjI3r17cXNzw8DAgLi4OJydnbU+Dg4Ohb4+gIuLC5MmTWLnzp289tprhISEAGBjY0NCQoJWknXy5Mlc53t6ejJjxgwOHz5M/fr1+f777wtss1GjRsTExGjFbW9vj7Ozszq6ZGRkRM+ePVm2bBkREREcOXJEHd3T19cnMzOzSP0sjOPHj5ORkUFQUBCvvPIKLi4u3LihnXzn1banpyeZmZkkJibm+nnY2dkVe5zi5SFThEKIsufq4dwjV//1J7X4CW8e/62P8vH/ntx766231F+YhoaGTJ8+nWnTpqGvr0/Lli25desWf/zxBwMGDGDu3LkMGTKEgIAAbt26xfjx4xk0aJA6PViQbdu2cfnyZdq0aYOVlRU///yzOg1lZmbG1KlTmTRpEllZWbRq1YqUlBQOHz6MqakpQ4YMKfD6Dx8+5N1336VPnz7UqFGDa9euERkZyeuvvw5kP61369YtFi1aRJ8+fdixYwfbt2/H3NwcgNjYWL744gt69uxJlSpViImJ4fz58wwePLjAtufMmUP37t1xcHCgb9++ZGZmcvDgQY4ePcrChQsJDQ0lMzOT5s2bY2xszDfffIORkRGOjo5A9tTf/v37eeONNzAwMKBSpUqFuqcFqVWrFhkZGXz66af06NGDQ4cO8X//939adZycnEhNTWX37t00bNgQY2NjXFxcGDBgAIMHDyYoKAhPT0+SkpLYs2cP7u7uNG/evFjiEy8fGcESQpQ9qTfzPWTPTVpzDF/28GqTqvj5+TF8+PBcydHs2bOZMmUKc+bMoW7duvj5+ZGYmIixsTG//vord+7coWnTpvTp04eOHTvy2WefFTo8S0tLNm3aRIcOHahbty5ffPEFkydPVp8kXLBgAXPmzGHhwoXUrVsXb29vfv75Z2rUqFGo6+vq6nL79m0GDx6Mi4sL/fr1o2vXrsybNw+AunXrsmLFCpYvX07Dhg357bff1CfmAIyNjfnzzz95/fXXcXFxYfTo0YwbN4633nqrwLa9vb3Ztm0b4eHhNG3alNatW/PTTz+pCZSlpSWrVq2iZcuWNGjQgN27d/Pzzz+r65nmz5/PlStXqFWrFjY2NoW+pwXx8PBg8eLFBAYGUr9+fb777jv1ycYcXl5ejBkzBj8/P2xsbFi0aBGQPV06ePBgpkyZgqurKz179uTYsWNFHlEU4kka5emJekFKSgoWFhYkJSWVy0WO6enphIWF4evrW+jHussK6VvZVaT+xR6ANd0LvuiQbVCjdfEE+C+U559dee4bwO3bt6lUqRLJycnqCKAQhSEjWEKIssfRK/tpwWe9HdC8anY9IYQoBZJgCSFeTFmZ2SNVZzZk/2/WE4uTdXSzt2IA8n07oM/H2fVEkdSrVw9TU9M8P999911phydEmSGL3IUQL57C7G/l1hP6fZ1PvY+fuQ+WyF9YWFi+2xoUdpG/EEISLCHEC0bz5zbYOIxcWzDk7G/V72vtJKtOt+ynClNvgqlt9rSgjFz9YzmL1YUQ/45MEQohXhxKFro73+dZ+1ux473c04U1WoN7n+z/zSe5ynmB8r1794o76hLVrl07/P39SzsMIUQRSYIlhHhhWKfGoPn7f9N9DzDkHM5coRopmAIKpFzPHrEqZ/JLADdt2sSCBQtKvH1J5IQoXjJFKIR4YRim39P6fo7abNNkv7PPWHnAu/x348hn7INV3lSsWLG0QxBC/AMygiWEeGE8qmCp9b0xZ5imrGCcspoxfPO/A6a2tGvXjvHjx+Pv74+VlRW2trZ88cUX3L9/n2HDhmFmZkatWrXYvn17nm2FhoZiaWnJtm3bcHV1xdjYmD59+nD//n3WrFmDk5MTVlZWjB8/vtCvdklLS2PatGlUrVoVExMTmjdvTkREhHr86tWr9OjRAysrK0xMTKhXrx5hYWFcuXKF9u3bA2BlZYVGo2Ho0KFA7pElJycnPvjgAwYPHoypqSmOjo789NNP3Lp1i169emFqaoq7uzvHjx9Xz7l9+zb9+/enWrVqGBsb4+7uztq1a9XjQ4cOZd++fSxduhSNRoNGo+HKlStA9supfX19MTU1xdbWlkGDBpGUlFSo+yHEy0wSLCHEC+O2qSuKmfb+VkY8wpp7mHGfp/e3WrNmDZUqVeK3335j/PjxvP322/Tt2xcvLy9OnDiBt7c3gwYN4sGDB3m29+DBA5YtW8a6devYsWMHERERvPbaa4SFhREWFsY333zDF198keslzPkZNmwYhw4dYt26dZw+fZq+ffvi4+PDhQsXAJg4cSKPHz9m//79nDlzhsDAQExNTXFwcGDjxo0AxMTEEB8f/8wXDS9ZsoSWLVvy+++/061bNwYNGsTgwYMZOHAgJ06cwNnZmcGDB6vvInz06BGNGzdm27ZtnD17ltGjRzNo0CCOHTsGwNKlS2nRogWjRo0iPj6e+Ph4HBwciI+Pp23btnh4eHD8+HF27NjBzZs36devX6HuhxAvM5kiFEK8ODQ6ZHb5CL2Nw8hOsp5c7J57f6uGDRsya9YsAGbMmMHHH39MpUqVGDVqFJD93ryVK1dy+vTpPJtLT09n5cqV1KpVC4A+ffrwzTffcPPmTUxNTXFzc6N9+/bs3bsXPz+/Z4Z+6dIl1q5dy7Vr16hSpQoAU6dOZceOHaxZswYvLy/i4uLo06cP7u7uANSsWVM9P2cqsHLlylhaWj6zLV9fX/W1Njl9bNq0KX379gVg+vTptGjRgps3b2JnZ0fVqlW1XpUzfvx4duzYwY8//kjz5s2xsLBAX18fY2NjrRccr1y5kkaNGvHRRx+pZatXr8bBwYHz588X6eXXQrxsZARLCPFCUep0z96Kwdxe+4B5Fe0tGoAGDRqof9bV1cXa2lpNXuB/+zYlJibm2ZaxsbGaXOXUd3JywtTUVKssv/OfdOLECfWFzk9uzrlv3z4uX74MwLhx4/jggw9o2bIlc+fOzTfxK8iT/c7p47P6nZmZyYcffkiDBg2wtrbG1NSUnTt3EhcX98x2oqKi2Lt3r1Z/6tSpA2QnlEKI/MkIlhDixVPI/a2efvedRqPRKtNoske9srKy8mymoPNzyvI7/0lZWVno6uoSFRWFrq52nAYGBpw4cYLhw4fj6+vLL7/8ws6dO1m4cCFBQUGMHz++wOvnF3dOH5/V76CgIJYsWUJwcDDu7u6YmJjg7+9PWlpagX3q0aMHgYGBuY7Z29vncYYQIockWEKIF1PO/lZlhKenJ5mZmSQmJtK6tXbcT+6M7uDgwJgxYxgzZgwzZsxg1apVjB8/Hn19fYBCL6gvigMHDtCrVy8GDhwIZCdOFy5coG7dumodfX39XG03atSIjRs34uTkhJ6e/LoQoihkilAIIYqBi4sLAwYMYPDgwWzatInY2FgiIyMJDAxUn2ScMmUKv/76K7GxsZw4cYI9e/aoSY6joyMajYZt27Zx69YtUlNTiy02Z2dnwsPDOXz4MOfOneOtt94iISFBq46TkxPHjh3jypUrJCUlkZWVxTvvvMOdO3fo378/v/32G5cvX2bnzp0MHz68RBJBIcoTSbCEEKKYhISEMHjwYKZMmYKrqys9e/bk2LFjVKtWDcgenXrnnXeoW7cuPj4+uLq6smLFCgCqVq3KvHnzeO+997C1tWXcuHHFFtfs2bNp1KgR3t7etGvXDjs7O3r37q1VZ+rUqejq6uLm5oaNjQ1xcXFUqVKFQ4cOkZmZibe3N/Xr12fixIlYWFigoyO/PoR4Fo2S8xyvUKWkpGBhYUFSUhLW1talHU6xS09PJywsDF9f31zrTco66VvZVWD/sjLL7DsHy/PPrjz3DbL3EKtUqRLJycmYm5uXdjiiDJFJdSHEiy96K+yYDin/e40O5lXAJ1DrqUIhhHhRlPoY74oVK6hRowaGhoY0btyYAwcO5Ft36NCh6i7DT37q1aunVW/jxo24ublhYGCAm5sbmzdvLuluCCFKSvRW+GGwdnIFkBKfXR699bmEceDAAa3tCp7+CCHEk0p1BGv9+vX4+/uzYsUKWrZsyeeff07Xrl2Jjo6mevXqueovXbqUjz/+WP2ekZFBw4YN1c31AI4cOYKfnx8LFizg1VdfZfPmzfTr14+DBw/SvHnz59IvIUQxycrMHrkir5UMCqCBHe9lb+lQwtOFTZo04eTJkyXahhCi/CjVBGvx4sWMGDGCkSNHAhAcHMyvv/7KypUrWbhwYa76FhYWWFhYqN+3bNnC3bt3GTZsmFoWHBxM586dmTFjBpC9u/O+ffsIDg7WeveWEKIMuHo498iVFgVSrmfXK+EtHYyMjHB2di7RNoQQ5UepJVhpaWlERUXx3nvvaZV36dKFw4cPF+oaX331FZ06dcLR0VEtO3LkCJMmTdKq5+3tTXBwcL7Xefz4MY8fP1a/p6SkANmLN5/cv6a8yOmT9K1sKc99g7z7p0m+Xqh/pDKSr6O8wPelPP/synPfoPz2S5S8UkuwkpKSyMzMVF/pkMPW1jbX/ix5iY+PZ/v27Xz//fda5QkJCUW+5sKFC5k3b16u8r1792JsbFxgLGVVeHh4aYdQYqRvZdeT/bP++wqtCnHO0bNXuH01rOSCKibl+WdXXvuW34vChShIqT9FmPNKhxyKouQqy0toaCiWlpa59nL5J9ecMWMGkydPVr+npKTg4OBA+/bty+02DeHh4XTu3LncPVYtfSu78uxfljfKZ2vg73g0eazDUtCAeRWa9/XPtQZLX1+fH3/8kV69ej2H6J+tsD+7+fPns3XrVo4fP/4co/t3yvvfy9u3b5d2CKKMKrUEq1KlSujq6uYaWUpMTMw1AvU0RVFYvXo1gwYNUl8vkcPOzq7I1zQwMMDAwCBXeYUKFcrlPxg5ynP/pG9ll3b/KkDXwOynBdGgvdhdgwbA52MqGBjmeS09Pb0X6l492TeNRsPmzZu1/iNx+vTp+Pv7l3jMAQEBbNmypVgX7ZfXv5flsU/i+Si1bRr09fVp3LhxrmHl8PBwvLy8nnnuvn37uHjxIiNGjMh1rEWLFrmuuXPnzgKvKYR4Qbn1hH5fg/lTLxc2r5JdXo72wTI1NS2Xo+ZCvIxKdR+syZMn8+WXX7J69WrOnTvHpEmTiIuLY8yYMUD21N3gwYNznffVV1/RvHlz6tevn+vYxIkT2blzJ4GBgfz5558EBgaya9cu/P39S7o7QoiS4tYT/M/CkG3w+lcwZBtOyx4QvPOyVjUPDw8CAgJynX7lyhU0Gg0//PADrVu3xsjIiKZNm3L+/HkiIyNp0qQJpqam+Pj4cOvWrUKHFRISQt26dTE0NKROnTrqa28g+0GecePGYW9vj5mZGaNGjSIwMBDIfu8fwKuvvopGo1G/BwQE4OHhoV5j6NCh9O7dm48++ghbW1ssLS2ZN28eGRkZvPvuu1SsWJFq1aqxevVqrbimT5+Oi4sLxsbG1KxZk9mzZ6uLtUNDQ5k3bx6nTp1S9xIMDQ0FIDk5mdGjR1O5cmXMzc3p0KEDp06dKvT9EEL8T6muwfLz8+P27dvMnz+f+Ph46tevT1hYmPpUYHx8PHFxcVrnJCcns3HjRpYuXZrnNb28vFi3bh2zZs1i9uzZ1KpVi/Xr18seWEKUdTq6/3orhrlz5xIcHEz16tUZPnw4/fv3x9zcnKVLl2JsbEy/fv2YM2cOK1euLPBaq1atYu7cuXz22Wd4enry+++/M2rUKExMTBgyZAjLli1j69at/PDDD9jb2/Pjjz+q7ySMjIykcuXKhISE4OPjg65u/nt47dmzh2rVqrF//34OHTrEiBEjOHLkCG3atOHYsWOsX7+eMWPG0LlzZxwcHAAwMzMjNDSUKlWqcObMGUaNGoWZmRnTpk3Dz8+Ps2fPsmPHDnbt2gVkb4GjKArdunWjYsWKhIWFYWFhweeff07Hjh05f/48FStW/Ff3XoiXjiJySU5OVgAlKSmptEMpEWlpacqWLVuUtLS00g6l2Enfyq6i9s/R0VFZsmSJVlnDhg2VuXPnKoqiKICyefNmRVEUJTY2VgGUL7/8Uq27du1aBVB2796tli1cuFBxdXUtVPsODg7K999/r1W2YMECpUWLFoqiKMr48eOVDh06KFlZWXn27cn4csydO1dp2LCh+n3IkCGKo6OjkpmZqZa5uroqrVu3Vr9nZGQoJiYmytq1a/ONddGiRUrjxo3zbUdRFGX37t2Kubm58ujRI63yWrVqKZ9//nm+1y7vfy+TkpIUQElOTi7tUEQZU+pPEQohxPPSoEED9c85D764u7trlSUmJhZ4nVu3bvHXX38xYsQIRo0apZZnZGSomyEPHTqUzp074+rqSpcuXahUqRK+vr5FjrlevXro6PxvNYetra3W8ghdXV2sra214t6wYQPBwcFcvHiR1NRUMjIyCnxRcVRUFKmpqbnWgD18+JBLly4VOW4hXnaSYAkhyiQdHR0URXvrhoI2hXzyibCcrVueLsvKyiqw7Zw6q1atyrX8IGe6r1GjRsTGxrJ9+3Z27tzJf/7zH06fPs2mTZsKvH5+MefEmFdZTkxHjx7ljTfeYN68eXh7e2NhYcG6desICgoqsE/29vZERETkOmZpaVmkmIUQkmAJIcooGxsb4uPj1e8pKSnExsY+l7ZtbW2pWrUqly9fZsCAAfnWMzc3x8/Pj9deew0HBwfmzZvHnTt3qFixIhUqVCAzM7PYYzt06BCOjo7MnDlTLbt69apWHX19/VxtN2rUiISEBPT09NRF90KIf04SLCFEmdShQwdCQ0Pp0aMHVlZWzJ49+5mLxYtbQEAAEyZMwNzcnK5du/L48WOOHz/O3bt3mTx5MkuWLMHe3h4PDw8yMzM5dOgQdnZ26miQk5MTu3fvpmXLlhgYGGBlZVUscTk7OxMXF8e6deto2rQpv/zyC5s3b9aq4+TkRGxsLCdPnqRatWqYmZnRqVMnWrRoQe/evQkMDMTV1ZUbN24QFhZG7969adKkSbHEJ8TLolS3aRBCiH9qxowZtGnThu7du+Pr60vv3r2pVavWc2t/5MiRfPnll4SGhuLu7k7btm0JDQ2lRo0aQPaeVoGBgTRp0gQvLy8SExP56aef1PVUQUFBhIeH4+DggKenZ7HF1atXLyZNmsS4cePw8PDg8OHDzJ49W6vO66+/jo+PD+3bt8fGxoa1a9ei0WgICwujTZs2DB8+HBcXF9544w2uXLlS4ObPQojcNMrTixgEKSkpWFhYkJSUVC43/UtPTycsLAxfX99yt0ux9K3sKs/9k76VXbdv36ZSpUokJycX+KCAEE+SESwhhBBCiGImCZYQQuTB1NQ038+BAwdKOzwhxAtOFrkLIUQenvUi5KpVqz6/QIQQZZIkWEIIkQdnZ+fSDkEIUYbJFKEQQgghRDGTBEsIIYQQophJgiWEEEIIUcwkwRJCCCGEKGaSYAkhhBBCFDNJsIQQQgghipkkWEIIIYQQxUwSLCGEEEKIYiYJlhBCCCFEMZMESwghhBCimEmCJYQQQghRzCTBEkIIIYQoZpJgCSGEEEIUM0mwhBBCCCGKmSRYQgghhBDFTBIsIYQQQohiJgmWEEIIIUQxkwRLCCGEEKKYSYIlhBBCCFHMJMESQgghhChmkmAJIYQQQhQzSbCEEEIIIYqZJFhCCCGEEMVMEiwhhBBCiGImCZYQQgghRDGTBEsIIYQQophJgiWEEEIIUcwkwRJCCCFEoQ0dOpTevXur39u1a4e/v3+pxBIQEICHh0eptF0QSbCEEEII8cIoSsI2depUdu/eXbIB/UN6pR2AEEII8TLLzFL4LfYOiX8/orKZIc1qVERXR1PaYb3QFEUhMzMTU1NTTE1NSzucPMkIlhBCCFFKdpyNp1XgHvqvOsrEdSfpv+oorQL3sONsfIm2m5WVRWBgIM7OzhgYGFC9enU+/PBDAM6cOUOHDh0wMjLC2tqa0aNHk5qaWuhrp6WlMW3aNKpWrYqJiQnNmzcnIiJCq86hQ4do27YtxsbGWFlZ4e3tzd27dxk6dCj79u1j6dKlaDQaNBoNV65cISIiAo1Gw6+//kqTJk0wMDDgwIEDMkUohBBCCG07zsbz9rcniE9+pFWekPyIt789UaJJ1owZMwgMDGT27NlER0fz/fffY2try4MHD/Dx8cHKyorIyEh+/PFHdu3axbhx4wp97WHDhnHo0CHWrVvH6dOn6du3Lz4+Ply4cAGAkydP0rFjR+rVq8eRI0c4ePAgPXr0IDMzk6VLl9KiRQtGjRpFfHw88fHxODg4qNeeNm0aCxcu5Ny5czRo0KDY70txkilCIYQQ4jnLzFKY93M0Sh7HFEADzPs5ms5udsU+Xfj333+zdOlSPvvsM4YMGQJArVq1aNWqFatWreLhw4d8/fXXmJiYAPDZZ5/Ro0cPAgMDsbW1fea1L126xNq1a7l27RpVqlQBstdJ7dixg5CQED766CMWLVpEkyZNWLFihXpevXr11D/r6+tjbGyMnZ1druvPnz+fzp07/+t78DxIgiWEEEI8Z7/F3sk1cvUkBYhPfsRvsXdoUcu6WNs+d+4cjx8/pmPHjnkea9iwoZpcAbRs2ZKsrCxiYmIKTLBOnDiBoii4uLholT9+/Bhr6+x+nDx5kr59+/6j2Js0afKPzisNpT5FuGLFCmrUqIGhoSGNGzfmwIEDz6z/+PFjZs6ciaOjIwYGBtSqVYvVq1erx0NDQ9V52yc/jx7l/xdZCCGEeJ4S/y7c76TC1isKIyOjfI8pioJGk/eIWX7lT8rKykJXV5eoqChOnjypfs6dO8fSpUsLbL8gTyZ+L7oiJVh79uzBzc2NlJSUXMeSk5OpV69egQnSk9avX4+/vz8zZ87k999/p3Xr1nTt2pW4uLh8z+nXrx+7d+/mq6++IiYmhrVr11KnTh2tOubm5urcbc7H0NCw8B0VQgghSlBls8L9TipsvaKoXbs2RkZGeW5v4ObmxsmTJ7l//75adujQIXR0dHKNSuXF09OTzMxMEhMTcXZ21vrkTPk1aNDgmVsr6Ovrk5mZ+Q969mIpUoIVHBzMqFGjMDc3z3XMwsKCt956i8WLFxf6eosXL2bEiBGMHDmSunXrEhwcjIODAytXrsyz/o4dO9i3bx9hYWF06tQJJycnmjVrhpeXl1Y9jUaDnZ2d1kcIIYR4UTSrURF7C0PyGxPSAPYW2Vs2FDdDQ0OmT5/OtGnT+Prrr7l06RJHjx7lq6++YsCAARgaGjJkyBDOnj3L3r17GT9+PIMGDSpwehDAxcWFAQMGMHjwYDZt2kRsbCyRkZEEBgYSFhYGZC+wj4yMZOzYsZw+fZo///yTlStXkpSUBICTkxPHjh3jypUrJCUlkZWVVez34Hko0hqsU6dOERgYmO/xLl268MknnxTqWmlpaURFRfHee+/lusbhw4fzPGfr1q00adKERYsW8c0332BiYkLPnj1ZsGCB1pBjamoqjo6OZGZm4uHhwYIFC/D09Mw3lsePH/P48WP1e84IXXp6Ounp6YXqT1mS0yfpW9lSnvsG5bt/0reyq6T6paujYW4PN97+9gQa0FrsnpN0ze3hVmL7Yc2ePRs9PT3mzJnDjRs3sLe3Z8yYMRgbG/Prr78yceJEmjZtirGxMa+//nqRBk9CQkL44IMPmDJlCtevX8fa2poWLVrg6+sLZCdhO3fu5P3336dZs2YYGRnRvHlz+vfvD2Qvih8yZAhubm48fPiQ2NjYErkHJU2jKEpeDzHkydDQkLNnz+Ls7Jzn8YsXL+Lu7s7Dhw8LvNaNGzeoWrUqhw4d0hqB+uijj1izZg0xMTG5zvHx8SEiIoJOnToxZ84ckpKSGDt2LB06dFDXYR09elSNIyUlhaVLlxIWFsapU6eoXbt2nrEEBAQwb968XOXff/89xsbGBfZFCCFE+fTgwQPefPNNkpOT85y9+bd2nI1n3s/RWgve7S0MmdvDDZ/69sXennh+ijSCVbVqVc6cOZNvgnX69Gns7Yv2F+LpRXPPWmCXlZWFRqPhu+++w8LCAsieZuzTpw/Lly/HyMiIV155hVdeeUU9p2XLljRq1IhPP/2UZcuW5XndGTNmMHnyZPV7SkoKDg4OtG/fXn3qoTxJT08nPDyczp07U6FChdIOp1hJ38qu8tw/6VvZdfv27RK9vk99ezq72clO7uVQkRIsX19f5syZQ9euXXMtGn/48CFz586le/fuhbpWpUqV0NXVJSEhQas8MTEx33lee3t7qlatqiZXAHXr1kVRFK5du5bnCJWOjg5NmzZVNzjLi4GBAQYGBrnKK1SoUC7/wchRnvsnfSu7ynP/pG9lz/Pok66Opti3YhClr0iL3GfNmsWdO3dwcXFh0aJF/PTTT2zdupXAwEBcXV25c+cOM2fOLNS19PX1ady4MeHh4Vrl4eHhuRat52jZsiU3btzQ2rL//Pnz6OjoUK1atTzPURSFkydPFnlkTQghhBDinyrSCJatrS2HDx/m7bffZsaMGeQs39JoNHh7e7NixYpCPWWQY/LkyQwaNIgmTZrQokULvvjiC+Li4hgzZgyQPXV3/fp1vv76awDefPNNFixYwLBhw5g3bx5JSUm8++67DB8+XF3kPm/ePF555RVq165NSkoKy5Yt4+TJkyxfvrwoXRVCCCGE+MeKvJO7o6MjYWFh3L17l4sXL6IoCrVr18bKygqA69evU7Vq1UJdy8/Pj9u3bzN//nzi4+OpX78+YWFhODo6AhAfH6+1J5apqSnh4eGMHz+eJk2aYG1tTb9+/fjggw/UOvfu3WP06NEkJCRgYWGBp6cn+/fvp1mzZkXtqhBCCCHEP/KPX5VjZWVF06ZN1e8JCQl8+OGHfPnll4V6ijDH2LFjGTt2bJ7HQkNDc5XVqVMn17Tik5YsWcKSJUsK3b4QQgghRHEr0hqse/fuMWDAAGxsbKhSpQrLli0jKyuLOXPmULNmTY4ePar12hohhBBCiJdRkUaw3n//ffbv38+QIUPYsWMHkyZNYseOHTx69Ijt27fTtm3bkopTCCGEEKLMKFKC9csvvxASEkKnTp0YO3Yszs7OuLi4EBwcXELhCSGEEEKUPUWaIrxx4wZubm4A1KxZE0NDQ0aOHFkigQkhhBDi+WnXrh3+/v6l0raTk1O5G6wpUoKVlZWltemarq4uJiYmxR6UEEIIIcqf0NBQLC0tSzuM56JIU4SKojB06FB11/NHjx4xZsyYXEnWpk2bii9CIYQQojzLyoSrhyH1JpjagqMX6OiWdlRFkpmZiUajQUenSOM25VqR7sSQIUOoXLkyFhYWWFhYMHDgQKpUqaJ+z/kIIYQQohCit0JwfVjTHTaOyP7f4PrZ5SXo/v37DB48GFNTU+zt7QkKCtI6fvfuXQYPHoyVlRXGxsZ07dpV65VzOSNR27Ztw83NDQMDA65evfrM8yIiIhg2bBjJycloNBo0Gg0BAQHqNR88eMDw4cMxMzOjevXqfPHFF1oxXb9+HT8/P6ysrLC2tqZXr15cuXJFPR4ZGUnnzp2pVKkSFhYWtG3blhMnThT/zSukIo1ghYSElFQcQgghxMsleiv8MBhQtMtT4rPL+30Nbj1LpOl3332XvXv3snnzZuzs7Hj//feJiorCw8MDgKFDh3LhwgW2bt2Kubk506dPx9fXl+joaHWp0IMHD1i4cCFffvkl1tbWVK5cmTfffDPf87y8vAgODmbOnDnExMQA2RuI5wgKCmLBggW8//77bNiwgbfffps2bdpQp04dHjx4QPv27WndujX79+9HT0+PDz74AB8fH06fPo2+vj5///03Q4YMYdmyZer1fH19uXDhAmZmZiVyH5+lyBuNXr16lZ07d5Kenk67du3URe9CCCGEKKSsTNgxnVzJFfy3TAM73oM63Yp9ujA1NZWvvvqKr7/+ms6dOwOwZs0a9Z2+OQnSoUOH1HcDf/fddzg4OLBlyxb69u0LQHp6OitWrKBhw4aFPs/CwgKNRoOdnV2uuHx9fdWNx6dPn86SJUuIiIigTp06rFu3Dh0dHb788ks0Gg2QPehjaWlJREQEXbp0oUOHDlrX+/zzz7GysmLfvn107969WO9hYRQpwdq/fz++vr48ePAg+2Q9PdasWUP//v1LJDghhBCiXLp6GFJuPKOCAinXs+vVaF2sTV+6dIm0tDRatGihllWsWBFXV1cAzp07h56eHs2bN1ePW1tb4+rqyrlz59QyfX19GjRooH4v7Hn5efJaOUlYYmIiAFFRUVy8eDHXSNSjR4+4dOkSAImJicyZM4c9e/Zw8+ZNMjMzefDggdYr956nIiVYs2fPpn379nz++ecYGRkxY8YMpk2bJgmWEEIIURSpN4u3XhEoSl6jZgUfVxRFHT0CMDIy0vpe2PPy8+QuBZCdZGVlZQHZuxg0btyY7777Ltd5NjY2QPa05q1btwgODsbR0REDAwNatGhBWlpagW2XhCItcj9z5gwLFy6kSpUqWFlZERQUxI0bN7h7925JxSeEEEKUP6a2xVuvCJydnalQoQJHjx5Vy+7evcv58+cBcHNzIyMjg2PHjqnHb9++zfnz56lbt26+1y3Mefr6+mRmZhY55kaNGnHhwgUqV66Ms7Oz1ifn4boDBw4wYcIEfH19qVevHgYGBiQlJRW5reJS5HcRVq5cWf1uYmKCsbEx9+7dK+64hBBCiPLL0QvMqwD5jexowLxqdr1iZmpqyogRI3j33XfZvXs3Z8+eZejQoeoWC7Vr16ZXr16MGjWKgwcPcurUKQYOHEjVqlXp1atXvtctzHlOTk6kpqaye/dukpKS1CVHBRkwYACVKlWiV69eHDhwgNjYWPbt28fEiRO5du0akJ04fvPNN5w7d45jx44xYMAAjIyM/uXd+ueKvGFFdHQ0p0+fVj+KonDu3DmtMiGEEEI8g44u+AT+98vTSdZ/v/t8XGL7Yf3nP/+hTZs29OzZk06dOtGqVSsaN26sHg8JCaFx48Z0796dFi1aoCgKYWFhuabxnlbQeV5eXowZMwY/Pz9sbGxYtGhRoeI1NjZm//79VK9enddee426desyfPhwHj58iLm5OQCrV6/m7t27eHp6MmjQICZMmKA1KPS8aZSCJmOfoKOjg0ajyXOeNadco9H8o+G/F0lKSgoWFhYkJSVhbW1d2uEUu/T0dMLCwvD19S3w/yxljfSt7CrP/ZO+lV23b9+mUqVKJCcnq7/Ii1X01uynCZ9c8G5eNTu5KqEtGsTzUaRF7rGxsSUVhxBCCPHyceuZvRVDGd/JXeRWpATL0dGxpOIQQgghXk46usW+FYMofUXeBysvFhYWODs7y4ufhRBCCCEoYoLVrl27fI/p6ury9ttvExQUVC7n4YUQQgghCqtICVZ++13du3eP3377jXfffVd9p5EQQgghxMuqSAlWzmZeeZU7Ojqir6/P+++/LwmWEEIIIV5qRd4H61kaNmzI1atXi/OSQgghhBBlTrEmWDdu3CjVTb2EEEIIIV4ExZZgJSYmMmvWLDp06FBclxRCCCGEKJOKtAbL09MzzzdiJycnc+3aNerWrcu6deuKLTghhBBClG1OTk74+/vj7+8PZL/5ZfPmzfTu3fu5x9KuXTs8PDwIDg4u8baKlGDldzPMzc2pU6cOXbp0QVdXdp8VQgghxPNRlIRt06ZNz20rqSIlWHPnzi2wTkZGBnp6RbqsEEII8dLKzMrkROIJbj24hY2xDY0qN0L3Ob4qJy0tDX19/efWXmlIT0+nQoUKVKxY8bm1WWxrsKKjo5k8eTJVq1YtrksKIYQQ5dquq7vw3ujN8F+HM/3AdIb/Ohzvjd7surqrxNps164d48aNY/LkyVSqVInOnTs/s/69e/cYPXo0tra2GBoaUr9+fbZt26Ye37hxI/Xq1cPAwAAnJyeCgoKKFM/169fx8/PDysoKa2trevXqxZUrV7TqrF69Wm3D3t6ecePGAdnTjwCvvvoqGo1G/R4QEICHhwerV6+mZs2aGBgYoCgK7dq1U6cqS9q/SrBSU1P58ssvadGiBQ0aNOC3337jvffeK67YhBBCiHJr19VdTI6YzM0HN7XKEx8kMjlicokmWWvWrEFPT49Dhw7x+eef51svKyuLrl27cvjwYb799luio6P5+OOP1eVAUVFR9OvXjzfeeIMzZ84QEBDA7NmzCQ0NLVQcDx48oH379piamrJ//34OHjyIqakpPj4+pKWlAbBy5UreeecdRo8ezZkzZ9i6dSvOzs4AREZGAhASEkJ8fLz6HeDixYv88MMPbNy4kZMnT/6Du/Tv/KO5vIMHD/Lll1+yceNGatSoQXR0NPv27aNly5bFHZ8QQghR7mRmZfLxbx+joOQ6pqCgQUPgb4G0d2hfItOFzs7OLFq0qMB6u3bt4rfffuPcuXO4uLgAULNmTfX44sWL6dixI7NnzwbAxcWF6Oho/vOf/zB06NACr79u3Tp0dHT48ssv1YfoQkJCsLS0JCIigi5duvDBBx8wZcoUJk6cqJ7XtGlTAGxsbACwtLTEzs5O69ppaWl88803ap3nrUgjWIsWLaJOnTq88cYb2NjYcPDgQU6fPo1Go8HKyqqkYhRCCCHKlROJJ3KNXD1JQSHhQQInEk+USPtNmjQpVL2TJ09SrVo1Nbl62rlz53INrrRs2ZILFy6QmZlZ4PWjoqK4ePEiZmZmmJqaYmpqSsWKFXn06BGXLl0iMTGRGzdu0LFjx0LF+yRHR8dSS66giCNY77//PtOnT2f+/PnytKAQQgjxD916cKtY6xWViYlJoeoZGRk987iiKLm2b1KU3KNy+cnKyqJx48Z89913uY7Z2Nigo/PPVzIVto8lpUiRz58/nx9//JEaNWowffp0zp49W1JxCSGEEOWWjXHhRlYKW6+kNGjQgGvXrnH+/Pk8j7u5uXHw4EGtssOHD+Pi4lKogZhGjRpx4cIFKleujLOzs9bHwsICMzMznJyc2L17d77XqFChQqFGy563IiVY77//PufPn+ebb74hISGBV155hYYNG6IoCnfv3i2pGIUQQohypVHlRtga26Ih9+bdABo02Bnb0ahyo+ccmba2bdvSpk0bXn/9dcLDw4mNjWX79u3s2LEDgClTprB7924WLFjA+fPnWbNmDZ999hlTp04t1PUHDBhApUqV6NWrFwcOHCA2NpZ9+/YxceJErl27BmQ/ERgUFMSyZcu4cOECJ06c4NNPP1WvkZOAJSQkvFC5yD8ae2vbti1r1qwhPj6et99+m8aNG9O2bVu8vLxYvHhxcccohBBClCu6Orq81yz7qfunk6yc79ObTX+u+2HlZ+PGjTRt2pT+/fvj5ubGtGnT1BGjRo0a8cMPP7Bu3Trq16/PnDlzmD9/fqEWuAMYGxuzf/9+qlevzmuvvUbdunUZPnw4Dx8+xNzcHIAhQ4YQHBzMihUrqFevHt27d+fChQvqNYKCgggPD8fBwQFPT89i7/8/pVGKMln6DGfOnOGrr77i+++/JzExsTguWWpSUlKwsLAgKSkJa2vr0g6n2KWnpxMWFoavr+9z29H2eZG+lV3luX/St7Lr9u3bVKpUieTkZPUXfnHadXUXH//2sdaCdztjO6Y3m04nx07F3p54fopty3V3d3eCg4P5z3/+U1yXFEIIIcq1To6daO/QvlR3chclo0hThHv27MHNzY2UlJRcx5KTk6lXrx5Hjx4ttuCEEEKI8k5XR5emdk3xrelLU7umzzW5+u6779TtEZ7+1KtX77nFUR4VaQQrODiYUaNG5TlMamFhwVtvvcXixYtp3bp1sQUohBBCiJLRs2dPmjdvnuex8jjl+zwVKcE6deoUgYGB+R7v0qULn3zyyb8OSgghhBAlz8zMDDMzs9IOo1wq0hThzZs3n5nR6unpcetWyWyKJoQQQghRVhQpwapatSpnzpzJ9/jp06ext7f/10EJIYQQQpRlRUqwfH19mTNnDo8ePcp17OHDh8ydO5fu3bsXW3BCCCGEEGVRkdZgzZo1i02bNuHi4sK4ceNwdXVFo9Fw7tw5li9fTmZmJjNnziypWIUQQgghyoQijWDZ2tpy6NAh6tevz4wZM3j11Vfp3bs377//PvXr1+fQoUPY2toWKYAVK1ZQo0YNDA0Nady4MQcOHHhm/cePHzNz5kwcHR0xMDCgVq1arF69WqvOxo0bcXNzw8DAADc3NzZv3lykmIQQQggh/o0ibzTq5OREWFgYd+/e5eLFiyiKQu3atbGysipy4+vXr8ff358VK1bQsmVLPv/8c7p27Up0dDTVq1fP85x+/fpx8+ZNvvrqK5ydnUlMTCQjI0M9fuTIEfz8/FiwYAGvvvoqmzdvpl+/fhw8eDDfR1GFEEIIIYpTkRKs4cOHF6re0yNK+Vm8eDEjRoxg5MiRQPY+W7/++isrV65k4cKFuerv2LGDffv2cfnyZSpWrAhkJ3xPCg4OpnPnzsyYMQOAGTNmsG/fPoKDg1m7dm2h4hJCCCFE/jQaDZs3b6Z3797Pve0rV65Qo0YNfv/9dzw8PJ57+4VVpAQrNDQUR0dHPD09+bevMExLSyMqKor33ntPq7xLly4cPnw4z3O2bt1KkyZNWLRoEd988w0mJib07NmTBQsWYGRkBGSPYE2aNEnrPG9vb4KDg/ON5fHjxzx+/Fj9nrNTfXp6Ounp6f+key+0nD5J38qW8tw3KN/9k76VXc+jX0pmJg+OR5Fx6xZ6NjYYN2mMRldelQMwdOhQ7t27x5YtW0o7lCIrUoI1ZswY1q1bx+XLlxk+fDgDBw5UR5KKKikpiczMzFxrtmxtbUlISMjznMuXL3Pw4EEMDQ3ZvHkzSUlJjB07ljt37qijZgkJCUW6JsDChQuZN29ervK9e/dibGxc1K6VGeHh4aUdQomRvpVd5bl/0rey58GDByV6/ZSdO7n50UIynvgdpWdnh+37MzDv0qVE286RlpaGvr7+c2mrNNt83oqUYK1YsYIlS5awadMmVq9ezYwZM+jWrRsjRoygS5cuaDSaIgfw9DmKouR7naysLDQaDd999x0WFhZA9jRjnz59WL58uTqKVZRrQvY04uTJk9XvKSkpODg40L59e6ytrYvcpxddeno64eHhdO7cudy9CkH6VnaV5/5J38qu27dvl9i1U3bu5PpEf3hqRijj5s3s8qXBJZJktWvXjvr166Ovr8/XX39NvXr12LdvX771L1y4wIgRI/jtt9+oWbMmS5cuzVXnzJkzTJw4kSNHjmBsbMzrr7/O4sWLMTU1Bf43EtW8eXM+/fRT9PX1uXLlyjPPCwgIYM2aNcD/fq/v3btXXRp0+fJlJk2axLFjx6hduzb/93//R4sWLdSYDh8+zHvvvUdkZCSVKlXi1VdfZeHChZiYmADw7bffEhwcTExMDCYmJnTo0IHg4GAqV65cLPe5yIvcDQwM6N+/P/379+fq1auEhoYyduxY0tPTiY6OVm9mQSpVqoSurm6ukaXExMR8n0S0t7enatWqanIFULduXRRF4dq1a9SuXRs7O7siXTOnTwYGBrnKK1SoUC7/wchRnvsnfSu7ynP/pG9lT0n1ScnM5OZHC3MlV9kHFdBouPnRQsw6diyR6cI1a9bw9ttvc+jQoWcu+cnKyuK1116jUqVKHD16lJSUFPz9/bXqPHjwAB8fH1555RUiIyNJTExk5MiRjBs3jtDQULXe7t27MTc3Jzw8HEVRCjxv6tSpnDt3jpSUFEJCQgCoWLEiN27cAGDmzJl88skn1K5dm5kzZ9K/f38uXryInp4eZ86cwdvbmwULFvDVV19x69Ytxo0bx7hx49RrpaWlsWDBAlxdXUlMTGTSpEkMHTqUsLCwYrnHRU6wnqTRaNBoNCiKQlZWVpHO1dfXp3HjxoSHh/Pqq6+q5eHh4fTq1SvPc1q2bMmPP/5IamqqmsidP38eHR0dqlWrBkCLFi0IDw/XWoe1c+dOvLy8ito9IYQQokQ8OB6lNS2Yi6KQkZDAg+NRmDRvVuztOzs7s2jRogLr7dq1i3PnznHlyhX19+xHH31E165d1TrfffcdDx8+5Ouvv1ZHhz777DN69OhBYGCgOsBhYmLCl19+qU4Nrlq1qsDzjIyMePz4MXZ2drlimzp1Kt26dQNg3rx51KtXj4sXL1KnTh3+85//8Oabb6rJYO3atVm2bBlt27Zl5cqVGBoaaj24V7NmTZYtW0azZs20cox/o0j7YEH2gvC1a9fSuXNnXF1dOXPmDJ999hlxcXFFDmjy5Ml8+eWXrF69mnPnzjFp0iTi4uIYM2YMkD11N3jwYLX+m2++ibW1NcOGDSM6Opr9+/fz7rvvMnz4cHV6cOLEiezcuZPAwED+/PNPAgMD2bVrV66MWwghhCgtGYV8b29h6xVVkyZNClXv3LlzVK9eXU2uAK1puJw6DRs2VJMkyB4QycrKIiYmRi1zd3fXWndV2PPy06BBA/XPOa/pS0xMBCAqKorQ0FBMTU3Vj7e3N1lZWcTGxgLw+++/06tXLxwdHTEzM6Ndu3YAxMXFFdh2YRRpBGvs2LGsW7eO6tWrM2zYMNatW/ev1ij5+flx+/Zt5s+fT3x8PPXr1ycsLAxHR0cA4uPjtTpqampKeHg448ePp0mTJlhbW9OvXz8++OADtY6Xlxfr1q1j1qxZzJ49m1q1arF+/XrZA0sIIcQLQ8/GpljrFdWTSc2z5DV9WJR1zk+WP91mYc/Lz5PTtzn1c2bTsrKyeOutt5gwYUKu86pXr879+/fp0qULXbp04dtvv8XGxoa4uDi8vb1JS0srsO3CKFKC9X//939Ur16dGjVqsG/fvnwXxW3atKnQ1xw7dixjx47N89iTc7c56tSpU+DTKn369KFPnz6FjkEIIYR4noybNEbPzo6MmzfzXoel0aBna4txk8bPP7gnuLm5ERcXx40bN6hSpQqQvR3S03XWrFnD/fv31STq0KFD6Ojo4OLi8sxrF3Sevr4+mZmZRY67UaNG/PHHHzg7O+d5/MyZMyQlJfHxxx/j4OAAwPHjx4vczrMUaYpw8ODBtG/fHktLSywsLPL9CCGEECJ/Gl1dbN+f8d8vT43W/Pe77fszSn0/rE6dOuHq6srgwYM5deoUBw4cyPXO4QEDBmBoaMiQIUM4e/Yse/fuZfz48QwaNOiZD5gV5jwnJydOnz5NTEwMSUlJhd6XbPr06Rw5coR33nmHkydPcuHCBbZu3cr48eOB7FEsfX19Pv30Uy5fvszWrVtZsGDBP7xLeSvyRqNCCCGE+PfMu3SBpcG598GytX2u+2A9i46ODps3b2bEiBE0a9YMJycnli1bho+Pj1rH2NiYX3/9lYkTJ9K0aVOt7RaepTDnjRo1ioiICJo0aUJqaqrWNg3P0qBBA/bt28fMmTNp3bo1iqJQq1Yt/Pz8ALCxsSE0NJT333+fZcuW0ahRIz755BN69uz5z25UHv7VU4RCCCGE+OfMu3TBrGPH57qTe0RERJHqu7i4cODAAa2yp9dmubu7s2fPnnyvkd8ATUHn2djYsHPnzlzlT7dvaWmZq6xp06Z5npsjZ8upZ13335AESwghhChFGl3dEtmKQZSuIm/TIIQQQojy4bvvvtPayuDJT7169Uo7vDJNRrCEEEKIl1TPnj3z3caoPO7M/zxJgiWEEEK8pMzMzDAzMyvtMMolmSIUQgghhChmkmAJIYQQQhQzSbCEEEIIIYqZJFhCCCGEEMVMEiwhhBBCiGImCZYQQgghRDGTBEsIIYQQophJgiWEEEIIUcwkwRJCCCGEKGaSYAkhhBBCFDNJsIQQQgghipkkWEIIIYQQxUwSLCGEEEKIYiYJlhBCCCFEMZMESwghhBCimEmCJYQQQghRzCTBEkIIIYQoZpJgCSGEEEIUM0mwhBBCCCGKmSRYQgghhBDFTBIsIYQQQohiJgmWEEIIIUQxkwRLCCGEEKKYSYIlhBBCCFHMJMESQgghhChmkmAJIYQQQhQzSbCEEEIIIYqZJFgvMX19fbZs2VLaYRRJQEAAHh4epR2GEEII8UySYIkXlkajyZUATp06ld27d5d425LICSGE+Df0SjsAIYrC1NQUU1PT0g5DCCGEeCYZwSrHnJycCA4O1irz8PBg/vz5uepeuXIFjUbDDz/8QOvWrTEyMqJp06acP3+eyMhImjRpgqmpKT4+Pty6davQMYSEhFC3bl0MDQ2pU6cOK1asUI+lpaUxbtw47O3tMTQ0xMnJiYULF6qxA7z66qtoNBr1+9MjS0OHDqV379589NFH2NraYmNjw7p168jIyODdd9+lYsWKVKtWjdWrV2vFNX36dFxcXDA2NqZmzZrMnj2b9PR0AEJDQ5k3bx6nTp1Co9Gg0WgIDQ0FIDk5mdGjR1O5cmXMzc3p0KEDp06dKvT9EEII8XKQESyhZe7cuQQHB1O9enWGDx9O//79MTc3Z+nSpRgbG9OvXz/mzJnDypUrC7zWqlWrmDt3Lp999hmenp78/vvvjBo1ChMTE4YMGcKyZcvYunUrP/zwA9WrV+evv/7ir7/+AiAyMpLKlSsTEhKCj48Purq6+bazZ88eqlWrxv79+9m/fz+jR4/mzp07tG3blmPHjrF+/XrGjBlD586dcXBwAMDMzIzQ0FCqVKnCmTNnGDVqFGZmZkybNg0/Pz/Onj3Ljh072LVrFwAWFhYoikK3bt2oWLEiYWFhWFhY8Pnnn9OxY0fOnz9PxYoVi+EnIIQQojyQBEtomTp1Kt7e3gBMnDiR/v37s3v3blq2bAnAiBEj1NGcgixYsICgoCBee+01AGrUqEF0dDSff/45Q4YMIS4ujtq1a9OqVSs0Gg2Ojo7quTY2NgBYWlpiZ2f3zHYqVqzIsmXL0NHRoWbNmsybN48HDx7w/vvvAzBjxgw+/vhjDh06xBtvvAHArFmz1POdnJyYMmUK69evZ9q0aRgZGWFqaoqenp5W23v27OHMmTMkJiZiYGAAwCeffMKWLVvYsGEDo0ePLtR9EUIIUf5JgiW0NGjQQP2zra0tAO7u7lpliYmJBV7n1q1b/PXXX4wYMYJRo0ap5RkZGVhYWADZ03udO3fG1dUVHx8funfvTpcuXYocc7169dDR+d9st6WlJfXr11e/6+rqYm1trRX3hg0bCA4O5uLFi6SmppKRkYG5ufkz24mKiiI1NRVra2ut8ocPH3Lp0qUixy2EEKL8kgSrHNPR0UFRFK2ynHVG+alQoYL6Z41Gk2dZVlZWgW3n1Fm1ahXNmzfXOpYz3deoUSNiY2PZvn07u3btol+/fnTq1IkNGzYUeP38Ys6v7Mm4jx49yhtvvMG8efPw9vbGwsKCdevWERQUVGCf7O3tiYiIyHXM0tKySDELIYQo3yTBKsdsbGyIj49Xv6ekpBAbG/tc2ra1taVq1apcvnyZAQMG5FvP3NwcPz8//Pz86NOnDz4+Pty5c4eKFStSoUIFMjMziz22Q4cO4ejoyMyZM9Wyq1evatXR19fP1XajRo1ISEhAT09PXXQvhBBC5EUSrHKsQ4cOhIaG0qNHD6ysrJg9e/YzF4sXt4CAACZMmIC5uTldu3bl8ePHHD9+nLt37zJ58mSWLFmCvb09Hh4e6Ojo8OOPP2JnZ6eOBjk5OanrvwwMDLCysiqWuJydnYmLi2PdunU0bdqUX375hc2bN2vVcXJyIjY2lpMnT1KtWjXMzMzo1KkTLVq0oHfv3gQGBuLq6sqNGzcICwujd+/eNGnSpFjiE0IIUfaV+jYNK1asoEaNGhgaGtK4cWMOHDiQb92IiAj1sfknP3/++adaJzQ0NM86jx49eh7deaHMmDGDNm3a0L17d3x9fenduze1atV6bu2PHDmSL7/8ktDQUNzd3Wnbti2hoaHUqFEDyN7TKjAwkCZNmtC0aVOuXLlCWFiYup4qKCiI8PBwHBwc8PT0LLa4evXqxaRJkxg3bhweHh4cPnyY2bNna9V5/fXX8fHxoX379tjY2LB27Vo0Gg1hYWG0adOG4cOH4+LiwhtvvMGVK1fU9WpCCCEEAEopWrdunVKhQgVl1apVSnR0tDJx4kTFxMREuXr1ap719+7dqwBKTEyMEh8fr34yMjLUOiEhIYq5ubnW8fj4+CLFlZycrABKUlLSv+rfiyotLU3ZsmWLkpaWVtqhFDvpW9lVnvsnfSu7kpKSFEBJTk4u7VBEGVOqU4SLFy9mxIgRjBw5EoDg4GB+/fVXVq5cqW44mZfKlSs/c1GxRqMp8NF+IYQQQoiSUmoJVlpaGlFRUbz33nta5V26dOHw4cPPPNfT05NHjx7h5ubGrFmzaN++vdbx1NRUHB0dyczMxMPDgwULFjxziunx48c8fvxY/Z6SkgJkP3FX0FN3ZVFOn/5t355cE5WVlYWiKOoar59//plWrVr9q+v/E8XVtxdRee4blO/+Sd/KrvLaL1HyNIry1HP8z8mNGzeoWrUqhw4dwsvLSy3/6KOPWLNmDTExMbnOiYmJYf/+/TRu3JjHjx/zzTff8H//939ERETQpk0bIPsR/IsXL+Lu7k5KSgpLly4lLCyMU6dOUbt27TxjCQgIYN68ebnKv//+e4yNjYupx+XPtWvXuHfvHnfu3CE9PR1zc3OqVq2KRqOhYsWK6macQghRVj148IA333yT5OTkAvfKE+JJpZ5gHT58mBYtWqjlH374Id98843WwvVn6dGjBxqNhq1bt+Z5PCsri0aNGtGmTRuWLVuWZ528RrAcHByIj4/PtalkeZCenk54eDidO3fOcw+pgqSkpBAZGcnvv/9ORkYG9evXp1mzZlSuXLkEoi2af9u3F1l57huU7/5J38qu27dvY29vLwmWKLJSmyKsVKkSurq6JCQkaJUnJiYW6YmsV155hW+//Tbf4zo6OjRt2pQLFy7kW8fAwCDP0ZYKFSqUy38wchS1fwkJCRw5coSzZ89SoUIFmjZtSvPmzTEzMyvBKP+Z8vyzK899g/LdP+lb2VMe+ySej1JLsPT19WncuDHh4eG8+uqranl4eDi9evUq9HV+//137O3t8z2uKAonT57Uet2LKDxFUbh8+TKHDx/m8uXLWFhY0LlzZzw9PWUKUAghhMhHqT5FOHnyZAYNGkSTJk1o0aIFX3zxBXFxcYwZMwbI3sfp+vXrfP3110D2U4ZOTk7Uq1ePtLQ0vv32WzZu3MjGjRvVa86bN49XXnmF2rVrk5KSwrJlyzh58iTLly8vlT6+sLKyeBAZCXfuomdjg3GTxmie2IQ0MzOTs2fPcuTIEW7evImdnR2vvfYabm5u/3iz0oiICNq3b8/du3fL1Ktl2rVrh4eHB8HBwaUdihBCiDKiVBMsPz8/bt++zfz584mPj6d+/fqEhYXh6OgIQHx8PHFxcWr9tLQ0pk6dyvXr1zEyMqJevXr88ssv+Pr6qnXu3bvH6NGjSUhIwMLCAk9PT/bv30+zZs2ee/9eVKm7dlHj40BuJCerZXp2dti+PwODtm2Jiori2LFjpKSk4OzsjLe3N05OTuq7Ccur/BLATZs2PZdpAknkhBCi/Cj1V+WMHTuWsWPH5nksNDRU6/u0adOYNm3aM6+3ZMkSlixZUlzhlTspO3eSMHkKek8925CSkkLU6tXE/vYbGYqCu7s7Xl5eL8TC9dJWsWLF0g5BCCFEGVPqr8oRz4+SmcnNjxaCovD0WNSBNq25XKsWtS7HMmHcOHr37l1gctWuXTvGjx+Pv78/VlZW2Nra8sUXX3D//n2GDRuGmZkZtWrVYvv27XmeHxoaiqWlJdu2bcPV1RVjY2P69OnD/fv3WbNmDU5OTlhZWTF+/PhCv/Q5LS2N0NBQnJycMDExoXnz5kRERKjHr169qr6b0cTEhHr16hEWFsaVK1fU/dSsrKzQaDQMHTpU7ae/v796DScnJz744AMGDx6Mqakpjo6O/PTTT9y6dYtevXphamqKu7s7x48fV8+5ffs2/fv3p1q1ahgbG+Pu7s7atWvV40OHDmXfvn0sXbpUfb3TlStXAIiOjsbX1xcrKyuGDBnC0KFDSUpKKtT9EEIIUTokwXqJPDgeRcZTT23maHXgIN1/2or74cPonc//icunrVmzhkqVKvHbb78xfvx43n77bfr27YuXlxcnTpzA29ubQYMG8eDBg7xjevCAZcuWsW7dOnbs2EFERASvvfYaYWFhhIWF8c033/DFF1+wYcOGQsUzcuRI/vzzT7799ltOnz5N37598fHxUZ8ifeedd3j8+DH79+/nzJkzBAYGYmpqioODg7qWLyYmhvj4eJYuXZpvO0uWLKFly5b8/vvvdOvWjUGDBjF48GAGDhzIiRMncHZ2ZvDgweTsgvLo0SMaN27Mtm3bOHv2LKNHj2bQoEEcO3YMgKVLl9KiRQtGjRpFfHw88fHx6lYhbdu2xcPDgyNHjjB37lwSExPp169foX9GQgghSkHpvaXnxVVe30V47+dtSrRrnQI/937eVqjrtW3bVmnVqpX6PSMjQzExMVEGDRqklsXHxyuAcuTIEfVdknfv3lUUJfu9kYBy8eJFtf5bb72lGBsbK3///bda5u3trbz11lsFxnPx4kVFo9Eoq1ev1novWseOHZUZM2YoiqIo7u7uSkBAQJ7nPx3fk/2cOHGi+t3R0VEZOHBgrj7Onj1bLTty5IgCPPM9mL6+vsqUKVPybUdRFGX27NlKly5dFEX53zvfLl++rL6Tszwpz++0k76VXfIuQvFPlfoaLPH86NnYFGs9gAYNGqh/1tXVxdraWmtLjJw9zRITE/PcpM/Y2JhatWpp1XdycsLU1FSrLDExscBYTpw4gaIojB07lvHjx6vljx8/VjeMnTBhAm+//TY7d+6kU6dOvP7661p9KKwnz8npY379trOzIzMzk48//pj169dz/fp1dXNbExOTZ7YTFRXF3r171fuRmZmpPsV56dIlXFxcihy7EEKIkicJ1kvEuElj9OzsyLh5E/LawF+jQc/WFuMmjQt9zaefrtNoNFplOU8eZmVl/aPzc8ryO/9JWVlZ6OrqEhQURPv27bWuk5OgjBw5Em9vb3755Rd27tzJwoULCQoK0krICiOvPj6r30FBQSxZsoTg4GDc3d0xMTHB39+ftLS0AvvUo0cPAgMDSU9PZ9++fbRt25YKFSo8c/83IYQQpUvWYL1ENLq62L4/A4Bc6dV/EwLb92do7YdVlnh6epKZmcm9e/dwdnbW+tjZ2an1HBwcGDNmDJs2bWLKlCmsWrUKyN78Fij0gvqiOHDgAL169WLgwIE0bNiQmjVr5nq7gL6+fq62GzVqxB9//IGTkxPOzs7Y29urfSpo9EsIIUTpkQTrJWPepQt2i4PIsLDQKteztaXq0mDMu3Qppcj+PRcXF/r378/SpUvZvHkzsbGxREZGEhgYSFhYGAD+/v78+uuvxMbGcuLECfbs2UPdunUBcHR0RKPRsG3bNm7dukVqamqxxebs7Ex4eDiHDx/m3LlzvPXWW7leE+Xk5MSxY8e4cuUKSUlJZGVl8c4773Dnzh369+9PZGQkCQkJhIeHM3z48BJJBIUQQhQPmSJ8CZl26kTso0e0s7XNdyf3surLL79k+PDhTJ8+nevXr2NtbU2LFi3UzWgzMzN55513uHbtGubm5vj4+Kj7plWtWpV58+bx3nvvMWzYMAYPHpxrL7Z/avbs2cTGxuLt7Y2xsTGjR4+md+/eJD+x2evUqVMZMmQIbm5uPHz4kNjYWJycnDh06BDTp0+nW7duPHjwgBo1atC1a1d0dOS/j4QQ4kWlUZS8FuO83FJSUrCwsCApKUldHF2epKenExYWhq+vb7l7kan0rewqz/2TvpVdt2/fplKlSiQnJ+f5oI4Q+ZERLFGilMzM7P23bt0qVyNlQgghxLNIgiVKTMrOndz8aKHW5qY57zx8cq1XZmYmf/31FxcuXMDS0pKmTZvmeb0DBw7QtWvXfNsrzjVTQgghxL8hCZYoESk7d3J9on+u7SAybt7k+kR/7v9nETerVePChQtcunRJ3ROqZcuW+V6zSZMmnDx5smQDF0IIIYqBJFii2D35zsMnpevpcb6OK/FVqnAnKgqioqhSpQqvvPIKLi4u2Nvbq/tH5cXIyAhnZ+eSDl8IIYT41yTBEsUuv3ce3q5UiUvOztgk3sL5wgUaz3ifym1al0KEQgghRMmSBEsUWmEXrGfcupXn+XYJCfTc8pP63SAlpcRiFeJFs3v3boYMGcK9e/dKOxQhxHMgG+mIQknZuZOLHTsRN2QIN6ZOJW7IEC527ETKzp256pbEOw+FKEucnJwIDg7WKmvVqhV//PFHibcdGhqKpaVlibcjhHg2SbBEgXIWrD897ZezYP3pJCvnnYfkt55Ko0HPzq5I7zwUoqwzMDCgcuXKpR2GEOI5kQRLPFN+C9YBMnR0cPvzHGumTCXt0SMyMjLIyMggC7CZ8R4KoDydZL0A7zwMCAjAw8OjVNp+3rKysggMDMTZ2RkDAwOqV6/Ohx9+CMCZM2fo0KEDRkZGWFtbM3r0aK2tLoYOHUrv3r355JNPsLe3x9ramnfeeYf09HS1zooVK6hduzaGhobY2trSp08f9ZiiKCxatIiaNWtiZGREw4YN2bBhg3o8IiICjUbD7t27adKkCRYWFkyfPp2YmBi1zqlTp2jfvj1mZmaYm5vTuHFjjh8/DuT9cwwODsbJyUmrjWbNmmFiYoKlpSUtW7bk6tWrhbp3P//8M40bN8bQ0JCaNWsyb948MjIy1OMBAQFUr14dAwMDqlSpwoQJEwBo164dV69eZdKkSWg0GvXBjd27d2PzxKhtTvyrV6+mevXqmJqa8vbbb5OZmcmiRYuws7OjcuXK6s8rx+LFi9UXhjs4ODB27Fj15xYREcGwYcNITk5W2w4ICAAgLS2NadOmUbVqVUxMTGjevDkRERGFuhdCiKKTNVjimfJbsJ5kbc3eTh1h/h8cqufGncDA3Ce/4af+UZOVlf2/ZL90WhMZCZGR6i+BJ38RPet7QXUgez+s69evo6Ojw9tvv83bb7+Nh4eHWsfGxobhw4ezevXqf9VWQeesW7eOY8eOsXTp0mLpX1ZWFgkJCRw8eBBdXd1cdXR1dWnQoAHGxsbqfZ8xYwarVq1iyZIltGrVivj4eP78808ePHiAj48Pr7zyCpGRkSQmJjJy5EjGjRun9XqgvXv3Ym9vz969e7l48SJ+fn54eHgwatQojh8/zoQJE/jmm2/w8vLizp07HDhwQD131qxZbNq0iZUrV1K7dm3279/PwIEDsbGxoW3btmq9mTNnEhQUhKWlJW+++SajR4/m8OHDAAwYMABPT09WrlyJrq4uJ0+eLPRu4RkZGfTu3ZtRo0axdu1a0tLS+O233575pGqOX3/9lYEDB7Js2TJat27NpUuXGD16NABz585lw4YNLFmyhHXr1lGvXj0SEhI4deoUAJs2baJhw4aMHj2aUaNGPbOdS5cusX37dnbs2MGlS5fo06cPsbGxuLi4sG/fPg4fPszw4cPp2LEjr7zyCgA6OjosW7YMJycnYmNjGTt2LNOmTWPFihV4eXkRHBzMnDlz1ETV1NQUgGHDhnHlyhXWrVtHlSpV2Lx5Mz4+Ppw5c4batWsX6p4KIQpPEizxTPktWK945w5eBw8C4PpnDD6+vhh5eJDz5iVFUbI/mZk8vnKVzL//RmNqir5jdXLGwtQ6//08XVbQ97zKsrKyuHz5svriZoCKFStSpUqVPM/5J20Uts7Dhw/JzMwkMTGx2K776NEjoqKi8uyDRqOhcuXK1KxZE4C///6bpUuX8tlnnzFkyBAAatWqRatWrVi1ahUPHz7k66+/xsTEBIDPPvuMHj16EBgYiK2tLQBWVlZ89tln6OrqUqdOHbp168bu3bsZNWoUcXFxmJiY0L17d8zMzHB0dMTT0xOA+/fvs3jxYvbs2UOLFi0AqFmzJgcPHuTzzz/XSrA+/PBD2rZtS3p6Oq+99hoffPABjx49wtDQkLi4ON59913q1KkDUKREICUlheTkZLp3706tWrUA1Bd7F+TDDz/kvffeU+9bzZo1WbBgAdOmTWPu3LnExcVhZ2dHp06dqFChAtWrV6dZs2ZA9t83XV1dzMzMsLOzA9Aa9XtSVlYWq1evxszMDDc3N9q3b09MTAxhYWHo6Ojg6upKYGAgERERaoLl7++vnl+jRg0WLFjA22+/zYoVK9DX18fCwgKNRqO2DdmJ3Nq1a7l27RpVqlQBst99uWPHDkJCQvjoo48KfV+FEIWkiFySk5MVQElKSirtUEpEWlqasmXLFiUtLa3AutXt7JX3bCor0a511I+rgYEy1tpaiXatowDKsipVldSjx5TY2FgFUNavX6+0atVKMTQ0VJo0aaLExMQov/32m9K4cWPFxMRE8fb2VhITEwsd7+rVq5U6deooBgYGiqurq7J8+XL12OPHj5V33nlHsbOzUwwMDBRHR0dl4MCBSlpamuLo6KgA6sfR0VFRFEWZO3eu0rBhQ/UaQ4YMUXr16qV8+OGHSuXKlRULCwslICBASU9PV6ZOnapYWVkpVatWVb766iutuKZNm6bUrl1bMTIyUmrUqKHMmjVLvachISFabQNKSEiIoiiKcu/ePWXUqFGKjY2NYmZmprRv3145efJkgfehKD83RVGUY8eOKYBy+fLlXMcmTZqktGvXTqvs3r17CqDs27dPvS++vr5adSZMmKC0b99eURRFSUlJUdzd3ZVKlSopAwcOVL799lvl/v37iqIoym+//aYAiomJidanQoUKSrNmzRRFUZS9e/cqgPp3IS0tTVm8eLECKFevXlUUJftnpaenp3Ts2FFZuHChcvHiRTWWp3+OiqIoS5YsUX/OiqIoQ4cOVQwMDJTu3bsrwcHByo0bNwp174yNjRVDQ0Ot2A0NDRVAuX//vhIXF6c4ODgo1apVU0aOHKls2rRJSU9PV893dHRUlixZon5PS0tTxo8fr1hYWGjF7+bmptXu4MGDc93zNm3aKJMmTVK/79mzR+nUqZNSpUoVxdTUVI0rNTVVUZTsv3tPtqMoivLDDz/k+fPQ09NT+vXrV6h7kp+i/r0sa5KSkhRASU5OLu1QRBkja7DEM2kM9NExN8t/wTqga2WptWB97ty5zJo1ixMnTqCnp0f//v2ZNm0aS5cu5cCBA1y6dIk5c+YUqv1Vq1Yxc+ZMPvzwQ86dO8dHH33E7NmzWbNmDQDLli1j69at/PDDD8TExBAaGqouJI6MjAQgJCSE+Ph49Xte9uzZw40bN9i/fz+LFy8mICCA7t27Y2VlxbFjxxgzZgxjxozhr7/+Us8xMzMjNDSU6Oholi5dqk7FAfj5+TFlyhTq1atHfHw88fHx+Pn5oSgK3bp1IyEhgbCwMKKiomjUqBEdO3bkzp07hbonhWVkZJTvMeW/I155ebL86em4nKlKyO7/iRMnWLt2Lfb29syZM4eGDRty7949tc4vv/zCyZMn1U90dLTWOqy82gDU8wMCAvjjjz/o1q0be/bswc3Njc2bNwPZU2XK05vZPjVSFBISwpEjR/Dy8mL9+vW4uLhw9OjRfO/Lk+3PmzdPK/YzZ85w4cIFDA0NcXBwICYmhuXLl2NkZMTYsWNp06ZNviNV+cnr/j7rnl+9ehVfX1/q16/Pxo0biYqKYvny5Xn2/en+6OrqEhUVpdWnc+fOsXTp0iLFLIQoHJkiFAUy69QJ9u3PTrKe/IX231/Eln36aC1Ynzp1Kt7e3gBMnDiR/v37s3v3bvU1OCNGjNBa5/MsCxYsICgoiNdeew3InhKJjo7m888/Z8iQIcTFxVG7dm1atWqFRqOhSpUqJCcnA6gLii0tLbWmS/JSsWJFli1bpk7LLFq0iAcPHvD+++8D2WuZPv74Yw4dOsQbb7wBZK8xyuHk5MSUKVNYv34906ZNw8jICFNTU/T09LTa3rNnD2fOnCExMREDAwMAPvnkE7Zs2cKGDRvUdT7FoXbt2hgZGbF7925GjhypdczNzY01a9Zw//59dYrw0KFD6Ojo4OLiUug29PT06NSpE506dWLu3LlYWlqyZ88eOnfujIGBAXFxcVrTgf+Ei4sLLi4uTJo0if79+xMSEsKrr76KjY0NCQkJWsliXq9S8vT0xNPTkxkzZtCiRQu+//57dbotP40aNSImJuaZbw4wMjKiZ8+e9OzZk3feeYc6depw5swZGjVqhL6+PpmZmf+q33k5fvw4GRkZBAUFoaOT/d/HP/zwg1advNr29PRUp6tbt5bNfYV4HiTBEgUyrFOHqq+9lvvFzba28Oc5jJ96kqtBgwbqn3PW8ri7u2uV5axLepZbt27x119/MWLECK3FwhkZGVhYWADZT7p17twZV1dXfHx88PHx+Ud9rFevnvoLKyfG+vXrq991dXWxtrbWinvDhg0EBwdz8eJFUlNT+f/27j2upuz/H/jrdNLldNedbtJVNd3oahiKTOM2U4QUhqEPueUymkGFcRty+2pm8qEwRmYmPvNBjFAumXGrTKNULuUyJYWaUOm0fn/4tT+OUylz0jl5Px+P83jYa6+99nqfFb2tvfbe9fX1UFdXb/E8ly9fRnV1NbS1tUXKnz17hhs3brxR35ujpKSEzz//HAsXLoSCggK8vb3x4MEDXL16FcHBwYiKisKECRMQHR2NBw8eYObMmQgJCeHG7HUOHTqEmzdvol+/ftDS0kJKSgoaGhpgbW0NNTU1zJ8/H3PnzkVDQwP69u2LqqoqnDt3DqqqqtzappY8e/YMCxYsQGBgIHr06IG7d+/i4sWLCAgIAPDibr0HDx5g7dq1CAwMxNGjR3HkyBFuDG7duoX4+HgMHz4c3bp1Q35+PgoKChAaGvracy9duhRDhw6FsbExRo0aBTk5Ofzxxx/IycnBihUrkJiYCKFQCHd3dwgEAuzevRvKysowNTUF8CLhPn36NMaMGQNFRUXu5/Wf6tmzJ+rr67FlyxYMGzYMGRkZ+Pbbb0XqmJmZobq6GidOnICjoyMEAgGsrKwQHByM0NBQrF+/Hs7OzigvL8fJkyfh4OAAf39/ifSPEPI/lGCRFjVehlEfPBhqPj54eukyeKNGQWPwYFh89y0gL/4j9PIljsaZhVfLGi95tKSxzrZt2+Du7i6yj///Z8xcXFxw69YtHDlyBMePH8e4cePQq1evNv/CaOulmt9//x1jxoxBTEwM/Pz8oKGhgaSkJKxfv/61MRkaGjZ5e3x7PBxyyZIlkJeXx9KlS/HXX3/B0NAQYWFhEAgE+PXXXzF79mz06dMHAoEAAQEBiI2NbXXbmpqa2L9/P6Kjo1FTUwNLS0vs3bsXdnZ2AF7MPurp6WHVqlW4efMmNDU14eLiws0Kvg6fz0dFRQVCQ0Nx//596Ojo4JNPPkFMTAyAFwvW4+LisHLlSixfvhwBAQGYP38+4uPjAQACgQDXrl3Dzp07UVFRAUNDQ4SHh2PatGmvPbefnx8OHTqEZcuWYe3atejSpQtsbGy4mUBNTU2sXr0aEREREAqFcHBwwMGDB7nEedmyZZg2bRp69uyJ2tpa1NXVtfp7bYmTkxNiY2OxZs0aREZGol+/fli1apVI0ujl5YWwsDAEBQWhoqICUVFRiI6ORkJCAlasWIF58+bh3r170NbWhqenJyVXhLSXDl0BJqVokfv/uLm5sQULFnDblZWVTFlZmUVFRTHGGAPADhw4wBhj3CL3rKwsrn7jQuZHjx5xZU0twm1O9+7d2bJly1pVlzHGDh06xACw0tJSxhhjXbp0YT///LNIneYWub+sf//+bPbs2SJlLy9cXrduHTM3NxfZP3nyZJG4vvrqK2Zvby9S59ixY4zP57Nbt261OqZGnX0xcWeOj2KTXbTInbwpmsEiLRo4cCASExMxbNgwaGlpYcmSJdzs0dsQHR2NWbNmQV1dHR9++CFqa2tx6dIlPHr0CBEREdiwYQMMDQ3h5OQEOTk5JCcnQ0tLi5sNMjMz49Z/KSoqQktLSyL9srCwwO3bt5GUlIQ+ffrg8OHD3OLrRo3PKcrOzoaRkRHU1NTg6+sLT09PjBw5EmvWrIG1tTX++usvpKSkYOTIkejdu7dE+kcIIaRj0V2EpEWNlyGGDh0Kf39/jBw5knum0NswZcoU/Pvf/0ZiYiIcHBzQv39/JCYmokePHgBePERxzZo16N27N/r06YPi4mIsWbKEW0+1fv16pKamwtjYmHtGkySMGDECc+fORXh4OJycnHDu3DksWbJEpE5AQACGDBmCAQMGQFdXF3v37gWPx0NKSgr69euHTz/9FFZWVhgzZgyKiopavfaJ/HN2dnZQVVVt8rNnz56O7h4hpBPgMdbEO1DecVVVVdDQ0EB5ebnYYuTO4Pnz50hJSYG/v3+rn4otKyg22fU24ysuLm72sQb6+vpQU1OT6Pk689h15tgAoKKiAjo6OqisrHztTSyEvIwuEZJmCRuEyCzLxIOnD6Ar0IWLngv4ch3z/kBCJKnxbj9CCGkvlGCRJh0vPo7VF1bj/tP7XJm+QB+L3BbB19RXYudpfE9aU44cOULP7CGEECKTKMEiYo4XH0dEegQYRK8elz0tQ0R6BGI/iJVYktXUgyEbde/eXSLnIIQQQt42SrCICGGDEKsvrBZLrgCAgYEHHtZcWIMBxgMkcrmwpSdlE0IIIbKK7iIkIjLLMkUuC76KgaH0aSkyyzLfYq8IIYQQ2UIJFhHx4OkDidYjhBBC3kWUYBERugJdidYjhBBC3kWUYBERLnou0Bfogwdek/t54MFAYAAXPZd270t6ejp4PB4eP37c7ueSpA8++ABz5szp6G4QQgjpQJRgERF8OT4WuS0CALEkq3H7c7fP6XlYaD4B3L9/P5YvX97u56dEjhBCpBclWESMr6kvYj+IhZ5AT6RcX6Av0Uc0dFZdu3aV+JPACSGEyBZKsEiTfE198WvAr9g2aBumvjcVUx2mYkXfFRhgPOCN2/zggw8wc+ZMzJkzB1paWtDX10d8fDyePHmCSZMmQU1NDT179sSRI0eaPD4xMRGampo4dOgQrK2tIRAIEBgYiCdPnmDnzp0wMzODnp4e4uPjIRQKW9Wnuro6LFy4EN27d4eKigrc3d2Rnp7O7S8uLuZedK2iogI7OzukpKSgqKgIAwa8+C60tLTA4/EwceJELs6XZ5bMzMywYsUKhIaGQlVVFaampvjll1/w4MEDjBgxAqqqqnBwcMClS5e4YyoqKjB27FgYGRlBIBDAwcEBSUlJ3P6JEyfi1KlT2LRpE3g8Hng8HoqKigAAubm58Pf3h6qqKvT19RESEoLy8vJWfR+EEEIkgxIs0qy0O2lYnLEY8X/EIz4nHlOOTYFfsh+OFx9/4zZ37twJHR0dXLhwATNnzsS//vUvjBo1Cl5eXsjMzISfnx9CQkLw9OnTJo9/+vQpNm/ejKSkJBw9ehTp6en45JNPkJKSgpSUFCQkJODYsWNITk5uVX8mTZqEjIwMJCUl4Y8//sCoUaMwZMgQFBYWAgBmzJiB2tpanD59Gjk5OVizZg1UVVVhbGzMnSM/Px8lJSXYtGlTs+fZsGEDvL29kZWVhY8++gghISEIDQ3F+PHjkZmZCQsLC4SGhqLx1aA1NTVwdXXFoUOH8Oeff2Lq1KmYNGkSCgoKAACbNm2Cp6cnPvvsM5SUlKCkpATGxsYoKSlB//794eTkhEuXLuHo0aO4f/8+Ro8e3eoxIoQQIgGMiKmsrGQAWHl5eUd3pV3U1dWx//znP6yurq7ZOqlFqcwh0YHZJ9qLfBwSHZhDogNLLUpt83n79+/P+vbty23X19czFRUVFhISwpWVlJQwAOy3335jaWlpDAB79OgRY4yxhIQEBoBdv36dqz9t2jQmEAjY33//zcXm7OzMPvvss9f25/r164zH47F79+6JlPv4+LDIyEjGGGMODg4sOjq6yeNf7d/Lcc6ePZvbNjU1ZePHjxeLccmSJVzZb7/9xgCwkpKSZvv74YcfshEjRnDj9up5GGNsyZIlbPDgwSJld+7cYQBYfn5+s21Lg9b8XMoqik12lZeXMwCssrKyo7tCZAw9yZ2Iac+nub/33nvcn/l8PrS1teHg4MCV6evrAwDKysqafHO9QCBAz549ReqbmZmJvNNQU1MTZWVlr+1LZmYmGGOwsrISKa+trYW2tjYAYNasWfjXv/6FY8eOwdfXFwEBASIxtNbLxzTG2FzcBgYGEAqFWL16Nfbt24d79+6htrYWtbW1cHNza/E8ly9fRlpaWpPveLxx44ZYrIQQQtoHJVhETHNPc+c38GFTaQNFoSLkmBy279kO9S7qqK+vR319PYRCocifBw4cCHt7e5E2unTpIrLN4/FEyni8F3cqNjQ0NNm31x3fqLnjX63D5/Nx+fJl8PmiiWJjgjJlyhT4+fnh8OHDOHbsGFatWoX169dj5syZr22/uX43xthS3OvXr8eGDRuwceNGODg4QEVFBbNnz37tWqqGhgYMGzYMa9asEdtnaGjYpj4TQgh5cx2+BisuLg49evSAkpISXF1dcebMmWbrNt4W/+rn2rVrIvWSk5PRq1cvKCoqolevXjhw4EB7h9GpNPeUdj7jQ6dGB5p1mlCpV8H9yvtgjEFBQQHq6urQ0dGBkZERLCwsYGdnx83KSCtnZ2cIhUKUlZXBwsJC5GNgYMDVMzY2RlhYGPbv34958+Zh27ZtAAAFBQUAaPWC+rY4c+YMRowYgfHjx8PR0RHm5ubcurBGCgoKYud2cXHB1atXYWZmJhaTioqKxPtJCCGkaR06g7Vv3z7MmTMHcXFx8Pb2xnfffYcPP/wQubm5MDExafa4/Px8kctHurr/e6r4b7/9hqCgICxfvhwff/wxDhw4gNGjR+Ps2bNwd3dv13g6i+ae0l7Hr8Mpw1PctpaiFr4Y/YXMPhPLysoKwcHBCA0Nxfr16+Hs7Izy8nKcPHkSDg4O8Pf3x5w5c/Dhhx/CysoKjx49wsmTJ2FrawsAMDU1BY/Hw6FDh+Dv7w9lZeUmL829CQsLCyQnJ+PcuXPQ0tJCbGws7t+/Dy0tLa6OmZkZzp8/j6KiIqiqqqJr166YMWMGtm3bhrFjx2LBggXQ0dHB9evXkZSUhG3btonN1BFCCGkfHTqDFRsbi8mTJ2PKlCmwtbXFxo0bYWxsjG+++abF4/T09GBgYMB9Xv6lsXHjRgwaNAiRkZGwsbFBZGQkfHx8sHHjxnaOpvNw0XOBlqJWs/tVnqtA9bkqHtU8kvmXPickJCA0NBTz5s2DtbU1hg8fjvPnz8PY2BjAi9mpGTNmwNbWFkOGDIG1tTXi4uIAAN27d0dMTAwWLVoEfX19hIeHS6xfS5YsgYuLC/z8/PDBBx/AwMAAw4cPF6kzf/588Pl89OrVC7q6urh9+za6deuGjIwMCIVC+Pn5wd7eHrNnz4aGhgbk5Dp8wpoQQt4ZPMaY+Ermt6Curg4CgQA//fQTPv74Y6589uzZyM7OxqlTp8SOSU9Px4ABA2BmZoaamhr06tULixcv5p5HBAAmJiaYO3cu5s6dy5U1rmUpLi5usi+NC4gbVVVVcbe8Ny527kyeP3+O1NRUDBo0qMn1SwCw7vI6/JD/Q5P7vO57oduzbqiWr0YP8x7wcfGBiYkJ5OU7fklfa2KTVZ05NqBzx0exya6KigoYGhqisrKyyRtvCGlOh/1GLC8vh1AoFFuno6+vj9LS0iaPMTQ0RHx8PFxdXVFbW4vdu3fDx8cH6enp6NevHwCgtLS0TW0CwKpVqxATEyNWnpaWBoFA0NbQZEZqamqz+5SeKzW777zueejV6MHgmQHUbqkhqSAJcnJyUFVVhYaGBtTU1Lj1SR2lpdhkXWeODejc8VFssqe5Z/IR8jodPuXQePdUI8aYWFkja2trWFtbc9uenp64c+cO1q1bxyVYbW0TACIjIxEREcFtN85gDRgw4J2dwRI2CHHov4dQ9lT8cQdCOSFKBaVo0GnAd8O+w8OKh7h+/TquX7+Ou3fvgjEGfX19bnG1oaHhW7s89XJs58+fx7Bhw5qt++jRo7fSJ0np7DMFnTk+ik12VVRUdHQXiIzqsARLR0cHfD5fbGaprKysTXefeXh44Pvvv+e2DQwM2tymoqIiFBUVxcq7dOnSKf/BaNRSfF3QBZFukYhIf5F4vvxMrMaXPi9yWwRlJWV0794d3bt3R//+/fHs2TNcv34dhYWFyMzMREZGBgQCASwsLGBpaQkLCwsoKTU/OybJ2Dw8PJCdnd1iHVn0Lv9cyjqKTfZ0xpjI29FhCZaCggJcXV2RmpoqsgYrNTUVI0aMaHU7WVlZIs/38fT0RGpqqsgarGPHjsHLy0syHX+HNL70efWF1SLPxdIX6ONzt8+bfOmzsrIyHBwc4ODggIaGBty9exeFhYUoLCzEH3/8AR6PBxMTE1haWsLS0hK6urotzi7+E8rKyrCwsGiXtgkhhJCWdOglwoiICISEhKB3797w9PREfHw8bt++jbCwMAAvLt3du3cPu3btAvDiDkEzMzPY2dmhrq4O33//PZKTk0XeOzd79mz069cPa9aswYgRI/DLL7/g+PHjOHv2bIfEKOt8TX0xwHgAMssy8eDpA+gKdOGi59KqRzPIycnBxMQEJiYm8PHxQWVlJZdspaen4/jx49DQ0IClpSWsrKxgZmZG/1skhBDSKXRoghUUFISKigosW7YMJSUlsLe3R0pKCkxNTQEAJSUluH37Nle/rq4O8+fPx71796CsrAw7OzscPnwY/v7+XB0vLy8kJSVh8eLFWLJkCXr27Il9+/bRM7D+Ab4cH30M+vzjdjQ0NNC7d2/07t0b9fX1KCoqQkFBAQoLC3Hp0iXIy8ujR48esLKygqWlJTQ0NCTQe0IIIeTt6/BF7tOnT8f06dOb3JeYmCiyvXDhQixcuPC1bQYGBiIwMFAS3SPtRF5enlsEzxhDeXk5l2ylpKSAMQY9PT1udsvIyIie40QIIURmdHiCRQiPx4Ouri50dXXh7e2Nmpoa3LhxA4WFhcjKykJGRgZUVFQwdepUeg4NIYQQmUAJFpE6SkpKsLOzg52dHRhjuHfvHkpKSqCsrNzRXSOEEEJahRIsItV4PB6MjIxgZGTU0V0hhBBCWo0WtRAiRRITE6GpqdnR3SCEEPIPUYJFSAcxMzMTewl5UFAQCgoK2v3clMgRQkj7okuEhEgRZWVlWmtGCCGdAM1gETQ0NGDNmjWwsLCAoqIiTExM8NVXXwEAcnJyMHDgQCgrK0NbWxtTp05FdXU1d+zEiRMxcuRIrFu3DoaGhtDW1saMGTPw/Plzrk5cXBwsLS2hpKQEfX19kUdoMMawdu1amJubQ1lZGY6Ojvj555+5/enp6eDxeDhx4gR69+4NgUAALy8v5Ofnc3WuXLmCAQMGQE1NDdra2oiIiMDly5cBANHR0XBychKJt/GBtS+fw83NDSoqKtDU1IS3tzeKi4tb9d0dPHgQrq6uUFJSgrm5OWJiYlBfX8/tj46OhomJCRQVFdGtWzfMmjULAPDBBx+guLgYc+fOBY/H455m/+rMUmP/d+zYgZ49e2LMmDEIDw+HUCjE2rVrYWBgAD09PW68GsXGxsLBwQEqKiowNjbG9OnTuXFLT0/HpEmTUFlZyZ07OjoawItnzS1cuBDdu3eHiooK3N3dkZ6e3qrvghBCyP/QDBZBZGQktm3bhg0bNqBv374oKSnBtWvX8PTpUwwZMgQeHh64ePEiysrKMGXKFISHh4s8oywtLQ2GhoZIS0vD9evXERQUBCcnJ3z22We4dOkSZs2ahd27d8PLywsPHz7EmTNnuGMXL16M/fv345tvvoGlpSVOnz6N8ePHQ1dXF/379+fqffnll1i/fj10dXURFhaGTz/9FBkZGQCA4OBgODs745tvvkFDQwMSExMhL9+6H+36+nqMHDkSn332Gfbu3Yu6ujpcuHChVa/v+fXXXzF+/Hhs3rwZ77//Pm7cuIGpU6cCAKKiovDzzz9jw4YNSEpKgp2dHUpLS3HlyhUAwP79++Ho6IipU6fis88+a/E8N27cwJEjR3Dw4EH8+OOPWLduHYqLi2FlZYVTp07h3Llz+PTTT+Hj4wMPDw8AL56iv3nzZpiZmeHWrVuYPn06Fi5ciLi4OHh5eWHjxo1YunQpl6iqqqoCACZNmoSioiIkJSWhW7duOHDgAIYMGYKcnBxYWlq26jslhBACgBExlZWVDAArLy/v6K60i7q6Ovaf//yH1dXVsaqqKqaoqMi2bdsmVi8+Pp5paWmx6upqruzw4cNMTk6OlZaWMsYYmzBhAjM1NWX19fVcnVGjRrGgoCDGGGPJyclMXV2dVVVVibVfXV3NlJSU2Llz50TKJ0+ezMaOHcsYYywtLY0BYMePHxfpAwD27NkzxhhjampqLDExUSw2xhiLiopijo6OIu1v2LCBmZqaMsYYq6ioYABYenr667+4V7z//vts5cqVImW7d+9mhoaGjDHG1q9fz6ysrLi+vMrU1JRt2LBBpCwhIYFpaGhw21FRUUwgELCqqioutsGDBzMzMzMmFAq5etbW1mzVqlXN9vXHH39k2trazZ6HMcauX7/OeDweu3fvnki5j48Pi4yMbLZtSXl17DoTik12lZeXMwCssrKyo7tCZAzNYL3j8vLyUFtbCx8fnyb3OTo6QkVFhSvz9vZGQ0MD8vPzoa+vDwCws7MDn/+/dxMaGhoiJycHADBo0CCYmprC3NwcQ4YMwZAhQ/Dxxx9DIBAgNzcXNTU1GDRokMh56+rq4OzsLFL23nvvibQPAGVlZTAxMUFERASmTJmC3bt3Y8CAAdDW1m51/F27dsXEiRPh5+eHQYMGwdfXF6NHjxZ5gXhzLl++jIsXL4pcnhMKhaipqcHTp08xatQobNy4kYvd398fw4YNa/XsWiMzMzOoqalxl1319PQgLy8v8mR7fX19lJWVcdtpaWlYuXIlcnNzUVVVhfr6etTU1ODJkyci4/myzMxMMMZgZWUlUl5bW9um75QQQghdInzntbSgmjHW7KWyl8tffUEzj8dDQ0MDAEBNTQ2ZmZlIT0/HsWPHsHTpUkRHR+PixYtcncOHD6N79+4ibSgqKopsv3yOxnM3Hh8dHY1x48bh8OHDSElJQXp6OrS1tTFq1CjIycmBMSbS1svrwwAgISEBs2bNwtGjR7Fv3z4sXrwYqamp3OW25jQ0NCAmJgaffPKJ2D4lJSUYGxsjPz8fqampOH78OKZPn46vv/4ap06datNLrZv6flv6zouLi+Hv74+wsDAsX74cXbt2xdmzZzF58mSx2F+Nh8/n4/LlyyIJM/C/S4iEEEJahxKsd5ylpSWUlZVx4sQJTJkyRWRfr169sHPnTpFZj4yMDMjJyYnNcrREXl4evr6+8PX1RVRUFDQ1NXHy5EkMGjQIioqKuH37tsh6qzdhZWUFKysrhIeHw8fHBzt37sSoUaOgq6uL0tJSkWQxOztb7HhnZ2c4OzsjMjISnp6e+OGHH16bYLm4uCA/Px8WFhbN1lFWVsbw4cMxfPhwzJgxAzY2NsjJyYGLiwsUFBQgFAr/UdxNuXTpEurr67F+/XpuluvHH38UqdPUuZ2dnSEUClFWVob3339f4v0ihJB3CSVY7zglJSV8/vnnWLhwIRQUFODt7Y0HDx7g6tWrCA4ORlRUFCZMmIDo6Gg8ePAAM2fOREhICHd58HUOHTqEmzdvol+/ftDS0kJKSgoaGhpgbW0NNTU1zJ8/H3PnzkVDQwP69u2LqqoqnDt3DqqqqpgwYcJr23/27BkWLFiAwMBA9OjRA0VFRSgsLERwcDCAF3frPXjwAGvXrkVgYCCOHj2KI0eOcO80vHXrFuLj4zF8+HB069YN+fn5KCgoQGho6GvPvXTpUgwdOhTGxsbcbNkff/yBnJwcrFixAomJiRAKhXB3d4dAIMDu3buhrKwMU1NTAC8u/Z0+fRpjxoyBoqIidHR0WvWdvk7Pnj1RX1+PLVu2YNiwYcjIyMC3334rUsfMzAzV1dU4ceIEHB0dIRAIYGVlheDgYISGhmL9+vVwdnZGeXk5Tp48CQcHB/j7+0ukf4QQ8i6gxzQQLFmyBPPmzcPSpUtha2uLoKAglJWVQSAQ4Ndff8XDhw/Rp08fBAYGwsfHB//3f//X6rY1NTWxf/9+DBw4ELa2tvj222+xd+9e2NnZAQCWL1+OpUuXYtWqVbC1tYWfnx8OHjyIHj16tKp9Pp+PiooKhIaGwsrKCuPGjYOLiwuioqIAALa2toiLi8PWrVvh6OiICxcuYP78+dzxAoEA165dQ0BAAKysrDB16lSEh4dj2rRprz23n58fDh06hNTUVPTp0wceHh6IjY3lEihNTU1s27YN3t7eeO+993DixAkcPHiQW8+0bNkyFBUVoWfPntDV1W31d/o6Tk5OiI2NxZo1a2Bvb489e/Zg1apVInW8vLwQFhaGoKAg6OrqYu3atQBeXC4NDQ3FvHnzYG1tjeHDh+P8+fMwNjaWWP8IIeRdwGOvLlAhqKqqgoaGBsrLyzvl4t7nz58jJSUF/v7+bVoLJAsoNtnVmeOj2GRXRUUFdHR0UFlZyc18E9IaNINFCCGEECJhlGAR0gw7Ozuoqqo2+dmzZ09Hd48QQogUo0XuhDQjJSWl2ccatHaRPyGEkHcTJViENKNxsTohhBDSVnSJkBBCCCFEwijBIoQQQgiRMEqwCCGEEEIkjBIsQgghhBAJowSLEEIIIUTCKMEihBBCCJEwSrAIIYQQQiSMEixCCCGEEAmjBIsQQgghRMIowSKEEEIIkTBKsAghhBBCJIwSLEIIIYQQCaMEixBCCCFEwijBIoQQQgiRMEqwCCGEEEIkjBIsQgghhBAJowSLEEIIIUTCKMEihBBCCJEwSrAIIYQQQiSMEixCCCGEEAmjBIsQQgghRMIowSKEEEIIkTBKsAghhBBCJIwSLEIIIYQQCaMEixBCCCFEwjo8wYqLi0OPHj2gpKQEV1dXnDlzplXHZWRkQF5eHk5OTiLliYmJ4PF4Yp+ampp26D0hhBBCiLgOTbD27duHOXPm4Msvv0RWVhbef/99fPjhh7h9+3aLx1VWViI0NBQ+Pj5N7ldXV0dJSYnIR0lJqT1CIIQQQggR06EJVmxsLCZPnowpU6bA1tYWGzduhLGxMb755psWj5s2bRrGjRsHT0/PJvfzeDwYGBiIfAghhBBC3hb5jjpxXV0dLl++jEWLFomUDx48GOfOnWv2uISEBNy4cQPff/89VqxY0WSd6upqmJqaQigUwsnJCcuXL4ezs3OzbdbW1qK2tpbbrqysBAA8fPiwLSHJjOfPn+Pp06eoqKhAly5dOro7EkWxya7OHB/FJrsafw8wxjq4J0TWdFiCVV5eDqFQCH19fZFyfX19lJaWNnlMYWEhFi1ahDNnzkBevumu29jYIDExEQ4ODqiqqsKmTZvg7e2NK1euwNLSssljVq1ahZiYGLFyKyurNkZFCCGkM6qoqICGhkZHd4PIkA5LsBrxeDyRbcaYWBkACIVCjBs3DjExMS0mPh4eHvDw8OC2vb294eLigi1btmDz5s1NHhMZGYmIiAhu+/HjxzA1NcXt27c75V+oqqoqGBsb486dO1BXV+/o7kgUxSa7OnN8FJvsqqyshImJCbp27drRXSEypsMSLB0dHfD5fLHZqrKyMrFZLQD4+++/cenSJWRlZSE8PBwA0NDQAMYY5OXlcezYMQwcOFDsODk5OfTp0weFhYXN9kVRURGKiopi5RoaGp3yH4xG6urqnTY+ik12deb4KDbZJSfX4TfdExnTYT8xCgoKcHV1RWpqqkh5amoqvLy8xOqrq6sjJycH2dnZ3CcsLAzW1tbIzs6Gu7t7k+dhjCE7OxuGhobtEgchhBBCyKs69BJhREQEQkJC0Lt3b3h6eiI+Ph63b99GWFgYgBeX7u7du4ddu3ZBTk4O9vb2Isfr6elBSUlJpDwmJgYeHh6wtLREVVUVNm/ejOzsbGzduvWtxkYIIYSQd1eHJlhBQUGoqKjAsmXLUFJSAnt7e6SkpMDU1BQAUFJS8tpnYr3q8ePHmDp1KkpLS6GhoQFnZ2ecPn0abm5urW5DUVERUVFRTV427Aw6c3wUm+zqzPFRbLKrs8dH2g+P0b2nhBBCCCESRav2CCGEEEIkjBIsQgghhBAJowSLEEIIIUTCKMEihBBCCJGwdybBiouLQ48ePaCkpARXV1ecOXOmVcdlZGRAXl4eTk5OIuWJiYng8Xhin5qamnbofcvaElt6enqT/b527ZpIveTkZPTq1QuKioro1asXDhw40N5hNEnSsUnTuAFt/7msra3Fl19+CVNTUygqKqJnz57YsWOHSB1ZHDvg9bFJ09i1JbaJEyc22W87OzuRetIyboDk45PVsQOAPXv2wNHREQKBAIaGhpg0aRIqKipE6kjT2BEpwt4BSUlJrEuXLmzbtm0sNzeXzZ49m6moqLDi4uIWj3v8+DEzNzdngwcPZo6OjiL7EhISmLq6OispKRH5vG1tjS0tLY0BYPn5+SL9rq+v5+qcO3eO8fl8tnLlSpaXl8dWrlzJ5OXl2e+///62wmKMtU9s0jJujL3Zz+Xw4cOZu7s7S01NZbdu3WLnz59nGRkZ3H5ZHTvGXh+btIxdW2N7/PixSH/v3LnDunbtyqKiorg60jJujLVPfLI6dmfOnGFycnJs06ZN7ObNm+zMmTPMzs6OjRw5kqsjTWNHpMs7kWC5ubmxsLAwkTIbGxu2aNGiFo8LCgpiixcvZlFRUU0mWBoaGhLuadu1NbbGJOTRo0fNtjl69Gg2ZMgQkTI/Pz82ZsyYf9zftmiP2KRl3Bhre3xHjhxhGhoarKKiotk2ZXXsWhObtIzdm/570ujAgQOMx+OxoqIirkxaxo2x9olPVsfu66+/Zubm5iJlmzdvZkZGRty2NI0dkS6d/hJhXV0dLl++jMGDB4uUDx48GOfOnWv2uISEBNy4cQNRUVHN1qmuroapqSmMjIwwdOhQZGVlSazfrfGmsQGAs7MzDA0N4ePjg7S0NJF9v/32m1ibfn5+r21TktorNqDjxw14s/j++9//onfv3li7di26d+8OKysrzJ8/H8+ePePqyOrYtSY2oOPH7p/8XDbavn07fH19uQcqA9IxbkD7xQfI5th5eXnh7t27SElJAWMM9+/fx88//4yPPvqIqyMtY0ekT4c+yf1tKC8vh1AoFHuBtL6+vtiLphsVFhZi0aJFOHPmDOTlm/6KbGxskJiYCAcHB1RVVWHTpk3w9vbGlStXYGlpKfE4mvImsRkaGiI+Ph6urq6ora3F7t274ePjg/T0dPTr1w8AUFpa2qY220N7xSYN4wa8WXw3b97E2bNnoaSkhAMHDqC8vBzTp0/Hw4cPubVKsjp2rYlNGsbuTWJ7WUlJCY4cOYIffvhBpFwaxg1ov/hkdey8vLywZ88eBAUFoaamBvX19Rg+fDi2bNnC1ZGWsSPSp9MnWI14PJ7INmNMrAwAhEIhxo0bh5iYGFhZWTXbnoeHBzw8PLhtb29vuLi4YMuWLdi8ebPkOt4KrY0NAKytrWFtbc1te3p64s6dO1i3bh2XhLS1zfYk6dikadyAtsXX0NAAHo+HPXv2QENDAwAQGxuLwMBAbN26FcrKym1usz1JOjZpGrs3/Y4TExOhqamJkSNHSqzN9iDp+GR17HJzczFr1iwsXboUfn5+KCkpwYIFCxAWFobt27e/UZvk3dHpLxHq6OiAz+eL/W+irKxM7H8dAPD333/j0qVLCA8Ph7y8POTl5bFs2TJcuXIF8vLyOHnyZJPnkZOTQ58+fVBYWNgucTSlrbE1x8PDQ6TfBgYG/7jNf6q9YntVR4wb8GbxGRoaonv37lwCAgC2trZgjOHu3bsAZHfsWhPbq2Tt7xxjDDt27EBISAgUFBRE9knDuAHtF9+rZGXsVq1aBW9vbyxYsADvvfce/Pz8EBcXhx07dqCkpASA9IwdkT6dPsFSUFCAq6srUlNTRcpTU1Ph5eUlVl9dXR05OTnIzs7mPmFhYbC2tkZ2djbc3d2bPA9jDNnZ2TA0NGyXOJrS1tiak5WVJdJvT09PsTaPHTvWpjb/qfaK7VUdMW7Am8Xn7e2Nv/76C9XV1VxZQUEB5OTkYGRkBEB2x641sb1K1v7OnTp1CtevX8fkyZPF9knDuAHtF9+rZGXsnj59Cjk50V+TfD4fwIsYAOkZOyKF3uaK+o7SeGvu9u3bWW5uLpszZw5TUVHh7nJZtGgRCwkJafb4pu4ijI6OZkePHmU3btxgWVlZbNKkSUxeXp6dP3++PUMR09bYNmzYwA4cOMAKCgrYn3/+yRYtWsQAsOTkZK5ORkYG4/P5bPXq1SwvL4+tXr26Q2/1l2Rs0jJubxLf33//zYyMjFhgYCC7evUqO3XqFLO0tGRTpkzh6sjq2LUmNmkZuzf992T8+PHM3d29yTalZdwYa5/4ZHXsEhISmLy8PIuLi2M3btxgZ8+eZb1792Zubm5cHWkaOyJd3okEizHGtm7dykxNTZmCggJzcXFhp06d4vZNmDCB9e/fv9ljm0qw5syZw0xMTJiCggLT1dVlgwcPZufOnWun3resLbGtWbOG9ezZkykpKTEtLS3Wt29fdvjwYbE2f/rpJ2Ztbc26dOnCbGxsRJKUt0nSsUnTuDHW9p/LvLw85uvry5SVlZmRkRGLiIhgT58+Fakji2PH2Otjk6axa2tsjx8/ZsrKyiw+Pr7ZNqVl3BiTfHyyPHabN29mvXr1YsrKyszQ0JAFBwezu3fvitSRprEj0oPH2P+f5ySEEEIIIRLR6ddgEUIIIYS8bZRgEUIIIYRIGCVYhBBCCCESRgkWIYQQQoiEUYJFCCGEECJhlGARQgghhEgYJViEEEIIIRJGCRYhhBBCiIRRgkUIIYQQImGUYBEiZSZOnAgej4ewsDCxfdOnTwePx8PEiRO5stLSUsycORPm5uZQVFSEsbExhg0bhhMnTogdv3LlSvD5fKxevVpsX3R0NJycnMTKHz9+DB6Ph/T0dABAUVEReDwe99HQ0ICHhwcOHjwoduyzZ88QFRUFa2trKCoqQkdHB4GBgbh69apY3aqqKnz55ZewsbGBkpISDAwM4Ovri/3794NeOEEIkTWUYBEihYyNjZGUlIRnz55xZTU1Ndi7dy9MTEy4sqKiIri6uuLkyZNYu3YtcnJycPToUQwYMAAzZswQazchIQELFy7Ejh07/nEfjx8/jpKSEpw/fx5ubm4ICAjAn3/+ye2vra2Fr68vduzYgeXLl6OgoAApKSkQCoVwd3fH77//ztV9/PgxvLy8sGvXLkRGRiIzMxOnT59GUFAQFi5ciMrKyn/cX0IIeZvkO7oDhBBxLi4uuHnzJvbv34/g4GAAwP79+2FsbAxzc3OuXuOM1oULF6CiosKV29nZ4dNPPxVp89SpU3j27BmWLVuGXbt24fTp0+jXr98b91FbWxsGBgYwMDDAV199hS1btiAtLQ329vYAgI0bN+K3335DVlYWHB0dAQCmpqZITk6Gu7s7Jk+ejD///BM8Hg9ffPEFioqKUFBQgG7dunHnsLKywtixY6GkpPTG/SSEkI5AM1iESKlJkyYhISGB296xY4dI0vTw4UMcPXoUM2bMEEmuGmlqaopsb9++HWPHjkWXLl0wduxYbN++XSL9fP78ObZt2wYA6NKlC1f+ww8/YNCgQVxy1UhOTg5z585Fbm4urly5goaGBiQlJSE4OFgkuWqkqqoKeXn6vyAhRLZQgkWIlAoJCcHZs2dRVFSE4uJiZGRkYPz48dz+69evgzEGGxub17ZVVVWF5ORk7vjx48fj559/RlVV1Rv3z8vLC6qqqlBSUsK8efNgZmaG0aNHc/sLCgpga2vb5LGN5QUFBSgvL8ejR49aFQchhMgKSrAIkVI6Ojr46KOPsHPnTiQkJOCjjz6Cjo4Ot79x4TePx3ttWz/88APMzc252SQnJyeYm5sjKSnpjfu3b98+ZGVl4b///S8sLCzw73//G127dm3VsS/3vS1xEEKIrKB5d0Kk2Kefforw8HAAwNatW0X2WVpagsfjIS8vDyNHjmyxnR07duDq1asil9oaGhqwfft2TJ06FQCgrq7e5GLyx48fAwA0NDREyo2NjWFpaQlLS0uoqqoiICAAubm50NPTA/Bi/VRubm6T/bl27RoXg66uLrS0tJCXl9diDIQQIktoBosQKTZkyBDU1dWhrq4Ofn5+Ivu6du0KPz8/bN26FU+ePBE7tjExysnJwaVLl5Ceno7s7Gzuc/r0aVy8eJG788/GxgZ3795FaWmpSDsXL16EnJwcLCwsmu1n//79YW9vj6+++oorGzNmDI4fP44rV66I1G1oaMCGDRvQq1cvODo6Qk5ODkFBQdizZw/++usvsbafPHmC+vr6lr8oQgiRMpRgESLF+Hw+8vLykJeXBz6fL7Y/Li4OQqEQbm5uSE5ORmFhIfLy8rB582Z4enoCeLG43c3NDf369YO9vT336du3Lzw9PbnF7oMHD4atrS3GjBmDjIwM3Lp1C7/88gvmz5+PsLAwqKmptdjXefPm4bvvvsO9e/cAAHPnzoWbmxuGDRuGn376Cbdv38bFixcREBCAvLw8bN++nbssuHLlShgbG8Pd3R27du1Cbm4uCgsLsWPHDjg5OaG6ulqSXyshhLQ7SrAIkXLq6upQV1dvcl+PHj2QmZmJAQMGYN68ebC3t8egQYNw4sQJfPPNN6irq8P333+PgICAJo8PCAjA999/j7q6OsjLy+PYsWMwNzdHcHAw7OzssGjRIkyZMgWxsbGv7efQoUNhZmbGzWIpKSnh5MmTmDBhAr744gtYWFhgyJAh4PP5+P333+Hh4cEdq6Wlhd9//x3jx4/HihUr4OzsjPfffx979+7F119/LXZ5khBCpB2P0SOSCSGEEEIkimawCCGEEEIkjBIsQgghhBAJowSLEEIIIUTCKMEihBBCCJEwSrAIIYQQQiSMEixCCCGEEAmjBIsQQgghRMIowSKEEEIIkTBKsAghhBBCJIwSLEIIIYQQCaMEixBCCCFEwv4fGlE1dY7T49MAAAAASUVORK5CYII=",
      "text/plain": [
       "<Figure size 500x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "dc.plot_metrics_scatter(\n",
    "    df, groupby='net', x='mcauroc', y='mcauprc'\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "68d852c7",
   "metadata": {},
   "source": [
    "Since this plot is too crowded, we can plot it again separating the methods by columns:"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7c0f8a55-0df8-480f-8a4f-a857e653b783",
   "metadata": {},
   "source": [
    "このプロットは混雑しすぎているので、メソッドを列で区切って再度プロットしてみよう："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "1a446ffd",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 700x200 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "dc.plot_metrics_scatter_cols(\n",
    "    df, col='method', x='auroc', y='auprc',\n",
    "    figsize=(7, 2), groupby='net'\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9a71056c",
   "metadata": {},
   "source": [
    "Or by selecting only one of the methods, in this case `consensus`:"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3d48bc31-7df8-4ef6-997c-0f86c4f079b4",
   "metadata": {},
   "source": [
    "あるいは、いずれかの方法（この場合はコンセンサス）のみを選択することによっても："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "b1851c3e",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 200x200 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "dc.plot_metrics_scatter(\n",
    "    df[df['method'] == 'consensus_estimate'],\n",
    "    groupby='net', x='mcauroc', y='mcauprc',\n",
    "    show_text=False, figsize=(2, 2)\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "16dc3380",
   "metadata": {},
   "source": [
    "As expected, the random network has a performance close to 0.5 while the rest show higher predictive performance. As previously reported, the collectri network has better performance than dorothea.\n",
    "\n",
    "If needed, we can also plot the distributions of the Monte-Carlo metrics, now grouped by network:"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fb9e460e-7661-42b3-b5d7-4e415c65f427",
   "metadata": {},
   "source": [
    "予想通り、ランダムネットワークは0.5に近い性能を示したが、その他のネットワークはより高い予測性能を示した。以前に報告したように、collectriネットワークはdorotheaよりも優れた性能を持つ。\n",
    "\n",
    "必要であれば、モンテカルロメトリクスの分布をプロットすることもできる："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "36aedb27",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 600x500 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axes = plt.subplots(\n",
    "    2, 2, figsize=(6, 5), sharex=True, tight_layout=True\n",
    ")\n",
    "\n",
    "dc.plot_metrics_boxplot(df, metric='mcauroc', groupby='net', ax=axes[0, 0], legend=False)\n",
    "\n",
    "dc.plot_metrics_boxplot(df, metric='mcauprc', groupby='net', ax=axes[0, 1], legend=True)\n",
    "\n",
    "dc.plot_metrics_boxplot(df, metric='rank', groupby='net', ax=axes[1, 0], legend=False)\n",
    "\n",
    "dc.plot_metrics_boxplot(df, metric='nrank', groupby='net', ax=axes[1, 1])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8ccaa001",
   "metadata": {},
   "source": [
    "### By source\n",
    "\n",
    "The benchmark pipeline also allows to test the performance of individual regulators using the argument `by='source'`. For simplicity let us just use the best performing network:"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ca6d36bc-556c-4117-b9d7-cee6320bc7e1",
   "metadata": {},
   "source": [
    "ベンチマークパイプラインでは、引数by='source'を使用して、個々のレギュレータのパフォーマンスをテストすることもできます。簡単のため、最も性能の良いネットワークだけを使用してみましょう："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "8f77f542-a512-4cd8-b6c9-7ccf72435b13",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Example on how to set up decouple arguments\n",
    "decouple_kws={\n",
    "    'methods' : ['ulm', 'mlm'],\n",
    "    'consensus': True,\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "389c20fb",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Extracting inputs...\n",
      "Formating net...\n",
      "Removed 109 experiments without sources in net.\n",
      "Running 279 experiments for 155 unique sources.\n",
      "Running methods...\n",
      "52 features of mat are empty, they will be removed.\n",
      "Running ulm on mat with 279 samples and 21933 targets for 766 sources.\n",
      "52 features of mat are empty, they will be removed.\n",
      "Running mlm on mat with 279 samples and 21933 targets for 766 sources.\n",
      "Calculating metrics...\n",
      "Computing metrics...\n",
      "Done.\n"
     ]
    }
   ],
   "source": [
    "# Run benchmark pipeline\n",
    "df = dc.benchmark(\n",
    "    mat, obs, collectri,\n",
    "    perturb='TF', sign=-1, by='source', verbose=True,\n",
    "    decouple_kws=decouple_kws\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "21a76e8b",
   "metadata": {},
   "source": [
    "We can plot the results with:"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f75d29c7-b2c5-474b-bd91-8a2ff769ba42",
   "metadata": {},
   "source": [
    "その結果をプロットすることができる："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "d3ff2d43",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 600x500 with 12 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "dc.plot_metrics_scatter_cols(\n",
    "    df, col='source',\n",
    "    figsize=(6, 5), groupby='method'\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6fc4336d",
   "metadata": {},
   "source": [
    "It looks like we get good predictions for HIF1A, RELA and SOX2 among others. On the other side, we worse predictions for STAT1."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "84eba649-b0c4-4183-9fc7-ce72d83c1f93",
   "metadata": {},
   "source": [
    "HIF1A、RELA、SOX2などについては良い予測が得られたようだ。一方、STAT1の予測は悪かった。"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3b3e1bbe",
   "metadata": {},
   "source": [
    "### By meta-data groups\n",
    "\n",
    "The benchmarking pipeline can also be run by grouping the input experiments using the `groupby` argument. Multiple groups can be selected at the same time but for simplicity, we will just test which knockout method seems to perform the best inhibitions:"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1bc2737b-e5d0-4597-b07e-2a6aed93511a",
   "metadata": {},
   "source": [
    "ベンチマークパイプラインは、groupby引数を用いて入力実験をグループ化して実行することもできる。複数のグループを同時に選択することもできるが、ここでは簡単のため、どのノックアウト法が最も抑制効果が高いかをテストするだけにする："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "6e7b0435-e57b-4452-8e1a-a2d7de2a9920",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Example on how to set up decouple arguments\n",
    "decouple_kws = {\n",
    "    'methods': ['ulm', 'mlm'],\n",
    "    'consensus': True,\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "0ad8a428",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Extracting inputs...\n",
      "Formating net...\n",
      "Removed 109 experiments without sources in net.\n",
      "Running 279 experiments for 155 unique sources.\n",
      "Running methods...\n",
      "52 features of mat are empty, they will be removed.\n",
      "Running ulm on mat with 279 samples and 21933 targets for 766 sources.\n",
      "52 features of mat are empty, they will be removed.\n",
      "Running mlm on mat with 279 samples and 21933 targets for 766 sources.\n",
      "Calculating metrics...\n",
      "Computing metrics for groupby Knock.Method...\n",
      "Done.\n"
     ]
    }
   ],
   "source": [
    "# Run benchmark pipeline\n",
    "df = dc.benchmark(\n",
    "    mat, obs, collectri,\n",
    "    perturb='TF', sign=-1,\n",
    "    groupby='Knock.Method', verbose=True,\n",
    "    decouple_kws=decouple_kws\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "82448007",
   "metadata": {},
   "source": [
    "Again, we can visualize the results in a collection of scatterplots:"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9fef9727-49c6-49f4-b11c-cc1b2d70468a",
   "metadata": {},
   "source": [
    "ここでもまた、散布図の集まりで結果を視覚化することができる："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "80030b93",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 700x200 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "dc.plot_metrics_scatter_cols(\n",
    "    df, col='method',\n",
    "    figsize=(7, 2), groupby='group'\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6f2f04d5",
   "metadata": {},
   "source": [
    "From our limited data-set, we can see that apparently shRNA works better than other knockout protocols."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2a42f467-49f0-48bf-83c1-8a62d9b89a7e",
   "metadata": {},
   "source": [
    "私たちの限られたデータセットから、shRNAは他のノックアウトプロトコールよりもうまく機能することがわかる。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2dc3f7f1-29f5-46e7-8e60-332ec26df7fc",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "py_scvi",
   "language": "python",
   "name": "py_scvi"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.14"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}