protein_analysis.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "86720c9b",
   "metadata": {},
   "source": [
    "1. Dataset Analysis: Explore the statistics and potential issues of the dataset.\n",
    "2. Method Explanation: You should explain the method(s) you used.\n",
    "3. Experiment Description: You should design a set of experiments in order to prove the advantage (if exists) of the proposed method.\n",
    "4. Result Analysis: You should resume your experiment results and analyze them.\n",
    "\n",
    "\n",
    "Multiple users in Kaggle have already conducted comprehensive analyses of the PFAM database. In this study, we leverage some of the existing tools to hilight certain aspects which have not be fully discussed yet. Bileschi et al. in their paper titled \"Using Deep Learning to Annotate the Protein Universe\" they present a remarkable convolutional neural network (CNN) surpasses the accuracy of both __BLAST__, which operates through pairwise alignment, and __hidden Markov models__ (pHMMs) based on scoring functions. This demonstrates the superior performance of the CNN in classifying the protein universe. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "c91d1fe2",
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "\n",
    "import numpy \n",
    "import pandas\n",
    "import tensorflow\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib.cm as cm\n",
    "from matplotlib.ticker import ScalarFormatter\n",
    "from mpl_toolkits.axes_grid1.inset_locator import inset_axes"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f4be045c",
   "metadata": {},
   "source": [
    "## Dataset Analysis:\n",
    "Start by loading and exploring the dataset from the Pfam seed random split. Analyze the statistics of the dataset, such as the number of protein samples, the distribution of Pfam families, and any class imbalance issues. We visualize the data to gain insights, such as plotting the distribution of protein sequence lengths, examining the relationship between different features!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "7a50515d",
   "metadata": {},
   "outputs": [],
   "source": [
    "split_dataset = '/home/or22568/Desktop/PFAM/random_split' \n",
    "train_dataset_folder = f'{split_dataset}/train/'\n",
    "test_dataset_folder = f'{split_dataset}/test/'\n",
    "dev_dataset_folder = f'{split_dataset}/dev/'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "2e1d7fbe",
   "metadata": {},
   "outputs": [],
   "source": [
    "# We use this help function to read the datasets\n",
    "def read_data_csv(data_set_folder: str) -> pandas.DataFrame:\n",
    "    files = []\n",
    "    for file_name in os.scandir(data_set_folder): # scandir to get all files in the folder\n",
    "        files.append(pandas.read_csv(file_name))  # array of pandas.DataFrame\n",
    "    csv_pandas = pandas.concat(files)             # Concatenate the pandas.DataFrames\n",
    "    return csv_pandas "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "9986371d",
   "metadata": {},
   "outputs": [],
   "source": [
    "# We read the train, test and dev datasets present in the three folders \n",
    "train_dataset = read_data_csv(train_dataset_folder)\n",
    "test_dataset = read_data_csv(test_dataset_folder)\n",
    "dev_dataset = read_data_csv(dev_dataset_folder)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "666bd526",
   "metadata": {},
   "source": [
    "Data Structure (see https://www.kaggle.com/datasets/googleai/pfam-seed-random-split?resource=download): \n",
    "\n",
    "Each fold (train, dev, test) has a number of files in it. Each of those files\n",
    "contains csv on each line, which has the following fields:\n",
    "\n",
    "**sequence**: These are usually the input features to your model. Amino acid sequence for this domain.\n",
    "There are 20 very common amino acids (frequency > 1,000,000), and 4 amino acids that are quite\n",
    "uncommon: X, U, B, Z. This can be the training feature! \n",
    "\n",
    "**family_accession**: These are usually the labels for your model. Accession number in form PFxxxxx.y\n",
    "(Pfam), where xxxxx is the family accession, and y is the version number. This is usually the Label!\n",
    "\n",
    "**family_id**: One word name for family.\n",
    "\n",
    "**sequence_name**: Sequence name, in the form \"uniprot accession id/start index // end_index\".\n",
    "\n",
    "**aligned_sequence**: Contains a single sequence from the multiple sequence alignment (with the rest of the members of\n",
    "the family in seed, with gaps retained. This can be the training feature! "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "3b962b8c",
   "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>family_id</th>\n",
       "      <th>sequence_name</th>\n",
       "      <th>family_accession</th>\n",
       "      <th>aligned_sequence</th>\n",
       "      <th>sequence</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>GNAT_acetyltran</td>\n",
       "      <td>R6RQF6_9CLOT/17-251</td>\n",
       "      <td>PF12746.7</td>\n",
       "      <td>AFLFSGR..REVMAD....ACLQGMM..GCVYG..........TAG...</td>\n",
       "      <td>AFLFSGRREVMADACLQGMMGCVYGTAGGMDSAAAVLGDFCFLAGK...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>MoaC</td>\n",
       "      <td>W5NKR5_LEPOC/505-640</td>\n",
       "      <td>PF01967.21</td>\n",
       "      <td>MVDVGGK.PVSRRTAAASATVLLG.EK..........AFWLV.......</td>\n",
       "      <td>MVDVGGKPVSRRTAAASATVLLGEKAFWLVKENQLAKGDALAVAQI...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Methyltransf_25</td>\n",
       "      <td>C0QLU8_DESAH/50-147</td>\n",
       "      <td>PF13649.6</td>\n",
       "      <td>VLDVACGT.C...D..VA...ME..AR.NQ.......T....G......</td>\n",
       "      <td>VLDVACGTCDVAMEARNQTGDAAFIIGTDFSPGMLTLGLQKLKKNR...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>EMG1</td>\n",
       "      <td>T1G7Q2_HELRO/22-222</td>\n",
       "      <td>PF03587.14</td>\n",
       "      <td>VVLERASLESVKV..G.................KEYQLLN....CD...</td>\n",
       "      <td>VVLERASLESVKVGKEYQLLNCDRHKGIAKKFKRDISTCRPDITHQ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Glyco_hydro_30C</td>\n",
       "      <td>C6VRM9_DYAFD/453-540</td>\n",
       "      <td>PF17189.4</td>\n",
       "      <td>GAVRVDVSGGLGTD...............AMVVSSYLN..TDKSLV...</td>\n",
       "      <td>GAVRVDVSGGLGTDAMVVSSYLNTDKSLVTVIVNADNQDRDISLAI...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         family_id         sequence_name family_accession  \\\n",
       "0  GNAT_acetyltran   R6RQF6_9CLOT/17-251        PF12746.7   \n",
       "1             MoaC  W5NKR5_LEPOC/505-640       PF01967.21   \n",
       "2  Methyltransf_25   C0QLU8_DESAH/50-147        PF13649.6   \n",
       "3             EMG1   T1G7Q2_HELRO/22-222       PF03587.14   \n",
       "4  Glyco_hydro_30C  C6VRM9_DYAFD/453-540        PF17189.4   \n",
       "\n",
       "                                    aligned_sequence  \\\n",
       "0  AFLFSGR..REVMAD....ACLQGMM..GCVYG..........TAG...   \n",
       "1  MVDVGGK.PVSRRTAAASATVLLG.EK..........AFWLV.......   \n",
       "2  VLDVACGT.C...D..VA...ME..AR.NQ.......T....G......   \n",
       "3  VVLERASLESVKV..G.................KEYQLLN....CD...   \n",
       "4  GAVRVDVSGGLGTD...............AMVVSSYLN..TDKSLV...   \n",
       "\n",
       "                                            sequence  \n",
       "0  AFLFSGRREVMADACLQGMMGCVYGTAGGMDSAAAVLGDFCFLAGK...  \n",
       "1  MVDVGGKPVSRRTAAASATVLLGEKAFWLVKENQLAKGDALAVAQI...  \n",
       "2  VLDVACGTCDVAMEARNQTGDAAFIIGTDFSPGMLTLGLQKLKKNR...  \n",
       "3  VVLERASLESVKVGKEYQLLNCDRHKGIAKKFKRDISTCRPDITHQ...  \n",
       "4  GAVRVDVSGGLGTDAMVVSSYLNTDKSLVTVIVNADNQDRDISLAI...  "
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# We visually inspect the pandas.DataFrame initial fields\n",
    "test_dataset.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "15398597",
   "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>family_id</th>\n",
       "      <th>sequence_name</th>\n",
       "      <th>family_accession</th>\n",
       "      <th>aligned_sequence</th>\n",
       "      <th>sequence</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>12751</th>\n",
       "      <td>C1-set</td>\n",
       "      <td>IGHG2_CAVPO/229-315</td>\n",
       "      <td>PF07654.15</td>\n",
       "      <td>VYTLPPSRDE.LS.KSKVSVTCLIINFFP..ADIHVEWASNRVPVS...</td>\n",
       "      <td>VYTLPPSRDELSKSKVSVTCLIINFFPADIHVEWASNRVPVSEKEY...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12752</th>\n",
       "      <td>Cofac_haem_bdg</td>\n",
       "      <td>F0LPX5_VIBFN/30-233</td>\n",
       "      <td>PF04187.13</td>\n",
       "      <td>FRHAIRHADVILVGEWH..THTGIHRFQTDLLQTL..............</td>\n",
       "      <td>FRHAIRHADVILVGEWHTHTGIHRFQTDLLQTLSQEERPVALSMEQ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12753</th>\n",
       "      <td>S-methyl_trans</td>\n",
       "      <td>A1B3D4_PARDP/21-337</td>\n",
       "      <td>PF02574.16</td>\n",
       "      <td>LI.....................LDGAMGTQI.Q....QL.........</td>\n",
       "      <td>LILDGAMGTQIQQLGLSETDFAGHGTGCACGCHPPAPGEHPQQGNN...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12754</th>\n",
       "      <td>Glyco_trans_1_4</td>\n",
       "      <td>B8FJ84_DESAA/245-385</td>\n",
       "      <td>PF13692.6</td>\n",
       "      <td>KNRLLVTN...SA.DT.....PLKGL.YHL..LKAV.HEVR....K...</td>\n",
       "      <td>KNRLLVTNSADTPLKGLYHLLKAVHEVRKKRDVTLTVIGAPKKHGG...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12755</th>\n",
       "      <td>OB_MalK</td>\n",
       "      <td>A0A0A0D4M6_9PROT/236-288</td>\n",
       "      <td>PF17912.1</td>\n",
       "      <td>GSPAMNFLPGT..V.V..RSDG.......LVARIDGQDL.PL...G...</td>\n",
       "      <td>GSPAMNFLPGTVVRSDGLVARIDGQDLPLGAYPFLSEPAAGDEIVL...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             family_id             sequence_name family_accession  \\\n",
       "12751           C1-set       IGHG2_CAVPO/229-315       PF07654.15   \n",
       "12752   Cofac_haem_bdg       F0LPX5_VIBFN/30-233       PF04187.13   \n",
       "12753   S-methyl_trans       A1B3D4_PARDP/21-337       PF02574.16   \n",
       "12754  Glyco_trans_1_4      B8FJ84_DESAA/245-385        PF13692.6   \n",
       "12755          OB_MalK  A0A0A0D4M6_9PROT/236-288        PF17912.1   \n",
       "\n",
       "                                        aligned_sequence  \\\n",
       "12751  VYTLPPSRDE.LS.KSKVSVTCLIINFFP..ADIHVEWASNRVPVS...   \n",
       "12752  FRHAIRHADVILVGEWH..THTGIHRFQTDLLQTL..............   \n",
       "12753  LI.....................LDGAMGTQI.Q....QL.........   \n",
       "12754  KNRLLVTN...SA.DT.....PLKGL.YHL..LKAV.HEVR....K...   \n",
       "12755  GSPAMNFLPGT..V.V..RSDG.......LVARIDGQDL.PL...G...   \n",
       "\n",
       "                                                sequence  \n",
       "12751  VYTLPPSRDELSKSKVSVTCLIINFFPADIHVEWASNRVPVSEKEY...  \n",
       "12752  FRHAIRHADVILVGEWHTHTGIHRFQTDLLQTLSQEERPVALSMEQ...  \n",
       "12753  LILDGAMGTQIQQLGLSETDFAGHGTGCACGCHPPAPGEHPQQGNN...  \n",
       "12754  KNRLLVTNSADTPLKGLYHLLKAVHEVRKKRDVTLTVIGAPKKHGG...  \n",
       "12755  GSPAMNFLPGTVVRSDGLVARIDGQDLPLGAYPFLSEPAAGDEIVL...  "
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# We visually inspect the pandas.DataFrame final fields\n",
    "test_dataset.tail()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0b12e715",
   "metadata": {},
   "source": [
    "One crucial aspect to examine first is whether all the protein families are reported in the three subfolders and the number of proteins present in each subfolder."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "fe4c7b68",
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_datasets_information(all_dataset: dict) -> pandas.DataFrame:\n",
    "    columns = ['dataset', 'num_samples', 'num_families', 'min_samples_per_family', 'min_family', 'max_samples_per_family', 'max_family', 'mean_samples_per_family', 'name']\n",
    "    data_analysis_info = pandas.DataFrame(columns=columns)\n",
    "    \n",
    "    for dataset_name, dataset in all_dataset.items():\n",
    "        num_samples = len(dataset) # Total number of points\n",
    "        num_families = dataset['family_accession'].nunique() # Number of nonunique family_accession\n",
    "        fam_sizes = dataset.groupby('family_accession').size() # Number of protein for each family\n",
    "        max_family = fam_sizes.idxmax() # Get the family corresponding to the maximum samples  \n",
    "        min_family = fam_sizes.idxmin() # Get the family corresponding to the minimum samples  \n",
    "        min_samples_per_family = fam_sizes.min()\n",
    "        max_samples_per_family = fam_sizes.max()\n",
    "        mean_samples_per_family = fam_sizes.mean()\n",
    "\n",
    "        data_analysis_info = pandas.concat([data_analysis_info, pandas.DataFrame({'dataset': dataset_name,\n",
    "                                                                                  'num_samples': num_samples,\n",
    "                                                                                  'num_families': num_families,\n",
    "                                                                                  'min_samples_per_family': min_samples_per_family,\n",
    "                                                                                  'max_samples_per_family': max_samples_per_family,\n",
    "                                                                                  'mean_samples_per_family': mean_samples_per_family,\n",
    "                                                                                  'name': dataset_name,\n",
    "                                                                                  'min_family': min_family,\n",
    "                                                                                  'max_family': max_family}, index=[0])],\n",
    "                                           ignore_index=True)\n",
    "    return data_analysis_info\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "692aeb75",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>dataset</th>\n",
       "      <th>num_samples</th>\n",
       "      <th>num_families</th>\n",
       "      <th>min_samples_per_family</th>\n",
       "      <th>min_family</th>\n",
       "      <th>max_samples_per_family</th>\n",
       "      <th>max_family</th>\n",
       "      <th>mean_samples_per_family</th>\n",
       "      <th>name</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>train</td>\n",
       "      <td>1086741</td>\n",
       "      <td>17929</td>\n",
       "      <td>1</td>\n",
       "      <td>PF00541.17</td>\n",
       "      <td>3637</td>\n",
       "      <td>PF13649.6</td>\n",
       "      <td>60.613587</td>\n",
       "      <td>train</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>test</td>\n",
       "      <td>126171</td>\n",
       "      <td>13071</td>\n",
       "      <td>1</td>\n",
       "      <td>PF00058.17</td>\n",
       "      <td>454</td>\n",
       "      <td>PF13649.6</td>\n",
       "      <td>9.652743</td>\n",
       "      <td>test</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>dev</td>\n",
       "      <td>126171</td>\n",
       "      <td>13071</td>\n",
       "      <td>1</td>\n",
       "      <td>PF00058.17</td>\n",
       "      <td>454</td>\n",
       "      <td>PF13649.6</td>\n",
       "      <td>9.652743</td>\n",
       "      <td>dev</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  dataset num_samples num_families min_samples_per_family  min_family  \\\n",
       "0   train     1086741        17929                      1  PF00541.17   \n",
       "1    test      126171        13071                      1  PF00058.17   \n",
       "2     dev      126171        13071                      1  PF00058.17   \n",
       "\n",
       "  max_samples_per_family max_family  mean_samples_per_family   name  \n",
       "0                   3637  PF13649.6                60.613587  train  \n",
       "1                    454  PF13649.6                 9.652743   test  \n",
       "2                    454  PF13649.6                 9.652743    dev  "
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "all_dataset = {'train': train_dataset, 'test': test_dataset, 'dev': dev_dataset}\n",
    "get_datasets_information(all_dataset)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a751090d",
   "metadata": {},
   "source": [
    "Interestingly enough the protein family with less instances (1) in the \"test\" and \"dev\" databases is not the same of the train databse! In the following cell we notice that there are 17 instances of PF00058.17 in the train set.\n",
    "\n",
    "PF13649.6 is instead the family more rapresented in all datasets. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "16663ae1",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of instances of protein with family_accession PF00058.17: 17\n"
     ]
    }
   ],
   "source": [
    "def get_protein_count(dataset: pandas.DataFrame, family_accession: str) -> int:\n",
    "    count = len(dataset[dataset['family_accession'] == family_accession])\n",
    "    return count\n",
    "\n",
    "family_accession = 'PF00058.17'\n",
    "count = get_protein_count(train_dataset, family_accession)\n",
    "print(f\"The number of instances of protein with family_accession {family_accession}: {count}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "18faef4b",
   "metadata": {},
   "source": [
    "Upon initial analysis, it is evident that certain protein families are exclusively present in the training data, as indicated by the considerably higher number of families in the train partition compared to the other partitions. However, it remains uncertain whether there are families exclusive to the dev or test data. To gain further insights, we will delve deeper into the dataset and investigate if any families are exclusive to these partitions as well."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "37fa5c81",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "dev and test familes are the same\n"
     ]
    }
   ],
   "source": [
    "def intersection_common_families(dataset: pandas.DataFrame, dataset2: pandas.DataFrame) -> pandas.DataFrame: \n",
    "    return dataset.intersection(dataset2)\n",
    "\n",
    "\n",
    "train_families = set(train_dataset['family_accession'])\n",
    "dev_families = set(dev_dataset['family_accession'])\n",
    "test_families = set(test_dataset['family_accession'])\n",
    "\n",
    "if dev_families == test_families: \n",
    "    print(f'dev and test familes are the same')\n",
    "    common_families = intersection_common_families(train_families, dev_families)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2c3a143f",
   "metadata": {},
   "source": [
    "Based on the analysis, it is observed that there are exclusively unique protein families present only in the training dataset. To ensure a consistent validation process, it is advisable to exclude these families from the analysis. By doing so, the number of families will be consistent across all datasets (train, dev, and test). This approach ensures that the mapping of families can be effectively validated using the dev and test sets without any discrepancies."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "848d4e33",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>dataset</th>\n",
       "      <th>num_samples</th>\n",
       "      <th>num_families</th>\n",
       "      <th>min_samples_per_family</th>\n",
       "      <th>min_family</th>\n",
       "      <th>max_samples_per_family</th>\n",
       "      <th>max_family</th>\n",
       "      <th>mean_samples_per_family</th>\n",
       "      <th>name</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>train</td>\n",
       "      <td>1064950</td>\n",
       "      <td>13071</td>\n",
       "      <td>8</td>\n",
       "      <td>PF00245.20</td>\n",
       "      <td>3637</td>\n",
       "      <td>PF13649.6</td>\n",
       "      <td>81.474256</td>\n",
       "      <td>train</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>test</td>\n",
       "      <td>126171</td>\n",
       "      <td>13071</td>\n",
       "      <td>1</td>\n",
       "      <td>PF00058.17</td>\n",
       "      <td>454</td>\n",
       "      <td>PF13649.6</td>\n",
       "      <td>9.652743</td>\n",
       "      <td>test</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>dev</td>\n",
       "      <td>126171</td>\n",
       "      <td>13071</td>\n",
       "      <td>1</td>\n",
       "      <td>PF00058.17</td>\n",
       "      <td>454</td>\n",
       "      <td>PF13649.6</td>\n",
       "      <td>9.652743</td>\n",
       "      <td>dev</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  dataset num_samples num_families min_samples_per_family  min_family  \\\n",
       "0   train     1064950        13071                      8  PF00245.20   \n",
       "1    test      126171        13071                      1  PF00058.17   \n",
       "2     dev      126171        13071                      1  PF00058.17   \n",
       "\n",
       "  max_samples_per_family max_family  mean_samples_per_family   name  \n",
       "0                   3637  PF13649.6                81.474256  train  \n",
       "1                    454  PF13649.6                 9.652743   test  \n",
       "2                    454  PF13649.6                 9.652743    dev  "
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_dataset = train_dataset[train_dataset['family_accession'].isin(common_families)]\n",
    "all_dataset['train'] = train_dataset\n",
    "get_datasets_information(all_dataset)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "07ca1443",
   "metadata": {},
   "source": [
    "### Protein Family Analysis\n",
    "\n",
    "In the paper of Bileschi et al. they noticed that some of the protein were much less represented than the others. We explore here some of the underlined statistics "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "398f601b",
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_protein_families(all_dataset: dict) -> None:\n",
    "\n",
    "    plt.figure(figsize=(30, 10))\n",
    "    plt.suptitle('Distribution of family sizes', fontsize=28, y=0.95)\n",
    "    # Define the colormap\n",
    "    cmap = cm.viridis\n",
    "\n",
    "    for n, (name, df) in enumerate(all_dataset.items()):\n",
    "        ax = plt.subplot(1, 3, n + 1)\n",
    "        ax.set_title(name, fontsize=26)\n",
    "        ax.set_xlabel(\"Family size\", fontsize=24)\n",
    "        ax.set_ylabel(\"Number of families\", fontsize=24)\n",
    "        plt.tick_params(axis='both', which='both', length=6, labelsize=22)  \n",
    "        df.groupby('family_id').size().hist(bins=200, ax=ax, color=cmap(n / len(all_dataset)))\n",
    "\n",
    "        # Create inset plot for magnified view\n",
    "        ax_inset = inset_axes(ax, width=\"70%\", height=\"70%\", loc='upper right')\n",
    "        plt.tick_params(axis='both', which='both', length=6, labelsize=16)  # Increase tick size and font size\n",
    "        ax_inset.set_xlabel(\"Family size\", fontsize=20)\n",
    "        ax_inset.set_ylabel(\"Number of families\", fontsize=20)\n",
    "        ax_inset.set_ylim(0, 10)\n",
    "        df.groupby('family_id').size().hist(bins=100, ax=ax_inset, color=cmap(n / len(all_dataset)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "13f33487",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAACaoAAAOsCAYAAAChihOHAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAAEAAElEQVR4nOzdd3gU5fr/8c+mkAAJhBpKgAACHhARVARBA9KRZkE9cAQNBlQQESsepVkoiseGdKWoiIqAIEW6dIHQO5LQIQmQgiFhk+zvD3+ZbzbZJBvYzWaX9+u6cjnPzDPP3Ds77u7N3vuMyWKxWAQAAAAAAAAAAAAAAAAAgJN4uToAAAAAAAAAAAAAAAAAAIBno1ANAAAAAAAAAAAAAAAAAOBUFKoBAAAAAAAAAAAAAAAAAJyKQjUAAAAAAAAAAAAAAAAAgFNRqAYAAAAAAAAAAAAAAAAAcCoK1QAAAAAAAAAAAAAAAAAATkWhGgAAAAAAAAAAAAAAAADAqShUAwAAAAAAAAAAAAAAAAA4FYVqAAAAAAAAAAAAAAAAAACnolANAAAAAADka+TIkTKZTMbfrWLmzJlWjzs6OjrXvq1atTL6tWrVqtBiLCoKcq7gXJGRkRowYIDuuOMOBQUFycvLy3heevTo4bDjHDt2TK+88oqaNGmismXLytvb2zjOXXfd5bDjuKtnnnnGOB+hoaF59g0NDTX6PvPMM4USX1F0q77XAAAAAABwq/BxdQAAAAAAAMB+0dHRqlmzZp59ihUrJn9/f5UrV06VKlVS3bp11ahRIz344INq0qQJX/4DHuydd97RBx984PTjTJ48WS+99JLS0tKcfiwAAAAAAAB4BgrVAAAAAADwMNevX9f169eVmJioqKgobdmyxdgWEhKivn376qWXXlJwcLALo7TfunXrtG7dOqM9cuRIl8Xi6TjX7m3WrFk2i9S8vb1tLt+odevW6cUXX5TFYrFanzlzmyT5+PDPjgAAAAAAALDGvxgBAAAAAODmsheeZGRk5CggyXTmzBl98MEH+vTTTzVq1CgNHTq0yM+wtm7dOo0aNcpoUzzlPJxr92WxWPTee+8Z7fLly2vSpElq3769SpUq5dBjjR492niNKV68uD7//HM98sgjKleunEOPAwAAAAAAAM/i5eoAAAAAAADAjQsLC1NaWprVX3p6uuLj4xUVFaU1a9bogw8+UJs2baz2+/vvv/Xaa6+pR48eMpvN+R5n5MiRslgsxt+t4plnnrF63KGhoa4OqcjiXLnWvn379Ndffxntjz76SI8//rjDi9QSEhKsZt1744039Nxzz1GkZsPMmTON/x+io6NdHY5buFXfawAAAAAAuFVQqAYAAAAAgIcxmUwqXbq0QkND1bp1a7399ttatWqV9u/frx49elj1/fXXX/Xiiy+6JlAADrN7926rdrt27ZxynD179lgVEDnrOAAAAAAAAPA8FKoBAAAAAHCLaNCggRYsWKCPP/7Y6naf06dP15QpU1wYGYCbFRsba9UODg526+MAAAAAAADA81CoBgAAAADALebVV1/Vq6++arVu5MiRSklJcVFEAG7W1atXrdo+Pj5ufRwAAAAAAAB4Hv4lCQAAAACAW9CYMWO0ZMkSHT58WJJ04cIFzZgxQwMHDnToca5cuaKdO3fq2LFjSkhIUHp6ukqUKKHg4GDVqlVLd955p0qUKOHQYxbE0aNHtWfPHp0/f15Xr15VaGioevXq5fDjpKWlacOGDYqKilJMTIzKli2r2267TQ8++CCFPv9fQkKCNmzYoLNnz+ry5csKDAxUcHCwmjdvrpCQEIcdJzU1VevXr9fJkycVFxenMmXKqF69emrRooWKFSvmsOPY4/Tp09qyZYtiYmKUlJSksmXLKiQkRA888IBKlSpVoLGy3o7TmRxxnISEBO3bt09Hjx7V5cuXdf36dQUFBSk4OFj33XefQ59v6Z/ble7fv19nz56Vl5eX6tWrp4ceekglS5bMdR+z2axNmzZp//79SkxMVJkyZdSwYUPdf//98vLit7+2FPXX+xsVGxurzZs36/z587p8+bJKlSqlSpUqqWXLlqpUqdINjZmWlqa9e/dq//79iouL099//y1/f38FBQWpRo0aatCggapWrergRwIAAAAAQBFgAQAAAAAAbiMqKsoiyfgLCwu74bG++uorq7Huu+++XPuOGDHCqm9+du7caXn44YctPj4+Vvtl//P29rY0bdrUMmnSpDwfp71/UVFRdsW9cOFCy913351j/9KlS1vt/8033+Q5flZhYWE5npe0tDTLe++9ZwkODrYZb7ly5SwffvihxWw253tOCxJLXufym2++yXP7jZ7rG41v+/btlo4dO+Z5rTRu3Njyyy+/2DWexWKx2nfEiBEWi8ViuXbtmuXNN9+0lC1b1uYxSpcubfdzcbN+/PFHS6NGjXJ9vL6+vpbOnTtbdu3alec4NWrUKNBzdqOvF1mvbXv+atSokWOMgwcPWoYPH265++67LV5eXnnuf8cdd1hmzpxpSU9Ptyu+3P4/nzt3ruXOO++0eYzAwEDL2LFjLRkZGVZjmc1my7hx4yzly5e3uV/16tUtCxcutCuuvn375nlOssr6XPbt2zfH9h9++MEqjm+//dauGLLK+jyWK1fOkpKSUuAxbLnZ1/us7HmvWbt27Q29ZuX1OmjLokWLLPfdd5/FZDLlOk6zZs0sK1assPtcJSQkWN544w1LhQoV8o0xJCTE8uKLL1piYmLsHh8AAAAAgKKOn/8BAAAAAHCLevrpp61m84qMjFRycvJNj/vVV1/p3nvv1W+//aa0tLQ8+6anp+vPP//Up59+etPHtdfgwYPVo0cP7dy506nHSUxMVJs2bfTuu+/q4sWLNvtcunRJb7/9tlq2bKmEhASnxlMUvfPOO2ratKmWL1+e57Wya9cuPfroo+revbuuXbtW4OOcOnVK9913n8aNG6fLly/b7JOQkKC3335bjz76aL7X7Y26evWqOnXqpCeeeEJ79uzJtZ/ZbNbSpUt199136/3333dKLIWtefPmGj16tHbu3KmMjIw8++7fv1/PPPOMunXrpsTExBs63uDBg/Xvf/9be/futbk9KSlJb731lsLDw63WtW3bVm+++abi4uJs7nfq1Ck98sgjmjFjxg3FdaMeffRRVa5c2WhPnjy5QPsfPnxY69evN9rPPvus/Pz8bjquov56fyPi4+PVoUMHde/eXdu2bctzFsGtW7eqQ4cOeuGFF/J9/EePHlXDhg01fvx4xcbG5hvHmTNn9NVXX+mvv/4q8GMAAAAAAKCo4t4SAAAAAADcogICAtSwYUPt2rVL0j/FMVu2bFGbNm1ueMy1a9fmuH1ozZo11aJFC4WEhMjf319Xr17V+fPntW/fPu3fv99m0YrJZJK3t7ckKSMjw6pQIHO9LSaTKc/4xo8fry+++EKSFBgYqLZt2+q2226Tt7e3oqOjtWXLFrsfa36ee+45ozAkODhYHTt2VLVq1ZSQkKDNmzdbFcpt27ZNHTp00B9//FHot5901rnOz8svv6zPP//cal1ISIg6dOigypUrKz4+Xhs3btTu3buN7b/++qvat2+vNWvWyNfX167jXL16VV26dNG+ffskSXfeeadatGihChUqKCEhQX/88Yfx/4AkLV68WGPHjtU777xzU48vu9TUVLVp00Z//vmn1fq77rpLLVu2VFBQkM6fP68VK1bozJkzkv55Pt59911dvXpVY8eOzTGmj49PgZ67vJ7PvHh7exv7WiwWq/9nbY2Z3+1s69atq/r16ys0NFSBgYGyWCyKjY3V7t279eeffxqP47ffflOfPn20cOHCAsU7btw44//zihUrqkOHDqpevbpSU1O1ZcsWbdq0yeg7c+ZMPfjgg+rbt68ef/xx4//ZevXqqVWrVqpYsaKuXLmiFStW6NixY8Y5GDhwoFq2bKl69eoVKLYb5evrq4iICI0ePVqStHHjRh04cEANGjSwa/+shW0mk0kDBgy46Zgc9XpfUFlfs+yV/f+P3MTGxiosLEyHDh0y1hUrVkwtWrRQw4YNFRQUpKtXr2rPnj1av369UZw2efJkJScna9asWTbHTU1NVdeuXXXq1CljXUBAgB588EHVrVtXpUuX1vXr13XlyhUdOnRIu3btuuEiTQAAAAAAijSXzeUGAAAAAAAKzJG3/rRYLJb+/ftbjTdx4kSb/ey99Wfr1q2NPiVKlLD8+OOPeR4/JibGMnXqVEvPnj1z7VPQ247mta+3t7dFkmXgwIGWhISEHP1TU1Ot2jd6608/Pz+LJIvJZLKMHDnScv369Rz916xZY6lSpYrV+MOHD891fGfd+jOrmznXBYlv4cKFVn2LFStm+fLLL3PchtFisVhWrFhhqVSpklX/N998M89Yso8tyRIaGmpZt26dzf7ffvutxdfX19inZMmSlsTExAI9/vwMGTLEKq6qVataVq1alaNfenq65X//+1+O2yguXbo0z/Fv5rkriBu9Djt06GCZNm2a5dy5c3n2O3HihKVr165Wx/jhhx/y3Cf7Y/f19bWYTCbLqFGjbN7ecv78+cZ1of9/O8/x48dbpH9uAfvTTz/l2CctLc3y5ptvWh2nd+/eecblyFt/WiwWy9mzZ62ui0GDBuU5ZqZr165ZypQpY+zXtm1bu/bLT1F7vc9NVFSUpXLlylbXx7Zt23L0y8jIsHTo0MHoZzKZLC+88ILl4sWLNsc9ceKE1TmQZJk1a5bNvrNmzbLqFx4ebvM9KNP169ctq1evtjz55JOWHTt23NgDBwAAAACgCOLWnwAAAAAA3MKy3kpOUq63RbSH2WzWH3/8YbSHDRumnj175rlPhQoVFBERoR9//PGGj1sQ6enpevXVV/Xll1+qVKlSObY7ajaz1NRUSdKYMWM0YsQIm7N/tW7dWqtWrVJgYKCxbuzYsbpw4YJDYiiqMjIy9Morr1it+/bbbzVw4ECbs7S1b99eK1eutDpPH3/8saKjo+063vXr11WlShVt3rxZYWFhNvv07t1b7777rtH++++/NX/+fLvGt8dff/1lNXtcmTJltGbNGpuzF3p5eWnIkCE5bi05ePBgu2aEKqqWL1+u5557LsdrTnY1a9bUwoUL1aVLF2NdQW8VaTab9b///U/Dhw+3eXvLRx99VK+++qrRPnXqlN566y35+vrq999/1+OPP55jH29vb40dO1bNmzc31v3yyy8OuV2yvapUqaIePXoY7Tlz5th1/Hnz5unKlStG+/nnn7/pWNzh9V6S4uLi1KFDB50/f17SP7OxffPNN2ratGmOvrNmzdKKFSuM9meffaavvvpKFStWtDl2zZo1tWLFCj344IPGupEjRyo9PT1H31WrVhnLdevW1bRp02y+B2Xy9fXVQw89pB9++EF33313/g8UAAAAAAA3QaEaAAAAAAC3sKCgIKv2zRSqxcXFWX1Bf88999zwWM4SGhqqDz74oFCO1aRJE73++ut59vnXv/6lESNGGO3r16/nKFDyNIsXL1ZUVJTRfvLJJ/MtcLnjjjs0atQoo52enq4vv/zS7mN++eWX+RZIDRo0yKpQ0ZG3gf3yyy+tbnn44Ycfqm7dunnu06dPH3Xr1s1oHz9+XL/99pvDYirKvLy8rP6/2Lp1qy5dumT3/k2bNtXLL7+cZ5/st77MLKC0VcCUVf/+/Y3la9euac+ePXbH5QhZb7WZkJCguXPn5rtP1tt+Vq5cWd27d7/pONzh9T45OVldunTR0aNHjXUfffSRevfunaOvxWLR+PHjjXanTp300ksv5XsMX19fTZ482SiyjYqK0vLly3P0u3jxorHcuHFjeXnxz/IAAAAAgFsTGTEAAAAAALewgIAAq3ZSUtINj1WyZEmr9q5du254LGcJDw+3OcOSMwwaNMiuYoT+/furePHiRvuXX35xZlgut2jRIqv20KFD7dpvwIABVrOqZR8nNyEhIVazUOWmTJkyatiwodE+dOiQXePbI2usQUFBevbZZ+3aL3uho72P2RNkfS4kadu2bXbvm70IzZYaNWooODjYap09M401a9bMqn3w4EG743KEVq1aqUGDBkY7axGaLXv37tXWrVuNdr9+/eTj43PTcRT11/u0tDT17NnT6rp55ZVXrGbSy2r79u1W/88PHjzY7mP961//srpe165dm6NP1vfaPXv2WBWuAgAAAABwK6FQDQAAAACAW1j2wrS8bkWWn1KlSun222832qNHj9a8efOK1BfyrVq1KrRjZb11YV4CAwOt4tq7d6+uXbvmpKhcL+tMZZUqVcp3BqtMJUqUUMeOHY328ePHFRsbm+9+DzzwgM1bitoSGhpqLGe9VeLNuHjxotUMch07drS7WLJly5aqUKGC0d68ebNDYnKlq1evas6cOQoPD9c999yjqlWrKjAwUL6+vvLx8TH+shdCnTlzxu5jPPDAA3b1CwkJMZarVKmimjVrFmgfSYqPj7c7Lkd54YUXjOUdO3Zo586dufbNWsjm7e2tiIgIh8RQ1F/vIyIitHTpUqP973//WxMmTMi1//r1641lk8mkli1bFuh4derUMZZ3796dY3vW17nDhw+rX79+NzWDKQAAAAAA7opCNQAAAAAAbmEJCQlW7bJly97UeK+88oqxnJKSoqeeeko1atTQiy++qB9//FFnz569qfFvVtbCCmeqUqWKVYFRfu666y5jOS0tTceOHXNCVK5nsVisHlvjxo0LtH+TJk2s2ocPH853n6pVq9o9ftbiqKtXr9ofWB6OHDli1b6Zx3zkyBFZLBaHxFXY0tLS9PHHH6tKlSrq06ePvvnmG+3cuVPnzp3T1atXlZaWpvT0dKu/rApSOGjvc571+b6RfSTHXScF0adPH6vZBXObVe3vv//Wd999Z7Q7d+6s6tWrOyyOovp6P2zYMM2cOdNot2nTRjNnzsyzYDXrjHAWi0VBQUFWhZP5/WWdCdPWbWrDw8NVunRpoz1z5kxVrVpV3bp10xdffKFdu3bluOYBAAAAAPBEFKoBAAAAAHALO3funFW7TJkyNzVe//79NWjQIKt1Z86c0aRJk/Tkk08qJCRENWvW1LPPPqtFixbp+vXrN3W8ggoKCiqU42S/pWBB+ztqNq+iJjEx0aoYo1KlSgXaP3t/e85T9sKivGQtZHFUQVj2GG/mMaenp9/U7XldJS0tTb169dLrr79+w/GnpKTY3bdEiRJ29cv6fN/IPpLjrpOCCAwM1NNPP220586dq8TExBz9vvvuO6v19tzatCCK4uv9l19+qbFjxxrtxo0ba8GCBSpWrFie+8XFxVm1sxdN5veX9TrIXgAuSRUqVND8+fOtitVSUlK0ePFiDR48WE2aNFFQUJA6deqkzz77TOfPn7/RUwAAAAAAQJFGoRoAAAAAALew7du3W7X/9a9/3fSYX3zxhZYvX64HH3zQ5gw20dHRmjlzpnr06KGaNWtq2rRpN31Me/n6+hbKcQpSHGWrvytmaSoM2YuUCnqeAgIC8hyvKLoVH3N2n3zyiX766Sej7efnpz59+ui7777T7t27FRsbq+TkZGVkZMhisRh/WbnrTHLOMnDgQGP577//1pw5c3L0mTJlirEcGhpqdetcRylKr/c//fSTXn75ZaNds2ZNLV261Gr2udw48hauud3+tE2bNtq/f78GDBhgM6arV69q+fLlGjJkiKpXr65+/frZnJ0NAAAAAAB3RqEaAAAAAAC3qKtXr2r//v1Gu1ixYmrWrJlDxu7QoYPWr1+vkydP6ptvvlG/fv1Uv379HIUM586dU//+/fXss8865LhFxd9//31T/bMXJ3mK7MUZBT1P2Qv47ClAcbVb8TFndf36dX344YdGu1KlSoqMjNSsWbPUq1cvNWrUSOXLl1fx4sWtXh88tVjTUerXr69WrVoZ7axFadI/RciRkZFGOyIiQl5ezvmn4KLwer9u3To9/fTTRpFY+fLltXz5crtnMMw6o17VqlWtCiYL+hcdHZ3rcUJCQjR58mRdvHhRy5Yt07Bhw9SqVSsVL17cql9aWpq+/vprNW7cWKdPny74CQEAAAAAoIiiUA0AAAAAgFvUnDlzlJaWZrTvueeeHF+W36xq1arpmWee0fTp03XgwAFdvHhRs2fPVps2baz6zZw5U99++61Dj+1KMTExBep/8eJFq7atW7De6C0Hi1LBT6lSpeTt7W20L1y4UKD9s/e/2VvVFobsMd7MY/b29na7QrUNGzZY3Qpx7Nixql+/fr77Zf9/AjllnVVt37592rRpk9GePHmysezr66t+/fo5PR5Xvd7v2bNHPXr0UGpqqqR/Zi387bffVLduXbvHKF++vLF87ty5At1q9kYUL15cHTt21Icffqi1a9cqISFBa9eu1YsvvmhVqHz69GmPK+QGAAAAANzaKFQDAAAAAOAWlJaWpi+++MJqXZ8+fZx+3AoVKujpp5/WqlWrNH36dKtthXkLUGc7e/as4uLi7O6/Z88eY9nHx0d16tTJ0Sf7LSOTk5PtGvvcuXN2x+FsJpPJ6rHt2rWrQPtnnSFKkm6//XaHxOVM9erVs2oX9DFn7V+vXj2bt1csyo4cOWLV7tSpk1377dixwxnheJQePXqoatWqRjuzOC0hIUE//PCDVb/g4OBCj68wXu9PnjypTp06GcWQPj4++vHHH9W0adMCjZO1eNJiseiPP/5wSHz28vX1VatWrTRx4kQdOHBAVapUMbatXr1aUVFRhRoPAAAAAADOQqEaAAAAAAC3oP/+9786dOiQ0a5SpYqeeeaZQo2hX79+uvvuu4121mKtrHx9fa3a6enpTo3LUZYsWWJXv6SkJK1bt85o33nnnTZntss+M9fJkyftGr8gBReFca7vv/9+Y/nChQv6888/7drv2rVrWrFihdG+7bbbVKFCBYfH52jBwcGqVauW0V6xYoUx81N+Nm3aZDU7X9Zz5y7i4+Ot2vbOgvfjjz86IRrP4uPjo4iICKP9888/69KlS5ozZ45VIesLL7zgivCs2Pt6XxCXLl1Shw4ddP78eWPdtGnT1Llz5wKPlX3Wt++///6m47tR1atX17Bhw6zWOeJ8AQAAAABQFFCoBgAAAADALeazzz7TRx99ZLVu1KhR8vPzK/RYateubSxfv37dZp/stzq8cuWKU2NylC+//NKu23NOmzbNqqjk0Ucftdkv++0S169fn+/YKSkp+vrrr/Ptl6kwznX37t2t2p988old+02bNk2JiYlG+5FHHnFoXM6U9TFfuXJF33zzjV37TZgwwartTo85U/ZrKjo6Ot999u3bp0WLFjkpIs/Sv39/o8A0JSVFM2fO1JQpU4zt9erVU+vWrV0VnhV7Xu/tlZycrIcffthqxr4PP/zwhguuW7RooRo1ahjt7777Tvv377+pGG9G1nMl3fz5AgAAAACgqKBQDQAAAACAW8ShQ4fUs2dPDRkyxKqAasCAAXruueduevwLFy7o1KlTdve/fv26tm7darRDQ0Nt9su+fvv27TcSXqHbuXNnjkKj7I4cOaJRo0YZbV9fX/Xr189m38qVK1udi5kzZxq3u8vNK6+8YjXbUH4K41x36dJFNWvWNNrz5s3Tzz//nOc+hw4d0rvvvmu0vb29NXDgQIfH5iwDBw6Ul9f//TPc22+/rWPHjuW5z3fffacFCxYY7Tp16th928yipEGDBlbt/G75eOXKFfXu3dttZk50tcqVK1sVML733ntWBVYDBgxwynGd9Xpvj7S0NPXs2VPbtm0z1g0cODDHLGQF4ePjo7ffftvqGI888kiBb528ceNGmzMm2jtzZKbsM2HezPkCAAAAAKAooVANAAAAAAAPY7FYlJiYqFOnTmn9+vUaO3as2rVrpwYNGuQoCHr00Uf15ZdfOuS4hw8fVq1atfTEE09o4cKFunbtWq59Y2Nj9dRTT1kVOuQ2k1jTpk2tinxee+01bdy4sUjPMJM5O90bb7yh999/X2lpaTn6/PHHH2rbtq3VLGHDhg1TpUqVch23T58+xnJMTIwef/xxXb58OUe/uLg49e3bV5MnT1axYsXsjrswzrWXl5f+97//Wa3r3bu3Jk2aZHMGulWrVqlNmzZW5+m1116zmv2oqKtdu7YGDx5stK9cuaKHHnpIa9euzdE3IyNDX3zxhZ599lmr9Z999plMJpPTY3W0Fi1aqHz58kZ7woQJ+uqrr2w+1zt27NCDDz6offv2qWTJkoUZplvLWrSZtXjV399fffv2dcoxnfV6b4/XXntNS5cuNdqPPfaYPv/88xseL1O/fv3Url07o338+HE1adJE3333XZ6Fk1euXNHXX3+tBx54QA888IDNc/HEE0+oYcOG+uyzz3T69Olcx8rIyNDUqVP18ccfG+uqVaume+655wYfFQAAAAAARYuPqwMAAAAAAAA3bv369fLxsU7vMzIy8r3lZEBAgEaPHq0hQ4Y4tPglPT1dP/30k3766Sf5+/vrzjvvVMOGDVW+fHkVL15cCQkJOnjwoNavX6+UlBRjv+rVq+v111+3OWalSpXUrVs3LVy4UJJ08OBBPfDAAzKZTCpevLhV/AcPHlT16tUd9nhuVLNmzVSxYkX99NNPevfddzVx4kR16tRJVatWVWJiorZs2ZJjtrL77rtP//3vf/Mc9+WXX9aUKVN08eJFSf8UcdWqVUtdunRRrVq1lJKSosOHD2v16tVKTk6Wt7e3vvzyS/Xv39+uuAvrXHfv3l2DBw82ikuuX7+uF198UWPGjFGHDh1UuXJlxcfHa9OmTYqMjLTat2XLlnrvvfdu6LiuNHbsWG3evNmYWenMmTN66KGH1KRJE7Vo0UJBQUG6cOGCli9fnqOQ5Y033nDL2dSkf4o233nnHQ0ZMkTSP69PAwcO1Geffaa2bduqfPnyunLlirZu3Wr1/8Rnn33mkJkebwUPPvig7rjjjhy3qnziiSdUtmxZpx3XGa/39ti9e7dVe+HChQUqyJWkr7/+2qrwV/pnpsZ58+apVatW2rt3ryTp4sWL+s9//qMhQ4YoLCxMoaGhCggIUFJSkmJiYrR7924dOnTIrhkA9+/fryFDhuiVV15R7dq11bhxY4WEhKh06dJKTU3VqVOntG7dOp09e9Zqv88++8yqgBgAAAAAAHdGoRoAAAAAAG6uILfIq1atmvr06aPBgwerYsWKToxKSklJ0Z9//pnvLc/q1KmjZcuWqXTp0rn2+eqrr3TkyBEdOnTIWGexWJScnGzVLyMj4+aCdqDp06frwoUL2rBhgy5cuKBvvvkm175NmzbV8uXL8y22KFu2rH744Qd17dpVV69elfTPDErfffddjr7+/v6aMmWKHnzwwQLFXVjn+rPPPlNAQIDGjBljFFaePn1a06dPz3Wfrl276ocffpCvr+9NHdsV/Pz8tHr1aj3++ONasWKFsT4yMjJHMV4mk8mkUaNGWd321B29/PLLioyM1OzZs411R48e1dGjR3P0NZlM+uCDD9SvXz8K1Qpg4MCBeuGFF6zWPf/884V2fEe+3hfUjdwmNrfXrzJlymjz5s3q37+/vv/+e2N9XFyc5s+fn++4Pj4+8vb2zrOPxWLR8ePHdfz48Tz7FStWTF999ZXVrV0BAAAAAHB3/BQLAAAAAAAP4+Pjo4CAANWoUUPNmjVTnz59NGHCBO3YsUMnT57U+++/75QitXvvvVezZs1Sz5497Rr/tttu07hx47R3717Vrl07z76VK1dWZGSkpk+fru7du6tmzZoKCAgo0rdCLFWqlNasWaMRI0aoQoUKNvuUK1dOH3zwgTZu3KigoCC7xm3VqpU2bdqk1q1b29xuMpnUunVrbdmyJceMQfYozHP9wQcfaNu2berQoUOOmQGzuuuuu/Tzzz/r119/VYkSJRweR2EJCAjQsmXL9MMPP+jOO+/MtZ+Pj486d+6snTt3un2RWqZZs2Zp4sSJud7a1svLS61bt9bq1as1bNiwQo7O/fXq1ctq1q0777xTzZs3d9rxnPl672olS5bUd999py1btqhHjx4qXrx4nv2LFSumsLAwTZgwQWfOnFFgYGCOPnPmzNErr7yiBg0a5PtaGhAQoKeffloHDhxQv379buqxAAAAAABQ1Jgs+d0LBAAAAAAA4AacOHFChw8f1smTJ5WQkKD09HQFBASoatWqatSokerUqePqEAuN2WzWhg0bFBUVpZiYGJUpU0Z16tRRWFhYngVa+Tl58qQ2bNig8+fPy8vLSyEhIWratKlq1qzpwOgLR3x8vDZs2KCzZ8/q8uXLCgwMVHBwsJo3b65q1aq5OjynOHXqlLZs2aKLFy/q6tWrKlu2rEJCQvTggw+qVKlSrg7PKcxms7Zt26a9e/cqPj5eQUFBqly5su677z5VqVLF1eG5raVLl+rhhx822pMmTSrUGdU8+fU+NTVV27Zt04kTJ3Tp0iWlpKQoICBAFSpUUN26ddWgQYN8i9myio+P1/79+3XixAnFxsbq2rVr8vf3V7ly5VS/fn01atRI/v7+TnxEAAAAAAC4DoVqAAAAAAAAAODGunfvrl9//VXSPzNynTt3zubMXgAAAAAAAK7ErT8BAAAAAAAAwE399ddfWrJkidHu27cvRWoAAAAAAKBIolANAAAAAAAAANzUf//7X2VkZEiSvLy89NJLL7k4IgAAAAAAANsoVAMAAAAAAAAAN3P27FkNGTJE8+bNM9b16tVL9erVc2FUAAAAAAAAuaNQDQAANzJy5EiZTCaZTCa1atXK1eEAAAAAAArR888/r4CAAPn7+yskJESfffaZsa106dIaM2aMC6MDAAAACs8zzzxjfF/yzDPPuDocAICdfFwdAAAAAAAAAAAgfykpKfr7779zrPfz89PcuXMVEhLigqgAAAAAAADsw4xqAACPx69qAAAAAHgiT851Zs6caTy20NBQV4dTJPn6+qpGjRp69tlntWvXLnXq1MnVIQEAAAAAAOSJGdUAAAAAAAAAwA3MnDlTM2fOdHUYAAAAAAAAN4QZ1QAAcCMjR46UxWKRxWLRunXrXB0OAAAAAAAAAAAAAAB2oVANAAAAAAAAAAAAAAAAAOBUFKoBAAAAAAAAAAAAAAAAAJyKQjUAgMcymUwymUyaNWuWsW7WrFnG+ux/M2fOtNp/5MiRxrZWrVoZ6/ft26c333xTjRs3VnBwsLy9vWUymRQfH2+1f0ZGhjZs2KBRo0bp4YcfVu3atRUYGChfX19VqFBBjRo10vPPP6/Vq1fb/Zhyiym7Z555xuj3zDPPGOv37t2rwYMHq0GDBipdurRKliypOnXq6LnnntOuXbvsjgMAAACA69xsrpPdyZMnNXbsWD300EOqXr26ihcvrlKlSqlu3brq27evfv311wLFd+rUKb3//vtq06aNqlSpouLFi8vX19cYs2PHjnrnnXe0fv16ZWRkWO3bqlUrmUwmPfvss1bx5fbYsuY7AAAAANxfdHS03n77bTVq1EhBQUEKDAzU7bffrmeeeUZ//PHHTY+/evVqvfTSS2rUqJEqVqyoYsWKqWLFimratKn++9//6tixY3nu36ZNGyMf6datW4GO/fPPPxv7enl56dSpUzfzUADALfm4OgAAANxFenq6hg8frrFjx+b4MiW7HTt2qGvXrrpw4YLN7XFxcYqLi9PevXs1ZcoUtWzZUj/88IOqVq3qjNCVnp6uESNGaMyYMTliP378uI4fP66vv/5aI0aM0IgRI5wSAwAAAICiJTU1VcOGDdPEiRN1/fp1q20pKSlKSkrSsWPHNHv2bN133336/vvvVatWrTzHHDdunEaMGKHU1NQc25KSkowxV6xYoQ8++EATJ07Uiy++6NDHBQAAAMA9TZ48Wa+++qqSk5Ot1h85ckRHjhzRrFmz9Nxzz+mLL74o8NgHDhzQ888/r40bN+bYFhsbq9jYWG3fvl3jx4/Xyy+/rHHjxsnb2ztH3759+2rNmjWSpGXLlik2NlYVKlSwK4bZs2cby61bt1b16tUL/DgAwN1RqAYA8FgdOnSQ9M8MaOfOnZMkValSRQ0bNrTZP78isddee02ffvqpJKlYsWJq0KCBgoKCdOHCBR0+fNiq74ULF6yK1EqUKKE6deooKChIXl5eunjxoo4cOaL09HRJ0saNG9WsWTNFRkbandAUxKBBgzR58mRJUkBAgBo0aKDixYsrKipKJ0+elCRZLBaNHDlSlStXVv/+/R0eAwAAAADHcESuc+XKFXXr1i3HlzS33XabqlatKrPZrMOHD+vy5cuSpG3btql58+Zau3at6tevb/M4Y8aM0dtvv53j2KGhoSpevLiuXr2qc+fOWc0akP2HNE2bNpW/v7/Onj2r/fv3S5L8/f0VFhZm85i5PWYAAAAA7uXTTz/VK6+8YrWuYsWKqlevnq5fv64DBw7o6tWrmj59upKTk+Xr62v32KtXr9ajjz6qxMREY52/v7/q16+voKAgXb58Wfv371daWprS0tI0YcIEHT9+XL/88ou8vKxvUvfYY49p4MCBunr1qtLS0vT999/r5ZdfzjeGuLg4LV++3Gj37dvX7vgBwJNQqAYA8FiZH/ifeeYZ45Y47dq1y/e2N7ZERkZq/fr18vPz03vvvacXXnhBAQEBxvaTJ0+qZMmSVvvUrVtX4eHh6ty5sxo0aJAjmbly5YqmT5+uUaNG6e+//9aZM2c0YMAA/fLLLwWOLy+//fab4uLiVK5cOU2YMEH//ve/VaxYMWP76tWr1atXL8XExEiS3njjDfXu3TvH4wEAAABQNNxsrmOxWNS7d2+jSM3Ly0tDhgzR0KFDrYraMjIy9Ouvv2rgwIE6d+6cYmJi1LNnT+3YsUPFixe3GjM2NlajRo0y2p07d9bHH3+sf/3rXzmOf+XKFa1YsULffvttjjxp/PjxkqSZM2cat/8MDg62+kIHAAAAgGf5888/9eqrrxrtSpUq6auvvlL37t2NnOHatWuaNGmS3n77bX3//fd2/+j/2LFjeuSRR5SUlCRJCgkJ0dixY9WzZ0+r70quXLmisWPH6qOPPpLFYtGiRYv0/vvva/jw4VbjlSxZUo899piRi82ePduuQrXvv/9eZrPZagwAuBV55d8FAAAkJSXJy8tLixYt0uuvv25VpCZJNWrUsPr1zkMPPaTDhw/rzTffVMOGDXN8+SJJZcqU0euvv67ly5cb00cvXLhQR48edWjscXFxCgoK0qZNm9S3b1+rxEuS2rRpo/nz5xvthIQEhxfLAQAAACg6pk6dqmXLlkmSfHx8tHDhQk2YMCHHzGteXl7q0aOHtm7dquDgYEnSwYMHNWnSpBxjrlixwrjdZ82aNbVgwQKbRWrSP7nQU089pSVLlmjAgAGOfGgAAAAA3NCgQYOM2ZbLlCmjdevW6ZFHHrH6bqV48eIaOnSofvjhB5lMJsXGxto1dt++fY0itfr162vXrl3q3bt3ju9KypQpo3Hjxhl3p5GkDz74wOruOZn69OljLEdGRurAgQP5xpH1tp+PP/44kwUAuGVRqAYAgJ2ee+454xY7+SlRooRMJpNdfVu2bKknn3xS0j8zGyxYsOCGY8zN+PHjVa9evTxjaN68udHesGGDw2MAAAAA4Hrp6enGrGWSNHToUHXt2jXPfapVq6aPPvrIaH/++ec5+pw+fdpYbtq0aY4vfXKT+aMdAAAAALem7du3a/v27UZ79OjReX6f0aNHD/Xq1cuusdetW6ctW7ZI+if3+OGHH1S+fPk89+nfv78eeughSdL169etCtcytW7dWtWrVzfaWYvQbDl06JB27txptLMWugHArYZCNQAA7PT88887beysRWJ//vmnQ8cOCAiwK+kJCwszlu359Q8AAAAA97NhwwadOHFC0j8zpr3yyit27ffEE08Yt/s8efKkjh07ZrU9661A9+7dq/T0dAdFDAAAAMCTZb3DS8mSJRUeHp7vPoMHD7Zr7JkzZxrLHTp0UMOGDe3ar2/fvsbyypUrc2w3mUx6+umnjfZ3331nzAhnS+ZtQiWpevXqat26tV1xAIAn8nF1AAAAuINSpUrprrvuuqF9ExMTtXLlSu3evVsnT55UUlKSUlJSZLFYjD5nz541ls+cOXOz4Vq555575Ofnl2+/kJAQYzk+Pt6hMQAAAAAoGtavX28s33HHHapUqZJd+/n5+alevXravXu3JGnHjh2qU6eOsf2+++4zlg8dOqTevXtr3LhxqlGjhmMCBwAAAOCRtm3bZiw/+OCDKlGiRL77NG3aVOXLl1dcXFye/bLmP23btrU7pkaNGhnLWWdCy6pv37764IMPJP3zHc+qVavUvn37HP0yMjL07bffGu2nn37a7jvyAIAnolANAAA71KxZs8CJw6VLl/T2229r9uzZSklJsXs/RxeJ2fvFU8mSJY3l5ORkh8YAAAAAoGjYu3evsXzu3Dl17NjR7n1PnjxpLMfGxlpta968uVq2bKmNGzdKkubNm6cff/xRTZs2VZs2bdSiRQs1b95cZcqUuclHAAAAAMCTHD161Fi2d8azzL5r167NdXtSUpKio6ON9qxZs7RixQq7xr527ZqxnJqaqsTERJUqVcqqT506ddS8eXPj1qKzZ8+2Wai2evVqq8kKuO0ngFsdhWoAANghewKSnxMnTqh169Y6depUgY+Vmppa4H3yYs9satllne0NAAAAgOe4dOmSsRwXF2f3FzXZ2fqBzc8//6zu3bsbMyJYLBZt27bNaHt5ealRo0Z67LHHFB4ersqVK9/QsQEAAAB4jitXrhjL5cqVs3u//PpmzX0kac+ePdqzZ0/Bgvv/4uPjbX5P1KdPH6NQbcGCBbp69aoCAgKs+syePdtYbt68uerWrXtDMQCAp/BydQAAALgDLy/73zIzMjL0xBNPGEVqJpNJ3bt318yZM7V7927FxcXp2rVrslgsxt8333zjrNABAAAAwPD33387ZJyMjIwc64KDg7V582bNmTNHLVu2zJFHZWRkaNeuXXrnnXdUq1YtffDBB/xIBgAAALjFXb9+3VguVqyY3fvl9yN9R+U+ku38R5KeeuopI47k5GT9/PPPVtuvXr2qBQsWGG1mUwMACtUAAHC4pUuXaufOnUb722+/1cKFC9W3b181atRI5cqVk7+/v9U+SUlJhR0mAAAAgFtQUFCQsdy7d2+rH9AU5G/kyJE2x/fy8tJ//vMfbdiwQZcuXdKvv/6qt956S/fff7+8vb2NfikpKXrnnXf09ttvO/kRAwAAACjKss5UVpDvShITE/PcnjX3kaSVK1fecP4TGhqa6zG6detmtLPOniZJ8+fPNwrm/Pz89OSTT9r9+ADAU1GoBgCAgy1fvtxYfvDBB9WrV6989zl9+rQzQwIAAAAASVKlSpWM5aNHjzr1WEFBQeratavGjBmjTZs2KSYmRp9++qnKlClj9Pn444/JhwAAAIBbWHBwsLEcFRVl934nTpzIc3uFChWsfizjrPynb9++xvK6deuMu+1I1oVr3bp1s8qFAOBWRaEaAMDjZb3dTGHcVubkyZPGctOmTe3aZ+PGjc4KBwAAAICHupFc5/777zeWIyMjFRcX5/C4clO2bFm9/PLL+vHHH411aWlpWrVqVY6+hZ3HAQAAAHCNu+++21jetm2bXfskJCToyJEjefYpVqyY1di///77jQWYjw4dOhjFdhaLRd9++62kfyYoWLdundEva0EbANzKKFQDAHi8gIAAY/natWtOP57ZbC5Q/4MHD2rLli1OigYAAACAp7qRXKdjx44ymUySpPT0dE2aNMkpseWlbdu2Kl26tNG+cOFCjj6FnccBAAAAcI2wsDBj+eDBg9q7d2+++8ybN09paWn59uvcubOxvHTp0nxnYbsRPj4+VnfWyZxFbc6cOcrIyJD0z6xxHTp0cPixAcAdUagGAPB4lStXNpadfWsbSapSpYqx/Mcff+TZNyMjQwMHDnR2SAAAAAA80I3kOjVr1tTjjz9utMeNG6c9e/bcdCwFmfUsNTXV6gc+ZcuWzdEn62OLjY1VfHz8TcUHAAAAoGh64oknVLJkSaP95ptv5tn/6tWrev/99+0a+8UXXzTGNpvNeu655+wqcCuorLOlHTlyRNu2bdOcOXOMdb169ZKPj4/DjwsA7ohCNQCAx8s6tfPevXu1YsUKpx7voYceMpb//PPPXGcoSE5O1n/+8x+rqZ8BAAAAwF43mut8+OGHKlWqlCTp77//Vrt27bR8+fJ894uJidH48eP1n//8J8e2l19+Wa+++qqOHz+e7zjvvPOOkpOTjXbr1q1z9Lnzzjvl6+trtD/66CNuAQoAAAB4oFKlSumll14y2suXL9err76q9PT0HH0TExP1yCOP6PTp03aNXaFCBY0YMcJor127Vp07d7Y5q3N2O3fuVN++ffX999/n27dRo0Zq1KiR0X7llVd0+PBho81tPwHg/1C2CwDweA899JCqVq2qs2fPymKxqGPHjvrXv/6l0NBQFStWzOg3ePBgqyKzG9WzZ0+9/fbbRqL04osv6vfff9cTTzyhkJAQJSUl6c8//9TXX3+t06dPy9fXV3369NGMGTNu+tgAAAAAbh03muvcdtttmjt3rnr06CGz2azY2Fh16tRJzZo1U/fu3XXHHXcoKChI165dU1xcnPbt26dNmzZp06ZNSk9Pt7o1T6bExETNmjVLn3zyiRo3bqxWrVrprrvuUqVKlVSyZEklJSXpwIED+uGHH7Rjxw5jv169eqlu3bo5xitZsqQeffRRzZs3T9I/xXUzZ85UgwYNVKJECatzMHjwYIecTwAAAACuMWLECC1atEiHDh2SJH3yySdau3at+vXrp3/9618ym83avn27pkyZojNnzqh8+fK66667tGrVqnzHfv3117V37159++23kqSVK1caM023bt1a1apVU/HixZWQkKDTp09r165d+v333xUdHS3J9g9rbOnbt6+GDh0qSdqyZYuxPnsRGwDc6ihUAwB4PB8fH82aNUvdu3fX33//LUk6dOiQkfBk6tGjh0OO5+fnpx9//FFt2rQxZglYuHChFi5caDO2SZMmydvbm0I1AAAAAAVyM7lO586dtWbNGvXs2dOYTWDr1q3aunXrTce1a9cu7dq1K99+7dq105QpU3Ld/sknn2j79u06ceKEJOncuXM6d+6cVZ+goKCbihUAAACA6/n7+2vlypUKCwvTX3/9JemfvGLQoEE5+pYsWVJz5841Cs/sMWvWLIWGhuqDDz6QxWJRSkqKvv322wKNkZ/evXvrjTfeyHFrUWZTAwBr3PoTAHBLaNOmjfbv369hw4apefPmKl++vNVtZBytWbNm2rp1q+6///48+2zatEn9+vVzWhwAAAAAPNvN5DotW7bUkSNH9P777yskJCTPvj4+PmrevLnGjx9v89Y3AwYMUP/+/VWzZs18j1u/fn1NmzZNy5cvV0BAQK79qlSpot27d+vTTz9Vu3btVLVqVRUvXjz/BwYAAADA7VStWlU7d+7U888/Lz8/P5t9HnjgAf35559q27Ztgcb28vLSe++9p927d+vJJ5/MN68oW7asevbsqfnz56tXr152HaNixYrq2LGj1TofHx+79weAW4XJYrFYXB0EAACe7NChQ9q8ebNiYmJUvHhxVa5cWU2bNrXrCxwAAAAAKCxHjhzRzp07FRcXp8TERBUvXlzly5dXnTp11LBhQwUGBto1zoULF7Rv3z5FR0fr8uXLMpvNCggIUNWqVdW4cWPddtttTn4kAAAAANxZQkKCVq9erZMnTyo9PV1VqlTRfffdp9q1aztk/NTUVG3dulV//fWXLl26ZOQsISEhuv3223X77bfLy4s5fwDAGShUAwAAAAAAAAAAAAAAAAA4FWXAAAAAAAAAAAAAAAAAAACnolANAAAAAAAAAAAAAAAAAOBUFKoBAAAAAAAAAAAAAAAAAJyKQjUAAAAAKKJ27dql119/XY0bN1aFChXk5+enqlWr6p577tGgQYP0888/Kz09Pc8xzp8/rxEjRqhJkyYqW7asSpQooVq1aqlv375av359gWNatGiRHn/8cdWoUUP+/v6qWLGimjdvrvHjxyshIaFAYzk6NgAAAABFCzkNAAAAgKxMFovF4uog4B4yMjJ07tw5BQYGymQyuTocAAAAwIrFYlFSUpKqVKkiLy/3/k1OYmKiXn75Zc2aNUv5pWxXrlxRUFCQzW0LFixQeHi44uPjc90/IiJCkyZNkre3d57HSUpKUq9evbRkyZJc+4SEhGjevHm6//778xzL0bHZg3wGAAAARZkn5TMSOU1BY7MHOQ0AAACKMntzGgrVYLczZ86oWrVqrg4DAAAAyNPp06cVEhLi6jBu2OXLl9WhQwft2LFDklSxYkU99thjatKkiUqVKqWLFy/qzJkzWrdunbZv367Lly/b/FJn1apV6ty5s8xmsyTp4YcfVrdu3VSyZElFRkZq+vTpSkxMlPTPlydTp07NNaa0tDR17txZK1eulCQFBwcrIiJC9evX1+XLlzV37lxt2rRJkhQUFKSNGzeqQYMGuY7nyNjsRT4DAAAAd+Du+YxETkNOAwAAgFtZfjkNhWqwW0JCgoKCgnT69GmVKlWqUI5pNpv1+++/q3379vL19S2UY6JoeuuttzRp0qQc69955x29/vrrue73448/6vnnn1d6erqqVKmi8uXL69ChQzKbzWrSpIl+++03lShRwiExcr3CXXCtwl1wraKgEhMTVa1aNcXHx6t06dKuDueGdezYUStWrJAkPffcc/rf//6ngIAAm33Pnz+vChUqyMfHx2p9SkqK6tWrp1OnTkmSvvjiCw0aNMiqz9GjRxUWFqYLFy5Ikn7//Xe1a9fO5nEmTZqkF198UZJUv359rVmzRsHBwVZ9XnvtNU2YMEGS1KJFC23cuNHmWI6OzV6uyGckXsvwf3LLaYYNG6aGDRvmeo0UZk6DoovXEuSHawT54Rop+jwln5HIaQoam73IaeBKeX1HM2TIkFyvEfIZSLyOIH9cI7AH10nRZ29O45PrFiCbzKmkS5UqVaiFaiVKlFCpUqV4sbnFValSRV26dFHTpk117733avr06Zo/f778/PxyvR6jo6M1aNAgpaena/z48XrttddkMpl08uRJdejQQZGRkXr//ff15ZdfOiRGrle4C65VuAuuVdwod74FysyZM40vdHr27Klp06bl2b9y5co218+YMcP40qRr1645vjSRpLp162rixIl67LHHJEnDhw+3+cVJenq6Ro8ebbTnzJmT4wsdSRo3bpxWr16t3bt3a9OmTVqxYoU6dOjg1NgKwhX5jMRrGf5PXjlNbtdIYec0KLp4LUF+uEaQH64R9+HO+YxETlPQ2AqCnAaulN93NLauEfIZZOJ1BPnhGoE9uE7cR345Te43BQWAIuSdd97R4sWL9e6776pjx465/gIvq48++kipqalq3769Xn/9deMFsUaNGvr6668lSVOnTtXFixedGjsAAIC9xo8fL0ny9fXVZ599dsPjzJs3z1geOnRorv169Oih0NBQSdLWrVsVHR2do8/69euNWQDCwsLUpEkTm2N5e3tr8ODBRnvu3LlOjw1wJ+Q0AADgVkBOU7DYAHdBPgMAAByFQjUAHslisWjBggWSpH79+uXYfv/99+v222+X2WzWokWLCjs8AACAHDZt2qRDhw5Jkrp06ZLrzAL5SUpK0qZNmyRJgYGBeuCBB3Lt6+XlpY4dOxrtZcuW5eizfPlyY7lz5855HjvrdltjOTo2wJOR0wAAAHdDTlPw2ABPRT4DAAByQ6EaAI906tQpnT9/XpLUokULm30y12/btq3Q4gIAAMjN+vXrjeU2bdrIYrHou+++U9u2bRUcHCx/f39VrVpV3bp107fffqv09HSb4xw8eFAZGRmSpMaNG8vb2zvP4957773G8v79+3Ns37dvn82+tgQHB6tatWqSpJiYGMXGxjo1NsCTkdMAAAB3Q05T8NgAT0U+AwAAcuPj6gAAwBmOHTsmSfLz81OVKlVs9qlVq5ZV39ykpqYqNTXV5raMjAxdvnxZ5cqVU1pampKTk5WYmMh9sVGkmc1mrlW4Ba5V2GKxWJSUlKQqVarIy8uzfnezY8cOYzkkJERt2rTR2rVrrfqcO3dO586d0+LFi/W///1PCxcuNL5EyXTkyBFjuWbNmvkeN2ufrPvezHinT5829q1QoYLTYgM8mStymsxb8aDo4XMR8sM1gvxwjRQNnpzPSOQ0NxIb4KnIZ5AVn0OQH64R2IPrpGhwRE5DoRoAj3TlyhVJUlBQUK7JSZkyZaz65mbMmDEaNWqUYwMEAAA35fTp0woJCXF1GA6V+UtjSXr99dd17Ngx+fn5qU+fPmrZsqV8fHy0Z88eTZ8+XZcvX1ZkZKQeeugh7dixQ6VLlzb2jY+PN5bLly+f73HLlStnc19njOfo2LLL6x+vExMTJf3zDxpmsznfsRwl81iFeUy4h8yZODL/m/0ayZy9IygoSGlpaTbHKFWqlCTp8uXLeV5j77//vt5///2bjhkAADhGVFSUqlatarXOEz4vktMUPLbsyGngLjLzmPT0dJvXCPkMAACe7WZyGgrVAHiklJQUSVKxYsVy7ePn5ydJunbtWp5jDRs2TEOHDrW5LSEhQdWrV1dUVJQCAwNvMFprZrNZa9eu1ZyBi2ROsT39PQrfD6cnO2SczOe3devWVPt7IJ5fz8VzW3QkJSWpZs2aDnvfLUqyFs8fO3ZM5cqV0+rVq9WoUSNjfa9evfTKK6+obdu2OnDggI4fP67//ve/+vLLL40+V69eNZb9/f3zPW7x4sWN5aSkpBzbHTmeo2PLzp4fGPz+++8qUaJEvmM52sqVKwv9mCjazpw5I0k6fvy4mjRpkuMaybz9TXp6upYuXWpzjMxZOS5dupRrH0lq1KiRvv/+e5vb/v77b0VERDgsp8l8z/xi/Sldz7Dc9HgomGJeJr0UVj3H+V8xJsKFUd06+MzoejwHrsX5z19mPrN582aVLFnSaltycrKLonIccpqCx5YdOQ3cRWY+c/ToUePayHqNuGs+I/3f+9kk3z26bspwyJiwXzGLl14wN8px/n9r9aYLo7p18HnOtTj/rsX5t48jchoK1QB4pMx/JLh+/XqufTJ/mZb1Hwts8fPzM4raclO2bFnj1z83y2w2//MPDaleEoVqRUbWXz/ejMznt1y5cnzI8UA8v56L57boyDz/nng7h8xfI2f63//+Z/WFTqZKlSrpu+++U+PGjWWxWDRjxgyNGTPG5j/IFuXz5IzY8vqBQWJioqpVq6b27ds77HObPcxms1auXKl27drx+gEr8+fPlyTddtttkpTjGsn8hx1vb2917tzZ5hiZt6QqV65crn3yk5iYqIiICIflNJnvmRk+fspIp1CtsGV4m2yef0flM8gbnxldj+fAtTj/+cs8Lx06dMjxvps5W5Y7I6e5eeQ0cBeZ+UzdunXVrl27HNeIu+YzUpacppivMkx8R1PYMizeKnE95/knpykcfJ5zLc6/a3H+7eOInIZCNQAeKfO2nvHx8bJYLDb/0SDzF36ZfQEAAFwp65cypUuX1r///e9c+zZq1EjNmjXTli1blJKSok2bNqljx46SpICAAKNffjPHZu9j64uhgIAA43NTSkqK1fgFHc/RsWVnzw8MfH19XfIPDa46LoouLy8vq/9mv0YqVKgg6Z+cxsfHx2ZOk/mPP2XLlr3h64vrEgCAwmfrs6EnvCeT0xQ8tuzIaeAuMvMYb29v47rIeo2QzwAA4NluJqfxckZAAOBqderUkfTPrGnnzp2z2efEiRNWfQEAAFwpa/H8nXfeKR+fvH9XdPfddxvLx48fN5aDgoKM5bi4uHyPe+nSJZv7OmM8R8cGeDJyGgAA4G7IaQoeG+CpyGcAAEBuKFQD4JGqV6+uSpUqSZI2bdpks0/m+vvuu6/Q4gIAAMjN7bffbizbc7uK0qVLG8tZp9SuV6+esRwVFZXvOFn7ZN3XGeM5OjbAk5HTAAAAd0NOU/DYAE9FPgMAAHJDoRoAj2QymfTII49IkmbMmJFj++bNm3X48GH5+vqqW7duhR0eAABADo0aNTKWExIS8u0fHx9vLGf9gqd+/frGLTh27dql9PT0PMfZvn27sXzHHXfk2N6wYUObfW25ePGiTp8+LUmqWLGicasPZ8UGeDJyGgAA4G7IaQoeG+CpyGcAAEBuKFQD4LFef/11FStWTL///rs++ugjWSwWSdLJkycVHh4uSXruueeMX/UAAAC4UqdOnWQymSRJe/fuldlszrP/zp07jeWsv8wPDAxUixYtJElJSUnauHFjrmNkZGRoxYoVVjFk17FjR2N52bJleca0dOlSY7lz5845tjs6NsDTkdMAAAB3Qk5T8NgAT0Y+AwAAbKFQDYBb2LRpk8qXL2/8/fDDD5KkMWPGWK3P/LWbJNWsWVPTpk2Tl5eX3njjDVWrVk1NmjRRnTp1dOTIEd1999366KOPXPWQAAAArISEhCgsLEzSP7e9mTt3bq59d+/era1bt0qy/qIk01NPPWUsT5gwIddxFi5caNyKplmzZgoNDc3RJywszPhH43Xr1ikyMtLmWOnp6fr8889txuCs2AB3kltOM378eD399NOqXLkyOQ0AAHBr5DQFjw1wF3l9R1O5cmUjpyGfAQAA+aFQDYBbMJvNunTpkvGXmpoqSUpOTrZan32q9T59+mjDhg3q0qWLrl27poMHD6pWrVoaOXKkNm7cqJIlS7ri4QAAANj04YcfGstDhw7V3r17c/S5ePGi/vOf/xjtwYMHq3jx4lZ9wsPDVb16dUnS4sWLNXHixBzjHDt2TAMHDjTa7733ns2YvL29NXz4cKPdp08fxcTE5Oj31ltvaffu3ZKkFi1aqEOHDjbHc2RsgDvJK6dJSkoipwEAAB6BnKZgsQHuIr/vaDJzGvIZAACQHx9XBwAA9mjVqpUxLXRB3X///Vq8eLGDIwIAAHC85s2b680339S4ceN06dIlNW3aVH379lXLli3l4+Oj3bt3a/r06bp8+bIk6e6779Y777yTYxx/f3/NmDFDnTt3ltls1qBBg7R8+XJ169ZNJUuWVGRkpKZPn66EhARJUkREhNq2bZtrXBEREVqwYIFWrlypAwcOqFGjRoqIiFD9+vV1+fJlzZ0717jlTenSpTVlypRcx3J0bIC7yC2nMZvNWrp0qTp37ixfX1+b+5LTAAAAd0FOQ04Dz5TXdzT55TTkMwAAICsK1QAAAACgCBk7dqy8vb01btw4paamaurUqZo6dWqOfh06dNDcuXPl7+9vc5y2bdtq3rx5Cg8PV3x8vJYsWaIlS5bk6BcREaFJkyblGZOPj4/mz5+vXr16acmSJbpw4YLNGQFCQkI0b948NWjQIM/xHBkbAAAAgKKFnKZgsQEAAAC3Em79CQAAAABFzAcffKCdO3fqpZde0u23367AwED5+/urevXqeuqpp7R06VItX75cZcqUyXOcRx55RAcPHtS7776ru+66S0FBQfL391fNmjX19NNPa926dZo6daq8vb3zjSkwMFCLFy/WwoUL9eijj6patWry8/NT+fLldd9992ncuHHav3+/7r//frseoyNjAwAAAFC0kNOQ0wAAAAC2MKMaAAAAABRBjRo10ueff37T41SuXFmjR4/W6NGjHRCV1L17d3Xv3t0hYzk6NgAAAABFBzkNAAAAgOyYUQ0AAAAAAAAAAAAAAAAA4FQUqgEAAAAAAAAAAAAAAAAAnIpCNQAAAAAAAAAAAAAAAACAU1GoBgAAAAAAAAAAAAAAAABwKo8rVGvVqpVMJpPdf9HR0fmOef78eY0YMUJNmjRR2bJlVaJECdWqVUt9+/bV+vXrCxzjokWL9Pjjj6tGjRry9/dXxYoV1bx5c40fP14JCQkFGsvRsQEAAAAAAAAAAAAAAACAo/m4OoCibsGCBQoPD1d8fLzV+qioKEVFRWn27NmKiIjQpEmT5O3tnedYSUlJ6tWrl5YsWWK1PjY2VrGxsdq6dau++OILzZs3T/fff3+hxgYAAAAAAAAAAAAAAAAAzuLRhWoLFizIt0/FihVz3bZq1So9+eSTMpvNkqSHH35Y3bp1U8mSJRUZGanp06crMTFR06ZNkyRNnTo117HS0tL02GOPaeXKlZKk4OBgRUREqH79+rp8+bLmzp2rTZs26cyZM3r44Ye1ceNGNWjQoFBiAwAAAAAAAAAAAAAAAABn8uhCtR49etzwvikpKerXr59RCPbFF19o0KBBxvbevXtrwIABCgsL04ULFzRt2jT17NlT7dq1sznetGnTjCK1+vXra82aNQoODja2Dxw4UK+99pomTJig+Ph4DRgwQBs3biyU2AAAAAAAAAAAAAAAAADAmbxcHUBRNWPGDJ06dUqS1LVrV6tCsEx169bVxIkTjfbw4cNtjpWenq7Ro0cb7Tlz5lgVqWUaN26c7rrrLknSpk2btGLFCqfHBgAAAAAAAAAAAAAAAADORqFaLubNm2csDx06NNd+PXr0UGhoqCRp69atio6OztFn/fr1unDhgiQpLCxMTZo0sTmWt7e3Bg8ebLTnzp3r9NgAAAAAAAAAAAAAAAAAwNkoVLMhKSlJmzZtkiQFBgbqgQceyLWvl5eXOnbsaLSXLVuWo8/y5cuN5c6dO+d57KzbbY3l6NgAAAAAAAAAAAAAAAAAwNk8ulCtS5cuqlq1qooVK6YyZcqoQYMGioiI0Nq1a/Pc7+DBg8rIyJAkNW7cWN7e3nn2v/fee43l/fv359i+b98+m31tCQ4OVrVq1SRJMTExio2NdWpsAAAAAAAAAAAAAAAAAOBsHl2o9ttvv+ncuXMym82Kj4/XwYMHNX36dD300ENq06aNzp8/b3O/I0eOGMs1a9bM9zhZ+2Td1xnjOTo2AAAAAAAAAAAAAAAAAHA2H1cH4AxlypRRu3btdM8996hq1ary9vbW2bNntWbNGi1btkwZGRlas2aNmjdvrq1bt6pSpUpW+8fHxxvL5cuXz/d45cqVs7mvM8ZzdGzZpaamKjU11ea2xMRESZLZbJbZbM53LEfIPE5hHQ8oKK5NAABcw9ZnUt6XAQAAAAAAAAAAii6PK1QbM2aM7r77bhUrVizHtqFDhyoyMlKPPfaYoqOjdfLkSYWHh2vp0qVW/a5evWos+/v753vM4sWLG8tJSUk5tjtyPEfHlt2YMWM0atSoPPv8/vvvKlGiRL5jOdLKlSsL9XiAvZKTk10dAgAAtyRbn0l5XwYAAAAAAAAAACi6PK5QrXnz5nlub9KkiVasWKE777xTqampWrZsmbZv3657773XZn+TyeSMMB3CGbENGzZMQ4cOtbktMTFR1apVU/v27VWqVCmHH9sWs9mslStXql27dvL19S2UYwIFkTnTIAAAKFy2PpPyvgwAAAAAAAAAAFB0eVyhmj3q1q2rPn36aNq0aZKkJUuWWBWqBQQEGMvXrl3Ld7ysfQIDA3NsDwgI0JUrVyRJKSkpVuMXdDxHx5adn5+f/Pz88uzj6+tb6EVjvr6+6uzXy2ivzPipUI8P5IYCSgAAXMPWZ1Lel1GUPfDKRF1PtxjtyEmvuDAaAAAAACiY1qtG67opXZL0Z8cPXRwNAAAA3JWXqwNwldatWxvLhw4dstoWFBRkLMfFxeU71qVLl2zu64zxHB0bAAAAAAAAAAAAAAAAADjbLVuoVr58eWM5Pj7ealu9evWM5aioqHzHyton677OGM/RsQEAAAAAAAAAAAAAAACAs92yhWqxsbHGcvaZxurXry8vr39Oza5du5Senp7nWNu3bzeW77jjjhzbGzZsaLOvLRcvXtTp06clSRUrVlSFChWcGhsAAAAAAAAAAAAAAAAAONstW6i2du1aYzn7TGOBgYFq0aKFJCkpKUkbN27MdZyMjAytWLHCaHfq1ClHn44dOxrLy5YtyzOupUuXGsudO3fOsd3RsQEAAAAAAAAAAAAAAACAs92ShWpHjhzRnDlzjHaXLl1y9HnqqaeM5QkTJuQ61sKFC43bazZr1kyhoaE5+oSFhalSpUqSpHXr1ikyMtLmWOnp6fr8889txuCs2AAAAAAAAAAAAAAAAADA2TyqUO3zzz/X5s2b8+yza9cudezYUampqZKk9u3b67777svRLzw8XNWrV5ckLV68WBMnTszR59ixYxo4cKDRfu+992we09vbW8OHDzfaffr0UUxMTI5+b731lnbv3i1JatGihTp06GBzPEfGBgAAAAAAAAAAAAAAAADO5uPqABxpzZo1evnll1W7dm21bdtWd9xxh8qVKydvb2+dO3dOq1ev1tKlS5WRkSFJqlGjhr755hubY/n7+2vGjBnq3LmzzGazBg0apOXLl6tbt24qWbKkIiMjNX36dCUkJEiSIiIi1LZt21xji4iI0IIFC7Ry5UodOHBAjRo1UkREhOrXr6/Lly9r7ty5xm08S5curSlTpuQ6lqNjAwAAAAAAAAAAAAAAAABn8qhCtUx//fWX/vrrrzz7dOjQQV9//bWqVKmSa5+2bdtq3rx5Cg8PV3x8vJYsWaIlS5bk6BcREaFJkybleTwfHx/Nnz9fvXr10pIlS3ThwgWbs5yFhIRo3rx5atCgQZ7jOTI2AAAAAAAAAAAAAAAAAHAmjypUmzBhgrp27aqtW7dqz549io2NVVxcnFJTU1W6dGmFhoaqefPm6tWrl5o1a2bXmI888oiaNWumSZMmafHixYqOjlZKSooqV66sli1bql+/fgoLC7NrrMDAQC1evFiLFi3S7NmztX37dsXExCgwMFC1a9fWo48+qgEDBqh06dKFHhsAAAAAAAAAAAAAAAAAOItHFarVrl1btWvXVr9+/Rw6buXKlTV69GiNHj3aIeN1795d3bt3d8hYjo4NAAAAAAAAAAAAAAAAABzNy9UBAAAAAAAAAAAAAAAAAAA8G4VqAAAAAAAAAAAAAAAAAACnolANAAAAAAAAAAAAAAAAAOBUFKoBAAAAAAAAAAAAAAAAAJyKQjUAAAAAAAAAAAAAAAAAgFNRqAYAAAAAAAAAAAAAAAAAcCoK1QAAAAAAAAAAAAAAAAAATkWhGgAAAAAAAAAAAAAAAADAqShUAwAAAAAAAAAAAAAAAAA4FYVqAAAAAAAAAAAAAAAAAACnolANAAAAAAAAAAAAAAAAAOBUFKoBAAAAAAAAAAAAAAAAAJyKQjUAAAAAAAAAAAAAAAAAgFP5uDoAAADcQTuvnlbtlRk/uSgSAAAAAAAAAAAAAADcDzOqAQAAAAAAAAAAAAAAAACcikI1AAAAAAAAAAAAAAAAAIBTUagGAAAAAAAAAAAAAAAAAHAqCtUAAAAAAAAAAAAAAAAAAE5FoRoAAAAAAAAAAAAAAAAAwKkoVAMAAAAAAAAAAAAAAAAAOBWFagAAAAAAAAAAAAAAAAAAp6JQDQAAAAAAAAAAAAAAAADgVBSqAQAAAAAAAAAAAAAAAACcikI1AAAAAAAAAAAAAAAAAIBTUagGAAAAAAAAAAAAAAAAAHAqCtUAAAAAAAAAAAAAAAAAAE5FoRoAAAAAAAAAAAAAAAAAwKkoVAMAAAAAAAAAAAAAAAAAOBWFagAAAAAAAAAAAAAAAAAAp6JQDQAAAAAAAAAAAAAAAADgVBSqAQAAAAAAAAAAAAAAAACcikI1AAAAAAAAAAAAAAAAAIBTUagGAAAAAAAAAAAAAAAAAHAqCtUAAAAAAAAAAAAAAAAAAE5FoRoAAAAAAAAAAAAAAAAAwKkoVAMAAAAAAAAAAAAAAAAAOBWFagAAAAAAAAAAAAAAAAAAp6JQDQAAAAAAAAAAAAAAAADgVBSqAQAAAAAAAAAAAAAAAACcikI1AAAAAAAAAAAAAAAAAIBTUagGAAAAAAAAAAAAAAAAAHAqCtUAAAAAAAAAAAAAAAAAAE5FoRoAAAAAAAAAAAAAAAAAwKkoVAMAAAAAAAAAAAAAAAAAOBWFagAAAAAAAAAAAAAAAAAAp6JQDQAAAAAAAAAAAAAAAADgVBSqAQAAAAAAAAAAAAAAAACcikI1AAAAAAAAAAAAAAAAAIBTUagGAAAAAAAAAAAAAAAAAHAqCtUAAAAAAAAAAAAAAAAAAE5FoRoAAAAAAAAAAAAAAAAAwKkoVAMAAAAAAAAAAAAAAAAAOBWFagAAAAAAAAAAAAAAAAAAp6JQDQAAAAAAAAAAAAAAAADgVBSqAQAAAAAAAAAAAAAAAACcikI1AAAAAAAAAAAAAAAAAIBTUagGAAAAAAAAAAAAAAAAAHAqCtUAAAAAAAAAAAAAAAAAAE5FoRoAAAAAAAAAAAAAAAAAwKkoVAMAAAAAAAAAAAAAAAAAOBWFagA8XkxMjF577TU1aNBAJUqUkL+/v2rXrq3+/fvr+PHjrg4PAAAAAPJETgMAAADAXZHPAACArHxcHQAAONORI0f04IMPKiYmRr6+vqpVq5Z8fX11/PhxTZs2Td99952WLl2qsLAwV4cKAAAAADmQ0wAAAABwV+QzAAAgO2ZUA+DRBg4cqJiYGLVo0UInTpzQ4cOHtW/fPp05c0bdunVTcnKynn32WVksFleHCgAAAAA5kNMAAAAAcFfkMwAAIDsK1QB4rOTkZK1du1aSNGnSJIWEhBjbypUrp5kzZ8pkMikqKkqHDx92VZgAAAAAYBM5DQAAAAB3RT4DAABsoVANgMe6fv26MjIyJEm1atXKsb1MmTIqW7asJCktLa1QYwMAAACA/JDTAAAAAHBX5DMAAMAWCtUAeKygoCBVq1ZNkrR58+Yc248cOaJLly4pKChIderUKezwAAAAACBP5DQAAAAA3BX5DAAAsMXH1QEAgDO9//776tu3r8LDw/Xpp5+qVatW8vHx0datWzVkyBCZTCaNHz9e/v7+uY6Rmpqq1NRUm9sSExMlSWazWWaz2SExZ47jW5yX6KLsRp/vzP0cdb2gaOH59Vw8t0UHzwGAW4275TSZYxTzNt30WCi4zPOe/fzz/lk4+MzoejwHrsX5zx/nBsCtxt3ymcyxJKmYhfleXCHzvGc//7yHFg4+z7kW59+1OP/2ccT5oQoCgEfr06ePAgIC9N577+nxxx+32nbnnXdq6dKl6tixY55jjBkzRqNGjcqzz++//64SJUrcdLxZhc94zKHjwbGWLl16U/uvXLnSQZGgKOL59Vw8t66XnJzs6hAAoFC5a07z6kOhDhsLBZf9/N9s/oKC4TOj6/EcuBbnP3fkMwBuNe6az0jSy+YmDh0PBZP9/JPTFC4+z7kW59+1OP95c0ROQ6EaAI9msVh04sQJXbp0Sd7e3qpZs6aKFSum48ePa//+/Zo6daqaNm2qsmXL5jrGsGHDNHToUJvbEhMTVa1aNbVv316lSpVySMxms1krV67U1/3my3wtzSFjwvEWxc+6of0yn9927drJ19fXwVHB1Xh+PRfPbdGR+UtZALhVuFtOk/meOWFNtK6nW256PBRMMW+TXn0o1K7zv+F/AwspqlsHnxldj+fAtTj/+SOfAXCrcbd8Rvq/97PPfCN13ZThkDFhv2IWL71sbpLv+V/bdnghRnXr4POca3H+XYvzbx9H5DQUqgHwaM8//7ymTp2q+++/X3/88YdCQ0MlSTExMerXr58WLFigv/76S5GRkfL29rY5hp+fn/z8/PI8jq+vr8PfsMzX0mS+xtSiRdXNPt/OuGZQdPD8ei6eW9fj/AO41bhrTnM93UKhmgvZc/55T3UePjO6Hs+Ba3H+c8d5AXCrcdd8RpKumzJ03ZTu0DFhv/zOP++pzsXnOdfi/LsW5z9vjjg33FwbgMfas2ePpk2bJl9fX/3www9GAiRJFStW1Hfffafy5ctr7969+vHHH10XKAAAAADYQE4DAAAAwF2RzwAAAFsoVAPgsTZt2iSLxaK6deuqWrVqObaXKlVKTZs2lSTt2LGjsMMDAAAAgDyR0wAAAABwV+QzAADAFgrVAHispKSkfPtYLP/cCiUlJcXZ4QAAAABAgZDTAAAAAHBX5DMAAMAWCtUAeKw6depIko4eParTp0/n2J6YmKjt27dLkurWrVuosQEAAABAfshpAAAAALgr8hkAAGALhWoAPFb79u1Vvnx5mc1mPfXUU4qOjja2xcTEqHfv3oqLi5O/v78ef/xx1wUKAAAAADaQ0wAAAABwV+QzAADAFh9XBwAAzhIQEKDZs2fr0Ucf1ebNm3XbbbepVq1a8vX11fHjx3X9+nX5+Pho8uTJqlq1qqvDBQAAAAAr5DQAAAAA3BX5DAAAsIVCNQAerVOnTtqzZ48mTJigNWvW6NSpU7JYLKpcubIefPBBDRkyRE2aNHF1mAAAAABgEzkNAAAAAHdFPgMAALKjUA2Ax6tbt66mTJni6jAAAADy1apVK61fv97u/lFRUQoNDc2zz/nz5zV58mQtXrxY0dHRSklJUaVKlfTAAw8oPDxcYWFhBYpx0aJFmjNnjrZv366LFy+qVKlSql27th555BENGDBApUuXtnssR8cGeCpyGgAA4C7IachpgOzIZwAAQFYUqgEAAACAh1qwYIHCw8MVHx9vtT4qKkpRUVGaPXu2IiIiNGnSJHl7e+c5VlJSknr16qUlS5ZYrY+NjVVsbKy2bt2qL774QvPmzdP9999fqLEBAAAA8EzkNAAAAIBnoVANAAAAAIqgBQsW5NunYsWKuW5btWqVnnzySZnNZknSww8/rG7duqlkyZKKjIzU9OnTlZiYqGnTpkmSpk6dmutYaWlpeuyxx7Ry5UpJUnBwsCIiIlS/fn1dvnxZc+fO1aZNm3TmzBk9/PDD2rhxoxo0aFAosQEAAAAomshpyGkAAACA7ChUAwAAAIAiqEePHje8b0pKivr162d8afLFF19o0KBBxvbevXtrwIABCgsL04ULFzRt2jT17NlT7dq1sznetGnTjC906tevrzVr1ig4ONjYPnDgQL322muaMGGC4uPjNWDAAG3cuLFQYgMAAABQNJHTkNMAAAAA2Xm5OgAAAAAAgGPNmDFDp06dkiR17drV6kuTTHXr1tXEiRON9vDhw22OlZ6ertGjRxvtOXPmWH2hk2ncuHG66667JEmbNm3SihUrnB4bAAAAAM9ETgMAAAB4JgrVAAAAAMDDzJs3z1geOnRorv169Oih0NBQSdLWrVsVHR2do8/69et14cIFSVJYWJiaNGlicyxvb28NHjzYaM+dO9fpsQEAAADwTOQ0AAAAgGeiUA0AAAAAPEhSUpI2bdokSQoMDNQDDzyQa18vLy917NjRaC9btixHn+XLlxvLnTt3zvPYWbfbGsvRsQEAAADwPOQ0AAAAgOeiUA0AAAAAiqAuXbqoatWqKlasmMqUKaMGDRooIiJCa9euzXO/gwcPKiMjQ5LUuHFjeXt759n/3nvvNZb379+fY/u+ffts9rUlODhY1apVkyTFxMQoNjbWqbEBAAAAKLrIachpAAAAgOwoVAMAAACAIui3337TuXPnZDabFR8fr4MHD2r69Ol66KGH1KZNG50/f97mfkeOHDGWa9asme9xsvbJuq8zxnN0bAAAAACKLnIachoAAAAgOx9XBwAAAAAA+D9lypRRu3btdM8996hq1ary9vbW2bNntWbNGi1btkwZGRlas2aNmjdvrq1bt6pSpUpW+8fHxxvL5cuXz/d45cqVs7mvM8ZzdGzZpaamKjU11ea2xMRESZLZbJbZbM53LEfJPFYxb5PN9UDmteCqa4JrEQCAwmfrM6knvSeT09gfW3ZFOqexeOVYB0iuzWm4FgEAcI2byWkoVAMAAACAImLMmDG6++67VaxYsRzbhg4dqsjISD322GOKjo7WyZMnFR4erqVLl1r1u3r1qrHs7++f7zGLFy9uLCclJeXY7sjxHB1bdmPGjNGoUaPy7PP777+rRIkS+Y7laK8+FGrVzv68AStXrnTJcZOTk11yXAAAbmW2PpN6ynsyOU3BYsuuKOc0L5ubGMvkM7DFFTmNp7x2AgDgbm4mp6FQDQAAAACKiObNm+e5vUmTJlqxYoXuvPNOpaamatmyZdq+fbvuvfdem/1NJpPN9UWBM2IbNmyYhg4danNbYmKiqlWrpvbt26tUqVIOP3ZuzGazVq5cqQlronU93WKs3/C/gYUWA4q2zGukXbt28vX1LfTjZ87MAQAACo+tz6Se8p5MTnNzinJO85lvpK6bMiRJa9sOL7Tjo+hzZU7jKa+dAAC4m5vJaShUAwAAAAA3UrduXfXp00fTpk2TJC1ZssTqS52AgABj+dq1a/mOl7VPYGBgju0BAQG6cuWKJCklJcVq/IKO5+jYsvPz85Ofn1+efXx9fV1SDHQ93WJVqOaKGFC0uera5FoEAKDw2Xrfv5Xek8lpclekcxpThq6b0o0YgOxccW1yLQIA4Bo3k9N45d8FAAAAAFCUtG7d2lg+dOiQ1bagoCBjOS4uLt+xLl26ZHNfZ4zn6NgAAAAAuCdyGgAAAODWRKEaAAAAALiZ8uXLG8vx8fFW2+rVq2csR0VF5TtW1j5Z93XGeI6ODQAAAIB7IqcBAAAAbk0UqgEAAACAm4mNjTWWs/8qv379+vLy+ifV27Vrl9LT0/Mca/v27cbyHXfckWN7w4YNbfa15eLFizp9+rQkqWLFiqpQoYJTYwMAAADgnshpAAAAgFsThWoAAAAA4GbWrl1rLGf/VX5gYKBatGghSUpKStLGjRtzHScjI0MrVqww2p06dcrRp2PHjsbysmXL8oxr6dKlxnLnzp1zbHd0bAAAAADcEzkNAAAAcGuiUA0AAAAA3MiRI0c0Z84co92lS5ccfZ566iljecKECbmOtXDhQuNWNM2aNVNoaGiOPmFhYapUqZIkad26dYqMjLQ5Vnp6uj7//HObMTgrNgAAAADuh5wGAAAAuHVRqAYAAAAARcDnn3+uzZs359ln165d6tixo1JTUyVJ7du313333ZejX3h4uKpXry5JWrx4sSZOnJijz7FjxzRw4ECj/d5779k8pre3t4YPH260+/Tpo5iYmBz93nrrLe3evVuS1KJFC3Xo0MHmeI6MDQAAAEDRQU5T8NgAAACAW42PqwMAAAAAAEhr1qzRyy+/rNq1a6tt27a64447VK5cOXl7e+vcuXNavXq1li5dqoyMDElSjRo19M0339gcy9/fXzNmzFDnzp1lNps1aNAgLV++XN26dVPJkiUVGRmp6dOnKyEhQZIUERGhtm3b5hpbRESEFixYoJUrV+rAgQNq1KiRIiIiVL9+fV2+fFlz5841bnlTunRpTZkyJdexHB0bAAAAgKKBnIacBgAAAMgPhWoAAAAAUIT89ddf+uuvv/Ls06FDB3399deqUqVKrn3atm2refPmKTw8XPHx8VqyZImWLFmSo19ERIQmTZqU5/F8fHw0f/589erVS0uWLNGFCxdszggQEhKiefPmqUGDBnmO58jYAAAAABQt5DQFiw0AAAC4lVCoBgAAAABFwIQJE9S1a1dt3bpVe/bsUWxsrOLi4pSamqrSpUsrNDRUzZs3V69evdSsWTO7xnzkkUfUrFkzTZo0SYsXL1Z0dLRSUlJUuXJltWzZUv369VNYWJhdYwUGBmrx4sVatGiRZs+ere3btysmJkaBgYGqXbu2Hn30UQ0YMEClS5cu9NgAAAAAuB45DTkNAAAAkB8K1QA41MWLF7VkyRLFxcWpZs2a6tq1q4oXL+7qsAAAAIq82rVrq3bt2urXr59Dx61cubJGjx6t0aNHO2S87t27q3v37g4Zy9GxAY5ATgMAAHBjyGkA1yOfAQAARR2FagDsdujQIY0YMUImk0lTpkxRUFCQ1fZff/1VvXr10rVr14x11apV06+//qo777yzkKMFAAAAAGvkNAAAAADcFfkMAADwBF6uDgCA+1i4cKF+/vlnXbx4MUcCFBMTo//85z9KTk6WxWIx/k6dOqWuXbvq77//dk3QAAAAAPD/kdMAAAAAcFfkMwAAwBNQqAbAbqtXr5bJZFKXLl1ybPvqq6909epV+fj46JNPPtGePXs0fvx4eXl56cyZM5o2bZoLIgYAAACA/0NOAwAAAMBdkc8AAABPQKEaALudOnVKktSoUaMc23755ReZTCb16dNHQ4YMUcOGDfXaa6+pX79+slgs+vXXXws7XAAAAACwQk4DAAAAwF2RzwAAAE9AoRoAu8XGxkqSKlSoYLU+Li5OBw4ckCT16tXLalu3bt0kydgOAAAAAK5CTgMAAADAXZHPAAAAT0ChGgC7JScnS5JSUlKs1m/cuFEWi0XFihVTixYtrLZVrlxZkhQfH18oMQIAAABAbshpAAAAALgr8hkAAOAJKFQDYLeyZctK+r/ppTOtXr1aknTPPffIz8/PaltaWpokKSAgoBAiBAAAAIDckdMAAAAAcFfkMwAAwBNQqAbAbo0aNZIkff/998a6a9eu6aeffpLJZNJDDz2UY5+TJ09KkoKDgwsnSAAAAADIBTkNAAAAAHdFPgMAADwBhWoA7PbUU0/JYrFo8eLFeuqpp/Tll1+qffv2iomJkclk0r///e8c+2zbtk2SVKtWrcIOFwAAAACskNMAAAAAcFfkMwAAwBNQqAbAbn369FHLli1lsVj0008/6eWXX9bmzZslSc8++6xuv/32HPv88ssvuf6SBwAAAAAKEzkNAAAAAHdFPgMAADwBhWoA7Obl5aVly5Zp6NChCgkJkY+Pj6pVq6Z3331XkyZNytF/8eLFio6OliS1a9eukKMFAAAAAGvkNAAAAADcFfkMAADwBD6uDgCAeylZsqQ+/vhjffzxx/n2bdmypaKioiRJNWrUcHZoAAAAAJAvchoAAAAA7op8BgAAuDsK1QA4TZkyZVSmTBlXhwEAAAAAN4ScBgAAAIC7Ip8BAABFEbf+BAAAAAAAAAAAAAAAAAA4FTOqAbghx48f1+zZs7VlyxZduHBB165d0/Lly3XbbbcZffbv369Tp06pZMmSCgsLc2G0AAAAAGCNnAYAAACAuyKfAQAA7opCNQAFkpGRoTfffFOffvqpMjIyZLFYJEkmk0nXr1+36nv69Gl16dJFPj4+ioqKUtWqVV0RMgAAAAAYyGkAAAAAuCvyGQAA4O649SeAAhkwYIA++eQTpaenq0qVKnr88cdz7dupUyfVqlVL6enp+vnnnwsxSgAAAACwjZwGAAAAgLsinwEAAO6OQjUAdlu3bp1mzJghSXr77bcVHR2tH3/8Mc99evbsKYvForVr1xZGiAAAAACQK3IaAAAAAO6KfAYAAHgCbv0JwG6TJ0+WJHXu3Fnvv/++Xfs0bdpUknTgwAGnxQUAAAAA9iCnAQAAAOCuyGcAAIAnYEY1AHbbsmWLTCaT+vXrZ/c+ISEhkqQLFy44KywAAAAAsAs5DQAAAAB3RT4DAAA8AYVqAOwWExMjSapZs6bd+/j4/DNxo9lsdkpMAAAAAGAvchoAAAAA7op8BgAAeAIK1QDYrXjx4pKk5ORku/c5deqUJKlMmTJOiQkAAAAA7EVOAwAAAMBdkc8AAABPQKEaALtl/kpn165ddu+zZMkSSVL9+vWdEhMAAAAA2IucBgAAAIC7Ip8BAACegEI1AHZr3769LBaLpk6dqoyMjHz779y5U3PmzJHJZFLHjh0LIUIAAAAAyB05DQAAAAB3RT4DAAA8wS1VqNa3b1+ZTCbjb+TIkXbtd/78eY0YMUJNmjRR2bJlVaJECdWqVUt9+/bV+vXrCxzHokWL9Pjjj6tGjRry9/dXxYoV1bx5c40fP14JCQkFGsvRsQF5GTRokIoXL659+/YpIiJCZrM5177z589Xx44ddf36dZUqVUr9+/cvxEgBAAAAICdyGgAAAADuinwGAAB4Ah9XB1BYli9frtmzZxd4vwULFig8PFzx8fFW66OiohQVFaXZs2crIiJCkyZNkre3d55jJSUlqVevXsY0u5liY2MVGxurrVu36osvvtC8efN0//33F2psgD2qVq2qzz//XBEREZo5c6Z+//13de3a1dg+Y8YMJScna9WqVTpx4oQsFotMJpOmTp2q0qVLuzByAAAAACCnAQAAAOC+yGcAAIAnuCUK1RITE41fCpQsWVJ///23XfutWrVKTz75pPGLhIcffljdunVTyZIlFRkZqenTpysxMVHTpk2TJE2dOjXXsdLS0vTYY49p5cqVkqTg4GBFRESofv36unz5subOnatNmzbpzJkzevjhh7Vx40Y1aNCgUGIDCqJfv34ymUwaPHiwzp49qylTpshkMkmSPv30U0mSxWKRJPn5+Wny5Mnq2bOnq8IFAAAAACvkNAAAAADcFfkMAABwd7fErT9ff/11nT59WiEhIRowYIBd+6SkpKhfv35GIdgXX3yhJUuWqH///urdu7cmTJig7du3q1KlSpKkadOmGUVotmTdXr9+fe3Zs0fvvfee/v3vf2vgwIHauHGjXn31VUlSfHx8nnE6OjagoMLDw3X48GENHTpUtWvXlsVisfqrWrWqXnjhBR06dEh9+/Z1dbgAAAAAYIWcBgAAAIC7Ip8BAADuzOML1dauXWvMKjZp0iQFBgbatd+MGTN06tQpSVLXrl01aNCgHH3q1q2riRMnGu3hw4fbHCs9PV2jR4822nPmzFFwcHCOfuPGjdNdd90lSdq0aZNWrFjh9NiAGxUSEqKPP/5YR48eVXx8vI4ePapDhw4pJiZGp0+f1sSJExUaGurqMAEAAADAJnIaAAAAAO6KfAYAALgrjy5US05O1nPPPSeLxaInn3xSXbp0sXvfefPmGctDhw7NtV+PHj2MD3pbt25VdHR0jj7r16/XhQsXJElhYWFq0qSJzbG8vb01ePBgoz137lynxwY4QqlSpXTbbbepXr16Kl++vKvDAQAAAIACIacBAAAA4K7IZwAAgDvx6EK1YcOG6cSJEypTpow+++wzu/dLSkrSpk2bJEmBgYF64IEHcu3r5eWljh07Gu1ly5bl6LN8+XJjuXPnznkeO+t2W2M5OjYAAAAAAAAAAAAAAAAAcDaPLVTbvHmzvvzyS0nShAkTbN5qMzcHDx5URkaGJKlx48by9vbOs/+9995rLO/fvz/H9n379tnsa0twcLCqVasmSYqJiVFsbKxTYwMAAAAAAAAAAAAAAAAAZ/NxdQDOkJKSovDwcGVkZKhNmzZ69tlnC7T/kSNHjOWaNWvm2z9rn6z73sx4p0+fNvatUKGC02IDbAkPD5ckmUwmzZgxI8f6G5F9LAAAAABwFnIaAAAAAO6KfAYAAHgyjyxUGz58uI4cOaLixYtrypQpBd4/Pj7eWLbnXu7lypWzua8zxnN0bNmlpqYqNTXV5rbExERJktlsltlszncsR8g8jtlslm9x3xzr4RwzZ86UyWSSJE2ePNnm+oKwWCwymUxWY3kKrkUAAFzD1mdS3pcBZMqau2T9MuZmcxq+2AEAAADgbOQzAADAk3lcodqOHTv0ySefSJJGjRql2rVrF3iMq1evGsv+/v759i9evLixnJSU5NTxHB1bdmPGjNGoUaPy7PP777+rRIkS+Y7lSCtXrtSA75802kuXLi3U499qypcvbyQ7Wc911vU3whOft+TkZFeHAADALcnWZ1LelwFkql69us3cJbf1AAAAAFBUkM8AAABP5lGFatevX9ezzz6r9PR0NW7cWEOHDr3pMYvyBz5nxDZs2LBcz1tiYqKqVaum9u3bq1SpUg4/ti1ms1krV65Uu3bt9HiF54z1i+JnFcrxb1Vnz54t0PpbWeZMgwAAoHDZ+kzK+zKATNHR0QVaDwAAAABFBfkMAADwZB5VqPb+++9r//798vb21vTp0+Xt7X1D4wQEBBjL165dy7d/1j6BgYE2x7ty5YokKSUlxWr8go7n6Niy8/Pzk5+fX559fH195evrm2cfR/P19ZX5mtmqDRQFXIsAALiGrc+kvC8DAAAAAAAAAAAUXV6uDsBR9uzZo7Fjx0qShg4dqiZNmtzwWEFBQcZyXFxcvv0vXbpkc19njOfo2AAAAAAAAAAAAAAAAADA2TxmRrWZM2fKbDbLy8tLvr6+ev/99232++OPP6yWM/vVq1dPPXv2NJYzRUVF5XvsrH2y7pt1XWafqKgohYaG3vB4jo4NAAAAAAAAAAAAAAAAAJzNYwrVLBaLJCkjI0MffvihXfusXbtWa9eulSR1797dKFSrX7++vLy8lJGRoV27dik9PT3P24hu377dWL7jjjtybG/YsKGWL19u9G3dunWuY128eFGnT5+WJFWsWFEVKlSw2u7o2AAAAAAAAAAAAAAAAADA2TymUM2RAgMD1aJFC23YsEFJSUnauHGjwsLCbPbNyMjQihUrjHanTp1y9OnYsaM++ugjSdKyZcv0xhtv5HrspUuXGsudO3d2emyALbVq1XL4mCaTSX/99ZfDxwUAAACA7MhpAAAAALgr8hkAAODJPKZQ7dNPP9Wnn36ab7+RI0dq1KhRkqQRI0Zo5MiRNvs99dRT2rBhgyRpwoQJuRaDLVy40Li9ZrNmzWze1jMsLEyVKlXShQsXtG7dOkVGRqpJkyY5+qWnp+vzzz+3isHZsQG2REdHO3xMk8nk8DEBAAAAwBZyGgAAAADuinwGAAB4Mo8pVHO08PBwjRs3TqdOndLixYs1ceJEDRw40KrPsWPHrNa99957Nsfy9vbW8OHD9eKLL0qS+vTpozVr1qhixYpW/d566y3t3r1bktSiRQt16NDB6bEBtvTt29fVIQAAAADADSOnAQAAAOCuyGcAAIAno1AtF/7+/poxY4Y6d+4ss9msQYMGafny5erWrZtKliypyMhITZ8+XQkJCZKkiIgItW3bNtfxIiIitGDBAq1cuVIHDhxQo0aNFBERofr16+vy5cuaO3euNm7cKEkqXbq0pkyZUmixAdl98803rg4BAAAAAG4YOQ0AAAAAd0U+AwAAPBmFanlo27at5s2bp/DwcMXHx2vJkiVasmRJjn4RERGaNGlSnmP5+Pho/vz56tWrl5YsWaILFy7YnOUsJCRE8+bNU4MGDQotNgAAAAAAAAAAAAAAAABwJgrV8vHII4+oWbNmmjRpkhYvXqzo6GilpKSocuXKatmypfr166ewsDC7xgoMDNTixYu1aNEizZ49W9u3b1dMTIwCAwNVu3ZtPfrooxowYIBKly5d6LEBAAAAAAAAAAAAAAAAgLPccoVqI0eO1MiRIwu0T+XKlTV69GiNHj3aITF0795d3bt3d8hYjo4NAAAAAAAAAAAAAAAAABztlitUA5C/U6dOGcvVq1e3uf5GZB0LAAAAAJyFnAYAAACAuyKfAQAAnoxCNQA51KxZU5JkMpmUlpaWY/2NyD4WAAAAADgLOQ0AAAAAd0U+AwAAPBmFagBysFgsBVoPAAAAAEUJOQ0AAAAAd0U+AwAAPBmFagBy+Oabbwq0HgAAAACKEnIaAAAAAO6KfAYAAHgyCtUA5NC3b98CrQcAAACAooScBgAAAIC7Ip8BAACezMvVAQAAAAAAAAAAAAAAAAAAPBuFagAAAAAAAAAAAAAAAAAAp6JQDQAAAAAAAAAAAAAAAADgVD6uDgCAe9qzZ482bNigEydOKCkpSenp6Xn2N5lMmjFjRiFFBwAAAAB5I6cBAAAA4K7IZwAAgLuiUA1AgRw6dEj9+vXTtm3b7N7HYrGQBAEAAAAoEshpAAAAALgr8hkAAODuKFQDYLcTJ06oZcuWio+Pl8VikSQFBgYqKChIXl7cSRgAAABA0UZOAwAAAMBdkc8AAABPQKEaALsNHz5cV65ckZeXl1577TW98MILCg0NdXVYAAAAAGAXchoAAAAA7op8BgAAeAIK1QDYbdWqVTKZTBoyZIjGjRvn6nAAAAAAoEDIaQAAAAC4K/IZAADgCZgHFoDdEhMTJUmPPfaYiyMBAAAAgIIjpwEAAADgrshnAOD/sXfn0VGUafvHr8pCgBB2ElCWAIoacCGCL7KIIgiCoICMiggYDYggOriBC6sjIuKMAoJsKuogMyIwIFtYhyCRKMsrIIuSCLIGISRsIUv9/uCXfhPTSTrp6vT2/ZyTc6q7nnrqprohddF3VwHwBTSqAXBYvXr1JElBQVyMEQAAAID3IdMAAAAA8FbkGQAA4AtoVAPgsM6dO0uStm3b5uZKAAAAAKDkyDQAAAAAvBV5BgAA+AIa1QA47MUXX1RYWJgmT56sM2fOuLscAAAAACgRMg0AAAAAb0WeAQAAvoBGNQAOa9Cggb755hudPXtWrVu31tq1a91dEgAAAAA4jEwDAAAAwFuRZwAAgC/gJuYASqRDhw7avn27Wrdurc6dO6tatWq67rrrVLFixSK3MwxD69atK6MqAQAAAMA+Mg0AAAAAb0WeAQAA3o5GNQAl8t133+mJJ57Q6dOnZZqmzpw5o23bthU63jAMmaYpwzDKsEoAAAAAsI9MAwAAAMBbkWcAAIC3o1ENgMP27t2r++67T5cuXZJpmipfvryuv/56Va1aVQEB3EkYAAAAgGcj0wAAAADwVuQZAADgC2hUA+CwcePG6eLFiwoJCdH777+vJ598UuXLl3d3WQ7Jzs7WvHnz9MUXX2jPnj06f/68IiIi1Lx5cz355JN68MEH3V0iAAAAABcj0wAAAADwVuQZAADgC2hUA+CwLVu2yDAMvfbaaxoyZIi7y3HY2bNn1bVrVyUkJMgwDDVp0kSRkZE6duyYli5dqqCgIEIQAAAA4AfINAAAAAC8FXkGAAD4AhrVADjs7NmzkqQuXbq4uRLH5eTkqEePHkpISFCvXr30wQcfqG7durb1v//+uw4dOuTGCgEAAACUFTINAAAAAG9FngEAAL6ARjUADqtbt65++eUXZWdnu7sUh82aNUvx8fG655579O9//1sBAQH51tetWzdfKAIAAADgu8g0AAAAALwVeQYAAPiCgOKHAMBV3bt3lyT997//dXMljvvggw8kSRMmTCgQgAAAAAD4FzINAAAAAG9FngEAAL6AMwIADnv55ZcVHh6uyZMnKzk52d3lFOvgwYPat2+fqlevrtatW2vp0qXq16+f7r33Xj366KOaM2eOMjIy3F0mAAAAgDJCpgEAAADgrcgzAADAF3DrTwAOi4iI0OrVq9WzZ0/9z//8j9566y316dNHVatWdXdpdv3444+SpBtvvFFPPPGEvvzyy3zrFy5cqClTpmjVqlVq0KBBofNkZGQUGpbS0tIkSZmZmcrMzLSk7tx5givwT7QnK+3rnbudVe8XeBZeX9/Fa+s5eA0AOINMU5DVmSZ3jnKBhtNzoeRyj7sjx5/fqdbjnNH9eA3ci+NfPI4NAGeQZwpy5Wc05Uyu9+IOuce9uOPP71TX4HzOvTj+7sXxd4wVx4cuCAAOa9SokSTp4sWLSklJ0TPPPKMhQ4aoZs2aqlixYpHbGoahX3/9tSzKtDl+/LgkKTExUd99952efvppvfHGG6pdu7bi4+M1aNAg7du3T71799a2bdsKvez0xIkTNW7cuCL3tWbNmmKPQUnFzO1t6Xyw1ooVK5zaPi4uzqJK4Il4fX0Xr637Xbx40d0lAPBiZJrCWZ1pXuwQadlcKDlHjr+zmQaF45zR/XgN3IvjXzjyDABnkGcK54rPaJ7PjLZ0PpRMccefPONanM+5F8ffvTj+RbMi09CoBsBhf76UtGmaMk1Tp06dKnZbwyj7b9NfuHBB0tWu3nbt2mn27Nm2dffee6+++eYbNW/eXD/++KO+/fZbde/e3e48o0aN0ogRI+yuS0tLU7169XTfffepcuXKltSdmZmpuLg4zXtqkTIvZVkyJ6y3NPWzUm2X+/p26tRJwcHBFlcFd+P19V28tp4j95uyAFAaZJqCrM40ub8zp6xP1pVs0+n5UDLlAg292CHSoeO/+e9Dy6gq/8E5o/vxGrgXx7945BkAziDPFOTKz2g+CN6uK0aOJXPCceXMAD2fGV3s8d/QcXQZVuU/OJ9zL46/e3H8HWNFpqFRDYDDBgwY4O4SSqR8+fK25eeff77A+ltvvVX33HOP1q9fr1WrVhUagkJCQhQSElLkvoKDgy3/hZV5KUuZl7i0qKdy9vV2xXsGnoPX13fx2rofxx+AM8g0hbP6d9yVbJNGNTdy5PjzO9V1OGd0P14D9+L4F47jAsAZ5JnCueJ3zxUjR1eMbEvnhOOKO/78TnUtzufci+PvXhz/ollxbGhUA+CwTz75xN0llEi1atVsyzfeeKPdMTfddJPWr19f4JtIAAAAAHwPmQYAAACAtyLPAAAAX2D/Zt8A4ANuuOEG23Jh37bJfT47m2/FAAAAAPAsZBoAAAAA3oo8AwAA7KFRDYDPat68ue3S0ocOHbI7Jvf5a6+9tszqAgAAAABHkGkAAAAAeCvyDAAAsIdGNQA+KzQ0VF27dpUkffbZZwXWnzhxQqtXr5YkdejQoUxrAwAAAIDikGkAAAAAeCvyDAAAsCeorHaUkZGh6dOn6z//+Y9OnjypWrVqqVu3bho+fLgqVKhQVmUAsNDZs2e1a9cunT59WpcuXZJpmkWO79+/fxlV9n9Gjx6tpUuX6quvvtJ9992nAQMGSJJSU1M1cOBAXbp0SY0aNVKfPn3KvDYAAOBdyDSA7yHTAAAAf0GeAXwPeQYAAHgjpxvVjh07platWkmSHnjgAX300UcFxly+fFkdOnTQ999/b3vuwIED2rJliz777DNt3LhR4eHhzpYCoIxs3LhRY8aMUXx8vMPbGIbhlhB06623atq0aXr22Wc1cOBAjR49WuHh4dq7d68uXryomjVratGiRSpXrlyZ1wYAADwDmQbwP2QaAADgK8gzgP8hzwAAAG/m9K0/16xZo99//11Hjx5Vjx497I4ZPXq0EhISJEmmaeb72bdvn/r27etsGQDKyIwZM9SxY0fFx8cX+Ptc3I+7PPPMM9q0aZO6d++uixcv6n//938VHh6uoUOHaufOnbrtttvcVhsAAHA/Mg3gX8g0AADAl5BnAP9CngEAAN7OkkY1SapYsaLuvffeAuvT09P10UcfyTAMSdKgQYP0n//8R1OnTlX16tUlSRs2bND69eudLQWAi/38888aPny4TNPUzTffrCVLlujbb7+VdPXbOL/++qt++OEHzZw5U9HR0ZKktm3bas+ePTp06JA7S1e7du30n//8RykpKcrIyFBSUpKmTZuma6+91q11AQAA9yPTAP6DTAMAAHwNeQbwH+QZAADgC5y+9eeBAwckSTfffLOCg4MLrF+6dKkuXrwowzD01FNPaebMmbZ1jRs3VteuXSVJCxcuVIcOHZwtB4ALTZ06VdnZ2apVq5Y2b96ssLAw7dmzx7a+YcOGatiwoaKjoxUbG6uRI0dq8uTJeu6557R27Vo3Vg4AAFA4Mg3gP8g0AADA15BnAP9BngEAAL7A6SuqnT59WoZhqHHjxnbX5z3xGTZsWL51Xbp0UcOGDSVJiYmJzpYCwMU2bdokwzA0fPhwhYWFFTnWMAxNmjRJHTp00IYNGzRv3rwyqhIAAKBkyDSA/yDTAAAAX0OeAfwHeQYAAPgCpxvVUlJSJEmVK1e2u37z5s2SpGuvvVa33HJLgfW33HKLTNPUb7/95mwpAFzs999/lyTbJaMl2S4ZL0mZmZkFthk0aJBM09QXX3zh+gIBAABKgUwD+A8yDQAA8DXkGcB/kGcAAIAvcLpRLScnR5KUkZFRYN2JEyeUlJQkwzDUrl07u9uHh4dLktLT050tBYCLXb58WZJ0zTXX2J4LDQ21LZ89e7bANtddd50kae/evS6uDgAAoHTINID/INMAAABfQ54B/Ad5BgAA+AKnG9WqV68uSUpOTi6wbv369bblwkJQbngqV66cs6UAcLHcv+8XLlywPVerVi3bN3YOHDhQYJvTp09LklJTU11fIAAAQCmQaQD/QaYBAAC+hjwD+A/yDAAA8AVON6pFRUXJNE0lJCTo3Llz+dYtXLjQtnzXXXfZ3f7o0aOSrp5IAfBsN954oyTp4MGDtucqVqyo66+/XpL0n//8p8A2uc/xdxwAAHgqMg3gP8g0AADA15BnAP9BngEAAL7A6Ua1Ll26SJIuXbqkQYMG6eLFi5KkBQsWaPny5TIMQ40bN1ZUVJTd7Xfs2CHDMGyXngXgudq2bSvTNPXf//433/O9evWSaZr68MMPNW/ePF24cEEpKSl67733NGvWLBmGoQ4dOripagAAgKKRaQD/QaYBAAC+hjwD+A/yDAAA8AVON6oNHDhQ1apVkyR9/fXXqlmzpmrWrKl+/frJNE1J0rBhw+xum5iYqDNnzkiSbr/9dmdLAeBiDzzwgCRp6dKlunz5su35F198UdWrV1dmZqZiY2NVuXJl1a5dW6+++qqysrJUvnx5jRw50l1lAwAAFIlMA/gPMg0AAPA15BnAf5BnAACAL3C6Ua1GjRqaN2+eAgMDZZqmLl++rLNnz9oCUJs2bTR06FC72+a97HS7du2cLQWAi/3P//yPPvnkE02aNElnz561PV+jRg2tXr1akZGRMk0z3094eLgWL16sm266yY2VAwAAFI5MA/gPMg0AAPA15BnAf5BnAACALwiyYpIHH3xQ3333nd566y1t2bJFaWlpql+/vh555BG9/vrrCgwMLLDNhQsXNG/ePElSSEgIl5wFvMSAAQPsPn/77bdr3759Wr9+vfbs2aOsrCxdf/316ty5sypWrFjGVQIAAJQMmQbwH2QaAADga8gzgP8gzwAAAG9nSaOaJLVo0UJLlixxeHxwcLB27NhhW65QoYJVpQBw0ocffihJeuKJJ2yXjXdEcHCwOnfurM6dO7uqNAAAAJch0wC+g0wDAAD8DXkG8B3kGQAA4Mssa1QrqXLlyqlBgwbu2j2AIrzwwgsyDEMdO3bMF4I6dOggwzA0b948/v4CAAC/R6YBPBeZBgAAoGjkGcBzkWcAAIAvc1ujGgDvs3HjRhmGoQsXLri7FAAAAAAoMTINAAAAAG9FngEAAL7AZY1qu3fv1pEjR3T27FllZWWpf//+rtoVAIuVL19eGRkZSk1NdXcpAAAAbkOmAbwXmQYAAPg78gzgvcgzAADAl1naqPbbb7/p3Xff1YIFC3Tu3Ll86/4cgk6ePKnhw4fLNE21aNFCr7zyipWlAHBCZGSk9u/fr+XLl6t169buLgcAAKDMkGkA30CmAQAA/og8A/gG8gwAAPBlljWqLViwQIMHD9aFCxdkmma+dYZhFBgfERGh06dPa8OGDVqxYoWeffZZVapUyapyADiha9eu2rdvnyZNmqR169apSZMmCg4Otq1/4403VLVq1RLNaRiG5s6da3GlAAAA1iHTAL6DTAMAAPwNeQbwHeQZAADgyyxpVFu0aJH69esnSTJNU1WrVtWdd96pX3/9VQcOHCh0u6efflobNmzQpUuXtGrVKj388MNWlAPASa+99pr+85//6JdfflFiYqJ++OEH2zrTNLV06dISzWeaJiEIAAB4NDIN4FvINAAAwJ+QZwDfQp4BAAC+zOlGtdTUVMXGxso0TQUEBOjNN9/UqFGjFBISoueee67IENSjRw8FBwcrKytL69atIwQBHqJ69er64YcfNG3aNK1bt05Hjx5VRkaGfvvtNxmGoTp16uT79g4AAIA3I9MAvodMAwAA/AV5BvA95BkAAODLnG5UmzVrllJTU2UYhsaMGaM333zT4W1DQ0MVFRWlXbt2adeuXc6WAsBClStX1muvvabXXnvN9lxAQIAkac2aNYqKinJXaQAAAJbypkwzYMAAzZ8/3/Z4zJgxGjt2bLHbHT9+XDNnztSyZcuUnJysy5cvq3bt2mrXrp1iYmLUvn37EtWxdOlSff7550pMTNTJkydVuXJlNW7cWD179tTgwYNVpUoVh+eyujYgF5kGAAD4A/IMeQa+iTwDAAB8ldONaitWrJAk1axZU6+++mqJt2/SpIl27dqlQ4cOOVsKAAAAAJSYt2SaVatW5ftQx1GLFy9WTEyMUlNT8z2flJSkpKQkzZ8/X7GxsZoxY4YCAwOLnCs9PV19+/bV8uXL8z2fkpKilJQUJSQkaOrUqVq4cKFat25dprUBAAAA/og8Q54BAAAAvInTjWr79++XYRhq27atypUrV+Ltq1WrJkk6d+6cs6UAcLENGzZIkho2bOjmSgAAAKzjDZkmLS1NgwYNknT1qgcXLlxwaLu1a9fqkUceUWZmpiSpW7du6tGjh0JDQ7V9+3bNmTNHaWlpmj17tqSrV2MoTFZWlnr37q24uDhJUkREhGJjYxUVFaUzZ85owYIF2rJli37//Xd169ZN8fHxatq0aZnUBjiKTAMAAHwNeYY8A/9BngEAAL7A6Ua1M2fOSJLCw8NLtf2VK1ck/d/lagF4Li5TDgAAfJE3ZJqXX35ZR44cUd26dfWXv/xF77//frHbXL58WU899ZTtg5OpU6dq2LBhtvWPP/64Bg8erPbt2+vEiROaPXu2+vTpo06dOtmdb/bs2bYPdaKiorR+/XpFRETY1g8dOlQvvfSSpkyZotTUVA0ePFjx8fFlUhvgKDINAADwNeQZ8gz8B3kGAAD4AqeTR5UqVSRdvWRyaRw+fFiSVKNGDWdLAQAAAIAS8/RMs2HDBtu38GfMmKGwsDCHtps7d66ttu7du+f74CRXkyZNNH36dNvj0aNH250rOztb48ePtz3+/PPP832ok2vSpEm67bbbJElbtmzR6tWrXV4bAAAA4M/IM+QZAAAAwJs43ah27bXXyjRN7dy5s8TbXrx4UQkJCTIMQ02aNHG2FAAAAAAoMU/ONBcvXtTTTz8t0zT1yCOP6IEHHnB424ULF9qWR4wYUei4hx56SJGRkZKkhIQEJScnFxizadMmnThxQtLVb3BHR0fbnSswMFDDhw+3PV6wYIHLawMAAAD8GXmGPAMAAAB4E6cb1e655x5J0s8//6wdO3aUaNtp06bp4sWLkqQOHTo4WwoAAAAAlJgnZ5pRo0bp0KFDqlatmj744AOHt0tPT9eWLVskSWFhYWrXrl2hYwMCAtSlSxfb45UrVxYYs2rVKtty165di9x33vX25rK6NgAAAMCfkWfIMwAAAIA3cbpRrW/fvrblwYMH20JNcdatW2e71HFQUJD69evnbCkAAAAAUGKemmm+++47TZs2TZI0ZcoUu7emKczevXuVk5MjSWrevLkCAwOLHN+yZUvb8u7duwus/+mnn+yOtSciIkL16tWTJJ06dUopKSkurQ0AAADwZ+QZ8gwAAADgTZxuVGvRooV69eol0zT1448/6q677tLWrVsLHX/s2DG98sor6tq1q65cuSLDMPT000+rfv36zpYCAAAAACXmiZnm8uXLiomJUU5Oju699149+eSTJdp+//79tuWGDRsWOz7vmLzbumI+q2sDAAAA/Bl5hjwDAAAAeJMgKyaZM2eO9u7dq3379mnHjh1q27atrr32WpmmaRvTu3dvHTx4UHv27JEk27rmzZvr/ffft6IMAAAAACgVT8s0o0eP1v79+1WhQgV9/PHHJd4+NTXVtlyzZs1ix9eoUcPutq6Yz+ra8srIyFBGRobddWlpaZKkzMxMZWZmFrtfq+Tuq1ygYfd5IPe94K73BO9FAADKnr1zUmd+J5NnfCPPSB6eacyAAs8BknszDe9FAADcw5lMY0mjWtWqVbVhwwY99thj2rhxoyTp6NGjkiTDuPqf8UuWLJGkfMGoQ4cO+te//qWQkBArygAAAACAUvGkTPPDDz/YPigaN26cGjduXOI5zp8/b1suX758seMrVKhgW05PT3fpfFbXltfEiRM1bty4IsesWbNGFStWLHa/VnuxQ2S+xytWrCjzGuDZ4uLi3LJfR28PBgAArGPvnNSZ38nkGd/IM5JnZ5rnM6Nty+QZ2OOOTEOeAQDAPZzJNJY0qklSRESE1q1bpy+++EJ///vftXPnzkLH3nTTTXr11VfVr18/BQQ4ffdRABaLjo6WYRj6+uuv812a/PDhw5Kka6+9VoGBge4qDwAAwCU8IdNcuXJFTz75pLKzs9W8eXONGDHC6TlzP5jyRFbXNmrUqEKPWVpamurVq6f77rtPlStXtnS/RcnMzFRcXJymrE/Wlez/+1Bw89+HllkN/uiOO+6QYRj66quvPD7T5L5HOnXqpODg4DLff+6VOQAAQNmxd07q7O9k8kzZc0VtnpxpPgjeritGjiRpQ8fRZbZ/f+RNeUZyb6YhzwAA4B7OZBrLGtWkqyflTzzxhJ544gmdOHFCW7du1bFjx3Tu3DmFhoYqIiJCrVq1UmRkpJW7BWCxnTt3yjAMXbp0Kd/zkZGRCggI0P/+7/8qKirKTdUBAAC4jrszzVtvvaXdu3crMDBQc+bMKfV/PFeqVMm2/OdzOnvyjgkLC7M739mzZyVJly9fzjd/Seezura8QkJCir0aRHBwsFuaga5km/ka1dxRgz/ZtWuXDMNQVlZWvmN9/fXXe2ymcdd7k/ciAABlz97vfSt+J5NnvDvPSB6eaYwcXTGybTXAdbwxz0jueW/yXgQAwD2cyTSWNqrlVbt2bfXs2dNV0wNwodxvguXk5BRYl/fS8AAAAL6srDPNrl279M4770iSRowYoejo6GK2KFzVqlVty6dPny52/B9//GF327zP5X6wc/r06WI/2ClqPqtrA+wh0wAAAH9Hnsk/H3kG3oQ8AwAAfJnLGtUAeK8qVaro3LlzOnLkiJo1a+bucgAAAPzCp59+qszMTAUEBCg4OFhvvfWW3XH//e9/8y3njrvhhhvUp08f23KupKSkYvedd0zebfM+lzsmKSmp2CswFDWf1bUB9pBpAAAAyhZ5pnS1AfaQZwAAgC+jUQ1AATfffLPi4+P11ltvqWHDhrr++uvzXaY999s8AAAAsE7ut6JzcnL09ttvO7TNhg0btGHDBknSgw8+aPtgJyoqSgEBAcrJydGOHTuUnZ1d5G13EhMTbcv2/hP85ptv1qpVq2xj77nnnkLnOnnypI4cOSJJCg8PV61atfKtt7o2wB4yDQAAQNkiz5SuNsAe8gwAAPBlAe4uAIDnefrpp2WaphISEtS0aVOVK1fOFoJM01SzZs0UGBhYop+gIPpiAQAAykpYWJjatGkjSUpPT1d8fHyhY3NycrR69Wrb4/vvv7/AmC5dutiWV65cWeS+V6xYYVvu2rWry2sD7CHTAAAAeC/yDPwdeQYAAPgyh85Kxo8fn+/x6NGjC13njLzzAnCfJ554Qj/99JP+/ve/Kzs7u8D63G/HAQAAeAtvyDT/+Mc/9I9//KPYcWPHjtW4ceMkSWPGjNHYsWPtjnv00Ue1efNmSdKUKVPUvn17u+OWLFliux1Nq1at7N4Gp3379qpdu7ZOnDihjRs3avv27YqOji4wLjs7Wx9++GG+GlxdG2APmQYAAPgS8gx5Bv6FPAMAAHyZQ41qY8eOzXcZ2bxh5c/rnEGjGuA53n33XQ0fPlwbNmzQ0aNHlZGRoXHjxskwDD3zzDMKDw93d4kAAAAO88dMExMTo0mTJunw4cNatmyZpk+frqFDh+Ybc/DgwXzPTZgwwe5cgYGBGj16tJ599llJUv/+/bV+/foC54QjR47Uzp07JUlt2rRR586dXV4bUBgyDQAA8BXkGfIM/A95BgAA+CqHr/Oa251vL/BY0bnP/dQBz1O3bl098cQTtse533QbOnSooqKi3FUWAABAqfhbpilfvrzmzp2rrl27KjMzU8OGDdOqVavUo0cPhYaGavv27ZozZ47OnTsnSYqNjVXHjh0LnS82NlaLFy9WXFyc9uzZo1tvvVWxsbGKiorSmTNntGDBAtttb6pUqaKPP/64zGoDCkOmAQAAvoI8Q56B/yHPAAAAX+RQo9qYMWNKtQ6Ab6lfv74Mw1C5cuXcXQoAAECJ+Gum6dixoxYuXKiYmBilpqZq+fLlWr58eYFxsbGxmjFjRpFzBQUFadGiRerbt6+WL1+uEydO2L0qQN26dbVw4UI1bdq0zGoDHEWmAQAA3og8Q54BJPIMAADwDTSqAXBYcnKyu0sAAAAoFX/OND179lSrVq00Y8YMLVu2TMnJybp8+bLq1Kmjtm3b6qmnnlL79u0dmissLEzLli3T0qVLNX/+fCUmJurUqVMKCwtT48aN1atXLw0ePFhVqlQp89oAR5BpAACANyLPkGcAiTwDAAB8g8O3/gQAAAAAuN/YsWM1duzYEm1Tp04djR8/XuPHj7ekhgcffFAPPvigJXNZXRsAAAAAz0WeAQAAAPwbjWoASuXMmTP65JNPtHbtWu3evVtnzpyRJFWvXl3NmjVTx44d9eSTT6p69epurhQAAAAACiLTAAAAAPBW5BkAAOCtaFQDUGIff/yxXnrpJV28eFGSZJqmbd3Ro0d17NgxrVmzRmPHjtWUKVM0aNAgd5UKAAAAAAWQaQAAAAB4K/IMAADwZjSqASiRd955R6+//rot+FSpUkXNmzdX7dq1ZZqmTp48qR07dujcuXO6cOGChgwZotTUVL3yyiturhwAAAAAyDSAlaKH/D3f4+0z/uqmSgAAAPwDeQaw1h2rXsv3eFuXt91UCQD4D4ca1f773/+6ug5J0l133VUm+wFQOrt379abb74p0zRVp04dTZ48WX369FFwcHC+cVlZWfr3v/+tl19+WceOHdMbb7yhbt26qWnTpm6qHAAA+DsyDQCJTAMAALwTeQaARJ4BAAC+waFGtbvvvluGYbi0EMMwlJWV5dJ9AHDOtGnTlJ2drVq1amnr1q2qX7++3XFBQUF67LHH1LZtW7Vs2VIpKSmaNm2aZsyYUcYVAwAAXEWmASCRaQAAgHcizwCQyDMAAMA3BDg60DRNl/8A8Gzr16+XYRgaNWpUoQEor3r16unVV1+VaZpat25dGVQIAABQODINADINAADwVuQZAOQZAADgCxy6otpdd93l8m/rAPB8R48elSS1bt3a4W3atGkjSTp27JhLagIAAHAEmQaARKYBAADeiTwDQCLPAAAA3+BQo9rGjRtdXAYAbxAYGChJJboEfO7YgACHL+AIAABgOTINAIlMAwAAvBN5BoBEngEAAL6BsxIADsu9lHRJLhGdO9aRy1ADAAAAgCuRaQAAAAB4K/IMAADwBTSqAXBYp06dZJqm3nvvPf3000/Fjv/f//1fTZ48WYZh6L777iuDCgEAAACgcGQaAAAAAN6KPAMAAHwBjWoAHPbCCy8oJCRE58+fV9u2bTV58mSdPn26wLjTp0/r3Xff1V133aXz588rJCREL7zwQtkXDAAAAAB5kGkAAAAAeCvyDAAA8AVB7i4AgPdo0KCBPv74Yz355JM6f/68Ro4cqVGjRikyMlIREREyDEMnTpxQcnKyTNOUaZoyDEMff/wxl5UGAAAA4HZkGgAAAADeijwDAAB8AY1qAEqkf//+qlGjhp555hkdPXpUpmnq0KFDSkpKkiSZpmkbe80112jWrFnq2rWru8oFAAAAgHzINAAAAAC8FXkGAAB4O4ca1QIDA23LhmEoKyvL7jpn/HleAJ6rW7duSkpK0uLFi7V27Vrt3r1bZ86ckSRVr15dzZo1U8eOHdWzZ08FBdEPCwAA3I9MAyAvMg0AAPAm5BkAeZFnAACAN3Po7CT30rB5u/AdWQfAdwUFBalPnz7q06ePu0sBAAAoFpkGwJ+RaQAAgLcgzwD4M/IMAADwVgGODiwq5BCAAAAAAHg6Mg0AAAAAb0WeAQAAAOALHLqiWk5OTqnWAQAAAIAnINMAAAAA8FbkGQAAAAC+wuErqgEAAAAAAAAAAAAAAAAAUBo0qgEAAAAAAAAAAAAAAAAAXIpGNQAAAAAAAAAAAAAAAACAS9GoBgAAAAAAAAAAAAAAAABwKRrVAAAAAAAAAAAAAAAAAAAuFWTlZJs3b9ZXX32lxMREJSUlKS0tTVlZWQ5taxiGw2MBAAAAwBXINAAAAAC8FXkGAAAAgKezpFEtJSVF/fr109q1a/M9b5qmFdMDAAAAgEuRaQAAAAB4K/IMAAAAAG/hdKPaxYsXddddd+nAgQOEHsDHxcTESJLuv/9+9enTx83VAAAAWINMA/gPMg0AAPA15BnAf5BnAACALwhwdoL33ntP+/fvlyQFBQVp0KBBWr9+vU6dOqXMzEzl5OQ49JOdne30HwaAa3322Wf67LPPVLlyZXeXAgAAYBkyDeA/yDQAAMDXkGcA/0GeAQAAvsDpK6p9/fXXtuV//etfeuihh5ydEoCHqlWrllJSUhQREeHuUgAAACxDpgH8B5kGAAD4GvIM4D/IMwAAwBc4fUW1Q4cOyTAM3XnnnQQgwMdFRUVJkn777Tc3VwIAAGAdMg3gP8g0AADA15BnAP9BngEAAL7A6Ua18uXLS5KaNm3qdDEAPFu/fv1kmqY+++wzd5cCAABgGTIN4D/INAAAwNeQZwD/QZ4BAAC+wOlGtYYNG0qSzp8/73QxADzbk08+qXvvvVdLly7VuHHjZJqmu0sCAABwGpkG8B9kGgAA4GvIM4D/IM8AAABfEOTsBA899JB+/PFHbd261Yp6AHiwzZs366WXXlJKSorGjx+vr776So888ohuueUWVatWTYGBgUVuf9ddd5VRpQAAAI4j0wD+g0wDAAB8DXkG8B/kGQAA4AucblR75plnNHnyZP3222/69NNPNXDgQAvKAuCJ7r77bhmGYXt84MABTZgwwaFtDcNQVlaWq0oDAAAoNTIN4D/INAAAwNeQZwD/QZ4BAAC+wOlbf9aoUUOff/65goKCNGTIEC1YsMCKugB4KNM0S/0DAADgicg0gH8h0wAAAF9CngH8C3kGAAB4O6evqCZJ3bt316pVqzRgwAD169dP77//vnr37q2oqChVqVIlX3d/UbjkLODZNmzY4O4SAAAAXIJMA/gHMg0AAPBF5BnAP5BnAACAL7CkUU2SWrRooccee0yTJ0/W9u3btX379hJtzyVnAc/Xvn17d5cAAADgMmQawPeRaQAAgK8izwC+jzwDAAB8gSWNajt27FCXLl10+vRp2zdzuIQsAAAAAG9BpgEAAADgrcgzAAAAALyF041qhw8fVocOHXTu3DnbcyEhIWrcuLGqVaumoCDLLtoGAAAAAJYj0wAAAADwVuQZAAAAAN7E6YTyt7/9TefOnZNhGKpTp46mTJmihx56SCEhIVbUB8BDpaWl6euvv9bWrVt14sQJXbx4UfPmzVODBg1sY44dO6bU1FSVL19ejRo1cmO1AAAAhSPTAP6JTAMAAHwBeQbwT+QZAADgrZxuVFuzZo0kqVy5clq/fr2aNGnidFEAPNv06dP1+uuvKz09XdLVy8gbhqELFy7kG7dp0yY9/vjjKl++vH7//XdVr17dHeUCAAAUiUwD+B8yDQAA8BXkGcD/kGcAAIA3C3B2gmPHjskwDHXo0IEABPiBsWPHavjw4UpLS1O5cuV0++23Fzr2kUceUZ06dZSRkaFFixaVYZUAAACOI9MA/oVMAwAAfAl5BvAv5BkAAODtnG5Uy+2+r1evntPFAPBsO3bs0IQJEyRJ/fr104kTJ7Rt27ZCxwcEBKhPnz4yTVNxcXFlVSYAAECJkGkA/0GmAQAAvoY8A/gP8gwAAPAFTjeqXX/99ZKklJQUp4sB4NmmTp0q0zR15513av78+apSpUqx29x5552SpJ9++snV5QEAAJQKmQbwH2QaAADga8gzgP8gzwAAAF/gdKPao48+KtM0tXHjRl2+fNmKmgB4qE2bNskwDA0bNszhbSIjIyVJR48edVFVAAAAziHTAP6DTAMAAHwNeQbwH+QZAADgC5xuVHvyySfVtGlTpaam6tVXX7WiJgAe6vjx45KkG264weFtQkJCJEkZGRkuqQkAAMBZZBrAf5BpAACAryHPAP6DPAMAAHyB041qFSpU0JIlS3Tddddp2rRpevrpp3X69GkraiuVxMRETZ8+XQMHDlTLli0VGRmpSpUqKSQkRBEREbr77rs1fvx4HT582OE5jx8/rjFjxig6OlrVq1dXxYoV1ahRIw0YMECbNm0qcY1Lly7Vww8/rAYNGqh8+fIKDw/XnXfeqXfffVfnzp0r0VxW1wYUpVy5cpKkzMxMh7fJDU5Vq1Z1RUkAAABO87RMA8B1yDQAAMDXkGcA/0GeAQAAviDI2QnGjx8vSXrggQc0ffp0ffLJJ/ryyy911113qWnTpqpSpYoMw3BortGjRztbju655x5duHDB7rpTp07p1KlT2rRpk95++22NGTNGo0aNKnK+xYsXKyYmRqmpqfmeT0pKUlJSkubPn6/Y2FjNmDFDgYGBRc6Vnp6uvn37avny5fmeT0lJUUpKihISEjR16lQtXLhQrVu3LvbPamVtgCPq1q2rn3/+WXv27NEdd9zh0DZr1qyRJF133XWuLA0AAKDUPC3TAHAdMg0AAPA15BnAf5BnAACAL3C6UW3s2LEFQk5GRobWrl2rtWvXlmguq0JQeHi47rjjDjVt2lS1a9dW7dq1ZZqmkpOT9e2332rLli3KyMjQa6+9pszMzEL3u3btWj3yyCO2byZ069ZNPXr0UGhoqLZv3645c+YoLS1Ns2fPliTNmjWr0JqysrLUu3dvxcXFSZIiIiIUGxurqKgonTlzRgsWLNCWLVv0+++/q1u3boqPj1fTpk0Lnc/K2gBHdejQQXv37tUnn3yiJ598stjxhw4d0ty5c2UYhjp16lQGFQIAAJScJ2YaAK5BpgEAAL6GPAP4D/IMAADwBU7f+lOSTNPM92PvueJ+rJKQkKATJ05o2bJleuedd/TCCy/o0Ucf1WOPPaZRo0YpPj5en332mS24TZgwQceOHSswz+XLl/XUU0/ZGsGmTp2q5cuXa9CgQXr88cc1ZcoUJSYmqnbt2pKk2bNn25rQ7Mm7PioqSrt27dKECRP02GOPaejQoYqPj9eLL74oSUpNTdXgwYMLncvq2gBHDRs2TEFBQdqyZYvGjh1b5NgffvhB9913n86fP6+QkJAi39MAAADu5kmZBoDrkGkAAIAvIs8A/oE8AwAAfIHTV1QbM2aMFXVYplmzZsWO6d+/v77++mstW7ZMWVlZWrVqlWJiYvKNmTt3rg4fPixJ6t69u4YNG1ZgniZNmmj69Onq3bu3pKvfNrL3jYTs7Gzb5bcl6fPPP1dERESBcZMmTdK6deu0c+dObdmyRatXr1bnzp0LjLOyNqAkmjRpojfffFNjxozRhAkTtHLlStt7TJJWrVqlZcuWac2aNdq4caMkyTAMvfPOO6pTp46bqi7ojTfe0N/+9jdJV5tV33jjDTdXBAAA3MnTMg0A1/GFTEOeAQAAeZFnAP/hC3lGItMAAODvfK5RzVFNmzbVsmXLJEknT54ssH7hwoW25REjRhQ6z0MPPaTIyEglJycrISFBycnJioyMzDdm06ZNOnHihCSpffv2io6OtjtXYGCghg8fbmuaW7Bggd1GNStrA0rqzTffVGZmpt5++20lJibqhx9+sF2h8OWXX7aNM01ThmFo9OjRGj58uLvKLeDnn3/W5MmT3V0GAADwIN6aaQCUjjdnGvIMAAD4M/IM4F+8Oc9IZBoAAGDRrT+90S+//GJbzr1FZq709HRt2bJFkhQWFqZ27doVOk9AQIC6dOlie7xy5coCY1atWmVb7tq1a5F15V1vby6rawNKY/z48UpISFCvXr1UoUKFApeJDw4O1v3336/Nmzd71H+UmKapwYMHKzg4WB06dHB3OQAAAADcxBszDXkGAAAAgOSdeUYi0wAAgKucvqKaN1qyZIm++eYbSVKFChXUrVu3fOv37t2rnJwcSVLz5s0VGBhY5HwtW7bUzJkzJUm7d+8usP6nn37KN7YoERERqlevno4cOaJTp04pJSVFtWrVclltQGm1aNFCX3/9tbKysrR3716dOnVK2dnZqlGjhpo2baoKFSq4u8QC5s6dq82bN2vSpEnau3evu8sBAAAA4EbelmnIMwAAAAByeVuekcg0AADgKp9uVPvvf/+rM2fOSJKuXLmiI0eOaPXq1YqLi5MkBQcHa9asWQoPD8+33f79+23LDRs2LHY/ecfk3daZ+Y4cOWLbNm+jmtW1Ac4KCgrSLbfc4u4yipWSkqJXX31VUVFR+utf/6rY2Fh3lwQAAADAA3hDpiHPAAAAALDHG/KMRKYBAAD/x6cb1V555RV9//33BZ43DEP33HOPxo8frzZt2hRYn5qaaluuWbNmsfupUaOG3W1dMZ/Vtf1ZRkaGMjIy7K5LS0uTJGVmZiozM7PYuayQu5/MzEwFVwgu8DzgqOeff15nzpzRv/71L0myXZkwOzvbqfcT70UAANzD3jkpv5cB+Kq//vWvOnPmjL755hsFBwcXvwEAAAAAeBAyDQAAyGV5o5ppmtq7d6927typ06dPKz093dYQUpzRo0dbXY5ddevWVYcOHRQZGWl3/fnz523L5cuXL3a+vJfPTU9Pd+l8Vtf2ZxMnTtS4ceOKHLNmzRpVrFix2LmsFBcXp8H/fMT2eMWKFWW6f9iXmpqq3bt367fffrO9NytVqqQGDRqoWbNmqlq1qnsL/P927dqlBQsWqH379jp//rxWrFih33//XZJ04MABp95PFy9etKpMAABQAvbOSa36vewNmQaANU6ePKmNGzdq9+7dtiuyV69eXc2aNdPdd9+tiIgIN1corVu3Tl9++aX69eun9u3bu7scAADg4cgzgP/whjwjkWkAAEB+ljWqZWZmavLkyfroo490/PjxUs1hdQhKSEiwLV+4cEEHDx7U0qVLNWXKFL3xxht6//339c9//lOdO3cudA7DMCytyUquqG3UqFEaMWKE3XVpaWmqV6+e7rvvPlWuXNnyfduTmZmpuLg4derUSQ/Xetr2/NLUz8pk/7DvyJEjevXVV/Wf//xHWVlZdscEBgbqwQcf1DvvvKP69euXcYX/5/Lly3rppZdUpUoVffHFF7ZgtmjRIklSkyZN1LVr1yLnKOpKg7kNoFZeaTB3nuAKPn3RS69X2tc775Ui4Xt4fX0Xr63nyH0N7J2T5l4B2Jm5PS3TAHCNI0eO6KWXXtKSJUuKzDQ9e/bU5MmT3ZZpLl++rGeeeUZVqlTRe++9V+p5yvLq6blzlAv03P9P8WW5x92R42/F6/3n/fj7uRLnjO7Ha+BeHP/iufLYkGcA/+EteUayJtOU9d2gbJnGDLBkPpRM7nEv7vhb9XqXMwNdMq+34nzOvTj+7sXxd4wVx8eSLogzZ86oU6dO2rlzp0zTLHa8YRgFxrm6ISw0NFS33XabbrvtNvXr10/t2rXT8ePH1aNHD23btk233nqrbWylSpVsy5cuXSp27rxjwsLCCqyvVKmSzp49K+nqCVne+Us6n9W1/VlISIhCQkKKHBMcHFzml+UNDg5W5qXMfI/hHps3b1b37t2Vnp5e5N/3rKwsffPNN4qLi9Py5cvVtm3bMqzy/4wbN06//PKLpk2bprp169qeDwi4eoIbGBhY7Pvpb3/7m1uuNBgzt7el88Fazl7ZMS4uzqJK4Il4fX0Xr6375V41zd45qTPniN6QaQBYoySZ5uuvv9bq1avdlmneeustW55x5moI7rh6+osdIi2bCyXnyPG34mr1ozo1tHxOX8A5o/vxGrgXx79wrro7A3kG8B/elGckazKNu+4G9XxmtKXzoWSKO/5WZY+X1cIl83o7zufci+PvXhz/olmRaSxpVPvLX/6iHTt2SJJq166tnj176ocfflBiYqIMw9Do0aOVnp6u3377TVu2bNGJEydkGIZCQ0M1dOhQh25haaXGjRtr4sSJGjhwoK5cuaK3335bCxcutK3Pe7vC06dPFzvfH3/8YXfbvM/lNqqdPn262Ea1ouazujagJI4eParu3bvbvqFy//33KyYmRnfccYctYJw8eVKJiYmaN2+eVqxYobS0NHXv3l179uzRNddcU6b1/vzzz5o8ebKio6M1ZMiQUs9T1lcazL2S4LynFinzkv1vQ8H9SnNlxwerDlBwhSDFzO1d6OvLFSO9W94rgdJU7Vt4bT2Hs1dNK4y3ZRoApeNNmcaqPCOVbabJ/Z05ZX2yrmQX/0E5rFUu0NCLHSIdOv6b/z60xPO3++t0y+f0JZwzuh+vgXtx/ItHngHgDG/KM5L3f0bzQfB2XTEcu3UyrFPODNDzmdHFHv8NHUt+FdB71o4vdkxp5vUlnM+5F8ffvTj+jrEi0zjdqLZmzRqtX79ehmGoRYsWiouLU+XKlfXcc88pMTFRkjRmzBjb+JycHC1evFgjRozQ77//rtWrV+vbb79VnTp1nC2lRLp162Zb3rhxY751N9xwg205KSmp2Lnyjsm7bd7ncsckJSUpMjKy1PNZXRtQEu+8847S0tIUGBioTz75RP369Sswpl69eqpXr5569eqlf/7zn+rfv7/S0tL0zjvv6MMPPyzTep999lllZWVpxowZtiuolYa7rjSYeSkr35UE4VlK83rnfT0Le3058fEN7rj6KMoGr637ueL4e2umAVBy3pRprMozknsyzZVsk0Y1N3Lk+Jfm9XbFnL6Ic0b34zVwL45/4cgzAJzhTXlG8v7PaK4YObpiZFs6JxxX3PEvVZ5x4PXkHOYqzufci+PvXhz/ollxbJy+ufZXX31lW547d26x3eoBAQHq3bu3fvjhB91www3atWuXHn30UeXklG1Het7bYKampuZbFxUVZTth2rFjh7Kzi/6llRv2JKlZs2YF1t988812x9pz8uRJHTlyRJIUHh6uWrVqubQ2oCRWrFghwzAUGxtrNwD9Wd++fTVo0CCZpqlvv/22DCrMb8eOHTIMQz169FDt2rXz/eReRXHSpEmqXbu2WrZsWeb1AQAAz+CtmQZAyXlTpiHPAAAAR5BnAP/hTXlGItMAAAD7nG5U27p1qySpadOmJWqEqlWrlr744guZpqn4+Hh9/fXXzpZSIgcPHsxXS15hYWFq06aNJCk9PV3x8fGFzpOTk6PVq1fbHt9///0FxnTp0sW2vHLlyiLrynvf665duxZYb3VtQEkcO3ZMktSnTx+Ht8kdm7ttWcvOztbJkycL/Fy+fFmSdP78eZ08eVIpKSluqQ8AALift2YaACXnbZmGPAMAAIpDngH8h7flGYlMAwAACnK6Ue3EiRMyDCPfVcMkyTAM23JGRobdbaOjo9WiRQtJ0pdffulsKSUyc+ZM23Ju41dejz76qG15ypQphc6zZMkS2+01W7VqZfe2nu3bt1ft2rUlXb3N6Pbt2+3OlZ2dne+yu3lrcFVtQElUq1ZNklSlShWHt8kdm7ttWUpNTZVpmnZ/BgwYIEmaMGGCTNNUcnJymdcHAAA8g7dmGgAl502ZhjwDAAAcQZ4B/Ic35RmJTAMAAOxzulHtwoULkqSqVavmez40NNS2/Odba+bVrFkzmaapn376ydlSNHPmTG3YsEGmaRY6Jjs7W++8844++ugj23PPPvtsgXExMTGqX7++JGnZsmWaPn16gTEHDx7U0KFDbY8nTJhgd5+BgYEaPXq07XH//v116tSpAuNGjhypnTt3SrraPNe5c2e781lZG1ASuf9pUZK/r7ljc7cFAADwNJ6UaQC4FpkGAAD4GvIM4D/IMwAAwBcEOTtBWFiYUlNTbZdozZW3Mz8pKUkRERF2t8/d7sSJE86WooSEBA0ZMkT16tVTp06ddPPNNys8PFzlypVTamqqdu/eraVLl+bryh81apTat29fYK7y5ctr7ty56tq1qzIzMzVs2DCtWrVKPXr0UGhoqLZv3645c+bo3LlzkqTY2Fh17Nix0NpiY2O1ePFixcXFac+ePbr11lsVGxurqKgonTlzRgsWLLDdxrNKlSr6+OOPC53L6toARw0fPlzLly/Xu+++qz59+qhixYpFjr948aImTZokwzD03HPPlVGVAAAAJeNJmQaAa5FpAACAryHPAP6DPAMAAHyB041qDRo0UGpqaoF7h99000225fj4eLVq1cru9rt27ZIklStXztlSbI4cOaJ58+YVOaZKlSqaOHGihgwZUuiYjh07auHChYqJiVFqaqqWL1+u5cuXFxgXGxurGTNmFLm/oKAgLVq0SH379tXy5ct14sQJu1c5q1u3rhYuXKimTZsWOZ+VtQGO6tixo8aMGaNx48bp7rvv1qxZs3TbbbfZHbtr1y4NGjRI+/fv15gxY9SpU6eyLRYAAMBBnphpALgGmQYAAPga8gzgP8gzAADAFzjdqHbLLbdo586d+vnnn/M936pVKwUGBionJ0czZ87Us88+W6Cz/8svv9TPP/8swzB0ww03OFuKpk2bpscee0ybNm1SQkKCjh07plOnTik9PV2hoaGKiIjQLbfcos6dO6tPnz4O3cO9Z8+eatWqlWbMmKFly5YpOTlZly9fVp06ddS2bVs99dRTdq/IZk9YWJiWLVumpUuXav78+UpMTNSpU6cUFhamxo0bq1evXho8eLDD95a3sjYgr/Hjxxe6zjAMtWjRQj/88INuv/123XzzzWrZsqXCw8NlGIZOnjypxMTEApeTHj9+fL5b4Lrbp59+qk8//dTdZQAAAA/gSZkGgDV8PdOQZwAAQC7yDOB7fD3PSGQaAAD8mdONanfffbfmz5+vQ4cO6fjx46pTp44kqVatWurSpYu+/fZbJSUl6a677tIbb7yhpk2b6uLFi1q6dKneeecd2zy9evVythRVqlRJnTt3VufOnZ2eK686depo/PjxRZ4YlsSDDz6oBx980JK5rK4NkKSxY8fKMIwixxiGIdM09dNPP9kCT16macowDP3www/64YcfJMmjQhAAAEAuT8o0AKxBpgEAAP6CPAP4HvIMAADwZU43qnXr1s32rZzFixfr2Wefta2bNGmS1q5dqytXrmjHjh3q3bu33TkaNWqkoUOHOlsKAAuZpun0OEfnAAAAcCcyDeCbyDQAAMAfkGcA30SeAQAAvsrpRrVatWrp/fff17FjxwpcNjoqKkpLlizRo48+qnPnztndvkmTJlq2bJkqVarkbCkALJKTk+PuEgAAAMoMmQbwPWQaAADgL8gzgO8hzwAAAF/mdKOaJD333HOFruvcubMOHjyo2bNna926dTp27JgCAgLUqFEjPfDAAxo4cKDKlStnRRkAAAAAUCpkGgAAAADeijwDAAAAwFtY0qhWnJo1a2rUqFEaNWpUWewOAAAAACxFpgEAAADgrcgzAAAAADxFgCODOnTooA4dOmjq1KmurgcAAAAALEemAQAAAOCtyDMAAAAAfIVDV1TbuHGjDMNQs2bNCqxr1KiRJGnAgAEaM2aMtdUB8HhpaWlKT09XdnZ2sWPr169fBhUBAAAURKYBUBgyDQAA8HTkGQCFIc8AAABv4/StP5OTk2UYhv744w8r6gHgBdasWaMZM2Zo8+bNOnv2rEPbGIahrKwsF1cGAABQcmQawP+QaQAAgK8gzwD+hzwDAAC8mUONauXKlVNmZqYuXLjg6noAeLhnnnlGs2fPliSZpunmagAAABxDpgGQi0wDAAC8DXkGQC7yDAAA8HYONarVrFlTx48f1969e11dDwAP9vHHH2vWrFmSpLCwMPXs2VO33nqrqlatqoCAADdXBwAAUDgyDQCJTAMAALwTeQaARJ4BAAC+waFGtdtvv13Lli3Ttm3b9Le//U0xMTGqU6eOq2sD4GFyA9BNN92k9evXKyIiws0VAQAAOIZMA0Ai0wAAAO9EngEgkWcAAIBvcKi9/oknnrAtjx49WnXr1lVgYKACAwMlXb207PTp023PleYnKMihnjkAbrRv3z4ZhqGxY8cSgAAAgFch0wCQyDQAAMA7kWcASOQZAADgGxxqVHv44YfVr18/maZZ4CeXvXUl/QHg2SpVqiRJuv76691cCQAAQMmQaQBIZBoAAOCdyDMAJPIMAADwDQ7fsHz+/Pn68ssv1blzZ9WqVcv27RrDMFxWHADPctNNN0mSjh8/7uZKAAAASo5MA4BMAwAAvBV5BgB5BgAA+AKHG9Uk6bHHHtPKlSt18uRJXblyRTk5OTJNU4ZhaNiwYcrJySn1T3Z2tqv+jAAsMmjQIJmmqQULFri7FAAAgFIh0wD+jUwDAAC8GXkG8G/kGQAA4AtK1KgGwL/17dtXvXv31pdffqlp06a5uxwAAAAAKBEyDQAAAABvRZ4BAAC+IMjZCerXry/DMFSjRg0r6gHg4b788ku99NJLeuGFF/TVV1/pL3/5i5o0aaKKFSsWu+1dd91VBhUCAACUDJkG8C9kGgAA4EvIM4B/Ic8AAABv53SjWnJysgVlAPAWwcHBuvXWW1W9enVt3bpVW7dudWg7wzCUlZXl4uoAAABKjkwD+BcyDQAA8CXkGcC/kGcAAIC3c7pRDYD/yMrK0mOPPaZvvvlGkmSappsrAgAAAADHkWkAAAAAeCvyDAAA8AU0qgFw2MyZM7Vo0SJJUmRkpAYMGKBbbrlFVatWVUBAgJurAwAAAICikWkAAAAAeCvyDAAA8AU0qgFw2Jw5cyRJrVq10rp161ShQgU3VwQAAAAAjiPTAAAAAPBW5BkAAOALaK8H4LBffvlFhmFo1KhRBCAAAAAAXodMAwAAAMBbkWcAAIAvoFENgMNCQ0MlSfXq1XNzJQAAAABQcmQaAAAAAN6KPAMAAHwBjWoAHHbrrbdKkn777Tc3VwIAAAAAJUemAQAAAOCtyDMAAMAX0KgGwGHPPPOMTNPU3Llz3V0KAAAAAJQYmQYAAACAtyLPAAAAX0CjGgCH9erVS88884yWL1+ul156SdnZ2e4uCQAAAAAcRqYBAAAA4K3IMwAAwBcEubsAAN5j/vz5atWqlXbs2KG///3v+vrrr9WrVy81adJEFStWLHb7/v37l0GVAAAAAGAfmQYAAACAtyLPAAAAX+BQo9r8+fMlSTfddJNatmzp0oIAeK6BAwfKMAzb4yNHjuiDDz5waFvDMAhBAADAbcg0ACQyDQAA8E7kGQASeQYAAPgGhxrVck98hg4dWiAEjR8/XpJ0xx13qEuXLtZXCMCjmKbp7hIAAABKjEwDIBeZBgAAeBvyDIBc5BkAAODtnL7159ixY20BiRAE+LakpCR3lwAAAGA5Mg3gP8g0AADA15BnAP9BngEAAL7AoUa1gIAAmaap7OxsV9cDwIM1aNDA3SUAAACUCpkGgESmAQAA3ok8A0AizwAAAN8Q4MigypUrS5KOHj3q0mIAAAAAwBXINAAAAAC8FXkGAAAAgK9wqFGtSZMmMk1Ta9eu1aFDh1xdEwAAAABYikwDAAAAwFuRZwAAAAD4Codu/dm1a1dt27ZNly5dUrNmzdSuXTvVqVNHAQH/1+e2bt06xcTElLoQwzA0d+7cUm8PAAAAAIUh0wAAAADwVuQZAAAAAL7CoUa15557TnPmzNHvv/+ujIwMrV27Nt960zS1b98+7du3z6liCEGAZ2vUqFGptzUMQ7/++quF1QAAADiOTANAItMAAADvRJ4BIJFnAACAb3CoUa1atWravHmzBg0apLi4OLtjTNN0qhDDMJzaHoDrJScnOzzWMIx8/y7wdxwAALgTmQaARKYBAADeiTwDQCLPAAAA3+BQo5okNWjQQKtXr9bJkye1fft2nT17VleuXFFMTIwMw9C9996rxx9/3JW1AnCzAQMGFDvmwoUL2r9/v3766ScZhqHmzZvr5ptvLoPqAAAAikamAUCmAQAA3oo8A4A8AwAAfIHDjWq5IiIidP/999sex8TESJJuvPFGh06QAHivTz75xOGxu3fvVkxMjPbs2aNRo0apd+/eLqwMAADAcWQawH+RaQAAgLcjzwD+izwDAAB8QYAVkzh7SWkAvqdZs2Zav3696tSpo/79+2vfvn3uLgkAAKBQnpRpEhMTNX36dA0cOFAtW7ZUZGSkKlWqpJCQEEVEROjuu+/W+PHjdfjwYYfnPH78uMaMGaPo6GhVr15dFStWVKNGjTRgwABt2rSpxDUuXbpUDz/8sBo0aKDy5csrPDxcd955p959912dO3euRHNZXRtgFTINAADwFuSZkiHPwB+QZwAAgKcq8RXV/mzDhg2SpGuvvdbpYgD4lkqVKmnEiBEaNmyYJk+erLlz57q7JAAAgAI8LdPcc889unDhgt11p06d0qlTp7Rp0ya9/fbbGjNmjEaNGlXkfIsXL1ZMTIxSU1PzPZ+UlKSkpCTNnz9fsbGxmjFjhgIDA4ucKz09XX379tXy5cvzPZ+SkqKUlBQlJCRo6tSpWrhwoVq3bl3sn9XK2gBXINMAAABPR54hzwCFIc8AAABP5HSjWvv27a2oA4CPatmypSRp3bp1bq4EAADAPk/MNOHh4brjjjvUtGlT1a5dW7Vr15ZpmkpOTta3336rLVu2KCMjQ6+99poyMzM1evRou/OsXbtWjzzyiDIzMyVJ3bp1U48ePRQaGqrt27drzpw5SktL0+zZsyVJs2bNKrSmrKws9e7dW3FxcZKu3nIoNjZWUVFROnPmjBYsWKAtW7bo999/V7du3RQfH6+mTZsWOp+VtQGuRKYBAACejDxDngGKQp4BAACexulGteKcP39e6enpCgsLU6VKlVy9OwAeJvey8ydPnnRzJQAAAKVT1pkmISFBTZs2lWEYdtePGjVK8+fP18CBA2WapiZMmKCnn35a11xzTb5xly9f1lNPPWX74GTq1KkaNmyYbf3jjz+uwYMHq3379jpx4oRmz56tPn36qFOnTnb3O3v2bNuHOlFRUVq/fr0iIiJs64cOHaqXXnpJU6ZMUWpqqgYPHqz4+Hi7c1ldG+BKZBoAAODNyDNXkWfgr8gzAADA0wRYPeFvv/2m119/Xa1bt1aFChVUpUoV1a1bV1WqVFGFChXUunVrvfHGG/rtt9+s3jUAD7R69WpJUpUqVdxcCQAAgGPcnWmaNWtW6Ic6ufr3768HHnhA0tUrA6xatarAmLlz5+rw4cOSpO7du+f74CRXkyZNNH36dNvjwq5kkJ2drfHjx9sef/755/k+1Mk1adIk3XbbbZKkLVu22M4FXVkb4GpkGgAA4E3IMwWRZ+DPyDMAAMDTWNaoduXKFb3wwgu67rrr9M477+j7779XRkaGTNO0/WRkZOj777/XxIkTdd111+mvf/2rrly5YlUJADzMV199pYkTJ8owDLVt29bd5QA+q1NAn3w/AIDS8bZMk/c2NPa+Gb1w4ULb8ogRIwqd56GHHlJkZKSkq1c/SE5OLjBm06ZNOnHihKSrtxaKjo62O1dgYKCGDx9ue7xgwQK746ysDXAlMg1QNqKH/D3fDwCg5Mgz5Bngz8gzQNm4Y9Vr+X4AAEWz5Nafly5dUqdOnbR161bbJWQLk7s+OztbH374oRITE7V27VqVL1/eilIAuFBMTEyxY3JycnT27Flt375dx44dk2maCgoK0siRI8ugQgAAgNLxxkzzyy+/2JZr166db116erq2bNkiSQoLC1O7du0KnScgIEBdunTRzJkzJUkrV67UkCFD8o3Je4WDrl27FllX3vUrV64ssN7q2oCSINMAAABfRJ4hz8A/kGcAAIAvsKRRLTY2Vt99953tcs5NmzZVTEyM2rRpo8jISIWGhurChQv67bfftGXLFn322WfatWuXTNPU1q1bFRsbq88//9yKUgC40KefflrsZdtz5f6HR+XKlTVnzhy1aNHClaUBAAA4xdsyzZIlS/TNN99IkipUqKBu3brlW793717l5ORIkpo3b67AwMAi52vZsqXtw5Pdu3cXWP/TTz/lG1uUiIgI1atXT0eOHNGpU6eUkpKiWrVquaw2oCTINAAAwBeRZ8gz8A/kGQAA4AucblTbtm2b/vnPf8owDAUEBGjy5Ml6/vnnC5wohYaGKjw8XC1bttQLL7ygadOm6a9//auys7P1z3/+U8OHDy82IABwr/r16xcbggICAhQWFqaGDRuqffv26tevn2rWrFlGFQIAAJScJ2ea//73vzpz5oykq7fyOXLkiFavXq24uDhJUnBwsGbNmqXw8PB82+3fv9+23LBhw2L3k3dM3m2dme/IkSO2bfN+sGN1bUBJkGkAAICvIc+QZ+A/yDMAAMAXON2olvdbNpMnT9YLL7zg0HbDhg2TaZp6/vnnJUnz58+nUQ3wcMnJye4uAQAAwHKenGleeeUVff/99wWeNwxD99xzj8aPH682bdoUWJ+ammpbduQ/pGvUqGF3W1fMZ3VteWVkZCgjI8PuurS0NElSZmamMjMzi92vVXL3VS7QsPs8ytbBgwdLtZ0rX6/cud31nuC9CABA2bN3Tlra38nkGd/JM5KHZxozoMBzKFuemGfyzu+O9wXvRQAA3MOZTON0o9rGjRslSddcc43DASjXsGHDNHnyZB09elQbNmxwthQAAAAAKDFvzDR169ZVhw4dFBkZaXf9+fPnbcvly5cvdr4KFSrYltPT0106n9W15TVx4kSNGzeuyDFr1qxRxYoVi92v1V7sEJnv8YoVK8q8Bni23CuLlLWLFy+6Zb8AAPgze+ekpf2dTJ7xnTwjeXameT4z2rZMnoE97sg05BkAANzDmUzjdKPa0aNHZRiG2rVrV+Jtc7dbsGCBjh075mwpAAAAAFBinpxpEhISbMsXLlzQwYMHtXTpUk2ZMkVvvPGG3n//ff3zn/9U586di6zRU1ld26hRozRixAi769LS0lSvXj3dd999qly5sqX7LUpmZqbi4uI0ZX2yrmSbtuc3/31omdUAz5b7HunUqZOCg4PLfP+5V+YAAABlx945aWl/J5Nn3McVtXlypvkgeLuuGDmSpA0dR5fZ/uH53JlpyDMAALiHM5nG6Ua1S5cuSZIqVapUqu1DQ0PzzQMAAAAAZclbMk1oaKhuu+023XbbberXr5/atWun48ePq0ePHtq2bZtuvfVW29i8fxZH6so7JiwsrMD6SpUq6ezZs5Kky5cvF3usiprP6tryCgkJUUhISJFjgoOD3dIMdCXbzNeo5o4a4Nnc9d7kvQgAQNmz93u/tL+TyTO+k2ckD880Ro6uGNm2GoA/c8d7k/ciAADu4UymCSh+SNFq1qwpSfrll19Ktf2vv/6abx4AAAAAKEvemGkaN26siRMnSpKuXLmit99+O9/6qlWr2pZPnz5d7Hx//PGH3W1dMZ/VtQEAAAD+jDxDngEAAAC8idONak2bNpVpmoqPj9ehQ4dKtG1SUpI2b94swzDUtGlTZ0sBYKHAwEBLf4KCnL6AIwAAgEt4a6bp1q2bbXnjxo351t1www225aSkpGLnyjsm77aumM/q2oDCkGkAAIA/IM+QZ+CbyDMAAMBXOd2o1rVrV0lSdna2+vbt6/A9Ry9cuKB+/fopKytLkvTAAw84WwoAC5mmafkPAACAJ/LWTJP3tjGpqan51kVFRSkg4Grc27Fjh7Kzs4ucKzEx0bbcrFmzAutvvvlmu2PtOXnypI4cOSJJCg8PV61atVxaG1AYMg0AAPAH5BnyDHwTeQYAAPgqp9vnn3rqKb3zzjs6efKkEhMTdfvtt+u9995T9+7dbSfreZmmqW+//VYvv/yyDhw4IMMwFBERoZiYGGdLAWChAQMGOLW9aZpasWKF/vjjDwIQAADwaN6aaQ4ePGhb/vOHJ2FhYWrTpo02b96s9PR0xcfHq3379nbnycnJ0erVq22P77///gJjunTposmTJ0uSVq5cqVdeeaXQulasWGFbzv3QzJW1AYUh0wAAAH9AniHPwDeRZwAAgK9yulEtNDRUs2bNUs+ePZWTk6Nff/1VvXr1Uo0aNfQ///M/atCggUJDQ3XhwgUdPnxY27ZtU0pKiqSrJ0lBQUGaM2eOKlas6PQfBoB1Pvnkk1Jvu2TJEo0ZM0Z//PGH7bl69epZURYAAIDlvDXTzJw507bcpk2bAusfffRRbd68WZI0ZcqUQj88WbJkie12NK1atVJkZGSBMe3bt1ft2rV14sQJbdy4Udu3b1d0dHSBcdnZ2frwww/z1WCPlbUBhSHTAAAAf0CeIc/AN5FnAACAr3L61p/S1UtCf/HFFwoNDZV0NdycPn1aK1as0IwZM/Tee+9pxowZ+vbbb3Xq1CnbJWYrVaqkL774wu63UgB4nxUrVqhFixbq3bu3du/eLdM0Vbt2bU2dOlUHDhxwd3kAAACF8pRMM3PmTG3YsKHIbztnZ2frnXfe0UcffWR77tlnny0wLiYmRvXr15ckLVu2TNOnTy8w5uDBgxo6dKjt8YQJE+zuMzAwUKNHj7Y97t+/v06dOlVg3MiRI7Vz505JVz9s6ty5s935rKwNsBKZBgAAeCPyzFXkGfg78gwAAPAGTl9RLdcjjzyili1baty4cfrXv/6ljIyMQsNISEiIHnnkEY0ZM0YNGza0qgQAbhIXF6fRo0dr27Ztkq7+R0h4eLheffVVDRkyROXLl3dzhQAAAMXzhEyTkJCgIUOGqF69eurUqZNuvvlmhYeHq1y5ckpNTdXu3bu1dOlSJScn27YZNWqU3W/wly9fXnPnzlXXrl2VmZmpYcOGadWqVerRo4dCQ0O1fft2zZkzR+fOnZMkxcbGqmPHjoXWFhsbq8WLFysuLk579uzRrbfeqtjYWEVFRenMmTNasGCB4uPjJUlVqlTRxx9/XOhcVtcGOItMAwAAvB15hjwD/0WeAQAA3sSyRjVJatSokT777DN9+OGH+u6777Rjxw6lpKTo/PnzqlSpkmrVqqXmzZurdevWqlKlipW7BuAGmzZt0ptvvqktW7ZIuhp+atSooZdfflnDhg3jlr4AAMDreEqmOXLkiObNm1fkmCpVqmjixIkaMmRIoWM6duyohQsXKiYmRqmpqVq+fLmWL19eYFxsbKxmzJhR5P6CgoK0aNEi9e3bV8uXL9eJEyfsXhWgbt26WrhwoZo2bVrkfFbWBpQWmQYAAPgS8kzhyDPwReQZAADgjSxtVMtVpUoV3X///br//vtdMT0AN/vuu+80evRobdiwQdLV8FO1alWNGDFCL7zwgipVquTmCgEAAJzjrkwzbdo0PfbYY9q0aZMSEhJ07NgxnTp1Sunp6QoNDVVERIRuueUWde7cWX369HHow6WePXuqVatWmjFjhpYtW6bk5GRdvnxZderUUdu2bfXUU0/ZvYKBPWFhYVq2bJmWLl2q+fPnKzExUadOnVJYWJgaN26sXr16afDgwQ5/6GVlbUBJkGkAAIAvI8/YR56BryDPAAAAb+aSRjUAvikxMVGjR4/WmjVrJF0NP2FhYXrhhRc0YsQIrpQIAADgpEqVKqlz587q3LmzpfPWqVNH48eP1/jx4y2Z78EHH9SDDz5oyVxW1wYUhUwDAADgOuQZwLXIMwAAwBfQqAagWDt37tSYMWNsly43TVOhoaEaNmyYXn75ZVWvXt3NFQIAAABA4cg0AAAAALwVeQYAAPgSGtUAFGrPnj0aM2aMFi9eLOlq+KlQoYKGDBmikSNHqmbNmm6uEAAAAAAKR6YBAAAA4K3IMwAAwBfRqAbArr59++pf//qXTNOUaZoKCQnR4MGDNXLkSNWuXdvd5QEAAABAkcg0AAAAALwVeQYAAPgqGtUA2PXVV1/ZlmvVqqXnn39edevW1Zo1a0o9Z//+/a0oDQAAAACKRaYBAAAA4K3IMwAAwFfRqAagUIZhSJJOnz6tN9980+m5CEEAAAAAyhKZBgAAAIC3Is8AAABfRKMagEKZpunuEgAAAACg1Mg0AAAAALwVeQYAAPgiGtUA2LVhwwZ3lwAAAAAApUamAQAAAOCtyDMAAMBX0agGwK727du7uwQAAAAAKDUyDQAAAABvRZ4BAAC+KsDdBQAAAAAAAAAAAAAAAAAAfBuNagAAAAAAAAAAAAAAAAAAl3L61p/R0dGSpAoVKmjjxo0KDg52uigAAAAAKCtkGgAAAADeijwDAAAAwJs4fUW1Xbt2adeuXapevToBCAAAAIDXIdMAAAAA8FbkGQAAAADexOlGtRo1akiSrrnmGqeLAQAAAICyRqYBAAAA4K3IMwAAAAC8idONanXr1pUknTt3zuliAAAAAKCskWkAAAAAeCvyDAAAAABv4nSjWteuXWWapr777jsr6gEAAACAMkWmAQAAAOCtyDMAAAAAvInTjWoxMTEqX768jh49qnnz5llREwAAAACUGTINAAAAAG9FngEAAADgTZxuVGvUqJH+8Y9/yDRNDR06VF999ZUVdQEAAABAmSDTAAAAAPBW5BkAAAAA3iTI2QkOHz6sLl266N1339Vrr72mxx9/XB988IEeffRR3X777apVq5YqVKjg0Fz169d3thwAAAAAKBEyDQAAAABvRZ4BAAAA4E2cblSLjIyUYRi2x6Zpatu2bdq2bVuJ5jEMQ1lZWc6WAwAAAAAlQqYBAAAA4K3IMwAAAAC8idONarlM05RhGLZAZJqmVVMDAAAAgMuRaQAAAAB4K/IMAAAAAG/gdKNa/fr1831bBwAAAAC8CZkGAAAAgLcizwAAAADwJk43qiUnJ1tQBgAAAAC4B5kGAAAAgLcizwAAAADwJgHuLgAAAAAAAAAAAAAAAAAA4NtoVAMAAAAAAAAAAAAAAAAAuBSNagAAAAAAAAAAAAAAAAAAlwpy1cS7d+/WkSNHdPbsWWVlZal///6u2hUAAAAAWI5MAwAAAMBbkWcAAAAAeCJLG9V+++03vfvuu1qwYIHOnTuXb92fQ9DJkyc1fPhwmaapFi1a6JVXXrGyFAAAAAAoMTINAAAAAG9FngEAAADg6SxrVFuwYIEGDx6sCxcuyDTNfOsMwygwPiIiQqdPn9aGDRu0YsUKPfvss6pUqZJV5QAAAABAiZBpAAAAAHgr8gwAAAAAbxBgxSSLFi1Sv379bAGoatWquv/++9WkSZMit3v66aclSZcuXdKqVausKAUAAAAASoxMAwAAAMBbkWcAAAAAeAunG9VSU1MVGxsr0zRlGIZGjx6t48eP69tvv1WnTp2K3LZHjx4KDg6WJK1bt87ZUgAgH9M0FR8fr5dfflmtWrVS1apVVa5cOV1zzTXq3bu3NmzY4O4SAQCAByDTAPBUZBoAAFAc8gwAT0WeAQAA9jjdqDZr1iylpqbKMAyNGTNGY8eOVUhIiEPbhoaGKioqSqZpateuXc6WAgD5rF+/Xu3atdN7772nxMRERUREqFmzZkpPT9c333yjDh066M0333R3mQAAwM3INAA8FZkGAAAUhzwDwFORZwAAgD1ON6qtWLFCklSzZk29+uqrJd4+99LThw4dcrYUAMjHNE1dd911+uijj3T69Gnt379f27dv1x9//KFRo0ZJkt566y0tX77czZUCAAB3ItMA8FRkGgAAUBzyDABPRZ4BAAD2ON2otn//fhmGobZt26pcuXIl3r5atWqSpHPnzjlbCgDkc8cdd+jnn3/WkCFDbP/WSFK5cuX09ttv6/7775ckzZ49210lAgAAD0CmAeCpyDQAAKA45BkAnoo8AwAA7HG6Ue3MmTOSpPDw8FJtf+XKlauFBDhdCgDkU7lyZQUFBRW6vlOnTpKkAwcOlFVJAADAA5FpAHgqMg0AACgOeQaApyLPAAAAe5xOHlWqVJEkpaenl2r7w4cPS5Jq1KjhbCkAUCKXL1+WJFWoUMHNlQAAAHci0wDwVmQaAABAngHgrcgzAAD4p8Lb2B107bXX6vTp09q5c2eJt7148aISEhJkGIaaNGnibCkA4DDTNPXvf/9bktSmTZsix2ZkZCgjI8PuurS0NElSZmamMjMzLaktd57gCk7/Ew0XKs3rHVwh2Pa6Fvb6lnZeZ+eANXKPPa+B7+G19RyueA3INAC8kadmmtw5ygUaTs+Fkss97o4c/9K83sXNW9r30J/n9dZzLs4Z3Y/XwL04/sUjzwDAVZ6aZ3LnkqRyJleadIfc417c8S9VnjEDix1jxbzefC7E+Zx7cfzdi+PvGCuOj9NdEPfcc4927dqln3/+WTt27FDz5s0d3nbatGm6ePGiDMNQhw4dnC0FABw2e/Zs7dixQ+XKldMLL7xQ5NiJEydq3LhxRY5Zs2aNKlasaGGFUszc3pbOB2utWLGixNsM/ucjtuXCXl9n5y3tHLBWXFycu0uAi/Daut/Fixctn5NMA8AbeXqmebFDpGVzoeQcOf6lyQ2jOjW0fE5783p7puGc0f14DdyL41848gwAXOXpeUaSns+MtnQ+lExxx780meFltSh2jBXzenuekTifczeOv3tx/ItmRaZxulGtb9+++sc//iFJGjx4sDZu3OjQicC6des0evToq0UEBalfv37OlgIADtm+fbuef/55SdJbb72lxo0bFzl+1KhRGjFihN11aWlpqlevnu677z5VrlzZkvoyMzMVFxeneU8tUualLEvmRNlYmvpZvscPVh1QYExwhSDFzO3t0tf3z3XYq8XeGFdw137dJffvb6dOnRQcHFz8BvAavLaeI/ebslYi0wDwNp6caXJ/Z05Zn6wr2abT86FkygUaerFDZKmO/+a/D833uN1fp5d4/3+ew6p57M1hb1+u8Od9F7Vfzhndj9fAvTj+xSPPAIBn5xnp/36ffRC8XVeMHEvmhOPKmQF6PjO6VMd/Q8fR+R7fs3Z8iffvijnszWNvjCvYq7+ofXM+514cf/fi+DvGikzjdKNaixYt1KtXL33zzTf68ccfddddd2nq1Km688477Y4/duyY/vGPf+iDDz5QZmamDMPQ008/rfr16ztbCgAUKykpSQ888IAuX76svn376qWXXip2m5CQEIWEhBQ5Jjg42PJfWJmXspR5iUuLepM/vweKev1c+fraey/+eV9ldYLlrv26myv+TYBn4LV1P1ccfzINAG/iLZnmSrZJo5obleb4//n1L83rZ+89ZMU89uYoq3OyP+/bkf1yzuh+vAbuxfEvHHkGgL/zljwjSVeMHF0xsi2dE44rzfEvkCNK8fq5Yg5785RZnrFTP5nG83H83YvjXzQrjo3TjWqSNGfOHO3du1f79u3Tjh071LZtW1177bUyzf/7j5zevXvr4MGD2rNnjyTZ1jVv3lzvv/++FWUAQJFOnDihTp066fjx4+rWrZs+/fRTGYbh7rIAAIAHINMA8AZkGgAAYA95BoA3IM8AAABJCrBikqpVq2rDhg26++67ZZqmTNPU0aNHdezYMdsJxpIlS7Rnzx7beknq0KGD1qxZU2wXPAA468yZM+rUqZN+/fVXtW/fXv/+97/phAYAADZkGgCejkwDAAAKQ54B4OnIMwAAIJcljWqSFBERoXXr1umzzz7TbbfdZgs7f/6RpJtuukmffvqp1qxZo+rVq1tVAgDYdf78eXXt2lW7d+9Wy5YttWzZMlWoUMHdZQEAAA9DpgHgqcg0AACgOOQZAJ6KPAMAAPKy5NafuQzD0BNPPKEnnnhCJ06c0NatW3Xs2DGdO3dOoaGhioiIUKtWrRQZGWnlbgGgUBkZGXrwwQf1/fffq2nTplq1apXCwsLcXRYAAPBQZBoAnoZMAwAAHEWeAeBpyDMAAODPLG1Uy6t27drq2bOnq6YHgGJlZ2fr0Ucf1fr169W4cWPFxcXxDUEAAOAwMg0AdyPTAACA0iLPAHA38gwAALDHZY1qAOBu//rXv7RkyRJJUkBAgPr06WN3XJ06dfTvf/+7DCsDAAAAgOKRaQAAAAB4K/IMAACwh0Y1AD4rIyPDtnzw4EEdPHjQ7rgGDRqUVUkAAAAA4DAyDQAAAABvRZ4BAAD2uKRR7aefftKGDRu0Y8cOpaSkKD09XWFhYapZs6aio6N1zz336Oabb3bFrgHAZuDAgRo4cKC7ywAAAF6ITAPAE5BpAABAaZBnAHgC8gwAALDH0ka1TZs26dVXX1ViYmKhYz7//HNJUsuWLTVx4kTdc889VpYAAAAAAKVGpgEAAADgrcgzAAAAADxdgFUTjRw5Uh06dFBiYqJM0yz2Z9u2berYsaNeeeUVq0oAAAAAgFIj0wAAAADwVuQZAAAAAN7Akiuqvf7663r33Xdtjw3D0J133qk777xT9erVU2hoqC5cuKDDhw8rISFBW7dutYWhKVOmKDAwUBMnTrSiFAAAAAAoMTINAAAAAG9FngEAAADgLZxuVNu+fbsmTZokwzBkmqY6d+6sDz/8UNdff32h2xw4cEDDhw/XmjVrZJqm3nvvPfXp00fR0dHOlgMAAAAAJUKmAQAAAOCtyDMAAAAAvInTt/6cMWOGcnJyJEmPPfaYVqxYUWQAkqQmTZpo5cqVevTRRyVJOTk5+uijj5wtBQAAAABKjEwDAAAAwFuRZwAAAAB4E6cb1davXy9JCg0N1cyZM2UYhkPbGYahmTNnqlKlSpKkDRs2OFsKAAAAAJQYmQYAAACAtyLPAAAAAPAmTjeqHT9+XIZh6J577lFYWFiJtq1cubI6dOgg0zR1/PhxZ0sBAAAAgBIj0wAAAADwVuQZAAAAAN4kyNkJKlWqpIyMDNWuXbtU20dERNjmARzVKaCPbTku599urAQAAADejkwDAAAAwFuRZwAAAAB4E6evqNawYUNJ0smTJ0u1fe52kZGRzpYCAAAAACVGpgEAAADgrcgzAAAAALyJ041qvXv3lmmaWr9+vdLT00u0bVpamtavXy/DMNSrVy9nSwEAAACAEiPTAAAAAPBW5BkAAAAA3sTpRrXY2Fhdc801unDhgoYOHVqibYcOHarz58/rmmuu0aBBg5wtBQAAAABKjEwDAAAAwFuRZwAAAAB4E6cb1apVq6ZFixapatWq+vLLL9WjRw8lJycXuU1SUpK6d++uL7/8UtWrV9fXX3+t6tWrO1sKAAAAAJQYmQYAAACAtyLPAAAAAPAmQY4Mmj9/frFjnn/+eb399tv69ttvtWLFCrVp00atWrVS/fr1VbFiRV28eFGHDx/W1q1b9d1338k0TYWEhOj555/X/v37tX//fvXv39/pPxAAAAAA/BmZBgAAAIC3Is8AAAAA8BUONaoNHDhQhmE4PGlOTo7i4+MVHx9vd71pmjIMQ1euXNHYsWMlSYZhEIIAAAAAuASZBgAAAIC3Is8AAAAA8BUONapJV4NLSRQ3vqTzAQAAAIAzyDQAAAAAvBV5BgAAAIAvcKhRbcCAAa6uAwAAAABchkwDAAAAwFuRZwAAAAD4Coca1T755BNX1wEAAAAALkOmAQAAAOCtyDMAAAAAfEWAuwuwWnp6uhYtWqRhw4apdevWqlWrloKDg1W5cmXdeOON6t+/v1atWlWiy1ofP35cY8aMUXR0tKpXr66KFSuqUaNGGjBggDZt2lTiGpcuXaqHH35YDRo0UPny5RUeHq4777xT7777rs6dO1eiuayuDQAAAAAAAAAAAAAAAACs5tAV1bzF+++/r9dff12XL18usC49PV379+/X/v379fnnn6tdu3b64osvVL9+/SLnXLx4sWJiYpSamprv+aSkJCUlJWn+/PmKjY3VjBkzFBgYWORc6enp6tu3r5YvX57v+ZSUFKWkpCghIUFTp07VwoUL1bp162L/vFbWBgAAAAAAAAAAAAAAAACu4lONagcOHLA1qdWtW1f33nuvWrRooVq1aunSpUv6/vvv9cUXX+j8+fPavHmz7r77biUkJCg8PNzufGvXrtUjjzyizMxMSVK3bt3Uo0cPhYaGavv27ZozZ47S0tI0e/ZsSdKsWbMKrS0rK0u9e/dWXFycJCkiIkKxsbGKiorSmTNntGDBAm3ZskW///67unXrpvj4eDVt2rTQ+aysDQAAAAAAAAAAAAAAAABcyaca1QzD0H333aeXXnpJ9957rwIC8t/ZdODAgRo5cqQ6d+6s/fv3KykpSSNHjtS8efMKzHX58mU99dRTtkawqVOnatiwYbb1jz/+uAYPHqz27dvrxIkTmj17tvr06aNOnTrZrW327Nm2JrWoqCitX79eERERtvVDhw7VSy+9pClTpig1NVWDBw9WfHy83bmsrg0AAAAAAAAAAAAAAAAAXMnyRrWTJ0/qxx9/VFJSktLS0mzNVI4YPXq0U/v+29/+purVqxc5pkGDBlq4cKFuu+02SdLChQs1bdo0VaxYMd+4uXPn6vDhw5Kk7t2752sEy9WkSRNNnz5dvXv3ttVvrxksOztb48ePtz3+/PPP8zWp5Zo0aZLWrVunnTt3asuWLVq9erU6d+5cYJyVtQEAAADIz52ZBgAAAACcQZ4BAAAA4Mksa1T77rvv9Oabb2rTpk0yTbNUczgbgoprUst166236sYbb9S+fft08eJF/fLLL7rlllvyjVm4cKFtecSIEYXO9dBDDykyMlLJyclKSEhQcnKyIiMj843ZtGmTTpw4IUlq3769oqOj7c4VGBio4cOHKyYmRpK0YMECu41qVtYGAAAA4CpPyDQAAAAAUBrkGQAAAADeIKD4IcX7xz/+oXbt2mnjxo3KycmRaZol/ilrYWFhtuVLly7lW5eenq4tW7bYxrVr167QeQICAtSlSxfb45UrVxYYs2rVKtty165di6wr73p7c1ldGwAAAADvzDQAAAAAIJFnAAAAAHgPp6+oFh8frxdffNEWZIKCgtS6dWs1a9ZM1apVU1CQ5XcXdVpGRoYOHDhge9ygQYN86/fu3aucnBxJUvPmzRUYGFjkfC1bttTMmTMlSbt37y6w/qeffso3tigRERGqV6+ejhw5olOnTiklJUW1atVyWW0AAACAv/PGTAMAAAAAEnkGAAAAgHdxOqG8//77Mk1ThmHo7rvv1meffaa6detaUZvLLFiwQOfOnZMkRUdHq3bt2vnW79+/37bcsGHDYufLOybvts7Md+TIEdu2eRvVrK4NAAAA8HfemGkAAAAAQCLPAAAAAPAuTjeqbd26VZJUo0YNLVmyJN8tNT1RSkqKXnnlFdvjN954o8CY1NRU23LNmjWLnbNGjRp2t3XFfFbX9mcZGRnKyMiwuy4tLU2SlJmZqczMzGLnskLufjIzMxVcIbjIMYA78P4DAMA97J2Tlvb3srdlGgAAAADIRZ4BAAAA4E2cblRLTU2VYRi69957PT4AXblyRb1791ZKSook6aGHHlLPnj0LjDt//rxtuXz58sXOW6FCBdtyenq6S+ezurY/mzhxosaNG1fkmDVr1qhixYrFzmWluLg4Df7nI3bXrVixokxrAfK6ePGiu0sAAMAv2TsnLe3vZW/KNAAAAACQF3kGAAAAgDdxulHt2muvVVJSkipXrmxFPS6Tk5OjmJgYbd68WZLUuHFjzZs3r9jtDMNwdWml5oraRo0apREjRthdl5aWpnr16um+++4rs9c7MzNTcXFx6tSpkx6u9bTdMUtTPyuTWgB7cq80CAAAypa9c9LS/l72lkwDAAAAAH9GngEAAADgTZxuVLvtttt06NAhHTp0yIp6XMI0TT3zzDP68ssvJUn169fX2rVrVa1aNbvjK1WqZFu+dOlSsfPnHWPvG0uVKlXS2bNnJUmXL1/ON39J57O6tj8LCQlRSEhIkWOCg4MVHGz/NpyuEhwcrMxL9m/lVNa1AHnx/gMAwD3snZOW9veyN2QaAAAAALCHPAMAAADAmwQ4O8GQIUMkSfHx8Tp69KjTBVnNNE09++yzmj17tiSpbt26Wr9+vSIjIwvdpmrVqrbl06dPF7uPP/74w+62rpjP6toAAAAAf+fpmQYAAAAACkOeAQAAAOBNnG5Uu/fee/XAAw8oIyNDAwYM0JUrV6yoyxKmaWro0KGaOXOmpKuXwN6wYYMaN25c5HY33HCDbTkpKanY/eQdk3dbV8xndW0AAACAv/PkTAMAAAAARSHPAAAAAPAmTjeqSdJXX32l++67T+vXr1fLli21YsUK5eTkWDF1qeU2qc2YMUOSdM0112jDhg267rrrit02KipKAQFXD82OHTuUnZ1d5PjExETbcrNmzQqsv/nmm+2OtefkyZM6cuSIJCk8PFy1atVyaW0AAAAAPDPTAAAAAIAjyDMAAAAAvEWQFZNUrFhRK1eu1IQJEzR27Fh1795dFSpUUJMmTVSlShUZhlHsHIZhaN26dVaUU6BJrU6dOtqwYYOuv/56h7YPCwtTmzZttHnzZqWnpys+Pl7t27e3OzYnJ0erV6+2Pb7//vsLjOnSpYsmT54sSVq5cqVeeeWVQve9YsUK23LXrl1dXhsAAAAAz8s0AAAAAOAo8gwAAAAAb2FJo5okTZo0SR988IEMw5Bpmrp48aJ27drl0LamaToUlBw1bNgwW5Na7dq1tWHDBjVp0qREczz66KPavHmzJGnKlCmFNoMtWbLEdnvNVq1aKTIyssCY9u3bq3bt2jpx4oQ2btyo7du3Kzo6usC47Oxsffjhh/lqcHVtAAAAAK7ypEwDAAAAACVBngEAAADgDSy59eeTTz6p119/XampqfmeN03ToR8rPffcc/roo48k/V+T2g033FDieWJiYlS/fn1J0rJlyzR9+vQCYw4ePKihQ4faHk+YMMHuXIGBgRo9erTtcf/+/XXq1KkC40aOHKmdO3dKktq0aaPOnTu7vDYAAAAAnpVpAAAAAKAkyDMAAAAAvIXTV1T7/PPP9dlnn9m+pXPTTTepT58+atasmapVq6agIMsu2lasN954Q9OmTZN09TLVzz//vPbt26d9+/YVuV10dLSt8StX+fLlNXfuXHXt2lWZmZkaNmyYVq1apR49eig0NFTbt2/XnDlzdO7cOUlSbGysOnbsWOg+YmNjtXjxYsXFxWnPnj269dZbFRsbq6ioKJ05c0YLFixQfHy8JKlKlSr6+OOPC53L6toAAAAAf+ZJmQYAAAAASoI8AwAAAMCbOJ1QZs+ebVt++eWX9c4777jtEtG5jV7S1W8KjRo1yqHtPvnkEw0cOLDA8x07dtTChQsVExOj1NRULV++XMuXLy8wLjY21nar0cIEBQVp0aJF6tu3r5YvX64TJ07YvcpZ3bp1tXDhQjVt2rTI+aysDQAAAPBnnpRpAAAAAKAkyDMAAAAAvInTjWq7du2SYRhq1KiRTwagnj17qlWrVpoxY4aWLVum5ORkXb58WXXq1FHbtm311FNPqX379g7NFRYWpmXLlmnp0qWaP3++EhMTderUKYWFhalx48bq1auXBg8erCpVqpR5bQAAAIC/8vVMAwAAAMB3kWcAAAAAeBOnG9Wys7MlSe3atXN7ANq4caNL5q1Tp47Gjx+v8ePHWzLfgw8+qAcffNCSuayuDQAAAPA3npRpAAAAAKAkyDMAAAAAvEmAsxNce+21kkQAAgAAAOCVyDQAAAAAvBV5BgAAAIA3cbpRrX379jJNUz/99JMV9QAAAABAmSLTAAAAAPBW5BkAAAAA3sTpRrVnnnlGgYGB+vHHH/X9999bURMAAAAAlBkyDQAAAABvRZ4BAAAA4E2CnJ0gOjpa48aN0xtvvKG//OUviouLU5MmTayoDQAAr9IpoI+7S5BUujrsbROX828rynHLfuztqzT7sWIOwJ94698ZMg0AAFL0kL97xBylnefP22yf8VdLavGU/ZR2X2VVL+ALvPXvC3kGAICr7lj1mlfP8efttnV52+lavH0/9o6lq+oFfEFZ/f12ltNXVJOk1157TR999JFOnjyp6Ohovfzyy9q6davS0tKsmB4AAAAAXIpMAwAAAMBbkWcAAAAAeAunr6gWGBiY7/GVK1f0/vvv6/333y/RPIZhKCsry9lyAAAAAKBEyDQAAAAAvBV5BgAAAIA3cbpRzTRNGYZhe5x32TRNZ6cHAAAAAJci0wAAAADwVuQZAAAAAN7E6UY1ibADAAAAwLuRaQAAAAB4K/IMAAAAAG/hdKNaTk6OFXUAAAAAgFuQaQAAAAB4K/IMAAAAAG8S4O4CAAAAAAAAAAAAAAAAAAC+jUY1AAAAAAAAAAAAAAAAAIBL0agGAAAAAAAAAAAAAAAAAHApGtUAAAAAAAAAAAAAAAAAAC5FoxoAAAAAAAAAAAAAAAAAwKWCnJ0gMDDQijpkGIaysrIsmQsAAAAAHEWmAQAAAOCtyDMAAAAAvInTjWqmacowDJmmaUU9AAAAAFCmyDQAAAAAvBV5BgAAAIA3seTWn6UJQIZhyDAMK3YPAAAAAE4h0wAAAADwVuQZAAAAAN7C6Ua1nJwch36ys7N15swZbdq0Sc8++6yCg4MVEhKiTz/91LYeAAAAAMqaJ2Wa9PR0LVq0SMOGDVPr1q1Vq1YtBQcHq3LlyrrxxhvVv39/rVq1qkQfRB0/flxjxoxRdHS0qlevrooVK6pRo0YaMGCANm3aVOIaly5dqocfflgNGjRQ+fLlFR4erjvvvFPvvvuuzp07V6K5rK4NAAAA8DeelGckMg2ZBgAAACia07f+dJRhGKpataratWundu3a6amnnlLnzp315JNP6tKlSxo8eHBZlQIAAAAAJebqTPP+++/r9ddf1+XLlwusS09P1/79+7V//359/vnnateunb744gvVr1+/yDkXL16smJgYpaam5ns+KSlJSUlJmj9/vmJjYzVjxgwFBgYWOVd6err69u2r5cuX53s+JSVFKSkpSkhI0NSpU7Vw4UK1bt262D+vlbUBAAAAKFpZfEZDpil9bQAAAIC/KLNGtT9r3ry5PvvsM3Xr1k0vvPCC2rdvrxtvvNFd5QAAAABAiVidaQ4cOGD7QKdu3bq699571aJFC9WqVUuXLl3S999/ry+++ELnz5/X5s2bdffddyshIUHh4eF251u7dq0eeeQRZWZmSpK6deumHj16KDQ0VNu3b9ecOXOUlpam2bNnS5JmzZpVaG1ZWVnq3bu34uLiJEkRERGKjY1VVFSUzpw5owULFmjLli36/fff1a1bN8XHx6tp06aFzmdlbQAAAABKzhWf0ZBpyDQAAABAcZy+9acz7r//ft144426cuWKZsyY4c5SAAAAAKDErMw0hmHovvvu05o1a/Tbb7/p008/1bBhw/TII49o4MCBmjFjhnbv3q0bbrhB0tVv6Y8cOdLuXJcvX9ZTTz1l+9Bk6tSpWr58uQYNGqTHH39cU6ZMUWJiomrXri1Jmj17tu0DG3vyro+KitKuXbs0YcIEPfbYYxo6dKji4+P14osvSpJSU1OLvBqD1bUBAAAAKB2rP6Mh05BpAAAAgOK4tVFNklq0aCHTNLVmzRp3lwIAAAAAJWZVpvnb3/6m1atXq1OnTgoIsB/VGjRooIULF9oeL1y4UBcvXiwwbu7cuTp8+LAkqXv37ho2bFiBMU2aNNH06dNtj0ePHm13n9nZ2Ro/frzt8eeff66IiIgC4yZNmqTbbrtNkrRlyxatXr3a7nxW1gYAAADAOVZ+RkOmKXltAAAAgL9xe6Na+fLlJUm///67mysBAAAAgJKzKtNUr17doXG33nqr7ZY8Fy9e1C+//FJgTN4PfkaMGFHoXA899JAiIyMlSQkJCUpOTi4wZtOmTTpx4oQkqX379oqOjrY7V2BgoIYPH257vGDBArvjrKwNAAAAgHOs/IyGTFPy2gAAAAB/4/ZGtX379klSod+uAQAAAABP5o5MExYWZlu+dOlSvnXp6enasmWLbVy7du0KnScgIEBdunSxPV65cmWBMatWrbItd+3atci68q63N5fVtQEAAABwjrs+oyHTAAAAAP7Jrd1hW7Zs0ZYtW2QYhho2bOjOUgAAAACgxNyRaTIyMnTgwAHb4wYNGuRbv3fvXuXk5EiSmjdvrsDAwCLna9mypW159+7dBdb/9NNPdsfaExERoXr16kmSTp06pZSUFJfWBgAAAKD03PUZDZkGAAAA8F9uaVS7ePGiZs+erR49esg0TUlS9+7d3VEKAAAAAJSYOzPNggULdO7cOUlSdHS0ateunW/9/v37bcuOfNiUd0zebV0xn9W1AQAAACg5d39GQ6YBAADA/2vvzuO0quv+8b9nYBgQWRRZ3EFSFHdcEw0yV1BU0PK2EkXRXLs1szRzV9I7tTQlRZKs5CZzC78sYqIFhmK47yYopiiLMCDbMHN+f/jjuhlnH65rrmWez8djHp65zuec857rfIbrvJz3dS5artYbu4NDDz20wWPXrVsXixcvjvfeey/WrVuXenyLLbaIiy++eGNLAQAAaLR8yjQLFy6MSy+9NPX9FVdcUW3M0qVLq9RVny5dutS4bSb2l+7avmrNmjWxZs2aGteVlZVFRER5eXmUl5fXu690WX+sNq2Kanwc1s+FbM0JcxEAml9N16RNfU3OpzwTIdPUJ6czTVJc7TGIyG6mMRcBIDs2JtNsdKPa008/HUVFRfUP3MD6d+hERGy11VYxceLE2GyzzTa2FAAAgEbLl0yzdu3aGDZsWOqjZ44//vg44YQTqo1bsWJFarlt27b17rddu3ap5eXLl2d0f+mu7atGjRoV11xzTZ1jnnjiidhkk03q3Ve6/ejQnlW+nzRpUrPXQG6bNm1aVo67cuXKrBwXAFqymq5Jm/qanC95JkKmyfdM88PyfqlleYaaZCPTyDMAkB0bk2k2ulEtomqoaYiioqLo06dPnHLKKXHhhRdGx44d01EGAABAk+R6pqmsrIwRI0bEP/7xj4iI6N27d/zud79rUJ25KhO1XXbZZbXeCaKsrCy23XbbOOKII5o1g5aXl8e0adPilqfmxdqK/5tn/7jtvGargdy2fo4cfvjhUVJS0uzHX39nDgCg+dR0Tboxr8m5nmciZJqGyuVM8+uSObG2qDIiIqYfdmWzHZ/cl81MI88AQHZsTKbZ6Ea16dOnN3hsSUlJdOjQIXr27BkdOnTY2EMDAABstFzPNEmSxA9+8IP405/+FBER2223XTz55JO13vFg0003TS2vWrWq3v1vOKamn2nTTTeNzz//PCIiVq9eXWX/jd1fumv7qtLS0igtLa1zTElJSVaagdZWJFUa1bJRA7ktW3PTXASA5lfT635TX5NzPc9EyDR11fZVOZ1piipjbVFFqgb4qmzMTXMRALJjYzLNRjeqDRgwYGN3AQAAkDW5nGmSJIlzzz03xowZExER22yzTTz11FPRs2fPWrfp3LlzannRokX1HmPx4sU1brvhY+v/qLNo0aJ6/6hT1/7SXRsAALR0uZxnImQamQYAAKoqznYBAAAAVJckSZx33nnx29/+NiIitt5665g+fXr07t27zu369OmTWp47d269x9lwzIbbZmJ/6a4NAADIXTKNTAMAAF+lUQ0AACDHrP+DzujRoyMiYquttorp06fH1772tXq37du3bxQXfxn1XnzxxaioqKhz/OzZs1PLu+22W7X1u+++e41ja/Lpp5/G/PnzIyKiW7du0bVr14zWBgAA5CaZpmG1AQBAS6NRDQAAIId89Q86W265ZUyfPj123HHHBm3foUOH6N+/f0RELF++PGbMmFHr2MrKypg6dWrq+6OPPrramKOOOiq1PHny5DqPPWnSpNTyoEGDMl4bAACQe2SahtcGAAAtjUY1AACAHHL++een/qDTo0ePmD59euy0006N2sfJJ5+cWr7llltqHffoo4+mPormwAMPjJ49e1YbM2DAgOjRo0dERDz99NMxZ86cGvdVUVERt99+e401ZKo2AAAg98g0Da8NAABamtYNHThixIhM1hFFRUUxduzYjB4DAABoufIh01xwwQVx1113RcT//UGnT58+jd7PiBEj4qabbooPP/wwJk6cGHfeeWecd955Vca8++67VR677rrratxXq1at4sorr4xzzz03IiJOPfXUeOqpp6Jbt25Vxv30pz+Nl156KSIi+vfvH0ceeWTGawMAgJYiH/JMhEzT2NoAAKClaXCj2rhx46KoqCgjRSRJolENAADIqFzPNFdccUX85je/iYgv/0j0wx/+MN56661466236tyuX79+sd1221V5rG3btjF27NgYNGhQlJeXx/nnnx9TpkyJIUOGRPv27WPOnDlx7733xrJlyyIiYuTIkXHYYYfVeoyRI0fGI488EtOmTYvXX3899txzzxg5cmT07ds3lixZEuPHj0995E2nTp3i7rvvrnVf6a4NAABaglzPMxEyjUwDAAD1a3CjWsSXYQUAACBf5XKmWf9HkYgv67zssssatN19990Xp512WrXHDzvssJgwYUKMGDEili5dGo8//ng8/vjj1caNHDky9bE8tWndunU89NBDccopp8Tjjz8eCxYsqPGOANtss01MmDAhdt111zr3l87aAACgpcjlPBMh08g0AABQvwY3qv3kJz9J64GfeeaZmDVrVhQVFeV8uAIAAPJfS8w0J5xwQhx44IExevTomDhxYsybNy9Wr14dW265ZRx88MFxxhlnxIABAxq0rw4dOsTEiRPjsccei/vvvz9mz54dn332WXTo0CF69+4dQ4cOjbPPPjs6derU7LUBAECha4l5JkKmAQCAQtPgRrVRo0al5YDPP/98XHHFFfHcc89VuU11+/bt07J/AACAmuR6pnn66ac3srKabbnllnHttdfGtddem5b9HXfccXHcccelZV/prg0AAApVrueZCJkGAACoX3FzHeiVV16J4447Lr7+9a/H3/72t4j48tbPpaWl8d///d/x73//u7lKAQAAaDSZBgAAyFfyDAAAkAsafEe1pnr77bfjqquuir/85S+RJEnqFtIlJSVxxhlnxBVXXBFbbbVVpssAAABoEpkGAADIV/IMAACQSzLWqDZv3ry45ppr4o9//GNUVlamwk+rVq3ie9/7Xlx11VXRs2fPTB0eAABgo8g0AABAvpJnAACAXJT2RrWPP/44rrvuurjvvvuivLw8FX6KioripJNOimuvvTZ22mmndB8WAAAgLWQaAAAgX8kzAABALktbo9rChQtj1KhR8dvf/jbWrFmTCj8REUOGDInrrrsudt9993QdDgAAIK1kGgAAIF/JMwAAQD7Y6Ea1pUuXxs033xx33HFHrFy5skr4Ofzww+P666+P/fbbb2MPAwAAkBEyDQAAkK/kGQAAIJ80uVFtxYoVcdttt8Wtt94aZWVlVcLPIYccEtdff30ccsghaSkSAAAg3WQaAAAgX8kzAABAPmp0o9rq1avjN7/5Tdx0002xZMmSKuFnv/32i+uuuy6OOOKItBYJAACQLjINAACQr+QZAAAgnzW4Ua28vDzuvvvuGDVqVCxYsKBK+Nl9993juuuuiyFDhmSkSAAAgI0l0wAAAPlKngEAAApBgxvVdtxxx5g/f36V8NOnT5+45ppr4tvf/nZGigMAAEgXmQYAAMhX8gwAAFAIGtyo9uGHH0ZRUVFERBQVFcUBBxwQ3//+92Pp0qVxzz33pKWYs846Ky37AfiqSZMmxa233hpz5syJNWvWRJ8+feL000+P8847L4qLi7NdHgDQDGQaIJ/JNADQsskzQD6TZwCA9RrcqLbe+iD03HPPxXPPPZfWYoQgIBN+8YtfxGWXXRYRETvssENsuumm8fLLL8eFF14YTz75ZDzyyCOCEAC0IDINkG9kGgBgPXkGyDfyDACwoUa/6idJkpEvgEz45z//GZdffnkUFxfHAw88EP/+97/j5Zdfjjlz5kT37t3jr3/9a9x6663ZLhMAaEYyDZBPZBoAYEPyDJBP5BkA4KsafEe1b3zjG6l36gDki+uvvz6SJImRI0fGf/3Xf6Ue33PPPePWW2+N7373u/GLX/wifvjDH0ZJSUkWKwUAMk2mOW9CzwAAaiJJREFUAfKRTAMARMgzQH6SZwCAr2pwo9rTTz+dwTIA0q+srCyefPLJiIg444wzqq0/6aST4pxzzonFixfH9OnT44gjjmjuEgGAZiTTAPlGpgEA1pNngHwjzwAANfGB30DBevHFF2Pt2rXRtm3b6NevX7X1JSUlsd9++0VExHPPPdfc5QEAANRJpgEAAPKVPAMA1KTBd1QDyDfvvvtuRERst9120bp1zf/c7bDDDvG3v/0tNbYma9asiTVr1tS4btmyZRERsWTJkigvL9/Iir9UXl4eK1eujCitjEjSsktySWllTpzfxYsXV32gbQPGNEUN+83IcRpw7CYdp5H7WP/7u3jxYreqLzDObQOl4/euHsuXL4+IiCTxIgkUvnzMNOtfM4vXrYniSv9WN7fipKjFPP9fvc4orlhT5/qm+up+66ojndeMNR23KT9Tpp6XXOW6Pbvy/flvjt8XeQZoSfIxz0RskGnKy6O4qDIt+6ThipPKWFle+M9/TdcZxV+sq3dMY311n/XVktZMk4afp6b6CznT5Pv1dL4rhOc/E/+OfFU6Mo1GNaBgff755xERsdlmm9U6Zv269WNrMmrUqLjmmmvqPFavXr2aUCEt0qqIyadMyHYVscUWf0nLmHTIp+M0V61QKDL5O7N8+fLo1KlTxvYPkAtkGprin3dnu4LmscWYyzZqfXPVkWvHas56Id9l8vdFngFaAnmGpvpHtgtoBlvE/6RlTDrk23Gaq14oBJn8fdmYTKNRDShYq1evjoiINm3a1DqmtLQ0IiJWrVpV65jLLrssLr744hrXVVZWxpIlS6JLly6xbt26eOKJJ+KII47YqC7rsrKy2HbbbWP+/PnRsWPHJu+H3JQL57e8vDwtc5XqcuH8FpJcmqvObe5IkiSWL18eW221VbZLAci45s40RUVFG1Htl7xmZkZDr4s8/9mVzec/l66ds8nvQO2aY454/usnzwAtST7mmQivZ5nQmOsQz392Zev5l2e+ZP7XLdPzxPPfMOnINBrVgILVtu2Xnz+2du3aWsesv110u3btah1TWlqaCks16dy5c0R8+eK4ySabRMeOHdPy4tixY0cvggUsm+c33XOV6vz+pkcuzlXnNje48wDQUjR3pkknr5np1djrIs9/dmXj+c/Fa+ds8jtQXXPOEc9/3eQZoKXI5zwT4fUsnZpyHeL5z67mfv7lmarM/5o11zzx/NdvYzNNcZrqAMg5DblldENuPQ0AAJANMg0AAJCv5BkAoCYa1YCCteOOO0ZExIcffhjr1q2rccz7779fZSwAAECukGkAAIB8Jc8AADXRqAYUrL333jtKSkpi9erVMWfOnGrry8vLY/bs2RERccABBzR3eQAAAHWSaQAAgHwlzwAANdGoBhSsjh07xmGHHRYREWPHjq22/sEHH4yysrLo0qVLDBw4sJmrAwAAqJtMAwAA5Ct5BgCoiUY1oKD97Gc/i6Kiorj33ntj/PjxqcdffvnluPjiiyMi4tJLL402bdpkq0QAAIBayTQAAEC+kmcAgK/SqAYUtP79+8d1110XlZWVccopp0Tv3r1jzz33jH79+sWnn34agwcPjh/96EfZLrOK0tLSuOqqq6K0tDTbpZABzm9hc34Ll3MLQLbkW6bxmpldnv/s8vxnn3OQXZ5/AL4q3/JMhNezbPP8Z5fnP7s8/9nl+W8+GtWAgvezn/0sJk6cGIceemgsXrw43nvvvdh9993jV7/6VTz22GPRqlWrbJdYRWlpaVx99dVeBAuU81vYnN/C5dwCkE35lGm8ZmaX5z+7PP/Z5xxkl+cfgJrkU56J8HqWbZ7/7PL8Z5fnP7s8/82ndbYLAGgOxxxzTBxzzDHZLgMAAKBJZBoAACBfyTMAwHruqAYAAAAAAAAAAEBGuaMaAAAANLN+59yWWp4z+qIsVgIAANA4+0+5PLX8/FE3ZrESAADyjTuqAQAAAAAAAAAAkFEa1QAAAAAAAAAAAMgojWoAOWLSpElx2GGHxeabbx7t27ePfv36xR133BGVlZXZLq3FmDt3bowZMyZGjhwZe+65Z7Ru3TqKiori+uuvr3fbf/7zn3HcccdF165do127dtG3b9+47rrrYvXq1XVu9+abb8Z3v/vd2HLLLaNt27bRu3fvuOSSS2Lp0qV1bvef//wnzjrrrNh2222jtLQ0tttuuzj77LPjP//5T2N+5BYjSZKYMWNG/PjHP44DDzwwOnfuHG3atImtttoqhg0bFtOnT69ze+c39z366KNx9tlnxz777BNbbrlltGnTJjp37hwHHXRQ/PrXv461a9fWuq3zCwDpIdNsnHzKI4Um3/JCIcqn6/mW4IorroiioqJ6/w3y3ANQSOSZjSPPZI88kxtkmtwi0+S4BBpo2bJlSUQky5Yta7Zjrl27Nnn00UeTtWvXJocVnVjjF+SKDedrY40aNSqJiCQikh122CHZY489kuLi4iQikiFDhiQVFRUZqJiv+uEPf5g6Dxt+XXfddXVu98c//jFp1apVEhHJ1ltvney9995JSUlJEhHJfvvtl3zxxRc1bvfUU08l7dq1SyIi6dq1a9KvX79kk002Sc2DBQsW1Ljd66+/nmy++eZJRCSdOnVK+vXrl3Tq1CmJiKRLly7Jm2++WWe9GzNX89WTTz6ZOp/FxcXJTjvtlOy9997Jpptumnr8iiuuqHHbfDu/haQxc7V///5JRCSlpaVJr169kn333TfZeuutU+d3n332ST7//PNq2zm/hSUb16vkj2zNj/X/lh1w3m3J3j+4tcYvWrZCuTaTaTZeXXmkrnnS3NczhSif8kJt8v3fkny5ns9nDZ0jb7zxRtKmTZt6/5+I5z795Bnqk+1M03/SFcl+ky+r9gX5fh2SJPJMOsgz2SPP5AaZJvNkmtzX0OtVjWo0mEY1qFtTL6KeffbZpKioKCkuLk4eeOCB1OMvvfRS0r179yQikv/5n/9Jd7nU4LrrrkuOOeaY5Nprr00mT56cDBs2rN5Gtblz5yalpaVJRCQ333xzUllZmSRJksybNy/p06dPEhHJeeedV227srKypGvXrklEJBdeeGFq3ixatCh1MTt48OBq261bty7p27dvEhHJsGHDUhdJK1asSIYOHZpERLLHHnvUGZwL4YK/saZNm5Z87WtfS+66665kyZIlqcfXrFmTXHbZZamL1YkTJ1bZLh/PbyFpzFy97777kunTp1cb+89//jPZZpttkohIzj333CrrnN/C4w871CXbf9TRqEZtCuHaTKZJj7rySG3zpLmvZwpVvuSFuuT7vyX5cD2f7xoyRyorK5NDDjkkad++fXLooYfW+v9EPPeZIc9Qn2xnGo1q1Cbfr0PkmfSQZ7JHnskNMk3myTS5T6MaaadRDerW1IuoQYMGJRGRnHXWWdXW/elPf0oivrzLTj5fnOWr4cOH19uodu655yYRkRxxxBHV1s2cOTOJiKSkpKRa9/zNN9+cRESyyy67JOvWrauy7oMPPkhat26dRETyr3/9q8q6P//5z6k5UVZWVmVdWVlZ0qVLlyQikocffrjWmgvhgr+xli1blpSXl9e6/uijj069O25D+Xh+C0m65ur653Wrrbaq8rjzW3j8YYe6ZPuPOhrVqE0hXJvJNJmxYR6pbZ409/VMocqXvFCXQvi3pDa5cj2f7xoyR8aMGZNERHLTTTfV+f9EPPeZIc9Qn2xnGo1q1Cbfr0PkmcyQZ5qPPJP7ZJr0kGlyX0OvV4sDgKwpKyuLJ598MiIizjjjjGrrTzrppOjYsWMsXry43s+Qp/klSRKPPPJIRNR8/g466KDYeeedo7y8PB577LEq6x5++OGIiDjttNOiVatWVdZtt912cdhhh0VExF/+8pcat/v2t78dHTp0qLKuQ4cOcdJJJ0VExIMPPtjUH6sgdezYMVq3bl3r+sMPPzwiIt55553UY85v4dh5550jImLlypWpx5xfAEgPmSZ7snE9U6jyJS+0VLlyPV/oFi5cGD/5yU+ib9++cdFFF9U6znMPQCGRZ7LHNUX6yDO5T6ZpHjJN/tCoBpBFL774Yqxduzbatm0b/fr1q7a+pKQk9ttvv4iIeO6555q7POrx4YcfxieffBIREf37969xzPrHNzx/69ati3/961+N3i4iYtasWU3ajrqtXr06IiLatWuXesz5LRz//Oc/IyKq/Dvr/AJAesg02ZON65mWKlfyQkuVK9fzhe6iiy6KJUuWxF133RUlJSW1jvPcA1BI5JnscU3RfOSZ7JNpmodMkz80qgFk0bvvvhsRX3ZY1/Zuhx122KHKWHLH+nNSWloaW221VY1jajp/8+bNi/Ly8irrG7Ld2rVr48MPP2zQdhseg7olSZK6g9WGF5bOb36rqKiIjz76KO6666645JJLon379jFq1KjUeucXANJDpsme5r6eaalyJS+0NLl2PV/o/va3v8Wf/vSn+N73vhcDBgyoc6znHoBCIs9kj2uK5iHPZI9M07xkmvyiUQ0giz7//POIiNhss81qHbN+3fqx5I7156Rz585RVFRU45iazt+Gy7Wd+5q2W7ZsWVRWVjZou8rKyigrK2vQz9HSjRkzJl588cVo06ZN/Pd//3fqcec3P/3qV7+KoqKiaN26dWy77bZx3nnnxbe+9a2YNWtW7L///qlxzi8ApIdMkz3NfT3TUuVKXmgpcvV6vpCtXr06fvCDH0SnTp3il7/8Zb3jPfcAFBJ5JntcUzQPeab5yTTNT6bJPxrVALJo/e1227RpU+uY0tLSiIhYtWpVs9REwzX1/K3frq5tN3a7r25LzebMmRM//OEPIyLi+uuvj969e6fWOb/5aeutt47+/fvH/vvvH927d4+IiOnTp8f48eOjoqIiNc75BYD0kGmyp7mvZ1qiXMoLLUWuXs8Xsuuvvz7ee++9uOGGG1LPeV089wAUEnkme1xTZJ48kx0yTfOTafKPRjWALGrbtm1EfPmRcLVZs2ZNRFT97HhyQ1PP3/rt6tp2Y7f76rZUN3fu3DjmmGNi9erVccopp8Qll1xSZb3zm59OOumkmDFjRjz33HOxYMGCmDVrVvTs2TNuvPHGOP/881PjnF8ASA+ZJnua+3qmpcm1vNBS5Or1fKF6880343/+53+iX79+cc455zRoG889AIVEnske1xSZJc9kj0zTvGSa/KRRDSCLGnL7z4bceprsWH9Oli5dGkmS1DimpvO34XJt576m7Tp16hTFxcUN2q64uDg6duzYoJ+jJVqwYEEcfvjh8cknn8TgwYNj3Lhx1W7v6/wWhgMOOCAmTZoUpaWlcc8998QHH3wQEc4vAKSLTJM9zX0905LkYl5oqXLler5QnXvuubFu3boYPXp0Kq/Ux3MPQCGRZ7LHNUXmyDO5RabJLJkmP2lUA8iiHXfcMSIiPvzww1i3bl2NY95///0qY8kd68/JmjVr4uOPP65xTE3nr2fPnlFSUlJlfUO2a9OmTWy33XYN2m7DY1DVkiVL4vDDD49///vfMWDAgHjwwQdrfK6c38Kx1VZbxV577RWVlZXx8ssvR4TzCwDpItNkT3Nfz7QUuZoXWrJcuJ4vVC+++GIUFRXFkCFDokePHlW+JkyYEBERN910U/To0SP222+/iPDcA1BY5JnscU2RGfJMbpJpMkemyU8a1QCyaO+9946SkpJYvXp1zJkzp9r68vLymD17dkR82XFPbtluu+2iR48eERExc+bMGsesf3zD89e6devo169fo7fb8PvGbseXVqxYEYMGDYrXXnst9ttvv5g4cWKtt991fgvL+v/RtP6/zi8ApIdMkz3ZuJ4pdLmeF1qyXLieL1QVFRXx6aefVvtavXp1RHz5e/Hpp5/GwoULI8JzD0BhkWeyxzVF+skzuU2myRyZJv9oVAPIoo4dO8Zhhx0WERFjx46ttv7BBx+MsrKy6NKlSwwcOLCZq6M+RUVFccIJJ0REzefv2WefjbfeeitKSkpiyJAhVdYNHTo0IiLGjRsXFRUVVdZ9+OGH8eSTT0ZExLBhw2rc7s9//nMsX768yrrly5fHgw8+GBERJ554YlN/rIK1Zs2aOO644+K5556LXXfdNaZMmRIdOnSodbzzWzjmzZuXepfSnnvuGRHOLwCki0yTPdm4nilk+ZAXWqpcuZ4vROs/6qamr+HDh0dExHXXXRdJksS8efMiwnMPQGGRZ7LHNUV6yTO5TabJHJkmTyXQQMuWLUsiIlm2bFmzHXPt2rXJo48+mqxduzY5rOjEGr8gV2w4XxtjxowZSVFRUVJcXJw88MADqcdfeumlpHv37klEJDfddFO6y6UBhg8fnkREct1119U65v3330/atGmTRERy8803J5WVlUmSJMm8efOSPn36JBGRnHPOOdW2W7ZsWbLFFlskEZFceOGFqXmzaNGipH///klEJEcffXS17datW5fsvPPOSUQkw4YNS7744oskSZJkxYoVybBhw5KISHbbbbekoqKi1pqbOlfz2bp165Ljjz8+iYikd+/eyccff9yg7fLx/BaShs7VF154IbnyyiuTf//739XWTZ48OfWcDho0qMo657fwZON6lfyRrfmx/t+yA867Ldn7B7fW+EXLVgjXZjJNZmyYR2qbJ819PVOo8iUv1CWf/y3Jl+v5fNfYOVLX/xPx3GeGPEN9sp1p+k+6Itlv8mXVviCfr0OSRJ7JFHmm+cgz2SfTNA+ZJvc19HpVoxoNplEN6rYxF1HXX399EhFJRCQ77LBDssceeyTFxcVJRCSDBw9O1q1bl4GK+aoZM2YkXbp0SX2VlpYmEZFssskmVR7/8MMPq2z3+9//PnW+tt5662TvvfdOSkpKkohI9tlnn2TFihU1Hu/JJ59M2rZtm0RE0rVr12SfffZJNtlkkyQikp49eyaffPJJjdu9+uqryWabbZZERNKpU6dkn332STp16pRERLL55psnr7/+ep0/Z75f8DfFAw88kPod23HHHZP+/fvX+HXiidVfV/Lt/BaShs7V6dOnp85vjx49kn333TfZY489ks6dO6ce32+//ZKFCxdW29b5LSz+sENdsv1HHY1q1KZQrs1kmo1XXx7p0KFDTuSRQpRPeaE2+fxvST5dz+ezdP5RJ0k895kgz1CfbGcajWrUJp+vQ9aTZzaePJM98kz2yTTNQ6bJfQ29XvXRnwA54Gc/+1lMnDgxDj300Fi8eHG89957sfvuu8evfvWreOyxx6JVq1bZLrFFKC8vj8WLF6e+1qxZExERK1eurPL4V2/leuqpp8Y//vGPOOaYY2LVqlXxxhtvxA477BBXX311zJgxI9q3b1/j8b71rW/FCy+8ECeffHIUFRXFq6++Gt27d4+LL7445syZk/p89K/abbfd4uWXX44zzzwzNt1003j11Vdj0003jZEjR8bLL78cffv2Te8TUwDWn8uIiHfffTdmzpxZ49fs2bOrbev85r4999wzfv3rX8eQIUOiffv28dZbb8Vbb70V7dq1i6OPPjruu+++ePbZZ2OLLbaotq3zCwDpIdNsvPryyPLly3MijxSifMoLhSifruf5P557AAqJPLPx5JnskWeyT6bJT5777ClKkiTJdhHkh7KysujUqVMsW7YsOnbs2CzHLC8vj0mTJsWgQYNiUOkpNY6ZVvlgs9QC9dlwvpaUlGS7HKiVuUq+MFdprGxcr5I/sjU/1v9bNmra3FhbUXP8njP6omarh9zj9Y6GME+ojzlCfcyR3CfPUJ9sZ5r/afNCrC2qqLb++aNubLZayE1eY6iPOUJ9zBEawjzJfQ29Xm3djDVBRhxefFJqWdMaAAAAAAAAAADkHh/9CQAAAAAAAAAAQEZpVAMAAAAAAAAAACCjNKoBAAAAAAAAAACQURrVAAAAAAAAAAAAyCiNagAAAAAAAAAAAGSURjUAAAAAAAAAAAAySqMaAAAAAAAAAAAAGaVRDQAAAAAAAAAAgIzSqAYAAAAAAAAAAEBGaVQDAPLSwIEDo6ioKAYOHFjj+qKioigqKoqrr766WetKh3HjxqXqnzdvXrbLAQAAMkCmAQAA8pU8AzSVRjUAKDBPP/106gK6pq82bdrE8ccfH23atImioqIYN25ctksGAABIqS/TfPVLpgEAAHKFPANQN41qAAAAAAAAAAAAZFTrbBcAAGTOOeecE+eee26Vx8rLy+Mf//hHHHLIIVFSUhLbbLNNlqrbOE8//XS2S8iY0047LU477bRslwEAAFlXU6b5Kpkm98g0AAAgz+QreQYyS6MaABSwbt26xW677VblsfLy8vjwww9jt912i5KSkixVBgAAUL+aMg0AAEA+kGcAqvPRnwAAAAAAAAAAAGSURjUAICIivvjii5gwYUKceeaZsddee0WnTp2ipKQkunbtGgMGDIhf/vKXsWLFijr3UVRUFEVFRXH11VdHRMT06dPj+OOPj6222iratWsXu+yyS1x33XXxxRdfVNlu0qRJMWjQoNS4vn37xqhRo2Lt2rW1HmvgwIFRVFQUAwcObNTP2a9fvygqKopddtml3rGLFy+O0tLSKCoqirPOOqtRx1nvkUceieOPPz622WabKC0tjQ4dOsQOO+wQhxxySPz85z+P559/vto248aNSz2X8+bNq7LutNNOS61ryFdtt99euXJl/OpXv4pvfvOb0b1792jTpk1069YtjjjiiLjvvvuioqKiST8vAABki0xTnUwDAAD5QZ6pTp6BwuSjPwGAiIgYPHhwPPPMM9UeX7RoUfz973+Pv//973HXXXfFpEmTYuedd653f7/4xS/i8ssvjyRJUo+99dZbceWVV8aUKVNi6tSp0b59+7jooovi17/+dZVt33zzzbj88svj73//ezz++OPRqlWrjf8B/39nnnlmnHfeefHWW2/FrFmz4sADD6x17J/+9KdUEBsxYkSjjlNRURH/9V//FQ8++GCVx9euXRsrVqyIuXPnxowZM2Ly5MnxwgsvNP4H2QizZ8+OE044If7zn/9UeXzhwoUxbdq0mDZtWvz2t7+Nv/71r9G9e/dmrQ0AAJpKpqlOppFpAADID/JMdfKMPENh0qgGAERExLp162L33XePIUOGxL777htbbbVVJEkSH3zwQTzyyCPx5z//OebOnRvHH398vPTSS9G2bdta9zV58uR4/vnn4+tf/3pccMEFsdNOO8WiRYvi17/+dUyePDmeffbZ+MUvfhGbb755/PrXv46jjz46zjzzzOjZs2d89NFHMWrUqJg1a1ZMmTIlxowZEz/4wQ/S9nN+97vfjUsuuSRWrVoV9913X50h6L777ouIiF122aXOcTUZPXp0KgAdfPDBceaZZ0bv3r1j0003jSVLlsRrr70WkydPjiVLljRqvzfccENccsklta5ftWpVDBs2LObPnx8lJSWx5ZZbVln/6quvxje/+c344osvolu3bnHOOefEIYccEl26dInPPvss/vrXv8bdd98dzz//fBx33HHxj3/8I0pKShpVIwAAZINMU51MI9MAAJAf5Jnq5Bl5hsKkUQ0ACthnn30Wr732WpXHysvL44MPPojXXnsttt566+jWrVtEfHnBv+OOO1bbxwEHHBDf/va344wzzogjjzwy3n777fjTn/4UZ5xxRq3Hff7552PYsGExYcKEKu+0Oeyww+Lggw+OWbNmxe233x7l5eXx3//933HbbbelxvTr1y8OO+yw6Nu3b3zwwQcxevTotIagTp06xUknnRT3339//O///m/86le/inbt2lUb99JLL8VLL70UEVHnz1qbP//5zxHx5fM3ffr0aN266mXXoYceGhdeeGGjQ9DWW28dW2+9dY3rkiSJk08+OebPnx8REXfeeWf06dOnyvrvfe978cUXX8See+4ZTz75ZGyxxRZV9nHEEUfEMcccE4MHD47nnnsu7r///ib9/AAAkA41ZZoNdevWTab51a+q5Y0ImUamAQAg2+SZ6uQZeQaKs10AAJA5o0ePjt13373KV79+/eKHP/xh9OvXL+66667U2JoC0IYOO+ywGDJkSEREPProo3WO3WSTTeKee+6pdjvoVq1axdlnnx0REcuXL4+uXbvGzTffXOP2w4cPj4iIV155JZYtW1bvz9oYZ555ZkRElJWVxcMPP1zjmN/97ncREVFSUhLf//73G32MBQsWRETEQQcdVGPIWm/zzTdv9L5rc+2116bC14UXXhgjR46ssv7//b//F6+88kpERNx///3VAtB6Rx11VJx44okR8X/vWAIAgGyoKdNs+CXTyDQ1kWkAAMgF8kzN5Bl5hpZNoxoAUKOFCxfGu+++G6+99lrqq2vXrhER8fLLL9e57eGHH17rxf0ee+yRWh46dGittyzec889U8tz585tbPl1OuSQQ1LvYqnpIn/t2rXxwAMPRETE4MGDU+9oaoz1t3OeOHFiLFq0aCOqbZgHH3wwrrnmmoj48h03t956a7Uxjz32WERE9OnTp8p5qMk3vvGNiIiYPXt2VFRUpLlaAADIPJlGpomQaQAAyE/yjDwTIc9QmHz0JwAUsKuuuiquvvrqKo+Vl5fHpEmTYtCgQdUCyMyZM+P222+PJ598ss7bHdd3Ub/TTjvVuq5z586NHrd8+fI6j9cUZ5xxRlx66aXx1FNPxQcffBDbb799at1f//rXWLx4cUREjBgxokn7Hz58ePz973+P9957L772ta/F0KFD4/DDD49DDjkkttlmm7T8DOvNmTMnTjvttEiSJPr06VPtdt7rvfDCCxER8fbbb0dRUVGD9r127dpYsmRJKgADAEBzqinT1KUlZ5oNyTT/R6YBACBb5JnayTPyDC2XO6oBABERcfXVV8fBBx8cf/7zn+sMQBERq1atqnP9JptsUuu64uLiRo/LxLtFhg8fHiUlJZEkSfz+97+vsm79LaW33HLLOProo5u0/xEjRsTll18erVu3jmXLlsV9990Xp5xySmy77bbxta99LS655JJ4//33N/rn+OSTT+K4446LlStXRufOneOvf/1rlQC5oc8++6xJx1i5cuVGVAgAAM2jJWeaP/zhD1XWyTRVyTQAAOQ6eeb/yDNVyTMUGo1qAED87W9/S92SeIcddoi77rorXnnllVi6dGmsW7cukiSJJEni5z//eZYrTZ9u3brFkCFDIiJi3LhxkSRJRER8/PHH8cQTT0RExKmnnhqtWzf9BrQ33HBDvPfee3HDDTfEoYcemgp9//73v+OWW26JnXfeOX772982ef+rV6+O448/Pj766KNo1apV/PnPf67zHVDrw2T//v3j1VdfbfDXVltt1eQaAQCgObT0TPOHP/xBppFpAADIU/KMPCPP0JL46E8AIMaMGRMRX97K+Z///Gd069atxnGff/55c5aVcWeeeWY89NBDMXfu3HjmmWdi4MCBcf/996fCQlNvKb2h7bffPi6//PK4/PLLo7y8PJ5//vl48MEH4+67747Vq1fHueeeGwcccEDsvffejd73iBEj4vnnn4+IiNtuuy0OP/zwOsd36dIlPv3001i4cGHstttuTfp5AAAgF8k0c+O1116LwYMHyzQAAJBn5Bl5BloSd1QDAOL111+PiIhDDz201gAUEfHCCy80V0nN4ogjjojtttsuIiLuu+++iPjy7moRX76jpa53vjRFSUlJ9O/fP371q1/FAw88EBERSZLEX/7yl0bv64Ybbojx48dHRMRZZ50VF1xwQb3brA9a77zzTnzwwQeNPiYAmdfvnNtSXwA0nEwT8dRTT0WETANA9uw/5fLUFwANJ8/IM9CSaFQDAGLdunURUffn3L/00ksxa9as5iqpWRQXF6fekfOXv/wlJk+eHG+//XZERJxxxhkZPfa3vvWt1PKiRYsate0jjzySusX3gAED4je/+U2Dtlt/G+2IiJtvvrlRxwQAgFwm00Q8++yzMWXKFJkGAADyjDwjz0BLolENAIgdd9wxIiJmzJgR77//frX1CxcujO9973vNXVazGDFiRBQXF8fKlStTgWjTTTeNk046aaP2+8c//jEVLmvyxBNPpJZ79erV4P2+9NJL8f3vfz+SJIkddtghHnrooSgpKWnQtsOGDYtddtklIiJGjx4dY8eOrXP8a6+9FhMnTmxwbQAAkC0yTXGsWbMmzjrrrIiQadaTaQAAyAfyjDxTE3mGQtU62wUAANl36qmnxsSJE2PFihUxYMCA+MlPfhL77LNPJEkSzz77bNx6662xYMGC+PrXvx7//Oc/s11uWm277bZxxBFHxJQpU2LBggUREfHtb387Nt10043a7/e///245JJLYujQoXHQQQdF7969o23btvHpp5/GtGnTYvTo0RHxZeBqTMAcMmRIfPHFF9GqVau48cYb45NPPolPPvmk1vG9evWK9u3bR0REq1atYsKECXHQQQfFihUr4swzz4wHH3wwTjnllOjTp0+UlJTEZ599Fi+++GI8/vjj8eyzz8aPfvSjOPbYYzfquQAAgExr6Znm8MMPj6lTp8o0Mg0AAHlInpFn5BlaEo1qAECceOKJcfrpp8d9990XH330UVxwwQVV1rdq1Spuu+22+PzzzwsuBEVEnHnmmTFlypTU9+m6pfSnn34ao0ePTgWer+rcuXNMmDAhttlmmwbvc/78+RERUVFRESeffHK946dPnx4DBw5Mfb/77rvHzJkz48QTT4x33303pk6dGlOnTq11+44dOza4NgAAyJaWnmlOP/30Ktf1Ms3/kWkAAMh18ow8Uxt5hkKkUQ0AiIiI3/3ud3HooYfGPffcEy+99FKsXbs2evToEd/4xjfi/PPPj/333z+uvvrqbJeZEUOGDIlOnTrFsmXLok+fPnHQQQdt9D7feuutmDZtWvztb3+Ld955Jz799NNYtmxZdOjQIfr06RNHHXVUnHPOOdGtW7c0/ASNs8cee8Qbb7wRDzzwQDzyyCPxr3/9KxYuXBiVlZXRpUuX6NOnTxx88MFxwgknRL9+/Zq9PgAAaIqWnGmOPfbY2GSTTWLlypUyjUwDAEAekmfkGXmGlqIoSZIk20WQH8rKylJ/xG+uzt3y8vKYNGlSDBo0KAaVntKobadVPpihqqBmG87Xhn4WOWSDuVrde++9FzvuuGNERNx0001x6aWXZrkiIsxVGi8b16vkj2zNj/X/lo2aNjfWVtQcv+eMvii13O+c22p8nMLl9Y6GME+oz5tvvhl9+/aNCJmGmvl3JPfJM9Qn25nmf9q8EGuLKqqtf/6oG1PL+0+5vMbHKWxeY6iPOUJ95Bkawr8lua+h16vFzVgTAEBOuu+++yIionXr1nHqqadmuRoAAIDG+f3vfx8RMg0AAJB/5BloWQquUa2ioiJee+21GDduXFxwwQXx9a9/PTbZZJMoKiqKoqKiOO200xq9z08++SSuuuqq6NevX2y++eaxySabxA477BDDhw+PZ555ptH7e+yxx+LEE0+M7bffPtq2bRvdunWLr3/963HzzTfHsmXLslobALQ0ZWVlcc8990RExPHHHx89evTIckUAAAANV1ZWFmPHjo2IiCFDhsg0AABA3pBnoOVpne0C0u3b3/52PPzww2nb3yOPPBIjRoyIpUuXVnl87ty5MXfu3Lj//vtj5MiRMXr06GjVqlWd+1q+fHmccsop8fjjj1d5fOHChbFw4cKYNWtW3HHHHTFhwoQGfe5yOmsDgJbks88+i7Kysvjkk0/immuuiUWLFkVRUVH85Cc/yXZpAAAA9aot0/z4xz/OdmkAAAB1kmegZSu4RrWKiooq32+++ebRpUuXePfddxu9ryeffDK+853vRHl5eUREDB48OIYMGRLt27ePOXPmxL333htlZWUxZsyYiIjU3Vhqsm7duhg2bFhMmzYtIiK6d+8eI0eOjL59+8aSJUti/PjxMXPmzPjoo49i8ODBMWPGjNh1112bpTYAaGkuvfTS1K2k1zvnnHNi3333zVJFAAAADVdTpjnqqKNin332yVJFAAAADSPPQMtWcI1q+++/f+yyyy6xzz77xD777BO9evWKcePGxemnn96o/axevTrOOOOMVCPYHXfcEeeff35q/Xe/+904++yzY8CAAbFgwYIYM2ZMnHTSSXH44YfXuL8xY8akmtT69u0bTz31VHTv3j21/rzzzotLLrkkbrnllli6dGmcffbZMWPGjGapDQBaqjZt2kTv3r1j5MiRccEFF2S7HAAAgEZZn2lGjBgRO+ywQ7bLAQAAaDB5Blqm4mwXkG6XX355jBo1Kk488cTo1atXk/czduzY+PDDDyMi4thjj63SCLbeTjvtFHfeeWfq+yuvvLLGfVVUVMS1116b+v4Pf/hDlSa19W666abYa6+9IiJi5syZMXXq1IzXBgAt0bhx4yJJklizZk288cYbcdFFF0Xr1gXXvw8AABSor2aaH/7wh9GqVatslwUAAFAveQZatoJrVEuXCRMmpJYvvvjiWscdf/zx0bNnz4iImDVrVsybN6/amGeeeSYWLFgQEREDBgyIfv361bivVq1axYUXXpj6fvz48RmvDQAAAAAAAAAAINM0qtVg+fLlMXPmzIiI6NChQxxyyCG1ji0uLo6jjjoq9f3kyZOrjZkyZUpqedCgQXUee8P1Ne0r3bUBAAAAAAAAAABkmka1GrzxxhtRWVkZERF77713vbeZ3G+//VLLr732WrX1r776ao1ja9K9e/fYdtttIyLis88+i4ULF2a0NgAAAAAAAAAAgEzTqFaDt99+O7Xcq1evesdvOGbDbTOxv3TXBgAAAAAAAAAAkGmts11ALlq6dGlqeYsttqh3fJcuXWrcNhP7S3dtX7VmzZpYs2ZNjevKysoiIqK8vDzKy8vr3Vc6rD9OeXl5lLQradK20Fw2nK+Qy8xV8oW5SmOZKwAAAAAAALlLo1oNVqxYkVpu27ZtvePbtWuXWl6+fHlG95fu2r5q1KhRcc0119Q55oknnohNNtmk3n2l07Rp0+LsB77TqG0mTZqUoWqgbtOmTct2CdAg5ir5wlyloVauXJntEgAAAAAAAKiFRrV6FBUVZbuEWmWitssuuywuvvjiGteVlZXFtttuG0cccUR07Ngx7ceuSXl5eUybNi0OP/zwOLHrmY3a9rGlv89QVVCzDedrSUnj7gAIzclcJV+YqzTW+jsAAwAAAAAAkHs0qtVg0003TS2vWrWq3vEbjunQoUON+/v8888jImL16tVV9t/Y/aW7tq8qLS2N0tLSOseUlJQ0+x+LS0pKonxV4z7KyR+0yZZs/I5AU5ir5AtzlYYyTwAAAAAAAHKXRrUadO7cObW8aNGiescvXry4xm03fGx9o9qiRYvqbVSra3/prg0AAIDs6nfObdkuAQAAoEn2n3J5tksAACCPFGe7gFzUp0+f1PLcuXPrHb/hmA23zcT+0l0bAAAAAAAAAABApmlUq0Hfvn2juPjLp+bFF1+MioqKOsfPnj07tbzbbrtVW7/77rvXOLYmn376acyfPz8iIrp16xZdu3bNaG0AAAAAAAAAAACZplGtBh06dIj+/ftHRMTy5ctjxowZtY6trKyMqVOnpr4/+uijq4056qijUsuTJ0+u89iTJk1KLQ8aNCjjtQEAAAAAAAAAAGSaRrVanHzyyanlW265pdZxjz76aOrjNQ888MDo2bNntTEDBgyIHj16RETE008/HXPmzKlxXxUVFXH77bfXWEOmagMAAAAAAAAAAMg0jWq1GDFiRGy33XYRETFx4sS48847q415991347zzzkt9f91119W4r1atWsWVV16Z+v7UU0+Nzz77rNq4n/70p/HSSy9FRET//v3jyCOPzHhtAAAAAAAAAAAAmdY62wWk29y5c2Ps2LFVHnvllVdSyy+++GJcccUVVdb369cvhg4dWuWxtm3bxtixY2PQoEFRXl4e559/fkyZMiWGDBkS7du3jzlz5sS9994by5Yti4iIkSNHxmGHHVZrXSNHjoxHHnkkpk2bFq+//nrsueeeMXLkyOjbt28sWbIkxo8fn/oYz06dOsXdd99d677SXRsAAAAAAAAAAEAmFVyj2gcffBA33HBDretfeeWVKo1rERHDhw+v1qgWEXHYYYfFhAkTYsSIEbF06dJ4/PHH4/HHH682buTIkTF69Og662rdunU89NBDccopp8Tjjz8eCxYsqPEuZ9tss01MmDAhdt111zr3l87aAACA3FBRURFvvvlmvPDCC/Gvf/0rXnjhhXj55Zdj1apVEfFldhk3blyj9vnJJ5/Eb3/725g4cWLMmzcvVq9eHT169IhDDjkkRowYEQMGDGjU/h577LH4wx/+ELNnz45PP/00OnbsGL17944TTjghzj777OjUqVPWagMAALJLppFpAACgLgXXqJZuJ5xwQhx44IExevToKkFjyy23jIMPPjjOOOOMBgeNDh06xMSJE+Oxxx6L+++/P2bPnh2fffZZdOjQIXr37h1Dhw5tVAhKZ20AAED2ffvb346HH344bft75JFHUm9u2dDcuXNj7ty5cf/996fe3NKqVas697V8+fLUG282tHDhwli4cGHMmjUr7rjjjpgwYUIcdNBBzVobAACQG2QamQYAAOpScI1qAwcOjCRJ0rrPLbfcMq699tq49tpr07K/4447Lo477ri07CvdtQEAANlTUVFR5fvNN988unTpEu+++26j9/Xkk0/Gd77znSgvL4+IiMGDB8eQIUOiffv2MWfOnLj33nujrKwsxowZExER99xzT637WrduXQwbNiymTZsWERHdu3ePkSNHRt++fWPJkiUxfvz4mDlzZnz00UcxePDgmDFjRp13iU5nbQAAQO6QaWQaAACoS8E1qgEAAOSr/fffP3bZZZfYZ599Yp999olevXrFuHHj4vTTT2/UflavXh1nnHFG6o8md9xxR5x//vmp9d/97nfj7LPPjgEDBsSCBQtizJgxcdJJJ8Xhhx9e4/7GjBmT+oNO375946mnnoru3bun1p933nlxySWXxC233BJLly6Ns88+O2bMmNEstQEAALlDppFpAACgLsXZLgAAAIAvXX755TFq1Kg48cQTo1evXk3ez9ixY+PDDz+MiIhjjz22yh9N1ttpp53izjvvTH1/5ZVX1rivioqKKndw/sMf/lDlDzrr3XTTTbHXXntFRMTMmTNj6tSpGa8NAADILTJN42oDAICWRqMaAABAgZkwYUJq+eKLL6513PHHHx89e/aMiIhZs2bFvHnzqo155plnYsGCBRERMWDAgOjXr1+N+2rVqlVceOGFqe/Hjx+f8doAAIDCJNMAAEBh0qgGAABQQJYvXx4zZ86MiIgOHTrEIYccUuvY4uLiOOqoo1LfT548udqYKVOmpJYHDRpU57E3XF/TvtJdGwAAUHhkGgAAKFwa1QAAAArIG2+8EZWVlRERsffee0erVq3qHL/ffvulll977bVq61999dUax9ake/fuse2220ZExGeffRYLFy7MaG0AAEDhkWkAAKBwaVQDAAAoIG+//XZquVevXvWO33DMhttmYn/prg0AACg8Mg0AABSu1tkuAAAAgPRZunRpanmLLbaod3yXLl1q3DYT+0t3bV+1Zs2aWLNmTY3rysrKIiKivLw8ysvL691Xuqw/VptWRU3elsK2/jw739TFPKE+5gj1MUdyn3Pzf2SaHM00SePufWFOtxxeY6iPOUJ9zBEawjzJfQ09NxrVAAAACsiKFStSy23btq13fLt27VLLy5cvz+j+0l3bV40aNSquueaaOsc88cQTsckmm9S7r3T70aE9G73NpEmT0l8IOWvatGnZLoE8YJ5QH3OE+pgjuWvlypXZLiFnyDS5mWl+WN6vUePlmZbHawz1MUeojzlCQ5gnuauhmUajGgAAQIEqKmr8XbyaSyZqu+yyy+Liiy+ucV1ZWVlsu+22ccQRR0THjh3TfuzalJeXx7Rp0+KWp+bF2oqkUdv+47bzMlQVuWT9HDn88MOjpKQk2+WQo8wT6mOOUB9zJPetv1sWVck0/yfbmebXJXNibVFlg7ebftiVGayKXOI1hvqYI9THHKEhzJPc19BMo1ENAACggGy66aap5VWrVtU7fsMxHTp0qHF/n3/+eURErF69usr+G7u/dNf2VaWlpVFaWlrnmJKSkqz8j4y1FUmjG9X8D5eWJVtzk/xinlAfc4T6mCO5y3n5PzJNjmaaospYW1TR4PHmdMvjNYb6mCPUxxyhIcyT3NXQ89K4D5QHAAAgp3Xu3Dm1vGjRonrHL168uMZtM7G/dNcGAAAUHpkGAAAKl0Y1AACAAtKnT5/U8ty5c+sdv+GYDbfNxP7SXRsAAFB4ZBoAAChcGtUAAAAKSN++faO4+Muo9+KLL0ZFRd0fzTJ79uzU8m677VZt/e67717j2Jp8+umnMX/+/IiI6NatW3Tt2jWjtQEAAIVHpgEAgMKlUQ0AAKCAdOjQIfr37x8REcuXL48ZM2bUOraysjKmTp2a+v7oo4+uNuaoo45KLU+ePLnOY0+aNCm1PGjQoIzXBgAAFB6ZBgAACpdGNQAAgAJz8sknp5ZvueWWWsc9+uijqY+iOfDAA6Nnz57VxgwYMCB69OgRERFPP/10zJkzp8Z9VVRUxO23315jDZmqDQAAKEwyDQAAFCaNagAAAAVmxIgRsd1220VExMSJE+POO++sNubdd9+N8847L/X9ddddV+O+WrVqFVdeeWXq+1NPPTU+++yzauN++tOfxksvvRQREf37948jjzwy47UBAACFSaYBAIDC1DrbBQAAAPCluXPnxtixY6s89sorr6SWX3zxxbjiiiuqrO/Xr18MHTq0ymNt27aNsWPHxqBBg6K8vDzOP//8mDJlSgwZMiTat28fc+bMiXvvvTeWLVsWEREjR46Mww47rNa6Ro4cGY888khMmzYtXn/99dhzzz1j5MiR0bdv31iyZEmMHz8+9ZE3nTp1irvvvrvWfaW7NgAAIHfINDINAADURaMaAABAjvjggw/ihhtuqHX9K6+8UuWPPBERw4cPr/ZHnYiIww47LCZMmBAjRoyIpUuXxuOPPx6PP/54tXEjR46M0aNH11lX69at46GHHopTTjklHn/88ViwYEGNdwTYZpttYsKECbHrrrvWub901gYAAOQOmabxtQEAQEuiUQ0AAKBAnXDCCXHggQfG6NGjY+LEiTFv3rxYvXp1bLnllnHwwQfHGWecEQMGDGjQvjp06BATJ06Mxx57LO6///6YPXt2fPbZZ9GhQ4fo3bt3DB06NM4+++zo1KlTs9cGAAAUJpkGAAAKi0Y1CtbhxSellqdVPpjFSgAAoGEGDhwYSZKkdZ9bbrllXHvttXHttdemZX/HHXdcHHfccWnZV7prAwAAskumAQAA6lKc7QIAAAAAAAAAAAAobBrVAAAAAAAAAAAAyCiNagAAAAAAAAAAAGSURjUAAAAAAAAAAAAySqMaAAAAAAAAAAAAGaVRDQAAAAAAAAAAgIzSqAYAAAAAAAAAAEBGaVQDAAAAAAAAAAAgozSqAQAAAAAAAAAAkFEa1QAAAAAAAAAAAMgojWoAAAAAAAAAAABklEY1AAAAAAAAAAAAMkqjGgAAAAAAAAAAABmlUQ0AAAAAAAAAAICM0qgGAAAAAAAAAABARmlUAwAAAAAAAAAAIKM0qgEAAAAAAAAAAJBRGtUAAAAAAAAAAADIKI1qAAAAAAAAAAAAZJRGNQAAAAAAAAAAADJKoxoAAAAAAAAAAAAZpVENAAAAAAAAAACAjNKoBgAAAAAAAAAAQEZpVAMAAAAAAAAAACCjNKoBAAAAAAAAAACQURrVAAAAAAAAAAAAyCiNagAAAAAAAAAAAGSURjUAAAAAAAAAAAAySqMaAAAAAAAAAAAAGaVRDQAAAAAAAAAAgIzSqAYAAAAAAAAAAEBGaVQDAAAAAAAAAAAgozSqAQAAAAAAAAAAkFEa1QAAAAAAAAAAAMgojWoAAAAAAAAAAABklEY1AAAAAAAAAAAAMkqjGgAAAAAAAAAAABmlUQ0AAAAAAAAAAICM0qgGAAAAAAAAAABARmlUAwAAAAAAAAAAIKM0qgEAAAAAAAAAAJBRrbNdADSHw4tPSi1Pq3wwi5UAAAAAAAAAAEDL445qAAAAAAAAAAAAZJQ7qgEAAEAO6nfObanlOaMvymIlAAAAjbP/lMurfP/8UTdmqRIAAHKJO6oBAAAAAAAAAACQURrVAAAAAAAAAAAAyCiNagAAAAAAAAAAAGSURjUAAAAAAAAAAAAySqMaAAAAAAAAAAAAGaVRDQAAAAAAAAAAgIzSqAYAAAAAAAAAAEBGaVQDAAAAAAAAAAAgozSqAQAAAAAAAAAAkFEa1QAAAAAAAAAAAMgojWoAAAAAAAAAAABkVOtsFwDN7fDik6p8P63ywSxVAgAAAAAAAAAALYNGNQAAAMhx/c65LbU8Z/RFWawEAACg8fafcnlq+fmjbsxiJQAAZJOP/gQAAAAAAAAAACCjNKoBAAAAAAAAAACQURrVAAAAAAAAAAAAyCiNagAAAAAAAAAAAGSURjUAAAAAAAAAAAAySqMaAAAAAAAAAAAAGaVRDQAAAAAAAAAAgIzSqAYAAAAAAAAAAEBGaVQDAAAAAAAAAAAgozSqAQAAAAAAAAAAkFEa1QAAAAAAAAAAAMgojWoAAAAAAAAAAABklEY1AAAAAAAAAAAAMkqjGgAAAAAAAAAAABmlUQ0AAAAAAAAAAICMap3tAiDbDi8+KbU8rfLBLFYCAADQOP3OuS21PGf0RVmsBAAAoPH2n3J5avn5o27MYiUAADQHd1QDAAAAAAAAAAAgo9xRDQAAAPLIhndRAwAAyDcb3kUNAICWxR3VAAAAAAAAAAAAyCiNagAAAAAAAAAAAGSURjUAAAAAAAAAAAAySqMaAAAAAAAAAAAAGaVRDQAAAAAAAAAAgIzSqAYAAAAAAAAAAEBGaVQDAAAAAAAAAAAgozSqAQAAAAAAAAAAkFEa1QAAAAAAAAAAAMgojWoAAAAAAAAAAABklEY1AAAAAAAAAAAAMkqjGgAAAAAAAAAAABnVOtsFQC45vPik1PK0ygezWAkAAAAAAAAAABQOjWpQC01rAAAAAAAAAACQHj76EwAAAAAAAAAAgIzSqAYAAAAAAAAAAEBGaVQDAAAAAAAAAAAgo1pnuwAAAABg4/U757bU8pzRF2WxEgAAgMbbf8rlqeXnj7oxi5UAAJAp7qgGAAAAAAAAAABARmlUAwAAAAAAAAAAIKM0qgEAAAAAAAAAAJBRrbNdAOSDw4tPSi1Pq3wwi5UAAADUr985t1X5fs7oi7JUCQAAQOPtP+Xy1PLzR92YxUoAAEgnd1QDAAAAAAAAAAAgo9xRDQAAAFqQDe+25k5rAABAPnGnNQCA/KZRDRrJx4ACAAAAAAAAAEDj+OhPAAAAAAAAAAAAMsod1QAAAKDAbfhxnwAAAPlkw4/7BAAgv2lUgzTZ8CNBN+TjQQEAAAAAAAAAaOk0qsFGqK05DQAAIN9seNe1OaMvymIlAAAAjfPVu649f9SNWaoEAIC6aFSDDPtqM5s7rAEAAPlEAxsAAJDPNmxi08AGAJBdxdkugI332GOPxYknnhjbb799tG3bNrp16xZf//rX4+abb45ly5ZluzwAAIA6yTQAAEC+kmcAAKDh3FEtjy1fvjxOOeWUePzxx6s8vnDhwli4cGHMmjUr7rjjjpgwYUIcdNBBWaqShtrwzmvuugYAQEsg02TfhndLS9d+3HUNAICWQJ7Jvq9+3Ge69uWuawAAmaNRLU+tW7cuhg0bFtOmTYuIiO7du8fIkSOjb9++sWTJkhg/fnzMnDkzPvrooxg8eHDMmDEjdt111yxXTUT1jwJtzHgNbAAAFAqZJrelq4ENAAAKkTyT+9LZxAYAQPpoVMtTY8aMSQWgvn37xlNPPRXdu3dPrT/vvPPikksuiVtuuSWWLl0aZ599dsyYMSNb5dJIjW1mAwCAfCPT5Cd3TgMAAHkmn7lzGgBAdmlUy0MVFRVx7bXXpr7/wx/+UCUArXfTTTfF3/72t3jppZdi5syZMXXq1DjyyCObs1QAAIBqZJrCUNtd1776uIY2AAAKiTxTOGq765pmNgCAzNGoloeeeeaZWLBgQUREDBgwIPr161fjuFatWsWFF14YI0aMiIiI8ePHC0EFpKF3XfNxoQAA5BqZpmWp7S5s7s4GAEA+kmdaltqa1r7a5KahDQCgYTSq5aEpU6aklgcNGlTn2A3XT548OWM10Tw29iNBN9x+wwa22vbbkDF1bQMAADWRaVquht6Fbb2NbWzL1H4BAGi55JmWq7Y7sNW1rrbmtoY0tjVkn03ZLwBANmlUy0Ovvvpqanm//farc2z37t1j2223jfnz58dnn30WCxcujK5du2a6RHJIbQ1mDWk8a0pjXG3NcI0dAwBA4ZJpSKfamtHSvU1t22puAwBoWeQZ0q2uBrh0jK9re41tAEBz06iWh95+++3Ucq9eveod36tXr5g/f35q27pC0Jo1a2LNmjU1rlu2bFlERCxZsiTKy8sbU3KTlZeXx8qVK2Px4sURbZvlkKTR4ZvU0ujWtgFjmuB/5/82tXzytj9o1OPpsOF8LSkpSeu+IZ3MVfKFuUpjLV++PCIikiTJciXUJ1OZJtfyTMT//VtWvG5NFFeam42171m/SC0XN+DxdB6vNl89XkO2qUub4qK4YMB2VV7vjrxsTGr91FEjN2r/FAbXRdTHHKE+5kjuk2fyR0v6G03EBpmmvDyKiyqb7biF4sCHLk0tFzfg8XQerza11dFUbZLiOKd8zyqvMYOfvim1/v8N/MlGH4P85jqE+pgjNIR5kvsammk0quWhpUuXppa32GKLesd36dKlxm1rMmrUqLjmmmvqHNOQ4AXZsMUWf0nL4wBAflu+fHl06tQp22VQh0xlGnmGfPXPu2tft8WYy5qvEAAg6+SZ3OdvNFDdP+pYt0X8T7PVAQBkX32ZRqNaHlqxYkVquW3b+m8z1q5du9Ty+g7G2lx22WVx8cUX17iusrIylixZEl26dImioqIGVrtxysrKUrfF7tixY7McE5rKfCVfmKvkC3OVxkqSJJYvXx5bbbVVtkuhHpnKNLmWZyL8W0b9zBEawjyhPuYI9TFHcp88kz9a0t9oIvz7Qf3MEepjjlAfc4SGME9yX0MzjUY1qigtLY3S0tJa13fu3Ln5itlAx44d/WND3jBfyRfmKvnCXKUx3HmgZcvVPBPh3zLqZ47QEOYJ9TFHqI85ktvkGWQa8pk5Qn3MEepjjtAQ5klua0imSefHntNMNt1009Ty6tWr6x2/atWq1HKHDh0yUhMAAEBDyTQAAEC+kmcAAKDpNKrloQ3fMbNo0aJ6xy9evLjGbQEAALJBpgEAAPKVPAMAAE2nUS0P9enTJ7U8d+7cesdvOGbDbQEAALJBpgEAAPKVPAMAAE2nUS0P7b777qnl2bNn1zn2008/jfnz50dERLdu3aJr164ZrQ0AAKA+Mg0AAJCv5BkAAGg6jWp56KijjkotT548uc6xkyZNSi0PGjQoYzUBAAA0lEwDAADkK3kGAACaTqNaHhowYED06NEjIiKefvrpmDNnTo3jKioq4vbbb099f/LJJzdLfelUWloaV111VZSWlma7FKiX+Uq+MFfJF+YqFC6ZBv6POUJDmCfUxxyhPuYIpE9LyjMR/v2gfuYI9TFHqI85QkOYJ4WjKEmSJNtF0HijR4+Oc889NyIidt1113jqqaeiW7duVcb8+Mc/jl/+8pcREdG/f/+YMWNGs9cJAABQE5kGAADIV/IMAAA0jUa1PLVu3boYNGhQTJs2LSIievToESNHjoy+ffvGkiVLYvz48anQ06lTp5g5c2bsuuuu2SwZAAAgRaYBAADylTwDAABNo1Etjy1fvjxOOeWUePzxx2sds80228SECRPioIMOasbKAAAA6ifTAAAA+UqeAQCAxtOoVgAee+yxuP/++2P27Nnx2WefRYcOHaJ3794xdOjQOPvss6NTp07ZLhEAAKBWMg0AAJCv5BkAAGg4jWoAAAAAAAAAAABkVHG2CwAAAAAAAAAAAKCwaVQjJz322GNx4oknxvbbbx9t27aNbt26xde//vW4+eabY9myZdkujxxSUVERr732WowbNy4uuOCC+PrXvx6bbLJJFBUVRVFRUZx22mmN3ucnn3wSV111VfTr1y8233zz2GSTTWKHHXaI4cOHxzPPPNPo/aVzPqe7NprP8uXL46GHHorzzz8/DjrooOjatWuUlJREx44dY+edd45TTz01pkyZEo250am5SqbMnj077rzzzjjttNNiv/32i549e8amm24apaWl0b179xg4cGBce+218eGHHzZ4n+Yr0JLIM4WjpeUNmqalXevTeC3t+pr0Gj58eOp1p6ioKK6++uoGbWeOABvD73zhkGmojzxDfeQZNpZMQzUJ5JCysrLkmGOOSSKi1q9tttkmmTlzZrZLJUcMHTq0zvkyfPjwRu3v4YcfTjp37lznPkeOHJmsW7eu3n2lez6nszaa1y233JK0bdu2znO3/uuQQw5JPvjgg3r3aa6SSe3bt2/QfC0tLU1uvPHGevdnvgIthTxTeFpS3qBpWtq1Pk3Tkq6vSa/JkydXOx9XXXVVvduZI0BT+Z0vPDINdZFnaAh5ho0h01CT1gE5Yt26dTFs2LCYNm1aRER07949Ro4cGX379o0lS5bE+PHjY+bMmfHRRx/F4MGDY8aMGbHrrrtmuWqyraKiosr3m2++eXTp0iXefffdRu/rySefjO985ztRXl4eERGDBw+OIUOGRPv27WPOnDlx7733RllZWYwZMyYiIu65555a95Xu+ZzO2mh+77zzTqxevToiIrbZZpv41re+Ffvuu2907do1Vq1aFc8991z88Y9/jBUrVsQ//vGPGDhwYMyaNSu6detW4/7MVZpDt27dYv/9949dd901evToET169IgkSWLevHnx//7f/4uZM2fGmjVr4vLLL4/y8vK48sora9yP+Qq0FPJMYWopeYOma0nX+myclnB9TXqVlZXFWWedFRER7du3jy+++KJB25kjQFP5nS9MMg11kWdoKHmGppBpqFWWG+Ug5a677kp1rPbt2zdZsGBBtTE/+tGPUmP69++fhSrJNTfccEPy05/+NHnwwQeT999/P0mSJLnvvvsa/W6gVatWJdttt11quzvuuKPamLfffjvp0aNHaswTTzxR6/7SOZ/TXRvN7wc/+EFyxBFHJE888URSUVFR45h58+Ylffr0SZ3D008/vcZx5irN4dVXX00qKyvrHPP73/8+KSoqSiIiad26dfKf//yn2hjzFWhJ5JnC1BLyBhunpVzrs3FawvU16XfWWWel3tl/8cUXN+juA+YIsDH8zhcmmYa6yDM0hDxDU8k01EajGjlh3bp1Vf4x+de//lXruL322is1bsqUKc1cKfmgKSHrN7/5TWqbY489ttZxDz30UGrcgQceWOOYdM/ndNZGdixevLhB41566aXUOdxkk02SL774otoYc5Vccuyxx6bO5dixY6utN1+BlkKeaVkKLW+wcVrKtT7NI5+vr0mvp556KvWHvokTJyZXXXVVg/6oY44ATeV3vmWRaVhPniGd5Bk2JNNQl+KAHPDMM8/EggULIiJiwIAB0a9fvxrHtWrVKi688MLU9+PHj2+W+ih8EyZMSC1ffPHFtY47/vjjo2fPnhERMWvWrJg3b161Memez+msjezYfPPNGzRuzz33jJ133jkiIlauXBnvvfdetTHmKrlkw9sff/rpp9XWm69ASyHPUJ9cfk1k47SUa32aRz5fX5M+K1eujDPPPDOSJInvfOc7ccwxxzR4W3MEaCq/89THa0xhkmdIJ3mG9WQa6qNRjZwwZcqU1PKgQYPqHLvh+smTJ2esJlqO5cuXx8yZMyMiokOHDnHIIYfUOra4uDiOOuqo1Pc1zcF0zud010bu69ChQ2p51apVVdaZq+SaDf+HRI8ePaqsM1+BlkSeoS65/JpI88rXa32aT75eX5Nel112Wbz//vux2Wabxa9//esGb2eOABvD7zx18RpDhDxD/eQZ1pNpqI9GNXLCq6++mlreb7/96hzbvXv32HbbbSMi4rPPPouFCxdmtDYK3xtvvBGVlZUREbH33ntHq1at6hy/4Rx97bXXqq1P53xOd23ktjVr1sQ777yT+n777bevst5cJZc8+uij8fDDD0dERLt27WLw4MFV1puvQEsiz1CXXH5NpPnk87U+zSOfr69Jn2effTZ+85vfRETELbfcEt27d2/wtuYIsDH8zlMXrzHIM9RHnmE9mYaGaJ3tAiAi4u23304t9+rVq97xvXr1ivnz56e27dq1a8Zqo/A1Zf7VtO3G7K+2+Zzu2sht48ePj2XLlkVERL9+/aq948RcJRv+/ve/x5IlSyIiYu3atTF//vyYOnVqTJs2LSIiSkpK4p577olu3bpV2c58BVoSeYa65PJrIs0nn6/1Sa9CvL4mPVavXh0jRoyIysrK+Na3vhWnn356o7Y3R4CN4XeeuniNQZ5hPXmGusg0NJRGNXLC0qVLU8tbbLFFveO7dOlS47bQFOmef+ncn9+NlmPhwoVx6aWXpr6/4oorqo0xV8mGSy+9NJ577rlqjxcVFcU3v/nNuPbaa6N///7V1puvQEvi3xXqksuviTSPfL/WJ70K8fqa9Ljyyivj7bffjnbt2sXdd9/d6O3NEWBj+J2nLl5jWjZ5hg3JM9RFpqGhfPQnOWHFihWp5bZt29Y7vl27dqnl5cuXZ6QmWo50z7907s/vRsuwdu3aGDZsWOo2sscff3yccMIJ1caZq+SSbbbZJg499NDo2bNnjevNV6Al8e8Kdcnl10QyrxCu9Wke+Xx9zcZ74YUX4tZbb42IiGuuuSZ69+7d6H2YI8DG8DtPXbzGtFzyDA0lzyDT0Bga1QA2UFRUlO0SapXLtdF0lZWVMWLEiPjHP/4RERG9e/eO3/3ud/Vul8vzIZdro/FmzZoVSZJEkiSxYsWKePHFF+Pqq6+OpUuXxhVXXBF77LFHTJ06tc595PKcyOXaACg8XndalkK81mfjFfr1NY23du3aOP3006OioiL23nvvuPjiizd6n+YIAJniNablkGeoiTxDTWQaGkujGjlh0003TS2vXr263vGrVq1KLXfo0CEjNdFybDj/Npxbtalv/qVzPqe7NnJLkiTxgx/8IP70pz9FRMR2220XTz75ZGy22WY1jjdXybb27dvHXnvtFVdddVW8+OKLseWWW8aSJUtiyJAh8fLLL1cZa74CLYk8Q11y+TWRzCmka30yp1Cur9k4119/fbz22mvRqlWruPfee6NVq1ZN2o85AmwMv/PUxWtMyyPP0BDyDOvJNDSWRjVyQufOnVPLixYtqnf84sWLa9wWmiLd8y+d+/O7UbiSJIlzzz03xowZExFf3hb5qaeeqvW2yBHmKrmld+/eMWrUqIj48t0yN954Y5X15ivQkvh3hbrk8msimVFo1/o0j3y+vqbpXn755fjFL34REREXX3xx9OvXr8n7MkeAjeF3nrp4jWlZ5BmaQp5puWQamqJ1tguAiIg+ffrE3LlzIyJi7ty5dV7srB+z4bawMTacQxvOrdrUN//SOZ/TXRu5IUmSOO+88+K3v/1tRERsvfXWMX369Ho/r91cJdcMHjw4tfz0009XWWe+Ai2JPENdcvk1kfQrxGt9mk++Xl/TdOPGjYvy8vIoLi6OkpKSuP7662sc9/e//73K8vpxffr0iZNOOim1vJ45AjSW33nq4jWm5ZBn2BjyTMsk09AU7qhGTth9991Ty7Nnz65z7Keffhrz58+PiIhu3bpF165dM1obha9v375RXPzlP4cvvvhiVFRU1Dl+wzm62267VVufzvmc7trIvvVBb/To0RERsdVWW8X06dPja1/7Wr3bmqvkmg1vfbx06dIq68xXoCWRZ6hLLr8mkl6Feq1P88nX62uaLkmSiIiorKyMG2+8MX7+85/X+DV9+vTUNtOnT089vv7juCLMEWDj+J2nLl5jWgZ5ho0lz7RMMg1NoVGNnHDUUUellidPnlzn2EmTJqWWBw0alLGaaDk6dOgQ/fv3j4iI5cuXx4wZM2odW1lZGVOnTk19f/TRR1cbk875nO7ayK6vBr0tt9wypk+fHjvuuGODtjdXyTXvvvtuavmrF+3mK9CSyDPUJZdfE0mfQr7Wp/nk6/U1ucEcATaG33nq4jWm8MkzpIM8w8YyT1qQBHLAunXrkh49eiQRkURE8q9//avWcXvttVdq3JQpU5q5UvLBfffdl5ojw4cPb9A2d955Z2qbY489ttZxDz30UGrcgQceWOOYdM/ndNZGdp177rmpc9SjR4/krbfeavQ+zFVyyXnnnZc6l9/+9rerrTdfgZZCnmlZCi1vkB6Ffq1P88jn62sy66qrrkqdg6uuuqrWceYI0FR+51sWmYavkmdIB3mGusg0bEijGjnjrrvuSv3y77rrrsmnn35abcwll1ySGtO/f/8sVEk+aErIWrVqVbLddtultvvNb35Tbcw777xT5cVs2rRpte4vnfM53bWRHeeff36VoPfmm282aT/mKpk2evTo5KmnnkoqKytrHbNu3bpk1KhRSVFRUepcPv3009XGma9ASyLPtByFljfYeC3hWp+maynX12RWQ/+oY44AG8PvfMsh07AheYa6yDOki0zDhoqS5P//0FjIsnXr1sWgQYNi2rRpERHRo0ePGDlyZPTt2zeWLFkS48ePT93esVOnTjFz5szYdddds1kyOWDu3LkxduzYKo+98sorMXHixIiI2GOPPeLYY4+tsr5fv34xdOjQavt68sknY9CgQVFeXh4REcccc0wMGTIk2rdvH3PmzIl77703li1bFhERI0eOjHvuuafWutI9n9NZG83viiuuiBtuuCEiIoqKiuLGG2+MnXfeud7t+vXrF9ttt121x81VMum0006L3//+97HtttvG4YcfHrvvvnt069Yt2rRpE0uXLo3XXnstHnvssZg3b15qm8suuyxuvPHGGvdnvgIthTxTmFpK3qDpWtK1Pk3Tkq6vyZyrr746rrnmmoiIuOqqq+Lqq6+udaw5AjSV3/nCJNNQF3mG+sgzpItMQxXZ7pSDDZWVlSXHHHNMqmO1pq9tttkmmTlzZrZLJUdMnz69zvlS01dd7xB6+OGHk86dO9e5/ciRI5N169bVW1u653M6a6N5DRgwoNHzNCKS++67r9Z9mqtkyvDhwxs8Rzt16pTcdddd9e7TfAVaCnmm8LSkvEHTtLRrfRqvpV1fkxkNvfvAeuYI0FR+5wuPTENd5BnqI8+QLjING9KoRk569NFHk6FDhybbbrttUlpammyxxRbJAQcckNx0003J0qVLs10eOSTdIStJkuTjjz9Ofv7znyd77bVX0rlz56Rt27ZJr169ku9///s13qq2Pumcz+mujeaRibCXJOYqmbF8+fJkypQpyWWXXZZ885vfTPr06ZNsttlmSevWrZNOnTolO+20U3LiiScmY8aMadS8MF+BlkSeKRwtLW/QeC3xWp/GaYnX16RfY/+okyTmCLBx/M4XDpmGusgz1EeeIV1kGjbkoz8BAAAAAAAAAADIqOJsFwAAAAAAAAAAAEBh06gGAAAAAAAAAABARmlUAwAAAAAAAAAAIKM0qgEAAAAAAAAAAJBRGtUAAAAAAAAAAADIKI1qAAAAAAAAAAAAZJRGNQAAAAAAAAAAADJKoxoAAAAAAAAAAAAZpVENAAAAAAAAAACAjNKoBgAAAAAAAAAAQEZpVAMAAAAAAAAAACCjNKoBQCMVFRWlvmozcODA1Jinn366+YprRi3hZwQAgEIjz3ypJfyMAABQiGSaL7WEnxEoTBrVAKjThhe6jfl66aWXsl06AADQwskzAABAPpNpACg0GtUAAAAAAAAAAADIqNbZLgCA/LHffvvF/vvv36CxXbt2zXA1AAAADSfPAAAA+UymAaAQaFQDoMEGDRoUV199dbbLyLokSbJdQk54+umns10CAAA0mDzzJXnmS/IMAAD5Rqb5kkzzJZkGyFc++hMAAAAAAAAAAICM0qgGAAAAAAAAAABARmlUAyDjPvvss7jvvvti+PDhsffee8fmm28eJSUl0blz59h5553j9NNPj6lTpzZoX1dffXUUFRVFUVFR6hbXK1eujLvuuisOOeSQ6NGjR7Rp0yZ69uwZZ511VnzwwQfV9rFo0aK44YYbYt99940uXbpE+/btY/fdd49rrrkmvvjii3prWH/8oqKiRj0P6916662p7Y866qgGb/fEE0+kttt2222jsrKyScdfb/bs2XH++edHv379YrPNNouSkpLYfPPNY6eddoqDDz44Lr744pg6dWqsWrWqxu0HDhyYqqemW0z37NmzynPV0K9x48bVWfebb74Zl19+eey///7RvXv3aNOmTXTt2jUOOOCAuPLKK+Pjjz/eqOcFAAA2JM9UJc/IMwAA5BeZpiqZRqYBsiwBgDoMGDAgiYgkIpKrrrqq0dv/+te/Tlq1apXaR11fhx56aLJo0aI693fVVVdVqefdd99Ndtttt1r32alTp+Rf//pXavu//vWvSadOnWod36dPn2TBggV11rDh+Nps+LxNnz69yrpFixYlpaWlSUQkxcXFyQcffFD/E5kkyXe+853UPn/+8583aJualJeXJ2eddVaDzklEJD/72c8a/TMmSZJsv/32DT7Ghl/33XdfjcdbvXp18oMf/KDe+dSuXbvkjjvuaPLzAwBA4ZBnqpNn6v8Zk0SeAQAgN8g01ck09f+MSSLTALmrdQBABn388cdRUVERERE77LBD7LLLLtG1a9do27ZtLF26NF599dV4/fXXIyLiqaeeisMOOyxmzZoVpaWl9e67rKwsBg0aFO+++2507tw5Bg4cGF27do358+fHU089FWvXro1ly5bFkUceGe+9917MmTMnhg0bFuXl5bH99tvH17/+9dh0003j9ddfj3/+858REfH222/H9773vZg2bVrGnpMuXbrE0KFDY/z48VFZWRnjxo2LK6+8ss5tlixZEo8++mhEfPluodNPP73Jx7/kkkvinnvuSX2/9dZbx/777x9bbLFFJEkSixcvjjfeeCPefvvtJh8jImL48OGxePHiesctXrw4/vd//zf1fU3vgvriiy/iyCOPjJkzZ6Ye69WrV+y7776x2Wabxeeffx7PPvts/Oc//4lVq1bFBRdcEGVlZXH55Zdv1M8AAEDLJs9UJ89UJc8AAJDLZJrqZJqqZBqg2WW5UQ6AHLex79YZO3ZscscddyQfffRRrWNefvnlZN99900d57rrrqt17Ibv1mnTpk0SEcl5552XfPHFF1XGvfnmm8lWW22VGvvjH/842WabbZI2bdok9957b1JZWVll/EMPPZS0bt06Nf7pp5+utYbY4J0htanvnSzTp09Pre/Zs2e1er7q9ttvT40/7LDD6hxbl4ULF6Z+zlatWiXjxo2r9dgff/xxcvvttyf33ntvjevr+xkbYu3atcnAgQNT+9l5552TpUuXVht36qmnpsb07t07mTZtWrUx69atS+66667UO6FatWqVPPvss02qCwCAwiDPVCfPfEmeAQAgH8g01ck0X5JpgHzljmoANNikSZNi0aJF9Y677LLLYuutt46IiBEjRtQ7fo899ognn3wydt5551iwYEHcddddcdlll0WrVq3q3G7t2rVx2mmnxW9+85tq63beeee45ZZb4r/+678iIuJ//ud/IiLiD3/4Q3zve9+rNn7o0KExYsSI1LtY/vd//zcGDBhQb+1NNXDgwNhpp53inXfeiXnz5sXf/va3OOyww2od/7vf/S61fMYZZzT5uLNmzYp169ZFRMTJJ58cw4cPr3XslltuGRdccEGTj9UQ5513Xjz99NMREbH55pvHxIkTo1OnTlXG/OMf/4j7778/IiK22WabmDFjRvTo0aPavlq1ahXnnHNOtG3bNkaMGBEVFRVx7bXXxuTJkzP6MwAAkB/kmfSRZ74kzwAA0JxkmvSRab4k0wDZoFENgAabPXt2zJ49u95xZ555ZioENVSnTp3ihBNOiNGjR8cnn3wSb7zxRuy+++51btOmTZu46aabal1//PHHR2lpaaxZsyYiIvr161djAFrv5JNPToWghvycG+vMM8+MSy+9NCIixo4dW2sI+te//hUvvfRSRHwZFE444YQmH7OsrCy13LVr1ybvJx1+9atfxZgxYyIioqSkJB566KH42te+Vm3crbfemloeNWpUjQFoQ6eddlrcfPPN8dZbb8XUqVNj0aJFscUWW6S3eAAA8o48k17yjDwDAEDzkmnSS6aRaYDs0KgGQLP57LPPYtasWfHmm2/G559/Hl988UUkSZJa/8ILL6SWX3rppXpD0De+8Y3o1q1brevbtm0bvXv3jjfeeCMiIoYNG1bn/nbbbbfU8ty5c+scmw6nnXZaXHHFFbF27dp45JFHYsmSJbH55ptXG7fhO3W+973vRWlpaZOPud1226WWH3roofjJT35Sb6jIhClTpsQll1yS+v7OO++MgQMHVhu3bt26mDZtWkREtG7dut5zGBFRVFQU3/zmN+Ott96KJEni2WefjSFDhqStdgAAWiZ5pip5Rp4BACC/yDRVyTQyDZAdGtUAaLCrrroqrr766kZv98Ybb8RPfvKTmDx5clRUVDRom4bcvnrXXXetd8xmm22WWu7bt2+Dx274rpZM6dq1axx33HHx4IMPxpo1a+JPf/pTtds4r169Oh544IHU92eeeeZGHfOAAw6I7bffPj744IOYP39+7LrrrjF8+PA49thj48ADD4x27dpt1P4b4o033ojvfOc7qblw0UUXxciRI2sc+8orr8QXX3wREV+G2h//+McNOsaG77aaP3/+RlYMAEAhkGfSS56RZwAAaF4yTXrJNDINkB0a1QDIqKlTp8Zxxx2XurVzQy1fvrzeMZ06dap3TOvW//dSV9/4DceuW7eu3n2nw1lnnRUPPvhgRHx5a+mvhqCHH344li5dGhER++23X73vYKpPSUlJ/PGPf4zBgwdHWVlZLFmyJG677ba47bbbok2bNtGvX7/4xje+EYMHD45DDjkkioqKNup4X7V48eI49thjUyFz0KBB8ctf/rLW8R9//HFqecWKFXHnnXc2+piff/554wsFAICQZ+ojz8gzAADkNpmmbjKNTAM0v+JsFwBA4Vq4cGF85zvfSQWgXr16xU033RQzZ86Mjz/+OFauXBmVlZWRJEkkSRJXXXVVatvKysp699/YC/R0X9Cnw7e+9a3o3bt3RES8/PLLMWfOnCrrx44dm1re2HfqrHfwwQfHyy+/HCNGjIj27dunHl+7dm3MmjUrbr755hgwYEDsvPPO8cgjj6TlmBER5eXlMWzYsHj//fcj4st3W/3v//5vFBfXfjmybNmyjT5ucwVaAAAKizxTP3lGngEAIHfJNPWTaWQaoPlpVAMgY8aMGZO6iN17773jlVdeiUsvvTQOOuig2HLLLaNdu3ZVgklD3qFTaIqKiqqEmw1Dz9y5c2P69OkREbHJJpvEySefnLbj9uzZM8aOHRsLFy6MJ598Mq688sr45je/WeW20u+8804MHTo0br311rQc85xzzolnnnkmIiK22GKLmDhxYnTo0KHObTYMaXvttVcqMDfmqym3QgcAAHmmfvKMPAMAQO6Saeon08g0QPPTqAZAxvztb39LLV9xxRWx6aab1jn+gw8+yHRJOen000+PkpKSiIh44IEHYtWqVRERcd9990WSJBERcdJJJ0XHjh3Tfux27drFt771rbjmmmviqaeeisWLF8dDDz0Ue+65Z2rMZZddFv/5z3826ji33nprKuC1adMmHnnkkejVq1e923Xv3j21/P777zfoXVwAAJAO8kzDyDO1k2cAAMgmmaZhZJrayTRAJmhUAyBjNvzs+l133bXOsRUVFTFz5sxMl5STunfvHscee2xERCxdujQefvjhqKysjHHjxqXGpOuW0vVp165dDB06NKZPnx49evSIiC9vNz116tQm73PSpEnx4x//OPX93XffHQcffHCDtt1rr72itLQ0IiLKysrin//8Z5PrAACAxpBnGkaeqZ08AwBANsk0DSPT1E6mATJBoxoAGbPh59qvXLmyzrGPPvpoLFiwINMl5ayzzjortTx27NiYNm1azJ8/PyIi+vTp0+DQkC6bbbZZHHTQQanvP/300ybt5/XXX4//+q//Sr3L5tJLL43TTjutwdu3a9cuDj300NT3t912W5PqAACAxpJnGk6eqZk8AwBANsk0DSfT1EymATJBoxoAGbPDDjuklh977LFaxy1cuDAuuuii5igpZx1++OHRs2fPiIh4+umn4+qrr06tO+OMM9J2nMWLFzd47Icffpha7tq1a6OPtWjRojj22GOjrKwsIiKGDBkSo0aNavR+fvKTn6SWH3rooSrvYqpPSw7WAABsHHmm4eSZ2skzAABki0zTcDJN7WQaIN00qgGQMcccc0xq+Re/+EX88Y9/rDZmzpw5MWDAgJg/f360b9++OcvLKcXFxamwkyRJzJo1KyIiWrduHaeeemrajnPHHXfEXnvtFXfddVd88sknNY5Zvnx5/OQnP4kXXnghIiJatWoVRx55ZKOOU15eHsOGDYu5c+dGRMQee+wRf/rTn6q8g6uhBgwYEMOHD099P2LEiPjxj39ca6Bbs2ZNPPbYY3HCCSfEkCFDGn08AACIkGcaQ56pnTwDAEC2yDQNJ9PUTqYB0q11tgsAoHCddtppceutt8Y777wTa9asie9///tx4403xp577hlt27aN1157LXWhveeee8aRRx4ZN998c5arzp4RI0bENddcE+vWrUs9duyxx0b37t3TepyXX345zjvvvDj//POjd+/esdtuu8UWW2wR5eXl8fHHH8ezzz4bX3zxRWr8T3/609h2220bdYxnn302/v73v6e+79mzZ/z0pz9t0Lbf//7344ADDqjy2N133x2ffPJJPPHEE5EkSfzyl7+M22+/Pfbbb7/o3bt3tGvXLpYtWxb//ve/49VXX43Vq1dHRMQ+++zTqLoBAGA9eaZx5JkvyTMAAOQKmaZxZJovyTRApmlUAyBjSktLY+LEiXH00UfH+++/HxERb775Zrz55ptVxvXv3z8mTJgQY8aMyUaZOWOrrbaKwYMHV7kFdzpvKR0R0aFDh9RykiTx3nvvxXvvvVfj2DZt2sTPfvazuPLKKxt9nCRJqnz/17/+tcHb7rvvvtVCUGlpaUyaNCmuueaauOWWW2LlypWxdu3amDlzZsycObPG/ZSUlMSBBx7Y6NoBACBCnmkseeZL8gwAALlCpmkcmeZLMg2QaRrVAMionXbaKV588cW488474+GHH46333471q5dGz169Ijdd989TjnllDjppJOidWsvSRERw4YNS4WgrbfeOo466qi07v9HP/pRDBs2LKZNmxbPPvtsvPrqqzFv3rwoKyuL4uLi6Ny5c+yyyy5x6KGHxqmnnhrbb799Wo+/MVq1ahXXXnttXHDBBXH//ffHk08+GW+88UYsWrQoysvLo2PHjrH99tvH7rvvHt/85jdj0KBB0bVr12yXDQBAHpNnGkeeqZ08AwBANsg0jSPT1E6mAdKlKPlqOy0AkDWnnXZa/P73v4+IiJ/97Gdx/fXXZ7kiAACAhpFnAACAfCbTAGSeRjUAyBFlZWWx5ZZbxsqVK6OoqCjee++92GGHHbJdFgAAQL3kGQAAIJ/JNADNozjbBQAAX7r33ntj5cqVERFxxBFHCEAAAEDekGcAAIB8JtMANA93VAOAHDBv3rzYZ599YsmSJRERMXXq1DjiiCOyXBUAAED95BkAACCfyTQAzUejGgBkySWXXBIRER9//HFMnDgxVqxYERER3/zmN+Opp57KZmkAAAB1kmcAAIB8JtMAZIdGNQDIkqKiomqPdevWLWbNmhW9evXKQkUAAAANI88AAAD5TKYByI7ibBcAAC1dq1atYuutt47TTz89Zs+eLQABAAB5Q54BAADymUwD0LzcUQ0AAAAAAAAAAICMckc1AAAAAAAAAAAAMkqjGgAAAAAAAAAAABmlUQ0AAAAAAAAAAICM0qgGAAAAAAAAAABARmlUAwAAAAAAAAAAIKM0qgEAAAAAAAAAAJBRGtUAAAAAAAAAAADIKI1qAAAAAAAAAAAAZNT/Bz3kmOlhqB7UAAAAAElFTkSuQmCC",
      "text/plain": [
       "<Figure size 3000x1000 with 6 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_protein_families(all_dataset=all_dataset)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c3775283",
   "metadata": {},
   "source": [
    "The family size distributions observed in the graphs indicate that the distributions are consistent across the different sets. However, there is a notable prevalence of families with a small number of representatives and a scarcity of families with a large size. This could introduce bias in our neural network and machine learning models, as they may prioritize the more represented classes.\n",
    "\n",
    "To address this issue, one potential solution is to reduce the number of protein families considered in the analysis by retaining only those that are well-represented in the dataset. However, it is important to note that this approach may not be optimal if our objective is to classify the entire proteome, including novel families that have not been previously classified. __Previusly some authors have already presented models featuring the 100 more frequent familes, here we will attempt to keep all familes and weight the NN to remove the overall bias__.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a4dcc95e",
   "metadata": {},
   "source": [
    "### Aminoacid distribution Analysis\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "6109c5f0",
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_protein_families(name: str, dataset: pandas.DataFrame) -> None:\n",
    "    # Extract protein sequences from the dataset\n",
    "    sequences = dataset['sequence']\n",
    "    # Concatenate all sequences into a single string\n",
    "    all_sequences = ''.join(sequences)\n",
    "    # Calculate the frequency of each amino acid\n",
    "    # Series in pandas has index, the value counts \n",
    "    aa_counts = pandas.Series(list(all_sequences)).value_counts()\n",
    "  \n",
    "    cmap = cm.viridis\n",
    "    # Plot the amino acid distribution\n",
    "    plt.figure(figsize=(12, 6))\n",
    "    plt.title(f'Amino Acid Distribution in: {name} dataset', fontsize=24)\n",
    "    plt.xlabel('Amino Acid Label', fontsize=20)\n",
    "    plt.ylabel('Frequency', fontsize=20)\n",
    "    plt.tick_params(axis='both', which='both', length=6, labelsize=16)  \n",
    "    # Format y-axis ticks in millions\n",
    "    formatter = ScalarFormatter(useMathText=True)\n",
    "    formatter.set_scientific(True)\n",
    "    formatter.set_powerlimits((-1, 1))\n",
    "    plt.gca().yaxis.set_major_formatter(formatter)\n",
    "    # Plot\n",
    "    plt.bar(aa_counts.index, aa_counts.values, color=cmap(0.7))\n",
    "\n",
    "    # Create inset plot for magnified view\n",
    "    ax_inset = inset_axes(plt.gca(), width=\"40%\", height=\"50%\", loc='upper right')\n",
    "    ax_inset.set_xlabel('Amino Acid', fontsize=12)\n",
    "    ax_inset.set_ylabel('Frequency', fontsize=12)\n",
    "    ax_inset.tick_params(axis='both', which='both', length=4, labelsize=10) \n",
    "    ax_inset.bar(aa_counts.index, aa_counts.values, color=cmap(0.8))\n",
    "    # Specify the threshold for magnification\n",
    "    threshold = aa_counts.max() * 0.001\n",
    "    # Set the limits for the magnified view\n",
    "    ax_inset.set_ylim(aa_counts[aa_counts < threshold].values[-1], aa_counts[aa_counts < threshold].values[0])\n",
    "    ax_inset.set_xlim(aa_counts[aa_counts < threshold].index[0], aa_counts[aa_counts < threshold].index[-1])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "e9d354d7",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1200x600 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_protein_families(name='test', dataset=test_dataset)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8eed3415",
   "metadata": {},
   "source": [
    "Here we have examined the frequency of each amino acid in the protein sequences. According to the dataset description, there are 20 commonly occurring amino acids and 4 uncommon ones (X, U, B, Z). We see that Z and B are extremely rare.  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "af419da3",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1200x600 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_protein_families(name='train', dataset=train_dataset)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "68e3d757",
   "metadata": {},
   "source": [
    "In contrast to the previous analysis, we observe that the amino acid distribution varies between the training and test datasets, without reducing the number of families included. However, this heterogeneity in distribution should not pose a significant problem, as the rarest amino acids are expected to have a minor impact in the classification task."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "132a01bc",
   "metadata": {},
   "source": [
    "## Method\n",
    "\n",
    "We can draw inspiration from existing literature to develop a convolutional neural network (CNN) for the task of protein annotation based on sequence information. This task shares similarities with natural language processing, where the meaning of words in a sentence depends on the context provided by other words. Similarly, in image processing, features are recognized based on the non-local topology of the image. In the case of protein sequences, each amino acid carries information that is dependent on the other amino acids present in the peptide string. By applying convolutional procedures, we can encode information in a non-local manner, considering the interdependencies between amino acids rather than focusing on individual amino acids alone."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "260d52ad",
   "metadata": {},
   "outputs": [],
   "source": [
    "def amino_acids_dict() -> dict:\n",
    "    # Encode the amino acids using one-hot encoding\n",
    "    amino_acids = 'ACDEFGHIKLMNPQRSTVWYXUBOZ'  # List of amino acids\n",
    "    num_classes = len(amino_acids)\n",
    "\n",
    "    # Create a dictionary to map amino acids to indices\n",
    "    dict_to_index = {aa: i for i, aa in enumerate(amino_acids)}\n",
    "    return dict_to_index"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "8cfd15a9",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'A': 0, 'C': 1, 'D': 2, 'E': 3, 'F': 4, 'G': 5, 'H': 6, 'I': 7, 'K': 8, 'L': 9, 'M': 10, 'N': 11, 'P': 12, 'Q': 13, 'R': 14, 'S': 15, 'T': 16, 'V': 17, 'W': 18, 'Y': 19, 'X': 20, 'U': 21, 'B': 22, 'O': 23, 'Z': 24}\n"
     ]
    }
   ],
   "source": [
    "dict_amino_acids = amino_acids_dict()\n",
    "print(dict_amino_acids)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e27b5350",
   "metadata": {},
   "source": [
    "### Econde Sequences\n",
    "\n",
    "Here we encode the sequence of amino acids for the protein family selected. We need to reduce the overall size of the dataset otherwise we would require a huge ammount of memory."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "992605e2",
   "metadata": {},
   "outputs": [],
   "source": [
    "# In the following I have faced some severe memory issue using the whole dataset. We cut some of the data. \n",
    "top_k = 100\n",
    "def reduce_set_size(all_dataset: dict):\n",
    "    \n",
    "    most_freq_families = set(all_dataset['train']['family_accession'].value_counts()[:top_k].index)\n",
    "    \n",
    "    for name, df in all_dataset.items():\n",
    "        all_dataset[name] = df[df['family_accession'].isin(most_freq_families)]\n",
    "    \n",
    "    return all_dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "8a7f4b9a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>dataset</th>\n",
       "      <th>num_samples</th>\n",
       "      <th>num_families</th>\n",
       "      <th>min_samples_per_family</th>\n",
       "      <th>min_family</th>\n",
       "      <th>max_samples_per_family</th>\n",
       "      <th>max_family</th>\n",
       "      <th>mean_samples_per_family</th>\n",
       "      <th>name</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>train</td>\n",
       "      <td>92983</td>\n",
       "      <td>100</td>\n",
       "      <td>708</td>\n",
       "      <td>PF02739.16</td>\n",
       "      <td>3637</td>\n",
       "      <td>PF13649.6</td>\n",
       "      <td>929.83</td>\n",
       "      <td>train</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>test</td>\n",
       "      <td>11565</td>\n",
       "      <td>100</td>\n",
       "      <td>88</td>\n",
       "      <td>PF01250.17</td>\n",
       "      <td>454</td>\n",
       "      <td>PF13649.6</td>\n",
       "      <td>115.65</td>\n",
       "      <td>test</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>dev</td>\n",
       "      <td>11565</td>\n",
       "      <td>100</td>\n",
       "      <td>88</td>\n",
       "      <td>PF01250.17</td>\n",
       "      <td>454</td>\n",
       "      <td>PF13649.6</td>\n",
       "      <td>115.65</td>\n",
       "      <td>dev</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  dataset num_samples num_families min_samples_per_family  min_family  \\\n",
       "0   train       92983          100                    708  PF02739.16   \n",
       "1    test       11565          100                     88  PF01250.17   \n",
       "2     dev       11565          100                     88  PF01250.17   \n",
       "\n",
       "  max_samples_per_family max_family  mean_samples_per_family   name  \n",
       "0                   3637  PF13649.6                   929.83  train  \n",
       "1                    454  PF13649.6                   115.65   test  \n",
       "2                    454  PF13649.6                   115.65    dev  "
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "all_dataset = reduce_set_size(all_dataset)\n",
    "get_datasets_information(all_dataset)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "89a61b43",
   "metadata": {},
   "outputs": [],
   "source": [
    "def convert_tensor(dataset: list):\n",
    "    return tensorflow.convert_to_tensor(dataset, dtype=tensorflow.float32)\n",
    "\n",
    "def reshape_list_and_convert(dataset: list) -> tensorflow.Tensor:\n",
    "    dataset = dataset.reshape(dataset.shape[0], dataset.shape[1], 1)\n",
    "    converted_dataset = convert_tensor(dataset) \n",
    "    return converted_dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "8a91c692",
   "metadata": {},
   "outputs": [],
   "source": [
    "from typing import Tuple\n",
    "\n",
    "def encode_sequence(all_dataset: dict, dict_amino_acids: dict, max_length: int) -> Tuple[numpy.ndarray]:\n",
    "    encoded_list = []\n",
    "    for n, (name, dataframe) in enumerate(all_dataset.items()):\n",
    "        print(f'encoding {name}')\n",
    "        set_list = []\n",
    "        for sequence in dataframe['sequence']:\n",
    "            sequence_list = [0 for i in range(max_length)]\n",
    "            for index, amino_acid in enumerate(list(sequence[:max_length])):\n",
    "                sequence_list[index] = dict_amino_acids[amino_acid]\n",
    "            set_list.append(numpy.array(sequence_list))\n",
    "        \n",
    "        encoded_list.append(numpy.array(set_list))\n",
    "    return encoded_list[0], encoded_list[1], encoded_list[2]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "d4f3b51f",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "encoding train\n",
      "encoding test\n",
      "encoding dev\n"
     ]
    }
   ],
   "source": [
    "# Encode the sequences in the partitions\n",
    "train_encoded, test_encoded, dev_encoded = encode_sequence(all_dataset=all_dataset, dict_amino_acids=dict_amino_acids, max_length=100)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "f9a90d45",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'numpy.ndarray'> \n",
      " [17  9  2 17  0  1  5 16  1  2 17  0 10  3  0 14 11 13 16  5  2  0  0  4\n",
      "  7  7  5 16  2  4 15 12  5 10  9 16  9  5  9 13  8  9  8  8 11 14 14  4\n",
      "  0 16  7 12  9 17  1  0 11  0  9  0  9 12  4 13 15 16  6  4  2  0 17  9\n",
      "  7  0  4  5  7 14 11  7 10  2 14  8  5  0  9  8 13  4  6  2  0  9  8 12\n",
      "  5  5  0  0]\n"
     ]
    }
   ],
   "source": [
    "# Encoded sequence of ammino acid\n",
    "print(type(test_encoded), '\\n', test_encoded[0])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d6a61286",
   "metadata": {},
   "source": [
    "### Encode Labels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "085d4f5d",
   "metadata": {},
   "outputs": [],
   "source": [
    "def dict_family(labels: list) -> Tuple[dict, int]:\n",
    "    dict_family_accession = {}\n",
    "    all_family_accession = set(labels)\n",
    "    for i, family_accession in enumerate(all_family_accession):\n",
    "        dict_family_accession[family_accession] = i\n",
    "    return dict_family_accession, i+1 \n",
    "        \n",
    "def encode_labels(all_dataset: dict, dict_family_accession: dict, num_classes: int) -> Tuple[tensorflow.Tensor, tensorflow.Tensor, tensorflow.Tensor]:\n",
    "    encoded_list = []\n",
    "    for name, dataframe in all_dataset.items():\n",
    "        print(f'encoding {name}')\n",
    "        labels = dataframe['family_accession']\n",
    "        labels_array = numpy.array([dict_family_accession[label] for label in labels])\n",
    "        #print(labels.iloc[0:5], labels_array[0:5])\n",
    "        encoded_labels = tensorflow.keras.utils.to_categorical(labels_array, num_classes=num_classes)\n",
    "        encoded_list.append(encoded_labels)\n",
    "        \n",
    "    return encoded_list[0], encoded_list[1], encoded_list[2]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "5aefc30f",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'PF02873.16': 0, 'PF03462.18': 1, 'PF00763.23': 2, 'PF01687.17': 3, 'PF01339.17': 4, 'PF14748.6': 5, 'PF00560.33': 6, 'PF00231.19': 7, 'PF02885.17': 8, 'PF00677.17': 9, 'PF02787.19': 10, 'PF00472.20': 11, 'PF01817.21': 12, 'PF01458.17': 13, 'PF02397.16': 14, 'PF07516.13': 15, 'PF00162.19': 16, 'PF00400.32': 17, 'PF02777.18': 18, 'PF01330.21': 19, 'PF01288.20': 20, 'PF02482.19': 21, 'PF01297.17': 22, 'PF13976.6': 23, 'PF01769.16': 24, 'PF02132.15': 25, 'PF00023.30': 26, 'PF10035.9': 27, 'PF01715.17': 28, 'PF02559.16': 29, 'PF02660.15': 30, 'PF01367.20': 31, 'PF00137.21': 32, 'PF01250.17': 33, 'PF01523.16': 34, 'PF01255.19': 35, 'PF00586.24': 36, 'PF00684.19': 37, 'PF13692.6': 38, 'PF01035.20': 39, 'PF02537.15': 40, 'PF13649.6': 41, 'PF02021.17': 42, 'PF02729.21': 43, 'PF04122.12': 44, 'PF02502.18': 45, 'PF00391.23': 46, 'PF08742.11': 47, 'PF02843.16': 48, 'PF05192.18': 49, 'PF02580.16': 50, 'PF00344.20': 51, 'PF17802.1': 52, 'PF01066.21': 53, 'PF00920.21': 54, 'PF01641.18': 55, 'PF02673.18': 56, 'PF03948.14': 57, 'PF01195.19': 58, 'PF01556.18': 59, 'PF13932.6': 60, 'PF02152.18': 61, 'PF01725.16': 62, 'PF00842.21': 63, 'PF02417.15': 64, 'PF03458.13': 65, 'PF13439.6': 66, 'PF13545.6': 67, 'PF02844.15': 68, 'PF02881.19': 69, 'PF04205.14': 70, 'PF02545.14': 71, 'PF01799.20': 72, 'PF05697.13': 73, 'PF00367.20': 74, 'PF06580.13': 75, 'PF02817.17': 76, 'PF02618.16': 77, 'PF13508.7': 78, 'PF02410.15': 79, 'PF10396.9': 80, 'PF00126.27': 81, 'PF13472.6': 82, 'PF07664.12': 83, 'PF01368.20': 84, 'PF03484.15': 85, 'PF01384.20': 86, 'PF02033.18': 87, 'PF02739.16': 88, 'PF04002.15': 89, 'PF03453.17': 90, 'PF01040.18': 91, 'PF13927.6': 92, 'PF17853.1': 93, 'PF00381.19': 94, 'PF00742.19': 95, 'PF02978.19': 96, 'PF13288.6': 97, 'PF01765.19': 98, 'PF07479.14': 99} {'PF02873.16': 0, 'PF03462.18': 1, 'PF00763.23': 2, 'PF01687.17': 3, 'PF01339.17': 4, 'PF14748.6': 5, 'PF00560.33': 6, 'PF00231.19': 7, 'PF02885.17': 8, 'PF00677.17': 9, 'PF02787.19': 10, 'PF00472.20': 11, 'PF01817.21': 12, 'PF01458.17': 13, 'PF02397.16': 14, 'PF07516.13': 15, 'PF00162.19': 16, 'PF00400.32': 17, 'PF02777.18': 18, 'PF01330.21': 19, 'PF01288.20': 20, 'PF02482.19': 21, 'PF01297.17': 22, 'PF13976.6': 23, 'PF01769.16': 24, 'PF02132.15': 25, 'PF00023.30': 26, 'PF10035.9': 27, 'PF01715.17': 28, 'PF02559.16': 29, 'PF02660.15': 30, 'PF01367.20': 31, 'PF00137.21': 32, 'PF01250.17': 33, 'PF01523.16': 34, 'PF01255.19': 35, 'PF00586.24': 36, 'PF00684.19': 37, 'PF13692.6': 38, 'PF01035.20': 39, 'PF02537.15': 40, 'PF13649.6': 41, 'PF02021.17': 42, 'PF02729.21': 43, 'PF04122.12': 44, 'PF02502.18': 45, 'PF00391.23': 46, 'PF08742.11': 47, 'PF02843.16': 48, 'PF05192.18': 49, 'PF02580.16': 50, 'PF00344.20': 51, 'PF17802.1': 52, 'PF01066.21': 53, 'PF00920.21': 54, 'PF01641.18': 55, 'PF02673.18': 56, 'PF03948.14': 57, 'PF01195.19': 58, 'PF01556.18': 59, 'PF13932.6': 60, 'PF02152.18': 61, 'PF01725.16': 62, 'PF00842.21': 63, 'PF02417.15': 64, 'PF03458.13': 65, 'PF13439.6': 66, 'PF13545.6': 67, 'PF02844.15': 68, 'PF02881.19': 69, 'PF04205.14': 70, 'PF02545.14': 71, 'PF01799.20': 72, 'PF05697.13': 73, 'PF00367.20': 74, 'PF06580.13': 75, 'PF02817.17': 76, 'PF02618.16': 77, 'PF13508.7': 78, 'PF02410.15': 79, 'PF10396.9': 80, 'PF00126.27': 81, 'PF13472.6': 82, 'PF07664.12': 83, 'PF01368.20': 84, 'PF03484.15': 85, 'PF01384.20': 86, 'PF02033.18': 87, 'PF02739.16': 88, 'PF04002.15': 89, 'PF03453.17': 90, 'PF01040.18': 91, 'PF13927.6': 92, 'PF17853.1': 93, 'PF00381.19': 94, 'PF00742.19': 95, 'PF02978.19': 96, 'PF13288.6': 97, 'PF01765.19': 98, 'PF07479.14': 99}\n"
     ]
    }
   ],
   "source": [
    "dict_family_accession, num_family_accession = dict_family(all_dataset['test']['family_accession'])\n",
    "print(dict_family_accession, dict_family_accession)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "c658dfa8",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "encoding train\n",
      "encoding test\n",
      "encoding dev\n"
     ]
    }
   ],
   "source": [
    "train_labels, test_labels, dev_labels = encode_labels(all_dataset=all_dataset, dict_family_accession=dict_family_accession, num_classes=num_family_accession)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "420b7833",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "family_id                                             Methyltransf_25\n",
      "sequence_name                                     G8YSU8_PICSO/41-136\n",
      "family_accession                                            PF13649.6\n",
      "aligned_sequence    VLDVGCGP.G...S..IT...VD..IA.SR.......VP..K.......\n",
      "sequence            VLDVGCGPGSITVDIASRVPKGHVTGIDTFEELISQGQQKASDLKQ...\n",
      "Name: 4, dtype: object \n",
      "\n",
      "\n",
      "family_id                                             Methyltransf_25\n",
      "sequence_name                                     D1C4C3_SPHTD/46-139\n",
      "family_accession                                            PF13649.6\n",
      "aligned_sequence    VLDLGCAN.G...A..TS...RA..LA.DL............G......\n",
      "sequence            VLDLGCANGATSRALADLGARVTGVDVSARLIELARQREAARPRGV...\n",
      "Name: 57, dtype: object\n"
     ]
    }
   ],
   "source": [
    "print(all_dataset['train'].iloc[1], '\\n\\n') # 'PF13649.6': 86\n",
    "print(all_dataset['dev'].iloc[4])           # 'PF13649.6': 86"
   ]
  },
  {
   "cell_type": "raw",
   "id": "71dfcc66",
   "metadata": {},
   "source": [
    "for i, value in enumerate(all_dataset['train']['family_accession']):\n",
    "    print(value, i)\n",
    "    if \"PF13649.6\" in value:\n",
    "        print('****')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "03ad2874",
   "metadata": {},
   "source": [
    "### CNN model\n",
    "\n",
    "My first CNN model was simple: \n",
    "1 convolutional layer;\n",
    "1 pool layer; \n",
    "1 flatten layer;\n",
    "2 dense layers one with relu activation and 1 with softmax activation \n",
    "\n",
    "However such a simple model was not giving the accuracy I was seeking for the validation set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "9b7d4fe2",
   "metadata": {},
   "outputs": [],
   "source": [
    "class CNNModel(tensorflow.keras.Model):\n",
    "    def __init__(self, num_family_accession):\n",
    "        super(CNNModel, self).__init__()\n",
    "        self.conv1 = tensorflow.keras.layers.Conv1D(filters=16, kernel_size=7, activation='relu')\n",
    "        self.pool1 = tensorflow.keras.layers.MaxPooling1D(pool_size=2)\n",
    "        self.conv2 = tensorflow.keras.layers.Conv1D(filters=32, kernel_size=5, activation='relu')\n",
    "        self.pool2 = tensorflow.keras.layers.MaxPooling1D(pool_size=2)\n",
    "        self.conv3 = tensorflow.keras.layers.Conv1D(filters=64, kernel_size=3, activation='relu')\n",
    "        self.pool3 = tensorflow.keras.layers.MaxPooling1D(pool_size=2)\n",
    "        self.flatten = tensorflow.keras.layers.Flatten()\n",
    "        self.dense1 = tensorflow.keras.layers.Dense(units=128, activation='relu')\n",
    "        self.dense2 = tensorflow.keras.layers.Dense(units=num_family_accession, activation='softmax')\n",
    "\n",
    "    def call(self, inputs):\n",
    "        x = self.conv1(inputs)\n",
    "        x = self.pool1(x)\n",
    "        x = self.conv2(x)\n",
    "        x = self.pool2(x)\n",
    "        x = self.conv3(x)\n",
    "        x = self.pool3(x)\n",
    "        x = self.flatten(x)\n",
    "        x = self.dense1(x)\n",
    "        output = self.dense2(x)\n",
    "        return output\n",
    "\n",
    "\n",
    "CNN = CNNModel(num_family_accession)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "24549086",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Compile the model\n",
    "CNN.compile(optimizer='adam',\n",
    "              loss='categorical_crossentropy',\n",
    "              metrics=['accuracy'])\n",
    "\n",
    "# Train the model\n",
    "train_encoded_tensor = reshape_list_and_convert(train_encoded)\n",
    "train_labels_tensor = convert_tensor(train_labels)\n",
    "\n",
    "dev_encoded_tensor = reshape_list_and_convert(dev_encoded)\n",
    "dev_labels_tensor = convert_tensor(dev_labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "2a1a4f31",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Shape: tf.Tensor([92983   100     1], shape=(3,), dtype=int32)\n",
      "Shape: tf.Tensor([92983   100], shape=(2,), dtype=int32)\n",
      "Shape: tf.Tensor([11565   100     1], shape=(3,), dtype=int32)\n",
      "Shape: tf.Tensor([11565   100], shape=(2,), dtype=int32)\n"
     ]
    }
   ],
   "source": [
    "print(\"Shape:\", tensorflow.shape(train_encoded_tensor))\n",
    "print(\"Shape:\", tensorflow.shape(train_labels_tensor))\n",
    "print(\"Shape:\", tensorflow.shape(dev_encoded_tensor))\n",
    "print(\"Shape:\", tensorflow.shape(dev_labels_tensor))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "3768f17d",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023-05-24 00:41:12.099201: I tensorflow/compiler/mlir/mlir_graph_optimization_pass.cc:116] None of the MLIR optimization passes are enabled (registered 2)\n",
      "2023-05-24 00:41:12.099536: I tensorflow/core/platform/profile_utils/cpu_utils.cc:112] CPU Frequency: 2688000000 Hz\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/10\n",
      "2906/2906 [==============================] - 8s 3ms/step - loss: 2.4695 - accuracy: 0.4156 - val_loss: 1.1186 - val_accuracy: 0.7179\n",
      "Epoch 2/10\n",
      "2906/2906 [==============================] - 8s 3ms/step - loss: 0.9873 - accuracy: 0.7469 - val_loss: 0.8989 - val_accuracy: 0.7677\n",
      "Epoch 3/10\n",
      "2906/2906 [==============================] - 9s 3ms/step - loss: 0.7636 - accuracy: 0.7998 - val_loss: 0.7601 - val_accuracy: 0.8011\n",
      "Epoch 4/10\n",
      "2906/2906 [==============================] - 10s 4ms/step - loss: 0.6282 - accuracy: 0.8325 - val_loss: 0.6592 - val_accuracy: 0.8285\n",
      "Epoch 5/10\n",
      "2906/2906 [==============================] - 10s 3ms/step - loss: 0.5313 - accuracy: 0.8595 - val_loss: 0.6582 - val_accuracy: 0.8321\n",
      "Epoch 6/10\n",
      "2906/2906 [==============================] - 10s 4ms/step - loss: 0.4642 - accuracy: 0.8744 - val_loss: 0.5759 - val_accuracy: 0.8500\n",
      "Epoch 7/10\n",
      "2906/2906 [==============================] - 10s 3ms/step - loss: 0.4044 - accuracy: 0.8904 - val_loss: 0.5351 - val_accuracy: 0.8682\n",
      "Epoch 8/10\n",
      "2906/2906 [==============================] - 10s 3ms/step - loss: 0.3623 - accuracy: 0.9002 - val_loss: 0.5075 - val_accuracy: 0.8696\n",
      "Epoch 9/10\n",
      "2906/2906 [==============================] - 11s 4ms/step - loss: 0.3145 - accuracy: 0.9132 - val_loss: 0.5163 - val_accuracy: 0.8749\n",
      "Epoch 10/10\n",
      "2906/2906 [==============================] - 10s 4ms/step - loss: 0.2923 - accuracy: 0.9161 - val_loss: 0.4912 - val_accuracy: 0.8805\n"
     ]
    }
   ],
   "source": [
    "history = CNN.fit(train_encoded_tensor, train_labels_tensor, validation_data=(dev_encoded_tensor, dev_labels_tensor), epochs=10, batch_size=32)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "id": "cdcbaf91",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'Loss')"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plotting the loss\n",
    "cmap = cm.viridis\n",
    "\n",
    "plt.plot(history.history['loss'], color=cmap(0.7), linewidth=4, label='Train')\n",
    "plt.plot(history.history['val_loss'], color=cmap(0.4), linewidth=4, label='Validation')\n",
    "plt.title('Model Loss', fontsize=18)\n",
    "plt.xlabel('Epoch', fontsize=14)\n",
    "plt.ylabel('Loss', fontsize=14)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "id": "7e9d1b0f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'Accuracy')"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plotting the accuracy\n",
    "plt.plot(history.history['accuracy'], color=cmap(0.7), linewidth=4, label='Train')\n",
    "plt.plot(history.history['val_accuracy'], color=cmap(0.4), linewidth=4, label='Validation')\n",
    "plt.title('Model Accuracy', fontsize=18)\n",
    "plt.xlabel('Epoch', fontsize=14)\n",
    "plt.ylabel('Accuracy', fontsize=14)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "04fed36f",
   "metadata": {},
   "source": [
    "The CNN model demonstrates remarkable accuracy on the __validation set__, indicating that it has successfully learned the significant and meaningful features within the dataset __without overfitting__."
   ]
  },
  {
   "cell_type": "raw",
   "id": "610b69a0",
   "metadata": {},
   "source": [
    "tensorflow.keras.utils.plot_model(CNN, show_shapes=True, show_layer_names=True)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7a030cd2",
   "metadata": {},
   "source": [
    "I have used a __CNN as deep learning architecture for protein family classification__. It consists of convolutional, pooling, and dense layers that learn features from protein sequences. CNNs are beneficial for capturing __local patterns and motifs__, aiding in accurate classification. However, CNNs may struggle with complex relationships and interpretability. Further experiments can compare the CNN's performance with methods like BLAST and HMMs. Adjusting the CNN's architecture, hyperparameters, and exploring transfer learning can optimize its performance. \n",
    "\n",
    "Summary of future experiments and methods:\n",
    "\n",
    "Compare CNN performance with BLAST and HMMs for sequence analysis.\n",
    "\n",
    "Experiment with different CNN architectures, filter sizes, pooling strategies, and regularization techniques.\n",
    "\n",
    "Fine-tune hyperparameters to enhance the CNN's predictive capabilities.\n",
    "\n",
    "Explore transfer learning by leveraging pre-trained CNN models on large-scale protein datasets.\n",
    "\n",
    "Consider data augmentation techniques to increase the diversity of the training set.\n",
    "\n",
    "Explore hybrid approaches combining CNNs with other machine learning models or techniques.\n",
    "\n",
    "Conduct comparative studies on different protein classification tasks to assess the generalizability of the CNN."
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.16"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}