Figures/Figure 3A-C.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "6fd8a1e8",
   "metadata": {},
   "source": [
    "### Interpretation of the LR models\n",
    "\n",
    "Here, we compare the coefficients learned by the different linear regression models.\n",
    "\n",
    "These coefficients can be downloaded from Zenodo."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "b76aae10",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "\n",
    "import seaborn as sns\n",
    "from matplotlib import pyplot as plt\n",
    "\n",
    "import matplotlib\n",
    "matplotlib.rcParams['pdf.fonttype'] = 42"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "33324f65",
   "metadata": {},
   "source": [
    "### Read and filter weights"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "a2911add",
   "metadata": {},
   "outputs": [],
   "source": [
    "h_glia_all = pd.read_csv('../../PSI_prediction/Zenodo/Human/HPC/LR_weights/LR_all_glia.csv', index_col=0)\n",
    "h_neur_all = pd.read_csv('../../PSI_prediction/Zenodo/Human/HPC/LR_weights/LR_all_neur.csv', index_col=0)\n",
    "\n",
    "h_glia_var = pd.read_csv('../../PSI_prediction/Zenodo/Human/HPC/LR_weights/LR_var01_glia.csv', index_col=0)\n",
    "h_neur_var = pd.read_csv('../../PSI_prediction/Zenodo/Human/HPC/LR_weights/LR_var01_neur.csv', index_col=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "e6f73dbe",
   "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>HPC - glia (all)</th>\n",
       "      <th>HPC - neur (all)</th>\n",
       "      <th>HPC - glia (var)</th>\n",
       "      <th>HPC - neur (var)</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>AATF_start</th>\n",
       "      <td>0.001789</td>\n",
       "      <td>0.002657</td>\n",
       "      <td>-0.028788</td>\n",
       "      <td>-0.015118</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ABCF1_start</th>\n",
       "      <td>-0.008563</td>\n",
       "      <td>-0.001489</td>\n",
       "      <td>-0.058113</td>\n",
       "      <td>0.005881</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>AKAP1_start</th>\n",
       "      <td>0.014131</td>\n",
       "      <td>0.022668</td>\n",
       "      <td>-0.001819</td>\n",
       "      <td>0.004053</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>APOBEC3C_start</th>\n",
       "      <td>0.000020</td>\n",
       "      <td>0.000248</td>\n",
       "      <td>-0.003887</td>\n",
       "      <td>-0.027109</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>AQR_start</th>\n",
       "      <td>-0.010190</td>\n",
       "      <td>-0.012180</td>\n",
       "      <td>-0.005953</td>\n",
       "      <td>-0.018397</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                HPC - glia (all)  HPC - neur (all)  HPC - glia (var)  \\\n",
       "AATF_start              0.001789          0.002657         -0.028788   \n",
       "ABCF1_start            -0.008563         -0.001489         -0.058113   \n",
       "AKAP1_start             0.014131          0.022668         -0.001819   \n",
       "APOBEC3C_start          0.000020          0.000248         -0.003887   \n",
       "AQR_start              -0.010190         -0.012180         -0.005953   \n",
       "\n",
       "                HPC - neur (var)  \n",
       "AATF_start             -0.015118  \n",
       "ABCF1_start             0.005881  \n",
       "AKAP1_start             0.004053  \n",
       "APOBEC3C_start         -0.027109  \n",
       "AQR_start              -0.018397  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "w = pd.concat((h_glia_all, h_neur_all, h_glia_var, h_neur_var), axis=1)\n",
    "w.columns = ['HPC - glia (all)', 'HPC - neur (all)', 'HPC - glia (var)', 'HPC - neur (var)']\n",
    "w.iloc[:5]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "19a1de3c",
   "metadata": {},
   "source": [
    "Read the RBP counts, so we can use this to filter the RBPs. We will only keep RBPs with at least 50 binding sites. If there are few binding sites, the coefficient of the RBP might not be estimated with enough confidence."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "0b2ff6b4",
   "metadata": {},
   "outputs": [],
   "source": [
    "# HPC\n",
    "PSI_glia = pd.read_csv('../../PSI_prediction/Zenodo/Human/HPC/PSI/PSI_glia_norm.csv', index_col=0)\n",
    "PSI_neur = pd.read_csv('../../PSI_prediction/Zenodo/Human/HPC/PSI/PSI_neur_norm.csv', index_col=0)\n",
    "\n",
    "# Check which exons are variable\n",
    "exons_all = PSI_glia['0'].notna() & PSI_neur['0'].notna()\n",
    "exons_var01 = PSI_glia.index[(np.abs(PSI_glia['0'] - PSI_neur['0']) > 0.1)]\n",
    "\n",
    "RBP = pd.read_csv('../../PSI_prediction/Zenodo/Human/HPC/RBP/RBP_peaks.csv', index_col=0)\n",
    "RBP_hpc_var01 = np.sum(RBP.loc[exons_var01] > 0)\n",
    "RBP_hpc_all = np.sum(RBP.loc[exons_all] > 0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "e412f6a0",
   "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>HPC (all)</th>\n",
       "      <th>HPC (var)</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>AATF_start</th>\n",
       "      <td>1545</td>\n",
       "      <td>381</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ABCF1_start</th>\n",
       "      <td>1043</td>\n",
       "      <td>260</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>AKAP1_start</th>\n",
       "      <td>3071</td>\n",
       "      <td>766</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>APOBEC3C_start</th>\n",
       "      <td>2358</td>\n",
       "      <td>528</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>AQR_start</th>\n",
       "      <td>11642</td>\n",
       "      <td>2759</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                HPC (all)  HPC (var)\n",
       "AATF_start           1545        381\n",
       "ABCF1_start          1043        260\n",
       "AKAP1_start          3071        766\n",
       "APOBEC3C_start       2358        528\n",
       "AQR_start           11642       2759"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "RBP_counts = pd.concat((RBP_hpc_all, RBP_hpc_var01), axis=1)\n",
    "RBP_counts.columns = ['HPC (all)', 'HPC (var)']\n",
    "RBP_counts.iloc[:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "934cd08f",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(732, 4)\n",
      "(618, 4)\n",
      "(191, 4)\n"
     ]
    }
   ],
   "source": [
    "print(np.shape(w))\n",
    "\n",
    "# There should be enough binding sites to estimate a coefficient \n",
    "w_filt = w[np.min(RBP_counts.iloc[:,:4], axis=1) >= 50] \n",
    "print(np.shape(w_filt))\n",
    "\n",
    "# There should be at least one model where a feature has a bit a high value, otherwise it's not interesting in the heatmap\n",
    "w_filt2 = w_filt[np.max(np.abs(w_filt), axis=1) > 0.05]\n",
    "print(np.shape(w_filt2))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7a464dce",
   "metadata": {},
   "source": [
    "#### Figure 3A"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "92b63326",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 900x150 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.clustermap(w_filt2.T, figsize=(9,1.5), dendrogram_ratio=[0.02,.3],\n",
    "              cmap='vlag', method='complete',\n",
    "               center=0, \n",
    "               vmin=-0.2, vmax=0.2,\n",
    "               cbar_pos=(0.85, 0.9, 0.1, 0.1), \n",
    "                   cbar_kws=dict(orientation='horizontal'),\n",
    "              xticklabels=False)\n",
    "# plt.savefig('../../Evaluate models/Figures/Heatmap.pdf', bbox_inches='tight')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "010b1e2c",
   "metadata": {},
   "source": [
    "#### Figure 3B - hnRNP"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "1d167c04",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 900x300 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "w_filt_genes = pd.DataFrame(w_filt.index)[0].str.split('_', n=1, expand=True)\n",
    "\n",
    "hnrnp = ['HNRNPA1', 'HNRNPC', 'HNRNPL', 'HNRNPM', 'HNRNPU', 'HNRNPUL1',\n",
    "       'PCBP1','PCBP2', 'PTBP1']\n",
    "\n",
    "sns.clustermap(w_filt.iloc[np.isin(w_filt_genes[0], hnrnp)].T, figsize=(9,3), dendrogram_ratio=[0.02,.3],\n",
    "              cmap='vlag', method='complete',\n",
    "               center=0, \n",
    "               cbar_pos=(0.9, 0.9, 0.0, 0.), \n",
    "               vmin=-0.2, vmax=0.2, cbar=False,\n",
    "              xticklabels=['HNRNPA1 - upstream', 'HNRNPC - upstream', \n",
    "                           'HNRNPL - upstream', 'HNRNPM - upstream',\n",
    "                           'HNRNPU - upstream', 'HNRNPUL1 - upstream', \n",
    "                           'PCBP1 - upstream', 'PCBP2 - upstream',\n",
    "                           'PTBP1 - upstream', \"HNRNPA1 - 3' SS\", \n",
    "                           \"HNRNPC - 3' SS\",\n",
    "                           \"HNRNPL - 3' SS\", \n",
    "                           \"HNRNPM - 3' SS\", \n",
    "                           \"HNRNPU - 3' SS\",\n",
    "                           \"HNRNPUL1 - 3' SS\", \n",
    "                           \"PCBP1 - 3' SS\", \n",
    "                           \"PCBP2 - 3' SS\",\n",
    "                           \"PTBP1 - 3' SS\", 'HNRNPA1 - exon', \n",
    "                           'HNRNPC - exon', 'HNRNPL - exon',\n",
    "                           'HNRNPM - exon', 'HNRNPU - exon', \n",
    "                           'HNRNPUL1 - exon', 'PCBP1 - exon',\n",
    "                           'PCBP2 - exon', 'PTBP1 - exon', 'HNRNPA1 - downstream',\n",
    "                           'HNRNPC - downstream', 'HNRNPL - downstream',\n",
    "                           'HNRNPM - downstream', 'HNRNPU - downstream', 'HNRNPUL1 - downstream',\n",
    "                           'PCBP1 - downstream', 'PCBP2 - downstream',\n",
    "                           'PTBP1 - downstream', \"HNRNPA1 - 5' SS\", \n",
    "                           \"HNRNPC - 5' SS\",\n",
    "                           \"HNRNPL - 5' SS\", \"HNRNPM - 5' SS\",\n",
    "                           \"HNRNPU - 5' SS\", \"HNRNPUL1 - 5' SS\", \n",
    "                           \"PCBP1 - 5' SS\", \"PCBP2 - 5' SS\",\n",
    "                           \"PTBP1 - 5' SS\"]\n",
    "              )\n",
    "# plt.savefig('../../Evaluate models/Figures/Heatmap_HNRNP.pdf', bbox_inches='tight')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7c50cc08",
   "metadata": {},
   "source": [
    "#### Figure 3C - cell type specific RBPs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "9ceb3137",
   "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>HPC - glia (all)</th>\n",
       "      <th>HPC - neur (all)</th>\n",
       "      <th>HPC - glia (var)</th>\n",
       "      <th>HPC - neur (var)</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>QKI_start_overlap</th>\n",
       "      <td>-0.098697</td>\n",
       "      <td>-0.004609</td>\n",
       "      <td>-0.154732</td>\n",
       "      <td>0.121338</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MATR3_start_overlap</th>\n",
       "      <td>-0.156540</td>\n",
       "      <td>-0.024425</td>\n",
       "      <td>-0.190873</td>\n",
       "      <td>-0.000462</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>U2AF2_exon_complete</th>\n",
       "      <td>-0.003861</td>\n",
       "      <td>0.010134</td>\n",
       "      <td>-0.071185</td>\n",
       "      <td>0.069887</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>PTBP1_start_overlap</th>\n",
       "      <td>-0.032418</td>\n",
       "      <td>0.001286</td>\n",
       "      <td>-0.048690</td>\n",
       "      <td>0.088437</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SSB_end_overlap</th>\n",
       "      <td>0.001870</td>\n",
       "      <td>0.007953</td>\n",
       "      <td>0.002986</td>\n",
       "      <td>0.112833</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>QKI_end</th>\n",
       "      <td>0.002426</td>\n",
       "      <td>-0.037638</td>\n",
       "      <td>-0.006597</td>\n",
       "      <td>-0.113609</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ZNF622_exon_complete</th>\n",
       "      <td>0.000597</td>\n",
       "      <td>-0.000077</td>\n",
       "      <td>-0.006748</td>\n",
       "      <td>-0.113012</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>STAU2_start_overlap</th>\n",
       "      <td>-0.126207</td>\n",
       "      <td>-0.085894</td>\n",
       "      <td>-0.115437</td>\n",
       "      <td>-0.011643</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>IGF2BP1_end_overlap</th>\n",
       "      <td>0.162657</td>\n",
       "      <td>0.124775</td>\n",
       "      <td>0.166455</td>\n",
       "      <td>0.067298</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>EWSR1_start</th>\n",
       "      <td>0.058579</td>\n",
       "      <td>0.097100</td>\n",
       "      <td>0.021596</td>\n",
       "      <td>0.117653</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                      HPC - glia (all)  HPC - neur (all)  HPC - glia (var)  \\\n",
       "QKI_start_overlap            -0.098697         -0.004609         -0.154732   \n",
       "MATR3_start_overlap          -0.156540         -0.024425         -0.190873   \n",
       "U2AF2_exon_complete          -0.003861          0.010134         -0.071185   \n",
       "PTBP1_start_overlap          -0.032418          0.001286         -0.048690   \n",
       "SSB_end_overlap               0.001870          0.007953          0.002986   \n",
       "QKI_end                       0.002426         -0.037638         -0.006597   \n",
       "ZNF622_exon_complete          0.000597         -0.000077         -0.006748   \n",
       "STAU2_start_overlap          -0.126207         -0.085894         -0.115437   \n",
       "IGF2BP1_end_overlap           0.162657          0.124775          0.166455   \n",
       "EWSR1_start                   0.058579          0.097100          0.021596   \n",
       "\n",
       "                      HPC - neur (var)  \n",
       "QKI_start_overlap             0.121338  \n",
       "MATR3_start_overlap          -0.000462  \n",
       "U2AF2_exon_complete           0.069887  \n",
       "PTBP1_start_overlap           0.088437  \n",
       "SSB_end_overlap               0.112833  \n",
       "QKI_end                      -0.113609  \n",
       "ZNF622_exon_complete         -0.113012  \n",
       "STAU2_start_overlap          -0.011643  \n",
       "IGF2BP1_end_overlap           0.067298  \n",
       "EWSR1_start                   0.117653  "
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Find top 10 with biggest different between neur and glia HPC var\n",
    "ct_specific = np.abs(w_filt['HPC - neur (var)'] - w_filt['HPC - glia (var)']).sort_values(ascending=False)[:10].index\n",
    "w_filt.loc[ct_specific]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "75182065",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAU8AAAE7CAYAAABdfgMlAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8pXeV/AAAACXBIWXMAAA9hAAAPYQGoP6dpAABdSklEQVR4nO3dd1gUV/s38O8uSK8qiII0QYoNlWA0CcUoiCZgjDwWNKLGHn1iSyxRJEI0MRpjJzF2xd7hsYti7EpRQFABsaHSexHO+wc/5nVlkdndQYr357r2SmZm556zJ/H2zMwpIsYYAyGEEJmI67sAhBDSGFHyJIQQOVDyJIQQOVDyJIQQOVDyJIQQOVDyJIQQOVDyJIQQOVDyJIQQOSjXdwFIw7Lx0E3BYkXcyxUs1t///ViwWACQlpEvWKzWLTQFi9WQ/bIrUtB4PcPXCxarw7hvBYtl7OLG63vU8iSEEDlQ8iSEEDlQ8iSEEDlQ8iSEEDlQ8iSEEDlQ8iSEEDlQ8iSEEDlQ8pSBn58fBg4cyG27urri+++/VzhuQkICjIyMkJeXx/uct69tbm6OlStXAgBKSkpgamqKW7duKVw2Qoh0dZ483044VcLDwyESiZCdnS2xXfUxMDCAp6cnoqOjJc578OABRo8eDRMTE6iqqsLCwgLDhg3DzZvCde7m6+DBg1i8eLHCcebPn48pU6ZAW1tbgFIBqqqqmDVrFn788UdB4hFCqmtwLc+EhAQ8f/4coaGhyMrKQr9+/ZCTkwMAuHnzJrp3747ExEQEBwcjLi4Ohw4dgq2tLWbOnPney9q8eXOFE96TJ09w9OhRjB49WqBSVfL19UVERATi4+MFjUsIqdTgkqehoSGMjIzg5OSE5cuXIy0tDVevXgVjDH5+frC2tkZERAQGDBiAdu3awcHBAf7+/jhy5IjC1w4MDIShoSG0tbXx7bffYs6cOXBwcKjx+2/fOu/YsQOOjo7Q1taGkZERhg8fjpcvX77zmnv37kWXLl1gYmLC7cvIyMCwYcNgYmICDQ0NdOrUCSEhITL9lhYtWqBXr14yn0cI4afBJc83qaurAwDKysoQFRWF2NhYzJw5E2Jx9WLr6ekpdK2dO3ciKCgIv/76K27dugVTU1OsXy/b2NvS0lIsXrwY0dHROHz4MJKTk+Hn5/fOcy5evAhHR0eJfcXFxejevTuOHz+Ou3fvYvz48Rg5ciSuXbsmU3mcnJwQEREh0zmEEH7ey8Qgx48fh5aWlsS+8vLyd56TkZGBgIAAaGtrw8nJCeHh4QAAW1vbOinj6tWrMXbsWO72eeHChTh16hTy8/lPIDFmzBju3y0tLbFq1So4OTkhPz+/2u+vkpKSgu7du0vsMzY2xqxZs7jtqVOn4sSJE9i3bx969OjBuzzGxsZISUmp8XhJSQlKSkok9pWVlaJZMxXe1yDkQ/VeWp5ubm6IioqS+GzcuFHqd01MTKClpYWWLVsiPj4e+/btg6GhIapWSBaJRDJde+fOndDS0uI+NbXEEhIS4OTkJLHv7e3aREZGwtvbG2ZmZtDW1oarqysAIDU1tcZzioqKoKamJrGvvLwcQUFB6Ny5M1q0aAEtLS2cOnXqnXGkUVdXR2FhYY3HlyxZAl1dXYnP/w5skekahHyo3kvLU1NTE1ZWVhL7njx5IvW7ERER0NHRgYGBAXR0dLj97du3BwDEx8e/8znk27y8vCRaa8bGxjV+9+3ELMuS9gUFBXB3d4e7uzt27NgBAwMDpKamwsPDA6WlpTWe17JlS2RlZUnsW758Of744w+sXLkSnTp1gqamJr7//vt3xpEmMzMTBgYGNR6fO3cuZsyYIbFv54m7Ml2DkA9Vg5vP08LCQurzSwcHB9jb22P58uUYMmRIteee2dnZUs/T1tbm9UbcxsYG169fx8iRI7l9snR/unfvHtLT07F06VK0bduW9/ldu3ZFXFycxL6IiAh4e3tjxIgRAICKigrcv38fdnZ2vMsDAHfv3kXXrl1rPK6qqgpVVVWJfXTLTgg/DfqF0ZtEIhE2b96MxMREODs7IywsDElJSYiJiUFQUBC8vb0Vij916lT8888/2Lp1K+7fv4/AwEDExMTwfkxgamoKFRUVrF69GklJSTh69CivPqAeHh64cuWKxDNgKysrnD59GpcvX0Z8fDwmTJiAtLQ0mX9TREQE3N3dZT6PEFK7RpM8gcpnkDdv3kS7du0wbtw42NnZwcvLC7GxsdzoGnn5+vpi7ty5mDVrFrp168a9KX/7eWRNDAwMsGXLFuzbtw/29vZYunQpfv/991rP69+/P5o1a4YzZ85w+xYsWIBu3brBw8MDrq6uMDIykjrQ4F2uXLmCnJwcDB48WKbzCCH8iJgsD/Y+MH379oWRkRG2b99ep9dZt24djhw5gpMnTwoW08fHB127dsW8efNkOo+W4ZAdLcMhn8a+DEeDe+ZZXwoLC7FhwwZ4eHhASUkJISEhOHPmDE6fPl3n1x4/fjyysrKQl5cnyBDNkpISdOnSBdOnTxegdIQQaSh5/h+RSISwsDAEBgaipKQENjY2OHDgAPr06VPn11ZWVsb8+fMFi6eqqoqffvpJsHiEkOooef4fdXV1ieeOhBDyLo3qhREhhDQUlDwJIUQOlDwJIUQOlDwJIUQOlDwJIUQOlDwJIUQONMKISCh9xxR2ssrOLxYslr6qwH/PN2smXKxa5qaVRcor4UY+AYC5kZ5gsQ5H3BcsFgCYXzsoWCyN1q0Fi9Vh3ERe36OWJyGEyIGSJyGEyIGSJyGEyIGSJyGEyIGSJyGEyIGSJyGEyIGSJyGEyIGSJyGEyIGSZwMwcuRI/PLLL7y/Hx4eDpFIhOzsbADAli1bJFYOXbNmDby8vAQuJSHkTYIkTz8/P6kLlL39h7xqu+pjYGAAT09PREdHS5z34MEDjB49GiYmJlBVVYWFhQWGDRsm01LAjUVMTAxCQ0MxdepUwWKOGzcON27cwKVLlwSLSQiRVC8tz4SEBDx//hyhoaHIyspCv379kJOTA6ByrfPu3bsjMTERwcHBiIuLw6FDh2Bra4uZM2fWR3EVVlZWVuOxNWvWwMfHR5C1i6qoqqpi+PDhWL16tWAxCSGS6iV5GhoawsjICE5OTli+fDnS0tJw9epVMMbg5+cHa2trREREYMCAAWjXrh0cHBzg7++PI0eOKHTdRYsWwcHBAdu3b4e5uTl0dXUxdOhQ5OXlcd9hjOG3336DpaUl1NXV0aVLF+zfv587/vYtMgAcPnxYYn33quts2rQJlpaWUFVVhbQpBCoqKrBv375qt9g7duyAo6MjtLW1YWRkhOHDh+Ply5cy/VYvLy8cPnwYRUVFMp1HCOGn3p95qqurA6hsnUVFRSE2NhYzZ86EWFy9aG8nLXk8fPgQhw8fxvHjx3H8+HFcuHABS5cu5Y7/9NNP2Lx5M9avX4/Y2FhMnz4dI0aMwIULF2S6zoMHD7B3714cOHAAUVFRUr8TExOD7OxsODo6SuwvLS3F4sWLER0djcOHD3NryMvC0dERZWVluH79eo3fKSkpQW5ursSnpKREpusQ8qESbAG448ePQ0tLS2JfeS2zzWRkZCAgIADa2tpwcnJCeHg4AMDW1laoYlVTUVGBLVu2cLfJI0eOxNmzZxEUFISCggKsWLEC586dQ8+ePQEAlpaWuHTpEoKDg+Hi4sL7OqWlpdi+fTsMDAxq/E5KSgqUlJRgaGgosX/MmDHcv1taWmLVqlVwcnJCfn5+tTquiaamJvT09JCSklJjuZcsWYKAgACJfT/Nm4cFAq7kSUhTJVjydHNzw/r1kovYX7t2DSNGjKj2XRMTEwBAQUEBrK2tsW/fPhgaGnK3tm/eAvOxc+dOTJgwgdv+3//+h88++0zqd83NzSWeL7Zu3Zq7JY6Li0NxcTH69u0rcU5paSm6du0qU5nMzMzemTgBoKioCKqqqtV+b2RkJBYtWoSoqChkZmaioqICAJCamgp7e3veZVBXV0fhO6aYmzt3LmbMmCGxTyTg9GqENGWCJU9NTU1YWVlJ7Hvy5InU70ZEREBHRwcGBgbQ0dHh9rdv3x4AEB8fDwcHB97X9vLyQo8ePbhtY2PjGr/b7K15HEUiEZecqv4ZGhpaLYaqqioAQCwWV3t+Ke2FkKamZq3lbtmyJQoLC1FaWgoVFRUAlX+huLu7w93dHTt27ICBgQFSU1Ph4eGB0tLSWmO+KTMz850JXFVVlftdVYScz5OQpqxe1m23sLCQ+vzSwcEB9vb2WL58OYYMGVLtuWd2drbU87S1tQV5W21vbw9VVVWkpqbWeKtrYGCAvLw8FBQUcAmypmeatan6CyIuLo7793v37iE9PR1Lly5F27ZtAUCuLloPHz5EcXGxzC1mQgg/9f7C6E0ikQibN29GYmIinJ2dERYWhqSkJMTExCAoKAje3t51en1tbW3MmjUL06dPx9atW/Hw4UNERkZi7dq12Lp1KwCgR48e0NDQwLx58/DgwQPs2rULW7Zsket6BgYG6Natm0R/TFNTU6ioqGD16tVISkrC0aNHsXjxYpljR0REwNLSEu3atZOrbISQd2tQyRMAnJyccPPmTbRr1w7jxo2DnZ0dvLy8EBsbi5UrV9b59RcvXoyFCxdiyZIlsLOzg4eHB44dOwYLCwsAQPPmzbFjxw6EhYWhU6dOCAkJwaJFi+S+3vjx47Fz505u28DAAFu2bMG+fftgb2+PpUuX4vfff5c5bkhICMaNGyd3uQgh70ZrGNWz4uJi2NjYYPfu3dwbfkXdvXsXn3/+ORITE6GrqyvTubSGkRxoDSO50BpGRCFqamrYtm0b0tPTBYv57NkzbNu2TebESQjhr15eGBFJsvQf5cPd3V3QeISQ6qjlSQghcqDkSQghcqDkSQghcqDkSQghcqDkSQghcqDkSQghcqCuSkTChFVXBYsVnSbc+Islg4XrBA0ArfRrn7iFrw4WLQWLZdpKT7BYAFAh4BCYR6+EnTRGJz5BsFgdxk+o/UsCo5YnIYTIgZInIYTIgZInIYTIgZInIYTIgZInIYTIgZInIYTIgZInIYTIgZLnW/z8/DBw4EBu29XVFd9//73CcRMSEmBkZIS8vDyFY9Vm8ODBWLFiRZ1fh5APmSDJ8+2EUyU8PBwikQjZ2dkS21UfAwMDeHp6Ijo6WuK8Bw8eYPTo0TAxMYGqqiosLCwwbNgwuRZCU9TBgwflWkPobfPnz8eUKVMEWaiuNgsXLkRQUBByc3Pr/FqEfKjqpeWZkJCA58+fIzQ0FFlZWejXrx9ycnIAVK4U2b17dyQmJiI4OBhxcXE4dOgQbG1tMXPmzPde1ubNmyuc8J48eYKjR49i9OjRApVKuqolkDt37gxzc3OJtZEIIcKql+RpaGgIIyMjODk5Yfny5UhLS8PVq1fBGIOfnx+sra0RERGBAQMGoF27dnBwcIC/vz+OHDmi8LUDAwNhaGgIbW1tfPvtt5gzZ84714h/+7Z9x44dcHR0hLa2NoyMjDB8+HC8fPnyndfcu3cvunTpAhMTEwBATk4O1NXVceLECYnvHTx4EJqamsjPr1zH5scff0T79u2hoaEBS0tLLFiwQGKN+EWLFsHBwQGbNm2CpaUlVFVVuTXlvby8EBISIkvVEEJkUO/PPNXV1QFUtpqioqIQGxuLmTNnVluzHYDUNdtlsXPnTgQFBeHXX3/FrVu3YGpqivXr18sUo7S0FIsXL0Z0dDQOHz6M5ORk+Pn5vfOcixcvwtHRkdvW1dXFgAEDqrUMd+3aBW9vb2hpaQGoXAp5y5YtiIuLw59//om///4bf/zxh8Q5Dx48wN69e3HgwAGJ9eOdnJxw/fp1lJSUyPT7CCH8CDYxyPHjx7k/9FXKa1lVMCMjAwEBAdDW1oaTkxPCw8MBALa2tkIVS8Lq1asxduxY7vZ54cKFOHXqFNfS42PMmDHcv1taWmLVqlVwcnJCfn5+td9fJSUlBd27d5fY5+vri2+++QaFhYXQ0NBAbm4uQkNDceDAAe47P/30E/fv5ubmmDlzJvbs2YMffviB219aWort27fDwMBAIr6xsTFKSkqQlpYGMzMzqeUqKSmpllzLX5dCSVmllloghAjW8nRzc0NUVJTEZ+PGjVK/a2JiAi0tLbRs2RLx8fHYt28fDA0NuVtOkUgk07V37twJLS0t7hMRESH1ewkJCXBycpLY9/Z2bSIjI+Ht7Q0zMzNoa2vD1dUVAJCamlrjOUVFRVBTU5PYN2DAACgrK+Po0aMAgAMHDkBbW1ti8bb9+/fj008/hZGREbS0tLBgwYJq1zEzM6uWOIH/36IvfMdSwkuWLIGurq7EJyacbvUJ4UOw5KmpqQkrKyuJj7GxsdTvRkREIDo6Gjk5OUhMTISHhwcAoH379gCA+Ph4ma7t5eUlkbTfvEV+29uJWZZl6wsKCuDu7g4tLS3s2LEDN27cwKFDhwBUtgBr0rJlS2RlZUnsU1FRweDBg7Fr1y4AlbfsQ4YMgbJy5c3A1atXMXToUHh6euL48eOIjIzE/Pnzq11HU1P61GqZmZkAIDWxVpk7dy5ycnIkPp1dh9VSC4QQoJ7m87SwsJD6/NLBwQH29vZYvnw5hgwZUu25Z3Z2ttTztLW1eb0Rt7GxwfXr1zFy5Ehunyzdn+7du4f09HQsXboUbdu25X1+165dERcXV22/r68v3N3dERsbi/Pnz0t0ifr3339hZmaG+fPnc/sePXrEu6x3796FiYkJWrasea5JVVVVqKqqSuyjW3ZC+Kn3F0ZvEolE2Lx5MxITE+Hs7IywsDAkJSUhJiYGQUFB8Pb2Vij+1KlT8c8//2Dr1q24f/8+AgMDERMTw/sxgampKVRUVLB69WokJSXh6NGjvPqAenh44MqVK9WeAbu4uKBVq1bw9fWFubk5Pv74Y+6YlZUVUlNTsXv3bjx8+BCrVq3iWrl8RERE0PrthNShBpU8gcpnkDdv3kS7du0wbtw42NnZwcvLC7GxsVi5cqVCsX19fTF37lzMmjUL3bp1496Uv/08siYGBgbYsmUL9u3bB3t7eyxduhS///57ref1798fzZo1w5kzZyT2i0QiDBs2DNHR0fD19ZU45u3tjenTp+O7776Dg4MDLl++jAULFvAqZ3FxMQ4dOoRx48bx+j4hRHYiJstDvyaob9++MDIywvbt2+v0OuvWrcORI0dw8uTJOr0OAKxduxZHjhzBqVOnZD539NJzgpWDluGQXQWT7WXp+7T6UHTtX5JBp9A/BYv12R/CDUfWMGrD63sf1BpGhYWF2LBhAzw8PKCkpISQkBCcOXMGp0+frvNrjx8/HllZWcjLy6vzIZrNmjXD6tWr6/QahHzoPqjkKRKJEBYWhsDAQJSUlMDGxgYHDhxAnz596vzaysrKEi9/6tL48ePfy3UI+ZB9UMlTXV292nNHQgiRR4N7YUQIIY0BJU9CCJEDJU9CCJEDJU9CCJEDJU9CCJEDJU9CCJHDBz/CiEi6GfSzYLG6TJ8lWKyn4cKNfAKAkswMwWIVpqUJFut8a2fBYgltoldnQeO9+veCYLHSoyIFi9V9Dr/+2NTyJIQQOVDyJIQQOVDyJIQQOVDyJIQQOVDyJIQQOVDyJIQQOVDyJIQQOVDyJIQQOVDyfI9GjhyJX375pc6vc+fOHZiYmKCgoKDOr0XIh0qm5Onn54eBAwdW2x8eHg6RSITs7GyJ7aqPgYEBPD09ER0tuQbKgwcPMHr0aJiYmEBVVRUWFhYYNmyYTMsBNxYxMTEIDQ3F1KlT6/xanTp1gpOTE/744486vxYhH6o6bXkmJCTg+fPnCA0NRVZWFvr164ecnBwAleudd+/eHYmJiQgODkZcXBwOHToEW1tbzJw5sy6LVWfKyspqPLZmzRr4+PjU+fpFVWUYPXo01q9fX225Y0KIMOo0eRoaGsLIyAhOTk5Yvnw50tLScPXqVTDG4OfnB2tra0RERGDAgAFo164dHBwc4O/vjyNHjih03UWLFsHBwQHbt2+Hubk5dHV1MXToUOTl5XHfYYzht99+g6WlJdTV1dGlSxfs37+fO75lyxbo6elJxD18+LDEGu9V19m0aRMsLS2hqqoKaVMFVFRUYN++ffDy8uL2zZ07V2Kd9iqdO3eGv78/AODGjRvo27cvWrZsCV1dXbi4uOD27dsS3xeJRNiwYQO8vb2hqamJwMBAAJVrxWdkZODCBeHGDxNC/r/39sxTXV0dQGXLKCoqCrGxsZg5cybE4upFeDtpyePhw4c4fPgwjh8/juPHj+PChQtYunQpd/ynn37C5s2bsX79esTGxmL69OkYMWKEzMnmwYMH2Lt3Lw4cOICoqCip34mJiUF2djYcHR25fb6+vrh27RoePnzI7YuNjcWdO3e4Ndzz8vIwatQoRERE4OrVq7C2tkb//v0l/hIAAH9/f3h7e+POnTsYM2YMAEBFRQVdunRBREREjWUvKSlBbm6uxKf09WuZfj8hHyqZF4A7fvw4tLS0JPbVdmuYkZGBgIAAaGtrw8nJCeHh4QAAW1tbWS/PW0VFBbZs2cLdJo8cORJnz55FUFAQCgoKsGLFCpw7dw49e/YEAFhaWuLSpUsIDg6Gi4sL7+uUlpZi+/btMDAwqPE7KSkpUFJSgqGhIbevY8eO6Ny5M3bt2oUFCxYAAHbu3ImPPvoI7du3BwD07t1bIk5wcDD09fVx4cIFfPHFF9z+4cOHc0nzTcbGxkhJSamxXEuWLEFAQIDEvnFurhj/uVuN5xBCKsnc8nRzc0NUVJTEZ+PGjVK/a2JiAi0tLbRs2RLx8fHYt28fDA0NuVvbN2+B+di5cye0tLS4z7taVebm5hLPF1u3bo2XL18CAOLi4lBcXIy+fftKxNu2bZtES5APMzOzdyZOACgqKoKqqmq13+vr64udO3cCqHyMEBISwrU6AeDly5eYOHEi2rdvD11dXejq6iI/Px+pqakScd5s0b5JXV0dhYWFNZZr7ty5yMnJkfj4uXz2zt9CCKkkc8tTU1MTVlZWEvuePHki9bsRERHQ0dGBgYEBdHR0uP1VLav4+Hg4ODjwvraXlxd69OjBbRsbG9f43WbNmklsi0QiVFRUAAD3z9DQ0GoxVFVVAQBisbja80tpL4Q0NTVrLXfLli1RWFiI0tJSqKiocPuHDx+OOXPm4Pbt2ygqKsLjx48xdOhQ7rifnx9evXqFlStXwszMDKqqqujZsydKS0t5lSEzMxPt2rWrsVyqqqrc762iovxBrUZNiNzq9E+KhYWF1OeXDg4OsLe3x/LlyzFkyJBqzz2zs7OlnqetrS3I22p7e3uoqqoiNTW1xlt0AwMD5OXloaCggEtONT3TrE3VXxBxcXESf1mYmJjA2dkZO3fuRFFREfr06YNWrVpxxyMiIrBu3Tr0798fAPD48WOkp6fzvu7du3cxePBgucpMCHm3emlmiEQibN68GX369IGzszPmzZsHW1tb5Ofn49ixYzh16lSdviXW1tbGrFmzMH36dFRUVODTTz9Fbm4uLl++DC0tLYwaNQo9evSAhoYG5s2bh6lTp+L69evYsmWLXNczMDBAt27dcOnSpWotbV9fXyxatAilpaXV+mVaWVlh+/btcHR0RG5uLmbPns29eKtNSkoKnj59ij59+shVZkLIu9XbCCMnJyfcvHkT7dq1w7hx42BnZwcvLy/ExsZi5cqVdX79xYsXY+HChViyZAns7Ozg4eGBY8eOwcLCAgDQvHlz7NixA2FhYejUqRNCQkKwaNEiua83fvx47vnmm3x8fJCRkYHCwsJqAxA2bdqErKwsdO3aFSNHjsS0adMkXjq9S0hICNzd3WFmZiZ3mQkhNaM1jN6T4uJi2NjYYPfu3dwb/rpSUlICa2trhISE4JNPPpHpXFrDSHa0hpF8aA0jwouamhq2bdsm0zNLeT169Ajz58+XOXESQvijV6vvkSz9RxXRvn17rkcDIaRuUMuTEELkQMmTEELkQMmTEELkQMmTEELkQMmTEELkQMmTEELkwQiRQXFxMfP392fFxcUNLl5DjSV0vA+lbA35dzLGGI0wIjLJzc2Frq4ucnJyJGbKagjxGmosKlv9x6qLeHTbTgghcqDkSQghcqDkSQghcqDkSWSiqqoKf3//ajPQN4R4DTWW0PE+lLI15N8J0JR0hBAiF2p5EkKIHCh5EkKIHCh5EkKIHCh5EkKIHGgmeSKT4uJi7NmzBwUFBejbty+sra0Vivf69WsUFxdDS0tLoBIK48KFCygoKEDPnj2hr69f38VpND6keqO37aRGs2fPRmlpKf78808AQGlpKXr06IHY2FhoaGjg9evXOH36NK8F7cLCwpCRkYGRI0dy+4KCgrB48WK8fv0avXv3xp49e2T+A3f27FmcPXsWL1++REVFhcSxTZs21Xr+smXLkJ+fj4CAAAAAYwyenp44deoUAMDQ0BBnz55Fhw4deJWnoqICFRUVUFb+/+2SFy9eYMOGDSgoKICXlxc+/fRTvj+PwxjD/v37cf78eam/9eDBgzLFa0j19uDBA+Tk5KB79+4S5QsMDERBQQEGDhyIefPmyfLzuDIJWWdvo9t2UqP//e9/+Pzzz7ntnTt34tGjR7h//z6ysrLg4+ODwMBAXrF+//135ObmctuXL1/GwoULsWDBAuzduxePHz/G4sWLZSpfQEAA3N3dcfbsWaSnpyMrK0viw0dISAjs7e257f379+PixYuIiIhAeno6HB0duQTBx9ixYzF58mRuOy8vDx999BHWrl2LkydPws3NDWFhYfx/5P/573//i5EjRyI5ORlaWlrQ1dWV+MiiodXb7NmzcfjwYW47OTkZX375JVRUVNCzZ08sWbJEruXIhawzqQSZXoQ0Sdra2uz+/fvc9tChQ9m4ceO47cjISNa6dWtesQwMDNjt27e57enTpzMPDw9uOzQ0lFlZWclUPiMjI7Zt2zaZznmbnp4ei4uL47b9/PzYiBEjuO0rV64wExMT3vGsra3ZyZMnue01a9aw1q1bs+zsbMYYYz/88ANzdXWVuZz6+vosNDRU5vOkaWj1ZmJiwi5fvsxtL168mHXp0oXb3rhxo8Q2X0LWmTTU8iQ1EovFYG881bl69So+/vhjbltPT493SyUvLw8tWrTgti9duoTevXtz2x06dMCzZ89kKl9paSl69eol0zlvKysrkxhxcuXKFYmYbdq0kWm56KdPn0o8Bz579iy+/vprrqUzatQoxMbGylxOXV1dWFpaynyeNA2t3tLT02FiYsJtnz9/Hl9++SW37erqipSUFJnLKGSdSUPJk9TI1tYWx44dAwDExsYiNTUVbm5u3PFHjx6hVatWvGK1adMG8fHxAID8/HxER0dLrCufkZEBDQ0Nmcr37bffYteuXTKd8zYrKytcvHgRAJCamorExESJJaKfPHkikfRro6amhqKiIm777b9w1NTUkJ+fL3M5Fy1ahICAAInY8mpo9da8eXM8f/4cQOUz45s3b6JHjx7c8dLSUom/xPkSss6kobftpEazZ8/GsGHDEBoaitjYWPTv3x8WFhbc8bCwMDg5OfGKNXjwYHz//feYN28ewsLCYGRkJJFUbt68CRsbG5nKV1xcjL/++gtnzpxB586d0axZM4njK1asqDXGpEmT8N133yEiIgJXr15Fz549JZ7lnTt3Dl27duVdpi5dumD79u1YsmQJIiIi8OLFC4kW9sOHD9GmTRve8ar4+PggJCQEhoaGMDc3r/Zbb9++zTtWQ6s3FxcXLF68GOvWrcO+fftQUVEh8Zd0XFwczM3N+f24NwhZZ9JQ8iQ1+vrrrxEWFobQ0FC4u7tj6tSpEsc1NDQkXo68i7+/P549e4Zp06bByMgIO3bsgJKSEnc8JCRE4laNj5iYGDg4OAAA7t69K3FMJBLxijFhwgQoKyvj+PHjcHZ2hr+/v8TxZ8+eYcyYMbzLtGDBAvTv3x979+7F8+fP4efnh9atW3PHDx06JNHi5svPzw+3bt3CiBEj0KpVK96/T5qGVm9BQUHo27cvzM3NIRaLsWrVKmhqanLHt2/fLvEXEF9C1pk01FWJEIHFxcXh9OnTMDIygo+PD8Ti//907K+//oKTkxOXvPjS1NTEyZMn5erm1BiUlZUhLi4OBgYG1Vrm0dHRMDExkenxCVD3dUbJk9QoMzMThYWFEg/zY2Nj8fvvv3P974YPH67QNYTqVP3kyROIRCIYGxvLdF5d9csUmq2tLfbu3YvOnTsLGrch15uiAyjqqs44dfYenzR6Q4cOZdOnT+e2X7x4wfT19VmHDh2Yl5cXa9asGe8uL7/99htbuHAht11RUcE8PDyYSCRiIpGItWrVit29e1em8pWXl7OAgACmo6PDxGIxE4vFTFdXl/3888+svLycVww/Pz+J7le5ubmsbdu2zMDAgHXu3JkpKyvL1N3l/v377ObNmxL7zpw5w1xdXdlHH33EgoKCeMd60/Hjx5mHhwdLTk6W6/w3NbR6Cw0Nrfb/UWBgIFNVVWVKSkqsb9++LDMzk/8P/D9C1pk0lDxJjczNzdn58+e57WXLlrF27dqxsrIybrtHjx68YnXt2pXt3r2b2967dy9TV1dnly5dYhkZGWzAgAHMx8dHpvLNmTOHGRgYsHXr1rHo6GgWFRXF1q5dywwMDNi8efN4xRC6X+bAgQPZTz/9xG0nJSUxdXV15u7uzqZNm8a0tLTYH3/8wTteFT09PaaiosLEYjHT0tJi+vr6Eh9ZNLR6c3NzY2vWrOG2//33XyYWi1lgYCA7cOAAs7W1lfhLnC8h60waum0nNVJXV8e9e/dgZmYGAOjfvz86dOiAZcuWAQASExPRs2dPZGRk1BpLX18fly9fhp2dHQBg9OjReP36NbZv3w6gskuPj48PHj9+zLt8bdq0wYYNG+Dl5SWx/8iRI5g8eTKePn1aawxNTU3cvXuX60UwaNAgGBsbY/Xq1QAqn1+6urri5cuXvMrUtm1b7N27lxuyGhgYiP379yMqKgoA8M8//2D16tXcNl9bt2595/FRo0bxjtXQ6s3Q0BAnT57k3s7PmDEDcXFxOHHiBIDKXh3//e9/cf/+fd6/ERC2zqSht+2kRjo6OsjOzuaS5/Xr1zF27FjuuEgkQklJCa9Y0jpV//e//+W2Ze2MDlQ+k7W1ta2239bWFpmZmbxiSOuXWfWXQ9VxWfpl8unwPXPmTN7xqij6B/1NDa3epA2gGDx4MLctzwAKQNg6k4Y6yZMaOTk5YdWqVaioqMD+/fuRl5cn0WUkMTERbdu25RVL6M7oQGWfyjVr1lTbv2bNGnTp0oV3jKrWrxD9Muuqw/ebioqKkJubK/GRRUOrt7oYQPE2RetMKoVv/EmTFRkZyVq0aME9N3rzWR5jjI0YMYJNmDCBV6wNGzYwTU1NNmbMGGZvb8969eolcXzx4sXsiy++kKl84eHhTFNTk9nZ2bExY8awsWPHMjs7O6alpcUuXrzIK8a5c+eYmpoas7S0ZOrq6mzMmDESxydNmsS++eYb3mUaNmwY++KLL1hqaipbvnw509LSYvn5+dzx/fv3s86dO/OOVyU/P59NmTKFGRgYcC953vzIoqHV2w8//MBsbW3Ztm3b2NChQ5mpqSl7/fo1dzw4OJh98skn/H/g/xGyzqSh5Ene6eXLl+zw4cPs6tWr1Y4dP36cJSUl8Y61ceNGNnDgQDZx4kT2/PlziWOTJk1iBw8elLl8T58+ZfPmzWODBg1iX331FZs/fz57+vSpTDFiY2PZypUr2e7du6u9bQ4ODmaRkZG8YyUlJbF27doxsVjMlJWV2bp16ySOe3t7s++//16m8jHG2OTJk5mdnR3bt28fU1dXZ5s2bWKLFy9mJiYmbMeOHTLHa0j1VlBQwEaMGMH09PSYra1ttQTu6urKli5dKlPZGBO+zt5GyZM0WoWFhTUee/bs2XssiaTS0lIWFRUlNRlFRUWx9PR0mWO2bduW6/nw5mxX27ZtY56enjLFaqj1JjQh60waeuZJGq2uXbtKHZ+8f//+uusYzUOzZs3QpUsXqc/8unTpIvOzXaDyJU/Vm20dHR3uxc6nn37KPUvmq6HWm9CErDNpKHmSRqtv377o1asXli5dCsYY8vPz4efnh1GjRmHhwoX1XTxBWVpactOy2dvbY+/evQCAY8eOQU9PT6ZYH0q9CVlnUincdiWkHv3vf/9jRkZG7NNPP2WWlpbMwcGBxcbG1nexBLdixQr2559/MsYqX9aoq6tzL/JWrlwpc7wPod6ErrO3USd50qhVVFRg6tSpWL9+PZSVlXHs2DF4eHjUd7HqXGpqKm7evIl27drx7l70pg+x3hSts2oUTr+E1JMHDx4wJycnZmpqyk6dOsXmz5/PVFVV2ezZs1lpaWl9F6/OFBUVKXT+h1hvitaZNNTyJDUSi8UQiURgjEEkEqG8vFzuWG/P7chnhcbaaGtrY8CAAdiwYQP3DOvy5cv45ptvoK2tjcjIyFpjWFhYSMzzmJSUpFCZUlNTJbZNTU0VilelvLwcv/zyCzZs2IAXL14gMTERlpaWWLBgAczNzSVGftWmodXbtm3bJLa/+eYbuWO9Scg6k4aGZ5IaJScnCxaraoinkNatWyexlDEA9OrVC5GRkfj+++95xdiyZYugZTI3NxfsL5w3BQUFYevWrfjtt98wbtw4bn+nTp3wxx9/yJQIGlq9bd68mft3kUgkWPIUss6kErwtS8h7VlJSwu7du8fN9tQUtWvXjp05c4YxxpiWlhZ7+PAhY4yx+Ph4pqenJ1fMpl5vdVFnb6KuSqRWb7eerl27hosXL6KsrKyeSlSpqKgIY8eOhYaGBjp06MDdMk+bNg2//vqrQrFfvHhR7Ra8Pj19+hRWVlbV9ldUVMj83+FDqTch60waSp6kRs+fP8enn34KVVVVuLi4ICsrC1988QV69uwJV1dXdOzYkZsEg4+NGzdi1KhR3G3anj17YGdnB0tLy2pr4PAxZ84cREdHIzw8HGpqatz+Pn36YPfu3bxi5OXlYcSIETAzM8OoUaNQWlqKKVOmoHXr1rCwsICLi4vCk0hERUVh3759uHTpktyTgnTo0AERERHV9u/bt0+mBeqAhl9vWVlZWLlyJaZMmYLAwECZpil8k5B1JpXCbVfSZI0cOZL16tWLHT16lA0ZMoT16tWLffbZZ+zJkycsNTWVffbZZ2zKlCm8Yv3xxx9MU1OTDRo0iLVu3ZoFBgayFi1asMDAQPbzzz8zXV1dFhwcLFP5TE1N2ZUrVxhjkrdl9+/fZ9ra2rxifPfdd8zW1patWrWKubq6Mm9vb9axY0d26dIldvHiRdaxY0feEwQzVjkxSG5uLmOMsby8PObu7s5EIhFTUVFhIpGIOTo6sqysLJl+J2OMHT16lOnq6rKlS5cyDQ0NtmzZMvbtt98yFRUVdurUKZliNbR6a926NTdkNSkpiRkZGTEjIyPWt29fZmJiwnR1dVl8fLxMv5ExYetMGkqepEatW7fm/pBlZGQwkUjEPUNirLLjsaWlJa9Ytra2bOfOnYwxxm7fvs2UlZXZxo0bueObNm1i3bt3l6l86urq3B/8N5NAVFQU09HR4RWjbdu27Ny5c4yxyskyRCIRO3r0KHc8NDSU2djY8C6TWCxmL168YIwxNmvWLGZhYcFu3brFGGPszp07zM7OTq5Z0Rlj7MSJE8zZ2ZlpamoydXV19sknn0jM5s5XQ6s3kUjE1dnQoUOZq6srKygoYIwxVlxczL744gs2ePBgfj/uLULVmTSUPEmN1NTUWGpqKretqanJTa7AGGOPHj1i6urqvGKpq6uzR48ecduqqqoSaxbdv39f5of4zs7ObNWqVYyxyiRQNcPTlClTmIeHB68YqqqqEr9RQ0ODJSQkcNspKSlMQ0ODd5neTAQdOnRge/bskTgeGhrKrK2tecdjjLGysjK2aNEiiXIqoqHV25t1ZmFhwc6ePStx/OrVq8zExIRXrCpC15k01FWJ1MjQ0BDPnz/nJjz+7rvv0Lx5c+54VlaWxPra76KhoYGCggJu28DAoNqqiK9fv5apfEuWLEG/fv0QFxeH169f488//0RsbCyuXLmCCxcu8IrRokULvHr1ivuN3t7eEuOe8/PzJWbA56Oq/+OLFy/QsWNHiWMdOnSQ+RmesrIyli1bJtjM6A2x3qrqrKSkBK1atZI41qpVK7x69Yp3LED4OpOGXhiRGjk4OODKlSvc9tKlSyWS56VLl3jPwmNra4uYmBhu+/HjxxJ9P+/duwdzc3OZyterVy/8+++/KCwsRLt27XDq1Cm0atUKV65cQffu3XnF6Ny5M27cuMFt79q1C4aGhtz2jRs3uHWX+FqwYAFmzJgBsViMtLQ0iWPp6elyLaXbp08fhIeHy3yeNA2x3j7//HN069YNubm5SExMlDiWmpqKli1b8o5VRcg6k4ZanqRGR44ceedxJycniaU03uXXX399Zys1NTUVEyZMkKl8QGWH59oW+nqXnTt3QiyuuQ3RqlUrBAUF8Y7n7OyMhIQEAJUz+bw90CAsLAwdOnSQuZyenp6YO3cu7t69i+7du1ery7cXc6tNQ6q3t3tavL3kxrFjx/DZZ5/JXEah6+xtNDyTkPcoKSkJKioqEovE8fGuRCXkSKampK7rjJInIYTIgZ55EtIIbNu2Teoyz6WlpdUm1iCV6rrOqOVJSCOgpKSE58+fS7yUASqX5TU0NKTbdinqus6o5UkavQcPHuDkyZMoKioCAIXXRW+I2P/N0vS2J0+eQFdXV66YTb3e6qLO3kRv24lMJk+ejJ9//lmuriNv+/fff+Ho6ChzP8oqGRkZGDJkCM6dOweRSIT79+/D0tIS3377LfT09LB8+XKFy1jfunbtCpFIBJFIhM8//xzKyv//j2x5eTmSk5PRr18/mWLWVb0tXboUEydOFGZ9IAXURZ1JQy1PIpMdO3YoPFFGFU9PTzx9+lTu86dPnw5lZWWkpqZKdG8ZMmQITpw4IUQRER8fD0tLS4ViDBgwQKYJVN40cOBAeHt7gzEGDw8PeHt7c5+hQ4ciODgYO3bskClmXdXbL7/8wq1QKaSsrCyZnlHWRZ1JVWdjl0iT9OZY6PqO1apVKxYVFVUtVlJSEtPU1BSkjFFRUUwsFisUQ4g627JlCysuLlYoRpW6qjch/994k7z/DYSsM2notp00WgUFBdU6VAOVo3j4PgqYMWPGO4/LOiywrvTu3RuvXr3i+odev34du3btgr29PcaPHy9TLCHqTUi13cnk5eXJFVfIOpOGkieRibz/I0sTHBxcbRyzLJydnbFt2zYsXrwYQGXH54qKCixbtgxubm68Yvz5559wcHCAjo6O1OP5+flyl6+KmZkZmjVrplCM4cOHY/z48Rg5ciTS0tLQp08fdOzYETt27EBaWppM660LUW/SxMXFoU2bNjKfp6enJ/XFThVWw4uf2ghZZzUVjJBGKTY2lhkYGLB+/foxFRUVNnjwYGZnZ8datWrFHjx4wCuGjY0N2759e43HIyMjFb5tF4Kenh67d+8eY4yxP//8k/Xq1YsxxtjJkyeZhYWFTLGEqDch6ejosF9//ZWFh4dL/fz9999y/TcQss6koZYnabTs7e0RExOD9evXQ0lJCQUFBRg0aBA3ozkf3bt3x61btzBixAipx6sWc6tvZWVl3C31mTNnuHHZtra2Mr+MEqLehNStWzcAqHGeBD09Pbn+GwhZZ1IpnH4JqSePHj1iFRUVNR7j4/nz5ywlJUXIYtUJJycn9uOPP7KLFy8yNTU17oXPlStXmLGxsUyxhKg3If3111/szz//rPF4WloaW7RokcxxhawzaSh5kkbrzVnb35Sent4gbrWFdP78eaanp8fEYjEbPXo0t3/u3Lnsq6++kinWh1JvQtaZNDQ8kzRaYrEYL168gIGBgcT+R48ewd7eXmLy5XfZt28fDh8+jLKyMvTp00eQN7F1oby8HLm5udDX1+f2paSkQENDo9oQxHcRqt6ExBjDgwcPUFZWhvbt20t0bFeEUHUmDT3zJDUaNGgQr+8dPHiQ1/cYY0hJSUHbtm2hrKyM0tJSHDp0CCUlJejfvz/vUUtV3YtEIhEWLFgg0e2mvLwc165dg4ODA69Yf/31FyZOnAhra2uoqanhwIEDSE5OxpIlS3idz1dpaSlKS0vlmgi5ipKSkkQSACDTBNJC1htQuYTxrVu30Lx5c9jb20scKy4uxt69e/HNN9/UGiclJQXe3t64e/cuAKBt27Y4ePAg9yxUEUpKSigrK0NERAREIhHat28v86TbNaERRqRGurq6vD58JCQkwMLCAlZWVrCzs0NycjJ69eqFsWPHYtKkSbCzs8P9+/d5xYqMjERkZCQYY7hz5w63HRkZiXv37qFLly7YsmULr1irV6/G/PnzkZCQgOjoaPzzzz9Ys2YNr3NrsnnzZkydOhU7d+4EAMydOxfa2trQ1dVF3759kZGRIXPMFy9eYOTIkWjTpg2UlZWhpKQk8eFDyHpLTEyEnZ0dnJ2d0alTJ7i6ukq8hMnJycHo0aN5xfrxxx9RXFyM7du3Y9++fWjdujUmTpzI69x3yc3NxciRI2FsbAwXFxc4OzvD2NgYI0aMQE5OjsLx6ZknUUhZWRmv73l7ezMvLy8WExPDvv/+e2Zvb8+8vb1ZaWkpKykpYd7e3mzEiBEyXdvPz4/l5OTIU2yOhoaGxKiY169fs2bNmrHnz5/LFS8wMJCpq6uzzz//nDVv3pxNnDiRGRkZsaVLl7LffvuNmZiYsIkTJ8oct1+/fsze3p6tW7eOHTp0iB0+fFjiIwsh6m3gwIHsiy++YK9evWL3799nX375JbOwsOBeOKWlpfF+ftq6dWsWHh7ObT9+/JiJxWJWWFioUBl9fHyYtbU1O3HiBMvJyWG5ubnsxIkTzMbGhvn4+CgUmzF6YUTeISQk5J3HS0tLmZeXF69YBgYGLDIykjHGWH5+PhOJRCwiIoI7fvnyZWZqaipXOe/fv89OnDjB/WGr6U2yNG+u3FhFkWGGVlZWbNeuXYwxxm7cuMHEYjHbt28fdzwsLEyu36mlpcXVn1AUqTdDQ0MWExMjsW/y5MnM1NSUPXz4UKbkKRKJWFpamsQ+TU1NlpyczLs80mhoaEj8P1bl4sWLMq2IWhN65klq5OfnB319fXh4eFQ79vr1a/j4+ODmzZu8YuXn53OLx2lqakJTU1OiT6GJiQlevHghU/kyMzPh4+OD8+fPKzQ70MaNGyWeRb5+/RpbtmyReAY7bdo0XrFSU1Px6aefAgAcHR2hrKyMTp06ccc7d+4sVx/Dtm3bCtbfVIh6KyoqqvZSZ+3atRCLxXBxccGuXbt4l0ckElVbMkMsFiv8e1u0aCH1sZKurm61Z8dyUTj9kiZr5cqVTFNTk12+fFli/+vXr9nAgQNZq1atWHx8PK9Y7dq1k2gFrFu3juXm5nLbt27dYkZGRjKVb+TIkczDw4M9fvxYorV48uRJZm9vzyuGmZkZMzc3f+dHltEob7dk327FytIie9PJkyeZu7u7wq0xxoSpt48++oht27ZN6rEpU6ZwXYT4EIlETE9Pj+nr63MfkUjEdHV1JfbJKjg4mPXp04c9e/aM2/f8+XPm7u7ONmzYIHO8t1HLk9Tov//9LzIzMzFgwABcvHgRHTt2RHl5Of7zn//g8uXLOH/+PGxtbXnF6tOnD+7du8e1yiZNmiRx/NSpUzK/XT116hROnjxZbTE1a2trPHr0iFeMlJQUma7JR1xcHLfkMGMM9+7d48bIp6enyxVzyJAh3FLBGhoa1cbKyzIVnBD19tVXXyEkJAQjR46sdmzNmjWoqKjAhg0beMXavHkzr+/Jav369Xjw4AHMzMxgamoKoPLOQFVVFa9evUJwcDD33du3b8scn/p5klpNnToVBw4cQHh4OObPn4/w8HCcO3dO4nZUUcnJyVBTU5NpeKC2tjZu374Na2traGtrIzo6GpaWlrhx4wb69esn11ttRYnF4hqHdFbtl2flxi1btrxzcoxRo0bxjtUQ660uBAQE8P7u28sf80HJk/AycuRI7N+/H1paWjh79iw6d+4s+DWePn0KY2Nj3t8fMGAAunXrhsWLF0NbWxsxMTEwMzPD0KFDUVFRgf379/OKk5eXh8TERNjY2EBLSwu3b9/GypUrUVRUhIEDB8LX15d3mfi23MzMzHjHFJpQ9VbXkpKSUFRUBDs7u3cuI1xf6Lad1OjNuS6rJmdwcHCo1hdwxYoVCl0nLS0NQUFB2LhxI7eeDh/Lli2Dq6srbt68idLSUvzwww+IjY1FZmYm/v33X14xLl68iC+++AL5+fnQ19dHSEgIBg8eDGNjYygpKeHgwYMoLCzEuHHjeMWrq6To6+sLV1dXuLi4oH379grFEqLe3NzcpLaEdXV1YWNjgylTpqBt27a8YpWWliIoKAi3b9/Gxx9/jDlz5mDEiBHYu3cvAMDGxgZhYWGCdW4XCrU8SY34zO0oEolw7ty5Wr+XnZ2NKVOm4NSpU2jWrBnmzJmD7777DosWLcLvv/+ODh06YMaMGRg2bJhMZUxLS8P69etx69YtVFRUoFu3bjLNDuTs7Axra2sEBARg8+bNWLFiBSZNmoRffvkFABAYGIj9+/cjKiqKV7zMzEwUFhZKPE+MjY3F77//joKCAgwcOBDDhw+X6TcCwIQJE3DhwgUkJibCyMgILi4ucHFxgaurK+/nzm9StN6mT58udX92djZu3bqFpKQkXLp0ideIpZkzZ2L79u3w8vLC+fPn0bFjRyQkJCAgIABisRiLFy9Gp06duEEHfFU9QqmJwiuOKvzKiRAeJk2axExMTNjMmTNZhw4dmFgsZp6enszNzU2ig/T7pqury/UYKCkpYWKxmJt9h7HKvpBaWlq84w0dOpRNnz6d237x4gXT19dnHTp0YF5eXqxZs2Y1vqXm4/nz5ywkJIRNmDCB2draMrFYLHMvhfdh8uTJzNPTk9d3TU1NWWhoKGOMsYSEBCYSiVhYWBh3PDw8XK5ZkN4eSLBv3z42b948ZmxszDZu3ChzvLfRbTvhLT09HSoqKjXOuv4uoaGh2Lx5M/r06YPJkyfDysoK7du3x8qVK2WKExMTw/u7fJ7L5ubmcv1PVVRUoKGhAW1tbe64trY2CgsLeV/z6tWrEm+Pt23bhubNmyMqKgrKysr4/fffsXbtWqlvqfnQ1taGvr4+9PX1oaenB2VlZRgZGdV6ntD1VpsJEyZI7R8szbNnz9ClSxcAQPv27aGqqgorKyvuePv27bneC7Lw9vautm/w4MHo0KED9uzZg7Fjx8ocU4LC6Zc0aVlZWWzy5MmsRYsWTCwWM7FYzFq1asXmzJnDCgoKeMdRVlZmT58+5bbV1dXZnTt3ZC6PSCRiYrGYiUSid3749jEUi8Xs5cuX3La2tjZLSkritmXtl6mmpiYxP6inpyebNWsWt52QkMCaN2/OO16VH374gfXo0YOpqakxR0dHNmPGDHbkyBGWlZXF63yh6602iYmJTE9Pj3fZ6qJvbE0ePHhAI4xI3crMzETPnj3x9OlT+Pr6ws7ODowxxMfHY/Xq1Th9+jQuXbqE6OhoXLt27Z2jcCoqKiT6JiopKUFTU1PmMiUnJ8v1W2rCGJNY27uwsBBffvklVFRUAFSONpKFjo4OsrOzuRdH169fl2jhiEQilJSUyFzOZcuWwcDAAP7+/vD29oadnZ1M5wtdb7U5deqUTC+2Tp48yY0GqqiowNmzZ7lZlrKzswUrV1FREVavXl2tj6s86IURqdH333+Ps2fP4syZM9UWaktLS4O7uztsbGxw6tQprFq16p19DcViMTw9PbllEY4dO4bevXtXS6B8p7cTCt++gHz7AX755ZcwNDTE33//jYMHD8LX1xdpaWnccMDQ0FDMmjUL8fHxMpUzOjoaFy5cQHh4OCIiIqCkpMS9MHJ1dZU5mSrq6NGjUvfn5OTgxo0b+Oeff7Blyxb4+PjUGotPNyR5+sbq6+tLvDBijCEvLw8aGhrYsWMHtyyHvCh5khqZm5sjODi4xmdXJ06cQP/+/eHv719rcvHz8+O1AqKso00SEhKwevVqxMfHQyQSwdbWFlOnToWNjQ2v81NTU2FiYiJYP8KoqCj06dMHeXl5eP36NebNm8etUglU9pfV1NTkPfqmJtHR0Vi5ciV27NiBiooKmROLovVWU31pa2vD1tYWs2bN4pU469LWrVsltsViMQwMDNCjRw9BxrZT8iQ1UlVVxcOHD2u8xXny5AnMzc1lvrUVyv79+zFs2DA4OjqiZ8+eACpf2Ny4cQO7du3i9YdXSUkJz58/V3hW8Te9evUKly9fhpGREXr06CFxLDQ0FPb29rCwsJA5bmRkJMLDw7nWZ25uLhwcHODm5oZly5bxjiNEvRHQCyNSszZt2kid0qvKxYsXWevWrXnFqmndHEVYWFiwBQsWVNu/cOFC3pN5SJuSriHS09NjysrKrHv37mzmzJns2LFjcs/JKUS91SY9PZ398ccfMp1TXl5e4/76WJiuNpQ8SY3GjBnDnJ2dWUlJSbVjxcXFzMXFRWJhrXepiySlrq7O7t+/X21/YmIiU1dXr5dyXb16VaKPImOMbd26lZmbmzMDAwM2btw4VlxcLHNcRZLl24SoN2kqKirYiRMnmI+PD1NRUWEtW7bkdV5OTg7z8fFhampqzNDQkC1cuJC9fv2aOy7023ah0Nt2UqOAgAA4OjrC2toaU6ZM4UayxMXFYd26dSgpKcG2bdvqrXyurq6IiIiQ6BMIAJcuXcJnn33GO87b83lKw3c+z0WLFsHV1RWenp4AgDt37mDs2LHw8/ODnZ0dli1bhjZt2mDRokW8ywcAX3zxBffvT548gUgkkmkegDcJVW9VUlJSsGnTJmzZsoXrmREaGsprhBoALFiwANHR0di+fTuys7MRGBiIW7du4eDBg1yvB9YAny7SM0/yTsnJyZg8eTJOnTrF/Q8sEonQt29frFmzptofwJqIxWJs3bq11jWPZHkDumHDBixcuBD/+c9/8PHHHwOofHa3b98+BAQEoE2bNrXGFYvFMDExeec6QCKRCElJSbzK1Lp1axw7dgyOjo4AgPnz5+PChQu4dOkSgMqVOv39/REXF8crXpWKigoEBgZi+fLl3PR22tramDlzJubPny/TCy8h6q2kpAQHDx7Exo0bcfnyZXh6emL48OEYNmwYoqOjqy0I9y5mZmbYunUrXF1dAQAZGRkYMGAAdHV1cfToUWRnZ6NNmzaKD6cUWv02fEljkZmZya5du8auXbvGMjIyZD6/ts7Z8nTQ5hOztrhC37arqqqy1NRUbvuTTz5hixcv5raTk5NlGu5ZZc6cOczAwICtW7eORUdHs6ioKLZ27VpmYGDA5s2bJ1MsIeqtRYsW7LPPPmPBwcEsMzOT26+srMxiY2NlKo+GhobEwATGGMvNzWU9e/ZkvXv3ZklJSQrfti9ZsoT3gAK+KHmS96KhvpgR+kWWqakpu3DhAmOscqy8uro6O3PmDHc8JiZGrlnRW7duzY4cOVJt/+HDh1mbNm3kL7Cc9PT0mLOzM/vrr78knsXKkzxtbGy4se1vysvLYz179mRdunRROHlqa2vLvS5VTRreJHmkSeLTx7M+MIGfWvXr1w9z5sxBREQE5s6dCw0NDYnniDExMWjXrp3McTMzM6XOnmRrayvTLPJCef78OcaPH4+QkBAYGRnh66+/xqFDh+T67+zu7i61f6+WlhZOnjwJNTU1hcsr9H9ngObzJO8Jn/95o6KieE1h9qbr168jPDwcL1++REVFhcQxPvOM+vv71/qySBaBgYEYNGgQXFxcoKWlha1bt3IvPQBg06ZNcHd3lzluly5dsGbNGqxatUpi/5o1a7hJNWShaL2pqanB19cXvr6+ePjwITZv3oxp06bh9evXCAoKgp+fH3r37s1rTfmAgAA8e/ZM6jFtbW2cOXMGt27d4vfD3iN6YUTei9GjR2PVqlUSMxYBlcP5du7ciY0bNyI6OlqmlwK//PILfvrpJ9jY2KBVq1YSrR6+84wWFhZi9uzZOHz4MMrKytCnTx+sWrVKYuVMeeTk5EBLS6ta8sjMzISWlpZEQuXjwoULGDBgAExNTdGzZ0+IRCJcvnwZjx8/RlhYmExvyYWot23btmHIkCHccFug8qXWyZMn8c8//+DYsWPQ1taWe80moT1+/Bht2rThlcx5E/QhACE8nT17lvn6+jJ1dXVma2vL5s+fz27fvi1TDENDQ7Z582aFyjFr1iymoaHBxo0bx6ZNm8ZatmzJBg8erFDMuvL06VM2b948NmjQIPbVV1+x+fPnS8xUxZcQ9Vbbs+KXL1+y5cuX84rVtm1blp6ezm2vXr1asD6tdYmSJ3lvHj9+zBYvXswsLCyYoaEh++677+R6wVDFyMiIJSYmKlQmS0tLFhISwm1fu3aNKSsrS3TSbmqEqDchXwC+HasuXu7UBbptJ+9F//79cenSJXzxxRfw9fVFv379oKSkhGbNmsncL7DKb7/9hmfPnsk8ofKbVFRUkJycLNHhXF1dHYmJibzX4KkrdTWBsRD1JhaL8eLFCxgYGMgd481YaWlp3PwCb67o2ZDRCyPyXpw6dQrTpk3DpEmTYG1tLUjMWbNmYcCAAWjXrh3s7e2rrWXOZ3q78vLyas8flZWV622ykzc5ODhILFdchb0xWKGKLM+Khag3oHKmrDefeUrzvqcYfJ8oeZL3IiIiAps2bYKjoyNsbW0xcuRIDBkyRKGYU6dOxfnz5+Hm5oYWLVrI1U2GMVYtCRQXF2PixIkSc43WRxJ4cwLjyMhIzJo1C7Nnz+ZmQrpy5QqWL1+O3377Taa4QtQbUNlCVFdXl+vct705RPb169fYsmVLtZd2fIfIvi90207eq8LCQuzevRubNm3C9evXUV5ejhUrVmDMmDHV3sTXRltbG7t378aAAQPkLs/o0aN5fU/WeUaF5uTkhEWLFqF///4S+8PCwrBgwQKZuvIIUW9v32orwtzcvNYELssQ2do8fvwY/v7+2LRpk2KB6vF5K/nA3bt3j82ePZsZGRkxNTU19uWXX8p0vqmpKbfyZVOnpqbG4uLiqu2Pi4tjampqMsUSot7qYorB9yUqKkqQWZqo5UnqXXl5OY4dO4ZNmzbVuLyDNJs3b8aJEyewefNmaGho1GEJ61+3bt1gZ2eHf/75hxtxU1JSgjFjxiA+Ph63b9/mHUuIehOy5Sm02v4fSkpKwsyZMxWeaISSJ2m0unbtiocPH4IxBnNz82ovPmRJKA3d9evX8eWXX6KiooIbURQdHQ2RSITjx4/DycmJdywh6u3ChQv45JNPuIXzFFFUVISzZ89y0+7NnTtXYpE8JSUlLF68mPcwTbFYzL1oq4k8ayK9jV4YkfdizJgxtX5HJBLhn3/+4R1z4MCBCpSocXFyckJycjJ27NiBe/fugTGGIUOGYPjw4TKvQipEvampqeH06dPcvKVA5agjf39/FBQUYODAgVi9enWtb+Orzjt+/DiXPNesWYMOHTpwL6Pu3buHNm3aYPr06bzK1rp1a6xdu7bG3xkVFYXu3bvzivUu1PIk74VYLIaZmRm6du36zhbBoUOH3mOpiLw8PT3h6uqKH3/8EUDlpM/dunWTmPR5woQJvCZ9dnZ2xvTp0/HVV18BqN7Pc8eOHVi7di2uXLnCq2xeXl5wcHDAzz//LPV4dHQ0unbtWm1Mv6yo5Unei4kTJ2L37t1ISkrCmDFjMGLECDRv3lyQ2Ldu3eJWgbS3t0fXrl0FidvUKVJvUVFREquC7t69Gz169MDff/8NAGjbti38/f15Jc/ExESJNd7V1NQkJnd2cnLClClTeJdt9uzZKCgoqPG4lZUVzp8/zztejRR+5UQIT8XFxWzXrl2sT58+TENDg/n4+LATJ06wiooKueK9ePGCubm5MZFIxPT19Zmenh4TiUSsd+/e7OXLlwKXvukQot6EnPRZTU2N3bt3r8bj8fHxTFVVlVes94nm8yTvjaqqKoYNG4bTp08jLi4OHTp0wOTJk2FmZsYtLSGLqVOnIjc3F7GxscjMzERWVhbu3r2L3NzcBtehuiERot5atWrFdeIvLS3F7du3uc77AJCXl1ftRVRNTExMcPfu3RqPx8TE1Lj8tbz279+veJD6zt7kw/To0SMWEBDALCwsmLGxMcvLy5M5ho6ODrt+/Xq1/deuXWO6uroClLJpEqLexo8fz3r27MkuXrzIZsyYwVq0aCGxyuqOHTuYo6Mjr1jTpk1j9vb2rKioqNqxwsJCZm9vz6ZNm8YrVpWysjJ29+5dlpCQILH/8OHDrHPnzkxFRUWmeNJQy5O8NyUlJQgJCUHfvn1hY2ODO3fuYM2aNUhNTZVrQuKKigqprZtmzZop/DKgIZs8ebJC82QKUW+BgYFQUlKCi4sL/v77b/z9999yT/o8b948ZGZmwsbGBsuWLcORI0dw9OhR/Pbbb7CxsUFWVhbmzZvH78ehcnXX9u3bo3PnzrCzs8OgQYPw4sULuLi4YNSoUejbty8ePHjAO16NFE6/hPAwadIkpq+vz7p06cJWrlwpMX+jvLy8vJizs7PEnJZPnjxhLi4ubODAgQrHb6gUnbJNyHrLzs6WOn1fRkaGREu0NklJSczDw4OJxWKJBeg8PDxk/q1ffvkl6927Nzt27BgbOnQoE4lEzNramgUEBLDc3FyZYr0LdVUi74VYLIapqSm6du36znHMskzA8fjxY3h7e+Pu3bto27YtRCIRUlNT0alTJxw5ckTw52QNhaJTtjXkesvMzORahVZWVnL1yDAyMkJYWBi6deuG7OxsNG/eHMHBwRg3bpygZaXkSd4LPz8/XrP3yDMBx+nTp7mO4/b29ujTp488RWw0hJrvsqnWm1gsxvPnz9GqVSsAlQvJ3b59W6I7lBAoeRJC6tWgQYN4fY/vXYmSkhLS0tK4iZp1dHQQHR0NCwsLucsoDXWSJ43WtGnTYGVlVa17zZo1a/DgwQOFZkpvyhpavenq6goajzGG9u3bc3c6+fn56Nq1q0THewAKL9lMLU/SaBkbG+Po0aPVxinfvn0bXl5eePLkST2VrGFr6vW2detWXt8bNWqUQtehlidptDIyMqS2WnR0dBrMkrcNUVOvN0WTIl/Uz5M0WlZWVjhx4kS1/f/73/8a/OJh9elDrLfi4mJs3boV69atw/379wWJSS1P0mjNmDED3333HV69eoXevXsDAM6ePYvly5fT8853aOr1Nnv2bJSWluLPP/8EUDl8tGfPnoiNjYWGhgZ++OEHnD59WmI4qTzomSdp1NavX4+goCA8e/YMQOV6OIsWLcI333xTzyUThtBvoqs05Xrr2LEjfvnlF3h5eQGo7P42c+ZMREZGwtTUFGPGjMHLly8RGhqq0HUoeZIm4dWrV1BXV5drmGdDVtcL1DXFetPR0cHt27dhZWUFABg2bBi0tbXx119/AaicTq9///7cXxzyott20iRU9elravgkRUXWmG+K9SYWiyUm3L569SoWLFjAbevp6SErK0vx6ygcgRBSZ3bv3v3O42VlZfj666/fU2kaB1tbWxw7dgwAEBsbi9TUVLi5uXHHHz16xI0+UgS1PAlpwPz8/KCvrw8PD49qx16/fg0fHx/cvHmzHkrWcM2ePRvDhg1DaGgoYmNj0b9/f4nRRWFhYTItmFcTankS0oD9+uuv+Prrr6ut31NeXg4fHx9cvXoVZ8+erafSNUxff/01wsLC0LlzZ0yfPh179uyROK6hoYHJkycrfB16YUSahCdPnqBNmzbVhuA1Bf7+/li9ejUuXryIjh07ory8HP/5z39w6dIlnD9/Hvb29nLHbsr1VteoxkiTYG9vj5SUlPouRp0ICAiAr68v3N3dkZiYiKFDh+LixYs4c+aMQokTaJr1ZmxsjJEjR2Lz5s3cUiF1gZ55kiahqd9ArV69GtnZ2ejSpQu0tLRw9uxZdOrUSeG4TbHeJk6ciAsXLuC7775DcXEx2rZti969e8PNzQ1ubm6CzVdKt+2kSRBqjsuGZsaMGdy/l5WV4e+//8Znn31WLXGuWLFCrvhNtd6Ayvq6cuUKwsPDceHCBVy5cgUlJSWwtLRE7969ERwcrFB8Sp6kSViyZAkmTZoEPT29+i6KoN7sYlMTkUiEc+fOyRW/qdabNFlZWVi+fDlWr16N/Px8lJeXKxSPkichpEkqLi7Gv//+i/DwcISHh+PGjRswNzeHs7MzXFxc4Ovrq1B8Sp6ENCLp6elQUVGBjo5OfRelwfL398f58+dx48YNWFpawsXFhfsYGRkJdh1KnoQ0cNnZ2Zg/fz727NnDDSs0MDDA6NGjsWDBAmhoaNRzCRuWqsUG58yZAx8fH7Ro0aJOrkPJk5AGLDMzEz179sTTp0/h6+sLOzs7MMYQHx+PXbt2wdbWFpcuXUJ0dDSuXbtWbWmND9GJEye4W/XIyEi0b98erq6uXOtTqPH8lDwJacC+//57nD17FmfOnKk2HjstLQ3u7u6wsbHBqVOnsGrVqvc2i3pjkZeXh4iICFy4cAHnz59HdHQ0rKys4ObmhjVr1igUm5InIQ2Yubk5goODpY5tBypbWf3794e/vz/8/f3fc+kaj/Lycly/fh1Hjx7FunXrBHnbTiOMSKNTVlaGH374AVZWVnBycqo2bduLFy+gpKRUT6UT1vPnz9GhQ4caj3fs2BFisZhX4vyQ6q2iogLXr1/Hr7/+Ck9PT+jr6+PTTz/Frl278NVXX2HTpk0KX4NGGJFGJygoCNu2bcOsWbOQnZ2N6dOn4+rVqxKdnpvKDVXLli2RkpJS46iY5ORkGBoa8or1odRb//798e+//yIvLw9t2rSBq6sr/vjjD7i5uQk7GIAR0shYWVmxY8eOcdsPHjxg1tbWzM/Pj1VUVLC0tDQmFovrsYTCGTNmDHN2dmYlJSXVjhUXFzMXFxc2evRoXrE+lHobOnQoCw4OZomJiXV6HXrmSRodDQ0NxMXFwdzcnNv37Nkz9O7dG46Ojvjtt9/Qtm1bhZ9pNQRPnjyBo6MjVFVVMWXKFNja2gIA4uLisG7dOpSUlODGjRswNTWtNdaHVG/vRZ2mZkLqgIWFBTtz5ky1/U+fPmXt27dnffr0aRItqCpJSUmsX79+TCwWM5FIxEQiEROLxczDw4Pdv3+fd5wPpd48PT1ZdnY2tx0YGMiysrK47fT0dGZnZ6fwdajlSRqdb7/9Fowx/PPPP9WOPX36FK6urkhKSmpyLaisrCxuzXErKys0b95cpvM/lHoTi8VIS0vjngXr6OggKiqKe9754sULtGnTRuHfSS+MSKOzYMEC3Lt3T+oxY2NjXLx4EadOnXrPpap7+vr6Ci0f8aHWW121D6nlSQhpUt5ueb497Z5QLU/q50kaHWdnZ2RnZ3PbR48eRVFRUf0VqJGpqKiocX9qaup7Lo3wRCIRRCJRtX2CX4danqSxqe2ZFpEuNzcX3377LY4dOwYdHR1MnDgRCxcu5DrGC9Uiq29isRienp5QVVUFABw7dgy9e/eGpqYmAKCkpAQnTpygZ56E0N///CxYsADR0dHYvn07srOzERgYiFu3buHgwYNQUVEB0DTq8u3x/SNGjKj2nW+++Ubh61DLkzQ6tT3TItKZmZlh69atcHV1BQBkZGRgwIAB0NXVxdGjR5Gdnd0kWp7vC7U8SaN08uRJ6OrqAqh8Vnf27FncvXtX4jteXl71UbQGKz09HWZmZtx2ixYtcPr0aXh4eKB///7YuHFjPZau8aGWJ2l0+KwxLhKJqAX1FltbW6xYsQL9+/eX2J+fnw93d3cUFhbizp07VG880dt20uhUVFTU+qEEUJ27u3u1mZQAQEtLCydPnoSamlo9lKrxopYnaXLKy8tx7NgxDBw4sL6L0qBkZWXh2bNnNU5xl5+fj1u3bsHFxeU9l6xxouRJmox79+5h06ZN2Lp1K7KyslBaWlrfRSJNGN22k0atoKAAmzZtwieffIIOHTrg9u3bCAoKwrNnz+q7aA3OkydPkJ6ezm1HRETA19cXn332GUaMGIErV67UY+kaH0qepFG6cuUKxo4dCyMjI6xZswaDBg2CSCTCqlWr8O2336Jly5b1XcQG5z//+Q9u3LgBADhy5AhcXV2Rn5+PTz75BIWFhXBxccHx48fruZSNB922k0bH3t4ehYWFGD58OEaMGAF7e3sAQLNmzRAdHc1tE0k6OjqIiYmBubk5Pv74Y3z11Vf48ccfueNr1qzBpk2bcPv27XosZeNBLU/S6Dx48ADOzs5wc3ODnZ1dfRen0RCLxcjNzQVQuXyHp6enxHFPT08kJCTUR9EaJUqepNFJTk6GjY0NJk2aBBMTE8yaNQuRkZF1MvlDU+Li4oKQkBAAQNeuXREeHi5x/Pz58zA2Nq6HkjVOdNtOGrVz585h06ZNOHjwIIqLizFr1ix8++23aN++fX0XrcGJj4/HZ599hgEDBsDa2hq//vorBg4cCDs7OyQkJGDPnj3YsGED/Pz86ruojQIlT9Ik5OTkYOfOndwzu44dOyImJqa+i9XgPHz4ED/99BNCQ0ORn58PAFBWVsZHH32E2bNnU99YGVDyJE1OVFQUNm3ahFWrVtV3URosxhhevnyJiooKtGzZEs2aNavvIjU6lDwJ+UBUVFTwmheA8EM1SRql6OhoBAYGYt26dRIdv4HKSX/HjBlTTyVruJo1a4aXL19y27Nnz0ZmZmY9lqhxo5YnaXROnTqFL7/8EtbW1sjLy0NhYSH27t0LNzc3AE1nRnSh0Qz8wqKWJ2l0Fi1ahFmzZuHu3btISUnBDz/8AC8vL5w4caK+i9aoULtJMTQZMml0YmNjsX37dgCV83bOnj0bJiYmGDx4MEJCQhRanpcQvih5kkZHVVVVYvVMABg2bBjEYjGGDh2K5cuX10/BGoGFCxdCQ0MDAFBaWoqgoCBuRv4qK1asqI+iNTqUPEmj4+DggPPnz6N79+4S+4cMGYKKiopqC4CRSs7OzhLDL3v16oWkpCSJ79AoLf4oeZJGZ9KkSbh48aLUY8OGDQMA/PXXX++zSI3C28MxiWLobTshhMiBWp6EfCB+/vlnXt9buHBhHZekaaCWJ2l0xGKx1GdzOjo6sLGxwQ8//IBBgwbVQ8katq5du9Z4TCQSISEhAcXFxdQ/lidqeZJG59ChQ1L3Z2dn4/r16xgxYgS2bt0KHx+f91yyhi0yMlLq/qioKMyZMwd3797FuHHj3nOpGi9qeZImZ+3atdi2bRuuXbtW30Vp0JKTk7FgwQLs2bMHgwYNQmBgIKytreu7WI0GjTAiTY67uzsSExPruxgNVnp6OqZOnQpbW1s8f/4cly9fxp49eyhxyohu20mTU1RUBDU1tfouRoNTUFCA33//HStWrICVlRWOHTsGd3f3+i5Wo0XJkzQ5f//99ztfjnyo2rVrh7y8PEydOhXDhg2DSCSSOmF0586d66F0jQ898ySNzowZM6Tuz8nJwc2bN/Hw4UNERERQAn3Lm3N5ikQiiYlBqrZFIhG9beeJWp6k0anprbGOjg769euHyZMnw8zM7D2XquFLTk6u7yI0KdTyJIQQOdDbdkI+EJ9//jkOHjxY4/H09HSaGFkG1PIk5AMhFoshFosxf/58BAQEVDtOM/DLhlqehHxA1q9fjz///BNfffUVt/QwkQ8lT0I+IN7e3rhy5Qri4uLQs2fPavN5Ev4oeRLygbGzs8P169fRtm1bfPTRRzhz5kx9F6lRouRJyAdIV1cXoaGhGDduHPr3748//vijvovU6FA/T0I+EG9P4ycSibB06VJ07doVY8eOxblz5+qpZI0TtTwJ+UDU1LFmyJAhuHTpEu7cufOeS9S4UcuTkA/E+fPn0bx5c6nHHBwccOvWLYSGhr7nUjVe1M+TEELkQLfthBAiB0qehBAiB0qehBAiB0qehBAiB0qehBAiB0qehBAiB0qehBAih/8HBiTo7HJFp+EAAAAASUVORK5CYII=",
      "text/plain": [
       "<Figure size 250x150 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=[2.5,1.5])\n",
    "sns.heatmap(\n",
    "    w_filt.loc[ct_specific].T, \n",
    "              cmap='vlag', \n",
    "               center=0, \n",
    "               vmin=-0.2, vmax=0.2, cbar=False,\n",
    "              xticklabels=[\"QKI - 3' SS\", \"MATR3 - 3' SS\", \n",
    "                           'U2AF2 - complete exon', \"PTBP1 3' SS\",\n",
    "                           \"SSB - 5' SS\", 'QKI - downstream',\n",
    "                           'ZNF622 - complete exon', \"STAU2 - 3' SS\", \n",
    "                           \"IGF2BP1 5' SS\", 'EWSR1 - upstream']\n",
    "              )\n",
    "# plt.savefig('../../Evaluate models/Figures/Heatmap_ctspecific.pdf', bbox_inches='tight')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a6a4caeb",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.11.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}