{
    "cells": [
     {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
       "FaST-LMM Manual\n",
       "=====\n",
       "Factored Spectrally Transformed Linear Mixed Models\n",
       "========\n",
       "\n",
       "##### Version 0.6.8\n",
       "\n",
       "FaST-LMM Team, April 25, 2024"
      ]
     },
     {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
       "### Introduction\n",
       "\n",
       "FaST-LMM, which stands for Factored Spectrally Transformed Linear Mixed Models, is a program for performing \n",
       "genome-wide association studies (GWAS) on datasets of all sizes, up to one millions samples.\n",
       "\n",
       "See [FaST-LMM's README.md](https://github.com/fastlmm/FaST-LMM/blob/master/README.md) for installation instructions, documentation, code, and a bibliography.\n",
       "\n",
       "### Contacts\n",
       "\n",
       "* Email the developers at fastlmm-dev@python.org.\n",
       "* [Join](mailto:fastlmm-user-join@python.org?subject=Subscribe) the user discussion and announcement list (or use [web sign up](https://mail.python.org/mailman3/lists/fastlmm-user.python.org)).\n",
       "* [Open an issue](https://github.com/fastlmm/FaST-LMM/issues) on GitHub."
      ]
     },
     {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
       "### Citing FaST-LMM\n",
       "\n",
       "If you use FaST-LMM in any published work, please cite [the relevant manuscript(s)](https://github.com/fastlmm/FaST-LMM/blob/master/README.md) describing it. "
      ]
     },
     {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
       "### Data preparation\n",
       "\n",
       "This version of FaST-LMM is designed for use with randomly ascertained data with Gaussian residuals.  If you have case-control data with substantial ascertainment bias, you should first transform your phenotype(s) using [LEAP](https://github.com/omerwe/LEAP) (Weissbrod _et al._, _arXiv_ 2014).  If you are analyzing continuous phenotypes with non-Gaussian residuals, you should first transform your phenotype(s) using [Warped-LMM](https://github.com/PMBio/warpedLMM) (Fusi et al., _Nature Commun_ 2014).\n",
       "\n",
       "FaST-LMM uses four input files containing (1) the SNP data to be tested, (2) the SNP data used to determine the genetic similarity matrix (GSM) between individuals (which can be different from 1), (3) the phenotype data, and (4, optionally) a set of covariates.\n",
       "\n",
       "SNP files can be in PLINK format (ped/map, tped/tfam, bed/bim/fam, or fam/dat/map).  For the most speed, use the binary format in SNP major order.  See the PLINK manual http://pngu.mgh.harvard.edu/~purcell/plink/ (Purcell et al., _Am J Hum Genet_ 2007) for further details.  FaST-LMM also supports Hdf5 file format http://www.hdfgroup.org/HDF5/whatishdf5.html.  See https://github.com/fastlmm/PySnpTools for more details.  Note that each SNP will be standardized to have mean zero and standard deviation one across all individuals before processing.  Missing values are mean imputed.\n",
       "\n",
       "The required file containing the phenotype uses the PLINK alternate phenotype format with no header.  The covariate file also uses this format (with additional columns for multiple covariates).\n"
      ]
     },
     {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
       "### Notebook preparation and general use\n",
       "\n",
       "To prepare this notebook to run analyses, please run the following script."
      ]
     },
     {
      "cell_type": "code",
      "execution_count": 1,
      "metadata": {},
      "outputs": [],
      "source": [
       "# set some ipython notebook properties\n",
       "%matplotlib inline\n",
       "\n",
       "# set degree of verbosity (adapt to INFO for more verbose output)\n",
       "import logging\n",
       "logging.basicConfig(level=logging.WARNING)\n",
       "\n",
       "# set figure sizes\n",
       "import pylab\n",
       "pylab.rcParams['figure.figsize'] = (10.0, 8.0)\n",
       "#pylab.plot([1,2,3],[4,5,6])\n",
       "\n",
       "# set display width for pandas data frames\n",
       "import pandas as pd\n",
       "pd.set_option('display.width', 1000)"
      ]
     },
     {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
       "If you would like to run any of the code below from the command line, first copy it into a file (_e.g._, `test.py`), and then run it by typing `python text.py` at the command line.\n",
       "\n",
       "If you would like to see all the options for a function just type `? <function_name>` to an ipython prompt."
      ]
     },
     {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
       "### Single-SNP association testing\n",
       "\n"
      ]
     },
     {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
       "#### Traditional analysis: LMM(all)"
      ]
     },
     {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
       "First, let's run a standard LMM analysis in which the GSM uses (almost) all available SNPs.  The model for this analysis is called LMM(all) in [Widmer et al., Scientific Reports 2014](http://www.nature.com/articles/srep06874).  We'll apply this model to the synthetic data in `tests\\datasets\\synth`.  The data has 500 samples with 5000 SNPs, and was generated from a Balding-Nichols model with FST=0.05.\n",
       "\n",
       "When using a linear mixed model for association analysis, it is important to avoid proximal contamination ([Lippert _et al._, _Nat Meth_ 2011](http://www.nature.com/nmeth/journal/v8/n10/abs/nmeth.1681.html)). To understand proximal contamination, first note that a LMM with no fixed effects, using a realized relationship matrix for the GSM (as FaST-LMM does), is mathematically equivalent to linear regression of the SNPs on the phenotype, with weights integrated over independent Normal distributions having the same variance (_e.g._, Hayes _et al._, _Genet Res_ 2009).  That is, a LMM using a given set of SNPs for the GSM is equivalent to a form of linear regression using those SNPs as covariates to correct for confounding.  This equivalence implies that, when testing a given SNP, that SNP (and SNPs physically close to it) should be excluded from the computation of the GSM.  If not, when testing a particular SNP, we would also be using that same SNP as a covariate, making the log likelihood of the null model higher than it should be, thus leading to deflation of the test statistic and loss of power.\n",
       "\n",
       "Excluding the SNP you are testing and those SNPs in close proximity to it from the GSM in a naïve way is extremely computationally expensive.  A computationally efficient approach for performing the exclusion is to use a GSM computed from all but chromosome $i$ when testing SNPs on chromosome $i$ ([Lippert _et al._, _Nat Meth_ 2011](http://www.nature.com/nmeth/journal/v8/n10/abs/nmeth.1681.html)).  We call this approach leave out one chromosome (LOOC).  The analysis here does this.\n"
      ]
     },
     {
      "cell_type": "code",
      "execution_count": 3,
      "metadata": {
       "scrolled": true
      },
      "outputs": [
       {
        "data": {
         "image/png": "iVBORw0KGgoAAAANSUhEUgAAA0oAAAKnCAYAAAC4d70FAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjguNCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8fJSN1AAAACXBIWXMAAA9hAAAPYQGoP6dpAAD/5klEQVR4nOzdd3hb1fnA8Vd2HMdJbGfvvYAECBBCAoFCgAApUGYXbRml7FJmaeHXlg1l71Gg7CQQAgTI3s504jjDzvKe8Z7ykK11fn+cSrauJVuy5chJvp/n8ZNIupKOZd1zz3vGe0xKKSUAAAAAALewUBcAAAAAALoaAiUAAAAAMCBQAgAAAAADAiUAAAAAMCBQAgAAAAADAiUAAAAAMCBQAgAAAAADAiUAAAAAMOgW6gJ0NqfTKQUFBRIdHS0mkynUxQEAAAAQIkopqampkWHDhklYWOtjRsd8oFRQUCAjR44MdTEAAAAAdBF5eXkyYsSIVo855gOl6OhoEdEfRkxMTIhLAwAAACBUzGazjBw50h0jtOaYD5Rc0+1iYmIIlAAAAAD4tSSHZA4AAAAAYECgBAAAAAAGBEoAAAAAYECgBAAAAAAGBEoAAAAAYECgBAAAAAAGBEoAAAAAYECgBAAAAAAGBEoAAAAAYECgBAAAAAAGBEoAAAAAYECgBAAAAAAGBEoAAAAAYECgBAAAAAAGBEoAAAAAYECgBAAAAAAGBEoAAAAAYECgBAAAAAAGBEoAAAAAYECgBAAAAAAGBEoAAAAAYECgBAAAAAAGBEoAAADo0mw2kcbGUJcCx5tuoS4AAAAA4I3VKrJsmUhysojTKTJ+vMgvfiESExPqkuF4wIgSAAAAuqQVK0T27BFxOESUEklPF1m4MNSlwvGCQAkAAABdjsMhkpTU8v78fJHS0iNfHhx/CJQAAADQJTmdgd0PBBOBEgAAALqc8HCRk05qef/AgSKDBx/58uD4Q6AEAACALunyy0XGjWu6PXiwyK9+Fbry4PhC1jsAAAB0ST17itx4o0hVlYjdLjJgQKhLhOMJgRIAAAC6tD59Ql0CHI+YegcAAAAABgRKAAAAAGBAoAQAAAAABqxRAgB0aVlZImlpelH3aaeJ9O4d6hIBAI4HBEoAgC5rzRqRzZubbm/ZInLzzeyhAgDofEy9AwB0SVVVOjBqzmIR2bAhFKUBABxvCJQAAF1SUZGIUi3vLyg48mUBABx/CJQAAF3SoEEiJpP3+wEA6GwESgCALqlfP5Fp0zzvi4gQueCCkBQHAHCcIZkDAKDLuvxykXHjRFJTRXr10oFTv36hLhUA4HhAoAQA6LJMJpHJk/UPAABHElPvAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADEIaKD3//PMyffp0iY6OlkGDBsnVV18tKSkpHscopeSJJ56QYcOGSVRUlFxwwQWyf//+EJUYAAAAwPEgpIFSXFyc3HPPPRIfHy+rV68Wu90ul1xyidTV1bmPefHFF+XVV1+Vt99+WxISEmTIkCEyZ84cqampCWHJAQAAABzLTEopFepCuJSWlsqgQYMkLi5Ofvazn4lSSoYNGyb333+//O1vfxMRkcbGRhk8eLC88MILcscdd7T5mmazWWJjY6W6ulpiYmI6+1cAAAAA0EUFEht0qTVK1dXVIiLSr18/ERHJysqSoqIiueSSS9zHREZGyvnnny9bt271+hqNjY1iNps9fgAAAAAgEF0mUFJKyYMPPijnnnuunHzyySIiUlRUJCIigwcP9jh28ODB7seMnn/+eYmNjXX/jBw5snMLDgAAAOCY02UCpT//+c+SlJQkCxYsaPGYyWTyuK2UanGfy6OPPirV1dXun7y8vE4pLwAAAIBjV7dQF0BE5N5775Uff/xRNm7cKCNGjHDfP2TIEBHRI0tDhw51319SUtJilMklMjJSIiMjO7fAAAAAAI5pIR1RUkrJn//8Z/nuu+9k3bp1MnbsWI/Hx44dK0OGDJHVq1e777NarRIXFyfnnHPOkS4uAAAAgONESEeU7rnnHpk/f7788MMPEh0d7V53FBsbK1FRUWIymeT++++X5557TiZOnCgTJ06U5557Tnr27Ck33HBDKIsOAAAA4BgW0kDpvffeExGRCy64wOP+Tz75RG6++WYREXnkkUfEYrHI3XffLZWVlTJjxgxZtWqVREdHH+HSAgAAADhedKl9lDoD+ygBAAAAEDmK91ECAAAAgK6AQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMCAQAkAAAAADAiUAAAAAMAgpIHSxo0b5corr5Rhw4aJyWSSxYsXezx+8803i8lk8viZOXNmaAoLAAAA4LgR0kCprq5Opk6dKm+//bbPYy677DIpLCx0/yxbtuwIlhAAAADA8ahbKN987ty5Mnfu3FaPiYyMlCFDhhyhEgEAAADAUbBGacOGDTJo0CCZNGmS3HbbbVJSUtLq8Y2NjWI2mz1+AAAAACAQXTpQmjt3rsybN0/WrVsnr7zyiiQkJMiFF14ojY2NPp/z/PPPS2xsrPtn5MiRR7DEAAAAAI4FJqWUCnUhRERMJpN8//33cvXVV/s8prCwUEaPHi1fffWVXHvttV6PaWxs9AikzGazjBw5UqqrqyUmJibYxQYAAABwlDCbzRIbG+tXbBDSNUqBGjp0qIwePVrS0tJ8HhMZGSmRkZFHsFQAAAAAjjVdeuqdUXl5ueTl5cnQoUNDXRQAAAAAx7CQjijV1tZKenq6+3ZWVpbs2bNH+vXrJ/369ZMnnnhCrrvuOhk6dKhkZ2fLY489JgMGDJBrrrkmhKUGAAAAcKwLaaC0c+dOmT17tvv2gw8+KCIiN910k7z33nuSnJwsn3/+uVRVVcnQoUNl9uzZ8vXXX0t0dHSoigwAAADgONBlkjl0lkAWbAEAAAA4dgUSGxxVa5QAAAAA4EggUAIAAAAAAwIlAAAAADAgUAIAAAAAAwIlAAAAADAgUAIAAAAAAwIlAAAAADAgUAIAAAAAAwIlAAAAADAgUAIAAAAAAwIlAAAAADAgUAIAAAAAAwIlAAAAADAgUAIAAAAAAwIlAAAAADAgUAIAAAAAAwIlAAAAADAgUAIAAAAAAwIlAAAAADAgUAIAAAAAAwIlAAAAADAgUAIAAAAAAwIlAAAAADAgUAIAAAAAAwIlAAAAADDoFugTGhsbZceOHZKdnS319fUycOBAOf3002Xs2LGdUT4AAAAAOOL8DpS2bt0qb731lixevFisVqv06dNHoqKipKKiQhobG2XcuHFy++23y5133inR0dGdWWYAAAAA6FR+Tb276qqr5Prrr5fhw4fLypUrpaamRsrLyyU/P1/q6+slLS1N/vGPf8jatWtl0qRJsnr16s4uNwAAAAB0Gr9GlC655BL55ptvpHv37l4fHzdunIwbN05uuukm2b9/vxQUFAS1kAAAAABwJJmUUirUhehMZrNZYmNjpbq6WmJiYkJdHADAMWbXLpH9+0WiokQuvlikT59QlwgA4EsgsUG7st5VVVXJRx99JI8++qhUVFSIiMiuXbvk8OHD7Xk5AACOSsnJIj/+KJKRIbJvn8hnn4nYbKEuFQAgGALOepeUlCQXX3yxxMbGSnZ2ttx2223Sr18/+f777yUnJ0c+//zzzignAABdTlqa5+3KSpGyMpGhQ0NTHgBA8AQ8ovTggw/KzTffLGlpadKjRw/3/XPnzpWNGzcGtXAAAHRlxlkbYWEivXuHpiwAgOAKOFBKSEiQO+64o8X9w4cPl6KioqAUCgCAo8F554mMHq3/Hx4ucuWVIuyQAQDHhoCn3vXo0UPMZnOL+1NSUmTgwIFBKRQAAEeDyEiRm28WqasT6d5d/wAAjg0BjyhdddVV8tRTT4ntf6tVTSaT5Obmyt///ne57rrrgl5AAAC6MpNJT7cjSAKAY0vAgdLLL78spaWlMmjQILFYLHL++efLhAkTJDo6Wp599tnOKCMAAAAAHFEBT72LiYmRzZs3y7p162TXrl3idDrljDPOkIsvvrgzygcAAAAARxwbzgIAAAA4LgQSGwQ8ovTUU0+1+vi//vWvQF8SAAAAALqUgAOl77//3uO2zWaTrKws6datm4wfP55ACQAAAMBRL+BAaffu3S3uM5vNcvPNN8s111wTlEIBAAAAQCgFnPXOm5iYGHnqqafkn//8ZzBeDgAAAABCKiiBkohIVVWVVFdXB+vlAAAAACBkAp569+abb3rcVkpJYWGhfPHFF3LZZZcFrWAAAAAAECoBB0qvvfaax+2wsDAZOHCg3HTTTfLoo48GrWAAAAAAECoBB0pZWVmdUQ4AAAAA6DKCtkYJAAAAAI4Vfo0oXXvttX6/4HfffdfuwgAAAABAV+BXoBQbG9vZ5QAAAACALsOvQOmTTz7p7HIAAAAAQJfBGiUAAAAAMAg4652IyKJFi2ThwoWSm5srVqvV47Fdu3YFpWAAAAAAECoBjyi9+eabcsstt8igQYNk9+7dctZZZ0n//v0lMzNT5s6d2xllBAAAAIAjKuBA6d1335UPPvhA3n77benevbs88sgjsnr1avnLX/4i1dXVnVFGAAAAADiiAg6UcnNz5ZxzzhERkaioKKmpqRERkT/84Q+yYMGC4JYOAAAAAEIg4EBpyJAhUl5eLiIio0ePlvj4eBERycrKEqVUcEsHAAAAACEQcKB04YUXyk8//SQiIrfeeqs88MADMmfOHPn1r38t11xzTdALCAAAAABHmkkFOAzkdDrF6XRKt246Yd7ChQtl8+bNMmHCBLnzzjule/funVLQ9jKbzRIbGyvV1dUSExMT6uIAAAAACJFAYoOAA6WjDYESAAAAAJHAYoOAp96NHTtW/vnPf8qhQ4faXUAAAAAA6MoCDpTuvfdeWbFihUyePFmmTZsmr7/+uhQWFnZG2QAAAAAgJAIOlB588EFJSEiQQ4cOyRVXXCHvvfeejBo1Si655BL5/PPPO6OMAAAAAHBEBWWNUnx8vNx1112SlJQkDocjGOUKGtYoAQAAABAJLDbo1pE32rFjh8yfP1++/vprqa6uluuvv74jLwcAAAAAXULAgVJqaqrMmzdP5s+fL9nZ2TJ79mz597//Lddee61ER0d3RhkBAAAA4IgKOFA68cQT5cwzz5R77rlHfvOb38iQIUM6o1wAAAAAEDIBB0qHDh2SSZMmdUZZAAAAAKBLCDjrHUESAAAAgGNdwIESAAAAABzrCJQAAAAAwIBACQAAAAAMAkrmUFNTI/Hx8WKz2eSss86SAQMGdFa5AAAAACBk/A6UkpKSZO7cuVJUVCRKKYmJiZFFixbJxRdf3JnlAwAAAIAjzu+pd3//+99l1KhRsmnTJtm5c6ecf/758uc//7kzywYAAAAAIeH3iNLOnTtl2bJlcuaZZ4qIyMcffyyDBg2S2tpa6d27d6cVEAAAAACONL9HlMrKymTUqFHu2/3795eePXtKaWlppxQMAAAAAELF7xElk8kkNTU10qNHDxERUUq57zObze7jYmJigl9KAAAAADiC/A6UlFIyadKkFvedfvrp7v+bTCZxOBzBLSEAAAAAHGF+B0rr16/vzHIAAAAAQJfhd6B0/vnnd2Y5AAAAAKDL8CuZQ11dXUAvGujxAAAAANCV+BUoTZgwQZ577jkpKCjweYxSSlavXi1z586VN998M2gFBAAAAIAjza+pdxs2bJB//OMf8uSTT8ppp50mZ555pgwbNkx69OghlZWVcuDAAdm2bZtERETIo48+KrfffntnlxsAAAAAOo1JKaX8PTg/P1+++eYb2bhxo2RnZ4vFYpEBAwbI6aefLpdeeqn8/Oc/l7Awv7dmOiLMZrPExsZKdXU1qcsBAACA41ggsUFAgdLRiEAJAAAAgEhgsUHXGv4BAAAAgC7A7/TgInrq3XvvvSdbt26VoqIiMZlMMnjwYDnnnHPkzjvvlJEjR3ZWOQEAAADgiPF76t3mzZtl7ty5MnLkSLnkkktk8ODBopSSkpISWb16teTl5cny5ctl1qxZnV3mgDD1DgAAAIBIJ61Rmj59upx77rny2muveX38gQcekM2bN0tCQkLgJe5EBEoAAAAARDppjdK+ffvkzjvv9Pn4HXfcIfv27fO/lAAAAADQRfkdKA0dOlS2bt3q8/Ft27bJ0KFDg1IoAAAAAAglv5M5PPzww3LnnXdKYmKizJkzRwYPHiwmk0mKiopk9erV8tFHH8nrr7/eiUUFAAAAgCPD70Dp7rvvlv79+8trr70m//nPf8ThcIiISHh4uEybNk0+//xz+dWvftVpBQUAAACAI6VdG87abDYpKysTEZEBAwZIRERE0AsWLCRzAAAAACASWGwQ0D5KLhEREaxHAgAAAHDM8juZQ1syMjLkwgsvDNbLAQAAAEDIBC1Qqq2tlbi4uGC9HAAAAACEjN9T7958881WHz98+HCHCwMAAAAAXYHfgdL9998vQ4cOle7du3t93Gq1Bq1QAAAAABBKfgdKo0ePlhdeeMFnCvA9e/bItGnTglYwAAAAAAgVv9coTZs2TRITE30+bjKZpB2ZxgEAAACgy/F7ROmpp56S+vp6n49PnjxZsrKyglIoAAAAAAglvwOlyZMnt/p4RESEjB49usMFAgAAAIBQC1p6cAAAAAA4Vvg9ouRy+umni8lkanG/yWSSHj16yIQJE+Tmm2+W2bNnB6WAAAAAAHCkBTyidNlll0lmZqb06tVLZs+eLRdccIH07t1bMjIyZPr06VJYWCgXX3yx/PDDD51RXgAAAADodAGPKJWVlclDDz0k//znPz3uf+aZZyQnJ0dWrVoljz/+uDz99NNy1VVXBa2gAAAAAHCkmFSAOb1jY2MlMTFRJkyY4HF/enq6TJs2Taqrq+XQoUMyffp0qampCWph28NsNktsbKxUV1dLTExMqIsDAAAAIEQCiQ0CnnrXo0cP2bp1a4v7t27dKj169BAREafTKZGRkYG+NAAAAAB0CQFPvbv33nvlzjvvlMTERJk+fbqYTCbZsWOHfPTRR/LYY4+JiMjKlSvl9NNPD3phAQAAAOBICHjqnYjIvHnz5O2335aUlBQRETnhhBPk3nvvlRtuuEFERCwWizsLXqgx9Q4AAACASGCxQbsCpaMJgRIAAAAAkcBig4Cn3rkkJibKwYMHxWQyyeTJk5lqBwAAAOCYEXCgVFJSIr/5zW9kw4YN0qdPH1FKSXV1tcyePVu++uorGThwYGeUEwAAAACOmICz3t17771iNptl//79UlFRIZWVlbJv3z4xm83yl7/8pTPKCAAAAABHVLv2UVqzZo1Mnz7d4/4dO3bIJZdcIlVVVcEsX4exRgkAAACASCfvo+R0OiUiIqLF/REREeJ0OgN9OQAAAADocgIOlC688EK57777pKCgwH3f4cOH5YEHHpCLLrooqIUDAAAAgFAIOFB6++23paamRsaMGSPjx4+XCRMmyNixY6WmpkbeeuutgF5r48aNcuWVV8qwYcPEZDLJ4sWLPR5XSskTTzwhw4YNk6ioKLngggtk//79gRYZAAAAAAIScNa7kSNHyq5du2T16tVy6NAhUUrJ5MmT5eKLLw74zevq6mTq1Klyyy23yHXXXdfi8RdffFFeffVV+fTTT2XSpEnyzDPPyJw5cyQlJUWio6MDfj8AAAAA8EeX2XDWZDLJ999/L1dffbWI6NGkYcOGyf333y9/+9vfRESksbFRBg8eLC+88ILccccdfr0uyRwAAAAAiHTChrNvvvmm328erBThWVlZUlRUJJdccon7vsjISDn//PNl69atPgOlxsZGaWxsdN82m81BKQ8AAACA44dfgdJrr73m14uZTKagBUpFRUUiIjJ48GCP+wcPHiw5OTk+n/f888/Lk08+GZQyAAAAADg++RUoZWVldXY5fDKZTB63lVIt7mvu0UcflQcffNB922w2y8iRIzutfAAAAACOPQFnvWtuy5YtHtPcgmnIkCEi0jSy5FJSUtJilKm5yMhIiYmJ8fgBAAAAgEB0KFCaO3euHD58OFhl8TB27FgZMmSIrF692n2f1WqVuLg4OeecczrlPQEAAABApB3pwZvraMK82tpaSU9Pd9/OysqSPXv2SL9+/WTUqFFy//33y3PPPScTJ06UiRMnynPPPSc9e/aUG264oUPvCwAAAACt6VCg1FE7d+6U2bNnu2+71hbddNNN8umnn8ojjzwiFotF7r77bqmsrJQZM2bIqlWr2EMJAAAAQKfq0D5K8+fPl6uuukp69eoVzDIFFfsoAQAAABDphH2UfGEKHAAAAIBjUYeSOQAAAADAsYhACQAAAAAMCJQAAAAAwIBACQAAAAAMCJQAAAAAwIBACQAAAAAMCJQAAAAAwIBACQAAAAAMCJQAAAAAwIBACQAAAAAMCJQAAAAAwIBACQAAAAAMCJQAAAAAwIBACQAAAAAMCJQAAAAAwKBbqAsAAAjMnj0i2dkiY8eKTJ0a6tIAAHBsIlACgKPIwYMiixfr/+/ZIxIVJTJpUihLBADAsYmpdwBwFCku9rxdVBSacgAAcKwjUAKAo8gJJ4hEROj/d++ubwMAgOBj6h0AHEWGDhW5806Rw4dFRowQ6dcv1CUCAODYRKAEAEeZ/v31DwAA6DxMvQMAAAAAAwIlAAAAADAgUAIAAAAAAwIlAAAAADAgUAIAAAAAAwIlAAAAADAgUAIAAAAAAwIlAAAAADAgUAIAAAAAAwIlAAAAADAgUAIAAAAAAwIlAAAAADAgUAIAAAAAAwIlAAAAADAgUAIAAAAAAwIlAAAAADAgUAIAAAAAAwIlAAAAADAgUAIAAAAAAwIlAAAAADAgUAIAAAAAAwIlAAAAADAgUAIAAAAAAwIlAAAAADAgUAIAAAAAAwIlAAAAADAgUAIAAAAAAwIlAAAAADAgUAIAAAAAAwIlAAAAADAgUAIAAAAAAwIlAAAAADAgUAIAAAAAAwIlAAAAADAgUAIAAAAAAwIlAAAAADAgUAIAAAAAAwIlAAAAADAgUAIAAAAAAwIlAAAAADAgUAIAAAAAAwIlAAAAADAgUAIAAAAAAwIlAAAAADAgUAIAAAAAAwIlAAAAADAgUAIAAAAAAwIlAAAAADAgUAIAAAAAAwIlAAAAADAgUAIAAAAAAwIlAAAAADAgUAIAAAAAAwIlAAAAADAgUAIAAAAAAwIlAAAAADAgUAIAAAAAAwIlAAAAADAgUAIAAAAAAwIlAAAAADAgUAIAAAAAAwIlAAAAADAgUAIAAAAAAwIlAAAAADAgUAIAAAAAAwIlAAAAADAgUAIAAAAAAwIlAAAAADAgUAIAAAAAAwIlAAAAADAgUAIAAAAAAwIlAAAAADAgUAIAAAAAAwIlAAAAADAgUAIAAAAAAwIlAAAAADAgUAIAAAAAAwIlAAAAADAgUAIAAAAAAwIlAAAAADAgUAIAAAAAAwIlAAAAADAgUAIAAAAAAwIlAAAAADAgUAIAAAAAAwIlAAAAADAgUAIAAAAAAwIlAAAAADAgUAIAAAAAAwIlAAAAADAgUAIAAAAAAwIlAAAAADAgUAIAAAAAAwIlAAAAADAgUAIAAAAAAwIlAAAAADAgUAIAAAAAgy4dKD3xxBNiMpk8foYMGRLqYgEAAAA4xnULdQHaMmXKFFmzZo37dnh4eAhLAwAAAOB40OUDpW7dugVlFMlqtYrVam1xf1hYmHTr1s3jOF9MJpNERES061ibzSZKqSN6rIhI9+7d23Ws3W4Xp9MZlGMjIiLEZDJ16rEOh0McDkdQju3WrZuEhYV1mWOdTqfY7Xafx4aHh7s7ELrCsUopsdlsQTm2+fnZWceKtH4uU0d4P5Y6gjqCOiLwY6kj2ncsdUTHju0K531XqiNaO++MunyglJaWJsOGDZPIyEiZMWOGPPfcczJu3Difxzc2NkpjY6P7ttlsFhGRV155RXr06NHi+IkTJ8oNN9zgvv3yyy/7/MBHjx4tN998s/v2G2+8IfX19V6PHTZsmNx2223u2++8845UV1d7PXbgwIFy9913u29/+OGHUlpa6vXY2NhYuf/++923P/30UykoKPB6bM+ePeWvf/2r+/a8efMkJyfH67ERERHy2GOPuW8vXLhQ0tLSvB4rIvL444+7///999/LgQMHfB776KOPuivEJUuWyN69e30e+/DDD0uvXr1ERGTlypWyc+dOn8fed9990qdPHxERWbt2rWzbts3nsXfddZcMGjRIREQ2bdokcXFxPo/905/+JMOHDxcRkfj4eI8RTaObbrpJxowZIyIiiYmJsnz5cp/H/va3v5VJkyaJiEhycrL88MMPPo+9/vrrZcqUKSIicvDgQVm0aJHPY6+66io57bTTREQkPT1dFixY4PPYuXPnyllnnSUiIrm5ufLZZ5/5PPbiiy+WWbNmiYhIYWGhfPTRRz6PPf/88+WCCy4QEZHS0lJ57733fB579tlnyyWXXCIiItXV1fLGG2/4PPbMM8+Uyy+/XERE6uvr5eWXX/Z57NSpU+Xqq68WEX0xf/75530eO3nyZPnlL3/pvt3asdQRGnVEE+oIjTpCo47QqCOaUEdoXbWOaGho8Hm8UZdeozRjxgz5/PPPZeXKlfLhhx9KUVGRnHPOOVJeXu7zOc8//7zExsa6f0aOHHkESwwAAADgWGBSrY2hdjF1dXUyfvx4eeSRR+TBBx/0eoy3EaWRI0dKaWmpxMTEtDieIXPvxzJkzpB5Vx0yD/axIkyrac+x1BHUEdQRgR9LHdG+Y6kjOnZsVzjvu1IdYTabZeDAgVJdXe01NmjuqAqURETmzJkjEyZMaHVIrjmz2SyxsbF+fRgAAAAAjl2BxAZdeuqdUWNjoxw8eFCGDh0a6qIAAAAAOIZ16UDp4Ycflri4OMnKypLt27fL9ddfL2azWW666aZQFw0AAADAMaxLZ73Lz8+X3/72t1JWViYDBw6UmTNnSnx8vIwePTrURQMAAABwDOvSgdJXX30V6iIAAAAAOA516al3AAAAABAKBEoAAAAAYECgBAAAAAAGBEoAAAAAYECgBAAAAAAGBEoAAAAAYECgBAAAAAAGBEoAAAAAYECgBAAAAAAGBEoAAAAAYECgBAAAAAAGBEoAAAAAYECgBAAAAAAGBEoAAAAAYECgBAAAAAAGBEoAAAAAYECgBAAAAAAGBEoAAAAAYECgBAAAAAAGBEoAAAAAYECgBAAAAAAGBEoAAAAAYECgBAAAAAAGBEoAAAAAYECgBAAAAAAGBEoAAAAAYECgBAAAAAAGBEoAAAAAYECgBAAAAAAGBEoAAAAAYECgBAAAAAAGBEoAAAAAYECgBAAAAAAGBEoAAAAAYECgBAAAAAAGBEoAAAAAYECgBAAAAAAGBEoAAAAAYECgBAAAAAAGBEoAAAAAYECgBAAAjmoldSWyIn2FlNaVhrooAI4hBEoAAOCotipjlcTnx8uqjFWhLgqAY0i3UBcAABA4q1Xks8/0vzffLNKrV6hLBITOlIFTpKSuRCYPnBzqogA4hhAoAcBRyGwWOXxY/7+8nEAJx7fTh54upw89PdTFAHCMIVACgKPQgAEi116rR5RGjQp1aQAAOPYQKAHAUerUU0NdAgAAjl0kcwAABKyxUaSyMtSlAACg8xAoAQAC9sEHIm++KZKVFeqSAADQOQiUAAABczo9/wUA4FjDGiU/OBwi330nopRePN2NTw3Ace7220Xq6nRSCQAAjkWMKPmhslJk/36RAwdEKipCXRoACL2oKIIkdA02h00sNkuoiwHgGMTYiB8GDBC5/HI9ojRoUKhLAwRfXnWeLE1bKtOHTZdpw6aFujgA4BeH0yHvJLwjddY6uWv6XdIvql+oiwTgGMKIkp+mTxc566xQlwLoHCnlKVJUWyTJJcmhLgoA+E2JEqvDKg7lELvTHuriADjGMKJ0HCotFfniC5GxY0WuuSbUpUFXMGvkLOke3l1OHHBiqIsCAH7rFtZN7p5+t1gdVkaTEFQWi8jOnSInncQ04+MZI0rHobIyEbNZJCcn1CVBsNkcNkk4nCBl9WUBPS8qIkp+NvpnMqgXc0sBdL7qhmrZkL1BqhuqO/xavbv3JkhC0G3bJrJ2rciyZaEuCUKJQOk4dNJJIr/9rcgf/hDqkiDYdhbslKVpS+WHQz+Euig4xjU2irz9tsh774nYmfGEAG3I3iAbsjfI+uz1oS4K4NUJJ4gMHy4ydWqoS4JQYurdceqEE0JdAnSGMX3GyJDeQ2TKoCmhLgqOcY2NOgtoWJiIzca2CQjMKYNPkXJLuZw6+NRQFyVgh8oOSWldqcwaNUvCTPQ3H6uGDxe57bZQlwKhZlJKqVAXojOZzWaJjY2V6upqiYmJCXVxAOCYUVSkAyWygeJ48tym58TqsMofTv2DjO83PtTFCaqfUn6SfHO+/O7U30lMJG0mHJsCiQ3oAwQAtMuQIaF9/8xMkaQkkdmzRWJjQ1sWHD8uGHOBFNcWy8jYkaEuStClVaSJudEs5fXlBEqAECgFVWWlyPz5IuPHi1x2WahLAwBHD6tV5L//1VP4br1Vj1S1ZfNmHSwNGCBy7rmdX8ZgU0ovFLdaRa66yr/fGaF3zshz2jzGqZySWJAow6KHyfCY4UegVMHxh1P/IJUNlTK279hQFwXoEgiUgqi8XKfeNplCXRIAOLpYrTojZ3i4Tg7RvXvbz7nwQpHBg0VOP73zy9cZrFadflgpPSrWp0+oS4RgSS1PlaVpS6V/VH+5d8a9oS6O3wb2GigDew0MdTGALoNAKYgmTNCZ5Pr3D3VJALikV6TLuqx1MmfcHHpJ22C3i2Rni4wZc+STM/TuLXLXXXpUxZ8gSURkxAj9c7SKjBS54QadDIMg6dgyMmaknND/BBnXd1zIypBvzpfvDn4nM0fMlLOGnxWycgBHMwb6g2z8eC54QFeSWp4qBTUFkl6RHuqidHkbNoh8+aX+NxQGDBDpd5xthzNxosjkyaEuBYKtV/de8ttTfiszRswIWRmKaoukwlIhOVVHdtPEWmut7C3aK3Yn+wbg6MeIEoBj2oVjL5Th0cPlxAEnhrooXd6IEbqjZ/jRs6QCXZDNYZNVGatkWPQwOX3okZ0XWVxbLH169JHIbpGd8voZFRmSW50r540+T7qFde0m1LSh06RPjz4yPPrIntAr0lfIvpJ9YrFbZOaImT6P21+yX/pF9ZOh0UOPYOkQaoWFukMssnNO0aDr2mc5AHRQj249ZOoQdgz0x4kn6h+EXmWlSEnJ0bnnXUFNgSQUJEhsZGynBko5VTlisVvcnSBZlVny2d7PZFzfcXLj1Bs75T1XpK+Q0vpSGR4zXCb1n9Qp7xEsJpNJJvSbcMTfd/LAyVLTWNPqtMOCmgL55sA3EhMZIw+e/eARLB1CKTVVJz074QSR3/421KXxD4FSOzkcIp99JhIRIfL735PAAQAQPIsWiRw+LHLjjSLjQrfMpV1GxY6SuRPmyqBenbvB1rzkeWJ1WOW+GfdJ36i+EhMZI3169JGhvTtvhGLO+DmSU5UT0rVHIiIldSUyoOeALrnh7eSBk2XywNbnkw7oOUCmDJzCaNJxpk8fvZVDqLeWCAQbzrZTQ4PIK6/oBc8PP6wzNQEAEAzx8SJpaSLXXivSq1eoSxMaxbXFUm4p99nojsuOE3OjWS6fdHmXDBg6S8LhBFmatlR+NvpncuHYC0NdHOCow4azR0CPHiL33KMzNBEk4Xi2v2S/xETGHJObLwZDo71RPtnziQzqNUiuPenaUBcHR4mZM/XP8eybA99IWX2Z3D7tdhkWPazF4+ePOT8EpQq9vlF9pVdEL+kf5TvFbll9mVgdVq+fGwD/HT9dMB3gdOo9koz69BEJ4iAVcNQpry+Xbw58Iwv3Lwx1UTpdg71BNmRvkLL6soCe1+holNK6UimsKeykkvnH6dTrXoCjxYzhM2Tq4KkysCf7+jQ3od8E+eusv7a69vLj3R/LR7s+klpr7REs2bFn4UK9EbY9CAn8rFaR7dtFqqs7/lodsWWLyJ49oS2DNzabyL59+nNqrw0bRN5+W6SqKlilYkTJL6tXi2zbJvKrX5HGFWiub1RfmTlipgzoOSDURel0+0r2yYbsDVJhqQhoZCgmMkbunXGvRIaHNsXPihUiO3bofXsmde016MeFxYtFsrJEbrtN7yGFlqYPny7Th08PdTGOSlMHT5Uaa41EdYsKdVGOaocPi9TX60Z8R/eW27NHZPlynfXt6quDUbrA1dToNm337iKnnRaaMviybZvIunUi550nctFF7XuNwkI9sFFbG7ytegiU/DBwYNMCNHRcY6M+WQcc+23rTtVgb5BdhbvklEGnSHRkdEjKEGYKk8smXBaS9z7SpgycIlUNVXLKoFMCfm6fHn2CX6AA9e+v67Do0HxVYGA26wZYMHqqj2Y2h01qrDXSL6rjG2hVN1RLVESUdA/3c8fiY9ilEy71+djS1KWSWp4qfzrjTyG7dhwt7rhDn6NRQYg3TzpJpLhY5IwzOv5a7RUdrYO0nj1DVwZfJk0Syc/3nemzvl6v3Zw6VV/PvLnuOt2+9PV4e5DMAUfc/Pl6kfLtt4sMJeFNu23J3SKrM1fL2SPObvWi2Bm+TPpSqhqq5I5pd0hEeMQRfe9jhc1hk/d2viexkbFy02k3hbo4x6z6et2DeuqpImPHhro0TRwOPcUkGA2wo9mb29+UxYcWyz3T75HrJl/X7tepsFTI2zveljF9xvidGnxTziYpqCmQKyZdIb26Hz8ZMxYkL5CMygy5e/rdQQlQ0bmSknTjf9asUJcktLZv1yNyZ54pcsUVHXstkjn4kJsr8tVXInPmiJx+ZPfAQzOjR4tYLJ3bs11VpYOxM844dpJt2J12Sa9Il/F9x0tEeIScMvgUqbHWyJnDzjziZam31YvFZhElx3Q/S6dSosTmsInVYZWy+jLZU7RHZo2cJVERx3nLOciys0V279ZTZ7pSoBQe7hkk1dToqZFnnnl8zV4Y1GuQRHeP7vD3vke3HjK412C/kxfkVufK2wlvi81hk8kDJ8spgwMfKe4Ih9Mh5ZbyTk+h7s2vpvxKrA4rdc1RYuVKkbo6PVXueM2AKaJHkhoadKfXkXRcjSgVFcXIggUiF16o50B2hNMpsnGj3sF+4sTglBUdV12tex1KSkTS00Wuv17k5JPb91oNDbqCmjw5uH/jbXnbpLKhUuZOmCumADbg2pq3VVZlrJLZY2bLtGHTJDI8MmSjOQ6nQ5zKyWhSBzmcDjGZTLI8bbkkFCTIlZOulGnDpoW6WMcUp1Nk/37dQRPopIKCAr17fDCncfiyaZPI2rUi558vMnt2579fV6KUCqguDIYGe4MsOrBIenTrIdeceI2Ehx3ZHrXlactl++HtcsMpN3T5jWsRWtnZOlCaMiU4r6eUXgs0aJDIKUe2f6DLYETJh0mTRP72N53au6PKynR2jWHDCJS6kj17RLZu1T0Ps2aJTOjApuSFhbonuqFB/40rKkTy8nRvRkeu6QkFCVLVUCUXjLlAeka0nChcVFskKWUpMmvULOkW1nSKTuw3UQ6bD8uImBHyRvwbMjxmuNx82s3tL0gHhIeFS7gcI0N1IeRqnJ076lzpF9XviPdqHw/CwtrXGGhs1NmuevcWeeCB4JfLaNo0XdautsD6SDjSQZKIHoH6/am/d9+usFSIiByxqWgjY0dKvjm/1RTfgIjImDHBfT2zWXfMDBx4/AZKgTiuAiWR4ARJIjoSv+46/UWDJ7NZZ1QJ1mcdiLPO0tNZTj214+8/dqwefdy/X/csx8WJpKToaTEdqbh+f+rvxWKzeA2SREQ2526WfSX7ZGj0UI+exoG9Bsovp/zSvTfGiJgR7S/EUWZ7/nbZWbBTfjXlVzKw17F30sX2iJWzR54d6mKgme7ddX1ypDLS9ezZOWsQzGY9Dfm0046dacid4YPED0RE5O/n/v2IvN/Jg06Wkwe1Pt1h4f6FYrFZ5MapN3oEk1UNVdItrJv07k66RAQuNlZnPz2epvh2xHEXKAWTP5G4UvpCdbx8IS0WkbfeEhk8WORPfwrOazY0iHz+uR7VaWtKSlSUbtwES2SkzlJTUKAbMYMHi4zoYHzSL6qfSCtTwy8ce6GMjh0tE/p5Hw7rHt5dbjn9loDe81DZIYnLjpNrT7q20wKNqoYq6d29t8coWLCU1ZfJ5tzNkl+TL3+f9XeJ7eH7hCqoKZDU8lQ5d9S5Xsvy/cHvpd5WLzecckNIerIRXK50t2ed1fFzszmTSeTSI5sjpVNs3Ciyc6fuOHJN3VFKZ/KK6ODM2VprrazJXCPThk476jecnja06015La8vF4tdrwU1ia6r7E67vJvwrkR3j5Z7Z9wb4hIGzuqwSpgprFOuE82lpOiO7H7kqvCKLSL8R6DUybZu1Rfx3/7Wd8rDY0n37nq625AhwXtNq1VPdQy0wktK0vseuPa+Ukpk0SK9TsHfBtBZZ+nRo8GD9e1RowIrQ3v0i+on/YYHt3YvqSuRwtpCqW6s7pRAqaSuRN5LeE+mDpkqV594ddBed3XGaokIj5C5E+dKmClMsqqy2lwXtTVvq+wr2SejY0fL2L4tV+8X1RZJva3eo/GBo1denj7Xe/cObqB0rJgxQ49WNZ+GvGiRSEaGyL33dmxxeF51nuwp2iM9uvVoNVBSSsn67PUyuNdgmTIoSAstgmzO+DmhLkILt0+7XZQoCTOFue8LN4XLKYNOOSpHk2wOm7we/7oM7Dkw4M6+QJSUiCxYIDJunMiN/iVA7LCGBpHNm/W0f2YahZbd3vE9r5ojUOpkgwfrhA99+4a6JEdGeLjIr38d3NeMiRF56KHAez9/+EGPCLkCJadTb/AYyIJuk6kpSAo1h9PR7gXH5406T04bcprERHZOivzo7tEyru84GR07OmivqZSSxMJEiQyPlAvGXCBzJ87163kXj7tYJvabKGP6jPH6+G3TbhOncno0PnD0Oukk3RgiSPJu4EA9hbi5Pn30NamjjYkTB5wofzj1DzKw50D5ZPcnMqn/JJk1quX8wTpbnWzK2STDood12UCpK/JW35tMJrnyhCtDUJqWGuwN0qOb7znuxtGj8LBwGdJ7iCilpNHeKJHdOmcT7gEDdFKU0cG7HLUpI0MHSna7yGUh3FowPV3//sHabPVok5kp8sUXOn34tCANEh9XWe/YR+n4kpmpA7fmlaXFou/r3s79CJXS+590pIFRVFskA3oOkG5h3cRis/iVonVX4S5ZkrpEfn/q72Vc33Htf/OjgNVhlQZ7g8RExkhVQ5WEmcI6LcA73jkc+l/WrqAjahpr5M3tb8qJA070uRdSbnWuRHePlr5Rx0mvYTPpFekyoOeALrHxdLCkV6TLvKR5cvmky71uUeFUTnlt22sSExkjt027zX1/TlWOfLLnE5k+bLpcPunyI1lkDxUVOrhxJVHpKKdTZN8+PXIbqs1cy8v10oeJE0V+97vQlCHUCgpEvvlGzxo68UTPx1JSdPbAM84ILDagSxUhtX+/yEcf6bTewTZuXMsepagoHSTl5OgTKlA//ijy0kv6ZGuPrMoseX/n+7Ihe4NkVmbKi1telB2Hd7T5vN7de0ufHn0kqtuxs++F1WH1ev+iA4vkze1vSp21Tvr06EOQ1Ik++EDkP/8J3fuXlYns3Ru690dwREdGy8PnPCzXnHSNz2NGxY7qUJCklJL5yfNlVcaqdr9GKFRYKuTLpC9laepSr4/bHLZ2vW55fbnU2+o7UrQOie4eLYN7D/YZ/IWZwmREzAiPpENLU5fKvpJ9Mm3otDYTWXS2jRtFli7VU3eDISxMJ5EKJEhKSNCzXFwOHBD58EO9D2R79O2r13Gfc077nn8s2LJF/y28ZaNetUpkyZKmDkJ/ESj5YDbrHoLycr0JoNMZ6hK1zmYT+fprnR47WGpqRN5/X2TXrsCeZ7fr1JPl5W0fW1amA5b6I1zff/GF7nVwSU0Vee01kcOHW3/egAF6Kp5rRMnp1POh/TWo1yAZ0HOAJBxOEJvDJsNjhvuVHnZS/0nylxl/kaHRQ/1/sy5sc+5m+ffmf0thTWGLx07of4JMGTil1SkdwVJvq5clqUukqLao09+rKxoypP1TS51O3SMb6EWnubVrRb7/PrBzCF1TZLfITp3O6lROKawplOLa4k57j87Qt0dfuWjsRXLuqHNFRI++7SnaI07llL1Fe+W5Tc9JRkVGQK/ZaG+UdxPela/3fe3zGKdyyta8rVJaV9qh8vsyuPdgufPMO30mHRIR+fXJv/aYMp1SniI51Tly5QlXyug+R3BenBc/+5nIlVeKjAxRDhKLRWTZMr2fUX29rgPLy/W2JO1tD4WF6SmHnbGxdkmJDjLa20l8pERG6sQ13vI0/fKXepp2oDMoWKPkRWGh7mk991z9pdi1S++X1NXmwDudOkCKjNQnXVqa/n+w9uGw2fSJW1MT2PNyc3UDqL6+7aQJ558vMnOmLveRdPnlempefb3uAbLZ9GfYVqNv1izPFL6bN+uK7sYb9QhWW3p17yUzR8yUZWnLpEe3HvKnM9qXGnBF+grJrsqWP53xp07PHpRRkSH9ovoFdcpM/6j+Mjx6uNcU6dOGTTtim64W1BTIzoKd0iuilwzprTOQHCg9IPtL9stVJ14l3cPbOUezC7BYdMfJGWf4Pr+u8T0A0KY9e/QI6xVXiJzZcuaNX2bP1j1/LH4+MlJT9fXsqqv06HqofZj4ofTo1kP+MPUPbR4bHhYu982876hbW2gymeS80U073G/J2yLx+fESGxkrMZExMqT3kIATM3QP7y4zR8yUQb0G+TymoKZAVmWskrL6MvnFCb9od/mD6e7pdwf092toEFmxQrdpRo3S1+eOZmp06dcvtBnxoqJE/vAHvWZ68WK9tuihh3TylfYuDWhOKZGVK3Xnbnvr5+ZSUnT2zIkT/U9M9vXX+tpz9dUdf39vlGoZEP2ila96e5OMHbeBUlKSHoa76SbPi7TVqkcLTjxRT9saNEhH58OHh66svvz0k566dv/9+mR74IHg7l3Ur5/Io48GPn937Fid0MHfhZRHOkgS0RVRcrL++553nk6b255drydM0MHkIN/XqxbOHHamTB08tc3sba1xKqc4nB3oyvdTrbVWvkj6Qib2myi/O9Vz0nODvUGSi5Nl6pCpAQcUJw08SU4aeFIwi9ouE/pNkD+d8Sd3kCSip0ceKjskF1svlu5RR2+glJysL5SRkTpYCraJE0XOPltPnausFJnzv6RhmZm6g8mf7RMGDQrs3OmovDxdVx4v2zUY5ebqYKmmxneg5HTqzp9x4/zr/OmIXt17BTRy7K1TyOawycL9C2VS/0kyffj0YBavU5w94mzpH9VfRsWOkjBTmNxx5h0Bv4bJZGozS9+ImBHy6ym/7lL77QU6S6CyUnfI9OypG+kZGbqdE4xAoitwnV9nnKHrwZ49O7aZvYvdrjt/d+/WnfzBCJTOOUdnAA5kBK6qqvP209y5UwfRt93W+Qm3jttAyWTSQ4m1tZ6B0tdf6+lXf/1r0/BcV925eORI3ePiqjSap3nNy9NZ366+umMjYe1Z5Ggy6UxU7bVmjX5fY6amYDrhBD3s7sqI117DhgXWK786Y7U4lEMum9CxtDg/n/jzDj1/U84miewWKWcNb33Tqd7de8uVk66Uwb0Gt8gUl1ScJMvSlklEeIScNuS0DpUnlIwNicsmXCbnjzn/iKbfdaUsD2aijtNO88z62FEWi15PePLJeiQoOlrkkkv04uGKiqbjVq/WdevJJwfnoh8sDQ0iH3+sGyd/aHsAo005Obq3tiPptY+0Cy/UI/itbaJbW6vn+VdXd36gdMMpN3jczqvOkxXpK+SqE69qdbSkObvTLnnmvFb3VgsGm8MmYaawdmcedYntESvTh0+XCkuFvJfwnlw49sJO22z6SHdGNdobZVfhLjltyGl+JSlqy9ChOoV9nz56PY/TGVgipZoavYzCW0d3ZqYeEf/Vr/R1PJROPLFl4oGO+Phj3en/5z8HFlQWFOjn/uIXer1Vc+HhgU9TvCPwPgC/9e6t2+5t/X61tTqgOuec9v+dj64x7CBwOPSCuT59dPR+8KDn45Mni5x++tGRBeqMM/TIjbeKw2bTXxBb+9aJBmzRIh3gdMT69XrtUGqqnkbYWSwWXUEOHtyyV3X37uCu8zLKqc6RrMqstg/sIHOjWd6If0N2FXpfYLazYKfPx4ymDZsm67LXyXsJ73ncf9qQ0+TqE6+WKQP9H4qLz4+X1PJUv48PhfCw8CO+R8mPKT/KvKR5EqwkpK7zf+rUllNVUlJE8vP1/9evb1kHepOTI/L883pNYfO60WTSDZnmWwJcd53IzTe3HSTt26cDr0DnvDc2isyfr5/fnNXa1DBqbGz5vB49RH7+c702oaMqK0U++UQHhc3Z7frz7Mi6raqqpjTDgSora/3xsLDWgyQRPeJ21116enJnarA3tEhGUGutleyqbKm0VPr9OlERUfLXc/4qJ/Q/Qb5M+lLqrMFfRFFvrZcn4p6Qr/Z9FbTX7B7eXYb0HnJMJatJKU+RlRkrZV/JvrYP9lP//rrOmTlTBzWBdN4uWaLrmIaGlo9VVuqp9119/Xlb6ut1ndfcxIl6Q9no6MBm7ERG6uCjeedPcbEOUruaE0/UgVhbW+9UVOhrRW5u+9/ruAuUDhwQWbhQ//GvvdZzvYmIThXZlXZjr6pqCnbWrBF5552WF9Dych1gFDVbjz5unMjf/945i/q8KSlpukiXlvrX+DKqq9Mn/K236h8R3euTlBS8corozzQ5WTf+jDZu1IkoXEpLdcMyGJzKKRP7TQzqhqytqWqoki25W7xmVbrjzDvk5tNulqzKLPkp5SefmZdSy1Plk92fuKeKNNc9vLucNuQ0v6cQOpVT1mSukfj8+MB/mS6i0lIp85LmSXx+vOwp2hO017184uXym5N/IyYf0UVKWYqU1HnPeFBfrxPONO8UWbFC5O23va8v/OYb/bjdrjfE9qdjoFcvPe3itttaBhrGIrsSnrz2ml6r6Etdna67Au3MqavT52lmpmdA8v33Im+8oX++/db7c6dPb31KcG6u7vSpr9eNq6VL9bXCqE8ffZ2YOdPz/qQkPSth//7Afqfm9u7Vdb0rmPVXUpL+m7en7jUaNKjzpsy4fL73c/kg8QOP+04aeJL06dFHtuRtCei1wsPCpay+TLIqs8RitwSzmCIisj57vWRVZkm/qOAtaundvbfcesatXWZfqbTyNJ9BpsPpkJ9SfpKUstYvhlMGTpHZY2ZLckmyVFgqWjyeU5UjK9JXeEwbTytPk3xzgF92L5KSPDPIiej23dy5Lb/LNpvI8uW6fdR8xk1mZvAy8BYU6LXuzdtlnWH+/JaZS2fP1qP9gerfXwcf48c33bd1q64HOyMz8ZEwapSerjljRvtf47iZeufqNTjhBN2rePLJnX8h6Ki6Oj2t5bTT9DSxyEjvc1jNZl1BVFX5v1gtJ0e/frCm5dx9d9P/167VwcXf/x5Yb8YVV+jFed9/r7/cZ54psmGDrmisVh2IdXQjt7o63QC57z7v6xRuucXz823v7+JNpaVS1mevF4vd0unZ62IiY2TWqFmyNW+r1FhrWlzgXUkUMiozZHfRbjlv9HnSJ7xPi9cxN5qlqLZIrjrxKo/XyK3OlazKLPnZ6J/5bNwbp+qFmcLkrjPvOiLZ7DpLdWO1pFekS1ZVloSZwoI25XB4jO9FkHanXb7a95VM6DehxToxER30L1+uv8+uRbannKJHS71NC7vxRv1Yt256NKi173VNje6VHDBA5I9/9P/3CQvTaxyjo30fM2NG6xcvh0PXf5GROnC56SZdDteawM2b9f9vuUUff+qp+vHw8PbPWS8o0A2P0lJd5yYk6N/D+Homk16fZTR5sq6r/F3s7M3ZZ+spLmPG+He8q7d34kRdZ3bF9bQuTuWURnujREVEybSh06TR0XLo7+yRZ7drRPfskWfL9OHTOyW5zVnDz5L+PfvLiOgRklycLKcM7qLz8duptK5U5iXPkxnDZ8jssbNledpymT58untacoO9QfYW75XwsHA5YYDvL3d4WLgM6jVI4nLipKax5XXnUNkh2XF4h8wcMdOdVvzbg99Kv6h+cvu02zv0OyxZots/zTuHR43SPyJ6TUtJiW7/RUToRFLN20sWi8jnn+tz+Fe/al8ZKir0+T9kiO5oKS/3PrptZDbrZA6nnx74dOUZM4KXNTg9XZe5eb08Z44uV1dY11lerq8tZrOuI/0dXexo2Y+bDWdTU6tl4sSja4jb6dRTY8aPb/ui6XAENl3w4491T+mjj3aoiF5VVuqf9sxvdzhEXn9dDxtfeaXuxWhs1J/D4cMiDz7Y9vMrKlpm0bJa9WO7dunpMjfd5N9oW0WF/pngOwNqC3anXcJMYV6z+xTUFEj/qP6dtiN5c07lFHOjWeYnz5cJ/SbIJeNbdjE5lVPqbfUeDRO70y755nwZ02eMz9dekrpEdhXukofOfkh6dW/ZGt+Su0XicuLk3rPulejIVlrLnexA6QFZkrpEbjntFhnYq32p1RxOhyxJXSInDTxJJvWfJDaHTSx2izTaG9v9moHKrMyU2MhY6d+zZSp5m00vcp40KTgbJ7okJYl8953/54o3TqceuWrP4mul9OiXyaQvkL/5jR7JUUpf0LOydMB33nltvlRAFi3SAebcuboe69MntGutDh3SjUBj4iGX99/Xn8lddx35sgVqZfpKSShIkAdmPuC13ujq5iXNk8zKTPm/n/1fwNn3rA6rWGyWTl9L1R5KKdldtFvG9R0nNodN3tv5nswZN8dj7ZTFZvE7BbzD6fC6lsvhdEiNtcZj76V8c75EhkfKwF4DZVNKspga+sm5UwOP9ktKdKeKr0bxokW6g/jBB32fz/v26U6R9mbhfOIJvXTg888DW0u1bp0eJb/jDr0uK1TmzRPJzhZ57LGutb7U5bXXdJ0cFqbXhsfE6BkC7bnuBbLh7HETKPnzYTRXV6dPukC+7O1RWqp7QY/0l7KqSgcg/va8utJoHwlOp+7JLi7WjT/Xff6kBo2L05XOn//sWdl9/rkOeO6+W1eWEyYE7zNfuFCX7be/1bdf2/aajIwdKddPvj44b9ABSilZsG+BjOkzRs4ZqXehq7fVS3x+vMwcMdNreu74/HhZkb5CbjntFp97XdgcNqm11vpMGZ5Wnia7CnfJNSdd0+EU23XWOqlqqGp11MWX7KpsWZe1Tq476bp2N1AsNou8Hv+6TB8+XS4ed3G7XqOrq6jQ073mzGma811RoRf1z56t17U4HDo4GTfO/wvTkiW68fHQQ8FL6+uL1ao7VKZN0/XHihU6uGprDntXl52tO3d++UsdtBm51l+0NkPCZtOfT6gTT2RUZEhKeYpcOv7SDidFCAVzo1nqrHXtmhHw3cHv5GDpQfnbuX/zOepld9olpSxFThxwYqd9PuZGs+Sb82XywJbTSWwOm0SER4jFZpEe3Xr4nC2QVZklhbWF7mtKIH449IPUWGvk96f+3uP+Rnuj3PzsErEWT5Sv3jg16PWF0xl4MghfXPtsGs/Hr77SnbFPP+3fDJSqKt3JERWlE3B52yS1sxw6pDufL7qo6T6LRbcLvdUzwZScrAPDm28OrE46eFB/7hUVerbC4sU6YJo61f/XsNv1LKFhw8zSr59/scFxtUbJ24I+bxwOPc99qZeNtK3W4JUnM1OvOQrWGpxA9jvq08f/ICkvT+Sllzo27z4QYWE64Jk/v2lIOSzMv4bW8OH65M/O9rx/6lTd05SfrysjY/1fVqYXq3/7rR7RqqvTf/+PP277PWNjPXuxTh50soyOHe11jnagqhuq3euHlFIBL/Y3mUxywyk3eFzQcqpyZGPORsmt9r668eRBJ8vcCXNbTSsbER7R6r5KMZExcsWkK4KyD9GqjFXy4MoHZXv+dq+P2512WXRgkdckEWP6jJE/nv7HdgdJSil5cOWDYrFZ5KKxF/k8zu60y9a8rVJrrW3X+/ir1lorHyZ+KOkV6UF93cpKfeFsPg+9Xz89quta/L9/v8iXX+oeU3+NH6+nDnd2h5OIvnjGxzetXbLZPBdqK+W56HnXLn1+H6mEN+3lWhvmq/HSo4f3IEkp/SOiGxRvvdV0u7mKCh3QBnvTb2/rXcb3Gy8/n/hzv4MAp3JKfH68VDd0jQUSMZEx7Z42fcbQM2T22NmtTg08UHpAvjnwjaRVBJ7NKC47Tlakr2jzuK15W2Xh/oUtPtP4/Hh5YcsLUt1QLVERUT6DJBGR3UW7ZX3W+nZtUdEzomeLqZXb8rbJS1tfkr/+fpr8857xbV7r16/XawF9sdv1tP3KZjlBwsKCUw9VV+uRo+Ztg/x8Xef85jciL77o/zT9BQv0KE6PHr6DpIMH9Xv5235tTX6+yLPP6hkIrn2Rmq/zjIoKTpDU2Kjbta76NznZM6mG629h/IpVVurBA19OOklv43LeeXqK5GWXBT7N+cABPVMhI4A9no+bQCkrS+SFF/SUjbaEh+s/xMkn6z90Xp6+PylJ5N//9r64tz2GD9fzZP2dovbii3qxrjcHDoi88krLACEY+vfXvbT+Dgnn5rZ+Umdk6AXLrZk9W6fwDXQUa8QIvcjauFarsVEPvS9c2HRfXFxT1qoePfRc5gEDdAOxWzcd/PizId2ll+p5zy5zxs8Rc6NZ3tr+ltQ0BrhbbzM2h03e2vGW++L39f6v5aNdH0mlpVIOlnpfsZ1YkCjzkuaJU/lO5XPSwJPknun3yIkDvOcj7d29t8wYMcNrY2ZN5hrZWbCz1XI7lVM+SPxAlqcv93nMgdID8mPKj34FfrNGzZJxfceJudHs9XGbwyaZlZmdsgO9zWmTqIgo6d+zf6sNh6LaIlmVsUoOlR3y+nittVbeTXjX59/NXw6nQyx2i9idOqOL3a7X0Xm7uBQXe28YezN+vJ5u0dompCeeqNPG9uihe0L9cdJJ+mIWzBHz/HzdcDAuLh4yRO8pN326HjH+y1903eWSkKCnbpT8LyeGa1rgsTinIiVF5KmndONk4UJ9+7zzvP8dCgpEEhM9G5Ud4XA6ZFPOJnlp60uSVu7Z4E8qTgroPC2rL5MV6SskuSS53eVxTS8Oteaj+q7tAHYV7pLD5sPuYyYPnCzXT75eJvYLfGihrL5MimvbbpycN+o8+cOpf2jReTQ8ericMugUv9J6XznpSrl3xr0trhEFNQVtdhDOGT+nRUKjETEj5NTBp8opY4bLaVPaHmJw7RPkS3m5DpSad+qUlOhzobZZX9a6db4Tv/iSkaHff9r/9kTPy9PnWWJi0zHr1+sEO2259FJdP1osumPYW/vNavWena+2tqlt6q9evXT9uXKl7gS7//7OyfDsmrbtyjRnrGunTNHTDI3tu++/151x/ujeXSfTCTTXwOTJenS+ecKKthw3yRwGDNBfbH83N3TNe09O1ifSjTfqQOG00/S8yGCIjNQBgb9SU/Xv4c2IETpPfPNRooIC/Zzzz/d+gXQ4RLZv11/a1ha79eypEy0019iov6jG162p0Y2Yc88VudjHLKX4eF1peRsudfWCRke3vhDclx499ElgdMIJetpd80WKZWVNAV3v3iK/M6yTP/fcwN/f5dTBp0pURFSH0kxHhEfIJeMvkeHResrZuL7jxGKzyJa8LZJYkCh/P/fvLdY6NToa/WoUtHdtTXpFusRGxsq0odN8Bg5hpjC5bvJ10j+q5Xoal6LaIkkrTxOnckq4yXtNXWmplBXpK+TSCZfK8xc/LyIiFZYKWZe1Ti4df6l77VNURJQ8MuuRdv0+rdmcu1k25myUZy98ts3Gw4iYEXLnmXf63PvFJCbpHt69wwvNY3vEym1n3CYldSW6tz0+TJYu1edo86mm2dkin36qM3sa98PwJTlZ7732pz9533ute3e9JcHzz+sOhDlz/O/ksdn0SK+/CQpaY7XqkSFXYyk8vGkqoLEeq6vTiS5+9jMdPJ17btNUvDPPDM5GjL5s2aI/R3833g6m5GQd+FxwgW5kzZrVMsOry8kn60aDtyC5vLyp08hfe4v3yvL05TKp/ySPjZytDqssPrRYzhh6hlwxqemCsvjQYqm11raYiiUiMqjXILlj2h0dWgu4JnONJBxOkL/O+qvHCLfFZpGSuhIZ3We05FXnSYO9QSb27/y5Tw6nQz5M/FCmDJoiB0sPyuSBk+WaGL0ZX7ewbnLyoJPb9brXTb7Or+N6de8l4/u1bCWOjB0pI2P92ygnIjxCIsIjxO60S4O9wX2dm5c0T4b0HiJ/mBrYRmX+vnd+vq6H5szRnd9ffily/fUtG8uDB+uOkubTbquqmtpDrlHytgIub04/Xddjrk7UYcN0R2nzDevT0/XIzFlniRQW6vJ4m67sqj/NZt1e8zYraOpU3Ym7dauux5YsEVm1Stf3TqfIk0/6P1LWt69eImAy6fJ0NEFVc42Nuq7p21d/Rn37NtV9U6e2Pj1OKb01y6xZnb+ZcLdu+m9lTKneGtYotaGxUV90jtT0ESOLRad+PP98fUFr3ihoy4YN+mL94IPeL4JlZXqEau7cwFInNjSIvPqqDswuuKDl4wcO6IwkvgKdxkZdQfXqpV+rtrYpAPz6a91TfLufCXCSk3U67z/+sX1ZDM1m/dm0NdSfnKzXPNx5Z/sCOH/sLNgp1Y3VMm3oNI/Frkb1tnqpaqiSYdFNu6fF58dL/6j+fl3oC2sKxe60+31RFNFBS421RtZlrpP0ynQZGTNSzh55tpw62M9W+P8opcRit3hdG2WUb86XeUnz5Hen/s49DTCzMlPmJ8+XP57+R4/fP1Bf7/taIsIj5NqTrvV5TG51ruwv2S+XjL8k4PUCB0sPyqbcTXLj1Bt9ZvmzOqwBT01Mr0iXd3a8I0qU3HLaLbJm3lQZMEBnShs6tOm7abPpEZTTTvN/VLauTvcETp/eel2Xna0v1FVVIo/4GZ9u3aqfc8897V8o7c1bb+nf+3ofywFLS0U+/FDv7dTWFI2qKpFt23Tnlbe6pHlDoC0Oh56uPGKEzto5e7ZuFHl7bn29DmovvDB4m046HLqOTU3VHX233BJ4wGY2izz8sO5gDGTjyAZ7gxwoPSBDew+VwppCOWPYGe7HSupKJCYyxuOc2JK7Rept9TJn/JzACuinfHO+ZFZmynmjzvPo3Pnh0A/ywa4P5OU5L0tCQYJUWirlvpn3+fWaWZVZMjR6aLszeB4qOyQDeg6QbmHdpGdEz3ZNUVZKyZLUJTKx/0QZ02eMmMTkM0lQbnWubM/fLledeFVQpkM7lVPe3/m+1Fprxe60y9/P/buEmcIktzpXorpFeQS2S1OXitVhlWtO8r0ze72tXr7Z/42cP+b8VpMIvf66Pjdvv11PSVuzRm8l0tbeYEeaUjoYycsT+e9/Ra6+WtfFgTpwQAdi2dm6E+vOO/VI1dKl+j1mzmzK+nkk7Nunry2nn97yse+/13+Txx4L/HUrKkTefFMHwL46dIItkNjguBlRaq/IyI71OiYk6NGT9m7e162bnqIXGxv4Yujzz9cBkK/pNAMG6F6XQBv+kZE6SPI1p9ZbynGbTQc0Z5yhe1q+/17f16ePnob3f/+nK5aTTvI+V95q1emAp0/3LG+PHnqErz1ZT5zOpvTrbf19+vbVmb86s7djVfoq2Ve6T2xTbXLZBN950HtG9PQINJRS8lPKTzJt2DS/AqXl6culwd4gd0/XOd3NjWZZmrpULhl/idesaiIin+z5RLqFdZN6W72cOexMSSxIlP/s/I+8OOfFVrNXpZSliBLlnua3s2CnLE9fLvfNuK/F1I/D5sMyLHqYuzEzImaE/O3cv3kcM67vOPnHz/7h9b2SipPkUOkhOXvk2W0GgYN6DfLY/8mpnFLTWONRplGxo1rsHeWv+cnzxWK3iEm8j7rlm/Plv7v+K78/9fdee3dF9Lqnelu9x4aUQ3oPkUsnXCp9evSRncunyKBBIlddpaeUnX12094ZDoceSaqr0/PQzzmn9eDHatXTMWbObLtDaMwY3SsZyJx5V3pZXyPiLo2Nei3U1Kn+TQk566zWRzsGDvT/wl1UpOvr00/3vs3CqlU6kPQnI1R4uJ7WUlenP9fly3WgdMMNLY8NC9OdRoHWLampenrleefpsnfv3tTLHR6ufyZO1L3d7UkZXl6uG6CBXv96dOshZww9Qx5a+ZBkVGbIvGvnibnRLJHdIr2OuM4aFfyWkVM5xe60S/fw7jIiZoTX9ZbTh0+XfaX7JDoyWq476Tr3dFYR3YmxLW+bTBs2zT1aopSShIIEGdJ7iHy29zOZPWa2nD/m/HaVz9e050AoUZJVlSV9evSRuOw4iYqIkhun3uj12OqGasmtzhWbw9auQOnDxA/lhAEnyM9G603UTGKSIb2HSGxkrORV50lScZKcNuQ0r/VlVESUx0h6UW2R9O3R1yOocyqn1Fprfe7n5/KLX+iOj82b9ehK81GcrsRVPwwbptsW7dkuwGrVywVmzNBT9CZO1PXElVfqn0OHPOu+7Gw9knXRRZ2XHGzPHr22MylJZ+Fs7uyz25+Mol8/HQS2dX3wpaBAP7fT2mfqGFddXa1ERFVXV/t1fGWlUnv3Bu/9165V6osvgvd6wbZ5s1LPP69UY6PvY+z2jr9PaalSTz6pVFKSvp2QoFR8vFJlZUodPNj28wsLlXriCaUOHOh4WZrbs0epoiL/js3JUcpsDu7751blqvL6cqWUUnXWOpVfna8abA1+PTexIFGty1yncqty1T/W/kPtKdzj81iLzaIKzAVKKaWqLFWqor7C/VhxbbF6ZesrKq86z+fz08vTVWFNoft2ZkWmenfHu+pgyUG1JmONcjqdXp/38a6P1ce7PnbfLq8vVxuyNii7w64qLZXqh0M/qHprvVq0f5F6cMWDan/JfqWUUnuL9qri2mK/PgeXlekr1S2Lb1FvbX8roOc5nA711va31F9X/VWZG/QfuMHWoD7b85nKrMhUSimVVZmlDpQcUCllKerFzS+q8rpylXA4QVlsFq+vuTpjtft32ZK7pcVnW2+tVyvSVqiaxhr3fUU1RR63l6ctV8/EPaMcTofX99iyRanERP3//Hyl6uv1/7/7TqnnntPn9T33KHXHHW1/x81mpV56San9+1s/rjVZWUq98Ub7zxGnU6m331bqj3/Uv8+RUFKiVG1t022rtfVj2/v5HD6sVEVF68csX67U9u36c/RxOqnUVKVWrtT/X7pUf95K6X874zpjt+t6Kbk4OeDnJhclq625W5VSSr2+7XX15d4v212Oopoin3WMNwv3LVT3Lb9P2Ry2dr1fgblA3fnTnerN+Dfd95kbzOrJDU+qrblbVXp5uqq31rfrtTvD5pzN6o34N1RtY23bBzfTaG/lwt/MD4d+UHuLvDeMPtn9ifp639d+vU6DrUE9sf4JtTpjtc9jDh1Sav587+0Op7Op3dCavXv1td2loiI47ZgjrbhYqQYfzYHPPtP1pUtcnFIvvKCUrZWvfEKCUh98oJTD+yWlTXa7/vyXLWvf8wNVWqrUiy8qtW2bLrfZrNRPP+l60qWmRqnHH9e/fyACiQ2Om2QOvjides6naxf0vXv1EGdrm4SVlfm3UE9ET6f4fcup1x3idLaeGcRfdruejjFjhu/RqsREnQSjoxmRBgzQ03RO+d8+fWeeqd+3f3//ppsMGaKff9JJHSuH0dSpvrP/1dV57vT9xRc6pWVbsiqzPNYJVVoqZX9Jy5SBNodN7l56t7yX8J6I6JGi4THD/d5jqbCmUHYX7ZYRMSPkptNu8jm3vaCmQBYfWiwf7vpQ7E67xPaI9chYN6jXIHnw7AdbzXI3vt94j/UGY/uOlbum3yUVDRWSWJjo0Rvb3O9O/Z3HJqn9ovrJ+WPOl/CwcKm0VMq+kn1S3VAtKeUpMmXgFFFKSaO9UZakLpHdhbtb/f2/PfCtHCg94L59yfhL5J7p98jsMU0L/4pri30mtqi0VMpHuz6SvOo8KakrkckDJ7t7j10bYzqUTgkUnx8vG7I3SJ8efWRCvwlS1VglS1KXSMLhBK+vffG4i2XywMniVE7ZnLu5xY72URFRcumESz3WsH2R9IWsylglIiJVDVXSPay7XHnClXKw9KBUNVS1eI9zztEjtCJ6xMA1clxfr3tar7pKT0n7v/9rO8NldLSeZtWRDagTEkR+/LHtZDc5OXo6WPNsSyJ6qtfGjU2j6MG0aJGu140++0yP+Li0Nmo/cGDT52O36xHy5gvDWzNsWNtT9ioq9NSWTz9tuh4ZFRXpxAxK6ZGiP/9Z3//rXzeNihuz/bkcOtS0uFpEv4e38h84oKfQiOhRqaTiJPn2wLcBZ547efDJ7n14fnvKb+XySb6H7e1Ou6zPWu81YUtRbZG8t/M92V/qf9rVflH9pN5WLyV1Je77tuVtk6TiphSzlZZKn7/T0Oih8qspv/I4P6Mjo+W+GffJjBEzZHy/8R7rFr9M+lI2ZG9w3260N8oHiR+0OO/bq9ZaK4sPLfZZ3hExI8ThdIjV4X9a3vj8eHlpy0vSYG8aGs6rzvOaEOIXJ/zC5zTrm6beJL+a4t8OrZHdIuW3p/xWZgz3PdffatXXXm+LQkympnaDN2VlerQpOVn/iOi23Ftv6ft9vd877+iRdyObTZ+T3s4nb2w2/T4WS9N9VVUt6zp/DRrkex3RpZfqOt7lZz/TdXhrMwJ69dKjN+0dcQoP15//3Lnte36goqL0SHzv3k0zDOrqPD/f3r11tkFXco1OEVgMdvRpK2psbFTqtdeU2r1b37bb9ahSazZtUurZZ1vvfQyWoqKW0f+OHXp0pfmvVFMTeHnefFP3SrampESpdet0GXbu1CM7zZWW6gjfn/eur/evJyM9XfeW+9LQoHtSEhN178n+/b5ft7FRqZSU1t9vxw6l0tJa3r9qlR4Fc/XQFBYqVVfX9LjD0fJ9G+2N6skNT6r1Wevd963JWON1ZMDpdKpF+xepQ6WHWi+gD7sLd6vH1z+uSutKVUltiXor/i2PUR+Xj3d9rJ7e8LRalb6qXe/TlkB6en2x2q2qrK5MPb7+cZVYkKhqGmtUnbXO52s7nU71ye5PVGJBovs+m8Omnt/0vFq0f5FSSo9ePb7+cbWrYJfX16i0VKr/7vqvKqktabPn2Wq3thjpSyxIVI+vf1xlVWa1+VzX75FcnKx2Ht7p9bj86nxVZalSSim1KWeTejruaVXXWKeejntarc1c2+p7eFNf33rvoi/Jybr3zvVcp9O/89vh0PVQW/btU+qdd1qOYqel6R7R8vLAy9yWLVs8eyFdMjP1/W2N9hiVlem6oa0R7txcpSwWpbKzlfrww6ZRP19sNt2j3pFT6q23lPrxx5b3v/uuUgsX6v87nUo9/bQexTL64gulvmw2+GN32AMe3Q1URX2FemrDU+pAScsP1OF0qD2Fe/weaXepbvC85n++53P146Ef3efif3b+R32257P2F7qZFWkr1O7C3e7bVrtVLUhe0Gbd4K/i2mL14uYXVX518IZaS2pL1KacTR517Ls73lXzkuZ1+LXzq/NVWZ0e9nE6nWpH/g6P0fJ2v26+Uv/9r+d1uLnt25V65hl93i1erNssSumZLFVVLY8vK1PqX/9S6oYbvJ/LBw/q0YrmI9xZWS1nHRUV6fcsKNBts/R0fX9jo1JPPdU02rFxY+ttGxenU7cz/akLS0uVev11/d4uDocefW7vyFEwpaToNmJrgj3a50/9GciI0nETKFVW+jf1zh9Op774taaqSk+RaN7AqKzUF0FfHA49NOp67fJyfZIaT8r6et3YaO7ll5VassTf30BLTNQX8NakpDRNjXvhBT2VsLnUVP3exgZSRkbTtCCl9O/23HNKbdjg+71WrtSfz5o1+qLui82mK8OHHlJq/Xql/vlPXZG5grnmQ9XvvKPUL3/ZevD7n/+0DBjNZv3Zb9igAzFvf+8vv1RqnpdrSoG5wOOibnPYVGJBYsBTNYpqitTS1KXK7mhZiyxLXaZWpq10X4grLZXqth9v85jmppSeLrI2Y62at3eeei/hPaWU8ph215maX4CrG6rVdwe+azE1xOF0eHwu+dX5yu6wK6vdqp7d+KzalrfN7/fblLNJ/WPtPzzeY3/J/oAaWPXWeveUubbYHDa1t2iv17+PL4sPLm4xBcnhdKituVs9GhIOp8N9u8pS1a4pRK+/7nmBqqz074KUkaHUokVNF9nly3WjurVgwlunQaDy83VDvrVpwP6yWtuuow8d0o2Yf/6zaTqbvxoadEPGblfq+++9T/uw2XQwtWGD/t2+/LLtMjVnt+vpvoHas8f78xoadDlcdXVBgfcGZzD+lu1RUV/R6XXT+qz16o14PV+xqKZIldaVdur7FdYUqgXJCwIO8vyxv2S/SipKCuprmhvMqs7qIwppJrcqV2VX+m48vLPjHfVV8lfu13xyw5MegaQ/tmzRnZjNHT6s1Kef+g6UlNLn/sqVSt10kz7/Wqvz0tP1MTu9910pp7Nl5/APPyj173/rKc6uzqQFC5R6T19eW5Rt376mTu2fflLqscdabwcqpdt4zzyj1C7vfXxKKd3G3LxZv/Y333i2cTIydPulrfZdW2w23WFsrPt37Wq7c2nHDl1H7tihO4lae4/nn9dT+jIz9efYVodSW2V+4YWW3x0jpt558fLLeiNCX1w7JIvoRcqt5QI0mdrOsFZerhe9NZ+ylpioc8v7Ulmps5nk5Ojb/frpIUXj1LSoqJaLGK+6Si/EDsQZZ7SdCWnXLv0jIvLAAy2z3E2cKPLQQ55ZZ2w2PT0kMVEPa+/Zo+//+c915j5fCgv13+Gii0Tuusv3cd26idx7r55usmGD/r2HDdOf+ZIlenqJzSby+ec669Tvf9/6Jmq33ea5D5KIzhR20UU6pfCCBSIffNBy35bp070vdB4aPdRj+tzqjNXyQeIHHtM+6m31sixtWaupvCsbKiWtPE1sTs8FrgdKD8g3+78Ru7K7MwT16dFH/nrOX+XCsRd6HLs8fbk8s+kZOWPoGXLr6bdKaV2pvLH9DUkpS5Faa60kHE7w2MuoqLZI9pXsc79PeX15i3I5lVMyKzNb3QNJKSWvbHtF4vPjRUSn4s2tzhWL3eJxXFx2nLyx/Q337eExwyU8LFy+O/id9OjWI6D9RM4YeobcdNpNHsklJg+c7HMq4/cHv5eN2RvlnR3vuPeG+in1J/lk9yd+pVjvFtZNTh18akAZ8a468SqPqYgiemPONZlrJKcqx31fmCnMPe0ntkdsu9KKX3qprjveeENni3r6aZ19ri3jxukscWFheopZerqe8vXZZ3rPJm+++krk/ff11Jf2Gj5cp/Zvz4Lc2lq9z9wLL+g6x5/NoocN0wlvzjnH89yuqdGv09r1Ii9PfxauHeKb7y5fUqKn6nTrplOtz5ypf7ff/a7160ZJiZ66aP/fLNYDB/TvEOj+Rq50wkaRkTqr6IYN+vbQod4zIoaFtS85ji8VlgpZkLygzc2YV2Wskq/3t7KLaBCM6ztOpg+bLiIig3sPlgE927l6XHQGzrb2yTtsPiyZlZniUA6x2CytHmvkq/5t/vj+0v2SXZUtxbXF8tq21yS3OlcqLBXy5vY3/dpTySg6MtqvjKQbczZ6TDU0+t0pv5MrT7jS/ZoPnf2QnDbktIDKUlamr+nNDRumkwi0lskzIkJPl7vmGj1dv7WkMOPHi/zrX76nbZlMLRO7XHGFrjM+/VRveSCiz3HXtFxj2aZMadpS5vLL9fKBtqbiRUXpsjfPLpeZ2VS/2my6nbN+vT5Xr7/es40zdqyue7zVA81t2OA5HdeooUG345ov9bDbdXIa1/RcX6qqdP04fbouizeuDMg/+5kuc1SUnlb57bdtv74v4eH67xEXp7OY+lLXcj9s39oftx0dXFHjsmXVHot2m6upUeqBB/SCMYdDT6vbqtegeh2ubW75ch3NuyQkKHXjjXqkxchub+rN89VjV1enFxevXOn/tBmLpWlB8vbt3kdP6up0r0mgUzqcTs+yfvCBUu+/33JkySUtTffQVFbq56al6aHo0k7quMvO1p/pv/+tp0+6Pl+rVS8Izc0Nznu88or318rL8907VFxbrKobqtWPh35Uy1KXKbvDrhIOJ6isyixVXFusXt7ycrumtNQ21qr4vHifi/zrrHXqg50fqMPmw8pqt6rNOZtVcpFejO10OlVycbKy2q0quThZPR33tMcIzKr0Ve4e1xc3v6jWZa5r8frp5enq8fWPu5ND+LI1d6vXqYCN9kb3aFN5fbnHCI7VblVrMtaozTmb1ZbcLWpd5roWv2dpXanfi5BdduTvUMtSm1ag5lblqlt/uFVtzN6o1mauVblV+o/71va31Df7v/H1Mp1mWeqyNqfptGeKo8Wi1Fdf6VHTuDj/psYppXtRP/5Y14NffqlHKdav1wuIm7+2K7lBaqpS992ne3F91bOdZd06PSL86qu6Z9fh0FNhMjL8e/6ePXoU28XVi9pW3V9Xp+uG5se5FhY3X0jur4wMPaXY1Zva3hGltWt9L7YuKWm9N74t2/O3q28PfBvQc0pqS9RHiR95TIVLOJyglqZ6DuPnVOWoktqS9hfO4EDJAfVW/FsqvzpfWWwW9WHih+7zPBie3fis1/qxuW8PfKv+s/M/6lDpIfXkhidVpaXS/VilpVJ9mPihe4qa0UtbXnK/vq9RHpvDpp7a8JTakLVBLUlZoiotlarOWqf+te5f6rsD37XvF/NDo71RNdgaVE5Vjvps92fqcPXhTnsvo127lProI8/zrvlI7YIFuk3WfFTCz1xebcrP169dU6PrCNd7bNqkZ9Z0RGFh08i0sd2llE7Y4po19Moruj5ubNQjJ/X1+jmHDumRntYSZJWX6zq9pkbP3Nm+XbfP4uObjqmr01N4XfVPebnn59nQ0P7pwXFxTSNlH3+s622jhQt9t6tam3FQVtZUv23YoEeoXO/ZfDrywYNK3XgjI0otzJrl2evXXO/euuegulpH57/4he4dKC7WaXdzcrw/T0RH7M03XRw4UPdSeOs5DA/X71VWJvLss94X7PbsqTdk/fe//Y+oly7Vox42m95p2tvuzjk5uhcg0KQMro3JXGbO1D2RxtEVl2HDdI/w0qV6B+uxY/WIU3vTPoro9/KVXGP0aP13Pe88/bdwjWxFROgUxiP93yrIp9GjdY+HsTd93z69I/eWLS2fY3fa5Z2Ed2TR/kXSPby7XDbhMgkPC5ek4iTJrsqWQb0GyUPnPORzg9L9Jfvlk92fuF+reW9sr+69ZMaIGRJmCpN6W717BMgl3BQufaP6Svfw7hIRHiE9uvWQ7w59J1aHVUwmk5w86GSJCI+Qkwed7N6o1TWacfG4i+WuM/Vw3l9m/MVr+ttxfcfJHdPukKHRQ72WfVfhLtmcu1nOHnm2RwIIEd0L+tjaxyQuJ05E9KLryQObMgjU2eoksTBRRvcZLf2j+ktiYaJYHVaP0asPEz+UbXm+u4oqLZUtEjiEmcIkzNT0Re7fs7/8/tTfy8wRM+XCsRe604nfeeadHnsrrUhfIXnV/m1/XtNYI3uL9sp3B78LuPe4sLZQKi2+hw6Si5Pl/9b+n5gbzJJVmSU/HPKSmcCLHj30yOvtt+vvsLf9RlwjGM1lZurRp8OH9UjI1Kl6NPmKK/SouNWq01MvWqTrlIkT9ULi66/X5+qaNfp19+3T9VEwks80V1qqR31KSvTITZ8+esT7N7/R9dXgwb43w3U4PJMYjB8vkpbWVB9366b383Cl3rXZ9KhadbX+PN5/X/dG9uypP4vt2/VxSulFxjfdpN/bn/p71SqRT/RpLuPG6RGuTZv07fDwtnuEvYmJ8T2CPnCgLvemTfp9m/ds19frRe15rXzde0b09EhX7y3JSIv37DVQbj3jVo/nhZnCPDaazqzMlI93f+xzE+v2yKrMko/3fCyLDy3WdaIhJXVH/fmsP8t5o8/zuC+9Il3qrE1d1VdOulJunHqjjO4zWq6cdKXERjblc+4W1k1iImN8jhbfe9a9cv6Y8yWzMlNe2vKSe3RJKeV+j25h3eTu6XfLrFGz5PJJl0ufHn2kZ0RPmTN+jtctEsrry2Vl+kqfCW5c2qq/uod3l8hukRJuCpcKS4W8n/h+q6NfwRQVpesVVxKW/fv1nmWuts3EiTpplCu5TUWFbselpwf+XgkJnhuT2mz63I+M1HWE6z3OPVfXP4EwTshYsECX0+nUiWe++EKP5LvcdpueJSCit4GYMkWP+KxYoeul0lL9GmvW6O1XfE34UErXzUrpmTtnnaXbhxs36vfOy9Mb+Q4frq8XtbV6ds66dU2vERnpX0IIb6M2lZVNs40OHtR/P6Nf/rIppfrhw02JNjIydNvZ1yj7/Pl6lE1E16Vjx+r/x8R4JtMxm33HA161GUod5YzzEF3Rs1FlZcsFyw6HXuDX0YVmdruOcl2RcE2N7vXzlfbR6dQ9la1F7Hv36kjc4dA9CK70v63NLQ/G3P9Dh3Rv5fz5vsuXmalH1h5/vClJQmvlqq3VPamHfXRKvfWW5xqirVt995gUFuq/4zff6Pm7HVVUpMuVldVyfdXevXp9Q1qa5/fnx0M/qte3va7+vvrvakPmBvXCphdUtaXaa3rV2sbaFouOldIjHivSViillFqXuU69svUVVVFfob478J2qslSpopoi9erWV9XGnI3qmbhnWh1hWZuxVr229TWvj72z4x311va31AubX3DfZ7X7XrlfaalUB0sPqvcS3vNIklBnrXOnyp6XNE8tSfG+YG5F2gr19va3VXl9ubI77D7TazfXaG9Uz2963j3yVGAu8Pq8msYa9fym59WDKx70mcDB6LD5sHp247Ne10Y4nU718a6PvS4w9yYuO079a92/1Ac7P3CnGS+vL1frs9a7R4MabA1qbeZav37vQ6WH3GuT9hbtVXf9dJeKy45TKWUpan7SfL/KpJRe/Nx8xCQurmnefVKS78Q027frNTxmc9PodkFBU1ICp9P7CFVKik41Xlure37feEOvcfJ33Ut1ddtreRoa9EiS8bjduz1T5hrV1OgF1Y8+2lR/NTToz+D55/XtkhLPGQHbtyt14YVK3Xyz/hx/+EHXpfn5ekTZ9dm4Fn5XV+vP9amndPnKy3WdlZ2tn5ee3nQNysnR9Yhr4fWuXU09n3a79zrbn1HBvLzWE/U884xSDz/sOWvBatU91v4mtiipLVGPr3+81bUqzdVb69ULm19Q6eXpLR6zOWxqf8n+dieGyavOU69te02V1ZV51IVbcraoV7a+4l5H6HA61MHSgy1GqTMrMtUrW19psY40uShZfbP/G/f53NyC5AXuOtrpdCq7w66eiXtGbcvbpg6WHvRaryul1PcHv1fxefFeH/OmqKZIfb7nc3eZvzvwnbrxuxvV85ue93pNOVh6UD214Smvo1AZFRnq5S0vq28PfOvzmpFUlKSejnva73VVdoddZVT4OXwbJNXVun7ZvVsnbdizR99nnMljt+vHN21qPSFNdXXL731jo76+r+qEHEhJSXoWTPPzr7y8aQQnN1fXb48/ruuj1jT/vcrLdT3ibY1PVZX+Pauq9CiUxaITmblGXVynXnGx5/qrF17QddLu3YFtQVBaqteRJxmW0ZWX67rx7bf1v20lt9i8Wc+IUErXvY895vtaUlrq/4h5IGuUjrsNZ+Pj9YiHw6GjcdcaHW89cGFhgaejzskR+eYbvQO9q7dh9Wp936hRIvfdp2+bzXoOZ//+OvJtzmRqfe2QUrrHd8oUXcbmkfKyZXpObUGB7gFuvuGwce5/aan+CSQlcE6O7iFOTdVreAZ5GRAZO1b/DB6sey6zs3WvwX0+Nj3v3l3/DsuWidx8s+41+O473dvRp0/T/FWXkhLvr6OUHuG59FI9kuTafDEQSons3q0/kx49dPmtVr1R5ODB+m/m+q6ceqpOuWu36/c96yxd1lMGnyInDDhBupm6ycqMlVJrq5W9xXslLjtOhvQe4jGKtCJ9hdRaa+Wm05p2b0srTxMlSi6doLuPpg+fLhP6TRBzo1l+SPlB6mx1csqgUySrMkvuPPNOmT5sunsTwQpLhfSL8vzFTx58slQ2VEpqeapM6j/J47FzR50rA6IGSEwP/UUpqCmQj3Z9JPdMv8fr5rMHSw/KtvxtMnPETI/RokUHFknPiJ5y6fhLJa08TX536u+ktK5U1mWtk2tPulYiwiMktzpXpg6ZKpdOuFSUUrIyfaVkVWXJXdP1CFZKWYrkVufKnPFzPN6ze3h3uWjsRTIyRveQ+hrJiuoWJWePOFsG9hooE/pN8HqMi8PpkM25m2XKoCly/ujzJTqy5a7LJpNJbjnd97bnVQ1VklycLOeOOldMJpOcO+pcOXPYme45/u8lvCdDeg+RwzWH5dxR50o3Uzeps9XJ3qK9curgU6VHt6ZhZ6WUNDoa3feZG83y1b6v5Pen/l5iImPku4PfyYVjL5S9xXvlZ6N/1uLv2JoZMzzrt0OH9PnmcOhRlMsv954W+/TTda/iAw/oUZdHH9Wjyc3nm3sbodq0SdcPmZkit96qz5+KCt/rXg4f1vXVdL10RL76So9MX3GF7gVeu1a/Z/O1BpGRIrNne77O1q165NqVMt3p1D3ArvS6drteq3Xmmbqnc/ly/RpRUXqzQ1e9kpSk39e1eWJVlf4sbrxR3xcVpevvpUv152Ox6Nc75RSRP/5R17mnnNI0syAjQ8+VN5v17xkZqevRl18WmTRJ13/vvy9yxx36fU46Sa/X/PJLPbrVvKe6uFjkvff0+osDB/QInutv53DoMp1zji5TWZmuz4w9v999pz/rOXM8H4uI0PVeYWHbqcxFRAb0HCC/O+V3rW7u7Brxnj5sukR2i5ShvYe61/9VWCokNjJWwsPCpVtYN5k8cLLE58eL3WmXc0edKyJ6ZCOjMsNj64Oqhiqpaaxx13mDew+W2MhYOWXQKfJT6k8S3T1arj3pWtlTtEfG9BkjTnG61xGmlafJV/u+krvOvEsG927Kmd89vLucMuiUFiNOy9KWSVFdkcwaOatFHTFl4BQJM4XJrsJdsi5rnQyPHi4T+k2Qs4afJS9vfVkuGHOBnDX8LK+fm3Gz7daU1pdKuaVptGZ49HCZMWKGTB823esGr2P6jJGrT7za61qjcX3HyUkDT5Lk4uQWm8/W2+olLjtOzh5xtlx70rXuz+KzPZ/J0Oihcsn4S8TutMvS1KVy7qhz3deH8LBwGdfX+/BteX25WOwW+Xzv53LHtDt8bmi+JXeL2J12SShIkEvGX+IzFbmLq13j2qx56lQ9s2PbNt2eOvlkfZ5ZrXqd0xln+E79b7HoWS/R0SKvvtp0f/fueiZSWpo+V5TSm0hPmqTbd944nSJvv61HfE48Ud+urfVsh4notuBFF3mm8u7XT/8sXarrjgsu0P8OHNjqR+Hxe7naPK62p4ge4d+3T7fBIiP1Z7Nhg8hpp+nPzfUcV10waJBu6+Tn683Ff/1rXa+Wl+t6rrBQXwfasm+fLtuk/12qCgr0NeHcc/W1JDVV14/jxun1RAMG6Lry1FP1+7rMmqV/rFZdN558sq4Hq6pats99zVyy2XS7bto0/zYyNzpupt6J6A/63nv1FyQx0feeOFVV+oJqXEh46JDOkV9V5Xm/3a7/EHFxuuFw9tlNF2dXYyQ2Vk9b69FDf0FGjtQn4NKlTa/jb659u11/6bztgtyrl26UvPxy23v+pKY2TfPwxlt5LrlEN35mzWqZr//gQc+h4r599Zdy0iRd0fgSEaEbLxaL/sy/+EL/br1768/6u+/07+Ry1VX6MzYymXRwdfrpOjhrXln4Y/lyPX3oww/1tB4RkauvFvnV/7aIiIvTjbjmfvUrXSH+4Q86UBLRF6pJ/SdJz+49JTYyVm49/VY5b/R5cv7Y893T6VwunXCpXH3i1R73pVWkSWp5qvt27+69ZWTsSBndZ7S8/fO35ReTfqEbA5OvldK6UgkzhYnD6ZDU8lR5c/ubHnuRKKVkUK9B0j28uxTV6j9OQU2BPB33tOwp2iOnDj5VhsUMcycOGNRrkFx1wlUe+yw1N6n/JLl7+t1yzshzPAKWKyZdIXPGzZHoyGh5+JyHZUK/CaLEc+x/U84m975DP6b8KAW1BfKLE37hftzutHvdB6TOWidTh0x1N1Rc0xIdTodkV2W7jwsPC5fzx5wvkwdObnP3eavDKnuK9khycbJY7JZWkyU02BtkU86mFntFVVgqZE/RHnEohyQVJ8nqjNUejZMBPQfI9OHT5c9n/dn9+v2i+skDZz/QYhF5QkGCvLD5BdmYs1GUUhITGSP3z7xfvjv4nVRYKtzTcGYODzBji+gOleb7Ev3ud3q62Pz5uh6ZOrXpsaoqPXVBKX1eZmc3TUVzsVhar6uuuEJffF0LaT/5xHPBsGvah0thoa5bXa6/vikI6tNH1x9tJRdwOPT0DKWa6obNm0X+85+mY7p10x0es2frBs/u3fr3F9HldSXNufDCpmAwOVnXKw89pC/gUVG6rn/jDV1nXXqpLn9Ojq7rcnObrhuuRd1TpujfydX59ctf6o4f13VkyBCRBx9sanzMn69fPyKiZdKe9ev1NOPBg/XjrsZNYaH+XCsr9XVu4kQd2HmbHjNxom6AeXssM1Pkp59an34nos8Jk8kkE/tP9JjS6lJQUyBf7P1C8s35sj5rvTQ6GiXMFCY9I3q69zZ6f+f7sjJ9pSxPW+5+3ra8bfLtgW+lwFwgHyZ+KPtL9suPKT+KzdGU0GZnwU6Znzxfnop7StZm6fnQ0ZHRctG4i2R83/GSVpHmTpZjdVpl5oiZklaeJlUNVbJg3wL59ZRfewRJIrrDqsHe0OJ3efCcB+WVS16R0vrSFgle6mx1sjRtqYzrO87duJ81cpaEmcLkgZkPuIOkmsYaj6nD5446V04c4Mfmgf+jlJI7pt0hYaYwsdgsklmVKb8/9fcyffh0KTAXyPqs9R7H9+jWQ04ZfIqsyVzjTlLTXHl9ucwZP8cj6Y2I/pt+d/A7eWP7G2ISk3yz/xuptFTK2SPPdgeqTuWUqoYqKa8v92vPpk/2fCLpFely4dgLWw0OI7tFSlRElFw49kIZHdtGhqlmTj1V5OKL9f/POUdPT1uzRrdDVq4UefJJ3e6y2bw//8cf9TkeFSVyi5c+sYsv1u0dEX2+nHGG78QPZrM+5swz9b9Op66T3nmn5bF2u26DetvHLCxMP99kat/UW6PSUj09b9Ysfd5PmSLy17/q+mP2bN02/e9/dXlc8vKa2lyu5Q2jRul2bYqXrcEqKkQWL/acxn3eeSL//GdTW7i8XAdCDQ36mlBf3zQ92mLRAZ2vPeBEdEfY0qX6OampetCj+ftv2OB7umFamsjf/66vCbt3t56szSv/BqmOXs2H1956S0+7UMr7YjmXxkY93Hr33Urdcou+3dCg1HXXKfXb33pO+0pL09MYSkv1FBPjtL66Oj3EWFSkhzC3b9f7lLj2JrJYdFkKCvT0j9b2/LHb9RBzfn7rqQ9ra3Va7JQU/3e4t1o9h4ETE5X6xz/aXtCslE5osXevfs/m091ci7qdTj2FsbWpN/HxTVNmMjP1wkjX4kJ/EkFUVDQthMzKUurii1sOE1dX68/NbPZcuOjy5z/roeDiYv33cdm7Vw9RWywtE2U4HJ5TELfkbnEnA3DtiePSaG90L1a2O+zqk92feN0XI7sy2+cUlKSiJPXcxufUUxueUnsK96hnNz6r9hTuUd8f/F7NT5rvsVg5syJTPbXhqRZTQHYV7FJPrH+iXXtyvLj5Rb+mjdRb69WajDUe0/gcTof79/rH2n+4p+cZp7uklad57C31UeJHHokYimqK1NbcrSqzIlM9sf6JgHajX3xwsVqZrnNB51XnqV0Fu9y3fSmuLVavbXut1X1ADpQccO+dZbVbVYG5QD0T94x7mtHS1KXq2wPfqrjsOLWveF+L59dZ69Sm7E3q3R3veqQbd+0p1WBrULWNtR7TZTIyfKe1NXI69XTYhgY9LeOzz/RCXmPihZwcnazFVRdYLHoa1+efN03vffBBfa60lmymokKpb7/VU86Skz3P4ffe0wuRXVNKbDadJCYvT9/esqXldJe0NN/T8Q4e1IubjaeM2dx6MgR/0mAnJrZMXLN3r/69rFb9/OZTV95/X0//qavTv9833+jrx8GD+v8Wi76mvPZay/dyOvXUks8+85wq2dzSpU2fk1K6DHv26IQ5e/e2Pl27qkon34iP169jteq/jes5+fk61flXX+nvii/l9eXqrp/uUhkVGWpf8b4WSRmUUqqsrkz9lPKT14QzqzNWq2Wpy1RxbbFKKU1RC5IXuB+raaxRS1KWqGc3PquWpi5VDbYG9/TTAnOBemHzC+qHQz+oNRlr1A8Hf2gxPayivkLtyNcXR9d7p5enqz/98CdVYC5QmRWZXstUaan0WY/UW+vV85ueVyllKepg6UGVUpbivt9V31rtVq/7QDmdTvXo6kfVewnvqR8P/eje482fKYbL05artZlr1V+W/cVd/9Vb69XX+7521+n/3vRvdduPt7mfsz1/u8qqyFJf7P1CfX/ge5VcnNzm+zS3Pmu9+mLvFyq7MlstSF6gyuu9b+TzXsJ76vM9n7dI+W1z2DwSVpTXl3tcA1alr/I69dIoO9vz3CwpaTvds1FZmU4lvmSJUv/3f/r8e+EFz0QrS5fqdtL27e3bc87ls890MpukJF3HPvGEft2GBu9LCurqdLvJOLXWtUzDX65kDr7qsZQU/fs7nfpcf+EF73XErl1N9bvRzp1KrV7ddLu4WE9/a542/fXX9dTl1qY3uj7f+Hil7r1X14n+pi8vLtbv52tfqcxMPT3P2+ewc6eerrx7ty7jE0/oayD7KDXT/MPIy2t7fvdXX+kvVGWl/lI99FDTXjnp6foPppT+gqam6n/b2nTQpbZW/yHz8vSF0nWSZGXpP97OnZ6NgcWLdd797dv1F6q0VAdTmzYp9cc/em/sN5eTo1+3eWPItVeT8Qu9YIF+P5d33lHqk0/8y2yyfr1udP3pT03ruZq/d0VFUwDSmubvlZur10K0tfHs3r36BNi1Sx/vavwsX96UudDlp590UJuertc9GdeerVunA9b0dF12i0U3Risrm/atcjr1Gi1Xw2/ePM9NG3OrctVXyV+pN+PfVE+sf8JnJjOn06n+m/hf9erWV1WDrUFtztms7A67qrJUqSfWP6EOm3XtWlpXqrIqs9S2vG3q/Z3vqzprnTpUekgV1xYru8OuahtrldPpdGfSa+6t+LfUf3f912vDoMHWoD7d/anPjEsuq9NXq+8Pfq+qLFUqvzpf7S/e777wpZWnqVe3vur19cvqytT7Ce+rvKq8Fo+5nlvbWKscTod6Ju4Zta94n0opS1EvbXlJrU5frZanNX2oFfUVPrM+NQ9e9hTuUZWWSrU8bbnPPavSytPcaype3vKySjjcSouwneLz4tUrW19RNodNOZwOta94n8qpylFZlVlqfdb6VhsvlZZKr+shFu5bqB5b85j6bM9nqtJSqWoaa1RCQtM+Sa56wpe6OqVefNH3OsDmPvjA87Xsdn2xraxsCqRcG3S3JjHRe92Yk6Of/+STTR1Fa9fqc3jZMn0Bbb5/iNOp5/Mne/nYtmzRF8n2ZJlzOPRn4u11W/Pdd0117+HDuq4wdih9+KE+ztu+UJs2ee/8cTiU+vprz40jfamr04FNZqau01wdbc333LPZmrJ9VVXpOvill3RjLjFRX4eeeaapYbZ3ry5zW+wOu3pk1SPqYOlBlV6eruKyvWwi9T8f7/pYHSzVPYuu4CCtPE3tLtjtc/8xi83i0YnjCoYa7Y0qPi9e/XPdP9Ujqx7xK9g4bD6snE6nejP+TXcGObvD3uYm34sPLnZv8vzSlpfUroJd7jrN1+bPuVW5HvudLU9druKy4tSO/B3q410fq/zqfJVblavK68vVM3HPeNS9W3K3qOVpy1VSUZKyOWyqrK5MJRcnq9SyVLUmY43alLOpxfvlVeepbXnbPD6H23+8Xd2//H717vZ31Rd7m3oKnU6n2pSzyaNuaWtfvwZbg891VhX1FSouO079cOgHj/u35m5Vr217zeO+Vemr1Ce7P1E/pfykFu5bqN7d8a67E7HR3ui+9rnU1upzqnknx759Ougxstt1h6i3DeOV0m2ahQt1I/vTT3VQkZKis749/HBTJ4fL4cP6PFq7Vq/D9tfGjbpOcwUDZWWBZ4XbtEmv6XzzTf0ZbN3q+RoHDrQMooqK9DntrV3rdOq2TPNOl/YsATx40LPT+ttvlbrnnqaybNumA6WSEl0OXx1ar76q27gOh35uYmJTpsC6Ot/XpuJi3TnnbRNtXxITdfvV4dBlS0/X9fxLLzXVxwRKzQTyYSilT6RVq5S66y7du1ZY2JQCt7n0dJ1SfO9e/UUqKWkZkRcX6y9OeXnrvQQOh97gsfkmsk6n/mK98or+ghg3nU1Pb30TVaV0cGVMaGA26y9naann4sXycs+LvdXa+klVU6NHgCor9RfvoYf0hdgVgDkcuqLbscNzYWBCgmcFVFXV1Atgt+seVVcjLTW1qUGxbFnLTR1zcnRD58MPmz675GTPRlxGRtNnv3ChPqldtz/9VL/usmWevUmuhtsnn+jfqXnDxen0bMwcPtyyl6PSUqlSylJUgbnAfRFLKUtRW3K3qJrGGpVcnKzejH9TVdRXqITDCaq0rlS9vf1tdzDQvGdzfeZ69WHih+qxNY+phfsWerxPSW2J2p6/XS1PXa4u++KyFgkCimuLvV4Msyqz1MtbXlbLUpd5HSXZkb/DHXR9lfyVemHTCyqvOk8t3LfQI0lDwuEEtTpDdzXVW+vdAZPVblWpZamquqFaPb7+cXeK8P0l+92b3pobzOqN+DdUpaVS5VblKqvdquqsdQH3gLo4nA716tZX1b7ifeqT3Z/4vMA3Z7FZfDa2fKVe90fzXtWyujJ16w+3qo8SP/LrufOT5qufUlpuY17dUK3yq/NVaV2p12QZrkbup596job6q7a2aYPolBRdd7z4YssFznl5nh0DbSkzxOF5ebohYLXqx5r3KmZm+t52oLa2ZfIbs1lvYu2rkdSazZt1oLZmTcfSZSvVcuF0TY1+fW8LqhMT9cwDo6oq3VllDKqqq1tusNnYqEf7n3tO33Z1+OTl6UaKq4G5dav+rJXSz1+2TP++ro0xlWqq4x0OHTR5C2xLaktUoblQWWwWVWAuaNEhU1Ff4e7YyanK8VjYvz5zvSqrK1O7Cnapt7a/5b7/lsW3qFe2vvK/MjhVRkWG+1wsqS1ROVX6lyirK1NPbniyRbKV1pLXuM7dSkulemL9E6q4tlh9s+8bde+ye90dPs/EPdOibsytynVvBp1WnqbyqvNUUU2R+ir5K/Xxro/V3qK96tPdXlrr/5Ndme1OImNz2NSdP92pnop7ymv5kouTPeqYvUV71aIDi9SHiR+qnYd3qpe3tJ1rOrk4uUUK8IyKDPXE+ifU/pL97pE1pXRw+J+d/3EHoOnl6erpuKd9bmS9p3CPemTVI+rpuKcD2lDbard6/K1qGmvUwysfVstSl6l5SfNUcU2x+nrf1+5jyuvL1dvb324xmmc8J8vLPTt3c3L0fU8/rb/XzRMeNG+7lJV5PpadrdQjjyh1zjm68W3s4LBadUdCYWHL9NTl5Z6N+V27Wk+Y0tyhQ7oz2aiioqn9s3On/ikp0Z0g777btKms06nbgs1nM+3fr9stixc3dQo3rz/+7/90+2XlypbtO29cI+ROp64jsrI8H8/N1W3jm2/Wde/Bg7qNtmCBnn2llA5Alyxp8dLq0091ANq8A87pbCrvtm26PVlbq+/btKmpE7CoSLfdXG00b7PB7HZ9v9Wqy11WptuFrhlkGzfq9qNrY3GdBIRAyc0VKKWkVLf6RWlsbPpDVFToD9J1Ybbb9RejOYdDqTvu0CMv8+bpYOaxx3TDPD5ef3E/+URnSvv0U32M2ex9j5G1a/WFzxX5O53tGwYuLPRs2MTHK3XnnZ5TXFzv7+q1Me467S+bTffmuho8//pXy89IKV2xNA80S0qaLsauHZRdQWBysp7e+Nhjnq9RXKzfa//+pkZVcrL3Btvnn+seD5c332yacqSUPlFcU/Ly8/XrffaZbpDs3Nm0/0hxsX6/1jIF1tXpYNr1vSoo8D70nF2Zrf6y7C/q410fqyc3PKnK68vd0zfasrtwt/r3pn+rb/Z/oyxWz0AotSxVLUheoGoaatzT4fKr89V/dv6n1YZ+vbXea7Ykl6/3fd2ix9XpdKpnNz7rUe5/rP2H2pita6J3dryjtuVtc/++z296XtkcNvdUw5LaEvXF3i/cgZXNYVPxefHK5rCp1LJUtTRlqVqRtsJd7oOlB9XmHP/SFiYVJak1GT7mKjVTYC7waDz4Ep8Xr/64+I8eU+QOlh5UedXeR8dcsiuzve4tVVxT7Nd0E6V0A3Br7lb1n53/8RrEZlRkqHprfYvMhE6nvjCsW6c7UZYv19/pkhL/sj9mZ+u6zNXASEvTnUW+MnP6o65O1zGHDzfVZ/X1ntO6VqzwHPnds8dz5NnVMPn+e92Z5GK16kbAo4823dc8+1zzTidvqqv1Rb35qLzV2nI0x25vGpWxWPQFNz+/qYPIm/37lbrmGt2AWb/e87G6Ou/TAWtqdAPDWO8vWqQzb7ne6/BhPcVk9WrvHWX5+U3Z9nbu1O9lNjc1LBoa9DWipqYpANu+XV/vqqt1J5Yx0F6aulS9uvVV9fKWl9Xig4s9RnuVUmpt5lp3Fsb1WevdmeAKzAXuqbE1jTUeAdS+4n3uc6XKUqWe2vCUe4Tly71fqs/3fK6U0vVOZkWm39nwCmsKPYIg1zm0KWeTx7Q4u8Outudvd0+FU0qpnYd3qpXpK5XD6VCf7flMFdcWq0X7F6mvkr9SSUWG1F1e7C7crZanLVfl9eVqR/4O9erWV1VWRVbAmfwcToc6XH1YvZ/wfos6YEPWBvX9we897nM6nerHQz+6P7/m75dYkKiyK7NbjMjnV+e3OoW6uKZYZZRntAiKU8pS1Ds73ml1unJpXalKLWtKP1fdUK3MDWb1w6EfVG1jrdqcs9nr9amktqRFnb9rl+5AefppHcjb7brhfOGF+jt76JA+Z/bs0Q3k3Fx9zXd1+iilv/+uwGvzZt3ovvVW39ns0tO970O2dq0OCvbv1+287Gz/RtZdZXBlf2to0PsHufZ/bK2u+uorpWbP1u2dhAQdUCilz+/nn9f11aFD+rxPStLtPVf7Ki5OH//ZZ7oeWbDA9/s0NOjsqK6OqoQEXeYlS5rK9+67+vXi4nRduX5907XFNYrk6tDKyGjqEN+yRf8dV670na3Y6dTlf/FF/XfJy/O9DGX5cv03rKpq6iT67391cJWaqj8X155Srr30vv3Wc1+99HSl4uLYR6mFjRtF9u7V/3c6PfOwr12rF/2tXq1v9+2rEz64FqHt3at3YXY5eFBnWHn+eZ084IYbRK68Ui943rxZL46LjdXPHztW7+2Rk6P3HPnkE52bv7mSEpERI/RiOqV0tqP33296fP9+vehQRC/Mcy12q6nR+y25FgW7FqqJ6AV3PXuKTJigF76J6IwuH32k/9+rl8hf/tJy12mjVav077tnj8hnnzXtKVBRobPCREToBdfXXKOz1hkXJw4b5plVb+BAnamkpkbkxRd1dihXVpQpU3T5HnnE8zX27NHv19Cg/1Z1dXpxnjHjXkKCXsh84olNizfvvVcvFF+y5P/b++7wqI6r/XfVe+8CSaiABJLovYMNGGNs4woY917ixCUuiVvs2LGdOLhXwMbGYGw6Fr0XgXrvvWtVV9L2cn9/vLm7WoHtJF9+5vvIvM+jZ7W7d+fOzJ1z5pwzp/D93LlM0AAwwD06ms/plVcY7L1nD4MOQ0LY76GZAgE+o/R0BiPqdJzXhgauhT17eM3mws0oaC8AAAz3HY77Jt6HTk0n1qSuQYB7AEYGjsQXOV8gpzXngvbTG9OtSQqSQ5KxOmU1zBYzVHoVqrurUdNTgyJlERICE3Br8q04UH0A/QZWiPdz80NSUBKqulk0QmvU2gUSA4C7sztMFpNdcHKxshglHSU4Xncc48PGY1QQixgcrD6IXWW7oFAosHzUcmu/NEYNCtoLrJmRFo5YiP1V+9Gp7oRKr8KDEx+Ek4MTgj2Zsie7NRu1PbWID4iH2WLGoepDSA5Jxsn6k+jUdAIKQKVXWfvare3GpsJN6Nf3X/gABqGutw6bCjbBz83P+pnOpMOh6kMXJF/o0/ehbaAN20q2WetGpVWmobTDvuDNqfpTmBo5FXEBcXb3ae1vvWgfJEmCRbLg+5Lv8erJVy+oJ+Lj5oPNRZvRPtD+s2MBGCx9sv4kQjxD4OroigNVB/Bd0XcAOOebCzez1lSmM155xfY7hYIJCIxGBrkeOwZs3UpaGZxg5acQHQ08/DD55BtvMAHNs8/aeGB7OxMj1Nb+Qv8t5HOdnQxKfvxxJjn4y19IK+7uDHaWsXixfWKW6mpb9fmBAfKDzk4mkVmyhIkODhxgUpWRI5k9qaAAyMkhPzGZyKPKyshjhsJo5J+PD5NaTJ1q+66ignVItm1jgDfAtmXev2UL6V6nA777jvccDLOZyXHi4oD332f/PDzIF+UAaA+Piwdoe3kx++DQBDnz53MOZT7k62sLYL9YptbISOCZZ5g0qKiI/HD9eo4D4PP092d7Tk7ke2fPko+7uTEAe9w4XqvT8XVpwlJMjJiIsWFjcW3itVgSv8TungtGLMCtybcCAObFzMMVsVdAb9LDy8UL3q7eMJgN8HLxssuMNiZkjDUZTIemA49Pe9yaEU1v1mN8+HgAzDw5wn/Ez9ZXymvLQ1Mfi2CFeIZY+1LVXWVNUjMrahaSgpkiy2Qx4UTdCdT31mNs2FgcqTmCE3UncLzuOGL9Y6GAAtG+0XB3cseKpBW4JfkWpISm/OT9ZYwLG4cl8Utwsv4kdpTtwPjw8fi64Gs09tlnxuhQd6Cyq/Ki/MAiWfBt4bcAyPvdnNxgkSw4XHMYaoMaScFJmBA+4YLfDU6cM3iulGolslqy8EnmJ9hRugNbCrdY56ZDw+JmOpMOOpPO1pYkYVvpNnRqO628DWASjaa+Jmvm1oPVB+2SbMio761HXlue9b2Pqw8kSDBbzFAb1SjtLEWnuhPvn38f55vOWxMXqY1qa6IPGb29pOOVKyk3yDXSPvyQMkBdHRMTnDoFZGRQ/rnpJvskKPv3MxETwIQGN94I/PGPpOVPPrkwi25nJ+ljKP9YsICynFJJGSE62kYrF0N5Ofnm+fOkv5R/LKHGRvJni4WZLi+WddhiIf0tWcLkVLm55B1y7R8XF/YlP5+838mJdLx8OWU+gBl4p00jr5s+nf09exb46CPyycH1NnfvZvsyP0xMZGKX4GDOeWsrswJOm8Z2e3uZdVPOsCnXDfX0JI/x8yMf1OuZGKu/n7zl5En7ulQyFArK5JMm2WosyXLhUEybxjXw/vuUDUtKgKVLOU8xMRyr2cw6THKyiBUr7OvqxcX9/LMbiv+a9OA332xL0VhZSYH2qaf4fvRoZqEbXJy0t5eb4Z13MiuUvPgALgR5wxmMefPs3/f1cXG5ulLoMJlIPPn5XJgrV1JwmDqV17q4MGOLgwMXSWUlMxQFBVE4lyRm4xs/nsTu5cUFIBdGTEy0KU0aDYWF8eNtC87Pj4qL0ciN8Z9JAevryw13wQIqbFu3UmlZtIjEA1AQklM2yqkXDxygUnPfffbZ+SSJxdA8PSkUBQaSwc2ZQ2IZuvl3dXEOY2KYFWrCBDKq/n6ObTAcHEi4kkRmIKcjHxiwPVv52Q3G/PlkEmPHAn/6E/s0FD09ZAbu7iTCsjIKPBERzHYXFUXClRXPSRGTrJnNHBQOGB08Gg4KB8T4xVjbXBK/BP5u/qjrrUOkdyScHZnnc1/VPvi6+uLJGU/CycEJw3yH4RbfW1CkLEJOaw5SQ1PtFIg50XOsv20daIWfmx+O1R7DyMCRWJe7DrOjZmNsmC21mSRJyGnNQbhXuDVLm9akhQIKuDu522WMmxg+EWaJKc6CPIKQ1ZIFi2SBh7MH3l/6PhTgphznH4cmVRP+fOrPmBQxCZ7OnvBztz3MJfFLsCR+CSRJQr+hH736XujNenyY8SFWpazC+PDxyGvLs6bxnRo5FcEewbhz5524c9yduCKWqY3cne1TGUZ6R8LXzddOODCajejSdqG+tx4uji7W9MWjgkZhVNAonGs6ZxWgIrwj7JQsALgl+RYEuAfAzckNdb118HLxukA4HIxDNYfQ3NeMJ6Y/gT59H/zc/NA20GYtcOnm5IY1qWt+srjwYLg5ueGZWc9Y3/9Q/AOCPINwS/It8HD2wHOzn4ODwgEdvqQLwKa4T5hAOjWbSfN799IAMzjV6unTXKvt7UyBvXKl7TuzmQaCOXO4qQ4bZvtOrydvGZztbutWXjNjBt+Xl3Mz8vUlj6utZea1vXuBG24gT/rgA+C22/jb++6zZabUaEiDK1Zw49uwgZkkn3ySfCc4mEanmBhmUvrb37gRurvTSBEczLSzTk6kx9JSCgIpQ+TbtDTyiWuuuXDux4zhvFVV2VLyjh1r418rVpB3uLiwqKyLC4UepZKZsEwmChdjx1LYKijgGLKzL0wPLOPIEabq9vEhPywro8IkIyCAY/n+e6Yw9/SkoDAUkkQ+Pdi4s2oVee3UqVQCZcUN4DWpqUzNrtHwd+vXk49PncrP161joeKwMCDQPRBHa4/C09kT04dfmHJUoVCgV9eLmu4alHeVw2QxoVPTiQZVwy9moDxWewxzoudYC9LePf7un71+KOTMn8N8uGBremqgMWpwuuH0BWUCclpzUN5ZjvPN5/H41Mcxwn8EnB2c4aBwsCuufbFC2/8slo9ajmtHXQuFQoGJ4RMvSCt+pvEMKroqMMxnGFalrLL7TlbSfNx8MDNqJk43nIa/mz+6NF0wWowX5SGyEetiWBK/BF2aLnye/TlUehUquiqwwrzCbnw/VvyI0s5SPDfrOZR1luF43XHcPf5uuDq54q0zb+HmMTcj1j8WJosJYV5hWDBiAaq7q/Fx5sfo1fVicsRkjPC31e+YGDERE8InwCJZrFkEfVx9cH3S9QCAeyfci11lu1DdUw2dSQdPF0+MDByJGL8Y6/5oNnMNytkvjxzhZwqFTQiWswfLpV6uu87esGk0MmPuFVfYZCTZEB0YyBTfNTUX0mZWFvnaAw9QIfslOU+r5X2Hpp0+fpy8ITmZPKS1lUWrW1qY1TI7m9+PHUvZYjCfzs2lXPTQQ+R1g9HTQ7quqKBi5OXFe8fGXphdWFa4ACoVMTHk421tNPw89RR55tKl5IsyP+7qYv+Skri3LFrE1+uvJw8yGjlPRUWUZZ97zpaZtK+P8pmcBfibb7hXmc22zKEXw5w5fJX3HouFc6tU2peH8fPjHFx1Fcfz1VecJ3d3GgZraykLh4byGQQF8RnISE9n20P3hp/FL545/R/HT8UodXfT7eunXM8MBrouDD01lwtavffehT6sF4NGc6H7SlcXM7C0ttIXtLqaR5hvv23zxSwrsx1dGo10T5GDfb/6Sh4bjzzT0ujmV1NjX9zr3DlbkPOBAzzuffrpiwcvd3ba/FkHF20bGKDboE7HY06djt9bLPS137jx4j7+TU28v/zdu+/yaPTYMbrbdHYyRiAtzXZ82tFhcxHR6znOkhK6FL31ln37eXl0oRvs6qbT8WjYYLD3S9brefQq+6v+FH6usPDGjXzmsmtMUxPdATZtssUo1fbUWoNua3tqf9FVS5Log//O2Xek+t56qUnVJBnNRslisVgTAfyzMJgMUq+2V1qXs06q7LIFbfRqe62FXS0Wi9X3fzAK2gouyNCnNWqlH4p/sGad6hjokGq6a6TXT75ujTk6VX/KGshtNBulD85/IO0s3fmTbiqt/a3S3Tvvltamr5U2F2yW+vX90tf5X0vnm85LlZ2VF2RQkvs2oB+Qtpdslx7+8WGpQ22Lgs9pyZHKOsqkqq4qu/gsvUkv6U166VT9Katr4GC09LVI6Y3pUlt/20XnYzD2lu/9xYQPfbo+6eXjL1uTRGgMGun5w89Lzx569hcDpnu0PdKp+lMXuDp2a7qlDzM+lDrVndLust3SztKdktqglnaU7pDyWknUcgCvHNA8NIPQuXO2ZCjV1XRLOXmS1zU325IqyP7dxcVc4ydO0DWhoUH6ScjutnLQspz1U3abMxjIn5qa6NP+6qv8rryc9xucba2mhrxY5jG9vYy/OX+e9Pz88zaeZTKRhwzmy729dOn48kv23WQif09PJ8+ReVBtLV1yNBryhKGxTWVlP11stb+f9zAY2Lfdu+mqWFFxoXudJNFtZfDzuFiAs8UiSfffb3N1Uyq551RUcG/IzuZ8tbTYCtMOTgJRUmJrV6djghrZHa+2lrz79ddtsZhqNfcZmQ8WF9OFT3aHaW3ldb29nLstW+z5bn1vvdWVtlPdKb187GXpdP1pyWKxSG39bdL6nPXS84eflx7d+6i0s3Sn9EPxDxe4bl04BxfnSQP6ASmrOUvKbc2V+nR9P+u+ltOSI7107CVJkuhS90PxD1KjqlHaXrLdGjslY0vhFulsw1k7l7t/FcoBpfRJ5ic/y5/lbHFmi1nq1/dLRe1Fkt6kl5r7mqW/nvnrBdnxTGaT1N7fLp1vOi8pB5RSv75fMpgMUnZLtl1CoIbeBjseKEmMp7pYUWydUSe9m/6udKjqkPR13teSzqiTTtWfsnOLkyTymq/zv5b2VeyTHt/3uLQ2fa1dvKmMwePt1/dLR6qPSC8effGibs9vnX7LGo8qSZK0p3yPnfux2WKWdEbdTxbebmy0X3t795LHqdV8PXHClm24q4subEMzNZpMDGcYTHsNDXRl/akwjE8/JR2YzaRxudj2xdzALBb284svLszIp1aTl6jVpMW77pKkpUspv+XkMCFOaSmvOXXKlrwmK4vvjcYL+9jUxLiq+nq+/o0hflJrK3mvxUKe8cUX5J29vbz2L3/hdbLMJcujMkmZTJwjWVbMyWGfH36YfFens7myDSZDed/YvduWFKOmhnLsY4/x2eh0NlfjoqJfTkA2GHKSiLfesslm7713YQFbGR0d5IHNzbZ4+ZwcJicavAZqaznH3d3C9e4X4e/PU6L33rO55GVl2U5knJ1pJVQoaOGUjwsPH6YWfcstPL4bCtnLqb6eJ1IbNtiOfQGexvj780QoLIxWwthY3isigpZOJydaQFpa2J7BQCunycRTlREjgLffZt76rCy643l50WKQksLPamtpGZRrpMTH01J5/fUX1l9KT+fplclEy83u3XS9AWg9WLWKllRHR74mJHBeRo9mH2R3nLw8W277yEjeX7ZgXnstrSZvv02LdmAgrQDTp7Pdmhrg+ed5NFxaylOgwED2+957aVk2GHii1tZGK0ttrc2KU1tLt5zNm2nZGFykzcWF1uYzZ+guc/687TuLBdi1i2vgww9p8amrs68HAPB5hYTw861baXGR3S7lgm0n607iWO0xdGu70dzXjAZVA7YWbUW3tht7yvfYuTcAQKemE++dfw/JIcmI8o3Ct4Xfoqi9CAqFAt3abrx5+k20D7RDkiRUd1dDpVNduOD+gUJlIbaXbsfd4++2s6L6uvnCZDHh0R8fxen609hcuBkaA13uJElCSUcJKroq0KvrRUZzBnaV7QIAOCockdmSiU5NJ843ncfa82vx4rEX8cysZ6yFZmdFzcKc6DnWez0y5RGEeYWhQTWoaM4/0K/vR5B7EEK8QpDdko0ubReO1x1HoHsgjtQcQXVPNfzd/HGm4QyKlcUA6DYY7BkMTxdPXDPqGtw/4X74utpqcSgUCpxuOI0/nfyT1T0EoLvgu+nv4mzjWUyMYNELk8WET7I+Qa+uFyaLCTqTDg2qBlR2VQIAzjedR3pj+gX9vnrk1QjzCkNuay6UaqV1fgaPy2gxYpj3MIR4hqC6uxqODo54Zf4reGrGUxecgJ1rPIff7PuN9X2Ptgf7q/Zba6HoTXoArAszJ3oOHBQOSG9Mh9akxffF38PX1RdV3VVobQVef53upu++S8vrt9/a9z08nDwFoDWzvZ3uD9HRpBONhr/XaGhd7O+nFbW7m/RYWUk6PHjQ1mZBAXD0KHnY7bfT8nnuHN3yHnjA5g7b1UVLbWQk8Npr5JebN/P3RiMtfPIhoMHAE6ft29mWry+LrE6axNOWF16wWQS1Wp58pKWRR+Xns+6ZqytdblpayId++1vy6YoKWhkBjqWjg24658+zP4NRXk7r5dAaG3KRxbw88ov58zmHajX3ANkV5digcjbe3ra6Sd3d5HuGIWVnFAq6JM5ibVUEB3MfyMriCdJTT7GPnZ3kmc3N9m4rJ09yPHo99xi5QHdBAeest5f7QVgY+33iBC2wZWX8bvRoWtvl+lE+PrxXWRn3BK2WBXSt89NZjh1lO7ieBlrRre2Gg8IBSrUSTxx4AsN8huGpGU/hnon3IKc1BymhKejV9dq5Ycmo7q7G+tz16NH14J30d7C3fK/d9wOGAeyv2o8zDWdwtPYothZvvaANGd3abowJHgOTxYS63jpE+UYhvy3fepp1qPoQ1AYugluSb8H04dMxK2qW9VRepVOhobcBRrMRepMen2R9Aq1R+5P383XzxYzhM6ynJX36PjT3NeO98+/BaDbCaDbik6xPoFQrUawsxtf5X2Nn2U6UdpTiSM0RLBixAEnBSWgbaIPOSLe3elU9NhZsxNairdheuh2bCjYhsyUTE8In2HkhlHSUoKa7Bu+kv2N13dtVtgsZLRkAWEhbRpe2C2VdZZgTMwc3jbkJRrPRWtNKZ9LBbDHjWO0xnKg/gckRk1HZXYk50XMQHxBv9RSQPRVa+lrw4rEXrfzJy8ULC2IX4JX5r2DG8Bmo7q62m6PJkZNxZazteCPKN8p6cl/WWYZnDz+Lut46u6Lb7QPtUA4o0dlJvnHDDTYX+rFjyb/S0kh3WVl0MQ0KIl0mJJAHDKYPR0cWdXWz3QLDh5M2fH1JN3q9/bOdOZO8ra2NbckyxdNPUzaQceQIaWTTJsoGg4tcNzZSzjhyxCb/zJxJ+Wb2bHrCLFxIOc7fn/Q/fjz5Tn095QlZBjx92uZCq9NxXnx8eKJ99dXkVwEBPD1pbLTVvmtpIV+PirJ5TgUGsiaU7FGjUPA3R47whCojg7T/4YeUoQIC6Ibt4sJrR43ia28vXRZralj8uqLCFg7i68uT67lzeV1tLa+pquKYZNfB1tYLXZeHIjaWf08/TV5UV8f+urjQy0CG2cz7+/mRnzk7syaWszP5X3S07bTr3Dm2df48eec/C4Uk/cull/5Poa+vD76+vjh9WoWZM+3PWC0Wbo6+vhQoXnuNCys6moty8WJet2sXH3JvL/3aARYnjYzkUWJFBZWpZcvoZhIYSIHhwAEKIk88wc3I1ZVCzQ03/HwhMdlnVq8nYet0NmKXq8tfcQXvLRcmGxhgH5KSKLR4eVEJAyhQODpyw9yyhUeWDQ3sk17PhZyayiPS2bPJnDIyKJA4OfH4cvhwbvgTJ3LRtbQwJuCKK3gUvW0bGdbMmRyf7DY42LXcYqGQIrvqALaq0iUl/O7sWW7gkydT8Hj1Vdu1JhPw4ovAo4+SiBUKzmlPD49+g4I4176+dL1bs8bela+mhorrjBnsd1cXGcyxYzwm1utJwDt28PN77rnwOL2sjM8jNpZM0MmJxLtzJ/DQIyZUdFXgRN0JPDT5IZgsJjx/+HmsSFqBfkM/5kTPuaDye0NvA545/Aw+XPohSjtL0dTXhFuSb4EkSejQdGBj/kasSlmFzOZMJAUnYWSgveNuYXshzJIZY0PHwiyZrYVNj9cdx+jg0QjxDMG7597FyfqTWLtkLRr7GtHS34IbR98InUmHL/O+xO1jb4eHswe6td34MONDrE5djVj/WOwt34tzzefw4pwXYbKYoNQoEeMXg6quKujMOiSHJMNsMUNr0mLtubV4cvqTON1wGi6OLnauHR3qDnye8zmem/UcDlQfQEpICiJ9ImGRLDBbzDhccxgLRizANwXfINY/FrH+sTCYDVAb1SjpKLnAPaWsswwao8bqrtiv70e0n61IodFsxKbCTfB08cT1iddb5+RE/QmMDRlr5xLYoGqAr6sv+g39OFV/Cr6uvlg6cqnd/XaU7oCbkxti/GLQp+/D1GG2wJa9FXvh7uQOPzc/jA8fjw25GxDqFQqzxYwJ4RMwYBhAUnASNEYN3J3c0anpxJGaI2hXtyPIIwguji6ID4jHsbpjuCHpBnya/SlGBo7E4rjFcHNyg4+rD2p7a/Fd0Xe4cfSNCPQIhFKtxEDtaPT0kNeEh5M/yFXNN22iS528SQ+GTmcrWurlZfu8u9vmuiBXN6+o4Ebt4UF/eLWafEattvl7FxTw/hUV3MhWryb/eOMN0t6aNcBnn9FFRC7+aDDQFTomhm2dO0f+O2IE6WlonM5QHDtGHq3RsADhbbexTdk9o72d/G3sWPrHd3eTnp9+mrxizx5+V1lJIcdkIh+RXXvefZd8bMIE8oRdu2iwmTLFJkzcfjvvtW8feURfHw0on356cZdmlYr9ra8nfxpaUb62lt+NH8/vLRaOydmZ7Ws05NMPP2xzXdm4kb9bsYK/XbKEc3f4MF1XEhNtY9Lr+ermxmeq05GX9/ZyLuVxZmbSJTIxkTxt+nS2N2cO4BOswidZn+DRKY9eUKy0ua/ZaoQYGzYWXZouuDm5oXWgFRqjBqmhqejWdMPFyQVeLl7Qm/RQqpUY7jscaZVpKOkowVMznrJrc2PeRhysOYhPl30Ko8Vo5yJboiyBxqTB+LDxUKqVaOlvQWV3JXRGHZJDkzEpgoFwkiRhf9V+zI6eDS8XLwwYBuDl4oVdZbswzGcYRgePRn57PrJbsuHi6IJrE6/Fn47/CX9d9Fc09TfB19XXGmdZ2VWJCO8IODs6429n/4ZHpzwKhUKBv6f/Hb+b/jv0aHuQ3pSOSO9ITAyfCEcHRzg5OEFv1uPbwm+hMWowO2o2UkNToVAo8MrxV+Di6IJov2isSlmFtoE2bMjdgLvH340PMj7Ai3NftCoqQ9GoakSkTyQcFA7o1HSipb8F/fp+nKg/gZSQFFwzir6lZosZLf0tWJe7Dj3aHqxIWgGD2YDMlkwEuQchzDsMc6LnwM3JDZ2aTqv74mCYLCbcvPVmjAsfhxfnvmj9fGfZTsyOmo1+fT9ON57Gbam3XbSvQ6E2qJHZnImxYWOR304r9byYeXj+yPMI9QxF36HHcc89lKWuucZWiPnoUYYwhITQ8DCYhiSJBt5x48h/YmIoDE+fTiE9NNTmfidj/36+Llli305GBg2xt99OOnBy4j1zc9leXx+NuXFx9nHSJSU03LS10VADkOba2ymDDS5UW1ZGV/1XXiGf/Pxz8uxDh3iP7GzKKT09VAzOnOFvEhJ4z4gI8vq1a8nDamtJu489xjEMdkE8e5Y0HhFBvnzvvWxjYIAKz3338Xo5rCAvj3LQ7NlU0mbOtJ83OfQgNpayanY2Xffc3Mhrv/2We4vstrxnD41wy5bxe3kem5o4V0lJnAPZsKTXU7709+e18+fzmbS2ch+pq6OiJLuQnzhBHv7CC1TEDh603f/HHymPysXA163jvC5ebNMNVCoVfH7KN/of+K9RlEpLVUhMvHAy5BMbg4FByDffzA0vNJQEJPufmkzc0BsauPn/6U/coOPjqZl+/jktmuHhJKyvvmJ7117L6/7wB2q548fzoTU1kbAffvjCPp84wQVw880k+rfeoj/tnj3cFD0H7VEWCwnXz49K3549XFijRnGRhYczacP999uSGOTmMkj6tde4cKur6YNqNttOU9raSDCLF3MDrajgQn72WRsRVlRQiCorI8Fce62t0v2RI7ZEEjffbBMeHB1pFfrxR37/u9/ZxqLVMnB69WoKCIcOcRxpaVTuoqLYx9ZWXiuflhmNDL6OjbUpRrW19pYEGY2NnLP0dBLvqVPAn/9sH5/2ww8UJm64gcxiMLZuJUO68kr2YflyttfZyTUDwM43eygaVY1WC15ZZxnGhY1Dfns+/N380avrtW5wtvUp2cXfSJKEvRV7EeIZgqnDpqKmpwaSJNklHgCAc03nkBCQgECPQLQNtCHYIxhakxbODs4wmA0X+M0DQEZzBpwdnBHpE4nSjlJoTUwEcVXCVQCAfZX7EB8Qj4yWDDT2NuLZ2c9iXc46OCocsTxxObxdvJFWmYalCUvh7OiMBlUDSjtKMTlyMjydPS9QEgGgV9eL/VX7rQHYMjbkbsCykcusQoqM9oF2bC/dDi8XL5R0lOBP8/+EzJZMuDu5WwPAAW7iC0cshLerN+p66xDhHYFD1YcwzGeYNV6rfaAdLx57EatSVmFuzFw09TVhQ+4GPD7tcWu8BAB8kfMF2tXt8HT2xL0T7rXGNw1+PiUdJXBycMLIwJE4VH0IZZ1lmBwxGVqTFnNj5uKd9Hfg7eKNByY9gF5tL47XHUdpZyk0Rg1emvcSzjedhwQJX+V9hQcmPoBRQaOwuWgzFoxYgPU567EwdiEWxi603rNd3Y6Cs2EoLydfKS3leg4Pp6L00kukvREj7BUmvZ4nEfPnX6iQmM3AX//KkyE/P/K111/nJnjFFdxkV63ihuviQqPLzJnknaGhbE826Fgs7NPo0RQW0tJIt15efD9rFpWkHTtoHQ0JIS1Om2bjL3V15AORkfb9PHOG/MnJie2dPcvNdvlyewHBbOaJTXQ0hYCjR3nKbTbT0JKTw/lxdCRPT0tj/2NjebqVmEgF56abbAkPTp0iP5XHqdXa4iYKC6lcubjw9Cciwv7U7L33OI/BwfZj+vRTnphVVtKYNVhwA6ioKJXcEwYb2QwG8p2gIPtxHz9u448BAdxr4uM53/KzX7uWp0Xff89n0dDA+b7tNu4vZrNNIauuZn/lMXdru+Hr6muNJ9xbsRf+bv6I8I5ATlsOJEmCi6MLIrwj0NTXhD5dH4ySEemN6bhlzC24Mm5IIMUg7Kvch6nDpkKlU8FkMaFeVY9ZUbPQqGqExqix0u5n2Z/BIllw+9jb0a3txrnGc+jSdkFtVGNl8kproojB6NJ04bPsz/DsrGdhlsxoVDWisa8RygElOjQdWJmyEn5ufvgg4wMsG7kMtT21iPCOQHxAPP6W/jdojVrcNf4utPS3oLanFitTGOBnMBtgtpjxfsb7uH/i/XB1dMWJ+hPQm/S4NvFaADy16tH14FjtMaxOXQ0XRxc09TVBa9QiyjcKzo6MkzKajXB2dMZfTv0FTo5OeGrGU2gfaEe3ttuqLEb52ltZTRYTslqy4OvqCycHJ1R1V2HqsKnIasnCorhFOFV/Co4OjhgbOhbOjs54J/0dxPnF4eqRV6OlvwW5bbm4NvFauDi6oEhZhGN1x7A0finiAuIgSRL+evavCPYMxqrkVXBxsi20c03nUNVVhZFBIzElcgp+rPgRsf6xqOmpwdUjr4bWqEWhshBODk4XTUAhr6U+fR9O1Z/Crcm3wqh3wt//rsCtt1KuSEqyD8I3GCifWCyk15tuIo11dFCpiIigASEujvzv+utpzElKso9zAWzxlrIxVD4V2baNSohORwNKYiLbev11yk0bNlBwf/tte0OsSmUzPPwSenrILyZOJK22tdmMr1VV5BVqNeWj+HgaZJYtI2339fFEavC9zWYqDykpVNY6OqiM1NWx/eJiyiw1NeTJchx2Vxf5YlUV7+XmRh5hNPIeg9Hfz7ny8KAClZXFGP7336d8Jsfxd3ZSvrvqKvJ1lYrtBQWxjc8+41hGjqQsZTJRiR09mvLftGls++abyXOHynAAlbxt2yhzFhXx2c2cyf7l5tr2nd//nork2LE0/K9cyX44OPxritJ/TYxSQYFKOnTowvSs5eWMEzIa6T9pNNp8L0tLWZTs44/p252Xx1gVSaLP/6uv0gdS9p/U6fhnNtP/vbycvqhdXYwxOnSI/uGZmbxGqaSv/BNP2KeZHBign6ncrlbL6woK7ONoCgqYgnb1aptvekMD/WVVKvrndnayxlGTrX6fpNHQ91ajkaR58+iLPrjNu++mH/Datbx3Tw9jccrL6VP7/fccw9NP26fsPXKEfvulpfSfz81lbZfNm+mbe/gwf//885J0330233g5/70ksb3+fo4zM5Pzdvw4/WAffdRWkFKOD+vupk/q0aO2Nt58075fksTn+sUXHHNHB/1ldTo+36ExBnJ17E8/tbWp1/P5nT7Nz4qK+P7DDy9Ycj+Lup466Yn9T0iNvY1So6pR6tX2WuNThvrh57flS819zVJFZ4W0pZAP6duCb6Uz9Wd+NvZFbVBb/di7NF3Sx5kfSyazSXr95OtSWkWalN6YLg3oB6zpZPUmvbSnfI/08vGXpaquKmt1+6zmLLuCgrU9tVJdb530SeYn1nii0o5S6XzTeclgMkgGk0HaWbpTUmlVkt6kl9Iq0qQnDzz5s/EAepNeKmgrkDrVndLh6sN2/vM6o07q1fbaxXptzNsovXHqDclisUjFymLJYrFITaomqUnVJG0q2GStr9Lc1yxVdlVKGoNG+rbgW+mD8x9ctMik0WyUVDqVNfZiV9kuqa3/whLlP+VLL+O9c+9JB6uYb/bN029Kea150nvn3pO+yf9G+ib/G6lX2ytlNWdJkiRJJcoSafUPq6W3Tr1lTe1uMpsYpzDQLikHlNIrx1+xPp+81jzJYDJIKp1K+ib/G6lH2yOtTV8r5eRYrLFBL7xgi8vLySHN7drFuCCzmb7jGzYwVi89nf/39PD15yDXIevqYl03s5lxL3199PXW6Zi6u/Ei4XilpfQFl///5htbDYyyMvIpOe21xcI4BLn2UH8//dyPH+f72lr7dNhaLfmkxUJeM7iWyMXQ0cGxyL7q6enkZ2VlHENvL2OBzpwhT6mpYepxmd+WltripAwGXmM0kk8OjUE1m/lM/snSfVJXl+0+Q0Nxamr4XDs7OceDU7jL169bd2HhzJwc/q6lhc+tsZFzu20b51GuhSJJ5M3ffccxGo2S9NJLLFfxUyUq1uest0v13afrs9ZR+3v636XKzkopuyVbOlV/SurX90u92l6pS9MlbS/ZLh2vPW79nUqnsqYar+iskL7M/VLKaMpg/EvNEWspgq/yvpJu23ab9P7596Vjtcek5w4/J9X31EsWi8Vaj8dgMkjnm85LaRVp1j4ZzUapV9trjYHq0nRJtd21kiTZaslxDi3WuM59lfukk3Un7Xix0WyUfij5QepSd0nKAaWkNqil0w2npbb+NqmgrUA6WnNUymjKkArbuEBO1Z+SVm9bLR2uPmyX7jq9IV16+8zbUlt/m7WIriRJ0vdF30tvnHzDbo61Bq1U1VUl1fbUSiXKEulsw1nphSMvWOs8DYZyQCl9lfeVdLbhrFTRWSFVdlVKbf1t0tairRekF6/tqbUr8fDysZelN0+9KRlMBsloNkpdmi7pQOUB6bWTr0l6k17SGrXS13lfS+Z/BA+fbzxvVxNOY9BInepOqb63XtpUsEl6YPcD0svHXpYkiXXzPs/6XPrg/AdSVVeVlNmceUEsrDy/g0sxqNWUN44fZ5zfYPlEhhxDLkmkn8JC8pr2dn63ezdlrZ9LvV1cTPqtryf9vfGGjSZOnGC8uFy6xGIhPXd0MI777bdZPqGy0paO++eywMtxkjod+ybHYhcXU6b83e8Yuy1JlEfOnqW8ObiumUZDnnjuHHnfT40tL88Wk5qWxn699x7l2EOHuDecP0/63rJFkn7zG8puAwOci8JCW3xqayv/VyrZPznmVS4aK0n8nVy4urKSbX36qW08GRm2OMyGBtZLGoyXXrLtQ6WlNl64eTPl58xM7g1D+ezBg5yTffts+09tLXlkeTlj2zdtYt8uNlf/So3V/5oTpXfeUWHqVB9ERlIDlTMLWSy0lK5cSdeGoCD6vsrYuZNWwuPHeeoyOJ22VsvTlmXLaNk8fJgnGcuXU6Me7N8N8JQlIYGWuuBgWgl1Omr0Wi19zI1GWhOSkmj9DAzkKc3XX9OCsGwZj3jDwmiViIujRj58+MU1b4DW0dmz6WubnMyjyfvv53dqNT/Ly2MbHh48mUpI4HddXXx//fW0Qt50Ey0EajWtACNG2NznDAZaMGprOca2NloqZsxg266uPFkLCKDfrJcXLdwZGbQOffQRrQHJyZyDu+/mSVRZGa00snthbS2trmFhtEA0NXGuQkN55F5dTct6WBj7JFtbq6rY33ffpdW8qopz8tJLF2Zi6emxnYL19dncXvr62P8DB/j7+++nz3BDA62uvfouBLgHWLNAGc1Gq7++1qhFmFcYTjecRlJwEnp1vXB1dLWm4gaAwzWHEeAegAnhE5DXlodgj2C0DbQhry0P90y4BwOGAbg5uVndyWRYJAs6NZ0I8gjClqIt6NP3ISUkBdOHT0fbQBsivCNgtpiR35aPQmUh3Jzc4Ovmi7nRc/HckefgAAfMjZ6Ls01n8eysZ1HfW4+k4CT06fvsTnXMFjPy2/PRoe7A4vjF1s/XnluLJXFLcLrxNMK8whDmFWZ1f6nvrcfB6oO4b+J9F12fGqMGvz/0e1ydcDVmDJ8BVydXHK09ioquCkT7ROPrwq8xOng0HpvyGPZX7cfkyMkYHTzarg1JkpDXlgcFFBgXPg7FymKkVaShaaAJV8VfhamRU+Ht6o1zTecwIXwCTtSdwMjAkYgLiMP6nPVo6GvAy/NeRnlnOaJ8o+Dk4IR30t/Brcm3Wt368tvycbz2OMJ9wnHj6BvRoGpAo6oRNT01mBM9B+He4XBzcoPZYkaPrgedmk5sLtoMX1dfPDH9CWtfTRYT/pb+Nzw48UF4u3pjd/luLI5bjHfPv4vHpjwGNyc3fFf8HeZGz0WkTyQ0Rg2O1x2H1qBlli5HZ7t0xXo9eZpGQ/ew7Gxa0Dw9aU0dGOB6XbWKvCIoiLQnp5DOzbXxGxn79tmnVx1KG0ajze3EZCJNjRvHrHvyKUpXF/lEVBSvUattaWl7enjvKVMYzzNmjI1/mc10JYuOtll7T5ygO9nF4kL5/O3v+8c/8qRYjh/cuZNjTEiw55NKJV18UlPJg999l26/t99OvidnuPr6a/K75ctpXT5/nqf7dXXkNd9+S5fgoTh50v6UTEZlJeegspLPSHZ51unYl4gIWmZPnOBzS0vjM5NPm7Zts2XA27iR8VyenrZT9sHIz2ffY2LoXi1bo2We9/zzfNZ33MH3Oh3n4b772KbBAMClH59kfYLkkGTrCfNgfFf0HebGzIXJYkKkdyR53z+ytB2uOYwRfiNwrO4Yantq8eeFf8aAYQCOCkd8lfcVbht7G841nkNZZxluGH3DBadBJosJpR2lCHAPQOtAK6q7qxEfEI8I7wi8eOxFXBl3JVJDU5Hdkg2TxYQ7xt2BrcVbkRySDGcHZ9SrmK7az9UPCYEJiAuIg9FsRFlnGaYOm2pNwd2v78c3Bd8gMTgR14y8Bt3abmzI3YBFsYsQ6RuJ3NZc7K3Yi2nDpiHYIxgakwYeTh5o6m9CkEcQAtwD4O3iDY1RY82ml9+ejwjvCEiShHfS38GasWvg5+aH3NZcrE6lH/+mgk3YV7kP1yddD19XX7g6uWJ2NF0ZdpTuwNRhU+Hq6Iptpdtw74R7oTao4ebkhi5tlzVeFGC2P28Xb2S2ZGJc2DgUKYuQHJKMCO8IujIb1OjSdsFgNmB08GgYzAY09zUjzCsMrk6ueOHoC1g2ahmmD5uOg9UHMdxnOOID4nG09iia+5uRFJiEHeU7MCliEpaPWg43JzfoTXr88egfoVAoMD96PjYXb8bzs59HYhAD31Q6Ff6e/nc8NvUxFCmLkN6UjrvH340QzxD06nrRqGrEyMCRqOmpQWlnKWYGrMDWrXQhA7j3l5fz5ECGXN7l3DnyiH37+Dp2LOW0mTPts5z9FDQaeq1kZ9O1VKPhnq9Wk84MBso6crxzURH3//p6ykPl5aSPwkLy0v5+8pioKHrqyCdYBgNPn55+mrzn+HHypdRUnopLEuUHgLx5sMfQrl38/U038f2RI5TnHnmEclBfH0+KgoJs2fDq622Z9saNo7zX1cVT7vvuI09paLBl2Vu4kHwoO5sy01X/IO+BAcppK1aQ914sjbmM4mJmNE1J4dz4+tp7Mshu22lp7Pf8+eRnkydzXhsbbfcFuAesXWs7zT52jB5QcsbPI0coC44Zw5Ms+ZS9ooJ7jNnM06rVqy/McixDuN4NgjwZnZ0qBAb6oKWFD3DwsaVaTYE7LIxCffggPn3iBN3ltm7lZnT2LF1IBsfOyDAYuOled91PKy0y+vupGAQHUyEwmykUnD3LzWzGDCovixaRCFQqCgKzZ5Mg+vsp5D/zDAUdPz8uGJOJCtuiRdz43d3JWE6e5LFrfT37l5BAourpIfGUlVGZkYWfykouftldD+BG3tBAl5WCAs6BSsXPRo9m//fv53fe3ozxqajgfUtKyADuuouE3tZGN5T+frp+9PfbPxO573l5JKLp0zkP6el0vYmLo4KSkGAjkkOHbLn+AVs64qefto+V6u4mo/r2W/4+IICudjNn2oSz995jjJebG5lVYiKFiU8/5fpZs4bC59Sp7PfHH5OZtIR/ao0lOVpzFCWdJZg5fCby2/IBBbAobhEivCPw3KHnkBqWirkxcxHhHQGj2YiKrgpE+kTCxdHFmrYbYF2facOmQaFQwEHhYHXr+yTzE0wfNh0KB4XVHeT+iffjQPUBeLt6I681D49Nfcxu3W0p2oKFIxZCZ9LhdMNphHuH42jNUfxm2m9gtphRpCxChFcEPsv5DE/OeBKbCzfj4ckPXxCTMBhqgxrt6na7OimDYTQb8dbZt/DEtCesyQ32VuxFpHek1V1OqVbC1dEV/YZ+eLt4I70pHSGeIVBpVSjuKEZKaAr83PywPnc9/jjnj3bKmyRJaFA1INQrFPsq9+G6xOugM+nwefbncHF0wcLYhUgITEBtTy1yWnNgkSzIb8tHqFcoUkJTMCF8gtXVTvb3Tw1Nxd7yvVCqlVieuBxBHkHYXrodkkVCSVcJ+vX9iPKOQox/DPZX7cdL815CsGcwTBYTDGaD9fn1anuxLncdbk2+FbW9tVBAgbiAOJyoO4Fbkm8BAByoOoAYvxhE+UbB3dkd2S3Z2Fy0GU/PeBqFykLMGDYDee152Fq8Fb+d+ltsL9uO21JvuyBV8IkTdLnw8SH9A/S1Dw8nP+rooKAvKzWycF1RQTqeM4f0X1zMgN5Ro2hAGlqvTE5jHR9PJS01lTyxsJDCSlAQDUVhYbYUsTLeeIPuJgkJNLD09NDV9YEHSE9XXkka3r+ftPXBB+QTCQn8fsECW/IBGSYTfc9vuIH3Nhq56U+dSkXIyYn8xtmZbiLLltniT3U6ukl3dJCXOjiwLtq0aeQDwcEUyIKCqHA8+yx/d/IklR+Z18g8BaDLYUcH5/PoUfIVebOurqaAsm0blaGmJluShrY2Jq2Ji+NnTU22eM7cXBrMrr6az+fUKc7tsWOMy2xroyBiNHLMsoK3bRvnLD+fBj8fHwoqJ07QbcZkoivRTTeRjzU08NrFi7knlZcDrW0WuI06AQBIb0rHc7Oeg8FswNHao1gSvwQKhQLNfc0I9QqFk4MTjtYeRUJAgjUtf1NfE/zc/Kwxei6OLvjD0T/gvaveQ3ZLNlJCU7C/ar/V+DI4eQFA/vJl3pe4Z8I96Nf3Q6lWYkwIC+VszN8IrVFrjceMC4izGioGKxF5rXnwcfWBr5svDlYfhFkyY5j3MPTqezHMZxgO1RzCwhEL4eLognFh4wAwscqzh59FjG8MPFw8sDJlJYxmI1r6WxDnH4d1uesQ4hmCcK9w9Bv6MTlyMrKaszA6ZDSifKMwYBjAhtwNuGnMTQjzCkP7QDsyWzKxbOQymCwmW+xk3Qm0q9tR3V2NsaFjYbQYkRKaglj/WEiShBP1J+DuxNp34d7hONNwBq5Orvi24FvsXLkT3xV9h/iAeGvimouhoqsC2S3ZmBczz6qIyspkrH8santq4ensiW5dNwxmA1wcXGC0GDE5cjLMFjO2Fm9Fn74PrQOtiPaNRlNfE24fezuO1h7FHWPvABR0xXN2cEZldyWujL3SyqM1Rg2KlEWYEjkFxcpijAlh4o3DNYehMWpwpuEMHpj4ABpUDRgfPh7a7kBr6urmZtLoYAPJyZMUlKdMscXdRUZyX9ZqqYgkJ9vHQgPcx2tqyGs6OynHzJ9PY4eXFxWu06cZ6zJmDHlPZSX5JcD/GxpIU6+/TnrZuZMusU5O5JWurqTb+nryMZ2O31VWkv5k5ev4cfKrwcqHHGoQHEzZ5J57GEoxerRt/BYLr5szh+P+7jvyg1dfJS9wcqKs1NpKOp8zh/dsaqK8J7vjzZpF/tzXRxfqe+9lXwfLYM3NNMLISWI2beK+4ufH+wwtbVZXR56ZmUk5avhw7hNz5nCPaWykjFRTw2dXWcm94PRpHghERXF8Mh/t7LTFlRmNHK8sV2u1/L+xkdeoVPyLiOB+ZzTy+d1zDz8fHFoh419RlH5BnL98UFXF16IiCsF5edxYtFouBh8fChGDs6bo9cze4eNDwTg1lQt53jz7DEZy8VEXF2rfF1OSiotJiAAD+Pbt48O8+mr+3tGRG+zx4/TNnDGDxB8YyD83N/q/vv46CV7ejPV6Lly5qKyDA/srE3xNDa0Jf/sbhaPERI59925uhr/9Lf1uR4+2D3YcMYLEJs9HfT0XZUYGhYMbbiBBGI2c0127eJ1SSUZ0++3c2OvrSSiypdvTk0S6YAG/mzqVBDc00NLJiQt8+3bOhbs7LSMjR5K4li7l2L/+mowjLY1C2uAicW5u9LmX48EAChx9fRSO5NpMe/ZQGNLp6M9aWMh5kX3y/f1t87tyJU+usrP5XJydOcc33QSopS5YJIu1cOLUYVOxNGGpVYi/NflWRHhHAAC6dd3wdPK0vu9Qd6CsswxZLVnIbc1FsbLYmuXuQPUBtPS34EDVAeS05KCmp8ZayyijOQMb8zciwD0A82LmQaFQYGnCUsT7x1+gJEmSBCcHJ6iNakR4R0BtVGO4z3D4uPlgT/keOCgcsDB2IY7UHcHNY27Gd0Xf4XfTfocjNUfw1IGnrFmidpbtRFlnGdoH2pHbmou2gTacaTiDA1UHUKQsgkWyoG2gDUq1EhvzN+K1k69Bb9IjuzXb2pcrYq+wOxUJ8giCu7M7DlUfgrerN5bEL8F3Rd/By9ULj059FLOiZqFb242xoWMR4B5g/d2O0h0oVhbju6LvcKz2GK5Puh4KhQLuzu6YNnwaFsUvgpeLF9oH2lHaWYqJ4RPh5OCEhyY/hLvH3w0/Nz/0aHtQ11uHAcMAvsz9EhaLBY2qRhgtRuS25eK6LdfBaDZiauRULE5YjAcnPoi7x92Npv4mHKw+iIWxC+Hk4ISSjhKcqj+F5488j/y2fPTr+/Fl/pe4b8J9+L7kezSpmtCh6UCxshjJwck4VH0I63IobEX7RVuVyFFBo/DCnBcQ4hkCD2cPuDm7YcbwGVi7ZC1i/GNwVfxVdtmiZHh6khfMn2/jT08+Sb4mSRSKDx6ktfbmm8kLoqOprBgMthPUyEjS58qVpEHZd19GXBz5lp8f/+SED/PnU4BpbCSvzMqyL+bY2Umaq6wk//LxoQXylVfIn+V4pFOn2H5LC41URiPpdMYMW6IJGefO8fu6OsZylpaynWnTeJ/8fLZTVkZlYvZs9k1GXh7vmZDAkxVJomAQHc05+egjChSLF9uUJMBWzFFGgG1JQqGgctnfz3l0daWQotWSbzQ2UgmRx/vkk5yDTz/lfGVm8lnKc2cwkHfn59M4tn07n0tfn01YlE/no6Nt9UHkOAgvL/LX/fu5T40Ywf2soYFGrfnzOf+bNpHvTp7Mz3/zG7a9eDHg6OCIqcOmIikoCV/kfIFjtccQ6BEIhUJhpS1Z8I8PiLejUZVOhQZVAxwdHBHqFQo3JzfcOfZONPU1YeqwqfBw9kC0bzTGhY2zU5IkScJn2Z+hpKMEtybfiuyWbOwo3YET9Ses12iNWtyafCtmR8/GqKBRcHJwQl1vHb7O/9p6TX1vPQqVhSjtLMU3Bd+grrcOCQEJOFB9ALF+sVDpVJgWOQ1TIqdgXNg4KNVK7C3fi1dOvIKX570MXzdfa6FsD2cP7Kvah3pVPZr7mzE5YjKmDJuCQ9WHcLTmKNKq0hDqGWrtW5++D0XtRThVfwqhXqFYNnIZAMDJwQn5bfkwW8yYGzMXs6NmY+bwmZgYMRFtA21wd3K38qWjtUext2IvHOCAQPdAtA60wt/NH3+54i9ce+4BUKqVeOX4K+jSdFmLdn+Y8SH69H2o6alBo6oRY8PGora31jovN46+EaODR0Nn0qGprwlRflEYFzYOPdoeKBQK7KnYg8L2QvRoe6zFxm8efTPGho3FmrFrEOUbhTvG3QEHBxrvpg2bBjcnNywYsQCBHoEwmA3IaclBkbLIenq1s2wnTBYmPWpUNWL6sOlYlbIKfm5+ONd8Dj9W/ohvN5utcT7h4TQeG422E+lZs+gFMmcOFZoFCyiXDR9O+eDWW22ZfU+ftmWrmzaNpxZubqRptZo0c/gwhfRHH6WRpL3dFvN51102upYTE1x3HWnqxAkaQmpryQe//ZZyXHQ0eZ2jI/ldTw/lCoWCckhrK2WzwUrSp5/alAV3d9vJ8TXX2Aq3Hj1KXrx4Ma9RKKi43HUXlQInJ8o3lZWUbzw9bckgQkI4b+HhNBRFRXF8AQE8dXFzI49Yu5bjaW6mLPncczblad48trtli02e/vhjytFy3Pf69exrd7dNTt2/n89l4ULKUU1NvK9SafNaiohgu999x3YbG3nf9HSOy8ODMvPRo/ze3Z1Gpy1byJd1OvLDhgYa6VpaGHPV328r9g1wf6qq4meyTPjP4P/EidJHH32Et99+G62trRgzZgzWrl2L2UOj7H8Cstb417+qsHq1D9LSuBG89BI3lBkzuDAWLaImv349td4bbuA18+dzgWRkcPEvXcoNdfFiLvT0dJub2cqVfDAXy3pkMnGTam7m4ggJsRV4LCuj8N/RwUVRU0PrQ3ExmcSkSVwAsbFcQDU1ZBL+/lSsfHzYdmqqrfBrdjbHZDCw77KFsamJBJCTw+vl05jWVgZyP/MM+/TQQ1QSFy2iEjJhAgUfmehl60NnJwlhwgS2OzRDCp8fN+aqKluWptpaLvSBARJyQID9SdCXX1JIKirinB86RCEmIoLE9thjDETu6SEDM5tpeQb4LHp7SbxffUXh8Ikn2Idjx9hOZiYtVkol+2IwUNFqaSHDld0J3d357EwmMqYPP+QJXFcXhZCkJJ6WBQeznZtuMaJR1Yj3Mt7DS3Nfgs6kw4n6E7gu8Tq4ObmhW9uN7aXbce+Ee+3m6FT9Kfi5+aG2pxaJwYno1faiub/ZesLUpGrC8brjGOE/Al6uXjCYDbhz7J3oN9DCmhiUiO2l29E20IZFsYtwpvEMTBYTJkZMRIB7AKJ8o7ClaAuaVE1YlboKEd4R6NP3wdPZEz3aHqzLXYfVqasxzGcYipXFqOiqQJBHEGZGzcSBqgNwVDiiorsCD056EB+c/wAPTX4IaqMaHeoOBHsE42TDSUR4R6CiowLby7fjtQWvwVHhiLreOnxf8j1uT70dGpMGUyOm4kTDCWtBRhl/PftX3Dj6RjtBqayzDAkBCTBLZhS0F6CsswyJQYnQGrWYHT0b+6v2w9XRFWqjGl2aLnRruzFt2DS0DbThusTrUN1Tjd1luzHMZxiyWrMQ4hGCacOnIas5C8N8h+HG0TcCoCBV01ODbk03EgITEB8Qj13luxDrH4vGvkZ4OXthZtRMfJr9Ka6KvwqR3pH4rvg7LIpbBIPZgLy2PBypOYLF8YtxpvEMRgeNxpVxVyLaLxpmixmODo4wW8zYUbYDzg7OiPCOgMaoQU5rDoxmI+bGzLVm0itWFuNY7TEkBCYgNTQVDgoH9Oh6rK4s6Y3pMFqMmD5sOjqVzjh8mOsasK35PXtI13Kxxk2byEvmzePmsHatLZFDUxN5zL59/NzZmZtScTHpOzubpzxDXVNffZXGkPBwBszOmUO+6e9P2goOJn/JyOBm/8EHpOfoaPLIgQFu6hYLDSYpKfz9zJnMVDdxIk9/nnyS9+7rIz+qraWhor+fYzx1isYWFxfSvUJBfvPUU+QPjY3k44DtWkdHm0ufXKS1r49j9/Qk//zqKworY8bQgnzFFTZlqKuLluR77rHNh8XCuXJzo1IUEsLx6nR0A3zmGSooVVVUmAYGOA6jke3ddx/5unz6l5/P/4cPpzfDlCncJ/R67lkuLjbh6dZbyQdLS7m3OTjQ1efxx+0LYVZWkn+qVBTKamqoaMpGp2++IQ+sq+P+p1RyTQ0bBuTrd2Bz0WY8MvkRdGu7cW3itdaTbZ1JZy1rEBcQd0EB54NVB+Hl4oXpw6dbaV5n0uF3+3+Hvy/5u53SrzfpUdNTg3pVPY7XHYefqx/uHHcnQrxCsCF3A8K8wjA6eDRO1J1AmHcYHBWOGOYzDEnBSXjx2It4YOID6FB3oGWgBb6uvgjyCIJSrYQECT3aHkT5RsHT2ROBHoEwWozwcPZAS3+Llb7MFjPaB9pR1V2FLm0XYvxiEOQRhLLOMvTp+3DD6BuYAVSthMaosSvFYJEsqOyqtHOlBoDPsz+HxqjB49MeR7++H08eeBLvL33fmpn0qUNPIcY3Bm5Obrh/4v08fXN2R6OqEXWqOjT0NiDMOwwdAx3Ibs3GqpRVmBgxEWaLGVuKtmBVyipYJAvONZ+Di4MLlGolfFx9MDFiIlwcXbC5cDNa+lvw1Iyn4OjgeNFkQ9tLtsNgNuDWFFtSHaVaidP1p6Ez6bA8cTmKlcVQqpWYGTUTb515C7+Z+hsUKYswKWISWvpb0KfvQ4+2B1ePvBofZX6Eso4yJAUlYULEBEwdNhUfnP8AXZouXBF3BWZGXURQAHCm4QxCjFPh5uKE4cNJy729pHUfHxp3Ro7ktefPUyCfO9fezaulhbR4zTWk79hY7vkvvEDDqbs7aSkujnynuJg8wMGBp9bV1VQeDAab293gNOLy6bBOZzPGRETwc1n+aWmh58uKFeRJDg48CWttpfwk02RXF8MRli4lXQ8MkO4rK6m4REbyty+9xPf33EO5o7vblgrb2Zn32rePio2ckMtiYb9jY3lqn5DA+Rs+nPOxcSP5g6srlSdnZ/I6Pz9bKMeIEVQ85LmQi8Lm5VGOTU2lbHrqFPcUuej3yJGUdcvLyfMv5jLd3895GjWK15aV8f9Dh/jMfX35l5hIPtzaytO9P/yBp3UFBfxs+HA+vw8/JP9KSuJakWW4jAzO/VNPsd+5uXJoSR+WLLlMXO++++47rFmzBh999BFmzpyJTz/9FF988QVKSkoQ9XM5tv8BWVHasUOF667zgSTxAb3+OoX7ffuoFcunB52d3DyWLOFCcHW1uV/JvqLd3RQugoKo3Li48LW11ZZm9/bbuREOFjB+/JHX3PsPGTk9nb9NTOTirq7mw3VyovYtSdx0k5N5jRxvU1PD/+PjqYEXFlIAkIni+HEufHd3Ep3cbmMj2z1/nm3KvrCSZPMhPXWKvykv5zXnz3MjDgkhY3jvPduxbWws+6JQ8L1ez9/Pm8eFPBQnT7JPkyfbPtu9m0SyZInNp1ej4aY/OJ3mZ59x866tpTAWF0cFrr2d95YttwCfhUpF5uriQmFQVi4BWm/c3WkVWrGCn2VkkLji45kyUz7Bi4ggA4uJIUM9c4bW7o8+YhvPPksB7cwZrqupi2vwVsZLuGvCXYjyjYIkSXBQOCAuIA67y3djxvAZ8HH1satWn9OaAw8nD8QHxuNIzREooMDJhpMobC/En+b/idniJMDNmT7hX+Z9iVuSb7G6iGQ2ZyLWPxaeLp7YUboD82PmI9QrFEdqj0Bj0GBG1AwEeQShfaAdWpMWZxvP4qbRN2FX+S5kNmdiUsQkTAyfiC5tFyZHTkZtTy0quysxKXwSGvoa4OXihYPVB5EakgqVXoV5MfOsrnhZLVnYWbaTme4cnBHtS1Ogp4sntEYt2tRtcHV0RYR3BE43nIZFsqCutw4PTnrQKjSpdCo0qZqgMWnQoe5AhE+EdWwAharslmwEewbDUeEILxcvtPa3Yl/VPtyQdAMSAhPw6slX0a/vx9iwsZgdNRuODo44UHUAPq4+aO5rhreLNxICE5DXngeVToVgz2DcOe5OOCgcUNdbh4ymDBR1FMHD2QNzoudAISlQ0V1hF1dV31OP1069hkemPILKrkpE+0ZjVNAoFLUXwc/dD3sq9uDOcXci2CMYXxd8jWtGXmM9XazrrUNBewHqeupwa/KtCPEKsY49rTIN04ZNQ6hXKHq0Pchry8P04dNR11sHF0cXqHQqq2BhtphRqCxEp6YTV8ReYT050utpBKip4Xr19eWmI6Ori3Th50eBPzeXwveBA3xtbSV9ennxtGmoX3dVFWnjzBnSSVcXNxsPDxpKRo/mRnf99eRZ8fE2+i0pIb+cO9dm6X38cZ50yfFIHh40Pn3zDU8yjh1jX5KSuGF/9BH5z+HD5EEODrZ4TxlFReSpzc00jLS0cMxubty4JYnj/+ADChI33si+1dRQqSkooMICkI/Lc1NXx/mcxJA7fPMNBaLgYLqt3X8/eWtdHRWQnTvJQx98kPN24ADH2thII09sLJ9BYiJ5XUsL/5dP+QsKeFL+xz/yvkeOkKfW1rKNuDieLL3wAo1XZWVUZpqbqdxotVwDTU3coyZPtp3EffQR5yIlhe2NH0++f+aMTQhVKtkXf38KnHv2AOu+0mHUaL21LltmcyZi/WIxYBzAZ9mfYXbUbJxsOIkAtwB0ajrx0OSHEOsfi6a+JqzLWYfK7krcO+FejA4eDQkSjGYqKZktmRgdPNqaya1B1YAdpTuwKG4RqruqERsQC4VCARdHF7g5uaFeVY/JEZNR0F6A1v5WjA4ejb+m/xXvLmHdtJlRM/FjxY9QQIE5MXPg6uiKDk0HPJw9EOTBSejWdqOyqxKTIibB0cERx2qPIS4gDkHuQfg853OkhqbC1ckVySHJsEgW+Ln5wSJZkNeWh/Fh46Ez6eDk4AQHhQNyWnPQqemEwWRAQmACgj2D4e/uD4vFgkJl4QXucKUdpTjdcBrLRy3HB+c/QLR/NFKCU5AYnIjWgVb0aHvQoekAJLppH6s7hunDp8PPzQ96kx6fZn2Km8bcZHWfa+6jMQ2gi+JnOZ9hxagVKOqwxSdZJAtaB1oR4xeDpr4mvJP+DhKDEnH72NutSuq67HXwdvXGzck32/X3ZN1JdGm6MC58HEb4j7B+XtheiDEhY9Dc1wwfVx+o9Co0qhoxM2omMpozEOoZiq/zv8as6FmYFzMPB6sO4kzDGUT5RuG6pOvg7eqNpr4m+Lv5W+dXzqJYV0flQ5J4ktrfTzpOTr7QDTgzk/QdH08BW+YpRUUUoHft4omLSkXac3Ii/5Ld9QDu62YzabavjzxFrj9nNJI+z53j72TeCrBNOfvbrbeS9mUDVGsrFRFXV8pm06eTro4fJ81FRZEnenuT18yezXZMJvLDCRNIl93d5DuurjaX5Pp6Gq+am9nnxETKa21tlHEKCsiPnZ2pcOn1tqyadXVUXMrLebpWXk7Zb/Jkyli33cYxTZ3K/WDECPbLw8Om8FRWcr7Xr+eYg4M5h3PmUB5ra2PdPLmmVESEvSJrMpHvOjmRf776Ko1BcoiFQsFnd/Ik5+XwYdves3w5lSGlkvL1jz/aylV4enKeoqK4P9TW8tlVVvLz5cv53A8coGGpr68Pv/nNZeJ698477+Cee+7Bvffei6SkJKxduxbDhw/Hxx9//C+1ExPDFN5VVbbYlJgYuliZTNwQ1q/nhjJvHhfN+fM2BWr5cj48eUPs6OBGdOoUH/revRQeHn6YD+3ll20FrTZv5gNbuNA+wUNqKj9/4AEuxqIiEsLYsbbTpMhIEtyxYzxlMRhsBR937ODCfeQRbpSALXWu2WwLbFMo2E+djoRfW8tFajBwwRw6ZEs73tREhrB6NYnl5pt53QsvsP3f/57MQg7ydnenQiEfY+7fz3vLAeaff06LNkCCHxxsLFtkjEZb4VyARDXRfn/BsmUkisxMjlUmnjVr2L+TJ21FyIKDbYS8fj0FB43Gdqq3YAGDQ5OTyYwHBvhMIyI4L/ffT6XsllvY1sGDVNwyM3m/kBCukz/8gWOXTymnTZNQ0JWNjSs2IiUkBRbJYg0eBoAJ4ROwq2zXBdY8lU6FdnU7jtceR4myhKdCcYuwJnUNMpozkBiUCHdnd3SoOxDjF4MI7wiMDbVNZJBHENyc3FDUXoQzjWdQ3FGMbm231YVCLtK6uWgzNuZvRJemCyaLCc4OTApw05ibEBsQC71ZD5VOhWi/aET5RqFH14OM5gw09zXDaKa/+riwcWhXt1vvrTfp8fhUptN2dnRGiFcIQrxC8Fn2Z/jjsT8iwjsCWS1Z2Ji/EfWqeowLG4eHJj8EhUKBBlUDSpQl+KbgG5xsOAmABVxLlCXIbc213iOtMg3xAfEY4TcCw32HQ2PU4M3Tb2JF0gr87dzfcKzuGJ6e8TTeuOINLIpbhEifSNT21CLQIxCjgkahT9+Hss4yZLdm45Yxt+D3M3+PKZEMnjGYDVQ64q6At4s3FscvhoujC/ZV70OcfxwMZmoiH2d+jJaBFjw982nktOQgpzUHbeo2FLYX4seqH+Hv7o8VSdS6u7XduC31Nvi6+aJ9oB0VXRVQQIEQzxAsjF0IDxcPawHM0w2nEeUThW8KvsGZhjPIa8vDxPCJ8HT2xITwCUgOScbEiIloUDVY+zIubByifKOgN+nx4INUOvLyyKOCg7mOc3K4UVVW0kqoUpGXOTuTH2RnU8mPiaHQfvAgaa6iwlYoW1bCGhrokpKXx+sTE7lpf/YZDRFyGQBnZ/62o4PCAEDBIzycltCEBLaVkcH306eTlpcvp/W3uZljcXAgj/PxYb9ycigw3Xwzrw0M5Ib86qvkVyUlvFdyMgV9Z2fSeU4ON9OtW/n/n/7E6yZMYJ8CAnhinpLCPl9/PXmE0UhBZ8ECbrpyXbk9e8gnbr2VAoSXF/nwbbdReRo+nO+nTKGSlJNDnrVgAYWrkBAqfhoN+ceuXRyfHHO1fDk3dDn+q7+fSpezM/n/6NE2IUCOO5KLix84wDarqviMZs7kmCIjafyTJLa1YgUt0bIRz8GB81lQYHMvTk/naaRWy+f0zc42nDN8BhdHF6gNapgtZpxtPIsdZTvw+qnXcee4O5Eckox+Qz/mj5iPKcOm4POcz2G0GDHcdzh83Xzx2JTHoNKrsC53HR5NexQ/VvwId2d3zIqahUhvSqxaoxabCjYhKSgJRcoipIalol5Vj25tNzRGDd499y60Ri0UCgWyW7MR5BkEFycXfHT1R3B2dMbcmLlwcnDCtYnXYnnicrg4uuC74u/wdf7X8HS2xVhuKtiELk0Xsluysa10G8aHj4eLgwvONJ7BuLBxmBI5Bfnt+dCb9PB09sTG/I2o663DhPAJ+DDzQ6RVpqG8qxwWyYJNBZtQ2F6I8u5y1PbW4mzjWfzhyB9Qp6qD0WwEAOwp32MtBOvt6o1xYeMQ6BGIaL9oDPMZBn8Pf/i6+WKE3wj06HqQHJKMPkMfclpzUN9bj7ONZ9E+0I6jtUeRHJKMUK9Q61h0Jh2O1R5DdXc1enW9UEABrUmLRXGLkNeWh6yWLHxX/B0cFY7Iac1BbmsuHpn0CBIDE9Gj7cHbZ97Gqh9WoU5VB1837hO5rbk413QOB6oO0G1Z3Y6jdUehNWrxefbnAICU0BSoDWoEugfiuSPPoVRZCm8Xb/Tr++Hr6ov8tnzcMPoG7Kvch6L2IowLH4d7J9yLBbF0y9Ob9CjvLMdzh59DZnMmdpbuhM6ow5aiLYiJIS1otZQl3Ny4xrdvZ0A/QN4kF1gdNcqW+t5g4G9++IE0c/XVPJnOzyevKC+nAeLoUcpJAPdwd3cqQ62t5J/bttlo5pNP+P2uXTSyAjQ+ywbisDDKf3v3MsRh925buZDwcPLe/HyeUpeUUO7JziaN5+XRPbC7m+OcP590npfH/oWFcZxyOYRlyyhjFhZSTo2P53hHjSLNKxTkL2o1x19RQd6jVFJmmz2bPHLpUp5qeXtTZvzwQ/Kb9et5T/kkrqeHbRqN5FMDA5wvFxfywMpK9n3xYn5XWUmjd04OPZS2baOs9sYbtnmrrua94uLIizQa7gVPPcU9KDraZiRKSOCeFh5OXhYdzef85psc45gxVPjkJBk7d3KdaDR0jTx+nONZs4Zy/+7d3Fd8fe0LEf8S/lefKBkMBnh4eOD777/H9ddfb/388ccfR15eHk6cOPEzvybkE6WPP1YhP98HcXEkgFWrOIGHDnFzUihshb/eeosLQdaOZY28sZGTu349FQYfHwrPdXVUMmShe8ECbly3306Cqajgg2tt5cMuLOQ1TU3cON3c+F14OO+p0VBJaW+n8uXqysU1aZItu4ns3pedzcXj4cGNUXbFOXeOwkV9PRdlcTEtK42NbFeuT+Tvz4Wn03HccuXl4GASS3ExiTc7m2OTrbCytbqoyFZg18XFpngtXkyi0el4ndlMwq6u5vV+fhxbRgb7fe21bHf6dBLXlVfytC8vj0qbnMWmu5sK4o03kikajWRsRqPN8trby2er1dqCAJ96is+8s9NWcLejg4wmPd2WBREgoU2aRKIKDbVl5Bo2jMxqyxbec9EiClZlZbTQT5oEKILL4OXsZbX4Zbdmo0fbg8Xxi3Gy7iQssCDGLwbODs7QmDQIdA9EXU8dPs76GCaLCQ9NfghuTm7QGXXMzlZ/AiGeIUgJTUFuay4SAhOsdXyMZiMcFA4oUhahW9uNII8ghHiGoK63Dj26Hjg5OKF9oB1Xxl6JEK8QNPQ2QKFQQKlWYlTQKJgtZpglMwLcA2CRLFh7bi1G+I1At7Ybc6PnwsfNB4HugdAYNda6Sx3qDuhMOhyuOYyVKSvh5uSGbwu+xezo2XBzckOwZzD69H04UnMEI/xHwNfVF93abowMHAmLZIGvmy/69f0oaC/A0dqjcHdyx53j74SXixfcnNxQ2F6IZw4+g9VjV1uzQn2Z+yWmD2cmJgBYlbIKzx1+DpMiJuGmMTchqyULs6JmwUHhAEmS8OrJVxEXEIc1qWvQoelAh7oDI/xGwGgx4v2M9zFt2DQkBCSgqrsK04ZNg5eLFz7J+gShnqGYN2IeyjrL0K/vx+yo2dheth0rk1eiT98HlV6FWP9YZLVkQaXjyVp5VzkSAhLg7OiMImUR6nvr4eLoYq0bcvfuu/HMjGfQ1N+ExMBEHKw5iEB3xnbE+cehSFmEYM9gXJ94PbaXbsfM4TPxRe4XGBU4Cg4KB3i6eGL6sOmo7qmG0WxEj64HC0csxLrcdZg1fBYinFJRX88N5vvvqZjcdhv5QFgYDTFjx1KgCA8nTaSl8TuVikJ1RQX5kFJJGk9K4rVZWeRxx47ZNhaFgt8PDNhox8mJ9Dx5Mumgt5e0sXUr+WdqKpUR2Yc+Ls7m5hwZSQOSnFDhiSdsfPXMGdJqXh75paw8tbSQH5SWUjior7e53Z48yf7V1pIv33QTeYzBwD6mpJB+5XhNSeIY4uO5sXp68h4BARTKIiPJ693caDgbHJC8aBE369ZWzn1YmO0kSaEA/v538s74eApuo0fbsnIaDBzrtm3koZJEZdFiIY8LCuK4/P2pfLa1Ubhpa2MfxowhD0tNJd81GtmngADy1/h4zv/NN3MtxMRwPm+7jXvDmDG8V16ezeVQ9jJQqTgHsoBkUehR3FGMcWHjsKN0B1wdXVHRVYF+Qz/aBtowL3oe5o6Yi051J5QaJSaGT4TRbIRJMtklVAAYd/RRxkcY4T8Cfu5+6FB3YEL4BAz3HQ7lgBK7Knbhvgn34f2M93HNyGsQ5hWGAcMA0irTkNOag9TQVNw9/m7ktuaiobcBvu7M3Pnaydfw+5m/t9Zq0xq1GDAM4HzzeXi7UDmRFYGSjhJEekeipqcGfm5+iPSJxI8VP2JpwlLU9NSguKMYMb4x8HLxwvH645gaORWhXqE423gWMb4x2FW+Cw9OehBKtRINqgbU9tTi3omsr2aymPBF9heYFT0L9b31uHrk1WhUNcLTxRNlnWUY0A/AJJlQ3V2NEf4jrDFLGqPGarzq0fagqqcKWqMWBcoCPDTxIbg5u2FPxR5MCp+Eg9UHodKpcPvY23G87jj6DH1QqpW4Y+wdCPIIwtbirXB2cEZScBLKOstQ1V2FZSOXoW2gDc19zUgKToLerMf3xd/jihFXIKMlA7OiZiHUMxTh3uEI9QqFJElIq0pDWkUa7ptwH8K9wxHuHW4tUG62mFHcUQxPZ0/EBcThq7yvUN5VjllRszAneg4quyoxYBiAs6Oztb6ck4MTvFy8YLaYUdlVifW56zExfCKWjVqG27bfhj/O+SPCXOLh5eyLXbvIy4xG8p66OsbxnT5NYV+lomxisZD3ffMN+c/77zNswMvLVnB50ybS1MAAvWHa2ymvyK7JLS38TqbR666jQjVqFI3Uw4aRZ95yi63oqSy3ODuT9+h0vObbbynkf/klP29u5quDA9uUM4CqVLY4qAULyGvvv5+yYUYG+cjs2aTVyZNtp2VubmxTPg3LyuI8jRlD/jR+PMdWXs4xf/UV8OKLvN+4cTY6/OYb8ulHH6USd+wYx7d+PeXHH36gkUSnI2+SFbYffuD8uLtTLj1xgvPk4WFzQUxIoCwnF/kNCCDP1WhoZHd15fVpaVSqNBqOobmZfWxu5mdeXjQ6lZeTp6nVbLO5mUpaQwP3EVmhvfpq8roVK7hOAgL4XWoq+VxPjy3eafx4oL6+D7Gxl4HrXUtLCyIjI3HmzBnMGJTC5PXXX8dXX32F8sFRWv+AXq+HfpBDqUqlQlRUFI4fb4SPjw+KirgA/f1tiRyuv56EEhfHxXXwIB9QezvjZj75hAK8ry8fVFcXH5x8nDg44xFgO20KC6OmfsUVNouiLKS4ubGt2lr2p7vblkJbxsAAF1BEBDfXzk77FL46HZWJQTqkNXuejD17yBx8fXl9cTH7un0750B2A5Qkjjs7m//fdhv7uGkTsyMNDND6kJbG/+WMSR98QKJes8aWBaW/3971rqaGysTSpbbNfGixtD//mXO9aBH7KTOEoiJbhioZWVm0imi1/IuJ4YmZbM3o6ODcbthAxrR6NecuN5fPLDHRVnBNRmMjmaCcpctg4JwnJZHATp3i0bKvL9uXi5bJSE/nGtlZuhOJwYnIbslGgHsATBYT5sXMQ0t/C5r6mhDjFwNPZ09YJAsyW+gyF+kdia3FW3H7uNtR11sHZwdnNPc1Y2bUTAoEqgbcOOZGawFEWRjIasmCn5sfipXFWBy/GMoBJYI8g5DVnAUvVy9E+0RjwDAAnVl3gd88QH/wMK8wu2K1P1b8iFGBo2CSTDCajQj1DEVVdxVmRM244PcAMzZVdlfirnF3wdHBEfsq92FK5BScbjhtLbZYrCyGv5s/g58jJ0Nn1GF93nrcPf5uuDi6XHDC1jHQAU9XT2vmOJ1JBzcnNygHlAjwCMC2km24Mu5KqA1qlHaWYnLEZJxuOA1fVwZda4waBHsGWwUjGZIkobmvGb5uvvB29UaHugMVXRWYOmwqCtsLrRn4Psr8CHeMvQP7q/bjhtE3WH+/o3QHYv1joTPpoDVq0a3txoLYBdaYDEmSkNuai+G+w9E2QHPlqKBRVIqNGrg5uaGmp4ZpxB3dsL96P+bFzLMrYHuq/hQSgxJhNBtR3lkOhUKBeSPmWb8vbC9EQmACXB1dra6LRUUUlgMDucnOns2Nb+FCKg6DC5jm5VEhuvJK+rOvXMnThHnzSD8LFtiuValIiyYTN2aNhvxg1Sq+jhlDOggIIP34+HDzra/niaycllxOFFNUxGvCw2nc2bLFlgFPrWa/urqoDCxbZuvHqVO8148/8jRnYMBmKOruJq2G2cvjVpjN3MyDg0n7cXGk695ejkcuvnj2LF/z8kjHsle3SsUN29GR17/8MvlAaCj5eUkJlbnGRvbr7FnObVQULdBJSexrUZEtKYUcA2AycePu66MSKCetyc3lPU6fJm9raaFA5uXF/k2bxnl3d6eS29PDfaG4mPxw5Ei2Jwd9d3RQkZPjl+TU4GazzcA2NJOgjG3b/pFYQ1ODlv4WzIiagd/s+w0envww1Ho1PFw8UNNTA1cnV1wRewU61Z0I9AiESq9CfW+9laZqumug0quQEpqC47XHMXXYVHRru6EcUOJM0xk8OuVRHKw+iD5dH65Puh6Hag7Bz9UPU4ZNgYujC/r0fTBZTDCZTdiYvxGzomahWFlMAwzoHpcSmmLlJU8ffBovzH0B5Z3l8HPzQ7h3OPaW74Wvm+8F6c2fPvg0Xp73MnaX74aDwgE+rj4o6SiBq5MrVqeshr+7P+p66lDWWYbYgFgEewTD08UT9b310Bq1cFA4YEzIGCgUCnSqO/H22bexMHYhFsUtsvKFvRV7YTQbMX34dIR7h6O1vxX+7v5W17fvi7+Ho4Mjs+6FjoPRYsSRmiNYnbramuRFb9Ljx8ofEe8fj9reWsyLmYedZTtxS/ItMFvMcHd2R0tfizV7aoxfDJQDSqgMKiSHJKOhtwHuzu4I8giCQqHA6frTqFfVWw1SXZoueDh7WO93su4kHBWOcHdxtxp+JEnC84efR0JgAu6eYHOPyWnJQWZLJpaPWo4BwwAskgXBnsHWpB4NvQ3o0nZhfPh4GMwG/OX0X3DXuLsACShQFsDJwQmL4xejtpZ0IRsjvvuO8Tn19RcvVZCeTpoOCSF9bdxIeWtw0Vc5k+M11+AfY+A95BgWtZqeMEFBFKpra2lUeOwxfmcwkM60WlusN0B+UF5OQ5SDgy29f3GxTc7IyuJYMjNtikpQkC1rW1kZZY6uLvKwjRspAw0uSL1hA8cohw7IBmFZztu3z6awlZVReZT7PXIkZcqcHPu+V1bSEPzEE5Tbyst5D3kv8fMj35Hdjf/2N94/Pp68PzKSvOn8eRqtByfZycnhWGT3RYAnO7m5nJPx49m/5ma2/+WX5JPjxpF/yvtVVRV5oOwpUVzM+W1tpcHr6FHOlV7PZyZnAl261H6NWCyUw8ePJ9/Oy+PcXXFFH4YPH47e3l74Ds0mNgT/JxSls2fPYvr06dbP//znP+Prr79GWVnZBb95+eWX8corr/ya3RQQEBAQEBAQEBAQ+D+ExsZGDJNjV34CTr9SX/4tBAUFwdHREW2yM+k/oFQqERoaetHfPPfcc3jiiSes7y0WC7q7uxEYGGiXZUvg59HXR227sbHxF48lBQT+JxBrTeDXglhrAr8WxFoT+LUg1tq/DkmS0N/fj4jBR18/gf/VipKLiwsmTpyIQ4cO2cUoHTp0CNcOLtU8CK6urnAdkrLJb7Cfl8C/BB8fH0F4Ar8KxFoT+LUg1prArwWx1gR+LYi19q/hl1zuZPyvVpQA4IknnsCaNWswadIkTJ8+HZ999hkaGhrw4IMPXuquCQgICAgICAgICAhcpvhfryjdcsst6Orqwp/+9Ce0trYiOTkZaWlpiB4crScgICAgICAgICAgIPAfxP96RQkAHn74YTz88MOXuhv/VXB1dcVLL710gRujgMB/GmKtCfxaEGtN4NeCWGsCvxbEWvv/i//VWe8EBAQEBAQEBAQEBAQuBRx++RIBAQEBAQEBAQEBAYH/LghFSUBAQEBAQEBAQEBAYAiEoiQgICAgICAgICAgIDAEQlESEBAQEBAQEBAQEBAYAqEoCdjh5MmTuOaaaxAREQGFQoGdO3de6i4JXIZ44403MHnyZHh7eyMkJATXXXcdysvLL3W3BC5DfPzxx0hNTbUWY5w+fTr27dt3qbsl8F+AN954AwqFAr/97W8vdVcELjO8/PLLUCgUdn9hYWGXuluXJYSiJGAHtVqNsWPH4oMPPrjUXRG4jHHixAk88sgjOHfuHA4dOgSTyYRFixZBrVZf6q4JXGYYNmwY/vKXvyArKwtZWVlYsGABrr32WhQXF1/qrglcxsjMzMRnn32G1NTUS90VgcsUY8aMQWtrq/WvsLDwUnfpssT/iTpKAr8errrqKlx11VWXuhsClzn2799v937Dhg0ICQlBdnY25syZc4l6JXA54pprrrF7/+c//xkff/wxzp07hzFjxlyiXglczhgYGMDq1avx+eef47XXXrvU3RG4TOHk5CROkX4FiBMlAQGBSw6VSgUACAgIuMQ9EbicYTabsWXLFqjVakyfPv1Sd0fgMsUjjzyCq6++GldcccWl7orAZYzKykpERERgxIgRuPXWW1FTU3Opu3RZQpwoCQgIXFJIkoQnnngCs2bNQnJy8qXujsBliMLCQkyfPh06nQ5eXl7YsWMHRo8efam7JXAZYsuWLcjJyUFmZual7orAZYypU6di48aNGDlyJNrb2/Haa69hxowZKC4uRmBg4KXu3mUFoSgJCAhcUjz66KMoKCjA6dOnL3VXBC5TjBo1Cnl5eejt7cW2bdtwxx134MSJE0JZEviPorGxEY8//jgOHjwINze3S90dgcsYg0MkUlJSMH36dMTFxeGrr77CE088cQl7dvlBKEoCAgKXDI899hh2796NkydPYtiwYZe6OwKXKVxcXBAfHw8AmDRpEjIzM/Huu+/i008/vcQ9E7ickJ2dDaVSiYkTJ1o/M5vNOHnyJD744APo9Xo4Ojpewh4KXK7w9PRESkoKKisrL3VXLjsIRUlAQOBXhyRJeOyxx7Bjxw4cP34cI0aMuNRdEvgvgiRJ0Ov1l7obApcZFi5ceEHmsbvuuguJiYl45plnhJIk8P8Ner0epaWlmD179qXuymUHoSgJ2GFgYABVVVXW97W1tcjLy0NAQACioqIuYc8ELic88sgj+Pbbb7Fr1y54e3ujra0NAODr6wt3d/dL3DuBywnPP/88rrrqKgwfPhz9/f3YsmULjh8/fkHmRQGB/ym8vb0viLP09PREYGCgiL8U+I/iqaeewjXXXIOoqCgolUq89tpr6Ovrwx133HGpu3bZQShKAnbIysrC/Pnzre9lX9c77rgDX3755SXqlcDlho8//hgAMG/ePLvPN2zYgDvvvPPX75DAZYv29nasWbMGra2t8PX1RWpqKvbv348rr7zyUndNQEBA4N9CU1MTVq5cic7OTgQHB2PatGk4d+4coqOjL3XXLjsoJEmSLnUnBAQEBAQEBAQEBAQE/jdB1FESEBAQEBAQEBAQEBAYAqEoCQgICAgICAgICAgIDIFQlAQEBAQEBAQEBAQEBIZAKEoCAgICAgICAgICAgJDIBQlAQEBAQEBAQEBAQGBIRCKkoCAgICAgICAgICAwBAIRUlAQEBAQEBAQEBAQGAIhKIkICAgIPAfR11dHRQKBfLy8i51VwQEBAQEBP4tCEVJQEBAQEBAQEBAQEBgCISiJCAgICDwvwJmsxkWi+VSd0NAQEBAQACAUJQEBAQEBP4HsFgsePPNNxEfHw9XV1dERUXhz3/+s/X7mpoazJ8/Hx4eHhg7dizS09Ot33355Zfw8/PD3r17MXr0aLi6uqK+vh49PT24/fbb4e/vDw8PD1x11VWorKy86O9GjRoFDw8P3HjjjVCr1fjqq68QExMDf39/PPbYYzCbzdbf/VK79fX1uOaaa+Dv7w9PT0+MGTMGaWlp1u9PnDiBKVOmwNXVFeHh4Xj22WdhMpms38+bNw+PPfYYfvvb38Lf3x+hoaH47LPPoFarcdddd8Hb2xtxcXHYt2+f3RyWlJRg6dKl8PLyQmhoKNasWYPOzs7/zAMSEBAQEPi3IRQlAQEBAYF/G8899xzefPNNvPDCCygpKcG3336L0NBQ6/d/+MMf8NRTTyEvLw8jR47EypUr7ZQLjUaDN954A1988QWKi4sREhKCO++8E1lZWdi9ezfS09MhSRKWLl0Ko9Fo97v33nsPW7Zswf79+3H8+HGsWLECaWlpSEtLw9dff43PPvsMP/zwg/U3v9TuI488Ar1ej5MnT6KwsBBvvvkmvLy8AADNzc1YunQpJk+ejPz8fHz88cdYt24dXnvtNbv5+OqrrxAUFISMjAw89thjeOihh3DTTTdhxowZyMnJweLFi7FmzRpoNBoAQGtrK+bOnYtx48YhKysL+/fvR3t7O26++eb//MMSEBAQEPjXIAkICAgICPwb6Ovrk1xdXaXPP//8gu9qa2slANIXX3xh/ay4uFgCIJWWlkqSJEkbNmyQAEh5eXnWayoqKiQA0pkzZ6yfdXZ2Su7u7tLWrVvtfldVVWW95oEHHpA8PDyk/v5+62eLFy+WHnjggX+63ZSUFOnll1++6Fiff/55adSoUZLFYrF+9uGHH0peXl6S2WyWJEmS5s6dK82aNcv6vclkkjw9PaU1a9ZYP2ttbZUASOnp6ZIkSdILL7wgLVq0yO5ejY2NEgCpvLz8on0REBAQEPh1IE6UBAQEBAT+LZSWlkKv12PhwoU/eU1qaqr1//DwcACAUqm0fubi4mJ3TWlpKZycnDB16lTrZ4GBgRg1ahRKS0utn3l4eCAuLs76PjQ0FDExMdYTIPkz+V7/TLu/+c1v8Nprr2HmzJl46aWXUFBQYNev6dOnQ6FQWD+bOXMmBgYG0NTUdNHxOjo6IjAwECkpKXZ9GjwH2dnZOHbsGLy8vKx/iYmJAIDq6uqLzqmAgICAwK8DoSgJCAgICPxbcHd3/8VrnJ2drf/LSsbghA3u7u52yockSRdtR5Iku+sGtyu3fbHP5Hv9M+3ee++9qKmpwZo1a1BYWIhJkybh/fffv+j9B7f5r/Rr6BxYLBZcc801yMvLs/urrKzEnDlzLtpnAQEBAYFfB0JREhAQEBD4t5CQkAB3d3ccOXLkP9bm6NGjYTKZcP78eetnXV1dqKioQFJS0v/3docPH44HH3wQ27dvx5NPPonPP//c+vuzZ8/aKVxnz56Ft7c3IiMj/+1+TZgwAcXFxYiJiUF8fLzdn6en57/droCAgIDA/xxCURIQEBAQ+Lfg5uaGZ555Br///e+xceNGVFdX49y5c1i3bt2/3WZCQgKuvfZa3HfffTh9+jTy8/Nx2223ITIyEtdee+3/13Z/+9vf4sCBA6itrUVOTg6OHj1qVaIefvhhNDY24rHHHkNZWRl27dqFl156CU888QQcHP79rfSRRx5Bd3c3Vq5ciYyMDNTU1ODgwYO4++677TL2CQgICAj8+hCKkoCAgIDAv40XXngBTz75JF588UUkJSXhlltusYtB+newYcMGTJw4EcuWLcP06dMhSRLS0tIucGv7T7drNpvxyCOPICkpCUuWLMGoUaPw0UcfAQAiIyORlpaGjIwMjB07Fg8++CDuuece/PGPf/wf9SkiIgJnzpyB2WzG4sWLkZycjMcffxy+vr7/IwVMQEBAQOB/DoX0U47bAgICAgICAgICAgIC/6UQ5ioBAQEBAQEBAQEBAYEhEIqSgICAgICAgICAgIDAEAhFSUBAQEBAQEBAQEBAYAiEoiQgICAgICAgICAgIDAEQlESEBAQEBAQEBAQEBAYAqEoCQgICAgICAgICAgIDIFQlAQEBAQEBAQEBAQEBIZAKEoCAgICAgICAgICAgJDIBQlAQEBAQEBAQEBAQGBIRCKkoCAgICAgICAgICAwBAIRUlAQEBAQEBAQEBAQGAIhKIkICAgICAgICAgICAwBP8PgObxU+H8iwgAAAAASUVORK5CYII=",
         "text/plain": [
          "<Figure size 1000x800 with 1 Axes>"
         ]
        },
        "metadata": {},
        "output_type": "display_data"
       },
       {
        "name": "stdout",
        "output_type": "stream",
        "text": [
         "lambda=1.0078\n"
        ]
       },
       {
        "data": {
         "text/html": [
          "<div>\n",
          "<style scoped>\n",
          "    .dataframe tbody tr th:only-of-type {\n",
          "        vertical-align: middle;\n",
          "    }\n",
          "\n",
          "    .dataframe tbody tr th {\n",
          "        vertical-align: top;\n",
          "    }\n",
          "\n",
          "    .dataframe thead th {\n",
          "        text-align: right;\n",
          "    }\n",
          "</style>\n",
          "<table border=\"1\" class=\"dataframe\">\n",
          "  <thead>\n",
          "    <tr style=\"text-align: right;\">\n",
          "      <th></th>\n",
          "      <th>sid_index</th>\n",
          "      <th>SNP</th>\n",
          "      <th>Chr</th>\n",
          "      <th>GenDist</th>\n",
          "      <th>ChrPos</th>\n",
          "      <th>PValue</th>\n",
          "      <th>SnpWeight</th>\n",
          "      <th>SnpWeightSE</th>\n",
          "      <th>EffectSize</th>\n",
          "      <th>Mixing</th>\n",
          "      <th>Nullh2</th>\n",
          "    </tr>\n",
          "  </thead>\n",
          "  <tbody>\n",
          "    <tr>\n",
          "      <th>0</th>\n",
          "      <td>52</td>\n",
          "      <td>snp495_m0_.01m1_.04</td>\n",
          "      <td>5.0</td>\n",
          "      <td>4052.0</td>\n",
          "      <td>4052.0</td>\n",
          "      <td>2.990684e-23</td>\n",
          "      <td>0.418653</td>\n",
          "      <td>0.040052</td>\n",
          "      <td>0.007373</td>\n",
          "      <td>0</td>\n",
          "      <td>0.451117</td>\n",
          "    </tr>\n",
          "    <tr>\n",
          "      <th>1</th>\n",
          "      <td>392</td>\n",
          "      <td>snp1422_m0_.49m1_.5</td>\n",
          "      <td>3.0</td>\n",
          "      <td>2392.0</td>\n",
          "      <td>2392.0</td>\n",
          "      <td>8.251922e-23</td>\n",
          "      <td>-0.416495</td>\n",
          "      <td>0.040300</td>\n",
          "      <td>0.082526</td>\n",
          "      <td>0</td>\n",
          "      <td>0.279710</td>\n",
          "    </tr>\n",
          "    <tr>\n",
          "      <th>2</th>\n",
          "      <td>650</td>\n",
          "      <td>snp1200_m0_.37m1_.36</td>\n",
          "      <td>3.0</td>\n",
          "      <td>2650.0</td>\n",
          "      <td>2650.0</td>\n",
          "      <td>3.048007e-14</td>\n",
          "      <td>0.328870</td>\n",
          "      <td>0.042021</td>\n",
          "      <td>0.054986</td>\n",
          "      <td>0</td>\n",
          "      <td>0.279710</td>\n",
          "    </tr>\n",
          "    <tr>\n",
          "      <th>3</th>\n",
          "      <td>3</td>\n",
          "      <td>snp433_m0_.14m1_.11</td>\n",
          "      <td>3.0</td>\n",
          "      <td>2003.0</td>\n",
          "      <td>2003.0</td>\n",
          "      <td>9.202499e-10</td>\n",
          "      <td>-0.268289</td>\n",
          "      <td>0.042973</td>\n",
          "      <td>0.016289</td>\n",
          "      <td>0</td>\n",
          "      <td>0.279710</td>\n",
          "    </tr>\n",
          "    <tr>\n",
          "      <th>4</th>\n",
          "      <td>274</td>\n",
          "      <td>snp2832_m0_.46m1_.1</td>\n",
          "      <td>4.0</td>\n",
          "      <td>3274.0</td>\n",
          "      <td>3274.0</td>\n",
          "      <td>7.069762e-04</td>\n",
          "      <td>0.170421</td>\n",
          "      <td>0.050003</td>\n",
          "      <td>0.013794</td>\n",
          "      <td>0</td>\n",
          "      <td>0.542046</td>\n",
          "    </tr>\n",
          "    <tr>\n",
          "      <th>5</th>\n",
          "      <td>13</td>\n",
          "      <td>snp1413_m0_.04m1_.03</td>\n",
          "      <td>3.0</td>\n",
          "      <td>2013.0</td>\n",
          "      <td>2013.0</td>\n",
          "      <td>8.161238e-04</td>\n",
          "      <td>-0.148719</td>\n",
          "      <td>0.044157</td>\n",
          "      <td>0.001805</td>\n",
          "      <td>0</td>\n",
          "      <td>0.279710</td>\n",
          "    </tr>\n",
          "    <tr>\n",
          "      <th>6</th>\n",
          "      <td>214</td>\n",
          "      <td>snp2804_m0_.16m1_.3</td>\n",
          "      <td>3.0</td>\n",
          "      <td>2214.0</td>\n",
          "      <td>2214.0</td>\n",
          "      <td>1.239806e-03</td>\n",
          "      <td>0.150705</td>\n",
          "      <td>0.046396</td>\n",
          "      <td>0.008374</td>\n",
          "      <td>0</td>\n",
          "      <td>0.279710</td>\n",
          "    </tr>\n",
          "    <tr>\n",
          "      <th>7</th>\n",
          "      <td>117</td>\n",
          "      <td>snp751_m0_.04m1_.25</td>\n",
          "      <td>1.0</td>\n",
          "      <td>117.0</td>\n",
          "      <td>117.0</td>\n",
          "      <td>1.527432e-03</td>\n",
          "      <td>-0.152430</td>\n",
          "      <td>0.047827</td>\n",
          "      <td>0.006603</td>\n",
          "      <td>0</td>\n",
          "      <td>0.614963</td>\n",
          "    </tr>\n",
          "    <tr>\n",
          "      <th>8</th>\n",
          "      <td>265</td>\n",
          "      <td>snp1440_m0_.35m1_.32</td>\n",
          "      <td>4.0</td>\n",
          "      <td>3265.0</td>\n",
          "      <td>3265.0</td>\n",
          "      <td>1.771049e-03</td>\n",
          "      <td>0.136281</td>\n",
          "      <td>0.043358</td>\n",
          "      <td>0.008860</td>\n",
          "      <td>0</td>\n",
          "      <td>0.542046</td>\n",
          "    </tr>\n",
          "    <tr>\n",
          "      <th>9</th>\n",
          "      <td>307</td>\n",
          "      <td>snp2162_m0_.61m1_.42</td>\n",
          "      <td>2.0</td>\n",
          "      <td>1307.0</td>\n",
          "      <td>1307.0</td>\n",
          "      <td>1.816577e-03</td>\n",
          "      <td>-0.143296</td>\n",
          "      <td>0.045700</td>\n",
          "      <td>0.009973</td>\n",
          "      <td>0</td>\n",
          "      <td>0.534262</td>\n",
          "    </tr>\n",
          "  </tbody>\n",
          "</table>\n",
          "</div>"
         ],
         "text/plain": [
          "   sid_index                   SNP  Chr  GenDist  ChrPos        PValue  SnpWeight  SnpWeightSE  EffectSize  Mixing    Nullh2\n",
          "0         52   snp495_m0_.01m1_.04  5.0   4052.0  4052.0  2.990684e-23   0.418653     0.040052    0.007373       0  0.451117\n",
          "1        392   snp1422_m0_.49m1_.5  3.0   2392.0  2392.0  8.251922e-23  -0.416495     0.040300    0.082526       0  0.279710\n",
          "2        650  snp1200_m0_.37m1_.36  3.0   2650.0  2650.0  3.048007e-14   0.328870     0.042021    0.054986       0  0.279710\n",
          "3          3   snp433_m0_.14m1_.11  3.0   2003.0  2003.0  9.202499e-10  -0.268289     0.042973    0.016289       0  0.279710\n",
          "4        274   snp2832_m0_.46m1_.1  4.0   3274.0  3274.0  7.069762e-04   0.170421     0.050003    0.013794       0  0.542046\n",
          "5         13  snp1413_m0_.04m1_.03  3.0   2013.0  2013.0  8.161238e-04  -0.148719     0.044157    0.001805       0  0.279710\n",
          "6        214   snp2804_m0_.16m1_.3  3.0   2214.0  2214.0  1.239806e-03   0.150705     0.046396    0.008374       0  0.279710\n",
          "7        117   snp751_m0_.04m1_.25  1.0    117.0   117.0  1.527432e-03  -0.152430     0.047827    0.006603       0  0.614963\n",
          "8        265  snp1440_m0_.35m1_.32  4.0   3265.0  3265.0  1.771049e-03   0.136281     0.043358    0.008860       0  0.542046\n",
          "9        307  snp2162_m0_.61m1_.42  2.0   1307.0  1307.0  1.816577e-03  -0.143296     0.045700    0.009973       0  0.534262"
         ]
        },
        "execution_count": 3,
        "metadata": {},
        "output_type": "execute_result"
       },
       {
        "data": {
         "image/png": "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",
         "text/plain": [
          "<Figure size 500x500 with 1 Axes>"
         ]
        },
        "metadata": {},
        "output_type": "display_data"
       }
      ],
      "source": [
       "# import the algorithm\n",
       "import numpy as np\n",
       "from fastlmm.association import single_snp\n",
       "from fastlmm.util import example_file # Download and return local file name\n",
       "\n",
       "# set up data\n",
       "##############################\n",
       "from fastlmm.util import example_file # Download and return local file name\n",
       "bed_fn = example_file('tests/datasets/synth/all.*','*.bed')\n",
       "pheno_fn = example_file(\"tests/datasets/synth/pheno_10_causals.txt\")\n",
       "cov_fn = example_file(\"tests/datasets/synth/cov.txt\")\n",
       "\n",
       "# run gwas\n",
       "###################################################################\n",
       "results_df = single_snp(bed_fn, pheno_fn, covar=cov_fn, count_A1=False)\n",
       "\n",
       "# manhattan plot\n",
       "import pylab\n",
       "import fastlmm.util.util as flutil\n",
       "pylab.rcParams['figure.figsize'] = (10.0, 8.0)#For some reason, need again (appears above too) to get big figures\n",
       "flutil.manhattan_plot(results_df[[\"Chr\", \"ChrPos\", \"PValue\"]].values, pvalue_line=1e-5, xaxis_unit_bp=False)\n",
       "pylab.show()\n",
       "\n",
       "# qq plot\n",
       "from fastlmm.util.stats import plotp\n",
       "plotp.qqplot(results_df[\"PValue\"].values, xlim=[0,5], ylim=[0,5])\n",
       "\n",
       "# print head of results data frame\n",
       "import pandas as pd\n",
       "pd.set_option('display.width', 1000)\n",
       "results_df.head(n=10)"
      ]
     },
     {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
       "For more on advanced features of single_snp see this [supplemental notebook](https://github.com/fastlmm/FaST-LMM/blob/master/doc/ipynb/fastlmm2021.ipynb). Topics covered include multiple phenotypes, filtering large output files, caching intermediate results, and controller muliti-processor runs."
      ]
     },
     {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
       "#### Improving power: LMM(all+select)"
      ]
     },
     {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
       "In [Widmer et al., Scientific Reports 2014](http://www.nature.com/articles/srep06874), we have shown that power can be increased while still maintaining control of type I error by using two GSMs: one based on all SNPs (`G0`) and one based on selected SNPs that are highly correlated with the phenotype (`G1`).  The model is called LMM(select + all).  This approach has greater computational demands (we recommend using a cluster computer when analyzing large data sets). Here is an example of how to apply this model to the synthetic data."
      ]
     },
     {
      "cell_type": "code",
      "execution_count": 3,
      "metadata": {
       "scrolled": true
      },
      "outputs": [
       {
        "data": {
         "image/png": "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",
         "text/plain": [
          "<Figure size 1000x800 with 1 Axes>"
         ]
        },
        "metadata": {},
        "output_type": "display_data"
       },
       {
        "data": {
         "image/png": "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",
         "text/plain": [
          "<Figure size 1000x800 with 1 Axes>"
         ]
        },
        "metadata": {},
        "output_type": "display_data"
       },
       {
        "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>sid_index</th>\n",
          "      <th>SNP</th>\n",
          "      <th>Chr</th>\n",
          "      <th>GenDist</th>\n",
          "      <th>ChrPos</th>\n",
          "      <th>PValue</th>\n",
          "      <th>SnpWeight</th>\n",
          "      <th>SnpWeightSE</th>\n",
          "      <th>EffectSize</th>\n",
          "      <th>Mixing</th>\n",
          "      <th>Nullh2</th>\n",
          "    </tr>\n",
          "  </thead>\n",
          "  <tbody>\n",
          "    <tr>\n",
          "      <th>0</th>\n",
          "      <td>52</td>\n",
          "      <td>snp495_m0_.01m1_.04</td>\n",
          "      <td>5.0</td>\n",
          "      <td>4052.0</td>\n",
          "      <td>4052.0</td>\n",
          "      <td>6.246233e-33</td>\n",
          "      <td>-0.395548</td>\n",
          "      <td>0.030731</td>\n",
          "      <td>0.006581</td>\n",
          "      <td>0.688079</td>\n",
          "      <td>0.545958</td>\n",
          "    </tr>\n",
          "    <tr>\n",
          "      <th>1</th>\n",
          "      <td>307</td>\n",
          "      <td>snp3970_m0_.53m1_.17</td>\n",
          "      <td>5.0</td>\n",
          "      <td>4307.0</td>\n",
          "      <td>4307.0</td>\n",
          "      <td>2.415292e-04</td>\n",
          "      <td>0.149395</td>\n",
          "      <td>0.040400</td>\n",
          "      <td>0.012465</td>\n",
          "      <td>0.688079</td>\n",
          "      <td>0.545958</td>\n",
          "    </tr>\n",
          "    <tr>\n",
          "      <th>2</th>\n",
          "      <td>177</td>\n",
          "      <td>snp3853_m0_.1m1_.04</td>\n",
          "      <td>5.0</td>\n",
          "      <td>4177.0</td>\n",
          "      <td>4177.0</td>\n",
          "      <td>2.162433e-03</td>\n",
          "      <td>-0.109143</td>\n",
          "      <td>0.035401</td>\n",
          "      <td>0.001465</td>\n",
          "      <td>0.688079</td>\n",
          "      <td>0.545958</td>\n",
          "    </tr>\n",
          "    <tr>\n",
          "      <th>3</th>\n",
          "      <td>46</td>\n",
          "      <td>snp2743_m0_.03m1_.1</td>\n",
          "      <td>5.0</td>\n",
          "      <td>4046.0</td>\n",
          "      <td>4046.0</td>\n",
          "      <td>2.178166e-03</td>\n",
          "      <td>0.109598</td>\n",
          "      <td>0.035574</td>\n",
          "      <td>0.001638</td>\n",
          "      <td>0.688079</td>\n",
          "      <td>0.545958</td>\n",
          "    </tr>\n",
          "    <tr>\n",
          "      <th>4</th>\n",
          "      <td>236</td>\n",
          "      <td>snp726_m0_.28m1_.31</td>\n",
          "      <td>5.0</td>\n",
          "      <td>4236.0</td>\n",
          "      <td>4236.0</td>\n",
          "      <td>2.831891e-03</td>\n",
          "      <td>0.104920</td>\n",
          "      <td>0.034969</td>\n",
          "      <td>0.004419</td>\n",
          "      <td>0.688079</td>\n",
          "      <td>0.545958</td>\n",
          "    </tr>\n",
          "  </tbody>\n",
          "</table>\n",
          "</div>"
         ],
         "text/plain": [
          "   sid_index                   SNP  Chr  GenDist  ChrPos        PValue  SnpWeight  SnpWeightSE  EffectSize    Mixing    Nullh2\n",
          "0         52   snp495_m0_.01m1_.04  5.0   4052.0  4052.0  6.246233e-33  -0.395548     0.030731    0.006581  0.688079  0.545958\n",
          "1        307  snp3970_m0_.53m1_.17  5.0   4307.0  4307.0  2.415292e-04   0.149395     0.040400    0.012465  0.688079  0.545958\n",
          "2        177   snp3853_m0_.1m1_.04  5.0   4177.0  4177.0  2.162433e-03  -0.109143     0.035401    0.001465  0.688079  0.545958\n",
          "3         46   snp2743_m0_.03m1_.1  5.0   4046.0  4046.0  2.178166e-03   0.109598     0.035574    0.001638  0.688079  0.545958\n",
          "4        236   snp726_m0_.28m1_.31  5.0   4236.0  4236.0  2.831891e-03   0.104920     0.034969    0.004419  0.688079  0.545958"
         ]
        },
        "execution_count": 3,
        "metadata": {},
        "output_type": "execute_result"
       }
      ],
      "source": [
       "# example for two kernel feature selection\n",
       "# this takes a couple of minutes to run on a 20-proc machine.\n",
       "from pysnptools.snpreader import Bed\n",
       "from fastlmm.association import single_snp_all_plus_select\n",
       "from fastlmm.util import example_file # Download and return local file name\n",
       "\n",
       "\n",
       "from pysnptools.util.mapreduce1.runner import LocalMultiProc\n",
       "import multiprocessing\n",
       "runner = LocalMultiProc(multiprocessing.cpu_count(),mkl_num_threads=4)\n",
       "    \n",
       "# define file names\n",
       "bed_fn = example_file('tests/datasets/synth/all.*','*.bed')\n",
       "snp_reader = Bed(bed_fn, count_A1=True)\n",
       "pheno_fn = example_file(\"tests/datasets/synth/pheno_10_causals.txt\")\n",
       "cov_fn = example_file(\"tests/datasets/synth/cov.txt\")\n",
       "\n",
       "# find the chr5 SNPs\n",
       "test_snps = snp_reader[:,snp_reader.pos[:,0] == 5]\n",
       "\n",
       "#select the 2nd kernel and run GWAS\n",
       "results_df = single_snp_all_plus_select(test_snps=test_snps,G=snp_reader,pheno=pheno_fn,GB_goal=2,do_plot=True,runner=runner)\n",
       "\n",
       "import fastlmm.util.util as flutil\n",
       "flutil.manhattan_plot(results_df[[\"Chr\", \"ChrPos\", \"PValue\"]].values, pvalue_line=1e-5, xaxis_unit_bp=False)\n",
       "pylab.show()\n",
       "\n",
       "results_df.head()"
      ]
     },
     {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
       "Aside: In some applications, you may want to use two kernels constructed from two pre-specified sets of SNPs (i.e., with no feature selection). Here we show how to do that and how to simultaneously find h2 and the mixing weight between the kernels."
      ]
     },
     {
      "cell_type": "code",
      "execution_count": 4,
      "metadata": {},
      "outputs": [
       {
        "data": {
         "image/png": "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",
         "text/plain": [
          "<Figure size 1000x800 with 1 Axes>"
         ]
        },
        "metadata": {},
        "output_type": "display_data"
       },
       {
        "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>sid_index</th>\n",
          "      <th>SNP</th>\n",
          "      <th>Chr</th>\n",
          "      <th>GenDist</th>\n",
          "      <th>ChrPos</th>\n",
          "      <th>PValue</th>\n",
          "      <th>SnpWeight</th>\n",
          "      <th>SnpWeightSE</th>\n",
          "      <th>EffectSize</th>\n",
          "      <th>Mixing</th>\n",
          "      <th>Nullh2</th>\n",
          "    </tr>\n",
          "  </thead>\n",
          "  <tbody>\n",
          "    <tr>\n",
          "      <th>0</th>\n",
          "      <td>52</td>\n",
          "      <td>snp495_m0_.01m1_.04</td>\n",
          "      <td>5.0</td>\n",
          "      <td>4052.0</td>\n",
          "      <td>4052.0</td>\n",
          "      <td>2.990684e-23</td>\n",
          "      <td>-0.418653</td>\n",
          "      <td>0.040052</td>\n",
          "      <td>0.007373</td>\n",
          "      <td>0</td>\n",
          "      <td>0.451117</td>\n",
          "    </tr>\n",
          "    <tr>\n",
          "      <th>1</th>\n",
          "      <td>236</td>\n",
          "      <td>snp726_m0_.28m1_.31</td>\n",
          "      <td>5.0</td>\n",
          "      <td>4236.0</td>\n",
          "      <td>4236.0</td>\n",
          "      <td>2.229561e-03</td>\n",
          "      <td>0.133405</td>\n",
          "      <td>0.043400</td>\n",
          "      <td>0.007144</td>\n",
          "      <td>0</td>\n",
          "      <td>0.451117</td>\n",
          "    </tr>\n",
          "    <tr>\n",
          "      <th>2</th>\n",
          "      <td>785</td>\n",
          "      <td>snp4608_m0_.35m1_.39</td>\n",
          "      <td>5.0</td>\n",
          "      <td>4785.0</td>\n",
          "      <td>4785.0</td>\n",
          "      <td>5.489019e-03</td>\n",
          "      <td>0.122558</td>\n",
          "      <td>0.043942</td>\n",
          "      <td>0.007127</td>\n",
          "      <td>0</td>\n",
          "      <td>0.451117</td>\n",
          "    </tr>\n",
          "    <tr>\n",
          "      <th>3</th>\n",
          "      <td>239</td>\n",
          "      <td>snp2615_m0_.49m1_.25</td>\n",
          "      <td>5.0</td>\n",
          "      <td>4239.0</td>\n",
          "      <td>4239.0</td>\n",
          "      <td>8.029457e-03</td>\n",
          "      <td>-0.123956</td>\n",
          "      <td>0.046572</td>\n",
          "      <td>0.007443</td>\n",
          "      <td>0</td>\n",
          "      <td>0.451117</td>\n",
          "    </tr>\n",
          "    <tr>\n",
          "      <th>4</th>\n",
          "      <td>937</td>\n",
          "      <td>snp1115_m0_.1m1_.1</td>\n",
          "      <td>5.0</td>\n",
          "      <td>4937.0</td>\n",
          "      <td>4937.0</td>\n",
          "      <td>8.030123e-03</td>\n",
          "      <td>0.116014</td>\n",
          "      <td>0.043589</td>\n",
          "      <td>0.002114</td>\n",
          "      <td>0</td>\n",
          "      <td>0.451117</td>\n",
          "    </tr>\n",
          "  </tbody>\n",
          "</table>\n",
          "</div>"
         ],
         "text/plain": [
          "   sid_index                   SNP  Chr  GenDist  ChrPos        PValue  SnpWeight  SnpWeightSE  EffectSize  Mixing    Nullh2\n",
          "0         52   snp495_m0_.01m1_.04  5.0   4052.0  4052.0  2.990684e-23  -0.418653     0.040052    0.007373       0  0.451117\n",
          "1        236   snp726_m0_.28m1_.31  5.0   4236.0  4236.0  2.229561e-03   0.133405     0.043400    0.007144       0  0.451117\n",
          "2        785  snp4608_m0_.35m1_.39  5.0   4785.0  4785.0  5.489019e-03   0.122558     0.043942    0.007127       0  0.451117\n",
          "3        239  snp2615_m0_.49m1_.25  5.0   4239.0  4239.0  8.029457e-03  -0.123956     0.046572    0.007443       0  0.451117\n",
          "4        937    snp1115_m0_.1m1_.1  5.0   4937.0  4937.0  8.030123e-03   0.116014     0.043589    0.002114       0  0.451117"
         ]
        },
        "execution_count": 4,
        "metadata": {},
        "output_type": "execute_result"
       }
      ],
      "source": [
       "# example script for two kernel without feature selection\n",
       "import numpy as np\n",
       "import pysnptools.util\n",
       "from pysnptools.snpreader import Bed\n",
       "from fastlmm.association import single_snp\n",
       "from fastlmm.util import example_file # Download and return local file name\n",
       "\n",
       "\n",
       "# define file names\n",
       "bed_fn = example_file('tests/datasets/synth/all.*','*.bed')\n",
       "snp_reader = Bed(bed_fn, count_A1=True)\n",
       "pheno_fn = example_file(\"tests/datasets/synth/pheno_10_causals.txt\")\n",
       "cov_fn = example_file(\"tests/datasets/synth/cov.txt\")\n",
       "\n",
       "# select data\n",
       "###################################################################\n",
       "snp_reader = Bed(bed_fn,count_A1=True)\n",
       "\n",
       "# partition snps on chr6 vs rest\n",
       "G1_chr = 6\n",
       "G0 = snp_reader[:,snp_reader.pos[:,0] != G1_chr]\n",
       "G1 = snp_reader[:,snp_reader.pos[:,0] == G1_chr]\n",
       "\n",
       "# test on chr5\n",
       "test_snps = snp_reader[:,snp_reader.pos[:,0] == 5]\n",
       "\n",
       "results_df = single_snp(test_snps, pheno_fn, K0=G0, K1=G1, covar=cov_fn, GB_goal=2, count_A1=True)\n",
       "\n",
       "\n",
       "import fastlmm.util.util as flutil\n",
       "flutil.manhattan_plot(results_df[[\"Chr\", \"ChrPos\", \"PValue\"]].values,pvalue_line=1e-5,xaxis_unit_bp=False)\n",
       "pylab.show()\n",
       "results_df.head()"
      ]
     },
     {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
       "#### DEPRECATED: Improving speed and memory use when there is little family structure: LMM(select)+PCs"
      ]
     },
     {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
       "*`compute_auto_pcs` is deprecated because, among other reasons, it no longer runs on AMD processos and so we can't test it. (It does run on Intel processors.*\n",
       "\n",
       "In the same publication, we have shown that a simpler, more computationally efficient model can be used when the data is  confounded only by population structure and not by family structure or cryptic relatedness.  Under these circumstances, we have found that a model with a single GSM based on selected SNPs in combination with principle components as fixed-effect covariates yields good control of type I error and power.  \n",
       "\n",
       "This model, called LMM(select)+PCs, should be used with caution.  Even if you explicitly remove closely related individuals from your data, cryptic relatedness may remain.\n",
       "\n",
       "To use this model, first identify the principle components to be used with the PCgeno algorithm as described in [Widmer et al., Scientific Reports 2014](http://www.nature.com/articles/srep06874). \n",
       "\n",
       "Then you can call single_snp_select with these PCs as covariates."
      ]
     },
     {
      "cell_type": "code",
      "execution_count": 5,
      "metadata": {},
      "outputs": [],
      "source": [
       "# DEPRECATED\n",
       "\n",
       "# from pysnptools.snpreader import Bed\n",
       "# from fastlmm.util import compute_auto_pcs\n",
       "# from fastlmm.association import single_snp_select\n",
       "# from fastlmm.util import example_file # Download and return local file name\n",
       "\n",
       "# # define file names\n",
       "# bed_fn = example_file('tests/datasets/synth/all.*','*.bed')\n",
       "# snp_reader = Bed(bed_fn, count_A1=True)\n",
       "# pheno_fn = example_file(\"tests/datasets/synth/pheno_10_causals.txt\")\n",
       "# cov_fn = example_file(\"tests/datasets/synth/cov.txt\")\n",
       "\n",
       "# # find number of PCs\n",
       "# pcs = compute_auto_pcs(bed_fn,count_A1=True)\n",
       "# print(\"selected number of PCs:\", pcs[\"vals\"].shape[1])\n",
       "\n",
       "# # test on chr5\n",
       "# test_snps = snp_reader[:,snp_reader.pos[:,0] == 5]\n",
       "\n",
       "# results_df = single_snp_select(test_snps=test_snps, G=snp_reader, pheno=pheno_fn, covar=pcs, GB_goal=2)\n",
       "\n",
       "\n",
       "# import fastlmm.util.util as flutil\n",
       "# flutil.manhattan_plot(results_df[[\"Chr\", \"ChrPos\", \"PValue\"]].values,pvalue_line=1e-5,xaxis_unit_bp=False)\n",
       "# pylab.show()\n",
       "\n",
       "# results_df.head()"
      ]
     },
     {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
       "### Epistasis"
      ]
     },
     {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
       "You can test for epistatic interactions between pairs of SNPs as well.  Here is an example analysis applied to the same synthetic data.  Note that this version of the code uses a likelihood ratio test based on maximum-likelihood estimates.  A REML-based version is in the works."
      ]
     },
     {
      "cell_type": "code",
      "execution_count": 6,
      "metadata": {},
      "outputs": [
       {
        "name": "stdout",
        "output_type": "stream",
        "text": [
         "lambda=0.9234\n"
        ]
       },
       {
        "data": {
         "text/html": [
          "<div>\n",
          "<style scoped>\n",
          "    .dataframe tbody tr th:only-of-type {\n",
          "        vertical-align: middle;\n",
          "    }\n",
          "\n",
          "    .dataframe tbody tr th {\n",
          "        vertical-align: top;\n",
          "    }\n",
          "\n",
          "    .dataframe thead th {\n",
          "        text-align: right;\n",
          "    }\n",
          "</style>\n",
          "<table border=\"1\" class=\"dataframe\">\n",
          "  <thead>\n",
          "    <tr style=\"text-align: right;\">\n",
          "      <th></th>\n",
          "      <th>SNP0</th>\n",
          "      <th>Chr0</th>\n",
          "      <th>GenDist0</th>\n",
          "      <th>ChrPos0</th>\n",
          "      <th>SNP1</th>\n",
          "      <th>Chr1</th>\n",
          "      <th>GenDist1</th>\n",
          "      <th>ChrPos1</th>\n",
          "      <th>PValue</th>\n",
          "      <th>NullLogLike</th>\n",
          "      <th>AltLogLike</th>\n",
          "      <th>H2</th>\n",
          "      <th>Beta</th>\n",
          "      <th>Variance_Beta</th>\n",
          "    </tr>\n",
          "  </thead>\n",
          "  <tbody>\n",
          "    <tr>\n",
          "      <th>0</th>\n",
          "      <td>snp2376_m0_.38m1_.49</td>\n",
          "      <td>1.0</td>\n",
          "      <td>20.0</td>\n",
          "      <td>20.0</td>\n",
          "      <td>snp1754_m0_.19m1_.42</td>\n",
          "      <td>1.0</td>\n",
          "      <td>38.0</td>\n",
          "      <td>38.0</td>\n",
          "      <td>0.000451</td>\n",
          "      <td>-700.878718</td>\n",
          "      <td>-694.724583</td>\n",
          "      <td>0.000399</td>\n",
          "      <td>0.151714</td>\n",
          "      <td>7.372408e-07</td>\n",
          "    </tr>\n",
          "    <tr>\n",
          "      <th>1</th>\n",
          "      <td>snp625_m0_.03m1_.07</td>\n",
          "      <td>1.0</td>\n",
          "      <td>NaN</td>\n",
          "      <td>NaN</td>\n",
          "      <td>snp3000_m0_.28m1_.34</td>\n",
          "      <td>1.0</td>\n",
          "      <td>5.0</td>\n",
          "      <td>5.0</td>\n",
          "      <td>0.001306</td>\n",
          "      <td>-700.387891</td>\n",
          "      <td>-695.220652</td>\n",
          "      <td>0.000399</td>\n",
          "      <td>-0.117448</td>\n",
          "      <td>5.272537e-07</td>\n",
          "    </tr>\n",
          "    <tr>\n",
          "      <th>2</th>\n",
          "      <td>snp250_m0_.23m1_.19</td>\n",
          "      <td>1.0</td>\n",
          "      <td>15.0</td>\n",
          "      <td>15.0</td>\n",
          "      <td>snp2377_m0_.06m1_.02</td>\n",
          "      <td>1.0</td>\n",
          "      <td>30.0</td>\n",
          "      <td>30.0</td>\n",
          "      <td>0.002279</td>\n",
          "      <td>-698.768488</td>\n",
          "      <td>-694.113511</td>\n",
          "      <td>0.000399</td>\n",
          "      <td>-0.145412</td>\n",
          "      <td>8.980881e-07</td>\n",
          "    </tr>\n",
          "    <tr>\n",
          "      <th>3</th>\n",
          "      <td>snp625_m0_.03m1_.07</td>\n",
          "      <td>1.0</td>\n",
          "      <td>NaN</td>\n",
          "      <td>NaN</td>\n",
          "      <td>snp376_m0_.17m1_.17</td>\n",
          "      <td>1.0</td>\n",
          "      <td>9.0</td>\n",
          "      <td>9.0</td>\n",
          "      <td>0.002287</td>\n",
          "      <td>-700.206861</td>\n",
          "      <td>-695.555229</td>\n",
          "      <td>0.000399</td>\n",
          "      <td>-0.132940</td>\n",
          "      <td>7.511802e-07</td>\n",
          "    </tr>\n",
          "    <tr>\n",
          "      <th>4</th>\n",
          "      <td>snp125_m0_.64m1_.41</td>\n",
          "      <td>1.0</td>\n",
          "      <td>21.0</td>\n",
          "      <td>21.0</td>\n",
          "      <td>snp1754_m0_.19m1_.42</td>\n",
          "      <td>1.0</td>\n",
          "      <td>38.0</td>\n",
          "      <td>38.0</td>\n",
          "      <td>0.002404</td>\n",
          "      <td>-700.873861</td>\n",
          "      <td>-696.267594</td>\n",
          "      <td>0.000399</td>\n",
          "      <td>-0.131984</td>\n",
          "      <td>7.477779e-07</td>\n",
          "    </tr>\n",
          "    <tr>\n",
          "      <th>5</th>\n",
          "      <td>snp2878_m0_.82m1_.58</td>\n",
          "      <td>1.0</td>\n",
          "      <td>42.0</td>\n",
          "      <td>42.0</td>\n",
          "      <td>snp1878_m0_.58m1_.36</td>\n",
          "      <td>1.0</td>\n",
          "      <td>48.0</td>\n",
          "      <td>48.0</td>\n",
          "      <td>0.002568</td>\n",
          "      <td>-698.579519</td>\n",
          "      <td>-694.033792</td>\n",
          "      <td>0.000399</td>\n",
          "      <td>-0.131828</td>\n",
          "      <td>7.560385e-07</td>\n",
          "    </tr>\n",
          "    <tr>\n",
          "      <th>6</th>\n",
          "      <td>snp1625_m0_.4m1_.47</td>\n",
          "      <td>1.0</td>\n",
          "      <td>23.0</td>\n",
          "      <td>23.0</td>\n",
          "      <td>snp1128_m0_.14m1_.13</td>\n",
          "      <td>1.0</td>\n",
          "      <td>35.0</td>\n",
          "      <td>35.0</td>\n",
          "      <td>0.003067</td>\n",
          "      <td>-700.445456</td>\n",
          "      <td>-696.061998</td>\n",
          "      <td>0.000399</td>\n",
          "      <td>0.129267</td>\n",
          "      <td>7.540972e-07</td>\n",
          "    </tr>\n",
          "    <tr>\n",
          "      <th>7</th>\n",
          "      <td>snp1753_m0_.17m1_.19</td>\n",
          "      <td>1.0</td>\n",
          "      <td>28.0</td>\n",
          "      <td>28.0</td>\n",
          "      <td>snp3_m0_.55m1_.46</td>\n",
          "      <td>1.0</td>\n",
          "      <td>32.0</td>\n",
          "      <td>32.0</td>\n",
          "      <td>0.004327</td>\n",
          "      <td>-700.735073</td>\n",
          "      <td>-696.664365</td>\n",
          "      <td>0.000399</td>\n",
          "      <td>0.126579</td>\n",
          "      <td>7.791090e-07</td>\n",
          "    </tr>\n",
          "    <tr>\n",
          "      <th>8</th>\n",
          "      <td>snp250_m0_.23m1_.19</td>\n",
          "      <td>1.0</td>\n",
          "      <td>15.0</td>\n",
          "      <td>15.0</td>\n",
          "      <td>snp3_m0_.55m1_.46</td>\n",
          "      <td>1.0</td>\n",
          "      <td>32.0</td>\n",
          "      <td>32.0</td>\n",
          "      <td>0.004486</td>\n",
          "      <td>-700.927439</td>\n",
          "      <td>-696.889564</td>\n",
          "      <td>0.000399</td>\n",
          "      <td>-0.120030</td>\n",
          "      <td>7.063126e-07</td>\n",
          "    </tr>\n",
          "    <tr>\n",
          "      <th>9</th>\n",
          "      <td>snp1753_m0_.17m1_.19</td>\n",
          "      <td>1.0</td>\n",
          "      <td>28.0</td>\n",
          "      <td>28.0</td>\n",
          "      <td>snp251_m0_.1m1_.49</td>\n",
          "      <td>1.0</td>\n",
          "      <td>33.0</td>\n",
          "      <td>33.0</td>\n",
          "      <td>0.006503</td>\n",
          "      <td>-700.728933</td>\n",
          "      <td>-697.026244</td>\n",
          "      <td>0.000399</td>\n",
          "      <td>-0.115623</td>\n",
          "      <td>7.152112e-07</td>\n",
          "    </tr>\n",
          "  </tbody>\n",
          "</table>\n",
          "</div>"
         ],
         "text/plain": [
          "                   SNP0  Chr0  GenDist0  ChrPos0                  SNP1  Chr1  GenDist1  ChrPos1    PValue  NullLogLike  AltLogLike        H2      Beta  Variance_Beta\n",
          "0  snp2376_m0_.38m1_.49   1.0      20.0     20.0  snp1754_m0_.19m1_.42   1.0      38.0     38.0  0.000451  -700.878718 -694.724583  0.000399  0.151714   7.372408e-07\n",
          "1   snp625_m0_.03m1_.07   1.0       NaN      NaN  snp3000_m0_.28m1_.34   1.0       5.0      5.0  0.001306  -700.387891 -695.220652  0.000399 -0.117448   5.272537e-07\n",
          "2   snp250_m0_.23m1_.19   1.0      15.0     15.0  snp2377_m0_.06m1_.02   1.0      30.0     30.0  0.002279  -698.768488 -694.113511  0.000399 -0.145412   8.980881e-07\n",
          "3   snp625_m0_.03m1_.07   1.0       NaN      NaN   snp376_m0_.17m1_.17   1.0       9.0      9.0  0.002287  -700.206861 -695.555229  0.000399 -0.132940   7.511802e-07\n",
          "4   snp125_m0_.64m1_.41   1.0      21.0     21.0  snp1754_m0_.19m1_.42   1.0      38.0     38.0  0.002404  -700.873861 -696.267594  0.000399 -0.131984   7.477779e-07\n",
          "5  snp2878_m0_.82m1_.58   1.0      42.0     42.0  snp1878_m0_.58m1_.36   1.0      48.0     48.0  0.002568  -698.579519 -694.033792  0.000399 -0.131828   7.560385e-07\n",
          "6   snp1625_m0_.4m1_.47   1.0      23.0     23.0  snp1128_m0_.14m1_.13   1.0      35.0     35.0  0.003067  -700.445456 -696.061998  0.000399  0.129267   7.540972e-07\n",
          "7  snp1753_m0_.17m1_.19   1.0      28.0     28.0     snp3_m0_.55m1_.46   1.0      32.0     32.0  0.004327  -700.735073 -696.664365  0.000399  0.126579   7.791090e-07\n",
          "8   snp250_m0_.23m1_.19   1.0      15.0     15.0     snp3_m0_.55m1_.46   1.0      32.0     32.0  0.004486  -700.927439 -696.889564  0.000399 -0.120030   7.063126e-07\n",
          "9  snp1753_m0_.17m1_.19   1.0      28.0     28.0    snp251_m0_.1m1_.49   1.0      33.0     33.0  0.006503  -700.728933 -697.026244  0.000399 -0.115623   7.152112e-07"
         ]
        },
        "execution_count": 6,
        "metadata": {},
        "output_type": "execute_result"
       },
       {
        "data": {
         "image/png": "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",
         "text/plain": [
          "<Figure size 500x500 with 1 Axes>"
         ]
        },
        "metadata": {},
        "output_type": "display_data"
       }
      ],
      "source": [
       "# import the algorithm and reader\n",
       "import numpy as np\n",
       "from fastlmm.association import epistasis\n",
       "from pysnptools.snpreader import Bed\n",
       "from fastlmm.util import example_file # Download and return local file name\n",
       "\n",
       "# define file names\n",
       "bed_fn = example_file('tests/datasets/synth/all.*','*.bed')\n",
       "bed_reader = Bed(bed_fn, count_A1=True)\n",
       "pheno_fn = example_file(\"tests/datasets/synth/pheno_10_causals.txt\")\n",
       "cov_fn = example_file(\"tests/datasets/synth/cov.txt\")\n",
       "\n",
       "# partition data into the first 50 SNPs on chr1 and all but chr1\n",
       "G0 = bed_reader[:,bed_reader.pos[:,0] != 1]\n",
       "test_snps = bed_reader[:,bed_reader.pos[:,0] == 1][:,0:50]\n",
       "\n",
       "# run epistasis analysis\n",
       "results_df = epistasis(test_snps, pheno_fn, G0=G0, covar=cov_fn)\n",
       "\n",
       "# qq plot\n",
       "from fastlmm.util.stats import plotp\n",
       "plotp.qqplot(results_df[\"PValue\"].values, xlim=[0,5], ylim=[0,5])\n",
       "\n",
       "# print head of results data frame\n",
       "pd.set_option('display.width', 1000)\n",
       "results_df.head(n=10)"
      ]
     },
     {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
       "### SNP-set association testing"
      ]
     },
     {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
       "SNP-set association testing is performed similarly to single-SNP testing, except we test sets of SNPs by putting them together into one GSM, separate from any (optional) background GSM that corrects for confounding.  Both LRT (\"lrt\") and score (\"sc_davies\") tests are supported.  The LRT is computed by default, but you can switch to the score test by setting test=“sc_davies”.  The score test can be more conservative in some settings (and hence can have less power), but it is closed form and often can be computed much faster.  The algorithms in their current form are described in [Lippert et al., Bioinformatics 2014](http://bioinformatics.oxfordjournals.org/content/30/22/3206).\n",
       "\n",
       "Here is an example that uses the LRT and no background GSM--note the inflation in the results."
      ]
     },
     {
      "cell_type": "code",
      "execution_count": 7,
      "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>SetId</th>\n",
          "      <th>LogLikeAlt</th>\n",
          "      <th>LogLikeNull</th>\n",
          "      <th>P-value</th>\n",
          "      <th>#SNPs_in_Set</th>\n",
          "      <th>#ExcludedSNPs</th>\n",
          "      <th>chrm</th>\n",
          "      <th>pos. range</th>\n",
          "      <th>Alt_h2</th>\n",
          "      <th>Alt_a2</th>\n",
          "    </tr>\n",
          "  </thead>\n",
          "  <tbody>\n",
          "    <tr>\n",
          "      <th>0</th>\n",
          "      <td>set_65</td>\n",
          "      <td>-703.888501</td>\n",
          "      <td>-707.609745</td>\n",
          "      <td>0.002564</td>\n",
          "      <td>10</td>\n",
          "      <td>0</td>\n",
          "      <td>1</td>\n",
          "      <td>650-659</td>\n",
          "      <td>0.029725</td>\n",
          "      <td>0.0</td>\n",
          "    </tr>\n",
          "    <tr>\n",
          "      <th>1</th>\n",
          "      <td>set_86</td>\n",
          "      <td>-704.240753</td>\n",
          "      <td>-707.609745</td>\n",
          "      <td>0.003768</td>\n",
          "      <td>10</td>\n",
          "      <td>0</td>\n",
          "      <td>1</td>\n",
          "      <td>860-869</td>\n",
          "      <td>0.030321</td>\n",
          "      <td>0.0</td>\n",
          "    </tr>\n",
          "    <tr>\n",
          "      <th>2</th>\n",
          "      <td>set_70</td>\n",
          "      <td>-704.525274</td>\n",
          "      <td>-707.609745</td>\n",
          "      <td>0.005158</td>\n",
          "      <td>10</td>\n",
          "      <td>0</td>\n",
          "      <td>1</td>\n",
          "      <td>700-709</td>\n",
          "      <td>0.029790</td>\n",
          "      <td>0.0</td>\n",
          "    </tr>\n",
          "    <tr>\n",
          "      <th>3</th>\n",
          "      <td>set_18</td>\n",
          "      <td>-704.920236</td>\n",
          "      <td>-707.609745</td>\n",
          "      <td>0.008020</td>\n",
          "      <td>10</td>\n",
          "      <td>0</td>\n",
          "      <td>1</td>\n",
          "      <td>180-189</td>\n",
          "      <td>0.024803</td>\n",
          "      <td>0.0</td>\n",
          "    </tr>\n",
          "    <tr>\n",
          "      <th>4</th>\n",
          "      <td>set_45</td>\n",
          "      <td>-705.410762</td>\n",
          "      <td>-707.609745</td>\n",
          "      <td>0.014031</td>\n",
          "      <td>10</td>\n",
          "      <td>0</td>\n",
          "      <td>1</td>\n",
          "      <td>450-459</td>\n",
          "      <td>0.023440</td>\n",
          "      <td>0.0</td>\n",
          "    </tr>\n",
          "    <tr>\n",
          "      <th>5</th>\n",
          "      <td>set_84</td>\n",
          "      <td>-705.507312</td>\n",
          "      <td>-707.609745</td>\n",
          "      <td>0.015693</td>\n",
          "      <td>10</td>\n",
          "      <td>0</td>\n",
          "      <td>1</td>\n",
          "      <td>840-849</td>\n",
          "      <td>0.024002</td>\n",
          "      <td>0.0</td>\n",
          "    </tr>\n",
          "    <tr>\n",
          "      <th>6</th>\n",
          "      <td>set_98</td>\n",
          "      <td>-705.623163</td>\n",
          "      <td>-707.609745</td>\n",
          "      <td>0.017965</td>\n",
          "      <td>10</td>\n",
          "      <td>0</td>\n",
          "      <td>1</td>\n",
          "      <td>980-989</td>\n",
          "      <td>0.022271</td>\n",
          "      <td>0.0</td>\n",
          "    </tr>\n",
          "    <tr>\n",
          "      <th>7</th>\n",
          "      <td>set_80</td>\n",
          "      <td>-705.728130</td>\n",
          "      <td>-707.609745</td>\n",
          "      <td>0.020327</td>\n",
          "      <td>10</td>\n",
          "      <td>0</td>\n",
          "      <td>1</td>\n",
          "      <td>800-809</td>\n",
          "      <td>0.023943</td>\n",
          "      <td>0.0</td>\n",
          "    </tr>\n",
          "    <tr>\n",
          "      <th>8</th>\n",
          "      <td>set_73</td>\n",
          "      <td>-706.000699</td>\n",
          "      <td>-707.609745</td>\n",
          "      <td>0.028153</td>\n",
          "      <td>10</td>\n",
          "      <td>0</td>\n",
          "      <td>1</td>\n",
          "      <td>730-739</td>\n",
          "      <td>0.019835</td>\n",
          "      <td>0.0</td>\n",
          "    </tr>\n",
          "    <tr>\n",
          "      <th>9</th>\n",
          "      <td>set_92</td>\n",
          "      <td>-706.166636</td>\n",
          "      <td>-707.609745</td>\n",
          "      <td>0.034473</td>\n",
          "      <td>10</td>\n",
          "      <td>0</td>\n",
          "      <td>1</td>\n",
          "      <td>920-929</td>\n",
          "      <td>0.019799</td>\n",
          "      <td>0.0</td>\n",
          "    </tr>\n",
          "  </tbody>\n",
          "</table>\n",
          "</div>"
         ],
         "text/plain": [
          "    SetId  LogLikeAlt  LogLikeNull   P-value  #SNPs_in_Set  #ExcludedSNPs  chrm pos. range    Alt_h2  Alt_a2\n",
          "0  set_65 -703.888501  -707.609745  0.002564            10              0     1    650-659  0.029725     0.0\n",
          "1  set_86 -704.240753  -707.609745  0.003768            10              0     1    860-869  0.030321     0.0\n",
          "2  set_70 -704.525274  -707.609745  0.005158            10              0     1    700-709  0.029790     0.0\n",
          "3  set_18 -704.920236  -707.609745  0.008020            10              0     1    180-189  0.024803     0.0\n",
          "4  set_45 -705.410762  -707.609745  0.014031            10              0     1    450-459  0.023440     0.0\n",
          "5  set_84 -705.507312  -707.609745  0.015693            10              0     1    840-849  0.024002     0.0\n",
          "6  set_98 -705.623163  -707.609745  0.017965            10              0     1    980-989  0.022271     0.0\n",
          "7  set_80 -705.728130  -707.609745  0.020327            10              0     1    800-809  0.023943     0.0\n",
          "8  set_73 -706.000699  -707.609745  0.028153            10              0     1    730-739  0.019835     0.0\n",
          "9  set_92 -706.166636  -707.609745  0.034473            10              0     1    920-929  0.019799     0.0"
         ]
        },
        "execution_count": 7,
        "metadata": {},
        "output_type": "execute_result"
       },
       {
        "data": {
         "image/png": "iVBORw0KGgoAAAANSUhEUgAAAb0AAAHACAYAAAA2krZbAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjguNCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8fJSN1AAAACXBIWXMAAA9hAAAPYQGoP6dpAABU50lEQVR4nO3deVzU1f4/8Ncwsi+ybwqCIAgqhOCeaZmW+TNbbbGulnUzrVQ0d1Nc0kzFut1s37u2l91vi9riUkYpAiKiAioIssg6rDPMzPn94Z1JZPsMzjDAvJ6Ph497+cxnPvPmhLw8n/M558iEEAJEREQWwMrcBRAREXUWhh4REVkMhh4REVkMhh4REVkMhh4REVkMhh4REVkMhh4REVkMhh4REVmMXuYu4FpotVpcvHgRzs7OkMlk5i6HiIjMRAiB6upq+Pv7w8qq9f5ctw69ixcvIiAgwNxlEBFRF3HhwgX07du31de7deg5OzsDuPxNuri4NHtdo9EgJycHISEhkMvlnV1et8F2ko5tJQ3bSTq2lTTttZNCoUBAQIA+F1rTrUNPd0vTxcWl1dBzcnKCi4sLf5jawHaSjm0lDdtJOraVNFLbqb2hLrM+yLJ27VrIZLImf3x9fc1ZEhER9WBm7+kNGjQIP/30k/5r/kuHiIhMxeyh16tXL/buiIioU5g99LKysuDv7w9bW1uMGDECzz//PPr379/iuUqlEkqlUv+1QqEAcPler0ajaXa+RqOBVqtt8TX6G9tJOraVNGwn6dhW0rTXTlLbT2bOTWR/+OEH1NXVISwsDMXFxdiwYQNOnTqFjIwMeHh4NDt/7dq1SEhIaHb8yJEjcHJyanZcq9WivLwc7u7ubc7bsHRsJ+nYVtKwnaRjW0nTXjvV1NRg2LBhqKqqavHBRh2zht7VamtrERISgiVLliA+Pr7Z6y319AICAlBeXt7q05vZ2dkIDQ3lWGEb2E7Ssa2kYTtJx7aSpr12UigUcHd3bzf0zH5780qOjo4YMmQIsrKyWnzd1tYWtra2zY7L5fJWf1isrKzafJ0uYztJx7aShu0kHdtKmrbaSWrbdam+tFKpRGZmJvz8/MxdChER9UBmDb3FixfjwIEDOHfuHP7880/cc889UCgUmDlzpjnLIiKiHsqstzfz8/PxwAMPoLS0FF5eXhg5ciSSkpLQr18/c5ZFREQ9lFlD75NPPjHnxxMRkYXpUmN6REREpsTQIyIii8HQIyIii8HQIyIii8HQIyIii8HQIyIii8HQIyIii8HQIyIii8HQIyIii8HQIyIii8HQIyIii8HQIyIii8HQIyIii8HQIyIii8HQIyIii8HQIyIii8HQIyIii8HQIyIii8HQIyIii8HQIyIii8HQIyIii8HQIyIii8HQIyIii8HQIyIii8HQIyIii8HQIyIii8HQIyIii8HQIyIii8HQIyIii8HQIyIii8HQIyIii8HQIyIii8HQIyIii8HQIyIii8HQIyIii8HQIyIii8HQIyIii8HQIyIii8HQIyIii8HQIyIii8HQIyIii8HQIyIii8HQIyIii8HQIyIii8HQIyIii8HQIyIii8HQIyIii8HQIyIii8HQIyIii8HQIyIii8HQIyIii8HQIyIii8HQIyIii8HQIyIii8HQIyIii8HQIyIii8HQIyIii8HQIyIii8HQIyIii8HQIyIii8HQIyIii8HQIyIii8HQIyIii9FlQm/Tpk2QyWRYsGCBuUshIqIeqkuE3pEjR/DGG28gKirK3KUQEVEPZvbQq6mpwYwZM/Dmm2/Czc3N3OUQEVEP1svcBcybNw9TpkzBzTffjA0bNrR5rlKphFKp1H+tUCgAABqNBhqNptn5Go0GWq22xdfob2wn6dhW0rCdpGNbSdNeO0ltP7OG3ieffIJjx47hyJEjks7ftGkTEhISmh3PycmBk5NTs+NarRbl5eXIzs6GlZXZO7VdFttJOraVNGwn6dhW0rTXTjU1NZKuY7bQu3DhAubPn4+9e/fCzs5O0nuWL1+O+Ph4/dcKhQIBAQEICQmBi4tLs/M1Gg2ys7MRGhoKuVxutNp7GraTdGwradhO0rGtpGmvnXR3/tpjttBLTk5GSUkJYmNj9cc0Gg0OHjyIV155BUqlstk3ZmtrC1tb22bXksvlrf6wWFlZtfk6XcZ2ko5tJQ3bSTq2lTRttZPUtjNb6E2YMAHp6elNjj3yyCMYOHAgli5dyv/4RERkdGYLPWdnZwwePLjJMUdHR3h4eDQ7TkREZAwcNSUiIoth9ikLV9q/f7+5SyAioh6MPT0iIrIYDD0iIrIYDD0iIrIYDD0iIrIYDD0iIrIYDD0iIrIYDD0iIrIYDD0iIrIYDD0iIrIYDD0iIrIYDD0iIrIYDD0iIrIYDD0iIrIYDD0iIrIYDD0iIrIYDD0iIrIYDD0iIrIYDD0iIrIYDD0iIrIYDD0iIrIYDD0iIrIYDD0iIrIYDD0iIrIYDD0iIrIYDD0iIrIYDD0iIrIYDD0iIrIYDD0iIrIYDD0iIrIYDD0iIrIYDD0iIrIYDD0iIrIYDD0iIrIYDD0iIrIYDD0iIrIYDD0iIrIYDD0iIrIYDD0iIrIYDD0iIrIYDD0iIrIYvaScdNddd0m+4FdffdXhYoiIiExJUk+vd+/e+j8uLi74+eefcfToUf3rycnJ+Pnnn9G7d2+TFUpERHStJPX03n33Xf3/X7p0KaZPn47XXnsNcrkcAKDRaDB37ly4uLiYpkoiIiIjMHhM75133sHixYv1gQcAcrkc8fHxeOedd4xaHBERkTEZHHpqtRqZmZnNjmdmZkKr1RqlKCIiIlOQdHvzSo888ggeffRRZGdnY+TIkQCApKQkbN68GY888ojRCyQiIjIWg0Nv69at8PX1RWJiIgoLCwEAfn5+WLJkCRYtWmT0AomIiIzF4NCzsrLCkiVLsGTJEigUCgDgAyxERNQtdGhyulqtxk8//YRdu3ZBJpMBAC5evIiamhqjFkdERGRMBvf0cnNzceuttyIvLw9KpRITJ06Es7MztmzZgoaGBrz22mumqJOIiOiaGdzTmz9/PuLi4lBRUQF7e3v98TvvvBM///yzUYsjIiIyJoN7er/99ht+//132NjYNDner18/FBQUGK0wIiIiYzO4p6fVaqHRaJodz8/Ph7Ozs1GKIiIiMgWDQ2/ixInYsWOH/muZTIaamhqsWbMGt912mzFrIyIiMiqDb28mJibixhtvRGRkJBoaGvDggw8iKysLnp6e2LVrlylqJCIiMgqDQ8/f3x+pqanYtWsXjh07Bq1Wi9mzZ2PGjBlNHmwhIiLqagwOvbq6Ojg4OODRRx/Fo48+aoqaiIiITMLgMT1vb2889NBD2LNnDxeYJiKibsXg0Pvggw+gVCpx5513wt/fH/Pnz8eRI0dMURsREZFRGRx6d911Fz7//HMUFxdj06ZNyMzMxOjRoxEWFoZ169aZokYiIiKj6NDamwDg7OyMRx55BHv37kVaWhocHR2RkJBgzNqIiIiMqsOh19DQgM8++wx33HEHhg4dirKyMixevNiYtRERERmVwU9v7t27Fx9//DG++eYbyOVy3HPPPdizZw/GjRtnivqIiIiMxuCe3h133IH6+nq8//77KC4uxhtvvNHhwNu5cyeioqLg4uICFxcXjBo1Cj/88EOHrkVERNQeg3p6arUamzdvxr333gs/P79r/vC+ffti8+bNCA0NBQC8//77mDZtGlJSUjBo0KBrvj4REdGVDOrp9erVC8uWLYNKpTLKh0+dOhW33XYbwsLCEBYWho0bN8LJyQlJSUlGuT4REdGVDB7TGzFiBFJSUtCvXz+jFqLRaPD555+jtrYWo0aNavEcpVIJpVKp/1qhUOjf29LODxqNptVdIehvbCfp2FbSsJ2kY1tJ0147SW0/g0Nv7ty5WLRoEfLz8xEbGwtHR8cmr0dFRRl0vfT0dIwaNQoNDQ1wcnLC119/jcjIyBbP3bRpU4vTInJycuDk5NTsuFarRXl5ObKzs2Fl1eEHVXs8tpN0bCtp2E7Ssa2kaa+dampqJF1HJoQQhnxwSx8mk8kghIBMJjP4XysqlQp5eXmorKzEl19+ibfeegsHDhxoMfha6ukFBASgvLwcLi4uzc7XaDTIzs5GaGgo5HK5QXVZEraTdGwradhO0rGtpGmvnRQKBdzd3VFVVdViHugY3NM7d+6coW9pk42Njf5Blri4OBw5cgQvvfQSXn/99Wbn2trawtbWttlxuVze6g+LlZVVm6/TZWwn6dhW0rCdpGNbSdNWO0ltO4NDz9hjeVcTQjTpzRERERlLh24gf/jhhxgzZgz8/f2Rm5sLANixYwd2795t0HVWrFiBQ4cO4fz580hPT8fKlSuxf/9+zJgxoyNlERERtcng0Nu5cyfi4+Nx2223obKyUj+G5+rqih07dhh0reLiYjz88MMIDw/HhAkT8Oeff+LHH3/ExIkTDS2LiIioXQbf3vzXv/6FN998E3fccQc2b96sPx4XF2fw2ptvv/22oR9PRETUYQb39M6dO4eYmJhmx21tbVFbW2uUooiIiEzB4NALDg5Gampqs+M//PBDq/PriIiIugKDb28+++yzmDdvHhoaGiCEwF9//YVdu3Zh06ZNeOutt0xRIxERkVEYHHqPPPII1Go1lixZgrq6Ojz44IPo06cPXnrpJdx///2mqJGIiMgoDA49AHj88cfx+OOPo7S0FFqtFt7e3saui4iIyOgMHtOrr69HXV0dAMDT0xP19fXYsWMH9u7da/TiiIiIjMng0Js2bRo++OADAEBlZSWGDx+Obdu2Ydq0adi5c6fRCyQiIjIWg0Pv2LFjGDt2LADgiy++gK+vL3Jzc/HBBx/g5ZdfNnqBRERExmJw6NXV1cHZ2RkAsHfvXtx1112wsrLCyJEj9UuSERERdUUGh15oaCi++eYbXLhwAXv27MGkSZMAACUlJW1u50BERGRuBofec889h8WLFyMoKAjDhw/X73K+d+/eFldqISIi6ioMnrJwzz334Prrr0dhYSGio6P1xydMmIA777zTqMUREREZU4fm6fn6+sLX1xcXLlyATCZD3759MXz4cGPXRkREZFQG395Uq9VYvXo1evfujaCgIPTr1w+9e/fGqlWr0NjYaIoaiYiIjMLgnt5TTz2Fr7/+Glu2bNGP5/3xxx9Yu3YtSktL8dprrxm9SCIiImMwOPR27dqFTz75BJMnT9Yfi4qKQmBgIO6//36GHhERdVkG3960s7NDUFBQs+NBQUGwsbExRk1EREQmYXDozZs3D+vXr4dSqdQfUyqV2LhxI5566imjFkdERGRMkm5v3nXXXU2+/umnn9C3b1/9lIW0tDSoVCpMmDDB+BUSEREZiaTQ6927d5Ov77777iZfBwQEGK8iIiIiE5EUeu+++66p6yAiIjK5Dk1OB4BLly7h9OnTkMlkCAsLg5eXlzHrIiIiMjqDH2Spra3Fo48+Cj8/P9xwww0YO3Ys/P39MXv2bP3mskRkmI+ScjFm8y/4KIk7lRCZksGhFx8fjwMHDuC///0vKisrUVlZid27d+PAgQNYtGiRKWok6vF27s9BQWU9du7PMXcpRD2awaH35Zdf4u2338bkyZPh4uICFxcX3HbbbXjzzTfxxRdfmKJGoh7vyfEh6ONqjyfHh5i7FKIezeAxvbq6Ovj4+DQ77u3tzdubRB3Q0NCAGwN6YdidfdC/v7+5yyHq0QwOvVGjRmHNmjX44IMPYGdnBwCor69HQkKCfi1OImqdVquFQqFAWVkZSkpKUFpaivr6elhbW8PLywvu7u7mLpGoxzI49F566SXceuut+snpMpkMqampsLOzw549e0xRI1G3p9FoUFpaitLSUhQWFqK6uhoqlQpyuRz29vZwdXVFdXW1ucsk6vEMDr3BgwcjKysLH330EU6dOgUhBO6//37MmDED9vb2pqiRqFvSaDQoLy9HSUkJ8vPzUV1dDY1GA1tbWzg4OMDV1RUymQzA5S27iMj0OjRPz97eHo8//rixayHq9rRaLSoqKlBcXKwPOrVaDTs7O7i6uqJXrw5PjSUiI+DfQCIjUKvVyM/Px9mzZ1FZWYnGxkbY2trCxcUF1tbW5i6PiP6HoUd0DVQqFfLz85GTk4OKigrI5XI4Oztzmy2iLoqhR9QBKpUKBQUF+p5dr1694OHhAblcbu7SiKgNDD0iA6hUKpSWluLs2bOorq6GjY0Nw46oG7mm0FMqlbC1tTVWLURdWnFxMdLT01FVVYVevXrBy8sLVlYGL2pERGZk0N/YPXv2YNasWQgJCYG1tTUcHBzg7OyMcePGYePGjbh48aKp6iQyOqmLPDc2NiIjIwN//PEHKioqYG9vDxcXFwYeUTck6W/tN998g/DwcMycORNWVlZ49tln8dVXX2HPnj14++23MW7cOPz000/o378/5syZg0uXLpm6bqJrJmWR5/Lychw+fBgnTpxAr1694OnpybAj6sYk3d58/vnnsXXrVkyZMqXFv/DTp08HABQUFOCll17CBx98wB0XqMt7cnwIdu7PaXGRZ41Gg+zsbJw+fRoNDQ3w8PBAr169IIQwQ6VEZCySQu+vv/6SdLE+ffpgy5Yt11QQUWd5aGQ/PDSyX7PjCoUC6enpKCgogJ2dHby8vPQrpxBR92bQgyzV1dVISkpCY2Mjhg8fDk9PT1PVRWQWly5dQnJyMqqqquDu7s6J5UQ9jOTQO378OCZPnoyioiIIIeDi4oIvvvgCN998synrI+o0Fy9exLFjx9DQ0ABvb2/27oh6IMkj8suWLUNgYCAOHTqEo0ePYty4cXjqqadMWRtRpxBCIDc3F0eOHIFSqYSHhwcDj6iHktzTO3r0KL7//nvExcUBAN555x14e3ujpqYGTk5OJiuQyJSEEMjJyUF6ejpkMhnc3d0ZeEQ9mOSeXmlpKQIDA/Vfe3h4wMHBgdMTqNsSQuDUqVNITU2FXC5vstVPZysrK0NubtvzBYno2knu6clkMlRXV+t3SxdC6I8pFAr9eS4uLsavksjINBoNMjIycPr0adjb25v1bkVpaSkWL14MhUKB4cOH4/rrrzdbLUQ9neTQE0IgLCys2bGYmBj9/5fJZNBoNMatkKgFHyXl6ufYtTTtoC1qtRppaWnIycmBk5MTHBwcTFRl+y5duoT4+Hjk5+fDy8uLQwVEJiY59H799VdT1kFkkCtXUzEk9LRaLY4fP47s7Gz07t1bf+fCHC5duoSFCxeioKAAvr6+eO6555oMIRCR8UkOvXHjxpmyDiKDtLWaSmt0Y3hdIfBKSkqwcOFCXLx4EX5+fnjxxRfNWg+RpZAUerW1tXB0dJR8UUPPJzJUa6uptOXs2bPIzMyEo6OjWQOmuLgY8fHxuHjxIvz9/bF9+3Z4eHigqqrKbDURWQpJT2+Ghobi+eefb3MXBSEE9u3bh8mTJ+Pll182WoFExpCfn4/09HRYW1ub9R9kxcXF+h6ev78/EhMT4ePjY7Z6iCyNpJ7e/v37sWrVKiQkJOC6665DXFwc/P39YWdnh4qKCpw8eRJ//PEHrK2tsXz5cvzzn/80dd1EkpWUlCA1NRVarRaurq5mq6OoqAjx8fEoLCyEv78/duzYAS8vL7PVQ2SJJIVeeHg4Pv/8c+Tn5+Pzzz/HwYMHcfjwYdTX18PT0xMxMTF48803cdttt3HbFepSKisr9UuLeXh4mK2OoqIiLFy4EEVFRejTpw8SExMZeERmYNCC03379sXChQuxcOFCU9VDZDS1tbU4evQoFAqFWXdKKCoqwoIFC1BcXIy+ffti+/btDDwiM2G3jHokpVKJo0ePoqysDJ6enmYLvMLCQn3gBQQEsIdHZGYG9fTy8/Oxc+dOHD58GEVFRZDJZPDx8cHo0aMxZ84cBAQEmKpOIsk0Gg1SUlJQWFho1p3OL168iIULF6KkpAQBAQHYvn07t+MiMjPJvw1+++03RERE4Ouvv0Z0dDT+8Y9/4KGHHkJ0dDS++eYbDBo0CL///rspayVqlxACp0+fRl5eHtzd3dGrl0H/rjOagoICLFiwACUlJQgMDERiYiIDj6gLkPwbYeHChXjssceQmJjY6usLFizAkSNHjFYckaHy8/Nx6tQpODo6wsbGxiw1FBQUYOHChbh06RL69euH7du3w93d3Sy1EFFTknt6J06cwJw5c1p9/YknnsCJEyeMUhRRR1RWViI9PR0AzDYXLz8/HwsWLGDgEXVRkkPPz88Phw8fbvX1P/74A35+fkYpishQDQ0NSElJQXV1Ndzc3MxSw4ULF7Bw4UKUlpYiKCiIgUfUBUm+vbl48WLMmTMHycnJmDhxInx8fCCTyVBUVIR9+/bhrbfewo4dO0xYKvV0Hd05QavVIj09HSUlJWZ7UjMvLw/x8fEoKytDcHAwtm3bZrbwJaLWSQ69uXPnwsPDA4mJiXj99df1WwjJ5XLExsbigw8+wPTp001WKPV8Hd054cyZMzh37hxcXV0hl8tNWGHLrgy8/v37Y9u2bWZd+YWIWmfQo2333Xcf7rvvPjQ2NqK0tBQA4OnpCWtra5MUR5alIzsnXLx4EZmZmbC3t4etra0Jq2tZXl4eFi5ciPLycgYeUTfQoee5ra2tOX5HRmfozgkKhQJpaWnQarVwdnY2YWUty83NxcKFC1FRUYGQkBBs27YNvXv37vQ6iEg6o83azcnJwU033WTQezZt2oRhw4bB2dkZ3t7euOOOO3D69GljlUQ9mEqlQkpKChQKhVkeFjl//rw+8EJDQxl4RN2E0UKvpqYGBw4cMOg9Bw4cwLx585CUlIR9+/ZBrVZj0qRJqK2tNVZZ1AMJIZCRkYGioiJ4eHh0+oMrubm5iI+PR0VFBQYMGICtW7cy8Ii6Ccm3N9vbI6+goMDgD//xxx+bfP3uu+/C29sbycnJuOGGGwy+HlmGvLw8nD17Fi4uLp2+4sq5c+ewdOlSVFVV6QPPxcWlU2sgoo6T/BtjwYIF8PPza3WVC5VKdc3F6HaObu12lVKphFKp1H+tUCgAXF5rUfc06ZU0Gg20Wm2Lr9HfulM7KRQKpKenw8rKCnZ2dhBCdNpnnz17FosWLUJVVRXCwsKwZcsWODs7G6UG3TW6y3+H9nSnnylzY1tJ0147SW0/mZD4NzY4OBgvvPBCq9MSUlNTERsb2+H/cEIITJs2DRUVFTh06FCL56xduxYJCQnNjh85cgROTk7Njmu1WpSXl8Pd3Z37/LWhu7STRqPBhQsXUFtbCwcHh069rXnu3DksW7YMCoUCoaGh2Lhxo1F7eFqtFkqlEkFBQbC3tzfadc2lu/xMdQVsK2naa6eamhoMGzYMVVVVbf7dlNzTi42NRXJycquhJ5PJrulfvE899RSOHz+O3377rdVzli9fjvj4eP3XCoUCAQEBCAkJafGb1Gg0yM7ORmhoqFnmb3UX3aGdhBA4ceIEqqqqOn0h6ZycHH3ghYeHY/369UYfS1Sr1aivr0dAQECPmNTeHX6mugq2lTTttZPuzl97JP/mWLduHerq6lp9PTIyEufOnZN6uSaefvppfPvttzh48CD69u3b6nm2trYtzsWSy+Wt/rBYWVm1+Tpd1tXb6cKFC8jJyYGzs3OnzgvNzs7GokWLoFAoMHDgQGzZsgXW1taQyWRGDT3dtXT/HXqCrv4z1ZWwraRpq52ktp3k0IuMjGzzdWtra/TrJ32OFXD5X+9PP/00vv76a+zfvx/BwcEGvZ8sg24cTyaTwcHBodM+NysrC4sWLUJ1dTUiIiKwZcsWODo6NhlXJqLuxTybjf3PvHnz8J///Ae7d++Gs7MzioqKAAC9e/fuEeMadO0aGxuRlpaG6upqeHt7d9rnnjlzBosXL0Z1dTUiIyP1gdeZD84QkfEZHHoxMTEt3taRyWSws7NDaGgoZs2ahRtvvLHda+3cuRMAMH78+CbH3333XcyaNcvQ0qiHEUIgMzMTFy9e7NT5eKdPn8bixYtRU1ODQYMG4YUXXjDbVkVEZFwGPyp066234uzZs3B0dMSNN96I8ePHw8nJCTk5ORg2bBgKCwtx8803Y/fu3e1eSwjR4h8GHgGX535mZ2fD2dm50x5cuTLwBg8erO/hEVHPYPBvktLSUixatAirV69ucnzDhg3Izc3F3r17sWbNGqxfvx7Tpk0zWqFkWWpqapCeng4hRKeN4506dQqLFy9GbW0thgwZgs2bN3fqGCIRmZ7BPb3PPvsMDzzwQLPj999/Pz777DMAwAMPPMA1NKnDNBoNjh8/3qnrap48eVIfeFFRUQw8oh7K4NCzs7NrcQf1w4cPw87ODsDlSYTm2OaFeoasrCxcuHAB7u7unTKOd/LkSSxZsoSBR2QBDL69+fTTT+t3UB82bBhkMhn++usvvPXWW1ixYgUAYM+ePYiJiTF6sdS9SdkZvaSkBKdOnYKDg0OnzMfLyMjAkiVLUFdXh+joaGzatIlPDhP1YAaH3qpVqxAcHIxXXnkFH374IQAgPDwcb775Jh588EEAwJw5c/Dkk08at1Lq9trbGb2+vh5paWlobGzslF0LTpw4gaVLl6Kurg7XXXcdnn/+eQYeUQ/XoUfiZsyYgRkzZrT6On9xUEva2hldq9XixIkTKC8vh5eXl8lva6anp2Pp0qWor69HTEwMnn/+ef3teSLquTr8HHhycjIyMzMhk8kQGRnJ25nUrrZ2Rj9//jzOnz8PV1dXky+6y8AjslwGh15JSQnuv/9+7N+/H66urhBCoKqqCjfeeCM++eQTeHl5maJO6iakjNtdraKiAidPnoSNjY3JH4A6fvw4li5dioaGBsTGxmLDhg0MPCILYvA/qZ9++mkoFApkZGSgvLwcFRUVOHHiBBQKBZ555hlT1EjdyJXjdlKoVCocP34cdXV1Jt+MNS0tTR94cXFx2LhxIwOPyMIYHHo//vgjdu7ciYiICP2xyMhI/Pvf/8YPP/xg1OKo+3lyfAj6uNq3OG53Nd0yY0VFRSafnpCamoply5bpA2/Dhg2cVkNkgQy+vanValt8lNza2hpardYoRVH31da43dU6a5mxlJQUrFixAg0NDRg+fDjWr18PGxsbk30eEXVdBvf0brrpJsyfPx8XL17UHysoKMDChQsxYcIEoxZHXdtHSbkYs/kXfJSUa/B7q6urceLECZNvF3Ts2DEsX74cDQ0NGDFiBAOPyMIZHHqvvPIKqqurERQUhJCQEISGhiI4OBjV1dX417/+ZYoaqYsydPxOR7fMWFVVlUl3CU9OTsaKFSugVCoxYsQIrFu3joFHZOEMvqcUEBCAY8eOYd++fTh16hSEEIiMjMTNN99sivqoC2tr3l1bsrKykJ+fb9JxPF3gqVQqjBw5EgkJCQw8Iur4PL2JEydi4sSJxqyFuhlDxu90OmOZsSNHjmDVqlVQqVQYPXo01qxZw8AjIgASQ+/ll1+WfEFOW+iZOjL/7mqdsczYX3/9hVWrVqGxsZGBR0TNSAq9xMRESReTyWQMvR6qvXUz26NbZqysrAze3t4mua15ZeCNGTMGa9as6ZRFq4mo+5AUeufOnTN1HdTFdXT8Tke3zJibm5tJlhn7888/sXr1ajQ2NmLs2LFYvXo1A4+ImrmmyVG///474uLiOMnXAnRk/E5Ht8yYtbW1SX5WkpKS8Nxzz+kD77nnnjPpvD9TUCqV5i6ByCJc0z+5J0+ejIKCAmPVQj1QY2Mjjh8/jtraWpOM4/3xxx/6wLvhhhu6XeA1NjaipKQESqUSQUFBnbKlEpElu6bfDkIIY9VBPZAQAidPnkRRURE8PDyMPo53+PBhrFmzBmq1GuPGjcOqVau6TeBptVpUVVVBrVbDx8cHERERnbKlEpGl6x6/IahbMuUyY7///jvWrl0LtVqN8ePHY+XKld0i8IQQqK2tRU1NDVxdXREeHo7AwEDI5XJzl0ZkEa7pt8Trr78OHx8fY9VCPUh1dTXS09NNsszYb7/9hoSEBKjVatx0001YsWJFtwgNpVKJyspK2NnZYdCgQQgNDeWGy0Sd7JpC78EHHzRWHdSDqNVqpKWlQaFQwNvb26jXPnToEBISEqDRaLpN4Gk0GlRUVAAAAgMDMXDgQJMuv0ZErev694OoWxFC4MyZM7h48aLRlxk7ePAg1q1bB41GgwkTJmD58uVdOvCEEFAoFGhoaICnpycGDhwIf39/jtsRmRFDj4yqqKgIp0+fhqOjo1HnyR04cADr1q2DVqvFzTffjGXLlnXpwKurq0N1dTWcnJwwdOhQBAUFcd4gURfA0KMWdWTZsdraWqSnp0Oj0cDR0dFotezfvx/r16+HVqvFxIkTsXTp0i4beI2NjaioqIC1tTUGDBiAsLAwODk5mbssIvofhh61yNBlxzQaDdLT01FRUQEvLy+j1fHrr79iw4YN0Gq1mDRpEpYsWdIlA0+r1aKyshIajQZ+fn4YOHAgPD09eSuTqIth6FGLDF12LDs7G3l5eUZdZuyXX37Bxo0bodVqceutt2Lx4sVdLvCEEKipqUFdXR3c3NwQHh6OgIAAkyy1RkTXjqFHLTJk2bFLly7h1KlTsLe3N9qOBj///DOef/75Lh14DQ0NqKqqgr29PYYMGYKQkBAuyUfUxTH06JrotgtSqVTw9PQ0yjV/+uknbNq0CVqtFrfddhsWLVrUpXpOWq0WZWVlkMlkCAoKwsCBA7l8GFE3wdCjDtNqtUhPT0dZWZnRltDat28fNm/eDK1WiylTpiA+Pr7LBJ5uCoJarYabmxsiIiLg5+fHcTuiboShZ+EuP6WZjbsGOsG7PA+vHTwn+YnNs2fPIjc3F66urkYJpr1792Lz5s0QQnS5wKuvr4dCoYCTkxO8vLwQGxvLW5lE3RBDz8JdfkqzAZ+mq2FtXYOCygZJT2yWlpYiIyMDtra2Rvnl/+OPP2LLli0QQmDq1KlYsGBBlwi8q6cgDBgwAAUFBd1inU8iao5/cy3c5ac0/9fT8/bW9/TaUl9fj9TUVKON410ZeLfffjvmz59v9sDT7YLQ2NjYZBcErVZr1rqI6Now9CzcQyP74YFhfZGVlYUBAwLx8OjgNs839jje999/j61bt0IIgWnTpmH+/PlmHSMTQuhXU+nduzeio6PRr1+/LvfkKBF1DEOPDGLMcbzvvvsO27ZtgxACd955J55++mmzBp7uVqaNjQ0iIiIQFhbGXRCIehiGHklmzHG8//u//8O2bdsAwOyBd+WtTD8/P0RERBht+gURdS3mf1KAOt1HSbkYs/kXfJSUK/k9V47jOTs7X9Pn//e//9UH3t13323WwKurq8OlS5dgZ2eHYcOGYfTo0Qw8oh6MoWeBrlxXU4orx/E8PDyuKaC+/fZbbN++HQBwzz33YN68eWYJPLVajZKSEiiVSoSFhWHcuHEIDg7m2B1RD8fbmxbI0HU1jTWOt3v3buzYsQMAcO+99+LJJ5/s9MATQqCqqgoqlarJU5mcYE5kGRh6FsiQdTVLSkqMMo739ddf4+WXXwYATJ8+HXPmzOn0oNGtlenk5IQhQ4agX79+nG9HZGH4N95CdGR/vLq6OqOsq3ll4N1333144oknOjXwNBoNKioqIJPJ0L9/f0RERHCPOyILxdCzEB3ZHy8tLQ1lZWXw9vbucEh9+eWXeOWVVwAADzzwAB5//PFOCzwhBGpra1FbWwsPDw9ERETA39+ftzKJLBgfZLEQT44PQR9Xe0njeEIInDlzBnl5eXB3d+/wON4XX3yhD7wHH3ywUwOvsbERly5dgkajweDBg3HDDTegT58+DDwiC8eenoUwZByvqKgIp06dgqOjY4f3x/v888/x6quvAgBmzJiB2bNnd0rgXPmgir+/PyIjI+Hu7m7yzyWi7oGhR01UV1cjLS0NWq22w+Nen332GXbu3AkAePjhh/HII490SuAplUpUVlbCyckJUVFRXD6MiJph6JGeWq1GWloaqqqq4O3t3aFrfPrpp3jttdcAdF7gabVaVFZWQqvVIigoCJGRkdc8gZ6IeiaGHgG4fFvw1KlTKCgogLu7e4eC6pNPPsHrr78OAJg5cyZmzZpl5Cqbq6+vR1VVFdzc3BAZGYm+ffty3I6IWsXQ68EMmaagUCiQn58PZ2dnWFtbG/xZu3btwhtvvAEAmDVrFmbOnNmhmqXSTUOwsrJCeHg4Bg4cyMWhiahdDL0eSBd2tUo1Kusb252mUFVVheLiYlhZWcHBwcHgz/v444/x1ltvAeicwKutrUVNTQ08PT0xaNAg+Pj4sHdHRJIw9HqQq8PO1d663WkKSqUSqampaGxshJubm+Gf+dFHePvttwEAjz76KB5++OEO198ejUaD8vJy9OrVCxERERg4cGCHny4lIsvE0OtBdBPQrwy7tnp4Wq0Wx48fR0lJCVxcXAzuLX344Yd45513AACzZ8/GQw89dE31t0a3sWtNTQ28vLwwePBgrpdJRB3C0OtBrlxIWsqcvOzsbJw7dw6urq4Gf9b777+P9957DwDw2GOPYcaMGQZfQwqNRoOysjLY2Nhg0KBBCAsLY++OiDqModcDXPnAyu/LbpL0nqKiIpw8eRL29vawtbWFUqmU/HlXBt4///lPPPDAAx0pu01XLiHm7e2t790REV0LLkPWAxi6P15NTQ3S0tKgVqsNmoAuhMC7775r8sDTaDQoLS2FWq3GkCFDcP311zPwiMgoGHo9gCHrajY2NiI1NRWVlZUGzcfTBd4HH3wAAJgzZ45JAq+2thaXLl2Ch4cHxowZg8jIyA5NoSAiaglvb/YAUtfVFELg5MmT+gnoUheS1gXehx9+CAB48sknMX369Guq+Wq6eXdyuRyRkZF8MpOITII9vW7so6RcjNn8Cz5KypV0fm5uLrKzs+Hi4iK59ySEwNtvv60PvLlz5xo98Orr61FaWgpXV1eMGjUKQ4YMYeARkUkw9LoxQ8byysrKcOLECcjlcskrlwgh8NZbb+Hjjz8GAMybNw/33nvvNdV8Ja1Wi9LSUtTX1yMsLAxjx46Fr68vpyIQkckw9LoxqWN59fX1SElJQX19PXr37i3p2kIIvPnmm/jPf/4DAHjqqadwzz33XHPNOg0NDfr5gSNHjsR1110HW1tbo12fiKglHNPrZj5KysXWPacBAItvCW93ioJarUZqairKysokT+gWQuD111/Hp59+CgB45plncOedd1578f+7dmVlJdRqNUJCQjB48GCumUlEnYY9vW5m5/4cVNY36tfUbIsQApmZmcjLy4Obm5ukB1dMGXiNjY0oKSmBra0thg0bhri4OAYeEXUqs4bewYMHMXXqVPj7+0Mmk+Gbb74xZzldmu6hldh+bnC1t4arvXW7tzXz8vJw5swZODs7S3owRAiBnTt36gNvwYIFRgk8IQSqq6tRXl6OPn364Prrr0dQUBDH7oio05n19mZtbS2io6PxyCOP4O677zZnKV3SlbcyAaCyvhEAkLpmUrvvLS8vR3p6uuSdE4QQePXVV/Hll18CABYuXIjbb7+9g5X/TbdItI2NDaKjozFgwADuZk5EZmPW0Js8eTImT55szhK6NN2tTACSdkzQqa+vx7Fjx1BfXw9PT892z9fd0ty9ezcAID4+HlOnTr224v9Xh0KhgLe3N4YMGSKpFiIiU+pWD7Iolcoma0QqFAoAl3sTGo2m2fkajQZarbbF17qDOTcEY+u+LMggsGjiADw4IhAA2vx+NBoNUlJSUFZWpg8ZIUSr5wsh8MorrzQJvP/3//5fm+9pj+5hFSEEwsLC9BPNu+t/hyt195+pzsJ2ko5tJU177SS1/bpV6G3atAkJCQnNjufk5LS4hqRWq0V5eTmys7Mlrz7SlQxzBz69L/B/XymRlZXV5vlCCJSUlKC0tBROTk5obGxs9/ydO3fiv//9L2QyGZ555hlMnDjRoMWnr6bRaNDQ0AAbGxt4e3vDzs4OubnSJs93B939Z6qzsJ2kY1tJ01471dTUSLpOtwq95cuXIz4+Xv+1QqFAQEAAQkJC4OLi0ux8jUaD7OxshIaGWsQ40oULF1BaWgpbW9t2x/GEEHj55Zf1gTd//nxMnTq1ww+XCCFQU1ODhoYG9OnTB4MHDzZoMevuwtJ+pjqK7SQd20qa9tpJd+evPd0q9GxtbVucwCyXy1v9YbGysmrz9a7m6nl4UtbUBP5eccXKyqrdsNFqtXjppZfw7bffQiaT4dlnn8WNN94ImUzWodC78mGVqKioHv+wSnf7mTIXtpN0bCtp2monqW3HvnQX8lFSLtbsPiF5Hp5OXV0djh07hrq6unY3hNVqtdixY4c+8JYsWYJbb721wzU3NDSgtLQUbm5uGD16NAYOHMi/uETUZZm1p1dTU4Ps7Gz91+fOnUNqairc3d0RGBjYxjt7pp37c6ARgAxAbwnz8IDLK66kpKSgvLy83RVXtFotEhMT8X//93+QyWRYunQpbrnllg49tCKEQFVVVZOVVezs7Ay+DhFRZzJr6B09ehQ33nij/mvdeN3MmTP1G5Vaio+SclGrVMPV3lrybU0hBDIyMpCfn9/uVkFarRbbt2/Hd999BysrKyxbtgwTJ07sUK0ajQZlZWVwcHBAdHQ0J5oTUbdh1tAbP378NT0a35Po5uT1cbWXPI539uxZZGVltbtVkFarxbZt2/D9999fc+DV19ejqqoKPj4+iI6Ohru7e4euQ0RkDhzT6yIM2f0cAIqLi3HixAlYW1u3uX6lVqvFiy++qA+85cuXdyjwhBCoqKhAbW0twsPDMWbMGAYeEXU73erpzZ7oo6Rc7NyfgyfHh7S7Y4JOdXU1UlNT0djYCA8Pj1bP02g02Lp1K3788UdYWVlh5cqVuOkmaZ9xJbVajbKyMjg5OWHw4MEIDAzk7Uwi6pbY0zMj3dOaUjeCBQCVSoVjx46hsrISHh4erYaPRqPBli1brjnw6uvrUVZWBl9fX4wZMwb9+vVj4BFRt8WenpnoAk8jALkMkm5rarVaHD9+HEVFRe0G3gsvvIB9+/bBysoKq1evxvjx4w2qT7eUmFarRXh4OAYNGtTmuCERUXfA0DMT3fQEuQxImDa43YdXhBA4c+YMzp49C1dXV/Tq1fJ/Oo1Gg82bN+Onn36ClZUVnnvuOYwbN86g2q58OnPIkCG8nUlEPQZDzww6Mj0hPz8fJ0+ehIODQ4ur0gCXw2rTpk34+eefIZfLsXr1aoMDr6GhAZWVlfD19UV0dDTc3NwMej8RUVfG0DMDQ6cnlJWVIS0tDQBaXWJMo9Hg+eefxy+//AK5XI41a9Zg7Nixkmu6crJ5WFgYBg8eLGnjWSKi7oSh14l062oq1RpJO58DlzfabW9vPI1Gg40bN+LXX3+FXC7H2rVrcf3110uuS6PRoK6uDtbW1oiNjeVkcyLqsRh6nejKTWHdHW3b7eU1Nja2u8SYRqPBhg0bsH//fvTq1Qtr1qwxKPCUSiUqKirg4uKC4cOHc6NXIurROGWhEz05PgSu9taSenlarRbp6ekoKChodYkxtVqN9evX6wMvISFBcuAJIVBdXY2qqioEBwcjMDCQ43dE1OOxp9dJdJPQpT64kpWVhZycHPTu3bvFqQK6wDt48CCsra2RkJCAUaNGSapFtxljr169EB0djf79+yMnR9o8QSKi7oyhZ2K6cbyq+kYIXL7F2V7oFRQU4OTJk7Czs2tx5wK1Wo1169bh0KFDBgdeY2MjysvL4erqiujoaPj6+kKj0XTkWyMi6nYYeiZ25TielEnoFRUVSEtLgxACzs7OzV5vbGzE+vXr9YG3bt06jBw5UlItdXV1qK6uRp8+fXDdddf1yJ3NiYjawtAzId18PHtrK9j2krd7a1O3GWxNTQ28vLyavd7Y2IiEhAT8/vvvsLa2xoYNGzB8+PB267hydZWIiAhERka2OrmdiKgn428+E7pyPl57i0nrntS8dOkSvL29mz2p2djYiLVr1+Lw4cMGBR5XVyEi+huf3jQhqdsF6Z7UzM/Ph4eHR7MnNVUqlT7wbGxssHHjRkmBp1KpcOnSJXh4eGD06NFcLJqILB57eib00Mh+ktfUbO1JTZVKhTVr1iApKUkfeHFxce1+dk1NDerq6hAcHIyoqKg299wjIrIUDD0T0D2xCaDdcbyCggJkZma2+KSmSqXCc889hz///BO2trbYuHEjYmNj2/xs3WavADB48GCEh4dDLpdf43dERNQz8PamCejG8irrG9vcJ6+srAypqaktPql5deA9//zz7QaeRqPBpUuXYGdnh+HDhyMiIoKBR0R0Bfb0jOyjpFyU16oAAPbWVq2O5+nW1Kyrq2v2pKZKpcKqVatw5MgR2NraYtOmTYiJiWnzc3XLifn4+CAmJgaurq5G+X6IiHoShp6RPLMrBf9NuwhxxbHW1tfU7X7e0pqaSqUSq1atwtGjR2FnZ4dNmzbhuuuua/OzdeN3ISEhiIqKanXrISIiS8fQM5Lvjv8deDIAvVtZX1O3+3lBQQE8PT2bPKlpaODpxu9kMhmioqIQFhbW4hqdRER0GUPPCD5KyoVNLznqGzWwt7bCyimRLfbwhBA4deoUzp49Czc3tyYTxBsaGrBq1SokJyfDzs4OmzdvRnR0dKufqdFoUFpaCmdnZ0RHR8Pf35/TEYiI2sHQuwZXr6vZ3iT03NxcZGZmwtHRscktyIaGBqxYsQIpKSmws7PDCy+8gKioqFavo1KpUF5eDm9vbwwdOpTjd0REEjH0OqCl8bv21tUsLi7G8ePHIZfL4ejoqD9eX1+PlStXIiUlBfb29njhhRcwZMiQVq9TW1uL2tpaBAcHIzo6usUFqYmIqGUMvQ64cvwOuDyGlzBtcKvz8aqqqpCSkgKVSgUPDw/98fr6eqxYsQKpqalwcHDACy+8gMGDB7d4Dd36mUIIzr8jIuoghp4BdD083dCZtZUMjra92pyAXl9fj2PHjqGqqqrJmpr19fVYvnw50tLS4ODggC1btmDQoEEtXkOr1aKsrAz29vaIjo5G3759OX5HRNQBDD0D6Hp4QrQ/fgdc3vcuNTUVxcXFTaYm1NfXY9myZTh+/DgcHR2xZcsWREZGtngN3f53Hh4eGDp0KNzd3Y39bRERWQw+326AKVH+kKHtSec6ukWk8/Ly4OHhob8VWVdXh6VLl0oKvPr6epSXlyMgIACjR49m4BERXSP29CT6KCkXybkVWH9H62N3OkIIZGVlITs7Gy4uLvpFpHWBd+LECTg6OuLFF19EREREi+9XKBRobGxEeHg4Bg8ezP3viIiMgD09ibbuOY2Cynr9QtJtKSgowMmTJ2FnZ6ff3aC2thZLlizRB97WrVtbDbyysjIAQExMDKKjoxl4RERGwt+mEnyUlIuq+kZJ55aWljZbRLq2thZLly5FRkYGnJycsHXrVoSHhzd7r27Cee/evRETEwMfHx+jfh9ERJaOodcG3eTzyv8FngyXtwpqTXV1NY4dO4b6+np4enoCuLwu5tKlS3Hy5Ek4OzvjxRdfbDHwdAtG+/n5ISYmBi4uLib5noiILBlDrw1XBh5weT3N1sbzGhoacOzYMVRUVOif1KypqcGSJUuQmZkJZ2dnbN26FWFhYc3eq5tw3r9/f0RHR3PBaCIiE2HoSeRqb91qL083NaGwsBBeXl6wsrJCTU0Nnn32WZw6dQouLi7YunUrBgwY0OR9QghUVVVBo9Fg0KBBGDhwICecExGZEEOvDYtvCcfO/Tl4cnxIqz08IQQyMjKQm5sLd3d3yOXyZoG3bds2hIaGNnmfbsK5nZ0dYmJiEBgYyAnnREQmxtBrxUdJue0GHgBkZWUhKysLLi4usLGxQXV1NZ599lmcPn261cBTq9UoLS2Fu7s7YmJimm0iS0REpsEpCy34KCkXq7850e4Uhfz8fGRkZMDW1hb29vZQKBRYtGgRTp8+jd69e2P79u3NAk+pVKK0tBT+/v4YPXo0A4+IqBOxp9eCnftzmiwo3ZKrpyYoFAosXrwYWVlZ+sDr379/k/fodjgPDQ1FVFQUbGxsTPdNEBFRM+zpteDJ8SFwtbdu9eEV3dSEuro6uLm5oaqqCosWLUJWVhZcXV2RmJjYJPB0O5yrVCoMGTIEQ4cOZeAREZkBe3oteGhkv1bH8ZRKZZOpCbpbmjk5OXBzc8O2bdsQHBysP//KHRKioqIQEBDAB1aIiMyEoWcAjUaD1NRUFBUVwcPDA9XV1YiPj8fZs2fh5uaG7du3IygoSH++Wq1GWVkZ3NzcMHToUP2EdSIiMg+GnkRCCJw4cQK5ublwc3NDTU0NFi1apA+8xMRE9Ov3d+9Qt8KKv78/hg4dCicnJzNWT0TdiRACarUaGo3G3KV0KUK097RF+xh6EmVnZyMrKwvOzs6ora3FokWLcO7cObi7uyMxMRGBgYH6c3UrrISEhCA6Oprjd0QkmUqlQmFhIerq6sxdSpcihEBjYyNUKpV+If+OYOhJkJ+fjxMnTsDGxgYNDQ2Ij4/H+fPn4eHhge3bt+sDTwiByspKaLVaDB48GAMHDoSVFZ8VIiJptFotzp07B7lcDn9/f9jY2PAZAFz+3ar7x0Bubi7CwsI6/LuVodeOsrIy/dQEtVqN+Ph45ObmwtPTE9u3b0dAQACApiusREVFcYUVIjKYSqWCVqtFQEAAHBwczF1Ol2JnZwetVovCwkKoVCrY2dl16DoMvTbU1NQgOTkZdXV1kMvlWLRokT7wEhMT0bdvXwBNH1jhCitEdK14h6hlxuhIMPRaoVQqkZycjIqKCsjlcsTHxyMvL69Z4F25JdDQoUP1e+gREVHXw9BrwZVTEwBg0aJFyMvLg5eXFxITE9GnTx8Alx9Yqamp4ZZARETdBEPvKrpdE86fPw+tVoulS5fiwoUL8Pb2xvbt29GnTx/9lkBqtVr/wAq3BCIi6vp44/gqZ8+exZkzZ6BSqfSB5+Pjo+/hCSFQVlYGKysrxMbGIjIykoFHRF3KudJavPDjKTy9KwUv/HgK50prO+VzL1y4gNmzZ+ufPO3Xrx/mz5+PsrKyJudlZGRg+vTp8PLygq2tLQYMGIDVq1d3yjQNht4VCgsLceLECSgUCqxYsaJJ4Pn7+0Oj0aCkpATOzs4YMWIEgoOD+YQmEXUpnx29gAnb9uONg2fx3fGLeOPgWUzYth+fH71g0s89e/Ys4uLicObMGezatQvZ2dl47bXX8PPPP2PUqFEoLy8HACQlJWHEiBFQqVT47rvvcObMGTz//PN4//33MXHiRKhUKpPWydub/1NZWYmUlBQUFxdj3bp1KCgogI+PD3bs2AFfX1+oVCqUl5fD19cXQ4cOhYuLi7lLJiJq4lxpLZZ9eRxaAUC3esn//nfpl8cxLMgdQZ6OJvnsefPmwcbGBnv37tVPHg8MDERMTAxCQkKwcuVKvPrqq5g9ezYiIiLw1Vdf6Z9S7devH8LCwhATE4PExEQsXbrUJDUC7OkBAOrq6pCcnIzz588jISEBBQUF8PX11QdeXV0dKioqEBwcjJEjRzLwiKhL+uzohVbvPslkMnxqot5eeXk59uzZg7lz5zZbLcXX1xczZszAp59+itTUVJw8eRLx8fHNpmVER0fj5ptvxq5du0xSo47Fh15jYyNSUlKQmZmJDRs24OLFi/Dz80NiYiJ8fHxQVVWFuro6REZGIi4ursMTIomITC2/or7V9SmFEMivqDfJ52ZlZUEIgYiIiBZfj4iIQEVFBc6cOaP/urXzdOeYikXf3tRqtUhPT0dKSgo2btyIoqIi+Pv7Y/v27fD29kZZWRmsra0xdOhQjt8RUZfX183+8u+pFoJPJpOhr1vH16y8FlIXihZCmHytYovt6QkhcObMGSQlJWHDhg36wEtMTISnpydKSkrg5OSEkSNHon///gw8IurypscFtNnTuy8uwCSfGxoaCplMhpMnT7b4+qlTp+Dl5YUBAwYAQJvnhYWFmaRGHYsNvaKiIhw4cAAJCQkoLi6Gv78/duzYgd69e6O0tBQ+Pj4YPXo0fHx8zF0qEZEkwZ6OeOHuKFjJALmVrMn/vnB3lMkeYvHw8MDEiRPx6quvor6+6S3UoqIifPzxx5g1axZiYmIwcOBAJCYmQqvVNjkvLS0NP/30E2bNmmWSGnUsMvRqa2uxb98+rF69GiUlJejTpw927NgBBwcHVFVVITg4GKNGjeIDK0TU7dwbF4BfFo3HP2/ojylR/vjnDf3xy6LxuNdEvTydV155BUqlErfccgsOHjyICxcu4Mcff8TEiRMRFhaG5557DjKZDG+99RZOnjyJu+++G3/99Rfy8vLw+eefY+rUqbjlllvwxBNPmLROixvT02g0+OGHH7Bo0SKUlpaib9++2LZtG3r16gWlUonBgwcjPDycE86JqNsK8nTE0lsHdupnDhgwAEeOHMHatWsxffp0lJSUQAiBu+66Cx9++KF+14gxY8YgKSkJCQkJmDx5sn7+3lNPPYXExEST/+61uJ7eL7/8gjlz5qC0tBQBAQHYtm0bZDIZrK2tERcXh4iICAYeEVEHBAUF4b333kNRURG0Wi2ee+457N27F2lpaU3OGzJkCL744guUlZVBo9Fg1qxZ+PLLL3Hu3DmT12hRPb2//voLDzzwAMrKyhAQEIAtW7ZACAF3d3fExMTAw8PD3CUSEfUYCQkJCAoKwp9//okRI0a0uGWSlZUV3n77bfzrX//CoUOH9A+7mIrFhN6JEycwZcoUlJWVITAwEOvXr7/8CG/fvoiJieGGjUREJvDII4+0e46VlRXmz5/fCdV0gdubr776KoKDg2FnZ4fY2FgcOnTI6J9x+vRp3HTTTSgtLUVgYCBWrVoFFxcXDBkyBCNGjGDgERFZCLOG3qeffooFCxZg5cqVSElJwdixYzF58mTk5eUZ7TPOnTuHG2+8EZcuXUJAQABWrFiBkJAQjBkzBpGRkejVy2I6u0REFs+sobd9+3bMnj0bjz32GCIiIrBjxw4EBARg586dRrn+6dOn8Y9//EM/LSEhIQGjR4/G9ddfDy8vL6N8BhGRsUldwcTSGKNdzNbNUalUSE5OxrJly5ocnzRpEg4fPtzie5RKJZRKpf5rhUIB4PI0BI1G0+TcU6dOIXb0OCgVZbD27AfZnRuRcNoVOJ0HoP2epL21FVZMHogHRwQa9o11QxqNBlqttlkbUnNsK2nYTtJd2VZWVlYQQqC2tpbr/F5FCIGGhgYAl8cAr/7ZkvqzZrbQKy0thUajabbiiY+PD4qKilp8z6ZNm5CQkNDseE5ODpycnJoc27Bhw+XA8wqCz/0bIXfobVB99Y1a/OvnMxjmrmz/5G5Oq9WivLwc2dnZLT5dRX9jW0nDdpLu6rZSq9UoLCxEY2Mj7OzsuAQi/g68kpISAJf37rtaTU2NpGuZfUDr6v+gQohW/yMvX74c8fHx+q8VCgUCAgIQEhLSbPWUd999F3+VyKAa9P8MDjzgck/v6QlhGDDAMnp62dnZCA0N5RzFdrCtpGE7SXd1WwkhUFxcjIqKCnOX1iUNHDiwxWcxdHf+2mO20PP09IRcLm/WqyspKWl1vUtbW1vY2to2Oy6Xy5v9xZLL5Tiz90NkZWVhwIAB/IvXDisrqxbbkZpjW0nDdpLu6rbq06cPfH190djYaObKug4rKyucPXsWvXr1avFnSurPmdlCz8bGBrGxsdi3bx/uvPNO/fF9+/Zh2rRp5iqLiKhL4D8YmjLW+LBZb2/Gx8fj4YcfRlxcHEaNGoU33ngDeXl5mDNnjjnLIiKiHsqsoXffffehrKwM69atQ2FhIQYPHozvv/8e/fr1M2dZRETUQ5n9QZa5c+di7ty55i6DiIgsgNlD71roJiq29tSORqNBTU0NFAoF7423ge0kHdtKGraTdGwradprJ10OtDeBvVuHXnV1NQAgIMC0myMSEVH3UF1djd69W5+mJhPdeL0brVaLixcvwtnZucW5fbp5fBcuXOAu6G1gO0nHtpKG7SQd20qa9tpJCIHq6mr4+/u3uSBCt+7pWVlZoW/fvu2e5+Liwh8mCdhO0rGtpGE7Sce2kqatdmqrh6fD9YGIiMhiMPSIiMhi9OjQs7W1xZo1a1pcuoz+xnaSjm0lDdtJOraVNMZqp279IAsREZEhenRPj4iI6EoMPSIishgMPSIishgMPSIishg9NvReffVVBAcHw87ODrGxsTh06JC5S+pyDh48iKlTp8Lf3x8ymQzffPONuUvqkjZt2oRhw4bB2dkZ3t7euOOOO3D69Glzl9Ul7dy5E1FRUfoJxKNGjcIPP/xg7rK6vE2bNkEmk2HBggXmLqXLWbt2LWQyWZM/vr6+Hb5ejwy9Tz/9FAsWLMDKlSuRkpKCsWPHYvLkycjLyzN3aV1KbW0toqOj8corr5i7lC7twIEDmDdvHpKSkrBv3z6o1WpMmjQJtbW15i6ty+nbty82b96Mo0eP4ujRo7jpppswbdo0ZGRkmLu0LuvIkSN44403EBUVZe5SuqxBgwahsLBQ/yc9Pb3jFxM90PDhw8WcOXOaHBs4cKBYtmyZmSrq+gCIr7/+2txldAslJSUCgDhw4IC5S+kW3NzcxFtvvWXuMrqk6upqMWDAALFv3z4xbtw4MX/+fHOX1OWsWbNGREdHG+16Pa6np1KpkJycjEmTJjU5PmnSJBw+fNhMVVFPUlVVBQBwd3c3cyVdm0ajwSeffILa2lqMGjXK3OV0SfPmzcOUKVNw8803m7uULi0rKwv+/v4IDg7G/fffj7Nnz3b4Wt16wemWlJaWQqPRwMfHp8lxHx8fFBUVmakq6imEEIiPj8f111+PwYMHm7ucLik9PR2jRo1CQ0MDnJyc8PXXXyMyMtLcZXU5n3zyCY4dO4YjR46Yu5QubcSIEfjggw8QFhaG4uJibNiwAaNHj0ZGRgY8PDwMvl6PCz2dq7caEkK0uP0QkSGeeuopHD9+HL/99pu5S+mywsPDkZqaisrKSnz55ZeYOXMmDhw4wOC7woULFzB//nzs3bsXdnZ25i6nS5s8ebL+/w8ZMgSjRo1CSEgI3n//fcTHxxt8vR4Xep6enpDL5c16dSUlJc16f0SGePrpp/Htt9/i4MGDkra0slQ2NjYIDQ0FAMTFxeHIkSN46aWX8Prrr5u5sq4jOTkZJSUliI2N1R/TaDQ4ePAgXnnlFSiVSu6i3gpHR0cMGTIEWVlZHXp/jxvTs7GxQWxsLPbt29fk+L59+zB69GgzVUXdmRACTz31FL766iv88ssvCA4ONndJ3YoQAkql0txldCkTJkxAeno6UlNT9X/i4uIwY8YMpKamMvDaoFQqkZmZCT8/vw69v8f19AAgPj4eDz/8MOLi4jBq1Ci88cYbyMvLw5w5c8xdWpdSU1OD7Oxs/dfnzp1Damoq3N3dERgYaMbKupZ58+bhP//5D3bv3g1nZ2f9XYTevXvD3t7ezNV1LStWrMDkyZMREBCA6upqfPLJJ9i/fz9+/PFHc5fWpTg7OzcbE3Z0dISHhwfHiq+yePFiTJ06FYGBgSgpKcGGDRugUCgwc+bMjl3QaM+BdjH//ve/Rb9+/YSNjY0YOnQoHy9vwa+//ioANPszc+ZMc5fWpbTURgDEu+++a+7SupxHH31U//fOy8tLTJgwQezdu9fcZXULnLLQsvvuu0/4+fkJa2tr4e/vL+666y6RkZHR4etxayEiIrIYPW5Mj4iIqDUMPSIishgMPSIishgMPSIishgMPSIishgMPSIishgMPSIishgMPeqRgoKCsGPHDnOX0arTp0/D19cX1dXVkt8zbNgwfPXVVyasqnuaNWsW7rjjDnOXQd0EQ49IooyMDNx9990ICgqCTCZrNVRfffVVBAcHw87ODrGxsTh06FCzc1auXIl58+bB2dkZALB//37IZDL9Hy8vL0yePBlpaWn696xevRrLli2DVqs1yffXmRhUZC4MPSKJ6urq0L9/f2zevBm+vr4tnvPpp59iwYIFWLlyJVJSUjB27FhMnjwZeXl5+nPy8/Px7bff4pFHHmn2/tOnT6OwsBDfffcdKioqcOutt+o3rZ0yZQqqqqqwZ88e03yDRBaAoUcWIS8vD9OmTYOTkxNcXFwwffp0FBcXNzlnw4YN8Pb2hrOzMx577DEsW7YM1113nf71YcOG4cUXX8T9998PW1vbFj9n+/btmD17Nh577DFERERgx44dCAgIwM6dO/XnfPbZZ4iOjm5xeyJvb2/4+vpi+PDh2LZtG4qKipCUlAQAkMvluO2227Br1642v9eCggLcd999cHNzg4eHB6ZNm4bz588DAE6dOgUHBwf85z//0Z//1Vdfwc7ODunp6QD+7oUlJCTA29sbLi4ueOKJJ6BSqfTvEUJgy5Yt6N+/P+zt7REdHY0vvviiSR0ZGRmYMmUKXFxc4OzsjLFjxyInJwdr167F+++/j927d+t7tvv372+3duDy9jvx8fFwdXWFh4cHlixZAq6kSIZg6FGPJ4TAHXfcgfLychw4cAD79u1DTk4O7rvvPv05H3/8MTZu3IgXXngBycnJCAwMbBJUUqhUKiQnJ2PSpElNjk+aNAmHDx/Wf33w4EHExcW1ez3dDg6NjY36Y8OHD2/xdqlOXV0dbrzxRjg5OeHgwYP47bff4OTkhFtvvRUqlQoDBw7E1q1bMXfuXOTm5uLixYt4/PHHsXnzZgwZMkR/nZ9//hmZmZn49ddfsWvXLnz99ddISEjQv75q1Sq8++672LlzJzIyMrBw4UI89NBDOHDgAIDL4XXDDTfAzs4Ov/zyC5KTk/Hoo49CrVZj8eLFmD59Om699VYUFhaisLAQo0ePbrd2ANi2bRveeecdvP322/jtt99QXl6Or7/+ut22JNIzzjrYRF1Lv379RGJiohBCiL179wq5XC7y8vL0r2dkZAgA4q+//hJCCDFixAgxb968JtcYM2aMiI6Obvf6OgUFBQKA+P3335sc37hxowgLC9N/HR0dLdatW9fkHN2OFxUVFUIIIUpLS8Xtt98unJ2dRXFxsf683bt3CysrK6HRaFqs6+233xbh4eFCq9XqjymVSmFvby/27NmjPzZlyhQxduxYMWHCBDFx4sQm58+cOVO4u7uL2tpa/bGdO3cKJycnodFoRE1NjbCzsxOHDx9u8tmzZ88WDzzwgBBCiOXLl4vg4GChUqlarHPmzJli2rRpBtfu5+cnNm/erH+9sbFR9O3bt9m1iFrTI/fTI8vx8ccf44knntB//cMPP2Ds2LFNzsnMzERAQAACAgL0xyIjI+Hq6orMzEwMGzYMp0+fxty5c5u8b/jw4fjll18MrkkmkzX5WgjR5Fh9fT3s7OxafK/ulmdtbS0GDBiAzz//HN7e3vrX7e3todVqoVQqW9zLLzk5GdnZ2foHZHQaGhqQk5Oj//qdd95BWFgYrKyscOLEiWY1R0dHw8HBQf/1qFGjUFNTgwsXLqCkpAQNDQ2YOHFik/eoVCrExMQAAFJTUzF27FhYW1u3+H22pL3aq6qqUFhYiFGjRulf69WrF+Li4niLkyRj6FG3dvvtt2PEiBH6r/v06dPsnKtDp7XjLYWVITw9PSGXy/WbzOqUlJTAx8enyXkVFRUtXuPQoUNwcXGBl5cXXFxcmr1eXl4OBweHVjev1Wq1iI2Nxccff9zsNS8vL/3/T0tLQ21tLaysrFBUVAR/f39J36NMJtM/Pfrdd981a2/dWGdHNteVWjvRtWDoUbfm7OzcrGdwtcjISOTl5eHChQv63t7JkydRVVWFiIgIAEB4eDj++usvPPzww/r3HT161KBabGxsEBsbi3379uHOO+/UH9+3bx+mTZum/zomJgYnT55s8RrBwcFwdXVt9TNOnDiBoUOHtvr60KFD8emnn+ofQGlJeXk5Zs2ahZUrV6KoqAgzZszAsWPHmgRVWloa6uvr9ceSkpLg5OSEvn37ws3NDba2tsjLy8O4ceNa/IyoqCi8//77aGxsbLG3Z2NjA41GY3Dtfn5+SEpKwg033AAAUKvVSE5ObrNNiJow681VIhO5csxNq9WKmJgYMXbsWJGcnCz+/PNPERsbK8aNG6c//6OPPhL29vbivffeE2fOnBHr168XLi4u4rrrrtOfo1QqRUpKikhJSRF+fn5i8eLFIiUlRWRlZenP+eSTT4S1tbV4++23xcmTJ8WCBQuEo6OjOH/+vP6cb7/9Vnh7ewu1Wq0/dvWYXmvGjRvXbDzwSrW1tWLAgAFi/Pjx4uDBg+Ls2bNi//794plnnhEXLlwQQghx7733ihEjRojGxkZRW1srwsPDxdy5c/XXmDlzpnBychIPPPCAyMjIEN9//73w8fERy5Yt05+zcuVK4eHhId577z2RnZ0tjh07Jl555RXx3nvvCSEuj0l6eHiIu+66Sxw5ckScOXNGfPDBB+LUqVNCiMvjnIGBgeLUqVPi0qVLQqVSSap98+bNws3NTXz11VciMzNTPP7448LZ2ZljeiQZQ496pKsfNMnNzRW33367cHR0FM7OzuLee+8VRUVFTd6zbt064enpKZycnMSjjz4qnnnmGTFy5Ej96+fOnRMAmv25MjyFEOLf//636Nevn7CxsRFDhw4VBw4caPK6Wq0Wffr0ET/++KP+mJTQy8/PF9bW1voAaE1hYaH4xz/+ITw9PYWtra3o37+/ePzxx0VVVZV4//33haOjozhz5oz+/KNHjwobGxvx3XffCSH+fsjkueeeEx4eHsLJyUk89thjoqGhQf8erVYrXnrpJREeHi6sra2Fl5eXuOWWW5p8r2lpaWLSpEnCwcFBODs7i7Fjx4qcnBwhhBAlJSVi4sSJwsnJSQAQv/76a7u1C3H5wZX58+cLFxcX4erqKuLj48U//vEPhh5JJhOCI8BELZk4cSJ8fX3x4YcfGv3ar776Knbv3m3QRPNnn30WVVVVeOONN4xez5VmzZqFyspKfPPNNyb9HCJz4JgeES7Pb3vttddwyy23QC6XY9euXfjpp5+wb98+k3zeP//5T1RUVKC6urrdMUkdb29vLF682CT1EFkK9vSIcHkawdSpU3Hs2DEolUqEh4dj1apVuOuuu8xdWqdjT496MoYeERFZDC5DRkREFoOhR0REFoOhR0REFoOhR0REFoOhR0REFoOhR0REFoOhR0REFoOhR0REFoOhR0REFuP/A1M7BbuGkkAAAAAAAElFTkSuQmCC",
         "text/plain": [
          "<Figure size 500x500 with 1 Axes>"
         ]
        },
        "metadata": {},
        "output_type": "display_data"
       }
      ],
      "source": [
       "# this runs in about 10 seconds on a fast machine\n",
       "\n",
       "# import the algorithm and reader\n",
       "import numpy as np\n",
       "from fastlmm.association import snp_set\n",
       "from pysnptools.snpreader import Bed\n",
       "from fastlmm.util import example_file # Download and return local file name\n",
       "\n",
       "# define file names\n",
       "test_snps_fn = example_file('tests/datasets/synth/chr1.*','*.bed').replace('.bed','')\n",
       "pheno_fn = example_file(\"tests/datasets/synth/pheno_10_causals.txt\")\n",
       "cov_fn = example_file(\"tests/datasets/synth/cov.txt\")\n",
       "set_list_fn = example_file(\"tests/datasets/synth/chr1.sets.txt\")\n",
       "G0_fn = None\n",
       "\n",
       "\n",
       "# run SNP-set analysis\n",
       "results_df = snp_set(test_snps=test_snps_fn, G0=G0_fn, set_list=set_list_fn, pheno=pheno_fn, covar=cov_fn, test=\"lrt\")\n",
       "\n",
       "# qq plot\n",
       "from fastlmm.util.stats import plotp\n",
       "\n",
       "plotp.qqplot(results_df[\"P-value\"].values, xlim=[0,5], ylim=[0,5], addlambda=False, legend=\"QQ\")\n",
       "\n",
       "# print head of results data frame\n",
       "import pandas as pd\n",
       "pd.set_option('display.width', 1000)\n",
       "results_df.head(n=10)"
      ]
     },
     {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
       "As just mentioned, we see inflation of the test statistic because no background GSM was used. When you include a background GSM (using `G0`) type I error is better controlled. As in the single SNP case, proximal contamination is avoided using LOOC."
      ]
     },
     {
      "cell_type": "code",
      "execution_count": 8,
      "metadata": {},
      "outputs": [
       {
        "name": "stdout",
        "output_type": "stream",
        "text": [
         "constructing LMM - this should only happen once.\n"
        ]
       },
       {
        "data": {
         "text/html": [
          "<div>\n",
          "<style scoped>\n",
          "    .dataframe tbody tr th:only-of-type {\n",
          "        vertical-align: middle;\n",
          "    }\n",
          "\n",
          "    .dataframe tbody tr th {\n",
          "        vertical-align: top;\n",
          "    }\n",
          "\n",
          "    .dataframe thead th {\n",
          "        text-align: right;\n",
          "    }\n",
          "</style>\n",
          "<table border=\"1\" class=\"dataframe\">\n",
          "  <thead>\n",
          "    <tr style=\"text-align: right;\">\n",
          "      <th></th>\n",
          "      <th>SetId</th>\n",
          "      <th>LogLikeAlt</th>\n",
          "      <th>LogLikeNull</th>\n",
          "      <th>P-value</th>\n",
          "      <th>#SNPs_in_Set</th>\n",
          "      <th>#ExcludedSNPs</th>\n",
          "      <th>chrm</th>\n",
          "      <th>pos. range</th>\n",
          "      <th>Alt_h2</th>\n",
          "      <th>Alt_h2_1</th>\n",
          "    </tr>\n",
          "  </thead>\n",
          "  <tbody>\n",
          "    <tr>\n",
          "      <th>0</th>\n",
          "      <td>set_60</td>\n",
          "      <td>-697.449026</td>\n",
          "      <td>-699.619196</td>\n",
          "      <td>0.016062</td>\n",
          "      <td>10</td>\n",
          "      <td>0</td>\n",
          "      <td>1</td>\n",
          "      <td>600-609</td>\n",
          "      <td>0.630039</td>\n",
          "      <td>0.024977</td>\n",
          "    </tr>\n",
          "    <tr>\n",
          "      <th>1</th>\n",
          "      <td>set_11</td>\n",
          "      <td>-697.979257</td>\n",
          "      <td>-699.619196</td>\n",
          "      <td>0.028124</td>\n",
          "      <td>10</td>\n",
          "      <td>0</td>\n",
          "      <td>1</td>\n",
          "      <td>110-119</td>\n",
          "      <td>0.618464</td>\n",
          "      <td>0.022458</td>\n",
          "    </tr>\n",
          "    <tr>\n",
          "      <th>2</th>\n",
          "      <td>set_14</td>\n",
          "      <td>-698.359356</td>\n",
          "      <td>-699.619196</td>\n",
          "      <td>0.042932</td>\n",
          "      <td>10</td>\n",
          "      <td>0</td>\n",
          "      <td>1</td>\n",
          "      <td>140-149</td>\n",
          "      <td>0.623297</td>\n",
          "      <td>0.019099</td>\n",
          "    </tr>\n",
          "    <tr>\n",
          "      <th>3</th>\n",
          "      <td>set_80</td>\n",
          "      <td>-698.401148</td>\n",
          "      <td>-699.619196</td>\n",
          "      <td>0.045045</td>\n",
          "      <td>10</td>\n",
          "      <td>0</td>\n",
          "      <td>1</td>\n",
          "      <td>800-809</td>\n",
          "      <td>0.619344</td>\n",
          "      <td>0.017966</td>\n",
          "    </tr>\n",
          "    <tr>\n",
          "      <th>4</th>\n",
          "      <td>set_70</td>\n",
          "      <td>-698.430620</td>\n",
          "      <td>-699.619196</td>\n",
          "      <td>0.046607</td>\n",
          "      <td>10</td>\n",
          "      <td>0</td>\n",
          "      <td>1</td>\n",
          "      <td>700-709</td>\n",
          "      <td>0.577376</td>\n",
          "      <td>0.017370</td>\n",
          "    </tr>\n",
          "    <tr>\n",
          "      <th>5</th>\n",
          "      <td>set_45</td>\n",
          "      <td>-698.442867</td>\n",
          "      <td>-699.619196</td>\n",
          "      <td>0.047275</td>\n",
          "      <td>10</td>\n",
          "      <td>0</td>\n",
          "      <td>1</td>\n",
          "      <td>450-459</td>\n",
          "      <td>0.592120</td>\n",
          "      <td>0.015451</td>\n",
          "    </tr>\n",
          "    <tr>\n",
          "      <th>6</th>\n",
          "      <td>set_86</td>\n",
          "      <td>-698.518404</td>\n",
          "      <td>-699.619196</td>\n",
          "      <td>0.051648</td>\n",
          "      <td>10</td>\n",
          "      <td>0</td>\n",
          "      <td>1</td>\n",
          "      <td>860-869</td>\n",
          "      <td>0.592057</td>\n",
          "      <td>0.017528</td>\n",
          "    </tr>\n",
          "    <tr>\n",
          "      <th>7</th>\n",
          "      <td>set_92</td>\n",
          "      <td>-698.547797</td>\n",
          "      <td>-699.619196</td>\n",
          "      <td>0.053477</td>\n",
          "      <td>10</td>\n",
          "      <td>0</td>\n",
          "      <td>1</td>\n",
          "      <td>920-929</td>\n",
          "      <td>0.621041</td>\n",
          "      <td>0.017506</td>\n",
          "    </tr>\n",
          "    <tr>\n",
          "      <th>8</th>\n",
          "      <td>set_65</td>\n",
          "      <td>-698.714102</td>\n",
          "      <td>-699.619196</td>\n",
          "      <td>0.065409</td>\n",
          "      <td>10</td>\n",
          "      <td>0</td>\n",
          "      <td>1</td>\n",
          "      <td>650-659</td>\n",
          "      <td>0.584255</td>\n",
          "      <td>0.016500</td>\n",
          "    </tr>\n",
          "    <tr>\n",
          "      <th>9</th>\n",
          "      <td>set_93</td>\n",
          "      <td>-698.820185</td>\n",
          "      <td>-699.619196</td>\n",
          "      <td>0.074734</td>\n",
          "      <td>10</td>\n",
          "      <td>0</td>\n",
          "      <td>1</td>\n",
          "      <td>930-939</td>\n",
          "      <td>0.618306</td>\n",
          "      <td>0.012515</td>\n",
          "    </tr>\n",
          "  </tbody>\n",
          "</table>\n",
          "</div>"
         ],
         "text/plain": [
          "    SetId  LogLikeAlt  LogLikeNull   P-value  #SNPs_in_Set  #ExcludedSNPs  chrm pos. range    Alt_h2  Alt_h2_1\n",
          "0  set_60 -697.449026  -699.619196  0.016062            10              0     1    600-609  0.630039  0.024977\n",
          "1  set_11 -697.979257  -699.619196  0.028124            10              0     1    110-119  0.618464  0.022458\n",
          "2  set_14 -698.359356  -699.619196  0.042932            10              0     1    140-149  0.623297  0.019099\n",
          "3  set_80 -698.401148  -699.619196  0.045045            10              0     1    800-809  0.619344  0.017966\n",
          "4  set_70 -698.430620  -699.619196  0.046607            10              0     1    700-709  0.577376  0.017370\n",
          "5  set_45 -698.442867  -699.619196  0.047275            10              0     1    450-459  0.592120  0.015451\n",
          "6  set_86 -698.518404  -699.619196  0.051648            10              0     1    860-869  0.592057  0.017528\n",
          "7  set_92 -698.547797  -699.619196  0.053477            10              0     1    920-929  0.621041  0.017506\n",
          "8  set_65 -698.714102  -699.619196  0.065409            10              0     1    650-659  0.584255  0.016500\n",
          "9  set_93 -698.820185  -699.619196  0.074734            10              0     1    930-939  0.618306  0.012515"
         ]
        },
        "execution_count": 8,
        "metadata": {},
        "output_type": "execute_result"
       },
       {
        "data": {
         "image/png": "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",
         "text/plain": [
          "<Figure size 500x500 with 1 Axes>"
         ]
        },
        "metadata": {},
        "output_type": "display_data"
       }
      ],
      "source": [
       "#Here we give a G0, the GSM, created via Bed file with all the chroms except chrom 1 (the test snps)\n",
       "\n",
       "#WARNING: Even on a fast machine, this takes 24 minutes (but, running with test=\"sc_davies\" takes seconds)\n",
       "\n",
       "bed_reader = Bed(example_file('tests/datasets/synth/all.*','*.bed'),count_A1=True)\n",
       "chrNot1 = bed_reader[:,bed_reader.pos[:,0] != 1]\n",
       "G0_fn = 'tempChrNot1.bed'\n",
       "Bed.write(G0_fn,chrNot1.read(),count_A1=True)\n",
       "\n",
       "# run SNP-set analysis\n",
       "results_df = snp_set(test_snps=test_snps_fn, G0=G0_fn, set_list=set_list_fn, pheno=pheno_fn, covar=cov_fn, test=\"lrt\")\n",
       "\n",
       "# qq plot\n",
       "plotp.qqplot(results_df[\"P-value\"].values, xlim=[0,5], ylim=[0,5], addlambda=False, legend=\"QQ\")\n",
       "\n",
       "# print head of results data frame\n",
       "import pandas as pd\n",
       "pd.set_option('display.width', 1000)\n",
       "results_df.head(n=10)"
      ]
     },
     {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
       "The `G1` option available for single-marker testing is not yet supported for set testing. With the same caveats as previously described, however, you can use PC covariates and a `G0` based on selected SNPs when your data is confounded, for example, by population structure."
      ]
     },
     {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
       "### Prediction\n",
       "\n",
       "We can train on one set of examples and make predictions on another set."
      ]
     },
     {
      "cell_type": "code",
      "execution_count": 9,
      "metadata": {},
      "outputs": [
       {
        "name": "stdout",
        "output_type": "stream",
        "text": [
         "Predicted means and stdevs\n",
         "[-0.01441869 -0.28643401 -0.25428545  0.08895201 -0.29511007 -0.39238035\n",
         "  0.06844375 -0.43282078 -0.21734716 -0.35522388]\n",
         "[0.95795615 0.96181176 0.95260426 0.94991531 0.9604778  0.95411448\n",
         " 0.96020394 0.9817148  0.95198989 0.96271045]\n"
        ]
       },
       {
        "data": {
         "image/png": "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",
         "text/plain": [
          "<Figure size 1000x800 with 1 Axes>"
         ]
        },
        "metadata": {},
        "output_type": "display_data"
       }
      ],
      "source": [
       "from pysnptools.snpreader import Pheno, Bed\n",
       "from fastlmm.inference import FastLMM\n",
       "import numpy as np\n",
       "from fastlmm.util import example_file # Download and return local file name\n",
       "\n",
       "# define file names\n",
       "bed_fn = example_file('tests/datasets/synth/all.*','*.bed')\n",
       "snp_reader = Bed(bed_fn, count_A1=True)\n",
       "pheno_fn = example_file(\"tests/datasets/synth/pheno_10_causals.txt\")\n",
       "cov_fn = example_file(\"tests/datasets/synth/cov.txt\")\n",
       "\n",
       "# Divide the data into train (all but the last 10 individuals) and test (the last 10 individuals)\n",
       "# (the cov and pheno will automatically be divided to match)\n",
       "train = snp_reader[:-10,:]\n",
       "test = snp_reader[-10:,:]\n",
       "\n",
       "# In the style of scikit-learn, create a predictor and then train it.\n",
       "fastlmm = FastLMM(GB_goal=2)\n",
       "fastlmm.fit(K0_train=train,X=cov_fn,y=pheno_fn)\n",
       "\n",
       "# Now predict with it\n",
       "mean, covariance = fastlmm.predict(K0_whole_test=test,X=cov_fn)\n",
       "\n",
       "print(\"Predicted means and stdevs\")\n",
       "print(mean.val[:,0])\n",
       "print(np.sqrt(np.diag(covariance.val)))\n",
       "\n",
       "#Plot actual phenotype and predicted phenotype\n",
       "whole_pheno = Pheno(pheno_fn)\n",
       "actual_pheno = whole_pheno[whole_pheno.iid_to_index(mean.iid),:].read()\n",
       "pylab.plot(actual_pheno.val,\"r.\")\n",
       "pylab.plot(mean.val,\"b.\")\n",
       "pylab.errorbar(np.arange(mean.iid_count),mean.val.flatten(),yerr=np.sqrt(np.diag(covariance.val)),fmt='.')\n",
       "pylab.xlabel('testing examples')\n",
       "pylab.ylabel('phenotype, actual (red) and predicted (blue with stdev)')\n",
       "pylab.show()"
      ]
     },
     {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
       "### Appendix"
      ]
     },
     {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
       "Columns in the output are as follows:\n",
       "\n",
       "`SNP` or `Set`\n",
       "The SNP or set identifier tested.\n",
       "\n",
       "`PValue`\n",
       "The _P_ value computed for the SNP tested.\n",
       "\n",
       "`NullLogLike`\n",
       "The log likelihood of the null model.\n",
       "\n",
       "`AltLogLike`\n",
       "The log likelihood of the alternative model.\n",
       "\n",
       "Global statistics, such as sample size and the number of SNPs tested are printed to stdout at the end of the run.\n",
       "\n",
       "###### Single-SNP testing only:\n",
       "\n",
       "`Chr`\n",
       "The chromosome identifier for the SNP tested or 0 if unplaced.  Taken from the PLINK file.\n",
       "\n",
       "`GenDist`\n",
       "The genetic distance of the SNP on the chromosome.  Taken from the PLINK file.  Any units are allowed, but typically centimorgans or morgans are used.\n",
       "\n",
       "`ChrPos`\n",
       "The base-pair position of the SNP on the chromosome (bp units).  Taken from the PLINK file.\n",
       "\n",
       "`SnpWeight`\n",
       "The fixed-effect weight of the SNP.\n",
       "\n",
       "`SnpWeightSE`\n",
       "The standard error of the SnpWeight.\n",
       "\n",
       "`EffectSize`\n",
       "SNPWeight^2 * “raw” test SNP variance / phenotype variance,  where \"raw\" means the original test SNP values of \"0,1,2\" (the count of allele 1 or 2 [it doesn’t matter])\n",
       "\n",
       "\n",
       "`Nullh2`\n",
       "The narrow sense heritability given by h2 $=σ_g^2/(σ_g^2+σ_e^2)$ on the null model.\n",
       "\n",
       "###### SNP-set testing only:\n",
       "\n",
       "`#SNPs_in_Set`\n",
       "The number of markers in the set.\n",
       "\n",
       "`Alt_h2`\n",
       "The value found in the alt. model, given by $\\sigma^2(h2(a2 \\cdot K_S+(1-a2) \\cdot K_C)+(1-h2) \\cdot I)$, where $K_C$ is the NxN matrix to correct for confounding, and $K_S$ is the matrix containing SNPs to be tested.\t\n",
       "\n",
       "`Alt_a2`\n",
       "See Alt_h2.\n",
       "\n",
       "`Null_h2`\n",
       "The value found in the null model, given by $\\sigma^2(h2 \\cdot K_C+1-h2) \\cdot I)$, where $K_C$ is the $N \\times N$ matrix to correct for confounding.\n"
      ]
     },
     {
      "cell_type": "markdown",
      "metadata": {},
      "source": []
     }
    ],
    "metadata": {
     "anaconda-cloud": {},
     "interpreter": {
      "hash": "821f95ec71e093120824af58e1c459369d357f87cd2635d0d001c262c60c5b22"
     },
     "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": "3.11.9"
     }
    },
    "nbformat": 4,
    "nbformat_minor": 1
   }