ipython-nbs/figure2_suppl1C.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "import matplotlib\n",
    "\n",
    "import sys\n",
    "from __future__ import division\n",
    "from collections import defaultdict\n",
    "# from pylab import *\n",
    "import numpy as np\n",
    "from itertools import izip\n",
    "import colorbrewer as cb\n",
    "import os\n",
    "from scripts import utils as sutils\n",
    "import cPickle as pickle\n",
    "import fnmatch\n",
    "from collections import defaultdict\n",
    "from matplotlib import pyplot as plt\n",
    "\n",
    "LSV_TYPES_DICT = {\n",
    "    's|1e1.1|1e2.1':'SE',\n",
    "    't|1e1.1|1e2.1':'SE',\n",
    "    's|1e1.1|1e1.2':'A3SS',\n",
    "    't|1e1.1|2e1.1':'A3SS',\n",
    "    't|1e1.1|1e1.2':'A5SS',\n",
    "    's|1e1.1|2e1.1':'A5SS'\n",
    "}\n",
    "\n",
    "BSIZE = 0.025  \n",
    "BINS = np.arange(0, 1, BSIZE)\n",
    "BINS_CENTER = np.arange(0 + BSIZE / 2, 1, BSIZE)\n",
    "                        \n",
    "def list_files_or_dir(file_or_dir_list, prefix='*', suffix='*', containing='*'):\n",
    "    if type(file_or_dir_list) != list: return [file_or_dir_list]\n",
    "    files = []\n",
    "    for file_or_dir in file_or_dir_list:\n",
    "        if os.path.isdir(file_or_dir):\n",
    "            for root, dirnames, filenames in os.walk(file_or_dir):\n",
    "                for filename in fnmatch.filter(filenames, '%s*%s*%s' % (prefix, containing, suffix)):\n",
    "                    files.append(os.path.join(root, filename))\n",
    "        else:\n",
    "            files.append(file_or_dir)\n",
    "    return files\n",
    "\n",
    "def grouped(iterable, n):\n",
    "    \"s -> (s0,s1,s2,...sn-1), (sn,sn+1,sn+2,...s2n-1), (s2n,s2n+1,s2n+2,...s3n-1), ...\"\n",
    "    return izip(*[iter(iterable)]*n)\n",
    "\n",
    "def calculate_ead_simple(psi_list1, psi_list2):\n",
    "    return abs(np.array(psi_list1) - np.array(psi_list2))\n",
    "\n",
    "def hex_to_rgb(value):\n",
    "    value = value.lstrip('#')\n",
    "    lv = len(value)\n",
    "    return tuple(int(value[i:i + lv // 3], 16) for i in range(0, lv, lv // 3))\n",
    "\n",
    "\n",
    "def rgb_to_hex(rgb):\n",
    "    return '#%02x%02x%02x' % rgb\n",
    "\n",
    "\n",
    "def intersect_sets(majiq1, majiq2):\n",
    "    \"\"\"Find common names and return their psi values in 2 lists\"\"\"\n",
    "    names1 = [m.get_id() for m in majiq1]\n",
    "    names2 = [m.get_id() for m in majiq2]\n",
    "    common_names = set(names1).intersection(set(names2))\n",
    "    return  np.array([mm.get_bins() for mm in majiq1])[np.array([name in common_names for name in names1])], \\\n",
    "            np.array([mm.get_bins() for mm in majiq2])[np.array([name in common_names for name in names2])], \\\n",
    "            np.array(names1)[np.array([name in common_names for name in names1])], \\\n",
    "            np.array(names2)[np.array([name in common_names for name in names2])]\n",
    "\n",
    "\n",
    "            \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def plot_delta_R(psi_dict_lists, replica_names, plotpath=None):\n",
    "\n",
    "    colors_list=['blue', 'red', 'green', 'purple', 'orange', 'yellow']\n",
    "    replica_names_joint = '; '.join([\"%s%s\" % (name1, name2) for name1, name2 in grouped(replica_names, 2)])\n",
    "    plotname=\"MAJIQ_Vs_Others_delta_expected_psi. \\nReplicates %s\" % (replica_names_joint)\n",
    "\n",
    "\n",
    "    fig = plt.figure(figsize=[8, 6]) # In inches\n",
    "    #figure out how many groups of events exist\n",
    "    max_difference=0\n",
    "    font = {'size': 10} #here also 'weight' and 'family'\n",
    "    matplotlib.rc('font', **font)\n",
    "\n",
    "    lthres = [0.05, 0]\n",
    "    for thres in lthres:\n",
    "        win_cdf_all = []\n",
    "        win_psi_all = []\n",
    "        lose_cdf_all = []\n",
    "        lose_psi_all = []\n",
    "        nbins=5000\n",
    "\n",
    "        for met_key, met_diff_list in psi_dict_lists.iteritems():\n",
    "            for jj, methods_diff in enumerate(met_diff_list):\n",
    "                palette=cb.Blues[7][::-1]\n",
    "\n",
    "                # methods_diff=np.mean(met_diff_list)\n",
    "                methods_diff.sort()\n",
    "                win_elems=methods_diff[np.where(methods_diff>thres)]\n",
    "                uu, ii = np.unique(win_elems, return_inverse=True)\n",
    "\n",
    "                win_cdf=[0]\n",
    "                for w_freq in np.bincount(ii):\n",
    "                    win_cdf.append(win_cdf[-1]+((1.*w_freq)/len(win_elems)))\n",
    "                win_cdf_final = np.dot(win_cdf[1:], len(win_elems)/len(methods_diff)) # np.dot(win_cdf[1:],len(win_elems))\n",
    "\n",
    "                aux=[]\n",
    "                aux_psi=[]\n",
    "                for aa in xrange(0,nbins,1):\n",
    "                    aux.append(win_cdf_final[int((aa/(nbins*1.0))*len(win_cdf_final))])\n",
    "                    aux_psi.append(uu[int((aa/(nbins*1.0))*len(uu))])\n",
    "                win_cdf_all.append(aux)\n",
    "                win_psi_all.append(aux_psi)\n",
    "\n",
    "                lose_elems=-methods_diff[np.where(methods_diff<-thres)]\n",
    "                uul, iil = np.unique(lose_elems, return_inverse=True)\n",
    "\n",
    "                _cdf=[0]\n",
    "                for w_freq in np.bincount(iil):\n",
    "                    _cdf.append(_cdf[-1]+((1.*w_freq)/len(lose_elems)))\n",
    "                lose_cdf = np.dot(_cdf[1:],len(lose_elems)/len(methods_diff))  # np.dot(_cdf[1:][::-1],-len(lose_elems))\n",
    "\n",
    "                aux=[]\n",
    "                aux_psi=[]\n",
    "                for aa in xrange(0,nbins,1):\n",
    "                    aux.append(lose_cdf[int((aa/(nbins*1.0))*len(lose_cdf))])\n",
    "                    aux_psi.append(uul[int((aa/(nbins*1.0))*len(uul))])\n",
    "                lose_cdf_all.append(aux)\n",
    "                lose_psi_all.append(aux_psi)\n",
    "\n",
    "\n",
    "                max_difference = max(max_difference, max(len(win_elems), len(lose_elems)))\n",
    "\n",
    "        lbl_majiq=\"MAJIQ winning |ddPSI|>%.0f%% - mean %.0f%%\" % (100*thres, 100*np.mean(win_cdf_all, axis=0)[-1])\n",
    "        lbl_miso=\"MISO winning |ddPSI|>%.0f%% - mean %.0f%%\" % (100*thres, 100*np.mean(lose_cdf_all, axis=0)[-1])\n",
    "        _ls = '-'\n",
    "        if thres:\n",
    "            _ls = 'dashed'\n",
    "            plt.axvline(x=thres, ymin=0, ymax=1, linewidth=2, color='grey', ls='dashed')\n",
    "\n",
    "        # MAJIQ wins\n",
    "        plt.plot(np.append(np.mean(win_psi_all, axis=0), 1),\n",
    "                 np.append(np.mean(win_cdf_all, axis=0), np.mean(win_cdf_all, axis=0)[-1]),\n",
    "                 # label=\"MAJIQ wins - Mean (%.2f) & STD (%.2f)\" % (np.mean(win_cdf_all, axis=0)[-1], np.std(win_cdf_all, axis=0)[-1]),\n",
    "                 label=lbl_majiq,\n",
    "                 color=rgb_to_hex(cb.Blues[3][-1]), lw=2, ls=_ls)\n",
    "\n",
    "        # MISO wins\n",
    "        plt.plot(np.append(np.mean(lose_psi_all, axis=0), 1),\n",
    "                 np.append(np.mean(lose_cdf_all, axis=0), np.mean(lose_cdf_all, axis=0)[-1]),\n",
    "                 # label=\"MISO wins - Mean (%.2f) & STD (%.2f)\" % (np.mean(lose_cdf_all, axis=0)[-1], np.std(lose_cdf_all, axis=0)[-1]),\n",
    "                 label=lbl_miso,\n",
    "                 color=rgb_to_hex(cb.Reds[3][-1]), lw=2, ls=_ls)\n",
    "\n",
    "        if not thres:\n",
    "            plt.fill_between(np.append(np.mean(win_psi_all, axis=0), 1),\n",
    "                             np.append(np.mean(win_cdf_all, axis=0)+np.std(win_cdf_all, axis=0), np.mean(win_cdf_all, axis=0)[-1]+np.std(win_cdf_all, axis=0)[-1]),\n",
    "                             np.append(np.mean(win_cdf_all, axis=0)-np.std(win_cdf_all, axis=0), np.mean(win_cdf_all, axis=0)[-1]-np.std(win_cdf_all, axis=0)[-1]),\n",
    "                             facecolor=rgb_to_hex(cb.Blues[3][-1]),\n",
    "                             alpha=0.4,\n",
    "                             linewidth=0.0)\n",
    "\n",
    "            plt.fill_between(np.append(np.mean(lose_psi_all, axis=0), 1),\n",
    "                             np.append(np.mean(lose_cdf_all, axis=0)+np.std(lose_cdf_all, axis=0), np.mean(lose_cdf_all, axis=0)[-1]+np.std(lose_cdf_all, axis=0)[-1]),\n",
    "                             np.append(np.mean(lose_cdf_all, axis=0)-np.std(lose_cdf_all, axis=0), np.mean(lose_cdf_all, axis=0)[-1]-np.std(lose_cdf_all, axis=0)[-1]),\n",
    "                            facecolor=rgb_to_hex(cb.Reds[3][-1]),\n",
    "                            alpha=0.4,\n",
    "                            linewidth=0.0)\n",
    "\n",
    "    plt.xlabel(\"Delta Delta PSI\", fontsize=11)\n",
    "    plt.ylabel(\"Number of LSVs\", fontsize=11)\n",
    "\n",
    "    plt.xlim(0, .35)\n",
    "    plt.ylim(0, .7)\n",
    "\n",
    "    plt.title(plotname, fontsize=13)\n",
    "    plt.legend(loc=2, fontsize=11)\n",
    "    plt.tight_layout()\n",
    "#     plt.savefig(\"%s/fig2_suppl1C_inset.pdf\"% plotpath, width=300, height=300, dpi=200)\n",
    "#     plt.clf()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "def plot_delta_expected_majiq_others(psi_dict_lists, replica_names, plotpath=None):\n",
    "\n",
    "    colors_dict = {\n",
    "        'miso': rgb_to_hex(cb.Paired[10][5]),\n",
    "        'majiq': rgb_to_hex(cb.Paired[10][1])\n",
    "    }\n",
    "    colors_list=['blue', 'red', 'green', 'purple', 'orange', 'yellow']\n",
    "    replica_names_joint = '; '.join([\"%s%s\" % (name1, name2) for name1, name2 in grouped(replica_names, 2)])\n",
    "    plotname=\"MAJIQ Expected PSI distribution.\\nReplicates %s\\n\" % (replica_names_joint)\n",
    "\n",
    "\n",
    "    fig = plt.figure(figsize=[10, 8]) # In inches\n",
    "    #figure out how many groups of events exist\n",
    "    max_difference=0\n",
    "    font = {'size': 10} #here also 'weight' and 'family'\n",
    "    matplotlib.rc('font', **font)\n",
    "\n",
    "    nbins=5000\n",
    "\n",
    "    for met_key, met_diff_list in psi_dict_lists.iteritems():\n",
    "        n_list = []\n",
    "        bins_list = []\n",
    "        for mm in met_diff_list:\n",
    "            n, bins = np.histogram(mm, 50)\n",
    "            n_list.append(n/n.sum())\n",
    "            bins_list.append(bins)\n",
    "        plt.plot(np.mean(bins_list, axis=0)[:-1], np.mean(n_list, axis=0), color=colors_dict[met_key], linewidth=2, label=met_key)\n",
    "        plt.fill_between(np.mean(bins_list, axis=0)[:-1],\n",
    "                     np.mean(n_list, axis=0) - np.std(n_list, axis=0),\n",
    "                     np.mean(n_list, axis=0) + np.std(n_list, axis=0),\n",
    "                     facecolor=colors_dict[met_key],\n",
    "                    alpha=0.4,\n",
    "                    linewidth=1.0, interpolate=True, label=met_key)\n",
    "\n",
    "    plt.xlabel(\"|E(PSI_1) - E(PSI_2)|\", fontsize=11)\n",
    "    plt.ylabel(\"Number of LSVs\", fontsize=11) \n",
    "    plt.xlim([0,.3])\n",
    "    plt.ylim([0,.7])\n",
    "    plt.title(plotname, fontsize=13)\n",
    "    plt.legend(loc=1, fontsize=13)\n",
    "    plt.tight_layout()\n",
    "#     plt.savefig(\"%s/fig2_suppl1C.pdf\"% plotpath, width=300, height=300, dpi=200)\n",
    "#     plt.clf()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Events in MAJIQ: rep1=13980; rep2=18015\n",
      "Events after intersection in MAJIQ: rep1=12885; rep2=12885\n",
      "Events in MISO: rep1=34369\n",
      "Events in MISO: rep2=35277\n",
      "Events in MISO after intersection with MAJIQ: rep1=9563; rep2=9568\n",
      "Events common in MISO: 9562\n",
      "ENSMUSG00000000159:96493816-96493867:target is not in all MISO replicates\n",
      "ENSMUSG00000005125:66948441-66948484:source is not in all MISO replicates\n",
      "ENSMUSG00000020265:77630726-77630761:target is not in all MISO replicates\n",
      "ENSMUSG00000026817:32629849-32629884:target is not in all MISO replicates\n",
      "ENSMUSG00000027506:8943551-8943577:target is not in all MISO replicates\n",
      "ENSMUSG00000027575:180977320-180978042:source is not in all MISO replicates\n",
      "ENSMUSG00000038807:74437431-74437463:source is not in all MISO replicates\n",
      "ENSMUSG00000040785:94410833-94410840:target is not in all MISO replicates\n",
      "ENSMUSG00000049044:72174423-72174476:target is not in all MISO replicates\n",
      "ENSMUSG00000058444:63216993-63217025:source is not in all MISO replicates\n",
      "ENSMUSG00000058444:63235289-63235318:target is not in all MISO replicates\n",
      "ENSMUSG00000061559:54731387-54731437:target is not in all MISO replicates\n",
      "ENSMUSG00000069769:88717705-88717745:source is not in all MISO replicates\n",
      "ENSMUSG00000092176:34617664-34617718:target is not in all MISO replicates\n",
      "Events after intersection of MAJIQ and MISO: 9562\n",
      "Events in MAJIQ: rep1=13980; rep2=15045\n",
      "Events after intersection in MAJIQ: rep1=12165; rep2=12165\n",
      "Events in MISO: rep1=34369\n",
      "Events in MISO: rep2=34676\n",
      "Events in MISO after intersection with MAJIQ: rep1=9033; rep2=9032\n",
      "Events common in MISO: 9029\n",
      "ENSMUSG00000000159:96493816-96493867:target is not in all MISO replicates\n",
      "ENSMUSG00000005125:66948441-66948484:source is not in all MISO replicates\n",
      "ENSMUSG00000020265:77630726-77630761:target is not in all MISO replicates\n",
      "ENSMUSG00000026817:32629849-32629884:target is not in all MISO replicates\n",
      "ENSMUSG00000027506:8943551-8943577:target is not in all MISO replicates\n",
      "ENSMUSG00000027575:180977320-180978042:source is not in all MISO replicates\n",
      "ENSMUSG00000028894:124741850-124741916:source is not in all MISO replicates\n",
      "ENSMUSG00000038807:74437431-74437463:source is not in all MISO replicates\n",
      "ENSMUSG00000040785:94410833-94410840:target is not in all MISO replicates\n",
      "ENSMUSG00000042590:106876935-106876968:target is not in all MISO replicates\n",
      "ENSMUSG00000049044:72174423-72174476:target is not in all MISO replicates\n",
      "ENSMUSG00000056486:73716364-73716419:target is not in all MISO replicates\n",
      "ENSMUSG00000058444:63216993-63217025:source is not in all MISO replicates\n",
      "ENSMUSG00000058444:63235289-63235318:target is not in all MISO replicates\n",
      "ENSMUSG00000061559:54731387-54731437:target is not in all MISO replicates\n",
      "ENSMUSG00000069769:88717705-88717745:source is not in all MISO replicates\n",
      "ENSMUSG00000070394:69838920-69838951:target is not in all MISO replicates\n",
      "ENSMUSG00000092176:34617664-34617718:target is not in all MISO replicates\n",
      "Events after intersection of MAJIQ and MISO: 9029\n",
      "Events in MAJIQ: rep1=14955; rep2=17490\n",
      "Events after intersection in MAJIQ: rep1=13105; rep2=13105\n",
      "Events in MISO: rep1=34220\n",
      "Events in MISO: rep2=35104\n",
      "Events in MISO after intersection with MAJIQ: rep1=9721; rep2=9723\n",
      "Events common in MISO: 9719\n",
      "ENSMUSG00000000159:96493816-96493867:target is not in all MISO replicates\n",
      "ENSMUSG00000005125:66948441-66948484:source is not in all MISO replicates\n",
      "ENSMUSG00000020265:77630726-77630761:target is not in all MISO replicates\n",
      "ENSMUSG00000026817:32629849-32629884:target is not in all MISO replicates\n",
      "ENSMUSG00000027506:8943551-8943577:target is not in all MISO replicates\n",
      "ENSMUSG00000027575:180977320-180978042:source is not in all MISO replicates\n",
      "ENSMUSG00000028863:125107517-125107546:source is not in all MISO replicates\n",
      "ENSMUSG00000032067:50525262-50525284:target is not in all MISO replicates\n",
      "ENSMUSG00000040785:94410833-94410840:target is not in all MISO replicates\n",
      "ENSMUSG00000041685:98749852-98749894:source is not in all MISO replicates\n",
      "ENSMUSG00000049044:72174423-72174476:target is not in all MISO replicates\n",
      "ENSMUSG00000058444:63216993-63217025:source is not in all MISO replicates\n",
      "ENSMUSG00000058444:63235289-63235318:target is not in all MISO replicates\n",
      "ENSMUSG00000059149:132041858-132041928:target is not in all MISO replicates\n",
      "ENSMUSG00000061559:54731387-54731437:target is not in all MISO replicates\n",
      "Events after intersection of MAJIQ and MISO: 9719\n",
      "Events in MAJIQ: rep1=9615; rep2=10800\n",
      "Events after intersection in MAJIQ: rep1=8589; rep2=8589\n",
      "Events in MISO: rep1=29869\n",
      "Events in MISO: rep2=30344\n",
      "Events in MISO after intersection with MAJIQ: rep1=6369; rep2=6371\n",
      "Events common in MISO: 6367\n",
      "ENSMUSG00000017119:101562297-101562319:target is not in all MISO replicates\n",
      "ENSMUSG00000020265:77630726-77630761:target is not in all MISO replicates\n",
      "ENSMUSG00000021326:21190092-21190420:source is not in all MISO replicates\n",
      "ENSMUSG00000021326:21191220-21191246:target is not in all MISO replicates\n",
      "ENSMUSG00000025188:42779738-42779978:target is not in all MISO replicates\n",
      "ENSMUSG00000025413:106671738-106671748:source is not in all MISO replicates\n",
      "ENSMUSG00000025413:106674411-106674488:target is not in all MISO replicates\n",
      "ENSMUSG00000026383:119605228-119605255:target is not in all MISO replicates\n",
      "ENSMUSG00000027746:53859616-53860273:target is not in all MISO replicates\n",
      "ENSMUSG00000028434:57068305-57068338:source is not in all MISO replicates\n",
      "ENSMUSG00000040811:19181159-19181223:source is not in all MISO replicates\n",
      "Events after intersection of MAJIQ and MISO: 6367\n",
      "Events in MAJIQ: rep1=9615; rep2=11295\n",
      "Events after intersection in MAJIQ: rep1=8540; rep2=8540\n",
      "Events in MISO: rep1=29869\n",
      "Events in MISO: rep2=30567\n",
      "Events in MISO after intersection with MAJIQ: rep1=6323; rep2=6325\n",
      "Events common in MISO: 6321\n",
      "ENSMUSG00000005125:66948441-66948484:source is not in all MISO replicates\n",
      "ENSMUSG00000007050:34982620-34983981:source is not in all MISO replicates\n",
      "ENSMUSG00000020265:77630726-77630761:target is not in all MISO replicates\n",
      "ENSMUSG00000025413:106671738-106671748:source is not in all MISO replicates\n",
      "ENSMUSG00000025413:106674411-106674488:target is not in all MISO replicates\n",
      "ENSMUSG00000026383:119605228-119605255:target is not in all MISO replicates\n",
      "ENSMUSG00000027365:126801123-126801148:target is not in all MISO replicates\n",
      "ENSMUSG00000027746:53859616-53860273:target is not in all MISO replicates\n",
      "ENSMUSG00000028434:57068305-57068338:source is not in all MISO replicates\n",
      "ENSMUSG00000037958:77076654-77076744:source is not in all MISO replicates\n",
      "ENSMUSG00000051989:92301286-92301343:source is not in all MISO replicates\n",
      "Events after intersection of MAJIQ and MISO: 6321\n",
      "Events in MAJIQ: rep1=11295; rep2=11520\n",
      "Events after intersection in MAJIQ: rep1=9904; rep2=9904\n",
      "Events in MISO: rep1=30567\n",
      "Events in MISO: rep2=30368\n",
      "Events in MISO after intersection with MAJIQ: rep1=7349; rep2=7351\n",
      "Events common in MISO: 7348\n",
      "ENSMUSG00000005125:66948441-66948484:source is not in all MISO replicates\n",
      "ENSMUSG00000020265:77630726-77630761:target is not in all MISO replicates\n",
      "ENSMUSG00000025413:106671738-106671748:source is not in all MISO replicates\n",
      "ENSMUSG00000025413:106674411-106674488:target is not in all MISO replicates\n",
      "ENSMUSG00000026383:119605228-119605255:target is not in all MISO replicates\n",
      "ENSMUSG00000028082:86075148-86075188:target is not in all MISO replicates\n",
      "ENSMUSG00000070394:69838920-69838951:target is not in all MISO replicates\n",
      "ENSMUSG00000089865:72997402-72997470:target is not in all MISO replicates\n",
      "Events after intersection of MAJIQ and MISO: 7348\n"
     ]
    }
   ],
   "source": [
    "\"\"\"\n",
    "PSI reproducibility plot.\n",
    "The order of each way within a LSV should be the same in MAJIQ and MISO.\n",
    "\"\"\"\n",
    "EXTENSION_TYPES = ['png', 'pdf']\n",
    "PLOTNAME = 'fig2_suppl1C'\n",
    "PLOTPATH = 'plots/' \n",
    "\n",
    "args = {\n",
    "    'majiq_dir': '../data/fig2_suppl1/', \n",
    "    'miso_dir': '../data/fig2_suppl1/miso/',\n",
    "    'rep_names': ['Hip1','Hip2','Hip1','Hip4','Hip5','Hip6','Liv1','Liv2','Liv1','Liv4','Liv4','Liv5']\n",
    "}\n",
    "\n",
    "psis_diff_dict = defaultdict(list)\n",
    "majiq_files = []\n",
    "miso_files = []\n",
    "names_all = []\n",
    "for rep_name in args['rep_names']:\n",
    "    majiq_files.extend(list_files_or_dir([args['majiq_dir']], suffix='psigroup.pickle', containing=rep_name))\n",
    "    miso_files.extend(list_files_or_dir([args['miso_dir']], suffix='miso_summary', containing=rep_name))\n",
    "\n",
    "for ii, majiq_file in enumerate(majiq_files):\n",
    "    if ii % 2:\n",
    "        majiq1 = pickle.load(open(majiq_files[ii-1]))\n",
    "        majiq2 = pickle.load(open(majiq_file))\n",
    "        print \"Events in MAJIQ: rep1=%d; rep2=%d\" % (len(majiq1.lsvs), len(majiq2.lsvs))\n",
    "        psi_met1_rep1, psi_met1_rep2, majiq1_names, majiq2_names = intersect_sets(majiq1.lsvs, majiq2.lsvs)\n",
    "        psi_names_met1 = defaultdict()\n",
    "        print \"Events after intersection in MAJIQ: rep1=%d; rep2=%d\" % (len(psi_met1_rep1), len(psi_met1_rep2))\n",
    "        # Discard LSVs with only one PSI\n",
    "        for i, psis_lsv_met1 in enumerate(psi_met1_rep1):\n",
    "            if len(psis_lsv_met1) > 1 or len(psi_met1_rep2[i]) > 1:\n",
    "                continue\n",
    "            psi_names_met1[majiq2_names[i]] = i\n",
    "\n",
    "        # Method1 (MAJIQ) psi scores\n",
    "        psi_list1_met1 = []\n",
    "        psi_list2_met1 = []\n",
    "\n",
    "        psi_lists_met2 = []\n",
    "\n",
    "        majiq_vs = 'miso'\n",
    "        miso_all = []\n",
    "        for miso_file_index in [ii-1, ii]:\n",
    "            miso_file = miso_files[miso_file_index]\n",
    "            miso_psis_dict = defaultdict()\n",
    "            num_events = 0\n",
    "            with open(miso_file, 'r') as miso_res:\n",
    "                for miso_line in miso_res:\n",
    "                    if miso_line.startswith(\"event\"): continue\n",
    "                    num_events+=1\n",
    "                    miso_fields = miso_line.split('\\t')\n",
    "                    if miso_fields[0] not in psi_names_met1:\n",
    "                        continue\n",
    "                    miso_psis_dict[miso_fields[0]] = miso_fields[1]\n",
    "\n",
    "            print \"Events in MISO: rep%d=%d\" % ((miso_file_index % 2) + 1, num_events)\n",
    "            miso_all.append(miso_psis_dict)\n",
    "\n",
    "        print \"Events in MISO after intersection with MAJIQ: rep1=%d; rep2=%d\" % (len(miso_all[0]), len(miso_all[1]))\n",
    "        miso_common_names = set(miso_all[0].keys()).intersection(miso_all[1].keys())\n",
    "        print \"Events common in MISO: %d\" % len(miso_common_names)\n",
    "\n",
    "        for miso_psis_dict in miso_all:\n",
    "            miso_psis = []\n",
    "            for psi_name in sorted(psi_names_met1.keys()):\n",
    "                if psi_name not in miso_common_names:\n",
    "                    print \"%s is not in all MISO replicates\" % psi_name\n",
    "                    del psi_names_met1[psi_name]\n",
    "                    continue\n",
    "                try:\n",
    "                    miso_psis_values = [float(miso_psi) for miso_psi in miso_psis_dict[psi_name].split(\",\")]\n",
    "                except KeyError, e:\n",
    "                    print \"LSV %s is in MAJIQ but not in MISO!\" % e\n",
    "                    del psi_names_met1[psi_name]\n",
    "                    continue\n",
    "                miso_psis.extend(miso_psis_values)\n",
    "\n",
    "            psi_lists_met2.append(miso_psis)\n",
    "\n",
    "        print \"Events after intersection of MAJIQ and MISO: %d\" % len(psi_names_met1.keys())\n",
    "        for psi_name in sorted(psi_names_met1.keys()):\n",
    "            psi_list1_met1.append(sum(psi_met1_rep1[psi_names_met1[psi_name]][0]*BINS_CENTER))\n",
    "            psi_list2_met1.append(sum(psi_met1_rep2[psi_names_met1[psi_name]][0]*BINS_CENTER))\n",
    "        psis_diff_dict['majiq'].append(calculate_ead_simple(psi_list1_met1, psi_list2_met1))\n",
    "        psis_diff_dict['miso'].append(calculate_ead_simple(psi_lists_met2[0], psi_lists_met2[1]))\n",
    "        names_all.append(sorted(psi_names_met1.keys()))\n",
    "# plot_delta_expected_majiq_others(psis_diff_dict, args['rep_names'], plotpath=PLOTPATH, plotname=PLOTNAME)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAsYAAAI6CAYAAAAkOI7XAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAIABJREFUeJzs3XmcZGV97/HPr3o71cswMAOKOGZGVFxicIlLjHgH446o\ncYmCC5Bc9cZ4ozFxiSaIxuUGvYnGlSSIiiJcb4yiiEuM45bEK2qIKCIoo+zrMFtPV093/e4f5/RQ\n0/RSvdQ01Xzer1e9puvUqfP8aumebz31nOeJzESSJEm6q6utdAGSJEnSnYHBWJIkScJgLEmSJAEG\nY0mSJAkwGEuSJEmAwViSJEkCDMaS1BUi4i8i4usrXUeriJiIiMdVP78wIv5zmY9/TERsa7n+0Yj4\nh+VsozruFyPiz5b7uJK6j8FY0qwiYktENCPiedO2P6rafuUM93lhddupsxzvTS3XmxHxmJbrRUT8\nVURcHhGjEXF9RJwVEfecp86tEbEnIna2XHZExMjiHvnyiYjTIuKrB6CdLRExVj322yLiBxHx7Jbb\nHxoRF0bEjdVz8/OI+Mcl1rlvIvzM/GRmPqTNWrdGxInzHjzzW5l58LT2ljT5/vT3XNXO0zLz3Us5\nrqTVwWAsaS4JXAq8dNr2l1bbZwopLwcuBv4gIqb/jZk12ERED3AB8CzgRGAYeDSwBvhuRNxjnjr/\nIDNHWi5rMnPnnI9udUngrZk5AqwDPgWcFxH3iYhh4KvAvwIbMnMN8ETguytYa8y1Q0T0dbD9OduW\ndNdlMJY0n88AD42ITQBVL+yzgbOYFjAi4gHAbwMvBA4GnrqAdk4AHgs8MzO/l5nNzNwKPB/YBZy2\nmOIj4vCq5/mFLdvOjIh/jdLGqhfxDyLisqq39bMRcWjL/oMR8e6I+EVE3FL1vB7ZcntfRLwxIn5a\n9cZeERHPiYjfA/4c2NzSi72xus8xEfHt6nhXRMRrptV9XET8pLrf54H17T7mzJwEPgT0AA8GjgIO\nAd6XmY1qn19kZtvDEiJiJCI+VtV7ZUScNO32kyPi8pbrL6jq31E9/x+ttn8euBfwj9Vj+1K1fUtE\n/G313G8H/iQiNkfE3mmlDEbExyNie/W8ndTS5n41VNv2Db+IiIurzV+p2v77lrZbv8n4jer9cWvV\ns/6mqQ95Le+XF0XEj6vH9+WIuHu7z6WkOy+DsaT5jAGfBP6gun4CsAW4boZ9XwZ8OzN/QhmoX76A\ndp4G/Edm/qJ1Y2ZOAOdVt89lxl7AzLyOMqh/ICLuHxEvqY71gsxs7b1+MXAMsAFoAp9oue0fgPsB\njwLuRtnT+oWqlxvgbVUbz616Y/8b8LPM/D/AO4Cvt/Rib42IB1L2jv81ZeA9DnhlRLwYoArd/1Qd\n9yDg74D/zvzDCKK6fz/wR8A4Ze/9ZcANwP+NiN9rDfUL8B7gSOABwNHAMymD9x2LiBgEzgZeUT0f\nmyifQzLzeOBX3N7D/5SWu/4+8J7MPAh43yyP7/eAL1F+8Ho58KGI+K056t73LUVmHl1te2LV9sum\n7xMRB1H2rn+N8rU+rqprvw8uVR3HAEcAQ8Bb56hBUpcwGEuaT1KGmlOqIPiy6vr03uKCMlyeXW06\nG3hqRBzRZjuHAtfMctt1zN1jGsAZEbGt5bLvRLDM/Brwt8D5lIHrhMy8cdox3pKZN1bDL14LPDEi\n7h4R6yk/DPxRZt5UBfW3AocDj4qIAF4B/FlmXlK1d01m/qiltumh/RXApzPz81m6DPgA8JLq9hcA\n383Mc6qe868Cn53hONOfgzdFebLaVcDxwHOqnuFdlKH+CuDNwGUR8cuI+O9zHO/2A5e9pScCf1k9\nRzuA181zt3HgARFxSGbuyczvzLN/Uj4nWwAyc88s+/17y/PyNcoPECfP9xDmub3VcUAjM9+emXsz\n86eUH2CmP1dvycxbq/fLOcBvLqANSXdSBmNJ88rMHwO/BE6lDKhfmmG351H2nP3f6vrXgeu5Y6CY\nzY3AbCfZ3YOZe6j3lQi8LDMPbrlMPxHsDMqv8H8wFb6m2dry8y+rf+9J2dsJ8F9ToRu4Beil7F0+\nlPJx/2yO+qbbBJzQGuQpn9upr+Pv2VLDTPXNJIG3VY/9bpn52My8YN+Nmb/KzFdn5oMoe1s/APx9\nRBzbRr2HAgPTapi1nswcpeyVfwpwRURcFBEntNHOrMecOjR3fF5+yezvm8XYMEMbv6i2t2p9P44C\nK36ip6SlMxhLatffA38BfGTaEIQpL6MMi5dGxHWUvb/rKU/Ca6fH7kuUPbD3bt0YEb2UX1t/YbGF\nVz2eH6+Ocb+IOGWG3Ta1/Lyx+vdqbg9J95kWvIcz8zzgJspgdL9Zmm/OsG0r5fPYeryDMvPBLe1u\nnHafjSxxRoYpmbkzM08HbqUcFjGfmyl7gGd6jmZr4xuZ+UzKEwHfBnwiqnHqzPyczLV9SszQ7kbK\nHnKAnZQfUlodwf7P23zP4VXAr03bdm/K4R+SVjmDsaS5tA4D+BTlTAZ/d4edyjGzv005o8TRLZdH\nUvaCHtdGW+cA/w58NiIeERE91Ylqn6LsrXxbG7XO5i8oe51fTDks4j0R8aDp+0TEYRGxhvKr869m\n5vXVkItzKMey3qN6vGsj4ncjYqj6kPBB4PSIeFCU7hkRUyH3OuBesf8sCx8EXhART4/yxL3eiHhg\nVHMCA+dSfkh4QXXbEyjH9M5nxucgIo6K8uTA+1TPaxERfwisBeYb4jB1Mt85wFtanqP/NWsR5T7P\niYiDqudnO2Ugnax2uZ6ZP0i08wHq0dXz0hMRj6c8EfRj1W3/CRwW5YmLtYj4XcpxwK1marv1fX4B\nMFA9X30RcRTlsJEz26hNUpczGEuaS+uJS43M/NfMvG36bZQnQX0/My+oxqBOXX4EfJqyN7n1mHds\nqAxfT6EcB3wusJvyK+xDgUdl5g3z1Do1y0HrPMYPqsLTnwLPq8a6fhM4Hfh0RNRb7v8J4FuUPYO9\nlCF6ykspT2DbEhE7gP8CntPyWN4E/B/KccA7KIeRTJ3g9mnKXsjrq1kOfq0amvJ04NXAtZQnxp1F\nNY46M38OPJdyeMU24FXAvjmH5zBbb+hO4IHAV4DbKHukX1g9J99rue9cvamvAq4Efkp5Qt/5wMS0\ntqfuX6McR31l9Xy9HzgpM6d6Xd8GvKh6Pi6Ydoy5HlNy+4mYt1KOdX9FZv477HveXkX57cYtwJMp\nxyC3ehPw1qrtD02vPTO3A08CnkD5unyJMnj/7Rx17vfcRcSHI+KLMzwWSXdyMfM3opK0/CLiB8CH\n2p0mrBry8B7gdzLzog7VtJEygN8zM6/tRBuSpO7Qu9IFSLpriIiHAA8C/qPd+2TmWRGxm3JGhY4E\nY0mSphiMJXVcRJxLOQb5z1umMWtLNRdwp/nVmSTJoRSSJEkSePKdJEmSBBiMdScVEc2IeEz18zHV\nAgirRjUV1Pl3lXbbUc0k8ai7Srvz8T1SiogPRcRMy0N3oq0vRsSfHYi25hMRL2xdvbHDbd1pXvOI\nuFf1O3n3+fdeclur7v8WLYPM9OJlQRdgCzBGOQXUdsqpm164zG00gcd0oPbTKOen7dRzsxnY2+72\nBRz3aOBCyjlxm8BvH6B2X0I5z+2tlAtZfBH49UW8X97U7vYFHPdtwA+AxkyvaafabTnO4dXzcvmd\n5D1ycvXe2Nly+WSn262OcSTwz5RTwd1GOR9171LfI4us5R+BS4C9wD/McPvW5fh7BTyacs7j66vH\nfBHwzAUeY9n+HgF/DHyXcprDO7wngY/O9Hwssc1l/R1YhnrOY9r/HdXvxYLqm+P4H6Vc6Kb1d+x/\nLPfj8LKyF3uMtRgJvDUzR4BDgI8AH4uI+69sWatag3Kp5adX1w/UyQHDlHPpHlFdfgB8Zdr8v/OZ\nbX7c+ebNnc8VwF9Szlm7kOMvtd0pZwDfX6ZjLZcrMnOk5fLCTjcYEYdSzv/8Q8plkw8G/ojbF/No\nx3K9JlB+UP8TynmWO/n6H0y5+MwDM3Mt8FfApyLiN5fh2ItxDeWiK2+f5fblfI6n3Gl+ByLi2ZT/\nH3VSAh+d9jv24Q63qQPMYKwlyXJRhn+kfC/tW0ksIp4VEd+PiG0R8ZOIOLHltpMj4oqIeF1EXBsR\nN0TEu6ulf+8gIjZHxN5p214WEf8VEdsj4lcR8UfV9qMj4hsRcVM1gf8Xo1piOCKeD/w5sLllAYiN\n1W3HRMS3I+KWqrbXtLR1cER8OiJujojbIuKSiHjsUp63iDgtIr7acn1rRPxlRHyrqu17rf/BZuZP\nM/PMzPz+AW73g5n5tSwXxhin7KW9O3BUyzG2RcQLllJXdZzW4TPzvkcy86OZeQHlQg7trJi2LO1W\n+72Y8j3/ieltR8QbIuLixdbTcpwFvVZTu61Au68BtmbmW7Ncajoz8weZ2brgxaLfIxHx0Yj4h+rn\nd0XEP0+7fXP1u1wHyMz3ZeZXKRdaWdDzERFbIuJN1c+fjoi/nXb7yRFxRdXOhZn5icy8tbr+OcpQ\n/tiW/T8fEe9f6GNuaevy6ufjqvdib8vtwxGxKyKOqdr/p8z8Z8oFY2Y97Cxt7XvNI+KPIuKH027f\nFBETEXGvlm0d+R2IiI3V7+Q9IuKQiBiLiKOn7bMlIk5tub4OeBflQjwLaWvf/y1RLgbUiIj1LbdH\nRPyieqyw/wqJWqUMxlqsAIiIfuAPKXs0v19teyJlWP7jzDwYOAl4/9Qf8Mq9KHuXNgG/BRwPvLat\nhsulbN9M+RXWQcBDKb9ChPJrtFMpl//dCOyi/MNNZp4HvAP4evVJf01mbo1yOeMLKJcBXk+5fPEr\nW/4YvhaoA/eqeoaeRbly2HJ7OeXXoQdT9g5/MSJGOtDOUtr9Haqvaqc2ZObBmXnuPG3M9J/JfP/B\nLPo90sl2oxz7+FfA/5jpWJn5vzLz6Onbl8lcr1UCGyLiuig/LH4qqg9+HWp3uLrtWODqiPhClB8s\nL46WD8LQ9ntkNq09nR8BntYaXoBTgPMyc88ijz9XWydO+1B0CuUKhXdQvS8eRBmOy4NlHp+Zr1yG\nur5Eucpg69LqzwOuzcxvLcPxW50D3H9aGD2Z8u/mr+DA/Q5UHzo+V7VP1fa9Kad+bH0d3g/8XWZu\nXUJbP6ZcUrz1W5bNwDrK1SuhfG88p3qfXxYRp0fE0GLb1J2TwViLEcCbojxpYRR4K/D0lj9KrwLe\nm5nfAchyydlPUo5XndIEXpvlMsO/oFyi9+Q22/+fwNsy89+q49+S1apomfmjzPxGZu7NzB1VbY+O\niKKl9ul/yF8BfDozP1/1dl0GfKCl3gblH8f7R0QtM6+Y5w9wT9VDtu8CfJ65v25M4MzM/GFmTmTm\nXwN7uH3oRDs62m5E3I8yLPxpZu5eQF373i/Tavvtee63lPdIJ9v9MHB6Zi7lw1EnXqtvUo7/Phx4\nBOV5AF+NiMEOt7seeDble+NQyuW3z4yI+Z7nhQiAzLyUcsjGiwCqDwXPqdpebl+hDKNPr9o6EngM\n5TjT/Ysrw9E/AV/IzK8vdyHVN3NnUwbzKbOG9CW2tY0yjJ4CZa8pZedG63O8HL8D7ToLeGFE9FTX\nTwb+NTOvqup7FrAxM9+7TG1Nf47Pzcyx6vr7gKMycx3wu8B/o1yWXKuIwViLkZTB9GDK/xQvpOyl\nnbIJeP20/3xPojxRY8qNLX9sAH4J3LPN9n8N+NlMN0TEkRHxmYi4OiK2A9+u6j10juNtAk6YVu+p\nlEMGoPyK7mvAx4Abo/xq97A5jjdZ9ZDtu1D+5zpfT+XWadd/RTmut10da7fqVf9X4F2Z+fcLqAla\n3i/Tavv2PPdbynukI+1WPaHrgQ8toI6ZLPtrlZlXZubU1/w3UH6tfDjlqoGdaHfqtdgJ/FtmfiYz\nm5n5L5Q9nM+Y55iLdRa3f1D5PeCqzPz35W6kCqMf5/agdDLwL5l5Tet+VTi/kPIkvNYP/8vtLOCp\nEbG+Cum/Rfk3qVNtTfWWPx44CPgMLOvvQLu+SnnC2zOmh/SIOAT4O6ohFNXtsPjhDucC94uIh1Sv\n69QHPgCqIUI3VT//BHg18NyI6Ftke7oTMhhrSTLzNuC/Aw+OiKmvoLYCb572H/CazGzthTws9j+B\nayNwVZvNbgXuN8ttH6acKePB1TCLx7J/L3FzluN9ZFq9B2Xmg6vHOJqZf1FdfxBlEHlXm7VOaecP\n9aZ9O5d/4O/F0odsLLndiHgY8HXgHZn57iXWsxDtvkeW+8Sfudp9IuUMITdGxE2U/ylvinJM+4OX\n2G6n3iPzHXcx7f5aS7s/nGH/YObfteVwHmV4eShlWJ2r13Sp742PAU+phg68ZHpb1djWr1E+F8/L\nzIkltjer6pus7wMvpnzcX83MucYTz3iYNvf7F8pvyo6v2vpUZjaq2zr5O3AHLR9QTqYM6SOUM6AA\n/Ablh7+vV7XcWG3/QkS8YxFt3QZ8lvLD0O8Bv8zM7859L8Bxx6uKwViLte8PQfXV298Ab66+7noP\n8JqIeGxE9EREf0Q8PCIe3nL/GvDXEVFUY8b+lPZ7Pz4AvDEiHh0RtaoHZepkoBHK4R3bq3GIb512\n3+uAe037hP9B4AUR8fSI6IuI3oh4YEQ8DiAijo+IB1SPbTflfxjL/R9gAKdExEOr2qbGNV+wb4fy\nuZoKbAPV9aX+Ds/ZbvV1+NeAN2bmB2c8QHmizHw9ZbP9xzHXfyhzvkeq16kA+oBaRAxExECH2/0T\nyhMPj64up1L2nh4NXFrVdVpEXDnH8RdrvtfquIi4Z5QOoXxf3wT8RwfaLbj9vXkG5XClZ1a/j8dS\nhqfP7jvA/O+RAPqq53zfpeW2farw8s+Usy88iml/N6q/NwXQC/RW74v+ae31T2tr6n2z31CrzPwp\n5TRsHwGGuD2QTY2z/QbwE+BFmXmHDwJRniQ233CHqfduaz2zvT/PAv6AMhzvN3yk+ls79fsQU8ds\n3aXl+Zj+HO+nJYy+inLIQGtby/Y7MMPj7pll148CTwVeB5yT5YnAAP9G+SFtqpaHVNtPppyho6Wp\nO7Q129/OsyjHGb+UO34QekFEHFT9fF/gfwOfa6lHq4DBWIs1vefhvZTjcF+c5RnhL6XsVb2J8izp\n/035H8uUX1L2slxJ+R/3hZRjOedtrwpo7wTOpOwd/j4wFYz/BDiG8oz0b3DH8ZOfpuz9uz7KWSt+\nrTrp4umUX4tdC9xA+Qdx6gSfe1NO/bS9qnc38Pp2ap1j+/Spk5Jy2rG/o5wX9HnAcZm5E8oztSkD\n/+5q369V11/UyXYpT7AZAd4T5YwEU5ffruq6F+WHhO/M0vZCaptuvvfIP1I+B2+kPAFsD9V/zJ1q\nNzNvy8xrpy7ANmCiuj71YelelD3sc+nEa/XfKE9C3Uk5j+9a4ImZOdrJdqsetRMpT17dQfm34CVT\nPW1tvkeS8oTa0ZbL7oi42wz1QPn7+RTgS9WwkVZfqe7/QsqAtIdyaEerj0xra2qRh7naOiczW2fH\neTnwQOC5lB/Ep3433tCyzwbmfi8kt793W+t53iy1nEvZez9EOQ641V9W9z2j2mcP5d+L1rZOntbW\n7igXt5ntcT8O+EVW53DAsv0OJNAzw+N+30y1ZObPgP8HPIH9hzaMt9bSMszlpizPMZlq694ztPVn\nszzuf6F83h5G+eGg1cuBX0TELuDLlMH8FLSqROZsfyelzoiIkykn87/vStdyZ1H1rrwpM8/ppnYj\n4hTgYZn5P5e5rpNZgffIcrQbEZcBj58+FnWpfI90l+rbhs9k5kPm3XmV6dTvgHQgzDhvrCS1IzPP\nogNnxnezzDxq/r3uOu6q75EsZzS5y4Vi8HdA3c2hFFoJnViBSavLSr1HfG9K0l2YQykkSZIk7DGW\nJEmSAIOxJEmSBBiMJUmSJMBgLEmSJAEGY0mSJAkwGEuSJEmAwViSJEkCDMaSJEkSYDCWJEmSAIOx\nJEmSBBiMJUmSJKDDwTginhIRP42IyyPi9TPc/mcR8cPq8qOImIiItZ2sSZIkSZpJZGZnDhzRA1wG\nPAG4BvgecEJmXjrL/k8HXp2ZT+hIQZIkSdIcOtlj/Ejgiszcmpl7gXOBZ86x/4nApzpYjyRJkjSr\nTgbjI4CrWq5fXW27g4gYBJ4M/FMH65EkSZJm1clgvJAxGscD387M2zpVjCRJkjSX3g4e+xpgQ8v1\nDZS9xjN5AXMMo4iIzgyEliRJ0qqRmbGU+3fy5LteypPvfge4Fvh/zHDyXUQcBPwCuGdm7pnlWNmp\nOtUZp512GqeddtpKl6E2+Xp1H1+z7uNr1n18zbpLRCw5GHesxzgzJyLilcCXgR7gzMy8NCJeXt1+\nRrXrs4AvzxaKJUmSpAOhk0MpyMwLgQunbTtj2vWPAR/rZB2SJEnSfFz5Th2xefPmlS5BC+Dr1X18\nzbqPr1n38TW76+nYGOPl5BhjSZIkzWU5xhjbYyxJkiTR4THGkiRJul3Ekjo0VenUSAKDsSRJ0gHk\n8NCl6eSHC4dSSJIkSRiMJUmSJMBgLEmSJAEGY0mSJC3ByMgI3/3ud1e6jGXhPMaSJEkHSDXX7kqX\n0dVmew6dx1iSJElaJgZjSZIk7bNx40be/va3c+yxxzIyMsLRRx/NJZdcwtlnn8197nMf1q5dy0tf\n+lImJycBqNVq/Nu//RsAW7du5clPfjIHH3wwhxxyCA9/+MP52c9+BsDExARvfetbOfLIIznkkEN4\nwhOewI9//OMVe5wzcR5jSZKkO4mrj9iwrMe75zVXLep+H//4xzn//PM58sgjOeWUUzj++ON50pOe\nxI9+9CNuvvlmfvM3f5PHP/7xnHDCCfvd741vfCMbN27kC1/4Ar29vfz4xz/m4IMPBuBd73oXZ599\nNhdeeCEbN27kne98J0984hO57LLLGBkZWfJjXQ72GEuSJGmfiOBlL3sZRx11FL29vZxwwgn88pe/\n5B3veAf1ep0NGzawefNmLrroojvcd2BggOuuu46f//znRAS//uu/zqGHHgrAWWedxRve8Abud7/7\n0d/fz6mnnkpPTw8XXHDBgX6Is7LHWJIk6U5isT28y+3www/f9/Pg4CA9PT2sW7du37Z6vc7OnTvv\ncL93vetd/NVf/RXHH388u3fv5rnPfS7vfOc7GRoa4uqrr2bTpk379o0INm7cyNVXX93ZB7MA9hhL\nkiRpQWZblnn9+vW8973v5fLLL+c73/kOW7Zs4fTTTwdgw4YNXHnllfv2bTabbN26lQ0blnf4yFIY\njCVJkrQgs005d95553HllVeSmaxZs4b+/n56e8sBCieffDKnn346l19+OePj47z97W9ncnKS4447\n7kCWPieHUkiSJGlO03uIZ+sx/s///E9e97rXccsttzAyMsIznvEMXvva1wLw2te+lkajwZOe9CS2\nb9/OQx/6UL7yla8wPDzc8frb1TULfOzdu3ffJw5JkqRu5AIfS+cCH8DY2NhKlyBJkqRVrGuCcaPR\nWOkSJEmStIp1TTC2x1iSJEmdZDCWJEmSMBhLkiRJgMFYkiRJArooGO/cPbrSJUiSJGkV655gPLpn\npUuQJEnSKtY1wXjXLoOxJElSN/nVr37FyMgI119/PQCf/OQnechDHrLCVc2ua1a++/uPncNLX3LC\nSpciSZK0aK58t3SufAfsdiiFJEmSOqhrgvEug7EkSVLHbdy4kbe//e0ce+yxjIyMcPTRR3PJJZdw\n9tlnc5/73Ie1a9fy0pe+lMnJSQBOOeUU7nWve7FmzRoe9KAH8alPfWrfsbZu3UqtVuPaa68F4KMf\n/Sj3ve99992+c+dOTjrpJNatW8emTZv42Mc+Rm9vL9/85jcP7IOudE0wHpto7nsBJEmS1Dkf//jH\n+fCHP8y2bdv4jd/4DY4//ni+/e1v86Mf/Ygf/ehHnH/++Zx33nkAHHPMMVx88cVs376dU089lZNP\nPplLL720rXZe/epX8/Of/5xLL72Uiy++mM997nM0m81OPrQ59a5Yyws0XhtgbGyMoaGhlS5FkiSp\nIzb++QXLeryt7zxuwfeJCF72spdx1FFHAXDCCSfwyU9+kne84x3U63U2bNjA5s2bueiiizjxxBP5\n/d///X33ff7zn8+73/1utmzZwgMe8IA522k2m5xzzjl88Ytf5LDDDgPg9NNP57Of/eyCa14uXdNj\nvLdWuMiHJEnSAXD44Yfv+3lwcJCenh7WrVu337Zdu3aRmZx66qnc//73Z+3atRx88MFcfPHF3Hzz\nzfO2cdNNN9FoNNi4ceO+ba0/r4Su6TFuRJ/BWJIkrWqL6eFdKZnJOeecw5lnnslXv/pVHvjABwLw\niEc8oq2ZN9avX09/fz9XXnklmzZtAsoxySupa3qMx+in0WisdBmSJEl3eVPBd+fOnfT29rJ+/Xom\nJib4yEc+wsUXX9zWMXp6ejjxxBN585vfzI033siOHTt4wxve0Mmy59U1wbiRPfYYS5IkrYCIuMP1\niOCkk07iUY96FPe5z3245z3vyaWXXsrjHve4OY/Teqz3vve9bNq0ifvf//4cffTRPOMZz6Cnp6dj\nj2M+XbPAx++++/P8xbH34GEPe9hKlyNJkrQod7UFPn7+859z3/velxtvvJH169e3dZ++vj6+9rWv\nzRqwXeADGM8edu4aXekyJEmS1KYf/OAHrF27tu1QvNK65uS78WaNXXtc5EOSJKkbvOY1r+G8887j\nfe9730qX0rauGUpxzDu/zCvuN8YLnv2MlS5HkiRpUe5qQyk6waEUwNhksnvUoRSSJEnqjK4Jxnv2\nJrtGHUohSZKkzuiiYNxk96jTtUmSJKkzuiYYTzSTXY0Jms3mSpciSZKkVahrgjHARE/hIh+SJEnq\niK6Zrg2gEf2MjY0xODi40qVIkiQtyvRV5HTn0VXBeIwBGo3GSpchSZK0KE7VdufWVUMpGtHrUApJ\nkiR1RFcF47E0GEuSJKkzuisYN3sMxpIkSeqIrgrG49nDbhf5kCRJUgd0VTBuZI2duw3GkiRJWn7d\nFYybwc7doytdhiRJklahrgrGYxOwa9RgLEmSpOXXVcF4z0STXY4xliRJUgd0VzDe22TXbmelkCRJ\n0vLrmmCfIaLkAAAgAElEQVQ80FujmbCzMUGz2VzpciRJkrTKdE0wHuwrS21Ev8tCS5Ikadl1TTCu\nV8F4vDbgIh+SJEladh0NxhHxlIj4aURcHhGvn2WfzRHxw4i4JCK2zHasoqf8txH9BmNJkiQtu95O\nHTgieoD3A08ArgG+FxHnZ+alLfusBT4APDkzr46I9bMdb6BWjitu0GcwliRJ0rLrZI/xI4ErMnNr\nZu4FzgWeOW2fE4F/ysyrATLz5tkO1h9lMB5r9jrGWJIkScuuk8H4COCqlutXV9ta3Rc4JCK+HhEX\nRcSLZztYf0wC0KDXHmNJkiQtu44NpQCyjX36gIcBvwMMAv8eEf+RmZdP37EeE0AfY80au1wWWpIk\nScusk8H4GmBDy/UNlL3Gra4Cbs7MPcCeiPgmcDRwh2B8yZfO4bbdvVxW1PhO7aE89rcf06m6JUmS\ndCe3ZcsWtmzZsqzHjMx2OnYXceCIXuAyyt7ga4H/B5ww7eS7+1OeoPdkYAD4LvD8zPzJtGPlH/3N\nOXzhxjUcPtzDn/76BM995tM7UrckSZK6T0SQmbGUY3SsxzgzJyLilcCXgR7gzMy8NCJeXt1+Rmb+\nNCK+BPwX0AT+YXoonjKQ5Ql3Y5Ow06EUkiRJWmadHEpBZl4IXDht2xnTrr8bePd8x+qfLE+427M3\n2T26ZxmrlCRJkrpo5bu+3EsAYxNNdu42GEuSJGl5dU0wHq73U+8vl7+7bWySTo2NliRJ0l1T1wTj\nocE69d5yPPU4/S7yIUmSpGXVNcF4ZGiQogrGjdqAi3xIkiRpWXVNMB4erFP0lMMnGtFnMJYkSdKy\n6ppgvGZ4iIGogjF9DqWQJEnSsuqaYDwyNEh/TAIwlr32GEuSJGlZdU0wHhysM7AvGPcYjCVJkrSs\nuiYYF0XBQEwAMNasuciHJEmSllXXBON6vU499gIw3qyxy2AsSZKkZdQ1wbgoCgZyHIBGM9i5e3SF\nK5IkSdJq0lXBuL9Zjisem0h2uSy0JEmSllHXBON6vU5fs5yibc9EsmvUHmNJkiQtn64Jxj09PQz1\n1ugJGJ9MtttjLEmSpGXUNcEYYGS4Tr2/B4Db9kySmStckSRJklaLrgrGw4N16r0BwBh9jI+Pr3BF\nkiRJWi26KhiPDA1RlB3GNKLfRT4kSZK0bLosGA8y0FMOn2hgMJYkSdLy6bJgXKefJgCN6DMYS5Ik\nadl0VTAeHhqkqE0CMJa9NBqNFa5IkiRJq0VXBeN6vc5ATADQaPbYYyxJkqRl01XBuCgKCspgPJY1\ndo86l7EkSZKWR9cF4wHKKdrGmzV27Xb1O0mSJC2P7gvGWQbjsSbsNBhLkiRpmXRVMK7X6/RPlifc\njU3A7j2OMZYkSdLy6Kpg3N/fT1Erxxjv2dtk527HGEuSJGl5dFUwBlg7VNDXE0wm3GYwliRJ0jLp\numA8NFin3luWfevoxApXI0mSpNWi64LxyNAg9d4AykU+xsfHV7giSZIkrQbdF4yHByl6EoBGbcBF\nPiRJkrQsui8YD9bpr5XBeJw+g7EkSZKWRdcF4zXDQwzEJABjBmNJkiQtk64LxvV6/fZgnL00Go0V\nrkiSJEmrQdcF46Io9s1l3Mgee4wlSZK0LLozGOdeAMaaNUZHnctYkiRJS9d1wbherzNAGYwbzWDn\n6OgKVyRJkqTVoOuCcVEU9DfLccWNyWD3qEMpJEmStHS9K13AQhVFwUCWi3qMTSY7XRZakiRJy6Dr\neowjgoOKsuw9ew3GkiRJWh5dF4wBDhoepOgNErh1l8FYkiRJS9eVwXh4cJCityz9ltGJFa5GkiRJ\nq0FXBuORoTpFNTp6rNnD3r17V7YgSZIkdb0uDcaDFLUEYCwGXORDkiRJS9aVwXjN8BD90QRgnD6D\nsSRJkpasK4Px4GCdojYJQCMMxpIkSVq6rgzGRVEwEGUwHsseg7EkSZKWrGuDcRHlbBSNZi+NRmOF\nK5IkSVK368pgXK/Xb1/9rlljzx7nMpYkSdLSdGUwbl0WupHBjt2jK1yRJEmSul3XBuO+Zjl8Ymwy\n2OWy0JIkSVqi3pUuYDF6e3sZ7q3mMZ5Ido168p0kSZKWpit7jAHWDRcE0JhMtttjLEmSpCXq2mC8\nZrhOva8s/+adzkohSZKkpenaYDw8OEhRDQTZNjq+ssVIkiSp63VtMF4zPETRU/48Ollj7969K1uQ\nJEmSulrXBuORoToDtfIEvPHagIt8SJIkaUm6NhgPDw0yUGsC0Ig+l4WWJEnSknRtMC6KgoJqWeg0\nGEuSJGlpujYY1+t1ipgEoJE9BmNJkiQtSdcG46IoGKA84W4sexxjLEmSpCXpaDCOiKdExE8j4vKI\neP0Mt2+OiO0R8cPq8hftHrsoCvopp2kba/awZ4+LfEiSJGnxOrYkdET0AO8HngBcA3wvIs7PzEun\n7fqNzHzGQo9fFAUDzbKXuNEMdu4eXWrJkiRJugvrZI/xI4ErMnNrZu4FzgWeOcN+sZiDDwwMUER5\n8t3YJOwctcdYkiRJi9fJYHwEcFXL9aurba0SeExEXBwRX4yIBy6kgXVDfQDsmUh27/bkO0mSJC1e\nx4ZSUIbe+fwA2JCZoxHxVOCzwP1m2vG0007b9/PmzZvZvHkza4fq9ARMNGHbLodSSJIk3VVs2bKF\nLVu2LOsxI7Od/LqIA0c8GjgtM59SXf9zoJmZfz3Hfa4EHp6Zt07bnjPV+Yn/8xn+14/r7Bpv8vtH\n3MSprzx5WR+DJEmSukNEkJmLGqI7pZNDKS4C7hsRGyOiH3g+cH7rDhFxt4iI6udHUgb1W+94qJmt\nGR6i6Cl/vnX33uWqW5IkSXdBHRtKkZkTEfFK4MtAD3BmZl4aES+vbj8DeC7whxExAYwCL1hIGyND\ndYqeXQDsngwmJibo7e3k6BBJkiStVh1NkZl5IXDhtG1ntPz8AeADiz3+muEh+mMHEIxHP41Gw2As\nSZKkRenale+gnMu4qDUBaNDnstCSJElatK4OxvV6nSImARgzGEuSJGkJujoYF0VBEeVJdw16DcaS\nJElatK4Pxv2UwXisWTMYS5IkadG6OhjX63UGmg2gDMaNRmOFK5IkSVK36upgPDAwwECOA9BoBjt2\n7V7hiiRJktStujoY12o1DqpW+GhMBrtG96xwRZIkSepWXR2MAdYN9QOwZyLZsctgLEmSpMXp+mB8\n8MgQfT1BM2HbboOxJEmSFqfrg/HwUJ16bwBw805PvpMkSdLidH0wXjM0SFFLAG4d3bvC1UiSJKlb\ndX0wHhkaZKAKxrv2wuTk5ApXJEmSpG7U9cF4aGiQgVoTgEb0O5exJEmSFqXrg3FRFBS1spd4PPpd\n/U6SJEmLsjqCcUwA0KDXYCxJkqRFWRXBeIDypLuxZo/BWJIkSYvS9cG4Xq/TXy0LPZY9jjGWJEnS\nonR9MC6KgoGc6jGu2WMsSZKkRen6YNzb28tQTzUrRTPYuWt0hSuSJElSN+r6YAywbqgPgLEJ2DXq\nstCSJElauFURjNevqQMwNgnb7TGWJEnSIqyKYHzQ0CBFbwBwyy57jCVJkrRwqyIYjwwNUvSUP9+8\na3xli5EkSVJXWhXBeM3wIAM9CcCto3tXuBpJkiR1o1URjIeHBilqZTDeOd6k2WyucEWSJEnqNqsi\nGJfLQk8CME6/i3xIkiRpwVZNMB6ognEj+l3kQ5IkSQu2KoJxvV6niAkAGvQajCVJkrRgqyIYF0VB\nP+VsFGPNHoOxJEmSFmzVBOOBZhWM02AsSZKkhVsVwbi/v3/fUIqxZs1gLEmSpAVbFcE4IjhksBeA\nxmSwe9TV7yRJkrQwqyIYA6wbLgAYm4Sduw3GkiRJWphVFYwDGG/CbbtGV7ocSZIkdZlVE4zXDA9S\n7w0Abtxhj7EkSZIWZtUE45HhQYpymDG37h5f2WIkSZLUdVZNMF4zNMhAJAC3ju5d4WokSZLUbVZN\nMK7X6wzUymWhd44nmbnCFUmSJKmbrJpgXBQFRa0JwFi6LLQkSZIWZtUE43q9vm+Rj0b002g0Vrgi\nSZIkdZNVE4yLomCAKhhjj7EkSZIWZnUF4yx7iRsOpZAkSdICrbJgXE7TNtasGYwlSZK0IKsmGNdq\nNdYU5cMZa9YcYyxJkqQFWTXBGOCQwX4AGs1g126XhZYkSVL7VlUwPnSkAGBsAnbudlloSZIktW9V\nBeNDRgbpCZhIuHWnPcaSJElq36oKxmuGB6n3BgA37XSMsSRJktq3uoLx0CADPeVS0DfvclYKSZIk\ntW9VBePBwfq+ZaG3je5d4WokSZLUTVZVMK7X6wxUwXh7I8nMFa5IkiRJ3WJVBeOiKChiEoAGPc5l\nLEmSpLatumA8QDmEohH9BmNJkiS1bfUG4+x1WWhJkiS1bVUF43q9zkBzHDAYS5IkaWHmDcYR8acR\nsbb6+eyIuCwintz50haut7eXod7yhLuxZs1gLEmSpLa102N8cmbeFhHHAocBvw+8o7NlLd7awT7A\nYCxJkqSFaScYT1b/Hguck5nfAaJzJS3N+uF+AMaawegeg7EkSZLa09vGPnsi4g3AicBjI6IG9He2\nrMU7bM0gAGMTsGPX7hWuRpIkSd2iraEUwOHA6zLzemAT8Ml2Dh4RT4mIn0bE5RHx+jn2e0RETETE\ns2fbZ2Jiop0mOWRkkL4aJHDTjj1t3UeSJEmaNRhHxKkRsSEzL8vMV2XmZwAy8+eZ+c75DhwRPcD7\ngacADwROiIgHzLLfXwNfYo4hGjt27Jj3wQCMDNUpesqfb9rpUApJkiS1Z64e43sDl0TElyPi+RGx\n0OETjwSuyMytmbkXOBd45gz7/U/g/wI3zXWw7du3t9Xo8NAgRU85M8Utu1zgQ5IkSe2ZNRhn5snA\nEcB5wCuB6yLifRHx0DaPfQRwVcv1q6tt+0TEEZRh+UNTzc52sHaDcb1ep6g1Adg22t7wC0mSJGnO\nMcaZuSszP5KZxwCPBnYAn4uIH7Zx7FlDbov3AG/IzKQcRjHrUIrbbpqzQ3mfoigoauVEGtsbk5SH\nliRJkubWzqwUU/ooZ6PoB9oZo3ANsKHl+gbKXuNWDwfOjQiA9cBTI2JvZp4//WAf/uAH2fKd7wCw\nefNmNm/ePGOj5bLQZTAey17Gx8cZGBhoo1xJkiR1iy1btrBly5ZlPeacwTgiDgJOAE4Bfg34BPD4\nzPxJG8e+CLhvRGwErgWeXx1rn8y8d0tbZwGfnykUAzz1scdw0p++Zt5G6/U6A7EXqNOIPsbGxgzG\nkiRJq8z0jtK3vOUtSz7mrME4Ij4FHAdsAd4JXFCdRNeWzJyIiFcCXwZ6gDMz89KIeHl1+xkLKbSx\nYzsTExP09s7dyV0UBQM5Xt4ne2k0PAFPkiRJ85srZf4n8CfV3MWLkpkXAhdO2zZjIM7MU+Y61lCt\nhx07dnDIIYfM2WZ/fz8F5Ul3Y9nrstCSJElqy1wn350F7Jy6EhHHRMR7I+JV1dzDB9RIrdbWzBQR\nwdp6Wd5Y1gzGkiRJastcwfifgbsDRMRRlD2/fcCzgHd3vrT9jZBtT9l2yFA5prgxaTCWJElSe+Ya\nSrE2M39e/XwC8OnMfEVEFMD3O1/a/kb6B9qesu3QkTIYjzWDPXsMxpIkSZrfXD3G4y0//xbwLwCZ\nOQa0fRLecllTr3Pb9e0Ndz5sTR2AxiTctnN3J8uSJEnSKjFXMN4eEU+LiIcAjwH+FSAiakBxIIpr\ntWZwkO03ttdjvGZ4kKIaBX3jjtEOViVJkqTVYq6hFK8CzqFcxvmtmXldtf144HudLmy6oYGBtqds\nO2h4iIGeXYxNwo07na5NkiRJ85s1YWbmxcCDZtj+uYj4RUermkFEMFTNTLFu3bo5963X6xS1Jtup\nccvu8Tn3lSRJkmDuoRRzuWBZq2jTSDWX8XyKoqColctCbxs94MOhJUmS1IUWG4xjWatoU7tTthVF\nwUCUwXj72GSny5IkSdIqsNhgvCLanbKtXq9TVMF4rNnD+LjDKSRJkjS3WccYR8RxQM50EyswKwXA\nQYODXN3GlG1FUTBQzSjXiH4ajQb9/f2dLk+SJEldbK7pHV7LzMEY4JIO1DKvkXqd2268cd79iqJg\nIMte4kb2MjY2xsjISKfLkyRJUheba1aKzQewjrYMDQwwvmPHvFO21Wo1hvvKYdBj2eOy0JIkSZpX\nV40xbp2ybT6HDJbBuZE1g7EkSZLm1VXBGNqfsm39cDkMemzSYCxJkqT5dV0wXhO01WN86JoqGDfD\nYCxJkqR5zRqMI+JT1b+vPnDlzG+4r7+tE/AOO2iQGrC3Cbfs3N35wiRJktTV5uoxfnD170kHopB2\nHTQ4yG033DDvfmuGhxiozs+74bY9Ha5KkiRJ3W6u6dq+FxHbgcGImL6qRmbmYR2sa1btTtk2NFin\nqCV7CG7c4VAKSZIkzW2u6dpOiYg3AF8HnsoKLQM9XbtTthVFQVFrAj3cstuV7yRJkjS3uXqMycwb\nIuJRmbnzQBU0n4hguKeH7du3s27duln3q9frFLVJoIdte/YeuAIlSZLUldqZlaKIiHMj4ubqck5E\nHNrxyuYwHLV5p2wrioKBmARg+9jkgShLkiRJXaydYHwG8DPgaOAhwOXVthXTzpRtRVEwQNlTvKdZ\nY+9ee40lSZI0uzmHUlSOzMxnt1x/c0Rc3KmC2tHOlG1FUTCQ5djicfoZGxujr6/vQJQnSZKkLtRO\nj3FExN1artyNFT4Rr50p2/r6+hjsSQAa9LrIhyRJkubUTo/xu4EfRMQFlIH4acAbOlrVPNa0OWXb\n2nov7ISx7KHRaByAyiRJktSt5g3GmfnxiPgBcCyQwHsy88cdr2wOg21O2XbIYDl0YqxZs8dYkiRJ\nc2qnx5jMvAS4pMO1tK3dKdsOHSkAaBiMJUmSNI92xhjfKQ0T807ZdreDymA8NhmMjo4eiLIkSZLU\npbo2GK+pxbxTth160DC9AZMJN+/cc4AqkyRJUjeaMxhHRC0innagilmIdqZsGx4apKhmprhhu8FY\nkiRJs5szGGdmE3j7AaplQdqZsq0oCopaGYxv3OEYY0mSJM2unaEUP4yIR3W8kgVqZ8q2oigY6GkC\ncOvu8QNRliRJkrpUO7NS/Cbwkoi4HNhVbcvMfGTnyprf1JRte/funXVFu3q9ThGTQC+37Zk4sAVK\nkiSpq7QTjP+441UswtSUbTt27Jh1yraiKBhgLzDAbWOTB7ZASZIkdZV2FvjYAhARh2bmTR2vaAGm\npmybLRjX6/UqGMPoZMy7IIgkSZLuuuYdYxwRj4qIXwI/rK4/IiL+vuOVtWG+Kdv6+/spKIdQNOhz\nkQ9JkiTNqp2T7/4WeBpwE0Bmfg94bCeLatd8U7ZFBGuK8iGO00uj0ThQpUmSJKnLtBOM+zPzx9O2\n3SmmeGhnyraDB/sBGMsee4wlSZI0q3aC8VhEjExdiYgHAneK1TLambLt0OEBAMaaNYOxJEmSZtXO\nmWjvAL4M3CMiPgY8BXhRR6tqUztTth22pgCaBmNJkiTNqZ1ZKb4YEZcBT642vS0zL+9sWe1pZ8q2\nww6qA7tpTMLu0dEDW6AkSZK6RjtDKQB+BXwL+CZwZefKWbipKdtmc8jIEP01SIKbdtwpRoBIkiTp\nTmjeHuOIeCzwKW4fV1xExAmZ+Z2OVtam+aZsq9frFD3JeDO4frvBWJIkSTNrp8f4g8CLMvN+mXk/\n4IXVtjuF+aZsq9frFLUmADftcIyxJEmSZtZOMM7M/EbLlW91sJ4Fm2/KtqIoGIhyOehbdt8pZpmT\nJEnSnVA7wfirEfEigCi9EPhKZ8tq33xTthVFQVEF49v27D1QZUmSJKnLzDrGOCJuarn6moj4x+rn\nfuAW4LWdLKxd803ZVhQF/ewF6mwfax74AiVJktQV5jr57hEHrIolmG/KtqIoKCh7indPBpOTk/T0\n9BzoMiVJknQnN2swzsytB7COJZmasm2mYNzT08NwH7AHGvQyNjbG0NDQgS9SkiRJd2rzjjGOiGMi\n4lsRcV1E3FRd5l6H+QCbb8q2tfUy/49nn6vfSZIkaUbtLAn9EeCNwA+Ayc6WszjzTdl28GA59ngs\nXRZakiRJM2snGG/LzE93vJIlOGhwkKvnmLLtsDV1AMaaPTQajQNVliRJkrpIO9O1nRcRfxgRh0TE\n4NSl45UtwHxTtt3toKlgHPYYS5IkaUbtBOPrgf8N3Azsqi47O1nUTHLP7IG2dcq2mdx97RABjDeD\nXaMuCy1JkqQ7aicYvxN4HNCXmbXqcsDnO5u87tpZb2udsm0mw4N1BnoSgOu37e5IfZIkSepu7QTj\nqzPzosxc0RPvJq+7bs7bp6Zsm0m9XmegVgbjG3baYyxJkqQ7aufku69FxF8D5wH7xjNk5k86VtUM\nJq+dvccY5p6yrSgKilqT7dS4aYcn30mSJOmO2gnGLwYS+L1p2zctfzmzm7x2nh7jOaZsq9frFDEJ\n9HLr6HgHqpMkSVK3mzcYZ+bGA1DHvCZvuIGcmCB6Zy55rinbiqJgICaAAbaNznyCniRJku7a2ln5\n7oEzXdo5eEQ8JSJ+GhGXR8TrZ7j9mRFxcUT8MCK+HxGPn/Vgk5M055iSba4p24qiYCDLnuIdjWY7\npUuSJOkupp2hFF+kHEoBUAB3A37JPEMpIqIHeD/wBOAa4HsRcX5mXtqy279k5ueq/R8M/DNwn9mO\nOXnddfTc4x4z3tY6ZVtfX99+t/X19VHvKQPxrr3QbDap1do571CSJEl3FfOmw8zcmJmbqsvhwBOB\nz7Rx7EcCV2Tm1szcC5wLPHPasVvnThumnCt5VnOdgDfflG0HDZQzzDXoc5EPSZIk3cGCu00z82vA\n7EMebncEcFXL9aurbfuJiGdFxKXAhcAfz3XAeU/Am2PKtkOG+gFo0GswliRJ0h3MO5Ri2njiGmVP\ncH8bx875d4HM/Czw2Yg4BjgbOGq2fSevu5bMJCJmvH2uKdvWTQXjZg+NhlO2SZIkaX8LHWM8AVwB\nnNTG/a4BNrRc30DZazyjzPxWRPRGxLrMvGX67X8ztgd2bGfgnHN43MMeyjEPuOP5fyP9A7OegHfY\nmgKYYCxr9hhLkiR1uS1btrBly5ZlPWYnp2u7CLhvRGwErgWeD5zQukNEHAn8IjMzIh5WtXeHUAzw\nugf/BhOXX87gQx5C/wyhGMqZKa6aZcq2u68ZBHYwNhkGY0mSpC63efNmNm/evO/6W97yliUfc9Zg\nPN+UbPOtfJeZExHxSuDLQA9wZmZeGhEvr24/A3gO8JKI2AvsAl4w2/F67nEPJi6/vDwB70EPmnGf\nuaZsO+zgIWqxnYkMtu9yWWhJkiTtb64e49YhFK1GgIMpw+6cMvNCypPqWred0fLz6cDp7RTac4/D\ngblPwJtryrZ6vU5Rg9FJuGH7aDtNSpIk6S5k1lkppk3Ttgn4deAsYBL4mwNV4JSew8v5iyevW9yU\nbUVRUFRzGd+w3R5jSZIk7a+dle96I+KPgZ9RnkD3sMx8bccrm6Z26Hro6yO33UZzdPYe39mmbCuK\ngiLKYHzTTscYS5IkaX+zBuMovQS4DHgscGxmvjQzrzlg1bXWU6vRc/e7A+UKeLOZbcq2er3OQEwA\ncMvu8c4UKUmSpK411xjj/wKGgLdQzjDR23pC3nwn33VCzz3uweRVVzF57bX0HXnkjPuM9A+wbYaZ\nKYqiYIAyGG/bs7ejdUqSJKn7zBWMRyhPvjttlts3LXs182jnBLw19TpXzTAzxcDAAEUVjLePNTtT\noCRJkrrWrMF4CfMXd0w7J+DNNmVbRDDcD4zB6AQ0m01qtQWviC1JkqRVqquSYc/hd4cImjfcSE5M\nzLhP65Rt0x08WE7hNkavy0JLkiRpP10VjKO/n9r69dBsMjnLCndzTdm2bqgfgEb2uvqdJEmS9tNV\nwRjKE/CAcgW8WYxEbcaZKdYPDwDQyB6DsSRJkvbThcF4/hPwRoIZe4zvtrYOwFiz5lAKSZIk7af7\ngnEbJ+DNNmXb4WuHABibDPbscfU7SZIk3a77gnFLj3E2Z552bU29zvYZZqZYOzxIbyRNgm27DMaS\nJEm6XdcF49rICDEyAo0GzW3bZtxntinbiqKg6EkArt+2u6N1SpIkqbt0XTAG6Dm86jWeZWnowYEB\nGjNM2Vav1ylqZS/zDTsdYyxJkqTbdWcwPqIaZ3zNzOOMI4KRGaZsK4qCgZgE4Oad/5+9O42NI83z\nO/99IiIjbybJTN6HpNJRpZJKdXR1q/quPqsP9zndY493XxgeA4M1bOwu9oWxC2O3jX23u17vC2PH\ns4sZeOAF7BlM9XRPz9EzdnerqqvrvnSfJVEU7yuZ9xkR+yIpipJIilUiq0Ty9wESSUZERj6ZKgE/\nPfU8/7+qUoiIiIjILdszGG9kA94qJdsikQjhpbbQ86X61g1QRERERLad7RmMP2DJtlYwbi2vWKzc\n3RlPRERERHavbRmMrXQaXJcgl8Mvrb6JbrWSbbZtk2h1hSZXXb2ihYiIiIjsTtsyGBvLwu7tBdbe\ngJeKRclNTd19PGIDUGpCEARbN0gRERER2Va2ZTCGe7eGTkaiLM7N3XW8I+oAUA0cdb8TERERkWU7\nNhivVbItnQgDUAscqlVVphARERGRlm0cjNffgLdWybbuZBSAamApGIuIiIjIsu0bjHt7wRj82VmC\nxuoVJlYr2dbXHgOg5ltaSiEiIiIiy7ZtMDahEFZXF/g+3h3VJ25qW6VkW/dyMDaUypUtH6eIiIiI\nbA/bNhjDvdcZJ1Yp2ZaIRXGtgADDfKG85WMUERERke1hhwTjjZdsi0ajRKxWDePpnGaMRURERKRl\nmwfjmxvwNl6yLRKJLAfjmbyCsYiIiIi0bO9g3LcUjCcnCfy7O9mtVrKt1Ra6CcBsQZvvRERERKRl\nWwdjK5HAtLVBvY6/sHDX+dVKtkWjUcKmFYznS/UPbawiIiIi8mDb1sEYNtAB746SbaFQiKjlAbBY\nWZwi8rQAACAASURBVL3Mm4iIiIjsPjsgGK/f6KPNcFct47awDUCuevfyCxERERHZnXZAMF6aMZ5c\nu2Tb4szMbcfaI61gXGpAEARbO0ARERER2RZ2TjAeXz0Yr1ayrTMeBqAa2NTrWmcsIiIiIjsgGFsd\nHRAOExQK+IXCXedXK9mWSboA1HCoVqsfyjhFRERE5MG27YOxsazbyrbdabWSbb1tUQCqvq1gLCIi\nIiLADgjGsP4GPGMMScu6rWRbX0ccgJpvKRiLiIiICLBTgnHf+hvwkpZ9W2WKdFsMQ0A9MJQqCsYi\nIiIislOC8cDNWsYbK9kWi0aJWK1qFLO58tYPUEREREQeeDsjGHd3g2Xhz84SrFJl4s6SbZFIhLDV\nqmE8rWAsIiIiIuyQYGxCIazubggCvDtKs8HdJdsikQiRpe5307nKhzZOEREREXlw7YhgDOtvwEtG\noizOzi7/Ho1GCdMEYK5Y+3AGKCIiIiIPtJ0TjNfZgBcLh6kVCssl28LhMGHTmjFeKDfuul5ERERE\ndp+dE4zfR8k2YwyJUOtcttL80MYoIiIiIg+uHRSMb84YTxL4/l3n7yzZlorYAORr3oczQBERERF5\noO2YYGzFYpj2dmg08O9oAQ13l2zrjLWmjIv14EMbo4iIiIg8uHZMMIb1l1PcWbItkwgDUA0c6quU\neBMRERGR3WVnBeN1NuDdWbKtK9kKxrXAVltoEREREdlhwfh9lGzra48DCsYiIiIi0rLDgvHN1tB3\nzxjHI5HbSrZ1p2IAVHxLwVhEREREdlYwtjo6IBIhKBbxC4W7zq8s2daeiGKbAC8w5ErqficiIiKy\n2+2oYGyMubWcYvzuWeOVJdui0Shhq1WRYiZX/vAGKSIiIiIPpB0VjGH9DXgrS7ZFo1EiS93vFIxF\nREREZOcF4w2WbItEIsttoWfyWkohIiIistttm2DsrdLNbjXrbcBbWbItEokQprURb66oOsYiIiIi\nu922CcYzK7rWrcfu7gbbxp+fJ6jVbju3smSbbdvEndYa44VSY3MHKyIiIiLbzrYJxhO5xQ1dZxwH\nu6cbggBvRUMPuLtkW1vEBiBbbW7uYEVERERk29k+wdjzNnzt8ga8VZZT3FayLeIAUKhu/N4iIiIi\nsjNtm2BcDoWo1De2FnjdDngrSrZ1xkMAFBrBJo1SRERERLarbROMBx55hInswoauXW8D3sqSbZl4\nGICqby8vrxARERGR3WnbBOM9hw8zWd5YvWGrb2nGeGqK4I4lGCtLtvWkogBUcdQWWkRERGSX2/Jg\nbIz5mjHmgjHmsjHmX6xy/r8yxpw0xpwyxvzGGHNstfsMDg4yucF1xlY02moP3Wziz83ddm5lybbe\n9jgANd9SMBYRERHZ5bY0GBtjbODfAl8DHgV+xxhz+I7LrgKfC4LgGPC/Av/PavdKJpO46TQLxeKG\n3vvWcorb1xmvLNnW1bY0Y+xbVCpq8iEiIiKym231jPEngCtBEIwEQdAA/hPwnZUXBEHwShAEN4sU\nvwYMrnWzPUePMr6w0XXGNzfg3b7OeGXJtrZEjJAJCDAsFNQWWkRERGQ32+pgPADcWPH72NKxtfwu\n8NdrnRx66CGmmhvbJLfuBjzbJp/PE41GiVitjnqzagstIiIisqs5W3z/DddBM8Z8AfjHwKdXO/+j\nH/0Iz/N499e/JlSu8OzRo+veb7mW8eQkQRBgjFk+lzAWuVyOTCZD2PIoeLaCsYiIiMg2cuLECU6c\nOLGp99zqYDwODK34fYjWrPFtljbc/b/A14IgyK52ox/96EcA/OQ//AcOLKx6ye33bE9holGCUokg\nn8ekUsvnbpZsGxwcJEwTcJlSMBYRERHZNp599lmeffbZ5d//1b/6V/d9z61eSvEmcNAYs9cY4wJ/\nH/iLlRcYY4aBHwP/dRAEV+51w+EjR5gs5O/5xsaYNRt93CzZ5rouEatV6WKuUNvQBxIRERGRnWlL\ng3EQBE3gnwF/C5wD/iQIgvPGmN8zxvze0mX/M9AB/L4x5h1jzOvr3XNoaGjD7aFvrTMev+34ypJt\nyVDrK1goq8GHiIiIyG621UspCILgb4C/uePYH6z4+Z8A/2Sj98tkMtSiUUrVKvFIZN1r1yrZ1haN\nLZdsS0VsKMJipbnRIYiIiIjIDrRtOt/dZIxh6PBhxjfQHnrlBryVYuHwcsm2jljr3wb5mr/5gxUR\nERGRbWPbBWOA4UOHmNxAQw6ruwtsG39+nuCOznY3S7alE2EAio0NF9AQERERkR1oWwbjwcFBpnyf\nIFg/zBrbxu7tBcCbnLrt3M2Sbd3JVjCu+jbNppZTiIiIiOxW2zIYx+Nx4j09zBUK97zW7luqTDF5\ne6OPmyXb+tpjANQCm+ods8oiIiIisntsy2AMsOexx5hcXLzndfbA0jrj8duDcTLcKtnW1dYKxlUF\nYxEREZFdbdsG48E9e5jYQHvotTbgtUVbJdu6UnEgoOYbSmU1+RARERHZrbZtMO7r6yNrDI17rAu2\n+5bWGE9NEayof3yzZFs8FiVsBYBhLl/eyiGLiIiIyANs2wZjx3HoPXjwnsspTCSClU6D5+HPzC4f\nj4XDNAoFfN8nbFql2mYUjEVERER2rW0bjAGGHnmEifvYgLfPCfHe5ctETGvWeTavpRQiIiIiu9W2\nDsbDw8NMePcusWb3LwXjiduD8ZHeXi6/8gourXvMKBiLiIiI7FrbOhh3dnbiJ5PkK+svgbD7B4C7\ng3EyGqW30cButgLxXOnem/lEREREZGfa1sEYYPjIESYWsutec2vGePKupiBHurrxC6320tlifWsG\nKSIiIiIPvO0fjA8cYKK2fv1h09aGiccJKhWCxdxt5zLJJMl6a8Z5sarOdyIiIiK71bYPxgMDA0x7\nPr7vr3mNMWbNDXgAe6I2ALna2vcQERERkZ1t2wfjaDRK++Ags/n8utfZ/UuNPibuDsbDcReAxbKW\nUoiIiIjsVts+GAMMH3mUidz69YxXrjO+U1eo9Vys+TTv0TBERERERHamHRGMh/buZaLprXvNejPG\nXa4BoI7D6Ojo5g9QRERERB54OyIY9/T0UHBsao21y61ZmQw4Dn42i1+5vV5xZikYe06EN196aUvH\nKiIiIiIPph0RjG3bpu/QISaya5dtM7aN3dsLgD95+3KKtGtjggDPCjHy7kkW79FmWkRERER2nh0R\njAH2HD7MRKm47jVrLaeIuC7hoLXxrsOEeff117dmkCIiIiLywNoxwXhoaIhJb/1ya2ttwIuGQoT9\nVjBOpNK89+qrlEqlrRmoiIiIiDyQdkwwTqVSWO0pFtcJtHbf0ozx5Cozxl4NgHxgs89YnH733a0b\nrIiIiIg8cHZMMAbYe/TouuuM7b5eMAZveoZgRVk227KILi2lmK0bHu3r4+wLL1Kvq66xiIiIyG6x\no4Lx0P79TK4TZk04jJVOg+fhzczcdi5BKyjPNw2JSIT+RoOzZ85s6XhFRERE5MGxo4Jxf38/s4GP\nt0576LU24LVZrTrIWa/VHvpITw8nf/UrPG/9+sgiIiIisjPsqGAcDodJ79nD9Drl1tbagNe+FIxz\nQSsYdyYStBeKXL58eYtGKyIiIiIPkh0VjAGGjhxhMp9b8/xaG/DSdmuWuRA4y8eOZtK8/ctfEgTB\nFoxURERERB4kOy8YDw8zsc7yh5UzxisDbzrUei5xKxj3tndgz8wwMjKyJWMVERERkQfHjgvG3d3d\nlMNhKmtswrPa2jCJBFSr+CsqWHQvBeOKCd22RvlIW4p3XnhhS8csIiIiIh+9HReMjTEMPvII4wsL\na16z2ga8tGsAqNthao3G8vE9XV2Ur11j8o420iIiIiKys+y4YAww/PDDTJbXafSxyga87nDrq6hZ\nLpX6rWBsjOFwIsk7v/nNFo1WRERERB4EOzIYDw4OMuX7a26aW20DXrvrYAcenrHJ1Zu3XX+gp4fp\n06dZWGcWWkRERES2tx0ZjJPJJOFMhoVicdXzq80YR9wQYf9m97vb6yDblsXhSJR3Xntti0YsIiIi\nIh+1HRmMAYaPHFmzPbSVyUAoRLC4iF8uAxAJucvBeK5+90zzob4+Rl5/g+IaYVtEREREtrcdG4yH\nHnqIyWZj1XPGsrD7bs4at5ZTuI5DxKsBML/Ky1zHYb9tc/Ktt7ZmwCIiIiLykdqxwbi/v5/5IKC5\nRk3j5eUU4+PLx2K01hbPNs2qrznc28uFl16iVqtt8mhFRERE5KO2Y4NxKBSi+8ABptZoD+3s3QtA\n7dcv4ZdaFSySphWMF9YIxvFIhEHf58ypU5s/YBERERH5SO3YYAwwdPgwE/n8qudCjz+OvXcvQaFA\n5c9/QhAEpKzW7HLWt9e855HuHk6dOEGz2VzzGhERERHZfnZ2MB4aYtJffSmFsSxif/+3wXVpnD5N\n45136bBb1+b8tb+W9nicdKnMhQsXtmTMIiIiIvLR2NHBOJPJUItGKVWrq563OzuJfvtbAJR/8hN6\nm63rioGz7n2PdnVx8sSJNeski4iIiMj2s6ODsTGG4SNHGM+u3ZjDffppnEcfhVqN/affBqDI+sG4\nO5XCnZ3l6tWrmzpeEREREfno7OhgDDB88CATlcqa540xxH7r+5hEgu7xEQCqxsX3/TVfA3C0o5N3\nTpzYxJGKiIiIyEdpxwfjgYEBptdpDw1gJRLEfuv7pGqt5h0NE6J2j811g+k0tdFRxleUexMRERGR\n7WvHB+N4PE6ir4+5NapT3BR69FHiDz8EQNVyKVdWX5d8kzGGRxNJ3vr1S5s2VhERERH56Oz4YAxL\n7aHXqGe8UuS5rxJrVPAtm/wL9w68+3t6WDh/jrm5uc0YpoiIiIh8hHZFMB7au5eJDdQdjiQShILW\ndYXTF2iOjKx7vWVZHI7GeOfVVzdjmCIiIiLyEdoVwbivr49Fy1C/RziOhEI4S22h826c8p/8KcE9\n2j8f6u1l9M23yN9jqYaIiIiIPNh2RTC2bZvegweZzGbXvc4YQzSoAzDX1Y+/sEDlZ3+57mtCjsPB\nkMPJN9/ctPGKiIiIyIdvVwRjgOHDh5koFu55XcK0ut9de+xJcBzqb7xB49y5dV9zuK+fSy+/TGWd\nsnAiIiIi8mDbNcF4aGiISW/19tArpUxrKcV0OEHkuecAKD//Y/xicc3XRF2X4SDgzMmTmzNYERER\nEfnQ7Zpg3NnZiZ9Mkq+U172u3WqF50XfJvyZT+Psf4igWKT8/I/XrYV8pLeX0y+8QKPR2NRxi4iI\niMiHY9cEY1gq27aw/jrjTqfV8S7n2xjLIvbD34ZwmOa5c9TffGvN17VFY3RXq5w/f35TxywiIiIi\nH45dFYz3HDzIRG39xh0ZpzUrXAwcAKyOdmLf/Q4Alb/4C7yFhTVfe6Srm5O//OU920mLiIiIyINn\nVwXjm+2h1wuumZABoISzfCz05JOEHnsM6vVWCbc1Xt/V1kZscZH33ntvcwcuIiIiIltuVwXjSCRC\n+8Ags+vUHE67rWBcsdzlNcXGGKLf+y4mmcQbGaH24q/XfP3Rzk7e/tWJTR23iIiIiGy9XRWMAfY8\ndpSJ3NrtobvDrWBcs1yqKzbSWfE4sR/8AIDq3/0d3sTEqq8f6Ezjj40xOjq6iaMWERERka2264Lx\n0J49TKxTti3tOhAE1C2XUv32ChOhRx7GfeY4eB6lP/lTgjUqUDzW3s6LP/0p8/Pzmzp2EREREdk6\nuy4Y9/T0ULAdqvX6qufjbohw0Aq8c7W7A3T0m9/ESqfxp6ao/t1/XvUee7u7OdZo8tN/9++4evXq\n5g1eRERERLbMrgvGlmUx8PAhJtZoDx0JhQh7NQBm63fXLTauS+wf/H2wLGq//jXNNYLvQ93dfKkz\nza///b/ntZdfXrcGsoiIiIh89LY8GBtjvmaMuWCMuWyM+RernH/EGPOKMaZqjPkftno8sNQeulRa\n9Zxj20SC1mzyasEYwBkeJvyFZyEIWksqKquXgEsnk3xz7z7Gfv5z/vrHP6ZWq23K+EVERERk821p\nMDbG2MC/Bb4GPAr8jjHm8B2XzQP/HPg/tnIsKw0ODq7bHjpOqy10trn2PSJf+hL2wADB4iLln/3F\n2te5Ll85cJDo+Qs8/0d/RHaNmWoRERER+Wht9YzxJ4ArQRCMBEHQAP4T8J2VFwRBMBsEwZvAh9ZL\nOZVKEersILvGrHFyKRjPrROMjW23llQ4Do233qZ+5sya11qWxSf27uVwqcyf//7vMzIycj/DFxER\nEZEtsNXBeAC4seL3saVjH7nhI0eYyK7exS5lt2aT55vrfz12dzfRb3wdgMrzP8YvFNa9/mBvL19o\n7+DEH/4Rb772mtYdi4iIiDxAtjoYP7DJb2j/fibXqEzRYbU62y169j3v437ykzgHDxKUy5T/7M/u\nGXa72tr45r59XPurv+LnP/0pjTVKvomIiIjIh8u59yX3ZRwYWvH7EK1Z4/ftRz/60fLPzz77LM8+\n++z9jIuBgQF+ASyWSrTH47edSzsBeJAL7h2MjWUR++EPKPybf0PzwkXqr79O+PjxdV8TdV2eO3CQ\n106f5vnpab7+D/4BqVTqfj6OiIiIyK5y4sQJTpw4san3NFv5v/ONMQ5wEfgSMAG8DvxOEATnV7n2\nR0AhCIJ/vcq5YCvGeeHCBV5+/nmeDEc42Nu7fPyP31vgf68+xJ7GHH/1xMb+7VB/9yTl//gfIRQi\n+d/9t9iZzIZed2lqkpNNjy/9w99heHj4A30OERERkd3OGEMQBOZ+7rGlSymCIGgC/wz4W+Ac8CdB\nEJw3xvyeMeb3AIwxvcaYG8B/D/xLY8yoMSaxleO66ZFHHuF7//SfcrmjnRNXLlNvtnbbpd3Wd1p6\nHxPq7hOPE3ricWg0KP/JnxKsU/VipUO9fXw+meQXf/iHvP3mm+//Q4iIiIjIptjSGePNslUzxjd5\nnserv/41V375Kz7b08O1SsDvzu8l7pX5/eQlntq3d0P38ctlCv/m/yLI54k891UiX/zihsdQrtX4\n1cg1Oo4f5wvPPUcoFPpgH0ZERERkF3rgZ4y3C9u2+fSzz/Ls7/5jXqyUmZufAqBhR/mbsUX+5vQZ\nmhuYAbZiMWI//CEA1f/8X6j8zc8JNri5LhYO87WDhwjefoc//+M/Jp/Pf/APJCIiIiLvm2aM71Au\nl/npn/wp/9PFTgJj8z92FxkbOUe/l+O7jx8lGY3e8x7VX/yC6n/+LxAEWJkMsd/6Ps5DD214DBcm\nJjhNwFf+4T9kcHDwfj6OiIiIyK6wGTPGCsarqNVqHPtf/pqacflHsVn2pOJMTI5iZq/y208cpre9\n4573aI5cp/z88/gzMwC4x48T/frXMdHIhsYwvbjIi3OzPPWd7/D4k0/e1+cRERER2ekUjLfQk//y\nx2S9MM8mFxms5tibTpPPLTI/corvHtrD4cF79ykJmk2qv/oVtV+dAM/DtLUR++53CB05sqExlKpV\nfnV9hMwnP8mzX/0qjrPV1fVEREREticF4y30mR/9mLFamH/0yWH8uRvkrl7lQCqFHfjcuPwun+5O\n8LmHD2FZ916m7U1NUf6z5/FutJoAhh57jOh3vo2VTN77tb7Py9euURoe4rkf/IDkBl4jIiIistto\n890WSiyVbKs0Ax45epR9x49zoVRmsVrnoUc/wSs5nz97610qa3TPW8nu7SXxT/8bot/+FrgujdOn\nKfzr/5PaG2/es1OebVl8dv9+hmfneP4P/oCJiYlN+XwiIiIicjvNGK/ht//1X/D6nM1zR3r5/KEu\nAKqVCudPnsTMzLC/s5PpyRvEcjf4wZNHyWxwJtdbWKDy5z+heekSAM6BA0S//z3sdPqer53MZnlp\nfo70oUN0DQ+T6ekhk8mQSqU2NHMtIiIislNpKcUW+ud/8Df8bMTHAB/f28lXH+0hFnYIfJ+Ra9eY\nPHuO/Yk4jUqJ4o2z/ODIAR7q6dnQvYMgoPHOO1R+9pcE5TKEQkS++hXCn/40xl6/DXWt0WCukGeh\nWGKh2STr+5Rti86+fjJ7hunq7yeTydDZ2alayCIiIrJrKBhvob/95Qv83++UOT3v4wcQDdl89UgP\nH9/biWUMi9ksF956i856nU7XZfK9k3xpKM3x/Q9hzMb+TPxikcrPfkbj3ZMA2IODxH7wW9h9fe9r\nrE3PI1ssMl8skm00WPA9cp5HoquLzNAQmaEhurq6SKfTxGKx9/1diIiIiDzoFIy30FtvvcXPz0wS\nGzjEz05NcHW2BMBAe5RvPd7PcGeMRqPBxTNnqIyOsjeZYGrkAo/FA7722BHc91FBonH+POU//wlB\nLgeWRfjZzxP54hcx9zHjGwQBuXKZhWKRbKXMgu+z4PtY8TiZwUG69+wh09NDOp0mlUptOMyLiIiI\nPIgUjLfQ2bNn+cmrF9n32NMEQcCZiTx/fXqSXKXVye5jwx08d7SXRNhhYnyca+++w6ATopqdIl2Z\n5beefIzU+5idDapVKj//W+qvvAKA1dXVagyyb9+mfq5yrcbCzdllzyMbBFQsi/RAP3379/PU8eNE\nIhurtSwiIiLyoFAw3kJXr17lP/7iDfY9/snlY/Wmz68uzvDS5Tm8ICASsvjy4R6O70tTrZQ5//bb\nRBZzxL0q/vQVfnjsEQY3sKlupebICOU/ex5/dhYA95lniH79a5gtDKuNZpNsqcTIwgKj0Qif+e53\nOXDgwJa9n4iIiMhmUzDeQhMTE/zRT37BQ09//q5zc4Uaf3l6gkvTRQB62yJ86/F+9nREeO/SJeYv\nX6KbgPLEZb6xv48n9ux5X+8dNBq3GoP4fqsxyPe+S+jRRzfjo61rNp/nlclJ2p98gs8/9xzxeHzL\n31NERETkfikYb6FsNsu//f9+zP7jX1n1fBAEnJ8s8FenJ8iWW8srnhhq52tHeqmXclw5eRKzmCWY\nusqzfW188fAjOPeoOHEnb3KK8vMrGoMcO9ZqDJJI3N+Huwff9zk9PsbFAI5/59s8+uijWoMsIiIi\nDzQF4y1UqVT4337/33PwM99c97qG5/PipVleuDRL0w9wHYsvPdLNJ/d1Mj83y7Vz55g9+xaP21X+\n0WeeIfE+l0QEvk/tNy9T/du/hUYDE4vhPvMMzt492ENDWFtYZSJbKvHq+Dihhx/mC3/vm6RSqS17\nLxEREZH7oWC8hYIg4Pf/8I+Zrdkke4bp6h/CXqfSxEKpzl+dnuT8ZB6ArmSYbx3rZ39XnLm5Od76\n5d/Be+/wu08e5amHNl7S7SZvYYHKj/+c5uXLtx23urqwh4dwhvfgDA9h9fTcsxby+xEEARcmJzld\nq/LE17/OE089pWYiIiIi8sBRMN5inucxOjrK26fOcua9Uey2bjoH9tKe7l7zNRenCvzlqQnmS61W\n0UcHUnzjaC+paIj3Lp7j/C9/ypNRm+cOPMRD3T3vKyAHQUDz4kWaV96jeWMUb2wcms3bL3JdnKFB\n7OFhnOFh7OHhTVl6UaxWeeXGDZp7hvnCt79NJpO573uKiIiIbBYF4w9RuVzm0qVLvPL2aabzVcKZ\nAXqGHiISvXspQ9PzeenKHL+6OEPDCwjZhi883M1nDmSoV4qcfenvSDdyDOLzWFuKgz09H2gWNmg2\n8aam8K5fpzl6A290FH9h4a7rrM7OpaA8hL1nD3ZvL+Z91Fle6cr0NG8X8jz6la/w9PHjOB/wPiIi\nIiKbScH4IzIzM8Ppc+d589QFak6cRM8w3f1DWHcsYVgs1/nrM1OcGc8BkI67/L1j/ezPRLh68jV6\nQlV6Ii75y5c5Ek9wqLcX+z6XKfiFAt6NGzRHR/Guj9IcG4N6/faLHAd7cABnaBh7T2tm2Xof64cr\n9TpvjN0gm8nwhe9+l/7+/vsas4iIiMj9UjD+iHmex8jICG+dOsv5q2PY7T2k+/eSSnfddt2VmSI/\nOzXBbKEGwOHeJN94rI/CjYu0Neb4ymc/ybXz55l8910OhyM83NdHaJNmYgPPw5+epjl6g+bodbzR\nG8s1klcyqRTOnj24xz+Bs3//hpZ43Jib47XsAg999rM887nP4brupoxZRERE5P1SMH6AlEolLl68\nxCvvnGa2UCfSNUD34L7lpRZN3+eV9+b5xYUZ6k0fxzJ87lAXj8Yr1MbP8Tvf+irt7e28/corXH/j\nTR52Qxzu6yd8H22h1+KXy3ijS7PKo6M0b9yAanX5vD0wQPjznyN09Og9N/I1mk3evDHKRCLB57/3\nPfbu3bvp4xURERG5FwXjB9T09HRrqcXpC9RDyVZVi75BLNsmX2nw87NTvHtjEYCOWIgv7m8jPnWS\nb3z6SY5/4uPk83neff11rrzyKgdsi0f7+olu4Wxs4Pv4s7M0zpyh9puXCUoloLU2OfzZz+I+/THM\nPd5/ajHLq7OzdH/843z2S18iGo1u2XhFRERE7qRg/IBrNpuMjIzw5skzXLw+gZ3qJTO4j7aONNfm\nSvzs5ART+dZM7YGuGEfNGM8eTPPN575CKBSiWCzy7htvcPE3v2FvAEf7+ohvYWtoaHXdq7/1NrUX\nX8Sfnwdo1U7+1CcJf+pTWOt0wvN8n5NjN3jPcfj0d7/LoUOHNndsQUC1WqVYLFIsFimVSlQqFQ4f\nPkxii5ueiIiIyINNwXgbKRaLXLhwkdfePcNcqUkkM0BmYB/vTJb4L+enqTZ8bGM4mmrw1e4iv/33\nvkpXV2utcrlc5tQ773D2xAsMeh77Ozvoakvd90a99QS+T+PcOWonXljuvEcohPv004Q/+xnsdHrN\n184XCrw8OUHyscf4/Ne+RjKZ3NB73gy9pVKpFX7zeQrz8xSzWQrzC5TzOUK+T8wYYpZN3BiM53E9\nHOYzv/V9Dh48uBkfXURERLYhBeNtampqitNnz/HmmUs03DaczCBvztm8PdpaXpEIGZ6ILvB0xueZ\nY49w8MB+0uk0tVqNM6dOcf30aeZv3KDTGHpth95Uikxb25YE5SAI8K5do/riizTPX2gdNIbQdAhg\n1gAAIABJREFUY48R/vzncAYH13zd2fFxzvkeT3/zmzz8yCPLgbdUKlEoFCjOz1OYn6e0uEgxu4jV\nbBK3LWKYpfBrEQ+HSUTCxMJh4uHIqp9xvlDgpYkJup85zue+/GXC4fCmfw8iIiLyYFMw3uYajQbX\nrl3jzZNnuTQ6yUJkgNcXo0wVGwC4tmFvwmefmeNYxubpx1ohubOzk0ajwdTUFOOjo4xduEj2xihp\ny6LXCdHT1kZXW9umd6jzpqapvfgi9XffBc8DwNn/EOHPfx7n0KFVK1nkymVeHR9nAVpB1zJEgbix\niIVdEuEI8UiEeDiMcx8d+zzf563RUcaTCb74gx8wMDDwge8lIiIi24+C8Q5SKBQ4f+Eir7xzhtfn\nXS432pmp3PrMbWGbh+IN9plZDnaG+NjRWyEZoF6v3wrK5y+wOD5GxrLocUKtGeVkctOCsp/LUfvN\nb6i9+hrUWiXorN5eIp//HKHHH9/UltTv18TCAr+Zn+ORL3+ZT3zqU9gf4VhERETkw6NgvAMFQcDk\n5CQnz5zjFyevctnr4lo1Qr7mL1/TlwyxL1pj2J9iuCNyV0iGVlCenJxk4saNpaA8TsYy9C4F5fQm\nBOWgUqX22mvUXnqJoFAAWvWQw5/9DOFPfALzES1pqDUavHJ9hPLgIF/+/vdv+15ERERkZ1Iw3uFq\ntRqXLl3i5bdO8+5sk2t+hqsFQ91rhWTLwEOdYfaEy/Q1Jhloj/Gxxw5z8MB+Ojo67rrX5OQkE6Oj\njF24QG5igq6VM8ptbRtq6rGaoNmk/s67rUoWMzMAmGgE95lnCH/601gb3Hy32a5MT/NWqcTT3/4W\nxx5//AN/PhEREXnwKRjvItPT05w6c5aXT17kGl1cb6a4vljHX/paIiGLhzNhhp0CndUp+jriPLU0\nk3xnSIZbQXn8+nXGLlwgPzlJt2XT64bo7+ikY52ybGsJfJ/mxYtUX3gB79pI66Bt4z71FO5TT2JS\nKaxk8p41kTdToVLhpRs3CD96mC9961vEP8DnEhERkQefgvEuVK/XuXLlCq+8dYrLMyVGnV6ulcNM\n5mvL13TEQhzuCjNoLRIrTdHbvn5IhlaptImJCcauXuX6mTN4izn6bIu+WJz+jg4i7zPMNq+PUnvx\nBRpnz8Gdf3au2wrIiQRWIoFJJpeeV/ycSGxaiA6CgNNjN7hoWXzuBz9g//79931PERERebAoGO9y\ns7OznDxzljdOXWDGdDBmZbg43yRfbS5fM9wZ5ZHOED3BPE5xmt72BB87+ggHDuynvb19zXvncjnG\nxsYYPX+e8YuXSDSb9Ds2/akU3W2pDa9P9mZnqb30G7zxcfxigaBQhGbz3i+86WaITt4RohNJrGQC\n09aG3deHcZx73mpuqaxb36c/xWe/+EXcD3HmWkRERLaWgrEArbJv7733Hq++fYqrU1kWooOMNhNc\nmC4vr0e2jeHh3iQPd1h0NGbxc1P0dbbxsSMP3zMk+77P9PQ0N65fZ/TsObI3Rum2LPrdMP2dnaRi\nsQ2PNQgCqNaWQnIBv1gkKBbxC63Q7BeLtx3fSIg2iQTuxz9O+PgnsNaYEb+p6Xm8OTrKZKqNL//w\nh/T19W147CIiIvLgUjCWu8zPz3PqzFleP3meop1kNtzHlbzhvdkSN7/BaMjm2GCKgymIlGZo5iY3\nHJKhtexifHyc0ffeY/TMWcjn6LNt+uMJ+jo6CIdCm/JZgiAgqFYJboblQpGgWLgtPPuzc/hzc60X\nGIPz8MOEn3kG5+FDmHVmtcfm53llYZ5Hn3uOp48fV1k3ERGRbU7BWNbUbDa5du0ar759ivfGZ2i0\nDTFtpzkzVWYqX12+Lh13eWKonf1Jn2BxkmZ+mpgDw/197Bvso6enm66uLmLrzApns9nWsosLF5i8\ndIm2pke/49Df3k7XfVS72IggCPBGrlN79VUap08vNx6xOjpwj38C9+Mfx0okVn1ttV7n5dHr1IaH\n+fL3vrfm+msRERF58CkYy4Zks1nOnD3PK++cpmrHqaWGGKmEOTWWo1C7tVRhbzrGk0MdDKccTDVH\nOZfFK+fwyznaYmH2DvaxZ6CXnp4eMpnMqq2XPc9rLbsYGWH07Dly4+P0WBZ94TBD6TTxSGTLPqdf\nLFJ/803qr72Ov7DQOmjbhI4eJfzMcex9+1YN6ZcmJ3m3WuHj3/kORx97TGXdREREtiEFY3lfPM9j\nZGSE194+xaUbUzidA5RifZybrXFuMk/Du/UdJ8MOQ52xpUeUzpBHrbBIKbdAUMnjlXNk2pPsG+xn\neKCXrq4uMpkMzh2b4CqVSms2+fJlrp85Q6xSYcBxGGzvuK/ayesJfJ/m5cvUXn2V5vkLy1UxrO5u\nws8cx33qY5jo7QE9Xynz0ugNYsce4wvf+IbKuomIiGwzCsbygeVyOc6eO88r75yhFIRxu4aZCdo4\nO1lgdKFMpeHddr1loKctwlBnjOHOGEMdUaJ+lUJugWo+i1/J4VeK9GY62DfYx2B/Kyyn0+nlChZB\nEDA9Pc31q1cZOX2a0uQk/bbNQDzBQGcn7gYqS7xffnaR2uuvU3/jjeXufIRCuE88gfvJZ3AGBpav\nDYKAU2M3uGTbPPvDH7Jv375NH4+IiIhsDQVjuW++73P9+nXeeOcU56+N43T0k0h340VSTBSa3Fgo\nc2OhtS7Zv+OPIBqyGeqMtsJyR4z+VBivUiCfnadeXMQv5zHNMgM9Xewb7GOgrxWW29vbMcZQLBYZ\nHR3l2tmzTF68SIfvMxByGezspH2TZ2wDz6Nx9iz1V1+l+d7V5eP20BDuM8dxjx1brpk8m8vx0vQU\n6ccfp3ffPjKZDOl0WrPIIiIiDzAFY9lUhUKBS5cuc2nkBiNjEzSMix1vJ5JKE02lyTYcbiyUGV0K\nyyvXJ9/UlQwzvBSUhzpjpGM2pXyWfHaBZmkRv5IjFDQZ7Otm30Affb09dHd3E41GmZiY4Prly1w7\nfRoWFxmwbIba2+lpb8feYN3kjfBmZqi/9hr1t94iqLQ2IppoBPdjH8N95hnsri4azSbX5+ZYrFbI\n+j6Lvo8fiZAZHCQzNESmt5fOzk46OjruWj7yYatUKhQKBQqFArVajT179ijEi4jIrqNgLFsmCALm\n5+eZnp7m6ugYV0bGyFXrOPEOQolO2jozeG6CscXqclieyFXx7phWdh2LwY4owx2tJRiDnTHCxqOw\nmKWQncMr5/DKOaKOYbC3h4GeDN1daXzfJ5/PM3XpEnPXRui1DAPRKIOdaWKrbPr7QJ+xXqd+8hT1\nV1/FGxtbPu7s34/7zDOEjjyKWVHGrVKvky0WWSgWWfSaZH2fApDMZMgMDZEeHCSdTpNOp0msUQnj\ng6jVauTzeYrFIvl8nsL8PLm5OQpzc+TnF7C9JgnLIo7BAcYCn6EnnuDIxz7GwMCANhOKiMiuoGAs\nH6piscjU1BSjY+NcuT7O5FwWK9qGHW8n2dlFPNXJbNlvzShnW7PK2XLjrvt0xt3ldcpDnTH6UlGa\n9SrFXJZSIU+znCeolfAqBRJRl3R7CifwqOay5MbG6LJs9sdiDHR0kEkmNyX4NcfGqL/6KvV3T0Kj\nNWaTSBA6dgz38WPYw8Or1kX2fZ/FcplsqUi2UmExCMh6Hn4kQnpggK7hYdI9PaTT6TVnl+v1+vKM\nb6FQIL+wQG52lsLcHIX5eag3SNgWcWORMJBwQsQjEZLRKIlwmNAd92w0m7w3M8PlUgm/K8ORz3yG\nRx55hMgWVgQRERH5qCkYy0eqXq8zPT3N5OQUl0fGGBmfxLMjWPF2Yqk0qXQXTTvMjYUKN7KtWeWx\nbPm26hcAjmUYaI8y0BElkwjTlQyTSYRpizjUa1VK+RzlYp56KY9fLVCYn6Gcy+LnF4k3qvQnYnRF\no3RHoqQTCRKRCPFIZNXQeC9+pULj7bepvfoa/szM8nGTSuEeO0bo8WPYg4P3DOM3Z5ezpRLZZoPF\nICDv+yQzGdJDQwS+T35ujvz8PF6lSsK2SBiLOJBwHBKRCIml8Hs/mxJnczkuLSxwI/DZ+/TTHH3q\nKXp7ez/w/URERB5UCsbyQPF9n/n5eaamprh6Y5wrI2MUqg2cRAehZJpUZ4ZYWwczhdryjPLoQoW5\nYm3V+7m2RSbhklkKyl2J1nMm4WL8BqVCjsW5OQqLc9TyC5TzC/jlIjELokGTkN8kGQqRicVIR6Ik\njCFmWSSWZlrjkQhR11015AZBgDc+TuPdk9RPnSLI5ZbPWZ2dhB4/hnvsGFZf34ZnrH3fJ1epsFAs\nYBmrFXwjESJLm/62Uq3R4Mr0NJcrZUL9Axz59Kc4dOgQ7ofw3iIiIh8GBWN54BUKBaampri+tPxi\nai6LHW/HjrWWX6TSGeq+4cZChel8ldlijblCjdlijXLdW/O+bRFnKSSHySTDdCVcMokwSRcqhTyl\nYp5qMU+ztEgln8WrlkjGIsRDNlF8rGaToF7DajSI2Q5xY4hbhjiGmOuSiIRpj8WJRyIEvo83eoP6\nyZM0Tp++VfYNsLq6bi236On5ML7S2wTNJt7UFN7YON7YGH42iz04gHPwIM7evZhVZpunFrNcXFhg\nynY48ImP8+gTT9DV1fWhj11ERGQzKRjLtlOr1ZiZmWF8YpLL129wfXwK34lixzsIJ9uJJZJE40nC\nkSjlepO5Yn05KM8Va8wWaiyU6jTvrB23xLYMnXF3eXa5K9kKzO0RC1MvUy7kqZby+NUifrVI0KiS\nisfoSMZoi0cIhULYvk+zVGJ25DoDzQbHevtIRqPAUvOQa9donDxF48wZglJp+b2t3l7cx48ROnYM\nO5PZ9O8uaDbxpqfxxsZaQXh8HG9qarkN9l1CIZz9+wkdOohz6BBWJnPb7HalXufK1BSXajUSe/dw\n9FOfYv/+/R95lQ0REZEPQsFYtj3f95mbm2NqaoqxyRlm5rPMZbMUK3UsN4YdiYEbw43GicaTxBJJ\nQpEouUpzOSjPFW8+6uQqd2/2uykaspfWL7vLs83pmEMkqFEr5qgUc/iVAl41T9Q2dKU7aJbyzF44\nzwHX5VN799GxotpE4Hk033uvFZLPnlku/QZgDwzcWm7R0fG+v5fA81ozwePjt0Lw5OTdIdgYrEwG\ne3AAe2AQq7MDb+Q6jcuX8Cenbr+0o53QwUM4hw7i7N+PFYu13isIGF9Y4NLiInNhl4efeYZHH3+c\njg8wbhERkY+KgrHsWI1Gg3w+Ty6XI5/PMzOfXX4UShWMG8FyY5hwjFA0QTTeelhujIVyoxWal5Zl\nzBXrzBZr1Jv+qu9lgFQ0REfcpTPm0hEP0RYyhIMaoUYRq7LI9PUrTF04w4Bj+Fh/HwcyHXQmEnQu\nbfbD82hevkz95Cka585B7da6aXt4eHkm2Wpru+v9A8/Dn56meTMAj421ZoKbd9eJboXgQeyBAZzB\nAeyBAcwa5ev8fJ7m5cs0Ll2mefnybbPbGIM9NIRz6CChQ4daGwptm2K1yuXpKa7U63QcOsTR48fZ\nt28f9oqydSIiIg8iBWPZlTzPI5/PLwfn2YUsM3OLzGazZHMFTCiCFY5BKIpzMzTHEnhOhPly87bA\nPFeskS3X7+rqt1LINnTEXNqjDlatSG12nGRlkX6qtNWzxLwKvck4Q20JepIJOlyX5MQEwZmzNM6f\nXy7/hjHYe/fiPn4MnBDe+NKSiMnJtUPwwMDSbPAATv8AJvrBSq4Fvo83MbEUlC/hjVwH/9Y/FEw0\ngrP/AM6hQ4QOHYRUitG5OS4XCizGohz+1KfY89BDRCIRXNclHA4/EEsuPM+jUqlQrVZve66Uy1Ty\neXzfp2/PHvr7+0mlUh/1cEVEZAspGIvcwff9Vi3gpeA8v7DI9PwCswuLzGUXCWwXOxzDhGJYkRjR\nRBI3mqDiOxQasFhpslCqs1Cuky3VWSjVqTTW3gQIEDU+KdMk5leJN0rEazli1QUG7AaHwgGHs7N0\nXh/BvTay5npgK52+FYIHB7H7+7GW1jW/H41mc0Ml6oJarbUMZGk22Z+bu308XV04B1uzyaW+Xq4s\nLjLj+zSCgHrQevYti3AsRiQaxY3FCC89IvE4bjzeOr4UosPh8PLPN59Xq+bRbDaXA+7KsFsuFqkU\nCpTzeaqFApVCkWqxgFevE7YsIsYQNqb1HND62XUJgoC5eo0pz4dkgsGHH2Zg/34FZRGRHUjBWOR9\nCIKAUqm0vDxjIbvIzHyW2YVFCqUSxXIVjIUJuVhOGOOECKwQDStMKXCpBA7FpkWhacjXArKVBovl\nBt46/2laBMSDBgmvTLKWZ092giPZcYYaBfraI3T2d2H6erGiMWzLwhiDZQy2ZWFZBttYWJZ165gx\nWCsajTQ9j8lslolCgUmvSdl2oNmkzbZIYGgzhmQkQioWIxmNEl2jPJu3sEDz0iWaly7TuHLltqUg\n2DbO3j3YQ8OYaBQTCWMiEQI3TDPk0LBt6o5D3XGoWYaG51NrNGgQUA+gbkwrTAM1P6ARBDQCH8d1\ncaMxwtEI9VqdSrEAjQZhy7oVco0hAoSNRcQNEQm5REMhwm6IaMh9X3WqC5UKk4tZZqpVppoetCWX\ng3JfXx/t7e0bvpeIiDx4FIxFNlEQBNTr9Vv/O37FzGW+UCJfLlMsViiUShRKFYrlMj4WJT/EWLbM\n1EIZ48QJYh1U7ChF41Ji/eBmBx7hoIkbNHCDJq7fIOw3Wr/79dbDqxPy67heDder41cLNKolyr5H\nzTK0pTN09vWS7uwklWqDIKDRqFOv1ahWq9SLRSqFPNV8AdOok7BtkrZFOuSQSSTobmujK5WiIx5v\n1XUOArwbN2gsBWVvbAw2+vfPGHBdTCRy+yPcCtMm2vodN4wfdmmGQjQdByeVIpJO4y5tCPwwFCoV\nphYXma5WloPywKFDDB44oKAsIrINKRiLfMTq9fpygM7lcpx8+23O/uZl2qpVupMpqlhMNBym/BBz\nQZi8HaMUSlK0YxTtKA3zwdfphm1D1LWJhmwijkXEMUQci7BjiDgQtlvXhG1wrYCwDbZXp1mrUq2U\nKebzlBfnqS4uUM0t4FVKuH6DtpBDOhQik4jSm0gwGHbpmpsjWigQ8jycZgO7Xieo1giq1dse1Ov3\n9X2aWAyrPYVpb8dKtWO1p7BW/Gza2jBbtBHwZlCeqbVmlP1EnMFDhxg4cID+/n4FZRGRB5yCscgD\nqFarceqddzj9qxMMNBoc67tVB7nRbFJtNKjW61TqDRbqDWYrHpM1j5lawGwTFuoBOd+ibLmUnSiN\nUBQTSRCEIviWQ80z1Jo+H/RvhGUgErKJuQ5x1ybm2sTCDhEbQiaAZhW/WqJZXKSRX6CRm8eu5HG9\nOqZRg0YVy2sSD9mk3BBtIYeOSJj2aJik69Jm262lHEstru1GA7vewKrXCWrVO8J0jaBaIcjl8XO5\ntWsy32QMJplsheX2dqxUCqujHZNKLR8z8fiGuxGuZzkoV6tMex5eIs7AwYMMHjxIX18fbW1tGGOW\nHyIi8tFSMBZ5gK0XkO+ULZWYyC4wWa8zFwS09feT3reP9o4OotEohUKRhVyB+cUci/kC2VyRSmDT\nsKM07Ag141K3wviWS9NyaAQ2jcCi0vCpNDwqda/13PDWLFt3LxHHIhpqzUyHHYOLj0MT26thGlVM\nvYxdL2PV8ljlHEFlEepVXMsQMhC1DYmQQ8J1CFk2rmURti1cY5FxQwx1dJIC7GIRu1DEKuSxCkVM\nPg/5HOTyUCzee1mH47QC880Z5pALIQcTcjEhB0IhTCgEztKzG8I4odvP3fls25RqtdbSi0qFac+j\nEvgEQUAABIC5uUbcsrEs0/p96Vjr2cKyVz631ovfvM6ybIxlsByHWDJJorOTeCpFPB4nFostP4dC\noQ/05ycistMpGItsA6sFZNdxGF9YYLJcYtLzsdtT7Dl6lKGlignhNWoTr1StVikWi5RKJYrFIvlC\nkfnFHAuLebL5ArlCkcA4WG4Uyw2DE8VyI4QiMQhFaOBQ9Q01D6oeVOoe5bpHqe5RrjcpLz2Xah7V\nhveBZqhvhWlwTUAIrxWm8bADHxM0Mb5Ho1ygkl3AaVRJBj5tIYuYbbD95tKjgeU3cJt14pUy8WqZ\nVK1Gsl6lrVolUasSr1aIVyq4jbWbvHxgxoDj3B6WHae1rMO2wbExtk1g2WBbrWNW61xgW0s/WwTW\nrd+Dpd9bx1sPLBvftqhaNhXHphxyKLphSo5DJQgo+x5OOEwslSLe0UGio4N4ezuJtrbl8ByPx4lG\no7dt0hQR2Q0UjEW2keWAfOIEvufR//DD7Dl8mMHBwS0pHRYEAZVKZTk4F4tFcvkC84t5srkC5WqV\naq1Ord6gVq+3KnJYNsZ2sGwHbAdjtQKej0Pd2FR9hzp26xHY1HxD3TNUPah6AZVmQKXhU663Zqm3\n4m+tTUDItJZ9OCbAARwCQsZvnfObhBs1Io0q0XoV12sQbtZwmw3cZg23Wcdt1Ig0a0vX3XxUl59j\njRqReo1Ys0a0XiMUfLBZ9k3jOJhYDBOLQjSKFw7juS71kEst5FB1w1QiYcrhMIWwS8l1sdJpoul0\nK0D//+3df4xmV1nA8e9z7n1/zY9lZ3fLtnRLAEOCNZii8suCFCVNbQiYAIEqEi0xakI0kT+IiRhi\nTAwhGoQoqCgQJNRIgPBHiS1/tBpBpNqFtpTCtluEbn9uZ6ftzL7vveecxz/Oed95Z3Zm95155513\nduf5JHfvufee8869c/bOPPfMc+89eJC5Q4cGj8nrp36cr7x+fqFys9mk3W4PnnVt6SXGmN1mgbEx\nFyHvPSKy594m572nruvBVFXVmuX+urPdXsqR7lX0ehXdXsXZXo9er6JXVXR7NVXtQRy1a1JJg642\n6FFSaUlXHQGXxo3VEZA8OYJCUMFH0mPffKCqA0Eh5jrTUKBp0kihARcDhQaK6CljoIx+MDWDpwye\nRqxphppG8DRCftpIqGmGinaoacdAJ9a0o6cTPK1QMxMqOr5itjrLbHeF2bMrtHpdXNxeYK5lSWy1\nCK0WvtXCt9uETpvQ7uDbLUK7je+08e0OPpdjWaIISMpaUYD0y2bNhY6qpjqkqVKlF5WeRoIIrdlZ\nOnNztOfm6MzNMXPgAO0DBwbB8/DU6XT2xAtjjDEXNwuMjTF7kqrivd80uB4Owru9im6VRq67vYpe\nr6Zb9ehVNVVd0+tVPLu8zFNPnWbxqdMsLy3TLpvMtWeYmZmHRofgSoJr4IuSIA2CK4hSEl1BlIIg\nBUFSQL4aiAte83ywDAGhVvI2phaM94nGQUBeakiBdvQ5yK5p+YpOqGj5Ko101z1m6x6dukvHVzSD\npxUqWr5OjwP0Na1+21DRzMv9dQ0NuEZBbLeJnQ7aaUNnBp3pIDOz6ckhc7Npmp2jnJujbDYpi4LS\nOcqiQFXp5RtNe3W60bRX12k5RnoCPaCnSk+VboxI2aA9O0tnfjWY7szP05ydRWNEYySGQPQ+vclx\n/TpVovdE1bQuhMF2YiSESIxhzWe5oqDV6dCYnaU9M0Or06E1P3/Oy2iazeaaaa9d1BpjEguMjTH7\nRj/YXl5e5uTJkzx433386N77mK0rnl8UHJmfp9NoUodAz3t6PtALkW4IaeTZB7xGokKMmp5bHQLP\ndrus9Hos+8BKzuMt2m06Bw7QmZ+n1Zml1e5QNFvUKvgAXiGKw+MIKsRBwD0cfPe3CV7dmjop6M5z\nFYKmzwwxB+eDaXrpCE1fDQXQKQAfBNlr1teUmkbNm7GmDIFSPQ1NeeSD0XZRShQnUOapcKS0mEIo\nAOcgfccUQVGN4FLaTAsoRSidyy/AcZSFw4mjcKsvwBFxCCAOhNWnhgggki5ynEvLESEIVAg14FEq\nhKpwVM5RFQVVWdAVoeeK9NKaGHGNJs1Om1a7fe5bH2dmaLbbNBoNGo0GZVlSluWm5bIsLe3EmB2y\n5wNjEbkB+ChQAJ9S1Q9vUOdjwK8CK8BvqerdG9SxwNgYc44QAqdOneLkAw/w4PHjtJaXuaosuerQ\nYY7Mzw/qxRg5s7LC0889x1L3LIsxciZGQrPJoSuu4MgLX8ihyy/n8OHDLCwsjHTzo6oSY34yxQXK\nMUZCCHjvB/Ph8vC6Xh5dreuQR1sDK13P2bpmpQ50q0DXR7p1oFtHeiFShzSyHaVIAXl/jhukq/ic\nolIFpY5KHZQQc0CugkoK3lUcUfbujXtlyOkrGlJZY05nCYO0lkYMNEKd56sj7I3oaQRPM8/72xo5\nBWZQL6yuK4e2NTSkfHYiJUpBpBSl0FQWJwTn0lQWBFcQigJfFniX50VJXTh8UeCLgqpwaLOFtFrQ\n7lB02mlUfnYGNzNDOTuLzM7S6nQo2+1B0F0UBWVZDubD00ZBeFEUFEVhN2WaS9qeDoxFpAAeAN4E\nPAJ8G7hJVe8fqnMj8D5VvVFEXg38taq+ZoPPssD4InPHHXdw3XXXTXs3zIguhf5SVZ544glOnjjB\nQ8ePUz3xJIeLgmc18hwwf9llHDl2jMPHjnH48GEOHTrE3NzcRTtaN9xnqrpp4H2+aThgXzyzxNLS\nEkvPPMPi6ad5+qmnWVx8hpWzPQopcTicFFA0oWggjRYeR6VCjIqPKThXVtNXVPpjv2mKOfiOIqvz\nvC7kgDysn1xKg/GuQPdwwA5QxIDTiKjiVHEacSiiKT/9uR/fx/OufBlO03ekiGl0fFBXFclzR8RF\nze3TOvrfSU0TgNMIQ9nfmv9VJM9Xtw7K/f/z0i9LesA5oP0bKp2kbeIGZXGpLM6lbc4hLo3iN0qh\nWRY0GgXNRkG72aDZatBpNWk2S9rtJs3C0W6WdJoN2q0GnWZJqyxpN0varQbtRip3Wk06rQatsqBR\nOBrF9J4Vfin8bNxPdiIwnuTdDq8CTqjqwwAicgvwVuD+oTpvAT4LoKrfEpGDInJUVR+f4H6ZXWA/\nTC4ul0J/iQhHjx7l6NGjvObaa1lcXOT06dMcPHiQhYWFSy4vdLjPRGQwQrjT6rpOAfOFLP+xAAAK\ntElEQVTSEouLi5x57DEWH32MpScex9U1885xQByFCCEGajw+KnWMRI0pT1tZLUfF54AtqBJUiar4\nGHFFgcsjm9IvO0eIKW/Y+/TZXtNodx0UHyJe042Cw0F3P1D3+cbNWlNqS91PZ8mpLB6HJ80Ho+w5\nH70fqKsUecplt1qOOaCPkkeI2fz/2ZnHHsS/9Jyxn4tfBKo8naO/YWWsL5H+OtC/0TX9xaDoX4QM\nLhjI5TyXfjnH8eme0hznyyD2d05wIoO5uFzWyD13fplrXv8QLgYkRlwIOA0QIhoVNOWskydRTctK\nWqe6OsV8aaJphzRfaKg46D+yMZcHy86hRbHmYsQ5oSj66UQFrnBpuUznTun6x5DWFy7d7O2KnIZU\nrP2M1L4YzIvCrSkXuR0MP6kmpSoNFhi63upfbNF/cs1w3aHPQFhzvSMyqNdv1y8X/eN2Q2WRwb6l\n/t2Zi6dJBsZXAj8eWv4J8OoR6hwDLDA2xoxlYWGBhYWFae/GRa/RaHDkyBGOHDlyzraVlRWWlpY4\nc+YMMcY1f67fatnll6Fs1/o0ln4Ky4XWjfKYuvWj7N77wWh7fzmEQK/2qKbgvw4e7wNV7QkxUvvA\n5/+v5O3XNqi9x/uID4HaB3wI+BCp89yHgK9T+xADdRXQqkarCg0BDXE1MMsBmSppJDn2R54ZlNHV\nUWaJ/RFtQFPg5mIKIPvb0XTTp+RtoEjM+d/Kmjagq+k74vD5rwXepQsF70qCK6hdQe1KalemFBLX\nwLuCqijTS4lcQV008jzXcwVVkW6m9UWJpwQunOY0FoV81y0AZ+IMS/FoWnB52gtinvyoDdYdmNnU\nJAPjUXMf1v8ktJwJY4y5CMzMzDAzM8MVV1wx7V0ZBLF7OYf2f79xJ+948/XT3o2J6V+crJ9ijOnJ\nIXmKdU2s60FZfVhd9v31q/WD99S1p1d5fJ1urq3rSF2np9tEH/BVRfCRUNf4uibWnugDwed5yE8w\nCZEYIhoCMWq6ETcqUdPFRf/G3KhCFOH2J3/AdQ98jSBCzMF+6I/s5mFolcHQM7omBSUtSx79lVwH\nJ+mCJergAqR/wbI6D6sXJhpXL3Ji+mtLWp1TZTTf84AMHrEYkRw7C/0knDi4tTXdPDyoJ0OpToPy\n8DpHdGNctJ4T5m1NzCPQ6a83q/u2ZjmnZf1orK+UTDLH+DXAh1T1hrz8x0AcvgFPRD4J3KGqt+Tl\n7wNvWJ9KISIWLBtjjDHGmPPayznGdwEvFZEXAaeAdwI3ravzVeB9wC05kD6zUX7xuAdpjDHGGGPM\nhUwsMFZVLyLvA/6N9Li2f1TV+0Xkd/P2v1PVW0XkRhE5ASwDvz2p/THGGGOMMeZ8LooXfBhjjDHG\nGDNpU71LQURuEJHvi8gPReQDm9T5WN7+HRF5xVbamp03Zp89LCLfFZG7ReS/d2+v97cL9ZmIvExE\nvikiXRF5/1bamskYs8/sPNtlI/TXb+Sfh98Vkf8UkZ8dta2ZjDH7zM6xKRihz96a++xuEfkfEfnl\nUduusdldpJOeSOkVJ4AXAQ3gOPDT6+rcCNyay68G/mvUtjbtrT7LyyeBQ9M+jv00jdhnlwG/APw5\n8P6ttLVpb/VZ3mbn2d7rr9cCz8vlG+x32cXbZ3nZzrG92WezQ+WXk96lseXzbJojxoMXgKhqDfRf\nADJszQtAgIMicvmIbc3O226fHR3abjdS7q4L9pmqPqmqdwH1VtuaiRinz/rsPNs9o/TXN1V1KS9+\ni/S8/pHamokYp8/67BzbXaP02fLQ4hzw1Khth00zMN7o5R5XjljnBSO0NTtvnD6D9HjFr4vIXSLy\nOxPbSzNslD6bRFuzfeN+3+08211b7a/3Ardus63ZGeP0Gdg5Ng0j9ZmI/JqI3A98DfiDrbTtm+Tj\n2i5kuy8AMdMzbp+9TlVPichlwO0i8n1V/Y8d2jezsXHurrU7c6dj3O/7tar6qJ1nu2bk/hKRNwI3\nA9duta3ZUeP0Gdg5Ng0j9ZmqfgX4ioi8HviciLxsq19omiPGjwBXDS1fRYriz1fnWK4zSluz87bb\nZ48AqOqpPH8S+DLpzxtmssY5V+w8m46xvu+q+mie23m2O0bqr3zz1j8Ab1HVxa20NTtunD6zc2w6\ntnSu5AuVEjiU643cdpqB8eAFICLSJL0A5Kvr6nwVeA8M3qTXfwHIKG3Nztt2n4nIjIjM5/WzwPXA\nPbu36/vWVs6V9SP9dp5Nx7b7zM6zqbhgf4nIC4EvAe9W1RNbaWsmYtt9ZufY1IzSZz8lIpLLPweg\nqqdHaTtsaqkUOsYLQDZrO50j2T/G6TPgcuBL+f9sCXxeVW/b/aPYX0bps3xD67eBA0AUkT8ErlbV\n5+w8233j9BnwfOw821Wj9Bfwp8AC8IncN7Wqvsp+l03HOH2G/S6bihH77G3Ae0SkBp4D3nW+tpt9\nLXvBhzHGGGOMMUz5BR/GGGOMMcbsFRYYG2OMMcYYgwXGxhhjjDHGABYYG2OMMcYYA1hgbIwxxhhj\nDGCBsTHGGGOMMYAFxsaYfUJEPiMib8jl60RkRUTuHpo+M1T3qIh8c2j5YRG5X0SOi8g9IvLOvL4p\nIn8rIveKyHfytpuGvsa3L7BP14vIXSLSFZGPrNv2ERF5xzaPda8c3weH2t4lItef7/hE5OHtHK8x\nxuyUqb3gwxhjdpnmqe8+VX3lJnU/AHxyXdu3qer3ROQa4Bsi8nXgZtJLAF6uqioiM8ALtrBPDwLv\nBd4OtNdt+yvgNuBft/B5w/bC8X0L+IiqdiW9XvdOEblcVXtsfHz2YH1jzFTZiLExZj9Z/9rrcyuI\nlMCvA1/caLuqHgeeBV4MXAk8rvlNSaq6su6Vv+elqg+q6ncAv8G2R4GnROQXR/28Uezy8d2mqt28\neA/p+384b5vI8RljzDgsMDbG7FdXr0s1+JO8/hXAKVVdXldfAETkjUAL+AHwKeAdIvJdEfmEiLx1\nh/fxG8CvbLPtXju+9wAnVPXU0Lpxjs8YY3acpVIYY/ar722SavBi4JF16wT4ooh0gSVS2sEzwL0i\n8hLgl4BrgY+LyA2q+vs7tI8/AV61zbZ75vhybvefAW9at2mc4zPGmB1ngbExxqy1UZ7rIAf3nA0p\nX/Z24HYRuTWXdyowhg3SP/JNbB/Oi/+sqn+5hc/b1eMTkdcCnwPeoqo/3KjKqJ9ljDGTZoGxMcas\n9TApt/aCROR1wAOq+mRe9fPAQ9v4mpsFh8c2+jxVvY1049p2PMwuHZ+IvBL4F1LQfXyDKhsenzHG\nTIsFxsaY/epqEbl7aPkRVX0zcBy4UkRmN8jDXe9FwMdEpAkE4HHg3UPbz/uUhRx4fgE4kBblXcDN\nqnp7rvJa4IOjHtC6rzv14wP+hpSv/Pcig9j/N1X13lze7vEZY8xESL7Z2BhjLmki8mngM6p65wh1\nPwrcraqfnfyebboPVwBfV9WfmcBn78njE5GTqvriae2TMcbYUymMMeZcfwH83pT34Y+AD03osy/1\n4zPGmG2xEWNjzL6QR4w/rar/vstf9xrg0xts+riq/tNu7ssk7OTxichDqvqSndkzY4zZOguMjTHG\nGGOMwVIpjDHGGGOMASwwNsYYY4wxBrDA2BhjjDHGGMACY2OMMcYYYwALjI0xxhhjjAEsMDbGGGOM\nMQaA/wc3MoO0EKmp5wAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x2acb6150>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "PLOTNAME = 'figure2A'\n",
    "PLOTPATH = 'plots/'\n",
    "plot_delta_expected_majiq_others(psis_diff_dict, args['rep_names'], plotpath=PLOTPATH, plotname=PLOTNAME)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjcAAAGoCAYAAAC68MSlAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nQgiRUf4fTtocYF3XwKIAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1156d2d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "better_worse = []\n",
    "for edx, ed in enumerate(psis_diff_dict['miso']):\n",
    "    better_worse.append(ed-psis_diff_dict['majiq'][edx])\n",
    "plot_delta_R({'miso': better_worse}, args['rep_names'], plotpath=PLOTPATH)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "----\n",
    "# Data for numbers quoted in MAJIQ paper\n",
    "\n",
    "*Defining Ψ reproducibility as the difference between replicates (R(Ψ) = |E[Ψr1]-E[Ψr1]|) we found __81.2%__ of the junctions in quantifiable LSVs were reproducible within 5% (R(Ψ) < 5%, Fig2A)*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "81.2%\n"
     ]
    }
   ],
   "source": [
    "n_total_juncs = np.sum([dd.size for dd in psis_diff_dict['majiq']])\n",
    "R_PSI_less_05 = np.sum([np.count_nonzero(dd <0.05) for dd in psis_diff_dict['majiq']])\n",
    "\n",
    "print \"%.1f%%\" % (100*R_PSI_less_05/n_total_juncs)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "*Comparing MAJIQ to the commonly used MISO algorithm [] we found MAJIQ improved Ψ reproducibility for approximately two thirds of the LSVs (P(ΔR = RMISO-RMAJIQ)> 0 = __XX__%) and over two fold more showed a significant improvement (P(ΔR>0.05) = __21%__), P(ΔR< -0.05) = __10%__)*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "61.7%\n"
     ]
    }
   ],
   "source": [
    "n_majiq_win = np.sum([np.count_nonzero(dd < psis_diff_dict['miso'][idx]) for idx, dd in enumerate(psis_diff_dict['majiq'])])\n",
    "print \"%.1f%%\" % (100*n_majiq_win/n_total_juncs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "21.2%\n"
     ]
    }
   ],
   "source": [
    "majiq_improved = np.sum([np.count_nonzero((psis_diff_dict['miso'][idx] - dd) >0.05) for idx, dd in enumerate(psis_diff_dict['majiq'])])\n",
    "print \"%.1f%%\" % (100*majiq_improved/n_total_juncs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "10.1%\n"
     ]
    }
   ],
   "source": [
    "miso_improved = np.sum([np.count_nonzero((psis_diff_dict['miso'][idx] - dd) <-0.05) for idx, dd in enumerate(psis_diff_dict['majiq'])])\n",
    "print \"%.1f%%\" % (100*miso_improved/n_total_juncs)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}