20200402_Hovestadt_Human_MB_scRNAseq.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "----\n",
    "# Re-anlaysis of Hovestadt et al.,\n",
    "\n",
    "L.Richards    \n",
    "April 2020    \n",
    "\n",
    "/mnt/work1/users/pughlab/projects/SinghLab_scMedullo/Brananvan_Wnt/PublicData_Analysis_April2020/\n",
    "\n",
    "----\n",
    "  \n",
    "\n",
    "  \n",
    "**Analysis Plan:** \n",
    "1. Download public human medulloblastoma scRNAseq data\n",
    "2. Cluster cells to generate tSNE using same methods as in Manoranjan et al. (similar to Figure 2)\n",
    "3. Score cells with GSVA as in Manoranjan et al. Figure 2A\n",
    "4. Determine significance threshold for cell classification based on gene signature scores\n",
    "5. Classify cells as Wnt + or Wnt -, as in as in Manoranjan et al. Figure 2b,d\n",
    "\n",
    "---\n",
    "# Updated Methods for Manoranjan et al., 2020\n",
    "---\n",
    "\n",
    "** Analysis of public scRNA-seq data**\n",
    "\n",
    "Publically available human medulloblastoma tumour scRNA-seq data was downloaded from (25 tumours, 7745 malignant cells)\n",
    "\n",
    "Public data from (log2(TPM/10+1)), we have to reverse the log2 and re-transform into ln for use in seurat. \n",
    "\n",
    "Relative gene expression was calculated by centering expression across all cells in the cohort using the ScaleData() function in Seurat. Identification of highly variable genes (4140 genes), principal component analysis (11 signicificant PCs determined by a scree plot) and SNN-Cliq-\n",
    "inspired clustering were performed in Seurat to generate t-distributed stochastic neighbor embedding (t-SNE) visualizations (Rtsne: Use the Rtsne package Barnes-Hut implementation of tSNE; default in Seurat)\n",
    "\n",
    "** Data Availability **\n",
    "\n",
    "Raw single cell RNA-sequencing data is available in the form of fastqs at [Singh Lab to insert accession and database information (ie. GEO)]. Processed single-cell RNA-sequencing data is available through the Broad Single Cell Portal (https://singlecell.broadinstitute.org/single_cell/study/SCP840).\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "----\n",
    "## 1.0 Download data\n",
    "----\n",
    "    \n",
    "    \n",
    "1. [Hovestadt et al. (Northcott Lab)](https://www.nature.com/articles/s41586-019-1434-6)  \n",
    "    - 25 human medulloblastomas (23 diagnostic, 2 recurrent, more detail in Figure 1B)\n",
    "    - scRNA-seq performed with Smart-seq2 (7745 malignant cells)\n",
    "    - Normalized (log2(TPM/10+1)) expression matrices are available at [GSE119926]( GSE119926)\n",
    "    - This dataset will be used for downstream analysis, excluding xenografts\n",
    "\n",
    "Data location: /mnt/work1/users/pughlab/projects/SinghLab_scMedullo/Brananvan_Wnt/PublicData_Analysis_April2020/data/Hovestadt"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---\n",
    "## 2.0 Process Hovestadt et al. Dataset\n",
    "---\n",
    "\n",
    "- 25 tumours (5 WNT, 3 SHH, 8 GR3, 9 GR4)   \n",
    "- 7745 malignant cells\n",
    "- Tumours subtypes based on DNA methylation  \n",
    "- More details in [Figure1B](https://www.nature.com/articles/s41586-019-1434-6/figures/1) of paper  \n",
    "\n",
    "> Combine expression matrices across 25 samples   \n",
    "> Make a metadata file to annotate  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "setwd(\"~/pughlab/projects/SinghLab_scMedullo/Brananvan_Wnt/PublicData_Analysis_April2020/\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "## reformat metadata downloaded from GEO\n",
    "\n",
    "meta.data <- read.table(\"data/Hovestadt/Meta_GSE119926_series_matrix.txt\", \n",
    "                        header = F,\n",
    "                        sep = \"\\t\"\n",
    "                       )\n",
    "meta.data <-t(meta.data)\n",
    "colnames(meta.data) <- meta.data[1, ]\n",
    "meta.data <- data.frame(meta.data[-1, ])\n",
    "#head(meta.data)\n",
    "\n",
    "###subset only tumours, no xenografts\n",
    "#table(meta.data$Sample_source_name_ch1) \n",
    "meta.data <- meta.data[-grep(\"xenograft\", meta.data$Sample_source_name_ch1), ]\n",
    "\n",
    "\n",
    "saveRDS(meta.data, file = \"./Clustering/Hovestadt_MB_Tumours_MetaData.rds\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---\n",
    "\n",
    "### Create merged expression matrix and metadata file"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "suppressMessages(library(\"Seurat\", \n",
    "                         lib =\"/mnt/work1/users/pughlab/projects/BTSCs_scRNAseq/Manuscript/lib/\" ))\n",
    "\n",
    "suppressPackageStartupMessages({\n",
    "    #library(GSEABase)\n",
    "    library(GSVA)\n",
    "    #library(Biobase)\n",
    "    library(RColorBrewer)\n",
    "    library(parallel)\n",
    "    #library(GSA)\n",
    "    library(qvalue)\n",
    "    library(taRifx)\n",
    "})\n",
    "\n",
    "setwd(\"~/pughlab/projects/SinghLab_scMedullo/Brananvan_Wnt/PublicData_Analysis_April2020/\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "### merge normalized expression matrices for tumour samples (n=25)\n",
    "### should have a total of ~7745 cells\n",
    "\n",
    "meta.data <- readRDS(\"./Clustering/Hovestadt_MB_Tumours_MetaData.rds\")\n",
    "meta.data <- remove.factors(meta.data)\n",
    "\n",
    "samples <- as.character(meta.data$Sample_title)\n",
    "samples\n",
    "\n",
    "files <- list.files(\"~/pughlab/projects/SinghLab_scMedullo/Brananvan_Wnt/PublicData_Analysis_April2020/data/Hovestadt/\", \n",
    "                    pattern = \"txt.gz\"\n",
    "                    )\n",
    "\n",
    "    file <- paste0(fastq.path, files[grep(samples[1], files)])\n",
    "    print(samples[1])\n",
    "    print(file)\n",
    "    raw.dat <- read.table(file)\n",
    "    print(dim(raw.dat))\n",
    "\n",
    "    sample.meta <- meta.data[meta.data$Sample_title == samples[1], ] #subset to metadata of loaded sample\n",
    "    sample.meta <-  sample.meta[rep(seq_len(nrow(sample.meta)), ncol(raw.dat)), ] #repeat metadata for # cells\n",
    "    rownames(sample.meta) <- colnames(raw.dat)\n",
    "\n",
    "    merged.exp <- raw.dat\n",
    "    merged.meta <- sample.meta\n",
    "    \n",
    "\n",
    "for (i in 2:length(samples)){\n",
    "    \n",
    "    file <- paste0(fastq.path, files[grep(samples[i], files)])\n",
    "    print(samples[i])\n",
    "    print(file)\n",
    "    raw.dat <- read.table(file)\n",
    "    print(dim(raw.dat))\n",
    "\n",
    "    sample.meta <- meta.data[meta.data$Sample_title == samples[i], ] #subset to metadata of loaded sample\n",
    "    sample.meta <-  sample.meta[rep(seq_len(nrow(sample.meta)), ncol(raw.dat)), ] #repeat metadata for # cells\n",
    "    rownames(sample.meta) <- colnames(raw.dat)\n",
    "    print(dim(sample.meta))\n",
    "\n",
    "    if(identical(rownames(merged.exp), rownames(raw.dat))){\n",
    "        \n",
    "        merged.exp <- cbind(merged.exp, raw.dat)\n",
    "        print(dim(merged.exp))\n",
    "        \n",
    "    } else {print(\"ERROR IN GENE ORDERING\")}\n",
    "    \n",
    "   if(identical(colnames(merged.meta), colnames(sample.meta))){\n",
    "        \n",
    "        merged.meta <- rbind(merged.meta, sample.meta)\n",
    "        print(dim(merged.meta))\n",
    "       \n",
    "    } else {print(\"ERROR IN METADATA NUMBERS\")}\n",
    "\n",
    "}\n",
    "\n",
    "saveRDS(merged.exp, file = \"./data/Hovestadt_MergedExpressionMatrix.rds\")\n",
    "saveRDS(merged.meta, file = \"./data/Hovestadt_MergedMetaData.rds\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---\n",
    "## 3.0 Cluster Hovestadt et al. Dataset with Seurat\n",
    "---\n",
    "\n",
    "Keep methods consistent with current clustering methods in Branavan's manuscript (aka those used to generate tSNE in Figure 2A-B).\n",
    "\n",
    "Publically downloaded data is already normalized, so we will skip the LogNormalize step in Seurat, and procede with ScaleData() and all subsequent steps as listed in methods. \n",
    "\n",
    "/mnt/work1/users/pughlab/projects/SinghLab_scMedullo/Brananvan_Wnt/PublicData_Analysis_April2020/Clustering"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "setwd(\"/mnt/work1/users/pughlab/projects/SinghLab_scMedullo/Brananvan_Wnt/PublicData_Analysis_April2020/Clustering\")\n",
    "\n",
    "suppressMessages(library(\"Seurat\", \n",
    "                         lib =\"/mnt/work1/users/pughlab/projects/BTSCs_scRNAseq/Manuscript/lib/\" ))\n",
    "\n",
    "suppressPackageStartupMessages({\n",
    "    #library(GSEABase)\n",
    "    library(GSVA)\n",
    "    #library(Biobase)\n",
    "    library(RColorBrewer)\n",
    "    library(parallel)\n",
    "    #library(GSA)\n",
    "    library(qvalue)\n",
    "    library(taRifx)\n",
    "})\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "### load in data\n",
    "\n",
    "log2.TPMs <- readRDS(\"~/pughlab/projects/SinghLab_scMedullo/Brananvan_Wnt/PublicData_Analysis_April2020/data/Hovestadt/Hovestadt_MergedExpressionMatrix.rds\")\n",
    "meta <- readRDS(\"~/pughlab/projects/SinghLab_scMedullo/Brananvan_Wnt/PublicData_Analysis_April2020/data/Hovestadt/Hovestadt_MergedMetaData.rds\")\n",
    "\n",
    "# Public data from (log2(TPM/10+1))\n",
    "# We have to reverse the log2 and re-transform into ln for use in seurat\n",
    "# This matrix can be used as input for normalized, log-transformed counts\n",
    "\n",
    "TPMs <- 2^log2.TPMs\n",
    "log.TPMs <- log(TPMs) ##this is the input for seurat@data slot\n",
    "\n",
    "\n",
    "##run Seurat on data\n",
    "##do not perform additional QC, since public data is already QC'ed\n",
    "\n",
    "Hovestadt <- CreateSeuratObject(raw.data = log.TPMs,\n",
    "                                min.cells = 0,\n",
    "                                min.gene = 0,\n",
    "                                normalization.method = NULL,\n",
    "                                meta.data = meta\n",
    "                               )\n",
    "\n",
    "Hovestadt <- FindVariableGenes(object = Hovestadt, \n",
    "                               mean.function = ExpMean, \n",
    "                               dispersion.function = LogVMR, \n",
    "                               x.low.cutoff = 0.0125, \n",
    "                               x.high.cutoff = 3, \n",
    "                               y.cutoff = 0.5\n",
    "                              )\n",
    "\n",
    "length(Hovestadt@var.genes) #6571 variabe genes\n",
    "\n",
    "Hovestadt <- ScaleData(object = Hovestadt, \n",
    "                       #vars.to.regress = c(\"nUMI\", \"percent.mito\")\n",
    "                      )\n",
    "\n",
    "Hovestadt <- RunPCA(object = Hovestadt, \n",
    "                    pc.genes = Hovestadt@var.genes, \n",
    "                    do.print = TRUE, \n",
    "                    pcs.print = 1:5,\n",
    "                    genes.print = 5\n",
    "                   )\n",
    "\n",
    "pdf(\"Hovestadt_elbowPlot.pdf\")\n",
    "PCElbowPlot(object = Hovestadt) ## 15 PCs\n",
    "dev.off()\n",
    "\n",
    "Hovestadt <- FindClusters(object = Hovestadt, \n",
    "                     reduction.type = \"pca\", \n",
    "                     dims.use = 1:15, \n",
    "                     resolution = 0.6, \n",
    "                     print.output = 0, \n",
    "                     save.SNN = TRUE\n",
    "                    )\n",
    "\n",
    "Hovestadt <- RunTSNE(object = Hovestadt, \n",
    "                     dims.use = 1:15, \n",
    "                     do.fast = TRUE\n",
    "                    )\n",
    "\n",
    "saveRDS(Hovestadt, file = \"Hovestadt_Seurat.rds\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "----\n",
    "## 3.0 Score Cells with Gene Signatures (GSVA)\n",
    "----\n",
    "\n",
    "Reviewer says they want \"WNT signature cells\"to be assessed. Is this referring to the Wnt signature (indicative of Wnt pathway activity) or Wnt Subgroup gene signature?\n",
    "\n",
    "Figure 2A of Manoranjan et al., is visualizing the Wnt subgroup score with GSVA. \n",
    "\n",
    "For information on gene signatures, see Table S2: https://docs.google.com/spreadsheets/d/1bi3wwzgs-CiGIrqao--raF06YBjipw_m2SjXDE49h9I/edit?usp=sharing\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "## load in gene signatures from Singh Lab\n",
    "load(\"~/pughlab/projects/SinghLab_scMedullo/Brananvan_Wnt/data/geneSets_SinghLab.Rdata\")\n",
    "str(gene.sets)\n",
    "\n",
    "## score genes signatures with GSVA on scaled data matrix \n",
    "\n",
    "dat <- data.matrix(Hovestadt@scale.data)\n",
    "\n",
    "EnrichmentScores <- gsva(expr = dat, \n",
    "                         gset.idx.list = gene.sets, \n",
    "                         method = \"gsva\",\n",
    "                         min.sz = 2, \n",
    "                         max.sz = Inf, \n",
    "                         mx.diff = TRUE, \n",
    "                         verbose = T, \n",
    "                         parallel.sz = 10\n",
    "                        )\n",
    "\n",
    "saveRDS(EnrichmentScores, file = \"EnrichmentScores_Hovestadt_ScaledData.rds\")\n",
    "\n",
    "meta.GSVA <- data.frame(t(EnrichmentScores))\n",
    "\n",
    "Hovestadt <- AddMetaData(Hovestadt, \n",
    "                         metadata = meta.GSVA\n",
    "                        )\n",
    "\n",
    "saveRDS(Hovestadt, file = )"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---\n",
    "## 4.0 Classify cells\n",
    "---\n",
    "\n",
    "Same method as currently used in Branavans paper. Create 100 random gene lists with same length as gene signature of interest. Score all cells for random gene signatures. Determine a 5% cutoff based on random scores. If gene signature of interset score in a cell surpasses cutoff, it is classified as being \"on\" for that gene signature.\n",
    "\n",
    "/mnt/work1/users/pughlab/projects/SinghLab_scMedullo/Brananvan_Wnt/PublicData_Analysis_April2020/GSVA/Random_Sigs"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "----\n",
    "### 4.1 Generate 100 random gene sets per gene signature\n",
    "----"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "suppressMessages(library(\"Seurat\", \n",
    "                         lib =\"/mnt/work1/users/pughlab/projects/BTSCs_scRNAseq/Manuscript/lib/\" ))\n",
    "\n",
    "suppressPackageStartupMessages({\n",
    "    #library(GSEABase)\n",
    "    library(GSVA)\n",
    "    #library(Biobase)\n",
    "    library(RColorBrewer)\n",
    "    library(parallel)\n",
    "    #library(GSA)\n",
    "    library(qvalue)\n",
    "    library(taRifx)\n",
    "})\n",
    "\n",
    "\n",
    "### load in data and gene signatures\n",
    "Wnt_medullo <- readRDS(\"../Hovestadt_Seurat_GSVA.rds\")\n",
    "load(\"~/pughlab/projects/SinghLab_scMedullo/Brananvan_Wnt/data/geneSets_SinghLab.Rdata\")\n",
    "str(gene.sets)\n",
    "\n",
    "\n",
    "### Generate random gene sets\n",
    "### 100 gene sets of same length of signature of interest\n",
    "\n",
    "for (i in 1:length(gene.sets)){\n",
    "    \n",
    "    sig.name <- names(gene.sets)[i]\n",
    "    print(sig.name)\n",
    "    \n",
    "    sig.length <- length(gene.sets[[i]])\n",
    "    print(sig.length)\n",
    "    \n",
    "    print(\"Generating random lists\")\n",
    "    random.sigs <- lapply(1:100, function(x) sample(rownames(Wnt_medullo@data),size = sig.length))\n",
    "    \n",
    "    names(random.sigs) <- paste0(\"Random\", 1:100, \"_\", sig.name)\n",
    "        \n",
    "    file1 <- paste0(sig.name, \"_random_lists.rds\")\n",
    "    print(file1)\n",
    "    saveRDS(random.sigs, file = file1)\n",
    "\n",
    "    \n",
    "}\n",
    "                        "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "----\n",
    "### 4.2 Score random gene sets with GSVA\n",
    "----"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "##### GSVA on random sigs \n",
    "\n",
    "dat <- as.matrix(Wnt_medullo@scale.data)\n",
    "\n",
    "files <- list.files(pattern = \"random_lists.rds\")\n",
    "\n",
    "for (i in 1:length(files)){\n",
    "    \n",
    "    print(\"---------\")\n",
    "    print(files[i])\n",
    "    print(Sys.time())\n",
    "    \n",
    "    random.sigs <- readRDS(files[i])\n",
    "    #print(head(random.sigs))\n",
    "    \n",
    "    print(\"Scoring random lists with GSVA\")\n",
    "    print(\"Scoring on scaled data\")\n",
    "    print(Sys.time())\n",
    "    \n",
    "    EnrichmentScores <- gsva(expr = dat, \n",
    "                         gset.idx.list = random.sigs, \n",
    "                         method = \"gsva\",\n",
    "                         min.sz = 2, \n",
    "                         max.sz = Inf, \n",
    "                         mx.diff = TRUE, \n",
    "                         verbose = T, \n",
    "                         parallel.sz = 10\n",
    "                        )\n",
    "    \n",
    "    print(\"Saving GSVA results\")\n",
    "    print(Sys.time())\n",
    "    \n",
    "    filename <- gsub(\"_random_lists.rds\", \"_GSVA_SCORES.rds\", files[i])\n",
    "    print(filename)\n",
    "    saveRDS(EnrichmentScores, file = filename)\n",
    "    \n",
    "    print(\"\")\n",
    "    print(\"\")\n",
    "    print(\"\")\n",
    "    print(\"\")\n",
    "    \n",
    "}"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "----\n",
    "### 4.3 Define 5% cutoff for cell classification\n",
    "----\n",
    "\n",
    "Each gene signature has its own cutoff"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "path <- \"/mnt/work1/users/pughlab/projects/SinghLab_scMedullo/Brananvan_Wnt/PublicData_Analysis_April2020/GSVA/Random_Sigs/\"\n",
    "files <- list.files(path, pattern = \"GSVA_SCORES.rds\")\n",
    "files\n",
    "\n",
    "for (i in 1:length(files)){\n",
    "\n",
    "    load.file <- paste0(path, files[i])\n",
    "    print(load.file)\n",
    "    EnrichmentScores <- readRDS(load.file)\n",
    "    print(dim(EnrichmentScores))\n",
    "    \n",
    "    sig.name <- gsub(\"_GSVA_SCORES.rds\", \"\", files[i])\n",
    "    print(sig.name)\n",
    "    \n",
    "    bb <- as.vector(as.matrix(EnrichmentScores))\n",
    "    length(bb)\n",
    "    \n",
    "    d <- density(bb)\n",
    "    cutoff <- qnorm(0.95, mean = mean(bb), sd = sd(bb))\n",
    "\n",
    "    filename <- paste0(path, sig.name, \"_GSVA_RandomListCutoff.pdf\")\n",
    "    print(filename)\n",
    "\n",
    "    pdf(filename)\n",
    "\n",
    "    plot_raw <- plot(d,\n",
    "                 main = paste0(sig.name,\" - Random Signatures\"),\n",
    "                 xlab = \"GSVA Enrichment Score\"\n",
    "                )\n",
    "    legend(\"topright\",\n",
    "       legend = c(paste(\"Mean:\", round(mean(bb),3)),\n",
    "                          paste(\"Sd:\", round(sd(bb), 3)),\n",
    "                          paste(\"Cutoff:\", round(cutoff,3))\n",
    "                     ),\n",
    "               bty = 'n',\n",
    "               cex = 1\n",
    "          )\n",
    "        abline(v=cutoff, lty = 2, col = \"red\")\n",
    "dev.off()\n",
    "    \n",
    "    \n",
    "}"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---\n",
    "### 4.4 Use cutoffs to classify cells in Hovestadt dataset\n",
    "---\n",
    "\n",
    "Canonical_Wnt_Pathway = 0.22  \n",
    "common.Wnt = 0.129  \n",
    "Group3_Subgroup = 0.402  \n",
    "Group4_Subgroup = 0.309  \n",
    "KEGG_Wnt_pathway = 0.139  \n",
    "Shh_Subgroup = 0.258  \n",
    "TGP_GSVA = 0.575  \n",
    "Wnt_Subgroup = 0.304  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "suppressMessages(library(\"Seurat\", \n",
    "                         lib =\"/mnt/work1/users/pughlab/projects/BTSCs_scRNAseq/Manuscript/lib/\" ))\n",
    "\n",
    "#### load in Hovestadt seurat object\n",
    "Hovestadt <- readRDS(\"Hovestadt_Seurat_GSVA.rds\")\n",
    "colnames(Hovestadt@meta.data)[51:58] <- paste0(colnames(Hovestadt@meta.data)[51:58],\n",
    "                                               \"_GSVA\"\n",
    "                                              )\n",
    "\n",
    "### Add cutoffs to meta.data\n",
    "Hovestadt@meta.data$Canonical_Wnt_pathway_Cutoff  <- 0.22\n",
    "Hovestadt@meta.data$common.Wnt_Cutoff <- 0.129\n",
    "Hovestadt@meta.data$Group3_Subgroup_Cutoff <- 0.402\n",
    "Hovestadt@meta.data$Group4_Subgroup_Cutoff <- 0.309\n",
    "\n",
    "Hovestadt@meta.data$KEGG_Wnt_pathway_Cutoff <- 0.139\n",
    "Hovestadt@meta.data$Shh_Subgroup_Cutoff <- 0.258\n",
    "Hovestadt@meta.data$TGP_Cutoff <- 0.575\n",
    "Hovestadt@meta.data$Wnt_Subgroup_Cutoff <- 0.304\n",
    "\n",
    "### Classify data\n",
    "### 0 = does not pass threshold\n",
    "### 1 = passes cutoff\n",
    "\n",
    "Hovestadt@meta.data$Canonical_Wnt_pathway_Classification <- Hovestadt@meta.data$Canonical_Wnt_pathway_GSVA > Hovestadt@meta.data$Canonical_Wnt_pathway_Cutoff\n",
    "Hovestadt@meta.data$common.Wnt_Classification <- Hovestadt@meta.data$common.WNT_GSVA > Hovestadt@meta.data$common.Wnt_Cutoff\n",
    "Hovestadt@meta.data$Group3_Subgroup_Classification <- Hovestadt@meta.data$Group3_Subgroup_GSVA > Hovestadt@meta.data$Group3_Subgroup_Cutoff\n",
    "Hovestadt@meta.data$Group4_Subgroup_Classification <- Hovestadt@meta.data$Group4_Subgroup_GSVA > Hovestadt@meta.data$Group4_Subgroup_Cutoff\n",
    "Hovestadt@meta.data$KEGG_Wnt_pathway_Classification <- Hovestadt@meta.data$KEGG_Wnt_pathway_GSVA > Hovestadt@meta.data$KEGG_Wnt_pathway_Cutoff\n",
    "Hovestadt@meta.data$Shh_Subgroup_Classification <- Hovestadt@meta.data$Shh_Subgroup_GSVA > Hovestadt@meta.data$Shh_Subgroup_Cutoff\n",
    "Hovestadt@meta.data$Wnt_Subgroup_Classification <- Hovestadt@meta.data$Wnt_Subgroup_GSVA > Hovestadt@meta.data$Wnt_Subgroup_Cutoff\n",
    "\n",
    "### bind tSNE results \n",
    "meta <- Hovestadt@meta.data\n",
    "meta <- cbind(meta, Hovestadt@dr$tsne@cell.embeddings[ ,1:2])\n",
    "\n",
    "### save\n",
    "saveRDS(Hovestadt, file = \"Hovestadt_Seurat_GSVA.rds\")\n",
    "saveRDS(meta, file = \"Hovestadt_Seurat_GSVA_meta.rds\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---\n",
    "## 5.0 Plot publication figures\n",
    "---\n",
    "\n",
    "Branavan_Hovestadt_Patient_tSNE.pdf\n",
    "> tSNE colored by patient ID. Samples subgrouped according to methylation results from Hovestadt et al. (Figure 1). All cells are malignant using annotations from Hovestadt et al.\n",
    "\n",
    "Branavan_Hovestadt_WntSubgroup_tSNE.pdf\n",
    "> (left) tSNE colored by Wnt subgroup GSVA score    \n",
    "> (middle) tSNE colored by Wnt subgroup classfication   \n",
    "> (right) Proportion Wnt Subgroup+ cells across samples  \n",
    "\n",
    "Branavan_Hovestadt_KEGGWnt_tSNE.pdf\n",
    "> (left) tSNE colored by KEGG Wnt pathway GSVA score    \n",
    "> (middle) tSNE colored by KEGG Wnt pathway classfication   \n",
    "> (right) Proportion KEGG Wnt pathway+ cells across samples  \n",
    "\n",
    "Branavan_Hovestadt_CanonicalWnt_tSNE.pdf\n",
    "> (left) tSNE colored by Canonical Wnt Pathway GSVA score    \n",
    "> (middle) tSNE colored by Canonical Wnt Pathway classfication   \n",
    "> (right) Proportion Canonical Wnt Pathway+ cells across samples  \n",
    "\n",
    "\n",
    "\n",
    "/mnt/work1/users/pughlab/projects/SinghLab_scMedullo/Brananvan_Wnt/PublicData_Analysis_April2020/figures"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "library(ggplot2)\n",
    "library(RColorBrewer)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\n",
       " BCH1031 (Group 4)  BCH1205 (Group 3)       BCH807 (WNT)   BCH825 (Group 3) \n",
       "               273                326                307                330 \n",
       "   MUV11 (Group 3)    MUV19 (Group 4)    MUV27 (Group 4)    MUV29 (Group 3) \n",
       "               431                304                314                314 \n",
       "   MUV34 (Group 3)    MUV37 (Group 4)    MUV39 (Group 4) MUV41 (SHH-infant) \n",
       "               294                190                241                338 \n",
       "       MUV44 (WNT)        SJ129 (WNT)     SJ17 (Group 3)    SJ217 (Group 4) \n",
       "               301                313                334                316 \n",
       "SJ454 (SHH-infant)        SJ516 (WNT)  SJ577 (SHH-adult)    SJ617 (Group 3) \n",
       "               293                507                493                400 \n",
       "   SJ625 (Group 4)    SJ723 (Group 4)    SJ917 (Group 3)    SJ970 (Group 4) \n",
       "               328                 75                333                 52 \n",
       "        SJ99 (WNT) \n",
       "               338 "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<table>\n",
       "<thead><tr><th></th><th scope=col>nGene</th><th scope=col>nUMI</th><th scope=col>Sample_title</th><th scope=col>Sample_geo_accession</th><th scope=col>Sample_status</th><th scope=col>Sample_submission_date</th><th scope=col>Sample_last_update_date</th><th scope=col>Sample_type</th><th scope=col>Sample_channel_count</th><th scope=col>Sample_source_name_ch1</th><th scope=col>⋯</th><th scope=col>Group3_Subgroup_Classification</th><th scope=col>Group4_Subgroup_Classification</th><th scope=col>KEGG_Wnt_pathway_Classification</th><th scope=col>Shh_Subgroup_Classification</th><th scope=col>TGP_Classification</th><th scope=col>Wnt_Subgroup_Classification</th><th scope=col>tSNE_1</th><th scope=col>tSNE_2</th><th scope=col>Methylation_Subgroup</th><th scope=col>Sample_Label</th></tr></thead>\n",
       "<tbody>\n",
       "\t<tr><th scope=row>BCH807.P02.A01</th><td>7777                       </td><td>12042.54                   </td><td>BCH807                     </td><td>GSM3905406                 </td><td>Public on Jul 30 2019      </td><td>Jun 25 2019                </td><td>Jul 30 2019                </td><td>SRA                        </td><td>1                          </td><td>Medulloblastoma, diagnostic</td><td>⋯                          </td><td>FALSE                      </td><td>FALSE                      </td><td> TRUE                      </td><td> TRUE                      </td><td>FALSE                      </td><td> TRUE                      </td><td> 0.1324481                 </td><td>-61.22671                  </td><td>WNT                        </td><td>BCH807 (WNT)               </td></tr>\n",
       "\t<tr><th scope=row>BCH807.P02.A02</th><td>6117                       </td><td>11166.15                   </td><td>BCH807                     </td><td>GSM3905406                 </td><td>Public on Jul 30 2019      </td><td>Jun 25 2019                </td><td>Jul 30 2019                </td><td>SRA                        </td><td>1                          </td><td>Medulloblastoma, diagnostic</td><td>⋯                          </td><td>FALSE                      </td><td>FALSE                      </td><td> TRUE                      </td><td> TRUE                      </td><td> TRUE                      </td><td> TRUE                      </td><td> 2.4042539                 </td><td>-54.95675                  </td><td>WNT                        </td><td>BCH807 (WNT)               </td></tr>\n",
       "\t<tr><th scope=row>BCH807.P02.A03</th><td>5717                       </td><td>10413.35                   </td><td>BCH807                     </td><td>GSM3905406                 </td><td>Public on Jul 30 2019      </td><td>Jun 25 2019                </td><td>Jul 30 2019                </td><td>SRA                        </td><td>1                          </td><td>Medulloblastoma, diagnostic</td><td>⋯                          </td><td>FALSE                      </td><td>FALSE                      </td><td>FALSE                      </td><td>FALSE                      </td><td>FALSE                      </td><td> TRUE                      </td><td> 6.3939187                 </td><td>-53.43443                  </td><td>WNT                        </td><td>BCH807 (WNT)               </td></tr>\n",
       "\t<tr><th scope=row>BCH807.P02.A04</th><td>6813                       </td><td>11307.03                   </td><td>BCH807                     </td><td>GSM3905406                 </td><td>Public on Jul 30 2019      </td><td>Jun 25 2019                </td><td>Jul 30 2019                </td><td>SRA                        </td><td>1                          </td><td>Medulloblastoma, diagnostic</td><td>⋯                          </td><td>FALSE                      </td><td>FALSE                      </td><td> TRUE                      </td><td> TRUE                      </td><td>FALSE                      </td><td>FALSE                      </td><td>-6.2815805                 </td><td>-47.67949                  </td><td>WNT                        </td><td>BCH807 (WNT)               </td></tr>\n",
       "\t<tr><th scope=row>BCH807.P02.A05</th><td>6338                       </td><td>11475.47                   </td><td>BCH807                     </td><td>GSM3905406                 </td><td>Public on Jul 30 2019      </td><td>Jun 25 2019                </td><td>Jul 30 2019                </td><td>SRA                        </td><td>1                          </td><td>Medulloblastoma, diagnostic</td><td>⋯                          </td><td>FALSE                      </td><td>FALSE                      </td><td>FALSE                      </td><td>FALSE                      </td><td>FALSE                      </td><td> TRUE                      </td><td> 2.6779469                 </td><td>-57.59939                  </td><td>WNT                        </td><td>BCH807 (WNT)               </td></tr>\n",
       "\t<tr><th scope=row>BCH807.P02.A06</th><td>7700                       </td><td>11877.21                   </td><td>BCH807                     </td><td>GSM3905406                 </td><td>Public on Jul 30 2019      </td><td>Jun 25 2019                </td><td>Jul 30 2019                </td><td>SRA                        </td><td>1                          </td><td>Medulloblastoma, diagnostic</td><td>⋯                          </td><td>FALSE                      </td><td>FALSE                      </td><td>FALSE                      </td><td>FALSE                      </td><td>FALSE                      </td><td>FALSE                      </td><td>-7.0491397                 </td><td>-47.44551                  </td><td>WNT                        </td><td>BCH807 (WNT)               </td></tr>\n",
       "</tbody>\n",
       "</table>\n"
      ],
      "text/latex": [
       "\\begin{tabular}{r|llllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllll}\n",
       "  & nGene & nUMI & Sample\\_title & Sample\\_geo\\_accession & Sample\\_status & Sample\\_submission\\_date & Sample\\_last\\_update\\_date & Sample\\_type & Sample\\_channel\\_count & Sample\\_source\\_name\\_ch1 & ⋯ & Group3\\_Subgroup\\_Classification & Group4\\_Subgroup\\_Classification & KEGG\\_Wnt\\_pathway\\_Classification & Shh\\_Subgroup\\_Classification & TGP\\_Classification & Wnt\\_Subgroup\\_Classification & tSNE\\_1 & tSNE\\_2 & Methylation\\_Subgroup & Sample\\_Label\\\\\n",
       "\\hline\n",
       "\tBCH807.P02.A01 & 7777                        & 12042.54                    & BCH807                      & GSM3905406                  & Public on Jul 30 2019       & Jun 25 2019                 & Jul 30 2019                 & SRA                         & 1                           & Medulloblastoma, diagnostic & ⋯                           & FALSE                       & FALSE                       &  TRUE                       &  TRUE                       & FALSE                       &  TRUE                       &  0.1324481                  & -61.22671                   & WNT                         & BCH807 (WNT)               \\\\\n",
       "\tBCH807.P02.A02 & 6117                        & 11166.15                    & BCH807                      & GSM3905406                  & Public on Jul 30 2019       & Jun 25 2019                 & Jul 30 2019                 & SRA                         & 1                           & Medulloblastoma, diagnostic & ⋯                           & FALSE                       & FALSE                       &  TRUE                       &  TRUE                       &  TRUE                       &  TRUE                       &  2.4042539                  & -54.95675                   & WNT                         & BCH807 (WNT)               \\\\\n",
       "\tBCH807.P02.A03 & 5717                        & 10413.35                    & BCH807                      & GSM3905406                  & Public on Jul 30 2019       & Jun 25 2019                 & Jul 30 2019                 & SRA                         & 1                           & Medulloblastoma, diagnostic & ⋯                           & FALSE                       & FALSE                       & FALSE                       & FALSE                       & FALSE                       &  TRUE                       &  6.3939187                  & -53.43443                   & WNT                         & BCH807 (WNT)               \\\\\n",
       "\tBCH807.P02.A04 & 6813                        & 11307.03                    & BCH807                      & GSM3905406                  & Public on Jul 30 2019       & Jun 25 2019                 & Jul 30 2019                 & SRA                         & 1                           & Medulloblastoma, diagnostic & ⋯                           & FALSE                       & FALSE                       &  TRUE                       &  TRUE                       & FALSE                       & FALSE                       & -6.2815805                  & -47.67949                   & WNT                         & BCH807 (WNT)               \\\\\n",
       "\tBCH807.P02.A05 & 6338                        & 11475.47                    & BCH807                      & GSM3905406                  & Public on Jul 30 2019       & Jun 25 2019                 & Jul 30 2019                 & SRA                         & 1                           & Medulloblastoma, diagnostic & ⋯                           & FALSE                       & FALSE                       & FALSE                       & FALSE                       & FALSE                       &  TRUE                       &  2.6779469                  & -57.59939                   & WNT                         & BCH807 (WNT)               \\\\\n",
       "\tBCH807.P02.A06 & 7700                        & 11877.21                    & BCH807                      & GSM3905406                  & Public on Jul 30 2019       & Jun 25 2019                 & Jul 30 2019                 & SRA                         & 1                           & Medulloblastoma, diagnostic & ⋯                           & FALSE                       & FALSE                       & FALSE                       & FALSE                       & FALSE                       & FALSE                       & -7.0491397                  & -47.44551                   & WNT                         & BCH807 (WNT)               \\\\\n",
       "\\end{tabular}\n"
      ],
      "text/markdown": [
       "\n",
       "| <!--/--> | nGene | nUMI | Sample_title | Sample_geo_accession | Sample_status | Sample_submission_date | Sample_last_update_date | Sample_type | Sample_channel_count | Sample_source_name_ch1 | ⋯ | Group3_Subgroup_Classification | Group4_Subgroup_Classification | KEGG_Wnt_pathway_Classification | Shh_Subgroup_Classification | TGP_Classification | Wnt_Subgroup_Classification | tSNE_1 | tSNE_2 | Methylation_Subgroup | Sample_Label | \n",
       "|---|---|---|---|---|---|\n",
       "| BCH807.P02.A01 | 7777                        | 12042.54                    | BCH807                      | GSM3905406                  | Public on Jul 30 2019       | Jun 25 2019                 | Jul 30 2019                 | SRA                         | 1                           | Medulloblastoma, diagnostic | ⋯                           | FALSE                       | FALSE                       |  TRUE                       |  TRUE                       | FALSE                       |  TRUE                       |  0.1324481                  | -61.22671                   | WNT                         | BCH807 (WNT)                | \n",
       "| BCH807.P02.A02 | 6117                        | 11166.15                    | BCH807                      | GSM3905406                  | Public on Jul 30 2019       | Jun 25 2019                 | Jul 30 2019                 | SRA                         | 1                           | Medulloblastoma, diagnostic | ⋯                           | FALSE                       | FALSE                       |  TRUE                       |  TRUE                       |  TRUE                       |  TRUE                       |  2.4042539                  | -54.95675                   | WNT                         | BCH807 (WNT)                | \n",
       "| BCH807.P02.A03 | 5717                        | 10413.35                    | BCH807                      | GSM3905406                  | Public on Jul 30 2019       | Jun 25 2019                 | Jul 30 2019                 | SRA                         | 1                           | Medulloblastoma, diagnostic | ⋯                           | FALSE                       | FALSE                       | FALSE                       | FALSE                       | FALSE                       |  TRUE                       |  6.3939187                  | -53.43443                   | WNT                         | BCH807 (WNT)                | \n",
       "| BCH807.P02.A04 | 6813                        | 11307.03                    | BCH807                      | GSM3905406                  | Public on Jul 30 2019       | Jun 25 2019                 | Jul 30 2019                 | SRA                         | 1                           | Medulloblastoma, diagnostic | ⋯                           | FALSE                       | FALSE                       |  TRUE                       |  TRUE                       | FALSE                       | FALSE                       | -6.2815805                  | -47.67949                   | WNT                         | BCH807 (WNT)                | \n",
       "| BCH807.P02.A05 | 6338                        | 11475.47                    | BCH807                      | GSM3905406                  | Public on Jul 30 2019       | Jun 25 2019                 | Jul 30 2019                 | SRA                         | 1                           | Medulloblastoma, diagnostic | ⋯                           | FALSE                       | FALSE                       | FALSE                       | FALSE                       | FALSE                       |  TRUE                       |  2.6779469                  | -57.59939                   | WNT                         | BCH807 (WNT)                | \n",
       "| BCH807.P02.A06 | 7700                        | 11877.21                    | BCH807                      | GSM3905406                  | Public on Jul 30 2019       | Jun 25 2019                 | Jul 30 2019                 | SRA                         | 1                           | Medulloblastoma, diagnostic | ⋯                           | FALSE                       | FALSE                       | FALSE                       | FALSE                       | FALSE                       | FALSE                       | -7.0491397                  | -47.44551                   | WNT                         | BCH807 (WNT)                | \n",
       "\n",
       "\n"
      ],
      "text/plain": [
       "               nGene nUMI     Sample_title Sample_geo_accession\n",
       "BCH807.P02.A01 7777  12042.54 BCH807       GSM3905406          \n",
       "BCH807.P02.A02 6117  11166.15 BCH807       GSM3905406          \n",
       "BCH807.P02.A03 5717  10413.35 BCH807       GSM3905406          \n",
       "BCH807.P02.A04 6813  11307.03 BCH807       GSM3905406          \n",
       "BCH807.P02.A05 6338  11475.47 BCH807       GSM3905406          \n",
       "BCH807.P02.A06 7700  11877.21 BCH807       GSM3905406          \n",
       "               Sample_status         Sample_submission_date\n",
       "BCH807.P02.A01 Public on Jul 30 2019 Jun 25 2019           \n",
       "BCH807.P02.A02 Public on Jul 30 2019 Jun 25 2019           \n",
       "BCH807.P02.A03 Public on Jul 30 2019 Jun 25 2019           \n",
       "BCH807.P02.A04 Public on Jul 30 2019 Jun 25 2019           \n",
       "BCH807.P02.A05 Public on Jul 30 2019 Jun 25 2019           \n",
       "BCH807.P02.A06 Public on Jul 30 2019 Jun 25 2019           \n",
       "               Sample_last_update_date Sample_type Sample_channel_count\n",
       "BCH807.P02.A01 Jul 30 2019             SRA         1                   \n",
       "BCH807.P02.A02 Jul 30 2019             SRA         1                   \n",
       "BCH807.P02.A03 Jul 30 2019             SRA         1                   \n",
       "BCH807.P02.A04 Jul 30 2019             SRA         1                   \n",
       "BCH807.P02.A05 Jul 30 2019             SRA         1                   \n",
       "BCH807.P02.A06 Jul 30 2019             SRA         1                   \n",
       "               Sample_source_name_ch1      ⋯ Group3_Subgroup_Classification\n",
       "BCH807.P02.A01 Medulloblastoma, diagnostic ⋯ FALSE                         \n",
       "BCH807.P02.A02 Medulloblastoma, diagnostic ⋯ FALSE                         \n",
       "BCH807.P02.A03 Medulloblastoma, diagnostic ⋯ FALSE                         \n",
       "BCH807.P02.A04 Medulloblastoma, diagnostic ⋯ FALSE                         \n",
       "BCH807.P02.A05 Medulloblastoma, diagnostic ⋯ FALSE                         \n",
       "BCH807.P02.A06 Medulloblastoma, diagnostic ⋯ FALSE                         \n",
       "               Group4_Subgroup_Classification KEGG_Wnt_pathway_Classification\n",
       "BCH807.P02.A01 FALSE                           TRUE                          \n",
       "BCH807.P02.A02 FALSE                           TRUE                          \n",
       "BCH807.P02.A03 FALSE                          FALSE                          \n",
       "BCH807.P02.A04 FALSE                           TRUE                          \n",
       "BCH807.P02.A05 FALSE                          FALSE                          \n",
       "BCH807.P02.A06 FALSE                          FALSE                          \n",
       "               Shh_Subgroup_Classification TGP_Classification\n",
       "BCH807.P02.A01  TRUE                       FALSE             \n",
       "BCH807.P02.A02  TRUE                        TRUE             \n",
       "BCH807.P02.A03 FALSE                       FALSE             \n",
       "BCH807.P02.A04  TRUE                       FALSE             \n",
       "BCH807.P02.A05 FALSE                       FALSE             \n",
       "BCH807.P02.A06 FALSE                       FALSE             \n",
       "               Wnt_Subgroup_Classification tSNE_1     tSNE_2   \n",
       "BCH807.P02.A01  TRUE                        0.1324481 -61.22671\n",
       "BCH807.P02.A02  TRUE                        2.4042539 -54.95675\n",
       "BCH807.P02.A03  TRUE                        6.3939187 -53.43443\n",
       "BCH807.P02.A04 FALSE                       -6.2815805 -47.67949\n",
       "BCH807.P02.A05  TRUE                        2.6779469 -57.59939\n",
       "BCH807.P02.A06 FALSE                       -7.0491397 -47.44551\n",
       "               Methylation_Subgroup Sample_Label\n",
       "BCH807.P02.A01 WNT                  BCH807 (WNT)\n",
       "BCH807.P02.A02 WNT                  BCH807 (WNT)\n",
       "BCH807.P02.A03 WNT                  BCH807 (WNT)\n",
       "BCH807.P02.A04 WNT                  BCH807 (WNT)\n",
       "BCH807.P02.A05 WNT                  BCH807 (WNT)\n",
       "BCH807.P02.A06 WNT                  BCH807 (WNT)"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "### load in meta.data\n",
    "\n",
    "file <- \"~/Desktop/Samwise/projects/SinghLab_scMedullo/Brananvan_Wnt/PublicData_Analysis_April2020/GSVA/Hovestadt_Seurat_GSVA_meta.rds\"\n",
    "meta <- readRDS(file)\n",
    "\n",
    "meta$Methylation_Subgroup <- gsub(\"methylation subgroup: \", \"\", meta$Sample_characteristics_ch1.2)\n",
    "meta$Sample_Label <- paste0(meta$Sample_title, \" (\", meta$Methylation_Subgroup, \")\")\n",
    "table(meta$Sample_Label)\n",
    "\n",
    "head(meta)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "----\n",
    "### 5.1 Patient + Subgroup tSNE\n",
    "----"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\n",
       "       SJ516 (WNT)        MUV44 (WNT)       BCH807 (WNT)         SJ99 (WNT) \n",
       "               507                301                307                338 \n",
       "       SJ129 (WNT) MUV41 (SHH-infant) SJ454 (SHH-infant)  SJ577 (SHH-adult) \n",
       "               313                338                293                493 \n",
       "   MUV11 (Group 3)     SJ17 (Group 3)    SJ917 (Group 3)    SJ617 (Group 3) \n",
       "               431                334                333                400 \n",
       "   MUV29 (Group 3)  BCH1205 (Group 3)    MUV34 (Group 3)   BCH825 (Group 3) \n",
       "               314                326                294                330 \n",
       "   SJ625 (Group 4)    MUV39 (Group 4)    SJ723 (Group 4)    SJ217 (Group 4) \n",
       "               328                241                 75                316 \n",
       "   MUV27 (Group 4)  BCH1031 (Group 4)    MUV19 (Group 4)    MUV37 (Group 4) \n",
       "               314                273                304                190 \n",
       "   SJ970 (Group 4) \n",
       "                52 "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "## order the same way as Figure 1 in Hovestadt et al.,\n",
    "\n",
    "order <- c(\"SJ516 (WNT)\",\n",
    "           \"MUV44 (WNT)\",\n",
    "           \"BCH807 (WNT)\",\n",
    "           \"SJ99 (WNT)\",\n",
    "           \"SJ129 (WNT)\",\n",
    "           \"MUV41 (SHH-infant)\",\n",
    "           \"SJ454 (SHH-infant)\",\n",
    "           \"SJ577 (SHH-adult)\",\n",
    "           \"MUV11 (Group 3)\",\n",
    "           \"SJ17 (Group 3)\",\n",
    "           \"SJ917 (Group 3)\",\n",
    "           \"SJ617 (Group 3)\",\n",
    "           \"MUV29 (Group 3)\",\n",
    "           \"BCH1205 (Group 3)\",\n",
    "           \"MUV34 (Group 3)\",\n",
    "           \"BCH825 (Group 3)\",\n",
    "           \"SJ625 (Group 4)\",\n",
    "           \"MUV39 (Group 4)\",\n",
    "           \"SJ723 (Group 4)\",\n",
    "           \"SJ217 (Group 4)\",\n",
    "           \"MUV27 (Group 4)\",\n",
    "           \"BCH1031 (Group 4)\",\n",
    "           \"MUV19 (Group 4)\",\n",
    "           \"MUV37 (Group 4)\",\n",
    "           \"SJ970 (Group 4)\"\n",
    "          )\n",
    "meta$Sample_Label <- factor(meta$Sample_Label, levels = order)\n",
    "table(meta$Sample_Label)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<ol class=list-inline>\n",
       "\t<li>'#4d004b'</li>\n",
       "\t<li>'#810f7c'</li>\n",
       "\t<li>'#8c6bb1'</li>\n",
       "\t<li>'#9ebcda'</li>\n",
       "\t<li>'#6baed6'</li>\n",
       "\t<li>'#67000d'</li>\n",
       "\t<li>'#ef3b2c'</li>\n",
       "\t<li>'#fc9272'</li>\n",
       "\t<li>'#662506'</li>\n",
       "\t<li>'#993404'</li>\n",
       "\t<li>'#cc4c02'</li>\n",
       "\t<li>'#ec7014'</li>\n",
       "\t<li>'#fe9929'</li>\n",
       "\t<li>'#fec44f'</li>\n",
       "\t<li>'#fee391'</li>\n",
       "\t<li>'#fff7bc'</li>\n",
       "\t<li>'#00441b'</li>\n",
       "\t<li>'#006d2c'</li>\n",
       "\t<li>'#238b45'</li>\n",
       "\t<li>'#41ab5d'</li>\n",
       "\t<li>'#74c476'</li>\n",
       "\t<li>'#a1d99b'</li>\n",
       "\t<li>'#c7e9c0'</li>\n",
       "\t<li>'#41ae76'</li>\n",
       "\t<li>'#99d8c9'</li>\n",
       "</ol>\n"
      ],
      "text/latex": [
       "\\begin{enumerate*}\n",
       "\\item '\\#4d004b'\n",
       "\\item '\\#810f7c'\n",
       "\\item '\\#8c6bb1'\n",
       "\\item '\\#9ebcda'\n",
       "\\item '\\#6baed6'\n",
       "\\item '\\#67000d'\n",
       "\\item '\\#ef3b2c'\n",
       "\\item '\\#fc9272'\n",
       "\\item '\\#662506'\n",
       "\\item '\\#993404'\n",
       "\\item '\\#cc4c02'\n",
       "\\item '\\#ec7014'\n",
       "\\item '\\#fe9929'\n",
       "\\item '\\#fec44f'\n",
       "\\item '\\#fee391'\n",
       "\\item '\\#fff7bc'\n",
       "\\item '\\#00441b'\n",
       "\\item '\\#006d2c'\n",
       "\\item '\\#238b45'\n",
       "\\item '\\#41ab5d'\n",
       "\\item '\\#74c476'\n",
       "\\item '\\#a1d99b'\n",
       "\\item '\\#c7e9c0'\n",
       "\\item '\\#41ae76'\n",
       "\\item '\\#99d8c9'\n",
       "\\end{enumerate*}\n"
      ],
      "text/markdown": [
       "1. '#4d004b'\n",
       "2. '#810f7c'\n",
       "3. '#8c6bb1'\n",
       "4. '#9ebcda'\n",
       "5. '#6baed6'\n",
       "6. '#67000d'\n",
       "7. '#ef3b2c'\n",
       "8. '#fc9272'\n",
       "9. '#662506'\n",
       "10. '#993404'\n",
       "11. '#cc4c02'\n",
       "12. '#ec7014'\n",
       "13. '#fe9929'\n",
       "14. '#fec44f'\n",
       "15. '#fee391'\n",
       "16. '#fff7bc'\n",
       "17. '#00441b'\n",
       "18. '#006d2c'\n",
       "19. '#238b45'\n",
       "20. '#41ab5d'\n",
       "21. '#74c476'\n",
       "22. '#a1d99b'\n",
       "23. '#c7e9c0'\n",
       "24. '#41ae76'\n",
       "25. '#99d8c9'\n",
       "\n",
       "\n"
      ],
      "text/plain": [
       " [1] \"#4d004b\" \"#810f7c\" \"#8c6bb1\" \"#9ebcda\" \"#6baed6\" \"#67000d\" \"#ef3b2c\"\n",
       " [8] \"#fc9272\" \"#662506\" \"#993404\" \"#cc4c02\" \"#ec7014\" \"#fe9929\" \"#fec44f\"\n",
       "[15] \"#fee391\" \"#fff7bc\" \"#00441b\" \"#006d2c\" \"#238b45\" \"#41ab5d\" \"#74c476\"\n",
       "[22] \"#a1d99b\" \"#c7e9c0\" \"#41ae76\" \"#99d8c9\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "### make custom color paletee\n",
    "### 5 shades blue = Wnt\n",
    "### 3 shades red = SHH\n",
    "### 8 shades orange = GR3\n",
    "### 9 shades green = GR4\n",
    "\n",
    "col.pal <- c(\"#4d004b\", #Wnt\n",
    "             \"#810f7c\",\n",
    "             \"#8c6bb1\",\n",
    "             \"#9ebcda\",\n",
    "             \"#6baed6\",\n",
    "             \"#67000d\", #SHH\n",
    "             \"#ef3b2c\", \n",
    "             \"#fc9272\",\n",
    "             \"#662506\", #GR3\n",
    "             \"#993404\",\n",
    "             \"#cc4c02\",\n",
    "             \"#ec7014\",\n",
    "             \"#fe9929\",\n",
    "             \"#fec44f\",\n",
    "             \"#fee391\",\n",
    "             \"#fff7bc\",\n",
    "             \"#00441b\", #GR4\n",
    "             \"#006d2c\",\n",
    "             \"#238b45\",\n",
    "             \"#41ab5d\",\n",
    "             \"#74c476\",\n",
    "             \"#a1d99b\",\n",
    "             \"#c7e9c0\",\n",
    "             \"#41ae76\",\n",
    "             \"#99d8c9\"\n",
    "            )\n",
    "col.pal"
   ]
  },
  {
   "cell_type": "raw",
   "metadata": {},
   "source": [
    "col.pal.2 <- c(\"white\", #Wnt\n",
    "             \"white\",\n",
    "             \"white\",\n",
    "             \"white\",\n",
    "             \"white\",\n",
    "             \"white\", #SHH\n",
    "             \"white\", \n",
    "             \"white\",\n",
    "             \"white\", #GR3\n",
    "             \"white\",\n",
    "             \"white\",\n",
    "             \"white\",\n",
    "             \"white\",\n",
    "             \"white\",\n",
    "             \"white\",\n",
    "             \"white\",\n",
    "             \"white\", #GR4\n",
    "             \"orange\",\n",
    "             \"blue\",\n",
    "             \"black\",\n",
    "             \"grey\",\n",
    "             \"red\",\n",
    "             \"yellow\",\n",
    "             \"green\",\n",
    "             \"white\"\n",
    "            )\n",
    "col.pal.2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "data": {},
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA0gAAANICAIAAAByhViMAAAABmJLR0QA/wD/AP+gvaeTAAAg\nAElEQVR4nOzdeZgdRb3/8XdV99nmzL5mluwrCSEEiECQRXZcwC0sIosCKqCogCzeq3KRnxsI\nior3ilcUBUVEIRIF4mWXTUwgMSHJJGSSzL6vZ+3u+v3RmWEyM0lOIskkM9/Xw8Nz5pyq6qpO\n5jmfVHVXK2MMQgghhBDi4KdHuwNCCCGEEOLdIcFOCCGEEGKMkGAnhBBCCDFGSLATQgghhBgj\nJNgJIYQQQowREuyEEEIIIcYICXZCCCGEEGOEBDshhBBCiDFiFIKd2p13vWJNTU1BQcHwAtdc\nc82eNvXuGnysd/G4Dz744KmnnjriUXbbjXHoIBr+rrt66qmnPvTQQ/uzP0IIIQ409mh3YKhw\nOPzuVkyn0+edd15nZ+fwjzZt2rR3xzqQ1dXVXX311U888cRod2Rf8ZPNPn1iyrt7iP3QYd+3\nv/3ts84664QTTigvL9/XxxJCCHFgGoUZOzOSVCq1ePFipdRvfvObd7fijTfe+Nprr434kR/s\n/vnPfw5v9l0Z6aj47Gc/e8wxxxx99NGj3RGxvy1atOjoo4/+zGc+M9odEUIIMWoOlGvsbrjh\nhpdeeummm2762Mc+9i5WXLp06V133VVVVTX8I8/zNm/eDMyYMWPv+nwAevTRR5ctW/aVr3xl\ntDsiRsf111//+OOPP/LII6PdESGEEKNkxGmw/ezJJ58EjjnmGMdx3sWKW7ZsKSwstCzrhRde\nGD7YLVu2AMXFxXvX5xUrVnzsYx8rLi6ORCKzZ8/+whe+sG3btuEduPLKK6dOnRoMBsvKyj7+\n8Y+/8MILgwsM7tWQHtbW1n75y1+eP39+bm5uNBqdO3fuDTfc0NLSsosueZ53yCGHTJ8+fcj7\nu/2DHrHA2rVrL7roosrKymAwWFFRcckll1RXVw98+qlPfQr43ve+N6TWd7/7XeBTn/rU8KPs\n9mzsdtS7/tu7atWqiy66aMqUKeFwePr06RdeeOHq1auHd+NXv/rVwoULI5FIZWXlTTfdlEwm\nh/8pZP4LsutB7Wlrux1CJo1Mnjz5sMMO23UZIYQQY9XoB7tYLDZ9+vRgMLh27dp3sWI6nT72\n2GOBb33rW2akb8Snn34aOProo/eizw8//HAgEBjynV1UVLRp06aBMs8++2xubu6QMkqpW265\nZaDMzoLdSy+9lJ2dPTwWTJo0acuWLTvr1bPPPgt87nOfG/L+XgS7pUuXhkKhIUePRqNPPPGE\nX2DZsmXAe97zniFNLVq0CFi2bNnwvu32bOx21MM/Gqj79NNPB4PBIZ/atr106dLB3bjtttuG\nlLn44ouH/ymMeIjhdjuoPWotkyHsthHTn7lfeeWVXRcTQggxJo1+sPuP//gP4Gtf+9q7W9Ff\njjzjjDM8zzMjfSPee++9wMc//vFvfvObs2bNCoVCBQUFp59++pAoMNzmzZuj0Shw00031dXV\n9fX1vfDCC3PnzgUuueQSv0xTU1NxcTHw6U9/es2aNfF4fOvWrd/61rds2waWL1/uFxseKfzX\nRx11FHD99dfX1NSkUqnOzs7ly5cfdthhwAUXXLCzjl177bXAAw88MOT93aaBIQW2bt2ak5MD\nXHHFFdXV1clksqam5utf/7rWOjc3t6amxhiTTCbz8/OVUlu3bh2o6E+C5ufnJ5PJwe1neDYy\nHPWIwznhhBOAU089de3atYlEYtOmTX5imz179kCZt956y7ZtpdRtt93W0tLS1NT07W9/27Ks\n4Q1mkp8yHFSGrWU4hEya+tWvfgXccMMNuz2iEEKIsWeUg11DQ0MkEikpKenu7n4XKy5btkwp\nVV5e3tzc7L8z/BvxpptuArQe4SrDK6+80o+DI7ruuuuACy+8cPCbK1asACZMmOD/+NWvftX/\nyh9S91vf+hawZMmS4b0a/NqfLRucmYwxa9euBYqKinbWscWLFwPD1x+HD3BEA+Wvv/564Iwz\nzhjSjp+Vr7rqKv/Hiy66CLjzzjsHCnz/+98HLrrooiEVMzwbGY56xHDjT/UNns7s6Ojw/3wH\n3vnCF74AXH755YMr+v882Itgl+GgMmwtwyFk0tTKlSuB448/frdHFEIIMfaMcrC75pprgO9/\n//vvYsVt27YVFRVprZ955pmBN4d/Iy5ZssSfDnnooYdaWlri8fiaNWuuvfZaf8blnnvu2dmh\n/TmkZ599dhfd88usW7duyPu1tbXApEmThvdq8Gs/ok2aNOn2229fsWJFOp3exbEGlJaWAnV1\ndUPe39NgN3/+fOCpp54a0o6fsaZNm+b/+NhjjwGLFy8eKOCvfT/22GN7dzYyHHUm4SaRSPjd\nG1zSn1V9/fXXB5dcv3793gW7DAeVYWsZDiGTprZu3QpMnTp1T48ohBBiDBjNYNfR0ZGVlZWb\nm7un03W7rvje974XGHzxltmTL9ef/OQnwKGHHrqzAv7MSltb2y4a8ddqdyYcDg/v1eDXW7Zs\nOfnkkwfKRyKRE0888bbbbtu8efMuDupfoRWPx4e8v9uxDyngd76pqWlIsXg8DgSDQf/HRCKR\nk5OjlPKjZG1trVIqJycnkUjs3dnIcNQjDqe3t/fHP/7xueeee/jhh/srpAMGykQiEaCzs3Nw\nxSE3T2R4xjIfVIatZTiETJrq6+vzz95ujyiEEGLsGc1g94Mf/AD44he/+O5W3MXXbSbfi93d\n3UAoFNpZAX9KLxaL7aKR4XceDKaUGtzV4a9969evv/vuu5csWTKwXUsgEHjooYd2fdDhHdvt\nqIcU8APi8NDc2NgI5OTkDLxzwQUXAHfffbcx5oc//CE7uQQww7OR4aiHD2flypUVFRWD25w1\na9bAMxiGjKurq2tw3ba2tuENZvL3JPNBZdJahkPIpKlYLMYu/wILIYQYw0Yz2PlLfiPuD/zv\nVNzF120m34v+SlZJScnOClRWVgJr1qzZRSNTp05lpFXREbs6/PVwGzZs+NznPgdMmTJlZ2X8\n5w0M33Vlt6MeUsDv/GuvvTakmL+5zMKFCwfe8fdLO+GEE4wxxx9/PPDII48Mbz/DszHciKMe\nPhx/CXjevHl33333a6+95s/J+fOLg0tOnDiRYUux//jHP/Yu2GU+qExay3AImTRVV1fHoMs9\nhRBCjCujFuyqq6uBGTNm7J+Kw78R/SvS3njjjSEl/ZsqPvrRj+6sKX+a6itf+crgN/2t8gau\nrLriiisY6RLARx99dHDndxbs8vPzgY0bNw6u29LSsuuZmPe9733AypUrh7y/p8Hus5/9LPDJ\nT35ySLEPf/jDwM033zzwTiwWi0ajWuuVK1dqraPR6IgTmRmejQxHPXw4/lTcwI0yvoEnkQy8\nc/755zPo5g/fwHMadnFCRpThoDJsLcMhZNLU6tWrgfe+9727LiaEEGJMGrVg599BOeRbdt9V\nHP6N6H+jT5069eGHH25tbfVvnrjmmmu01lrrXWwD9vLLLyulAoHAPffc09HR0dPTs2zZMn/d\ncGCPifXr14fD4Ugk8p3vfOftt9+Ox+PV1dW33HKL/0Dbe++9d3ivBr8+99xzgcWLF7/44ot9\nfX3xeHzFihUf+chHgBNPPHFnHbv55puB+++/f7dj33WBdevW+euMN9xwg7/zSE1NjX+rbHZ2\ndm1t7eC6/j0oCxcuZMdbQQfL8GxkOGp/g5I333xz4B1/Ku7GG29sbGyMx+OrVq266aabwuGw\nX3Jgf+PnnnsOsCzrrrvuam9v7+7uvvPOO7Oysoafn+GH2OtBZdhahkPIJNg98MADwPXXX7/r\nYkIIIcakUQt2/vTPLq4YMzv5GsukYiZNtba2+ku6QyilBm/hMaJbb711eMVDDz108PVbf/rT\nn/yr9Yf4zGc+M7CXys6C3YYNGwoKCobXzc3NHT7FOMBfVRy+AceeBjtjzC9/+cvhG8HYtv3w\nww8PqTtwHdiu/1AyORsZjvqQQw4Z+Mh/58c//vGQKlrrO+64w1/fHDw0f6u/wW6//fbhwx9+\niL0eVIatZTiEIY2M2OaVV14JPP/887vouRBCiLFq1IKdf6n48AvCBhvxeyuTihk2lUgkfvaz\nny1evLi4uDgYDFZVVZ177rkZfiP+4Q9/OO6446LRaF5e3sKFC2+//fbe3t4hZdavX3/ppZeW\nl5f7T+X6wAc+MGQrkF18Z2/cuPHTn/60/6yqYDA4ZcqUyy67bMgy5XCHHXbY8IvwdhtQRizw\n97///UMf+lBhYWEgEKisrLzggguGL/IaY3p7e/1wE4lEhp+BwXZ7Nkxmo16+fPmsWbP8qayB\nN//whz8sWrQoOzu7pKTkwx/+8IsvvmiM+ctf/lJUVFReXj64+j333DN79uxQKDRnzpz//u//\nHnH4Ix5irweVYWuZDCGTYDdjxox58+btuttCCCHGKmUy3uds/0smk+Xl5e3t7aPdkYPG73//\n+/POO++ZZ5456aSTRrsvYhT87W9/O+200x599NFzzjlntPsihBBiFBzQwe7ee+997LHHHn/8\n8dHuyMHktNNOCwaD/rNcxXhz+umnB4NB+ZURQohx64AOdkceeeTvf//76dOnj3ZHDibbtm07\n7LDDnnrqqUWLFo12X8R+9corr5x11lmrV68e2AJQCCHEeHNABzuxdx588MH77rtv+fLlo90R\nsV+dcsopl1122Sc+8YnR7ogQQohRI8FOCCGEEGKMGLqlhRBCCCGEOEhJsBNCCCGEGCMk2Akh\nhBBCjBES7IQQQgghxggJdkIIIYQQY4QEOyGEEEKIMUKCnRBCCCHEGCHBTgghhBBijLD358Ga\nmppWr169P48ohBB757jjjotEIqPdCyGE2DP7NditWrXq5ptv3p9HFEKIvbN06VIJdkKIg85+\nDXY+y7ImTJiw/48rhBC71djY6LruaPdCCCH20igEu7Kysscee2z/H1cIIXbr7LPPrq+vH+1e\nCCHEXpKbJ4QQQgghxggJdkIIIYQQY4QEOyGEEEKIMUKCnRBCCCHEGCHBTgghhBBijJBgJ4QQ\nQggxRkiwE0IIIYQYIyTYCSGEEEKMERLshBBCCCHGCAl2QgghhBBjhAQ7IYQQQogxQoKdEEII\nIcQYIcFOCCGEEGKMkGAnhBBCCDFGSLATQgghhBgjJNgJIYQQQowREuyEEEIIIcYICXZCCCGE\nEGOEBDshhBBCiDFCgp0QQgghxBghwU4IIYQQYoyQYCeEEEIIMUZIsBNCCCGEGCMk2AkhhBBC\njBES7IQQQgghxggJdkIIIYQQY4QEOyGEEEKIMUKCnRBCCCHEGGGPdgeEGKP6+vjf/+WllwiF\nOOMMzj0XW37dhBBC7FvyTSPEPuB5XH89jz1GOo0xPPccDQ1cd91od0sIIcQYJ0uxQuwD69bx\n5JOk02RlUVJCMsmvfkUiMdrdEkIIMcbJjJ0Q+8CLL7JtG65LWxtak51NPE5DA1OnjnbPhBBC\njGUS7ITYB+6+G8fZ/tp16eoiGKSgYFT7JIQQYuyTpVgh3m21taxbN/TNtja+9CWMGY0OCSGE\nGC9kxk6Id0l1NX/+M93d1GzkLI9pUAt/hTgAnsevf81Xv8qsWaPcTyGEEGOXBDsh3g0vL+f+\nK3A6WK/5Ui+LDBFIwmq4ENoB8Dx+9CN+9KNR7qoQQoixS4KdEP++anIu4GsdBCw0ju3+Jc6G\nJKWas48g/zPwnf6C27aNZjeFEEKMdRLshPi3uK7b2/1TVVAejaUsnHRl8rKweTabmCLo8T/w\n2zOY9EeoBgOLF492f4UQQoxlEuyE2Hvd3d0bN1Ynk8fhHh0yjbPyf/Gk+tcziqDFZEhoquHH\nx/O9V+CfcEU2V1wx2l0WQggxlsldsULsJc/zNm7cGI/HtfaU5cS9KdVdl64NkoACUBABDW9Y\nkAvvg4fghRdGu9dCCCHGMgl2QuwpA8uSyavXrftFd3en4zjpdJ5lO1YwltRlBU6+hlR/OQcq\nAA0WHNnLb77MW2+Nbu+FEEKMYbIUK8SeSW37cX3sXw1tJ6dSBWCM8RwnaExFINCFcs7wQvfD\nZghAGgrgXEABYMEdW/jb3zjkkNEdghBCiLFKgp0Qe8B5+OG3StM96izHzQFlMEoBxvOCjlOW\nl9c3M9z8c7gbVkMlXArvH1x/okvFG6PTdSGEEOOABDshMtbZ2f7Esr6LP2hbCdfLAjAaPLCM\nMdnZOTNmbAZzGPx8Zy0omLGFZcuYaZiVDVNgyv7qvRBCiLFPgp0QGVu/PhkIGy9o212pdL5r\nbK1RymitQqHo/PnzLesVjIL+tdcRPfs6nM/8JDGLrAhMhFPgVDhTrnkVQgjxb5JgJ0TGCgrC\n7e2qL+0GIsFgazJdbLyg0iYUyp4+fbplWXAmrtrVb5WrmJriuAQBg3IgCR2wGn4Eh8PzENl/\nwxFCCDHmyAyBEBmbMaMoHM5Zs9Fpj7h9OVYqFXS6p0ydsWDBgsLCQkjBH3B3Xt0o1lqckCBs\nsEAbMP4H4MDrcBp4+2ksQgghxiKZsRMiY6lW/cW5h76wtGtFqM0+znbtkpUrs5qaOOkkrruA\n4ptgOSFnpJoaAqyezNRqAmbnB3gV7odL91H3hRBCjHkS7ITITN9W/n4evZvVRC9/sp3fvI3v\n9tHnojVP/JzL/4fibnY6X6dgMtEjiGzABb2zi/Ac+KEEOyGEEHtNlmKFyMzG/6H3bUIlRCdh\nZaNqODzBxIlUVnKxR2E3Zqd5DWwIMP0pLNA7j38o2AD/t49GIIQQYsyTYCdEZrrX4bmkWunZ\nQKKZgEdp//Vwkz2MwbN2FuzS6UhTU0Vd3QmdXfNRYPdfXDeU38Ir+2gEQgghxjxZihUiM5Fy\nnB7SngGMUUHjFbhdU6aYQCB7W2OQHiiGFkgNSW2x2JS1a29IJkuN0ZYVLyv7v2nTfrE9AZrB\nUVBBDlgQ2K/jEkIIMYZIsBMiM7mzjfEwnp/FjNaxwyvWHXm+IhV1Fla0PVNU+E8Igw2xwdmu\npuaCRKI8GOxQKp1O5zc3n1Jc/Epuztr+SKcGFe6GHJgJL8A/IAinwpz9PVIhhBAHLQl2QmTE\n1dlGZRmlNI6nQgalVbI49HpF/E+W7jWrQm7fPGvBBFQdvNyf1VzQfX3TtE5rnQICga5Uqqiv\nd2pu7lrYHhH7j2DAgzjcCGlohDRYcDL8AaKjMmohhBAHF7nGToiMpIJVKG2ZpFG2UQFlnJTO\nr4o/GHCbjUE73WbjusSaYyAJNMFdmOvhHrxYuMmY7f+C8ryAUm4g0LWTg/gPKHsb6iAJLqTg\nSTh1f41SCCHEwU1m7ITISLj9/4wX1yal3QR0Oyq7Lzg9GqtJ2iUGDdlhp6lv09Ia++zIrDfP\n08ka/yGyMG3WXV9f9V2SpTGrJ265lYHa/II3Rz6GMSgXDLgxeBWVgAVQwT/hVTh6vw5YCCHE\nQUiCnRAZiNermt+Ca5TCKDCauPIcwLxz+4PCTbW1zftj79TNuetKIQQx2BRqf3H2Pdti+c/l\nvoWVmKnjt9mJ94x8mO232a6Hq2AjxoNC+E/cJayQYCeEEGK3JNgJkYFYLelOZYyxQhjA1cYJ\neR2uzgq57Y7O1iSN0p3BQ103Um0sD0IAZEEL5sncNduyU7ZHwFErHa6xWaooHX6U/ltlv6ZY\nDQVgQwPcincspmo/DlcIIcRBSq6xEyID2dNAoYwyaeUlleegSAeK67LOcXTUMjFjdFvomLro\nOa4bKOyb5DrZvYkJjpOTAg31pJVSZR6FtilNqtqkemXnj4TthlUQgRyIQCl0oFayfD+OVggh\nxMFKZuyEyEComOKjaVyO6/TPq+n8qFpvnd8cPjHi1KV0fl9gqlIaY95X/6EX89c2BjoUaC84\nVad6CfQpAx7KtXJdD+JuEB3s3xhlh33vgooA9Pb/6ILGZPEPSEFwP49bCCHEwUVm7ITIzLRP\noWyURmmUjQ6E3aaJEycm7QkdoSP6AtNAGWOAQ7ySb26++APtixZ0zTu79bjfknUs0SRed1DH\nLZo0+Sp4ZGAG5G9veccnjIXhVHCgCdqgFabAkdTD+ft7yEIIIQ42MmMnRGa8FMECrDDGQ4fw\nkqQ6p0yqaG5uTiQSA6W01lpPnBkPfXnbnFQqx7b1jJIbvq621eOtx4lBCdyosmcQgGKo2/Hh\nE9t9DTyCT5Fy4Vi4FfIx8Gfogrz9OmohhBAHFQl2QmQm9xDsHLwk4VK8FPEuShZrO7xw4cIN\nGzZ0d3drrXNzc7u6ulIpNxCYYEyZMemsrDylrprKlx/DvAa9cBhU0QHdEAKN5w4PdoWoHxLq\nJtUH5e+87cIqOH5/DloIIcTBRZZihchM0SKmXIAO0LeVRCvRycy92XXd6urqjo6OdDqdSqW0\n1sXFxUqpZDKZSqWCweCUKVNgJXhh1AnwfqgCMOBAH55HcsSDGejL3SHV+Y7at2MUQghxkJMZ\nOyEydtg3mXAq7SsJ5pnyMxs6zJZ1rw2swxpjmpqa8vLyZsyYEYvFbNsuKioKh8Pw9+F3SGzn\nGdIQGPFfWMPum00VE4y8yyMSQggxtkiwE2JPlJ5I6YlAzebNDQ0NyeQOE27GmM7OzqysrNmz\nZ/e/lwIPNDhDm3JBQRSszA7dGqDCk1l2IYQQuyBfEkLssXQ63dzc7Hkjb0bX2NjY3t7e/1MQ\nDoPwCLdIuKpNs8aih5Gn8waVBA9yGml+6d/suRBCiLFNgp0QeyyVSnmep7VWSimlgIH/G2M8\nz1u1atXmzZv7i98MM8AGNRDvXNStNscrzlacoPgFu8t2HriG39657wYlhBBiDJBgJ8QeC4fD\nlmUZY/wkBwz833/HGFNTs3nbthUAHAHL4FqYCIUQAftX2D/XdBm0oR6+rXhx2IzeOyzQ0Gbx\nl43sZJpQCCGEQK6xE2IvWJZVVVW1detWfzXWT3KDC/g/btmyrbz8S7Z9NmyCLrgK8iEJqef5\nZZzNkwEVy4Vt8By8dxeHVOBAdgi1iwAohBBivJNgJ8TeqKioyMrKamxs9C+2M8Zorf08p7Zn\nL+M40d7eWH7+Df6P8FuwoBBmOpRBEyoHavxV2GH3VmiMjUnjb2FsDGUeXzcS7IQQQuyCLMUK\nsZds2+7o6HBd109y/jos7wQ7BSqVyuvf68R/04VW+MdiVgXRLTh9qAbIhmN3aFvBJBRo/z+D\nUahs5scgtp+HKYQQ4iAiwU6IvVRfX59Op4GBWyhMPwA8rZORSF1/8cFrte7ldH6M6UFCMXQ+\nXAWnD22+A9Lv1LIhz0LbGW+OIoQQYjySpVgh9lIikVBK+SuwWmv/ejullD9Fp3U6N3ddNLp5\npKpWEPduKjbwy0Zumc5DlUO3I1bQOygL+m32wMkQ2qeDEkIIcVCTYCfE3jDGRCKR7u7uQCCQ\nTqf9ddhAIBAO28nkVstKRqMbp0+/T+vUSLU9UJAzi1mz+BHUwwuDHjWhdvKkigB8aR8OSQgh\nxMFPgp0QeyYWi23ZsqW3t9efovM3tAPC4fD8+fPDYZ1OX2hMdzDYuvO96SyIwvsAKILl8Bf4\nGtRBV/8jKcywKlVQsS9HJoQQ4qAnwU6IPeC67oYNG3p7ewfugY1Go+FwOBqNlpeXBwIP0Xdj\nINKAMttvc/V/w7wAWkMaPLChFN4H5/e3GoBz4Ez4CvwvpEbKdiH4GAT352CFEEIcdCTYCbEH\nuru7Y7GYbdu2bQPJZNJxnDlz5iil4DG6ryXc4pessah1mWqoBHD6b3oIwGfhbDhl2K1LISiF\nCAShqz/VKbDBwMlw834bphBCiIOUBDsh9oDjOIC/9uq/8DzPdV3btkn/mkgrGlfzNcMjiqQm\nYrgooW5GqYhC5cI18F87b34WRCAABroBsCEAVXAX5O778QkhhDi4yXYnQuyBaDRqWVY6nfbz\nnOu6WVlZ/uwd5u/YBo/fGH6tiEFY0Q0/c3nsPsU1hXx0Mg/M2mXzH4bFkIJk//0TaYhBHXwd\navfLEIUQQhzEJNgJsQeysrKqqqps206n067rRiKRqVOnAtCLncaAy0uGBJRBNpQbYlHz0hKX\nDb0s38TXv86zz+68+SDcDzdDesf34/CULMUKIYTYLVmKFWIXUnCvMX9w3bRt58BEWFBZeXFB\nwfzu7m7LsgoKCrZP12GhS0h1oL13fqk8TAqCWDlwY4wzbWprWbaMk07a+RFDsIGh29oZsOAf\n0AZF+2CYQgghxggJdkLswk9qa//V0HCR5wXC4cbJk3+bn38ffDcr65msrOk7lozA+/Dq6Ok9\nPszj2aZeEXXpscl2OLFHMQkqNNvSrF49wnEch0SC7GwA3uh/BJkBHCerufnkRGJGKNRXUpII\nyn2xQgghdk6WYoUAoGst9X+hc9Wgt7yGhupt285MpQo9L9DbO3PDhqtjsQqog8+O1MQtWOeQ\nipzbpj9XrwocXEWJw5ebOdOBJHQYgOxsurtpbNxeKRbj1ls58kgWLeK881i9GjzQ/akuumbN\nLZs3X1pff1JNzYdXr25MJpP7+EQIIYQ4iMmMnRj3jMubN7HtUdwEOkTFmRxxJzoIiba2Ga4b\nDoebwPW83lSqsKPjyKysWtg+6+Y4TlNTUzweDwaDZWVlodD9hK7TF3/o5n/Wf+aPqm6xmaRN\nfjHEDY9qEhAO09nJ0UfjusyZw623snQp//3fAIEATz9NQwNPHUp4NcTBffvtyzo75xujlbKM\n0Y7Tu2nTprlz547m6RJCCHEAk2Anxr2tD1PzEEAgF6ePbY9ScDjTL4cs1y1WyvVLae2Bct2o\n4+R6XmkwiOM4a9eu7e7tfrrw6VdzX3X6nNP16Z99MJn9XAO2KeozRV7/k12TijpNJE1+muef\nISef3FxefplrrqGnB6CiAiAWY/NmXv0UJ66Ezc3NRzQ1neZ5If95ZQCYtra2vr6+aDS6/8+T\nEEKIA58EOzHutf0DN0F0EoAVoW8Lra8y/XIgJ2d6b2+742RpnXCcHKXSTU3Hbdv2Ya3DeXlr\no9FoT0/Po1WPPlj+YEzHgJWsfKss8TM8/R7DkaD6HyGRbfimw+chCHWG23A8XvwAACAASURB\nVNp5K0JxMZs3Y9vY/b+Gto3r0pgF34LrGhrO8LwQ0J/qMMZ4ntfR0SHBTgghxIgk2IlxzwqD\nwRiUwk9QVsT/ZOLE9/T1vdnbG0mnE1r3ptO5qVSRUgp0a2trd3e3o5w/l/w5pVIVqQrP83oD\n3c8cE1s9nwWVUAg22x8sFoJCcKAP5sH34aP1xCZiDJMns3o1dXUAiQRlZV2HVm3856Z4/BbH\nyR3+0Fitteu6+/P0CCGEOIhIsBPj3oRT2PqQF9uWNkHtxl0Vqums6nr11Wg0WlVVNW/eEV1d\nXel0urW1obm5XSmU0v7MWTqdbg+1x3U8bMJ+S1lEY0FTW2kWNPQ/Qsxje1gEeqETeqEMjjf8\nuo45c7jkEq65hsM7OBY8XZf/yeqWlDET++uYIZ21bTt7+82zQgghxFAS7MS4V3ZyZ+XnrU0/\ntU1fWhfWZ32w0X6PSiSSyWRHR0dJSUllZWVBQUFtbS2oAZ7nGWMqVEW2k90Qaoh6Ua10LBCP\npKzpAY8A9EIuDOxOYvo3p/OjXhg8j4ULee45vuxxuUWOiSUmbt5wivHo3+4EMFFna9DriFsV\nCatUKVVYWFhUJFvZCSGEGJkEOzG+vfSS+/LLqxYsoOQnAaclpQtdAho8z1NKGWOam5u7urpm\nzZoVCoUAY7ZPoRljAoHAoXMPvarvqjuCd7RGWpVW2Spybpo5d4IFLiTBAReyQIMNESji9SR3\nnaPe+hyV8b9++m/ZHzk/RZaiKdidmO+RNTBLZ5nkjO4fF6RWauO4KtIQ/cDWrI/73RBCCCFG\nJMFOjGM/+hE//GHzokXuEUcYz7jWBPyL2oyh/5YFrXU6na6rqysuLu7s7Eyn0/6nWutp06ZF\no9HPRz9/FEc9yZMJEsew9kNFf6UDYpAFSegEB7aABVMgl7d7uCJM7SITTOltke51h/SEs9yz\ngCyPGg/jQMAYoxQVsaXFiZc9FXBV2DY9Fb1/6rVnGDNlFE+YEEKIA5wEOzFeNTTw05/S19e0\n4wO+BvYWMcYYYxzH0Vr39fXNmzfPdd36+nrHcQKBwNSpUweWRI/hmGM4BjzMe/A0HQCkoAy+\nB0uhAWzNQo8ili2i/moq6gikPVOZs2VC9x8xZwEBN2/SqmBXWyJZ4e8cnpf6F3hpnQe4KhR2\nm3PTa3NzL96fJ0kIIcTBRYKdGK+qq+nqwhgntMOWIoP3FvFfOI7jed6//vWviRMnHnXUUa7r\n2vaIvziKjVvISQ78hIFuqPdbMbwG0HY0niaQBlAFMQvTiMIYFJHsbdOn37dx01eSyYgxxqiA\n6l+WVQZQkWhuYWHhu38qhBBCjBXySDExLm38F69/llvq+VpLouydqOTP0tEf7/zL7Pz/d3R0\nbNiwIRaL7STVQSzOH3tQUKbIh3Jogpd2LKPU3LdUOEX7ZMwE+gIOsAALFQALVDjSrXWB1hro\nCB7uqWDQa7e93pDXQrCgbP65A7lTCCGEGE6CnRh/+npYdxYXrOcjbDvlHCeSNfCJUsoPVQzK\ndv4Lz/NisVh9ff1Om62p4U7N7xUxA7AeboCtmuJicnOJRtEa+MgmdXIWbhFbK+jRZh58AQUB\nsEG1t5+ZTjuhUEgp1ZB1Vn3k/a6KonBCFYEjblPFR+/D0yKEEOLgJ0uxYvxZ9wfmN5KGBtUd\nOxSjti+FAoOusRswMIdnjGlsbIxEIuXl5ZZlDWk1XVzccsqHUytCkWe3lWx4UVcnSULAIhCg\nuxvbJt9whQlcY36Rx1NQDRPg/ZCNC/6ew1muezpgWVYwGEynVU3OJfXZH5tQoKfNPwlbnjYh\nhBBiNyTYifHHqyFk6NTgBXW3/97geyYArbXneQM1/I3rAMdx3n777ZaWlkMPPTQQCAwUSCaT\na2pr4xdfbBIJnU43b3jfvNtu0yqJ69LcjNZUFG+9u/53xzuNUeYozoezduhTIcyDb2Vnz9B6\nYzqdDgQCSinXdSsqD506dapf6JlnePllbJtTT+WII/btSRJCCHEwkqVYMf5UHkfKItsAE4JP\nap0yqIFZOn8p1o9xA9NyAzdS+B/19vbW+Q8B61dXVxeLxQL5+eHOTiuZ7J4zp+msswiFUArb\nprJy/RLrnFO9OyLWr1G3wPnQu0OfpsNyeg8rWbGiuKuLdDqVSgH5+fmTJk3yS9x5J5dfzl13\ncfvtXHABv//9Pjk3QgghDmoS7MT4U3EaPSdiKSabnFnrivJeG7z2OnD/BLCzp7J6ntfX1zf4\nnVgspkDHYkSjVjJJIBCvqMDztl9dFwrduSRYo7LyEpXFscmRdO4b8Mcdmmyl5hk++EF1xRWz\nLr987ve+N33Lljlz5sybN88Pl3V13HsvfX2UlVFZSWcn3/0uqdS7e16EEEIc9GQpVoxDinnL\neOyTFP4xeUhhZ89hgyfkGLTjyRADy7VKqWAwOPijYCplurpMY6Myxti2cd1gYyOOQzyO53VP\nmbK63FPGso1lUIFkcaeV2KgHcpkFNt/9OWvWUFyMbee9/nrexo0sXEhBgV9i7Vqvro5EQnV0\nYFnk5NDVpWpqmDXr3T43QgghDmYyYyfGpwBqCYVqS/1FaTefHRdbh/On8QZWabXWpaWl2z9r\nbeW//mvC1VcHWluTxcXJwsJkfn6kqankxRfxPJJJPK9z/vyyRLnBNdrR2k1pR3mBqneaD+Ge\nxgtrCYf96T1KS+nqYuXKgaMvW9YRiyn/qr9UirY2XJfy8n1xZoQQQhzEZMZOjFcnzKAv0LF5\ngZ/oht8J68e4wbN0A7fHlpSU5OXlAVRXc9FFvPFGbjI5OxZrOPPM5IQJ0c2bK//851BTE36b\nrutFo+9vOevNkn81hboAZdTUZPFH7AQAOfBF1JfIOp22tu2Hdxwsi9xccOH5RKJx7dpTLMt4\nnvIXh41RM2Z4OTnyDzMhhBA7kGAnxqv8uUQmq20eaQNDoxvD7pMdmNLTWr+TAu+8k40bMQbI\nX7Uqf9WqHQ7RXyV36opDE+/9j+qrl5c+3RLsnRyf+VnruJJoD8yGM8FCw4c+xE9/Sn09gQCx\nGLNns/hw+BC8nk5PT6dPiEQc21bJpFbKuK465hgXdlgOFkIIISTYiXErSOhLhcXrY7WVoI1R\nA9HNX3gdyHlDDEzmAbz5JoEAu34axPEUfvSVivg03X7GIT2XaZ0uCPTMPOqyoRdCXHcdnscj\nj5BKceSRfOMbFH4GloOxbTseV729AaXQ2oRCpqDAOfHEwE6OJ4QQYvySYCfGsyumTXuip6er\nqyvfmB1WYwc/OnZwvPMfRJGbm7v95/JyqqvJz6epaacHWQQ5TO76bVnkmbhXGSxvjda00PIR\nysp2KJaVxTe+wVe/Sl8f+fnwDLwIHoTuu29JTU04EHAdx3IcZYz1iU+kTj019K6dBiGEEGOF\nBDsxnimtzzr8cNP+6qs169f3VlV5euhVa4NTndZaax0IBHJycra/dd55vPEGXV2EwyQSIxwh\nAmVgg2XCNIWtJiIePcHU8397YklOLbWTmHQGZwTon34LBMjPB2AlOKDB+cv/ndTtuWruKjsR\nDcULdVfR3LlZIxxLCCHEuCfBTox3SinrkUdK2tud00+PlZbSf3Xd4Gvs/Bdaa9u2J0yYEIlE\ntldesgRjeOABmpqYPJnycn7zG9rbt396BNwBsyEbDoFmgwZP9W6zLjrzO/+kK0EiTHgRi37N\nr7MYktWyIARpj/SbqWBSGUsZIr2xcNzqDafS2bDL9V8hhBDjkgQ7Ma4ZY6qrq9uOPdbTWjmO\n3dXl5OYO3Abrl+l/oTdsKEok8o4+umjKlP76SnHeeZx33jstHnEEl12G6xKEO2A+dABQAkUG\nBxz9i/OTr+auy07mFcdyeqLuy8GXf8kvr+KqwR1LJE7s6joH6tvyV8Te+yKvH2XaiqxI3Oko\nILdjxjEJKNm3p0YIIcRBSIKdGNdaW1vbamtx3WB7uxsKmezs0hdfdKLR7rlz06EQ/akuFrPu\nvHP22rV5jqPCYXPuuYk77giPfMvEEUdQUUFTEzPTTDZ0Qjd0QxzKQYNmTZZO4+R67TR053m6\nc5r1r6x/DW6jra1t06aedPoS6Hi84rG+z/+E9bPdF453e7Otgo7cK389/bgLoAQehwehERbC\n1TBtf5wyIYQQBzDZB0uMa30vv+z29ATa25Xr2rGYAde2F9x664S//c0a9DyxpUsrV64sUMpk\nZzvJpHrwQfXzn9e3DWw7N9jcuZx8Mvn5WCHMoPVSDWEIQZYpshzAC0K28vBUIlm05Z0nx7qu\nu3nz5nQ6HQxmE1RLi192o33qF58O/Okj6peX8tTpU75w70xmwlL4MiyHtXA/XAqt+/BMCSGE\nOBhIsBPjmv3EE3ie8bcsUQqtgz095OQUrF0bTKcHtjVZty7XGHJynGDQKyhIxeP61Ved6urq\n+vr6oS1aFj/4AV/6EsE5VEMeFEABFAJgIM3ZaVWk2GaZxlxvW5UpbuWcF4sGGujt7U0kEq7r\nJpPJrcF4Z6ArP50bUJj5q/XJT+uq2kuYGiYMD0AHTIIqKION8OT+OWlCCCEOWBLsxDhmTOHz\nzwfb21NFRamCgkRJid3XV/ziizQ0FD//fKlSA5fZRSKu521/7XlKKaJRlUqlNm7cuHr16paW\nlh2azc/n5ptZdCzXaV4GAwY2gf9s2CDHBMwP4VgoCHqL10Tu/o+co+LzBmo3NDQ4juN5Xjqd\n1nGtjRU0genGqoQCmACn4++TUgfB/inBEKShYR+fLyGEEAc6CXZiHFMqy3Vn/ehHuWvXWrFY\nVk3N1PvuK371VUKh+IQJZsOGgWB31FHtkYjb0hLq7g60toby8pwFC1o8z/M8r7Ozc9OmTUOz\nHZBOs16xxOaDFmfBddAKaYU2wBmGpSn1Zlw99qp7avVkTj7ZrxSPxzs7O/2ZQqVUSaJkZs/s\npNXXrRx/YfgomEExAIdCuj8tdkAWzN7nZ0wIIcSBTW6eEOPbjTfmX3ll/n/+p7FtlUqhFJaV\nzM5ee+21sUmTvP7L7E4+ubm7O/DXv5YnEtakSfFPfrJ+2rRuQCkViUTi8XhTU1NJyY63qR5y\nCNFoX2VloqAgq64u8tIWXlKcZQA8cKABnQ0nR1j0I6Ztv+8hlUoZY4LBoOu6juNYmM9vvvSx\n/FWvEQNOgf9CW8wE4EvwOtSCCxE4Gc7aX2dNCCHEAUqCnRjfLr2U3Fy+/321eTOtrXgeeK0f\nXBybNsW1g8YY/3FhlqUuvLD9Yx9rcd28YLDL8xzXNVrrYDAIaK2TyeSQhr1zztmYTLZPnerZ\ntk4kSp97buqV96l7DR8wxC26NT1QHubw98LxA7XC4bDWynWdUCgcDAZTqY6J2r3Pjvn3VuQC\n5MGHAZgDj8OfoA3mwtny6yyEEEK+CcS499GP8tGP0tLCokU4W/meSc0ucbtDCqXAA/qfHhsM\nqtmzy9Pp4ubm5s7OTqWUbdue5xljsrOzh7TaEA63LF6su7ut9nY3Emk69dSc9etL/vMlDrUp\n9bAMEzwIw/sHVUqEQt8rL2+vqzsukciCsG0nq6oe1TrZ/wgzBVVwEeTCx+HTcOX+OEVCCCEO\nEhLshACgpISrrqLkZk4zVkMPCoOnAGMDxph4PK617uzsnDp1amlp6bp16zo7O5PJpFIqHA5P\nnDhxSHs9PT3YdsAYkkkrkUiWlfXMmVNy3wt8U3GVptQlESHn83DZoEp3w88mTSI7+62urokK\nk5/3Rn7hm0D/TRIGVoMGBStgG3xnf5wcIYQQBwkJdmK8M8Y0Nja2trZ6J5+Yqz5V2fdIeeUT\nWzovdtwQygwu6Xneli1benp6FixYMG3atC1btsTj8XA4PHny5HA4PKRZDV5Pj2lqUsaglAE9\nLc1zUOKQVDwSInoLV1+3Y0+eUMqDssKcZwvz0mzPln4fBu+GbIOBGPwWrpNHUAghhBggwU6M\nd7W1tbW1ta7rKkWP+VBf56R5VbdNnPhAbe25rhc1BgY9LhZob29vaGioq6tLJpPGmL6+Ptd1\n58yZo3Uckv0b1pG/Zk2b66aKiqxEwo1E7HRv/ln/pAJ6FfmaC1P86Q76VhGsRk1KOFds3VrV\n2XmFUonigjcmVlbbOo2tBsU548FSeAXWko5BHpxM26epiUiwE0II0U+CnRjXjDH19fXpdFop\npZTWmp74rO6+mZOn/C6S1VhT88l4fIrnvZPq/IS3YcMGz/OCwWAoFHJdt7Ozva3t9pKSX4KB\nBfBfMKf0mWeSTU2NZ53lBoOh5uaJvY/kn7SaOrANUZcy+EwjqQfoUp5+o3rzUV0xpa08CNU3\nnuh5yenTf7bDgyvgBngYWiGBsdAzqVpHfDO/uYNFo3LqhBBCHIAk2IlxraurK5FIeJ4HKKW0\nzrIsN5UqUsqUlT2TShVXV39Wa23M9jVZP+H55dPpNBBMp9P0xrqrKeru6pnX3l7qeX/Kzb2o\nZPXqic8/X/n448mCgnBrq7ouDWDBFAiDBRbg0mz1dk3vTVTZbpsdqYKaVDLQ1nHMFPd+y0kS\nNnigeQJ+Bw54cAyLvsE3pzMNzCv8o5HGCUwYndMnhBDiACPBToxrbW1t/iScUsoY43mW1nZW\nVp1/ZVtR0QubNl3heQzM2Pn8DYSNMU4iYbc06opIMNXYtOqIzfGLHDeiFM3NG3tOOmna88/r\nRCLS0ADwOvTCRAjjZzUADfmu0xoxRr+Rv/K1smVpnZ7TOe3kLXOctXnWxBaCnuupr1vWLy23\nCaMhn8Lv88OZzOyhJ0TwDM5wcYcNSwghxDglwU6Ma8lk0rIsf8sS/52srNpodEv/64acnA3d\n3XOG1DL9tFKp7NxoYFthZMXq1m+4JhSKdCjlplLZzQsXlk+cGNm2bXudv8Ev4IugwYUYhEBB\ngOzI5icLlv9v5eN9wQSwPD/4dtapi689nJw1fLv+gRLvfu04/t0YcARHVTGxhZY4MQurjPLi\n7Q+iEEIIIeSRYmJ8i0Qi/n4lwWDQtm3bVmVlz7N9DkyBNWvWPdnZW7RWQypqrf1HyRbWbKly\nHnfCIYeopeJKpTGW1drpuW6somKHOrfBLdAOLVANcdAQQJe0/qn89zErVpooLUuUKaWeq3xl\n/cM/4JYjyTUveSZhmOFlncmJZ3JGOeUK5eEBGqJELaz9cqqEEEIcBGTGToxr5eXl7e3t8Xgc\nUEpFowVlZT2g/FQHdk7OugULbunq+l08XhkMBt9+++1UKgUYYzQYY9pmzWqNzwhvaFOWcUzQ\nMinVGvASaW1MqKNj6PHuhePgeJgICWiBpdRXET+nO0/ZYbtZ2bmFVmGP6nmbZ2eX/oWgZylm\nqjk/UD+pYgboBKl88iZQmiam8EJkA/4QgsFgUVGRv0wshBBifJJgJ8a1cDg8f/785ubmZDIZ\niURKS0tt+0JY3n8dXBpUMGhKSqZDqeu6W7ZsCQQClmUBiXjcaO2Axoo55Xaszw6qZKpCpWPK\nTheteCm6qXr7YZTCX+pNwaXwYZirafT4I8xmwn1ka+qVk0+XZcc6yQnFgpM2/S+zHTzeq9UH\n1f+bzmFdtCusSUzWaIVnkePfNLt+/frm5mbP87TWOTk5c+fO9R90JoQQYhySYCfGu2AwWFVV\nNeiNT8Dv4a/ggIJc+AhkAZZlZUejHU1NqqvLU8rk5wNKa2OMAjcYnbRmTbylxZvs5B3yrwmf\nfEp9yPAD+B3bU51/B4aredgDg6GpjLrzKC6kIsbaHNq0sUjlpTvPezhrbt4qpnoEOc/K72Be\ngr4+ekqZoEDhQh0UQm5nZ2dnZ6dlWa7rOo7T3d1dV1c3derUUTiPQgghDgAS7MTBr6+NbS+R\nPZmqw/asoufx9NNs2EBpKWeeyTvPe30EfgdPQDXUwePwPHweLpv2979vSKVipaXpwQ+HVcrf\nci7ruecm577Ipc1kG2JQBN+Eenge6I93rgt4AevbN3oPfMIkKujNJWnI9kgrUkrlps1NP81T\nn0+hW0l72k4WKdfDyocwAY3/AFsLcoGcnJwjjjjCdd2urq633347mUx2dnb+2ydUCCHEwUqC\nnTjIVZ/L5EeY5eEqVkxjxsvkZvYkBtflqqt4/S9kxWgKc89sHniA8nIALLgQFPwVeti+U/At\nmKmRH/94/qZN3TNn1i5Z0rJ4sT8J52+VgjG5b73FF3vINWwFA+0wCU7rD3YArJnHD7/Isye5\ndZWE0xRF6FR4iqmGPI+4p7ocs+KEQOXsLoIGDToGT2uWZFEKFtuD3fY7YS3LsizLtu2SkhLP\n89avXz9kZxYhhBDjilxnLQ5mLf/JzIcJemgIGI7YxPqzMq277E8s+B0PN/G7Pv7Yzsx/cvfd\nO5b4OXRAGtKQMqapqe7Z9aefvunyy53s7KzNm5Xr0r/1CRDNqrff39IdndKbmOF5NoABA4Om\n9mqr+NR9/Okj1FeQCNOXg2dhGwz0KNDYqaCbVD0LupiSoClIp8IouAX+ADHohDZwIDJkNEqp\noqIi27aLior27lwKIYQYA2TGThzQEonEhg0benp6gKKiopkzZ/o3LmxXcMfQCvPewBgymbUK\n/4qP9qItklDpcEOau57bscR68AZ+2Lz50sbG+eaEoAeNp50W6O5GawVq+17Fxg61v/nRryd7\niqnxosGtMwvviUS2kYTX+6+uM+Yv72fLZMobaJxAMoRr0QlZLkmbtEfS0GLp/HhoYU8c2yXl\nElJgoA2ugTIIQjfcDZ8c3FF/ylBrnZeXN3HixD04v0IIIcYWmbETBy5jzBtvvNHe3p5Op9Pp\ndGNj45o1awZ9/v+wk0PrBL2h7+zMzFoC0GrTa9EcJN/jWGfHEu/8mEhMaGk5Howd6zChkBuN\nJiZMMEphPIWjSFs61ds9LeZUaZVWeN2pORvjn1lZwe9CPJ1U6ZD2L7BrLsXTBFNk96I9PEVC\nYSmCBm1oh4J05NrwVw/JOo4smGDIUijT34sOqIMk/JFhT5vwV2DnzZuX6fCFEEKMRTJjJw5c\nra2tiUTCf62U8jyvvb3dcRzbtmEpfGOEOskI0cwuMisrIqFJp1EKDEpz+JBUNA2a/VeJRKnn\nBS0r7hTlu9pWjtM/XWcCdlfAdMe9CZ6xw1a7lRUngOfo2wtf/TtWvMQN3m8WPG3uuyxQ2JSe\ns45wgo5CCi16PdrBUhRZXO1xtKKnwzrcu3hG6TQuXjuoG/6knQdpCEEI+hhpU2Lbll9nIYQY\n72TGThy4GhoaBh7e5Xnbp+L6+vqgDS4ePmsFwD2Ztp51Gjl5VIbI0UzUBIuZd8WOJS4YeBUO\nN2mdct2ICYYBtFaep1zXM8r1wmmdr7Vj6RQoNCheKXjzb8WvpDDFEAjz6rHejz6XBs5eyonP\nkS6hpgo7yELD/Yb/S/P1FGd4fDzPm1F4H1zieX9sbV1cV/eh9vaFxmh/vzowEIQZ8M09Oo1C\nCCHGD/knvjhAeZ7X19cHdAY7u+yuCckJITeklMrJyYEfQ2//VNZgZxC9JOMjfJ5ANcXPUpyG\nbLgaFu9YYAlc68fHcLiptPTZxsbTPR1SFsqYoOcZz0tjK2VbdveEotc6W6bH4pONpQ1qfaQu\nYcemKQ8IKrojvHo0QDDFfd9m2SlsSFPaw9llFNrgQB1MAssQ63LCOWvf+o/e3imeF7CsZH7+\nqjlzfqCUBg1T4XEo/fdOrRBCiDFLgp04gKRSqaamplQqFYlEcnJyHMv5+Yyfv1D0gqOc3HTu\n+dvOX8ISrTW0AaDBG5TtJsGDe3K0KNwL66AZZkH5sALl8D542j/ElCm/ikb7Wlpu7OjoM87/\nZ++846sq0v//nnPO7Te9kAIJvYmAvaysoIIVC65YUexrXfVr97fq2tsW67qLrsIqithAQRQB\nBSwoKL2EEEoK6f3WU+b3xyExJKEJWNbzfgVeN3NmnnlmSLif+8zMM4ap68I0gxs39pk61X/R\nQM27uWn2vA2DLo4OyBYeK91TqylRw/4Fk1gayS3Z5VxZnKlDHUQhE+S2O19RwQRVlpaObmrq\nrWnNLleDaQbq6g6qrDy2S5cCCEGsNdGJg4ODg4NDRxxh5/BLIRqNrly5MhaLAaZhYFlvdXvn\n4y4fSyFVSy33lb/a89VTlVMBGAIeiIHSsv8sCIsgdQ/7FDAABuy4wkS4CRaCLkR6ZubgzMys\npqZgxb/+FY9G/cXF2dOneyorWbAASJBysHJ3pGeO8rqelF0+HasUAhAVBCxOndFisghi0BU0\n0Fo2y3lAgC6IyuZQL1BUNQJC05pisS7Nzf27dJkPMQjDLfAwBAAogWboAZ49HLiDg4ODw/8m\njrBz+KVQWloajUYVj2KFopaFFGJR6qJGrRFARUWtVWu/Ub45jMPgNIvjK/g0jbh7261fEyFr\nPziVA2/Cf+ExaIDnYWqCfmfChAnU12Pf8eByEY/bOVYUKQPrNvNPBrzI31Ueh2JBBly6hfPf\naDG5EUqhR8vGOUCAF3T7gIRwueqltO1JKVXA5aqEKGjghymQBtfB7TAfTMiBv8Bx+2H4Dg4O\nDg6/Mhxh5/BLYYOx4YXeL6wKLlcs9eiaY/5Q/Idif7GhGJqlCYQudEPo5ZQDH/LJY1RXkuTD\nuISDb+JFhf13O2oUXoBq6AJAKYlP4IOKGJaF242uAygKtOSrS4IoJ6icoFIrSZKoW37IiGdp\nfNSVb3U8OqfExZAAuCUm6KBJDJGuf10rDo/Hs4TQLSvgdkfT0r6HLEgDN5TAbKiCd8ANPlgL\nt8EHkLPfJsHBwcHB4deBI+wcfhFEiT7W9bHvXd+5DMV0iem508Na2BCGkAKQSPtvL3XLWX47\nt1dTHSRYSeRp1qYx/5L9KOwKoQpSwEu0EVVHLeJPA7hpK4rENOy7XxECjwfLwjSJQxiqwEWq\nCenQ/IO5ex8Rr+XJsAoak7zyCYXRUhDyURmnq8n33tR/+Hs93m+rW4vFYj5foGvXZYFALWS1\n/LYq0AivQKTldRqUwyI4a79NgoODg4PDrwNH2Dn8IljO8vXa2qAe12Z/RgAAIABJREFU8JsB\nTFHrrl2UuijJSGp2NetCl0ICLimyxJx5dK+mOpdcFRXYxKaP+fgSdv8w7J6SACpmjMr1pDSh\nSoBzVjFMI2wRM5mp8DdoliQno2mUlTHfoAYyoRGCEIfZAgFSrh7IlLOkqZMnMA1K3TwhOM1U\nRVYPMlyYxYzI4fgZGX7/9vfdJsNWSIUmsLO9REGAAgZUQ1LbdMoODg4ODr9ZnDx2Dr8ApCz7\n5pO4GdOkJqRESk1qpmIMbBhoYQFCCgGWkLMpDtUvlYaumNsOw6qozW0DYvuefDicaDWpTbhl\n3BKTIuIWX+O9veuXdFfpBldZ3AuqipR4PIwcyUqFe6AIgHp4CSa6SUqiT5/CfmrER1IZRFAV\nAlAJlY0WG0ppqMaKscnP4sXbel61ioce4raJfHEssgtshUqIw1YwwYQ4SNDBC4e38fkb+Be8\n1ppg2cHBwcHhN4ITsXP4BfDaa5nFW/0DAo1aXVos0VRkVAn3CeWeXHP43C5zpLAUhEtaASG/\nNyNjJswKnt9c4VmT7MmKJGoCcSRH7k/nFHiKOXM5KfKdFJcIucGPEHiRb/nMv9e5TvbCGJVn\nffQfzAsv0LMn55zDu+8yU5IvqHeRlMe//8KQITQ25tw8zB0nouDdAAEimaQI0qIWSfUY9RQq\nXLeO4vFccgnDh3PttVRVISVvePi//txcAx4IwMY27pmgwnjIbyl5DF6CJlDhafg3HLQ/58fB\nwcHB4ReEE7Fz+HlpjLz71/L33wq4e5y//sxA3FeSVLUluSKqxnKF7HHAvelKvKul9ZT0FVaC\nxNTlwXO8l7yf5G8y66NbpREfzvDruX4/O5nNP/p9uFmcKuRqlTjEIQZVqvxHuoWhoEky3dTU\n0LMnwJQp3Hcf3XoRzyOrHyNG0NyM38/y5UOX8ruvlFCALblsSsVbz2V/Qbtf41XBX2GMoLE7\n8TivvcYDD1BZSW4u+fm43VR/TzwMmY0UT4F/wmc/ZPDzwk0trq6AlyECOZAKm5xrKhwcHBx+\nUzgRO4efC0vKvy7/3lufOkRePxRLjqwdvLDrl7XBRr/p0qT2ReJGLZLnFVS66lONBEMNN2j1\nA9erPWq73P+qOPfTpvXJVZkXXHn4qP+n7I/PJ42NvP02xcVGj24Tz4089uZ3m9ItQ0GAKtAg\nBm7BFr8Z1qR/vZfNJsN7bmurKNx7L7ffzqOPRj/5pHHrVjF9etJ//uMeNUqzlGdvlP++Wvvy\nCDPQLE+dwbnvuREa0SgSsOgfIzWV6moKCvB6UVWApCSqazDEJnfZRRgbQYcAnAl/BYUgpLT4\nvaqGpkqS8hABvC1nZsPg3/dT5ODg4ODwy8MRdg4/F1O+/z65oaG3/Y1Q2OqrKErYmqgnJupB\nVUQbfdGv1NjYstEfZH1crzULRG510qPXNqpVtaSlDSjyDKj2Mqr3fok6V1Uxdixr12IY/7w6\n/kA0Wpuh2+ExCQYICYKYIEviVy3CUUzYuJGHH+bGG0lIACgtrdiwYdN11xkJCUJKT01Nn1mz\ngt1yQ0bxRa9x25OqsCSWhSZx2aYlilKYXv/q6JrSLDOvJnDpJLW7lAhBKMSSICHPX32bCgQZ\n4IEaeB9OhRNIsL02MB7kw9epaaA6Ge0+ci4iDgng3fdT5ODg4ODwi8QRdg4/C1Zl5SeNjedL\nKe3Ub5aUja5GUxh+ol4lIpXmBlwRT83L+ZMVqSQbwUNrDr25uM/Ry/5OpAQhaGoiK4sj98vu\nutBr/3r+uMWzHrMMt7IpOxZ2G4qFsDA0pECCLhDgl+KCqJ+qGFkmg3QWfsf33/PSSzz/PKec\nEl+7dvOJJ+qJiZ5QyIJoZub8MYMnPlq4zFWKaR600vXo5AN7LiilvBzTREpgQy/GPr21pCtS\nIGTjR8eIdy5c063CjddLRi+uLf/udTQ3XqiCeojAwzCA/rkA/If//I1pv2fk7zhGR3+Vucms\nPo1znR0XDg4ODr8dHGHn8LMwpbnZDaoQQggJEkS3ULcEI1AfqA4KzxaUZuISGlUDqNeaS3Jm\nebWKowfBVwaVlQwYwL33btvTtk/ZyPeXjntiSUqzIoU3qtT6LGEBaCamaksuFIFP4kF9yqPP\nSLUeCotBSRaqimlSUsKddzJwYCglxYhGXc3NwjRV0KPNjw1/f11WcdDMJB6f1y12y6ld3t0w\nSbn0ctasQQiysv47dnVJV7K34tKJu9mcJ984O377MwqHHEJtLZ83ddngXdc/UiGoAhMkLEFc\nw+YpRHz4JjLxeq6/mquDBIELufAlnj1tv29AdHBwcHD4BeF8lHf4WZilqhFF0UFKiZQChN/0\nX7thRIZQtxJroCWdSQuJIjGWmXff1HFXLzjgP5dK4/JLGD16n7u1lG9HM2JhelNYJaJJw2+6\nLUxVCikMDSFRLPxh+oXwSRQh4lhf+K0bglbTJoGmbdsSV1jIRRepa9YITZOArhOPlyRWbUoo\nTayKp9WQ5slOcAVWqkuK+lbwxRcUFrJ2LUOHbumumm6XDKYYCQluXVgKm/t6SEykoICKCoLB\nsz9LSdCVCjBBgA+RTmA5DUtYAhgY4xnvxr2ZzVvYkkDChVwut11G6+Dg4ODwm8ARdg4/C5GU\nlO88nhohTAAESL8Wu27gP6fxrwsZKtrcpAoMYcgMZvxDPHNd9hO3H/Xuxgcv/bf/tf3h1tPc\nX0azinBJFEmzQlCgWsKluKUQAtLrxJ+f9Mla1SPVTEgTpJmiKMrSNJASw0CaJEdY8lXwT38K\nbNlipKTEU1JiaWlxl6XEo2okTG05sdUKlZKqGJfBBIJBpk835s3z9TxOev2h9IRoRkZDXqZi\niR6lbpKSiMUA+jaff2LdbaZwS1RIQsknJUCyTqySSrCG0NOHr5FGQCKbaAoShK/2x0Q5ODg4\nOPwycYSdw8/CkYmJG3v2nJCQsM7jqfJ4yrt3n3nEsENdLnceF64mKH/I5QFwH/f1pnc99XVG\nZbJMuiThhoXDxA5t7wkW1mQmn87pueQGCPyXWREsBWEKoSAsiKj0MHP/Ik96oNHzUp0oqOXy\n5LipeFUlC7oiuqpNyVZECXtMYjFGSz6XfA6LpHJVuM9TT6QXFqo5Oe5o5LCqtXn+UF03q7mP\n2ezV66WVg2sDtW9xX8FLZ3HzzVuPOeb45rOzm9LqXXUV/qqahGhebcq4fzazZQsuF6efzjkR\nJS16bUgeY5IoyUF6aKhla4C6AyiHp89mZZy4v+UArB+/QkywdJ9MlIODg4PDrwJnj53Dz8If\nYVFGxkfp6V/qerLLFRRiImQAk5i0hBWijbBLIKE//cOEI0RUVdTpVQmutKyMA6NEvXt93nMS\nkx7ggWKKjZYruUxQwIWICykhKNx3eC68rP4ljDhhlUQ4V/av17/MbPa6slTUqtRQTn23IXqQ\nozbxWIg0aIYukpvw1ZT2v/9+a9E7ouwGcXz2ZO75SiYv1lZ8wJt5ptGsiusImTQmjvnoppUM\n9+dlxro8OevKmX0XVQTrMmT+hVM8XUongEFdHXfeif4xwXWVyEMUsULIEqSKlYR2BVo/JkD5\niVQvZEYvzulOPgidcDrvQde9nCIHBwcHh18RjrBz+FnwwxuwWIgVbncmnAhuoIGGf/BwPfVt\nw3FhwlGiySQDlio1020qZkogd+9VHTCZybXUGttftGohFfBBDu6XeGQYUdRGDMiy7AqPFHBd\nMG1jt4iFlUPO/cn3Z31yOqsuJ2MSxRYm1EI+nCjRG5TMSxjdH+1feTKzm1TOkrFbxKibxJXz\niKShaqZZ5ZP/uEn0mlzig4xI2vjvT0JVo2lpOVumYZ8ZjkSYOpUbMtZQcAkUI2OgwEB4BH6H\nDmvt0xTHcH8jBQ0c48fownzBWnhr72fJwcHBweHXgiPsHH4uFDh8+xtOeYzxG9kskS6wWoJ2\nJub7vH81V+eSa2D43L4NbDidM/feAwurgop2qs7Gi+cajjqH6wdzJjyE18Jnbdv4p8oDepoz\n3r980Q1HxokPZWgWWSgwqCfNbszoNhMS0gX3SLQa1CegC5TEhdkgkuHouPL7CDM9oJlK5lZZ\nnSxLwl8e0HByLCNDGIbUNHddXeYXX2wzZRisXw95T/HFRqwshBuqYQsytq0nvWW24ol8lUga\nJIAbXoTsvZ8oBwcHB4dfC46wc/ilYGF9Ys13C5eBaQrLBfGW9dineMrAGMMYL94iivrTP5PM\nve9RQRnEoEIKOz7KpfcMat/mri78406OOkW1D3hI7EXiIIERA47juO3aiEMJdqFrGdU6PoEl\nqXYzSCccwD8Q4lKYxRCmsSvdetArjCxF5IdSDaVKMemyuqTvJ38vHT06kpvrLynJnTYtUFiI\nEEiJlGRnS/KXMcUNPenVhdwSihazZQ3WcSjghxjocD48CYnbxse+2Yno4ODg4PBrwRF2DvsL\nC6uEEg0tm2yxGwqjtGxjKClJc5Eq1XpXg6UYws7bC2HCj/DIYzzmxn0/9/+e3+8rJ2/n9uUs\nX8nKtoUCUU21RPrwrWHNrRR3U8SBtJFJCS4G5XYwdhJiHL5/k1ZDHL7xMt3N8CYSVCiC7Cgi\nigzgjxOrpkRAPZY3SYkqav818sivzeSaZcnLlm1n0p4AReG000QoIVf5y3jvn08XZ7rxRIjM\n4eM8XoNqEC33wz4ASS0bBR1V5+Dg4PCbwzkV67DvaWzk8keWBH+/qOew4l4Pvnx8/Vlb2LLz\nJtFodGtJ8eDGgabQVciIpfsMXxfUtvfASixTmkMZug9dHcKQj/n4Ai5o7UhB6UMfA6MrXd24\nffhKZdl7oa61tYcZRiIIhECTYHYwJuBBAp/y2Tnc0Y1bulCSh+iLpw7+BXUu8rPJTyDle5ZN\nZ5YAgeJWfCN8Jz//9bjUWttGZ2rM7SY1lWvuveO9c/8gz9ekL2rFs8m7lGtPYCZMhSHQF2ZD\nFgjQcNLXOTg4OPwmcSJ2Dvueq+5ZM+XtXKQAzMJen23JvuS58+d5vthJk3A4bBiNl5efXOct\nX+PfrCvxzFjauUrCU0phTMakAJCQXS5OrOvGwH3pbQ45r/P6K7zyOq8303wqp97ETYtYtJnN\nDTRI5HHiuHJ/r6mhPoct9/Xu/XRi4joIQlPn5tQDOe8NTqimooK8PBI2wzCYDVeVcPY3pC9l\n2Wu8HCVsQSYpS1iS7E4m603ExG2rrq2kw+/Br7DezeuvM2XKUcOejUW8W9ic6cl3CZciFIEG\nPeAhCEDKvpwXBwcHB4dfIY6wc9jHbN4an/qJhjtORhVAdbr8bPi3mx8v61uWQ86OWmmaJUQ8\nKZb+8IbrN/i21FlKXyO1YuhSl1FkSWEpAJbAwnio6LIxA/8zcN+KO3DjvpRL7ddHcMR85jfR\npKH9h/8MY5hP+GKZsaLUojXfPTig/z2JiY3Qb2fm0tNJTwfgABgIa2DBx3x2B8TZdlhDQyRC\nImvhSBIT8XiIRiVybX/qUujnI+1RyAUNGpuY9t+KbH312cZQPwaKXwQUyxICg5hOnZdUsYMQ\nXT319oFiBwcHB4ffAs5SrMM+5pmqyZau4Ylt+94TQ3fFatNiFTtbjQ0GU4PBasPwGvHkvPqh\nQ5sG9UwqqhUub1zpsVnJqXJ54gIoz7CeOea7sYz9an9eqHA91+eQY2Gdx3kjGGFillEWJdpf\n61/eS1u9+s/19QMgoV0rifyIj67m6nM59wmeqGXb2irxk7EAKw0C0A3yIB8SkN2pV7gInuCo\no+jbty6FSyYy+gMumMwJ7zG5P9RDlSBVMGbDn2fLN1Lmh0UoQ2QLVKmoOuYWGjdjVSNDGFGa\nTEzZJrWzhRUgUELJ/psrBwcHB4dfFI6wc9jHFPeeR3I99ckYGqZKfTJJDb7uBfn1STtppShK\n374pXbosdrvrvN6K3NxZ+fnFAzktYPkiPlw6cZcUkNKg5DYmVFDxLM/uvyEkkXQapwnEIAa5\ncddRZ2I20qii+oI+w0gsKTkOVrRr9RZv3cAN7/DOXOb+jb9dwRWxeBPXXEPf53guhsnvIR3K\noQkaIIh6LlkQgZdIKWXy5Cf/n/vjE7EUvDqlfu7zsQzAha9rUxdjQW8+EVPflf9soiZOzEI2\n0mRCOpkllBdR1UzUPuErkRIZI2ZgKCiOsHNwcHD47eAIO4d9y/pu/q/EXY+SXk1ZDqW5pNRx\nx+M3fY3Sq8/OW3o8Y/v0OeyQQxYecsis/Pw8VX1hBCeeoY5RVFd5qm4q0hulS61LTUl3417L\nWrndrWO7RV1dXWlpaUVFhWF0kruuLaOsY8/dLAZX1ntNXFIF7IO9YbNZUWKRSLaUrnZNJjCh\nkcZ88vPJTyRxCUsW/O0sXnqJ4ipuiTW+xo2SGjCgAfzwZzLGk8G2qypWyr59Pvs9qklqLXGN\nkGCjxum9eSFLlW50L7rp1/WEF6OTLomPvsE6+wM+qKfBjWs962/n9ld4RUPrQQ+BsF114XLj\nVlACBPZ0ohwcHBwcfqU4e+wc9iEm0etO08pePXt93QGr5GfDkYJjPx80YOX9hb3RdvnDJuB0\nIU5v/V6Bv/tePL1pxBtrHnu3/+rUiM+TnS2DgTi1Pem5OylUWpFSbtiwoaqqyjRNIURJScnA\ngQN9Pt8OxhE+/ssJg2tUd/BD3xEX9vfmVmsxl/DUU69U11hWptdbDYPbtogT38pWL17bK6/l\nbaB+c9kiLAuvF3hhffSziAz6lJ5CKcesRD5I7QyiN5IwHI1Ygpj9sdoFqYjmoCzL3nbmtt7F\nk8mhLkbJ4ZVjuvo3L01co6CETH2lunASc/rRz49/DWs11Du4Q6BIpImporbKOwtrAAN2f6Ic\nHBwcHH7VOBE7h33Hm49S9fmIstjfakTPAWv817yYeM2//9CrcknIq+THoPpHmBSI4zLPe/rQ\nL48MjmzM8G4O1mxmcyqpV3HV7huJx+MFBQVlZWWGYSiKAoTD4eLi4h022Pi6qP4mRcutMjZt\nXXWjrP5Kizcuk8s+K/84dUsm0NR06MKF33/99ddlZWVATU1N0dqi9Kb0kBXSLT0ajVbr1Wpc\ntY67svaww2yTX45I0C2ftz4/rOc2o8agDv1zGi+j9JMmn37hc2Vvv31I+QGm212eJQwNLFwG\neTGa4JNyZcvmc6/Y8Mc+zb0MYUSUWEI8xcJazZrv+P5Yjp3AS4dzWAMNhRQWUmgvxRoYESJx\n4prz+c3BwcHhN4PzP77DvmH1qqkfhP8WStT7Gupir2bGkxOlyI5mH7dp3CrpPvDAZzSCP9q4\nH/9EJr7ES8tZnk76OZwzjGG72TYWi61cuTIUCpmmCRiGIYQAampqdtimcS2W4dJy+pAWr69U\nS8d7s4/om3d2bpMV8lrNzSlS2paMgoKC4uLiWCwmpRyjj9nce3OxVixV6Zbuo+qO7KceW3jl\nAcFNm2oPOyw68DnTs8iQrnqqdSkVoQgpwlgNyAuswqMeSjL9sTJ/LKhmVnmqIOqJkVWOpwGR\nLxrWp+haQlcj4ZFV920IFkWE8Ur3iSW+MoF4nMfP4Zx00n343HjixC0sgZBIC8vHDkKSDg4O\nDg7/ozjCzmFvWcGKv/P3ab3eMvPCbh8x1dDRgnE01KLApmd6vpS14sFA0fi+fb1700s66Xdy\n55610ZtQtIqKqmg0qqqqLex+eKjrpmmqamdZQryZAJaBormlBmqid2Ag+9K1DWtDoWoprdaK\nUspQKCSEEEIMrh78F/0vC1IWhF3hfuF+x1Yf61JCka5dw927SyEOqztsecryKk91RIlYwtLQ\nIpYppUQQ8ouP+n3nMV2+uNtUpN8MemLS1xjzRqlJRdPlwRvSRT9hSemx3AMa+ltoJb5HBBzF\nYWMYo6BUUtmVri5c2WTbmZYFwo17j2fZwcHBweFXjiPsHPaKF3jhFm6JEcOL8JAcUiIBLCyf\n6fOaXrfprvZUr0ksyKg8tHdvy14G/SloLmLZn6n+ApSEwNGa+3zhzdB1XUoJSCmFEIqiNDY2\npqR0ltS36xlsmkykFMWFpeNJJ+8PZWVlDQ0NUrY/sWGbsixLCJHfmJ/XkGeXS2TU7weElAJO\nqDy+0dU4K3tWtadaV3S3LmKapSog0IQWI65JNTOebJhmVbC5b1l6jSgt6UrUiy/CF4duzdAK\netT1NU1NCEzF1FAU1NGclU56PfVNNOlEfCQ45yQcHBwcfuM4e+wcfjz/5b/Xc32MbSnrpKA+\nYGEpEuJKHFBQpJBxJa7r7mXLloXD4Z/CLSvGFxdRPJVwCeHi5Kq3BlbemRya77Ya7EVYj2xM\n1NdqsdJ169bZq7TtLST254iXyDkVfx5ZJ3Do86QfZS/mdqpNbVUHdJR9gBRCqCpSjCkZ88Li\nF174+m9HlA9RsCwFKQiCKu2LygQaqtBAHrg+4dwpuHQCIRLgy6E1/+j/QIO/WFFMIaRwGR7V\n/TTPXMGVQYK55Paip0KFczWsg4ODg4MTsXP4kSzgwyu4rF3OESmQQoA0hBFTYw2uhuR4cp/G\nPgLC4fDGjRsPOOCA/e5Z9SLqvsOyUF1gClNPiK/yVD6WraRvSLo2YGzKbZ7qkk1SKk1NB61N\nu7vQNA888MD2ii3tMI56tW2BqqpCCJfLFYlE9tSjVsHnklrX+vRnJo3+4JKlz2ZiKaRDuaIL\nhMtyW7ginogmxVB3+fvjEF561AkyZVRS5aku7/vKgMpjIH9mTniEesI4xtlb6FTUAAHoDgKo\np76a6m508+D5kRPo4ODg4PCrxRF2Dj+GSh4+3/pLXOkkG5wqRYKRpEglrIZTY6kXbbkoJ5qD\nlEKI5uZmwzC0Xec92TvqVyBNFIEVZ9t+OKloXp9ZOaDhry6jUpExkCDSIvOOKvky6soyYme7\nD7oXV+JOrKamptbU1JimqWla2zR4QgjZgh232xEC3LW1eW+95astvvGUoUn+lS8lxkLQRVgx\n6Yoo0QpfudviEFOO+X3ka+3E4eIIMxBdxKyy2EpTg+RNfZM3QMa/GHEmZwYImJgmJiAQBljE\nHuOxKUyJE08j7Q7uGMOYfTKjDg4ODg6/Fhxh57DHLNXfGSUer9L0jo8Ui5M3Druw8poKb0VY\nDfds7pkaTwUQwrIsl8v1U2yzS+wLYBpt89xpGHhTtHBxy02t9slRFKL++GaKnia6gd+/s5PN\nCWlpabqul5WV6brudrs9Ho+iKFLKxsZG0zRbV2M7xev1eiKR1LffTvzyyw3XXBNLTRVV1rBo\n9Wl9n29IWp0NfqG/o+klkl4xzgwxN/2+h8X4AEGEPI9L7/fc2cysofihDhIyyEwjTSBa85gI\nRBVVF3FRAQUS6cO3kY33cE8veg1hyD6ZVAcHBweHXwWOsHPYM9bVrTs+eGWt1tTxkZDc+Kxy\nXuD8cK+0tFhaa7ki40G90GOawYzf/RTCLv1okD+oOgkC4rWoCtJqWTr+YQ1ZClVYkvJPaVhD\n0s5WirOysrKystoGHSsrK0OhkMvlikajlmWpqmpZVqcNuz/4YOzzz7998kk9IUFIqRiGFfGL\nDdccNPRWRYkBl9tVddYl9h0sLvTgKmULKDnk3ihu/9T87Ai1ATQ44wxOKaCAFnFqJyJOJXUI\nQwooyCYbSCChlNLP+dwRdg4ODg6/KRxh57AHGIbxdPjp+uQGgUAixQ8b7NyW8uwdgasm+dZe\nWhjp0UO2CDifUTKg4Qm/XqwqiFhXEh4n59Qf2b3eSP0yzDDebJIPZIc3T0jaZCTZVkvqxCMo\nbswo2+8LFPYdDVaMSOnOhZ1N26VkW8/Zq7FCCPtai7aW7fLi4uJ4//41I0caSUmKZSGlpbhk\nVImpGeFQt2BCoe31fMm/E0mizz34m6h3I03MCM1ZMuOy6jQyAsuV4f+koICZF3OxhaW0iS+6\ncd/FXYdy6L3c68Fjb79rPdfi4ODg4PAbwTkV67AHhMPhYlcxIKRAIOQ2EZNoJM5+dcRV72Vi\nWd3mzPFVVYmW4wIDGx5PjK92ySbFbBSNq/lyHBv+Te13e9ZxwdO8k87bKXx6LJ+dxpxRfHUp\nRnPnlfXGbdJNgGyJ2AlB97H4u3dSX6hYFoqHhH575hV4PB4gHo9vs9RG1dm77gDLsgzDKDv4\n4Gh6OkJIRRFSCsuULtWSqqLEASSLdP4o+ViwTmyNEdfwSVDBi8+IR7LOVDYUPDiehW/z7ipW\nvcEbDTTEiLVeMhEnHib8O36XTfZ61hdQkEDCERyxpyNycHBwcPhV4wg7hz1AVdU+4T722p/Y\ntgaIKtU7Vl9xzNYRvP46f/5z4KqrhhxySFZOjsvlCij1CfF1QlhggcSSxOtY9Ec+Hc6XFyPN\nXfQH1HzFO+ksvolYDdJCSiyDeDVlH1DwXOdNfNkIF7RKOpDgzeLwf9HzQhQXqNtF+6wYikq3\nMQTy93RC0tLS3G53p8uvtsizcxcD2IvQUkohLE2TqopQvFT41FJpKJUVw1+sO7LcCmREMhvM\ntUv4wk8gg7wudDeIRz56V60w30uaX0ppDjkZZKxi1QpWhAlbWCamRDbR1ESTB083uhkYIUK5\n5O7+/RwODg4ODv8bOEuxDnuA3+8ftPjixEnZDTIsj/qKkbNdUrt22ZgbHy1WnruDzEyOOALw\nQn/QdV3f8qEo7XhyVmKE2PIWXY6l1+U7669pPZ8ehxndrlCAtLDiVC5kYKfNBDmjKJ2xXRN/\nHkDOKRS+hBXFlYTQaC4CSWJ/8s9l4B5eawGApmldu3Zdt25dJ04I0TatnQShaVgWIBVFIFUl\n1CfpJaGzaetFJaVnlfR5zkIzjeQmI+EZ912r3KtOkMekyJB36ZzB10ynV6/yrd/L5FCd2Frl\nbdLRb+TGh3joBE5w4w4RChNOIy1KdDObXagm1lq+nc0jo7jnR4zLwcHBweFXiiPsHHbNWta+\nxVvVVEffOW3u3cfLmkFuEZGvNHU9f/rTQ5ac2nwE//4DycntWrlcLldsA2xbF5WS+9/hlfk0\nRshPsx6/QD8pd8YuhN3KhzE77BKzxZKUqDvO02bGWyJ1NgqNF5UGAAAgAElEQVTNRWz6L+lH\nkXU8pTOJ1WDFkAZCJV6LNFrie3uM1+sRglYJ15qpuHXvHS0iz5JSKApSaprm0+p79vxHYnKz\nrhxYUTnSMIL54bwlKUtiatRjejeb+r+sJ7Ib/37VNC9/T6F7dyor+04ul3dFyhPDWEJR1OUs\nP53Tb+Kmh3goSDCBBAtrKUtX8K2KJSBO7Bv+NorBMPpHjMvBwcHB4deII+wcdsESllzCJZVU\nYqr1T1ymNDT3z09QCFZUBNV3/zjgSui548aqu1Vg3fcOj03HtABWljD2GWth//rBO18qbN7Q\n7qDDD2g+sk7YYcOmAgDsxWIFLOI1fHs9WiIH3k3yEMqmU/EZ0kSahEtY8SCKmwG379SbTojF\nSjZu/FJVkw3DZZe0Runa7bdr/VtT1ZSNGwMFSz2jt9J7Y0Tta5o+iZURSxWIrd5yRSoB3Xf4\nxvwLu8/jyAhvSm67jcLCC5Z2fUEW1mq6ghQSj3DH0TPJjBO3N9spKOmkH8Xv5zHPh+JCeInC\nfx1h5+Dg4PDbwdlj57BTmtZvXvXHPy4tunqjt9uWg2RZsh6v21RW1NjYkJwswmHWr99p8+yT\n0Hz2y0kLMCy8LrxuXBqhKM98tKs9dgl9QLQk9GiDO5U+19Hris5b6Q3ojQDYRwtMpASJ3kSk\nhG+up3Q6RgxpITyoXhQPlsGG/+zGdLSnsnJqNOrxeiu83kpbg7amKe54yZhdYhlGQ3Jy6eHD\nV275c830YV6rQhHx2VlzJ3afrAtjMAeOF5dcaV42af0jAS2f/v3Jz6eoiIDPf0PVxX4rAKmS\nHBhAcgopx3CMibnVKG5o2FLdtClIYAhDfQgXpKMdhx82/4hxOTg4ODj8SnGEncOOqfmW+WcN\nXbxM/1e0/pZS46wFZkXIavA01TQVFxeXldW53eTk7NRCYl8G3oHqk5L6MEJsW+1UFUCUNPh2\n4UC3MSianUl4G76uHDOV0ws58H4UV+et6pYhlO1WYmGbtpOARfmnVH+JlNg5WezQWrxuF85A\nKBTaunVrRUVFLGYvEMdisTpbyGmarqoWbbbWtS7CtnHB7h7T7Tbd7mhKxrrUP9VvGpocWDEz\na2ZYCV9vXv8WUx8UD93sv917cktSGJ+PlJQtQ5vPH1X7rMeKQ50gDhXU+vBkkZWuJx2wRulZ\nRJ/KhCQjcL486zbr+oNk+sN0ORhg0C7H5fBTIjqQmZk5atSozz//fI8s7LxkL93bZbUZM2Yc\nfPDB7VolJCS0/Gr8QDweT0pK2lMP9/kYd7/5fp3ejlRWVt55551Dhw5NSEjwer09evQYOXLk\nq6++uvs3a+9X9/YTBx988EcfffRze/E/iyPsHHZMwbPl5VvPe1Le84p8fpq1fFmzNCYhdWnm\n6Xp2Q0M8O3vtgAGdHAjdjkH3cvISccBd3TJ9SExLgBo3VaEobd8YduDACyhuFDeKhqIR6M4p\n35P3B9zt9/NthzQQ7s63zP2QtdgEsHQAy0RAQu9OjZmmuWXLlqVLl37xxRfffffdhg0bCgsL\nV6xY0dDQAJbXW60o0rLcQkhFiXZqof3irBCW12u63ZaqxhKT1jTcVdxwSL27Lp+8y9yXJYqE\nGqOiVtS6cLf4LBg37s5bI/M8FhYJFhLqhExBuYaxPekpDZ3sbHJy6N7DJTwHm0Pv4f/NNv89\nBhO6wc07n2OHn4XWa+gMw1iwYIFhGCNGjHj99dd/br92C8Mwbrnllssvb79Btrm5ec6cOe0K\n586d29jY+FO59itj1apVgwYNmjp16n333VdYWNjY2Pjpp58ec8wxV1xxxYEHHrhp06af28H9\nxeWXX/5///d/be9mdNiHOMLOYcc0rp3waXzVeiMatw8HuOHvcL+Un0r5FTy1du0pL7ywg5wj\nbUkcwNBH/vbvaT5/QDdFJG5JSffu3e+9996dtWpYQ+XnmFEsA8vEkuh11C7ddXfJQ3D5d7g5\nz0ayLY+xGUXquBI45NlOKxYVFRUXF9fX18fjccMwdF0XQkQikY0bN4IvK0vx+Up03ROLJYPl\ndjerpiHaLL8qitLxw7SUsvUGDgvVZaTkxtJ6Kj0CItCoV8dl1LDibetXXzVmyZBEtxQZWv87\n1Mcmidf/zAOXMvxq7q6mutlrxFKDZloyioKUSI9iJQnz9/A8THMidr9wVFXt16/fiy++KKX8\ny1/+8uOMtCZN/Gl46aWXioqKzj333I6P3n///V2W/OrYf9N77bXXVlVVvfHGG2eddVaXLl3c\nbnevXr3uu+++Bx54oKio6E9/+tP+6PSXwHnnnbd+/foJEyb83I78b+IIO4cdE8hfVhSO6kh5\nEXwG38MUWAfnCXEG/LW+fsszzzwTiUR2x9jIkSO//vrrCy+88IQTTrj11ltXrFjh9Xp31mDz\nG1gxZEsOPEz0Br4YS93yXfTkSWPIE7s44mo/DPYm6UCCvcg+uX1SFQDi8XhNTU1r/uHWQlVV\nY7GYruuadvegQYvy82dmZn7VNfervPVf5n78SfLKlaIls92OUtxtnwlFnLdqjBmJxYnjciFI\nink8/HDm11QkCclZ6uAJ4o3xXDaKE6/kqkt4TcMdJlxORRNNQnNJKTFNLAsUXIlwOnTbxVw5\n/DLIy8sDNm/etiEyFArdddddPXv29Hg8Xbp0OeOMM5Yu3faRpvVzQtsFuHaLcYZhPPbYY336\n9PF6vf369Xv88cfbhkbsygsWLBgxYkQgEEhOTr788svthb8dGWyLaZqPPvroiBEj0tPT2z3q\n0aPHBx98sN3PtpTTp0/v2bOTA1Y7cXInbnz44YdHHHGE3+9PT0+/8cYb7d/NxYsXCyH69u3b\nrouDDjpICNE6q63s5fSGQqG7777bbt69e/e77rqr7bLpTqa3I4sWLbLnrV35uHHjgHnz5u1o\nHjqWfPDBB0cccYTP50tLS7vssstqamraPm1oaLj66qvT0tICgcDRRx89c+bMjgNsbm4+77zz\nWv9b3p1h7sgl+/XixYtPOOGExMTEpKSkcePGVVVVtVZOS0sbPnz4o48+apq7kc3UYU+RPyGf\nfPLJIYccMnr06J+yU4cfT+GEc44QQoyBDVAJxVAFS6GvEELTNFVVg8HgunXr9kvvX14k33DJ\n12n/9V6eDBXvsJWly8Z1smmDnDuqk7YdvyarcrIiJyvyDbf8/s52xhobGxcuXDh37tx58+bN\nbcP8+fO/+eYb0zSllFIaUq4yjEXL5n268P33P5s9e96cOXM//XTuTpkzZ07bb+ZPe++tLx/9\nvvnbJrM6YpRKvaGdJ2fIMx6Rj1TJymVy6WK5eIVcoUu9QTaslqtXy9WFxjpLWpZlSV2Xpikt\na9/+U/ymGD169CGHHHLIIYeUlpbuD/ud/sdrC4vBgwfb39rv6/fcc09jY+NTTz3V9lGnFtqV\nnHfeeaqqTpo0KRQKvf322263e9SoUYZhtK2ck5OzYMGC5ubmK664Arj99tt34l5bZs+eDTz6\n6KMdHbjpppuAr776qrX866+/Bm6++eaOZnfHyY5dZGVlzZ49u7q6+sorrwTuv/9+KaVlWbm5\nucCaNWta69t6bujQoR0N7s306rp+zDHH+Hy+adOmhUKhKVOmaJp25JFHRqPR3ZnedvTv3x84\n++yzt27dupM537lL9muXy/XKK6+Ew+FZs2YFAoEhQ4a0uhSPxw8//HCv1/vuu++GQqF33323\nVb21tWBPi124m8PcuUtpaWmzZs0KhULTp0/3+XyDBg0Kh8Ot9R944AFgzpw5Oxm4w4/Didg5\n7JjEgSce7IPTIQm2QAWUQjacqGma2+0GDMPo2rXrfundm4VwtY+7CYiWUzaj8yb1y/lsNPNO\nYd6JNBagemn9TNl2y912hyrMbUFBK87qJ1j9ZFt7fr+/9cKIdl2lp6e3LKeqMLC8PLfJkJbL\nJRWFloOxbZHtdtptNyiBT80xBihr1vrq3/FShbYRittWOYETutENhI4BxIlLpIZmYpqYumpZ\nWAjQNBSFX9tO6t8yuq4vWbLkyiuvFEI89NBDduG0adOAW265JSEh4ZprrgFWrVq1mwbfe++9\nN9988/LLLx83bpzf7z/77LOvv/76Tz755JVXXmlb7b777jvmmGMCgcCdd95pt9pN+7awO/DA\nAzs+OuOMM9h+7dV+feaZZ/44Jzvy5JNPnnDCCWlpafZGjilTpgBCiNNPPx2YPn16a017Dm2X\n2rE30zthwoSFCxdec801p59+ut/vHzt27BVXXPH1118//fTTbavt5vS+/PLLWVlZ77zzTrdu\n3U499dSJEyfW19fvpiftuP3228ePH+/z+U488cTrr79+2bJlzz//fKvP33zzzc0333zWWWf5\n/f6zzjrrtttu62jhs88+mzt3rn38ZTeHuXPuuOOOE0880e/3jx49+rbbblu5cmXb5oMHDwbm\nzp3748brsBMcYeewY5L6jz2ue6IvDVqj5QYoipImpbQXQfr27ev3+/dL73l/wJvWfqecrc+i\nFZ3UNyMsuYnaxSCxTCKlWPoPzQVYbV63Y1s1i7V/pXZJa7Gqqnl5eW2XGKSUQoicnBx77cwm\nHo9v3bpVV1XD55NCSEWR29enjZ6THU7LAqrHCPrW57v/oUbuQ14CI+GHdwIdfSITiym2sOwl\nWi/eOHEDI5lkP/500muokZ1vKzRhI1R29sjhZ0O04Ha7hw0b5nK5ZsyYMXr0toyDDQ0NUsrU\n1FSgrKwM2P0Vq5dffhkYP358a8k555wDTJ06tW21VrGVn58PlJaW7qb9BQsWAHaoqR3Dhg1L\nS0uzZZPNtGnT0tPTf/e73/04Jzty3HHH2S+ysrJos3htD+eDDz5o2zU7EHZ7M722lGwrVS++\n+OKOnu/m9B599NEbNmx49dVXTznllPnz548fPz4zM/Okk0564YUXmpt3cBf2Dmi75fHss89u\ndRV46623AFv72owZM6ajhWeeeWbEiBH2J/bdHObOaf3Hau397bffbi0ZMGAALT9ODvsWR9g5\n7Bh3asKh95xxVKGiIESCEJqiZLhcute7zuVyuVyuLl267OIAxN6QdgRHvooW2L5UoAZI6ORN\nhYY1NG/Ek4I7BU8qCKSJkG2b7hqjiZpv2hbk5ub26tVLVdXWt+F+/frZJa11NmzYEA6HZWtY\nTkraxOdkm0UK0ZLcrrVQVVVN07p2Xzrk0CeTBkNuLprt6w/JXLawpY66mcwspjiTzHzy0khb\nydLX+G8zzRbWSlb+H/+no3fQdhZUwIMwDK6B2t2YAoefgrbrJuFw+Isvvjj55JNbn5aXlz/7\n7LPjxo076KCDDjjggD2y/O2337K98LLfQZctW9a2WusOOU3TgN1PrmEroZSUlI6PVFU99dRT\n165dW1BQAKxfv37NmjWjR49u+8uyR052JDMzs1O3hw8fnpiY+NVXX1VXVwP19fWff/55t27d\nDjrooI5G9mZ6V6xYAbTdz2dbaHev4O5Pr9/vv+SSS6ZNm1ZbWzt//vwbb7xxxYoV1113Xb9+\n/dauXbv7jvXu/cO5fntT45o1a+xvV65cCfTr16/Tyq0cf/zxra93c5g7p+3eSttU2+bJycnA\n//DJ358RR9g57IwPViR8unqClEul9EMO4PfPHTcu+bjjjhs7duyECRPsj4b7iy7Hcepq/Dkt\nokyg+cg5mW7tV3aAbddIRGpoKqCpsE2Abk+wdCrmUvdd27Ju3bodddRRgwYNOuCAA4488sjs\n7Oy2TyORSENDg6ZpmmKmxL7LiHzuN4tbRV7bNVxb5NkCUVEUj8cTDAZ9Pp+iKIrSBzQoh3oo\ngzQ4trVhJplu3BvZeAM3/IO/v8kbT/L4mYy5juuP4qiRjDyJk6YwpZlmibS2G7gOyXAyROE9\nePDHzInDT8u33357wAEH3HzzzR6P59lnn7XfYnef2tpaIDU1tTUoaL+DtttN33oue0+xlVMw\nGOz0qR3jsVdg7fXHTmNmu+lkR2yd1BG3233yySebpjljxgxg5syZhmG0jVG1spfT29DQYHve\nWmLvV4tGtzt99SOm1+VyDRs27KmnniosLDz33HPLyspuvPHGTmt23BkC2JE2m4SEhLYu2cu7\ndqGNz9dJDtG2FXZzmDt3qa1Be2GnbZrDpKQkoO2JCod9hXOlmMMO2bhx4z333FNVvV5RxsBI\ny8rwere43d8ffPCDL774wk/kRCCP09ZT8j7ln6AF6XI8uad1npo42BMzhNGyQ2Vb8Iwf/t5J\nxM5e4RVgGRS/R9ks+v+JwQ+1Ptc0reMZQBtd16WUPup61D7ojW9UMA185cHRWwJj2T5fcbtw\nna7rhmEIIVwuV0LCSbAW3oU49IY7YWBrFwkknMM5L/PyetYvZpGB1Tqmeurr2TbkBSwYzWiV\nttER+2jtGFgJL8MciIMbh18wt956a21t7XPPPXfdddf9iOaBQKChoaG+vt5+49zn2HswOn0j\nB0aNGuX1eqdNm3b77be///77fr9/1KhRP42TZ5555pQpUz744AM7AEZne/vY6+n1+XzNzc1N\nTU2toscWqZ2GMHfHWjQabTcPPp/v8ccfnzJlysKFCzttFQqFOhY2Nze3GrFdSktLa+tzc3Nz\ncsuN3rZu27ljezTMTl2Kx+OtQrypqaldc/v/w47RXIe9x4nYOXRCEUVf8dWn335aXV3tdrul\njFjWNCknRCKf1NXVVlbu7YatgoKC6667rm/fvl27dj3ssMPuv//+ioqKt99++6yzzho2bNgN\nN9ywvu1VZZqf7hdw5Ksc+hzdztrhhRONa37ITiet7Y5IiO1V3Tadp6L6QfxwrsICBNLCaGb1\nkxS9ujtj8fv9LperS82rwfg6qXh0kajJSE5oeoayMS0tre35/9alNztEIaW0LMs0zbS0tKSk\ndHgIvoBZ8Bm0jzTczd13c3cXulhtRtCOiUw06DThpwp3QT8wQd+dQTn8jHz33XfABRdcYH+7\np7vp7WMNhYWF+9wxG/u9eUc5hwOBgJ3YaPny5YsWLRo1alSnwaH94eTJJ5/scrk++eSTpqam\nWbNmpaSkDB8+vGO1vZzegQMHAkVFRa0ldsxv0KAfkzDSXh7tGDW0P0a2nl11uVy0EdOd7tjb\nuHFj62t7Kbw1A7y9DNqxwk7Y5TB3x6W2iWbsZWX7wISNLfV29IHZYW9whJ3DdsSI3cRNJ3HS\n+Zz/AA+ECauqallWa9jJMIyJEyfuTcbw8vLy8ePHv/LKKxs2bCgrK1u8ePGDDz7Yq1eviy++\n+KOPPvruu+8mTZp02WWX7XGIvmYReh1yB0cI2iLsP3Jb7jpFwZ2AUBAC0aIILZ1Vj2K1vx+p\nI5pKfzG7S/MM1WpymQ1CSNOd7iLSM7m+e/fulmW1Hp5o+8IO1Hm9Xnv3XouxFOjdMaJmYU1i\n0tu8vZ71FpaG6PT3toSSMOEdjN8N58KhEOjsqcMvCDvu0po98eOPP25XYedBjosuugh48803\nW0vKysqEEIcffvju9L7LCEqfPn3YqR4644wzLMu6+uqrLcvqNGa2O07+iEBOUlLS8OHDm5qa\n7r333sbGxtGjR3e6bruX0zt27Fi2P/k7a9as1vI9xb6947nn2qd5t22OHDnS/jYnJweoqNh2\naMw+mNyOtgOx3bMnmZbIZduDC7u8zmuXw9wdl9qeZbFPY/zhD39oLbF/hHJ2cSulw49iX+dP\n2RlOHrtfPs/IZ3JkTp7M613dO/E/iVp3zeP1tPuZURRl6tSpP7qLl19+OSMjw+12d5oBVVEU\nW/Eceuihl1566YwZM3bLaOEE+YZ/uwR1ryEnt7zoNIPdD+VCvqHJ10X7Cm+ny7qlu+563fPW\ne92syS5rsmJNVq03/XJ6f/lOltz438LCwp0ksZs3b94XX3yxYMGC1atXRyKReDzeatKQxnw5\n/y351lK5dKacOUgOckmXJjUhBRIhUSS0+cqTeUPkkIFyYK2s3YGXppRzpNy0W5P52+ZnyWPX\nlltvvRW48cYbm5qaPvzww6ysLHv7VHl5uV3BPmfw7bffdmrTMIyRI0d6vd4333wzEokUFBSM\nGTMmMzOzNd9kRwfalnQ03g47Kd20adN2ZKGiosLeYaaqak1NTaed7tLJnY9xRyV2gg9bmX34\n4YedVt7L6Y1GowcddFAwGPzoo48ikcjMmTMDgcBhhx3W+vu7O362YpqmLZWuuOKKtWvXxmKx\nsrKy559/PjExsXfv3lu2bLGr3XLLLbbPoVBo4cKF9lF92O4YVlJS0ocffhgOhz/88EOfz3fs\nsce2JNqUTU1N/fv3DwaDc+bMCYVCkydPtrdIti4jdPRwl8PcHZcSExNnzZoVDocnT56sadrh\nhx/emgZPSmnLvjvuuKPTmXHYGxxh57AdY+XYdJneb0G/wGEBV1eXSBOqu5PPryNHjvzRXTz8\n8MPJycmapu0otb1NIBBIS0vr3bv3rkVkrEZOTZKTt1dmkzu86ETSIV8XcrL91aHaW0my9vtd\nDsf69Dj9zZTolExrsmZNVq3XFevNoPzoEBnaUlBQ0KmwmzdvXtsExfPmzfv888+//vrrdevW\nxWKxBtlwjjwnX+Zny+xusluaTPNKryIVl3SxvZ5DkiATXpAvFMiCIln0jfymVO5Ii+hS6nv8\nT/Wb5GcXduFw+Oqrr05NTU1KSjrllFMWLFjw8MMP+/3+YDBoV5g9e3bfvn1t+dKpzWg0et99\n93Xv3t3lcuXk5IwfP75VInTqQNuSjsbbYcdmHnzwwZ0Mys5vMnz48J10unMndznGTkv+P3vn\nHV9Flf7/z5m5Pb33hISeQEACitISQEXqokAQUVbBVQTxZ1u/qODCoqCruOLa0EXBEAkqoewS\npYUm4FIkdAg1pBHSy+13zu+PCcNwWwIJCeW8X7x4zTxz5pznTJ5757mnPM/Fi/XRHxMSEoRr\nw3RLhZv+eMvLy6dNmxYaGqpUKtu0afPKK69UVla60arBv3haWtrAgQP9/PyUSmV0dPSgQYOW\nLFkij+VbV1c3efJkX19fT0/PlJSU3bt3S/E1pfrXr1/fq1cvjUYTFBQ0depUMaSLRElJyZNP\nPunt7S3+Zv7Pf/4DwN/f342G7rvZGJXS0tK6d++u0WgiIyNnzJghv51S+u677wLYtGmTmyfD\nuDGuSW10s9m4cePMmTPDw8PlYSQZtw4Wi2Wcbdx2st00yGQ6ZuJ0nJW3KsuUpjr7Gck2bdrI\nV2xcFz///PNLL71UVFTkNN2WhEKh6Ny5c1FRUa9evdavX++uxnPfY88zgA2Oxkyu+EBOfEji\nfK2aVJjXYUwpeCcrhGSFBet/uguGQgvvpbRVKWx1BFaTOlrTbzmCHrh8+bIY+NTVp0weDIXj\nOJ7ng4ODf+7w8yf4xBOeWmjzkV+DGjXUZpgVUJhgsptpnY3ZL+AFE0xmmP3h7wUvJRzXIFqB\nKiDAXUcYVxg5cqQY0WPt2rVsnsgRm80WExMTHx+/YcOG1tbFCT179ty/f/+KFSucprJliJSV\nlQUGBnbv3v2PP/64GfVLo3duygwePDg3N/fs2bNs/0Szw9bYMeqpra3NyclpW9CWnqKGUwZb\nnc1cYqbF1GaxAd2AecC/gKmAFwCO49y7ZW4YMWJE//795ZvznWK1WgsLC9Vq9cWLFxsIH2oz\nXlk258BVr47Un9b/T64JcScifhquJqiwoeqYeyVBOJNnJyJYeGq28Z423sPMB54LmoSgXgAC\nAgKkbWhwlnNC/sUnCALHcZWVlXvpXgrqC1811B7woKA8eAJihZWCcuBIfX8IgEEYBKASlWIi\nCoXDPncB5kr8BLh1TxmMRsPz/CuvvLJ161Zxm+QthV6vLykpueeee25sxdudSmRkJCFE/lNc\nzFErBcRueUpLS7dt2zZz5kzm1d0MmGPHqOfChQtGo3F0+ejeZ3vTKkptlFNwCk4hWB8AfgD+\nAowDZgFLOU7bt2/fG46DpVKpvvvuu1mzZjWYsqK8vLyoqAhO03DJ8UmA0seJY3c1ygkBAE4D\nXnllkwRxktPCzlMVzChYg4awtB9tVIVzglFhrRKIqthjEA1QAecAcByXkJAQFxcXHBwsBVaF\n6/RiFovFYrF4Uk/blVQfGmgICAFRQ02vbPq9oi8FwIHTQNMRHduirTe8ybW7f80wFaBwAt5Z\nhb5AAwEOGIxGMn369JiYGCmxwa1AeXm50Wh87bXXCgoKFi1a1MA3xl2GuB7uhRdeOHfunNFo\n3LZt2/Tp09u2bfvqq6+2lko//PBDXFycmEiX0ewwx44BAIIg1NXV8TyvJurBZwariIoDx1t5\nIhBKpwJBhBQCeUAt0MvPb/zMmTOb0pxKpXr11TenT9/F8+8BLwFRTotRSgVBOHbs2EMPPeQu\ncmnQA4ibDLV/fSQUyWOrH4EjAuHXH1a/k6l8b7V17xkBhII6DDeKL4JrvD2K/HWwOonPJMcn\nolthhzHn/Cef9/3zcf/XS4Ifjo7OBoJE/SsrKy0Wi4eHR1BQkHxj7DUtXzkVBMFqtfYs6+kF\nr3zkl6K0GtV+8POFrxe8/OHvBz8OnHw29jAOe8JTDTXn8FmmgA3WGlTkIG8ODpdgkfuOMBiN\nRKVS/fOf//zmm29aW5GrREZGenl5bdmy5bvvvuvbt29rq3Nr8corr6xcudJkMt17772+vr7P\nPPPMqFGj9u7de5MiHTaGJUuWfPzxx67CTTOaCHusDODKXlRxYKlNm7jIqK7z3//46NFNHDhC\nOvK84O/vX1VVBZiB0BkzFnbq5Ntwpa4xm/HnP2PHjm5KZZzNZgAmAU8CzvNwC4KwY8eOuXPn\nuss/fc8ChD+CyhyU/Q95P9eHKamfeKV/X6P6bqul1mAEsHQb5o0jY+8DQK9Zgec0iHHFYewY\ng74rwatRsgOmUvjEw/ea9OeEdOyQcPFycEVdXUyA4nBQ0B8azcOAH4CzZ88WFxeLe9OUSiV1\nyBJLr81OwfM8z/P3nr331YBXl3BL6lDXBm3+ir9GI2OLUI0AACAASURBVPoUToUgJBCBiUi0\nwUZARPfuIA4Sp3PQV+Zq07HIB6iAcAh7Brt8fAzG9TFs2LBhw4a1thZXaXxKtLuTsWPHitl4\nW4YG1+7fpLV9DBHm2DHqiWwbqdFp9NWC0eg5MGVgaEjfR1PTa/G6qvIix3WMiAiOioqqqYFe\nD1mMyRtk/Xr89hvUaurjU2WxVNls0cCLwPOuylut1uzsbLPZ7HplHkFIMkKSYSpF0UaYrsTA\nozhdwi3fo7Jw2pigCkEQCsrx4X/po/eCt3PmHL0jChAbLv+G04tRvBnVRyFYoPBA7JPoIs+Q\nyxPyUXDwP4EtgBJ4ApgGQK/X5+fnSysRTSaT3UAdz/OUUmn5IAElgoGHCrz2acPTz3s8X4GK\nQASK/lkXdAFwAAd00FlkQYbvwT2SY0dBpWMbbGaYilGwBMvNoArAE0GuHi+DwWAw7hjYVCwD\nAPKRf9n7st5gvnBOeemS4vJl0ratunfSGItxaruxmyMiNBcv4vx5VFXhnnvgLEXQ9XH6NAwG\n1Nbml5eXC4IAcEB/D4+oxx57zGlmSQBWq7WBLRQi6kD4dgXvCaUXlJ7QhZ8p9zSYBB//ECg8\nOQJPDSmvRYGrZd/y35mij2Stw4V0lO8Dp4Y6EFYDzi5FyfZrbwsA/g78BmwF/ioGAT537pwU\n2Lm+bkrFIC/S8kRKKccRQqyE2AixCiBWWsvXlGm1WgUUQQiSj8bVoGY3diugkGZdu6Lrg3hQ\npm99YSushSjQQ78POUWwlMNaCforNA0/PQaDwWDc5jDHjgEAi7CIB2+pVgIqgFgshOc5X19P\nanskeNmRtDRu0iSMHIk33sDSpWhoz0PDhIcDsJSV6S0WD0rbAb5Au7q67WfOdBEzz9hBKW3X\nrp3T3EROCOwNhQ6acHi2A6cMC/RQaXR6vR66KBCF3kx1aoR4Qxwjq5+KlbAbtyMABFSfBuGh\n9AangjYE1hqUH2hQCzGuut3Eq1ar9fT0FIfxpJhDHLGAQqC8QHmON0aVf8s5ZN0oRemf8Kd3\n8a4eegAERAnlQAz0hnc5yuWr7iioACEQQVbYfsRKsXkbuKXYcAZnGvUAGQwGg3HbwqZiGQDw\nO35/gD7QVx2r0xGLRQgKIjU19NAhAkOYH/zuvReNS0fUWIYMgVp9wWYLAHxRn7SeB6IPHnxd\nocgGTtqV53neYml0ktOOM1D2P1SfgKUCCo8uD4xOOWz873/XX8ivpYKnp9o8qb9RrRTA8aBX\nhgCllXZ2se3EY2sNqAE0FISHYAU4qLwb1EJKCEsplfZMiOkRxWE8cQCSUgpOS0EBQmCNC/s6\n1LATx44hJERe27/x76M46ge/MIQVo7gSlZ3QSRzSy0NeHer84a+FtgY1VagKRrDo+b2G14/g\n6Hmct8BSicqTONkWbRv7GBkMBoNxG8IcOwYABCP4fcv7wbn9YqIDeJ6rqaGffSYcPkw4z8oR\naP5YR/7+iI19p6BgKDAG4AABMAMKQGO1vg3Yz/VyHHfhwgWLxSJmnm4ATQj6r0HxRhgvwbsT\nF9zv4061CQlddu3apVarhw4dOu4+igPT6re7yl06OI9YDFAINtSchjoA1jpoIxCc0rAWGo3J\nZIJs0E4ejri+3vrNE5KA1BnbovB33GMfTPgETthg84BHBSqMMBph/AN/eMP7cTweilAAaqgJ\niAYaHXSXcOkyLqugikPcX/CXN/GmGAxPLMlgMBiMOxjm2DEAYDwdP3/5/JFvLu5//0teXsoD\nB3DkCAVsvR67/CSevBktenlVcNzHgvAQ4IeruwEoEA1oAYO8sNlsLi0tbZRXJ6LQIfLqWj0v\nL6+XX35ZzHFZT/ggXFyJc+ko/wMwO69E7uoJFJwFhIN/Erq8Dc/YBlUICQmpqqqSR3J23Ckm\nm6glAKXgSsvvbXviNCZ0tisp+mRncVaMRSwKt2LrV/jqOTwXjWgCUo1qK6z+8PeGdwlKTDAJ\nEDqhkwkmAtILvRLR5G0vDAaDwbi1YWvsGADwcN3DK5NXLv5s8ObNF5ctqzpyRE9IbXR09cq5\nQxwDpDULvXv39vK6CJwFOEABKAEOsAClgNGxfFVVlZjoqXnQhqLDDEAANQPOk1ZAwDVL1wQz\nrHUAB5VfY1rQ6XQajUbKn+hYQJqrlQQAMVqi9FP/DQcXdizGesJT7tWJfIAP3sf7Vai6gAtn\ncEYPPQXVQKOEUkxEUYACHXTDMTwNaY55KRgMBoNxh8EcOwYAGAwGKtABA+JWrhT8/JYR8lOn\nTlk7doRER9+sFmfMmDF4cD+d7kWgCOAAcsWrW+Z0QlQQhH379jWnBpe2oPxg/bHjujpc6+2J\n6StsZpTvxYFXQa0NVu/r6+vr6ysmzHGMgy8tv5PvkAUgED63osKxth7oMRADnTZ0AifqUGeF\nFUAlKm2w8eBDEBKJSAUUXdH1KI6uxmp/+DeoM4PBYDBud5hjxwCAkpKSysrKS5cudevmN2JE\nDiF/CQj48uZ5dQB8fX0zMjJ+/XXhc8/929NzCSG7CdnA87M4bslNbFXOpWzAWfwUx8V24sYG\nUHA8VD6oyUXN6QarJ4R06tQpNjZWTF4mxSImMo+R4zjH9GIVFRUGg8Gxwt7oLSWKlbMf+0/g\nhB/8QhASjOBKVO7H/iM48h/851t8ey/ujcbN/EMyGAwG41aCOXYMABAHlmw2GyFEPI6MjGyB\nRvv27fvll7MPHRrcq9drhDxK6fcajSogIMAxMzTHcUlJSc3ZvELnaq8E4DBcRykohbm6/paG\n4qrX30eIn58fR8yEUJ4TACqmqFWYqqNOnPD19ZU2zMqhlJaUlDjWNhIjAxFIHXSuQ90LeGEd\n1lWhqgAFi7BoIAYOxuCn8fQilkaMwWAw7jKYY8cAgNDQUE9Pz9DQ0IqKit9//93Hx+e5555r\nsdatVqvBYBAzMej1+oqKitDQUD8/P3IFABzH7dq1qzlbLd1zjX9m5y859dyoBYZieMbBqx0A\nVJ/CwTex+0kcmo3ac04bUavVGnU5pQKlEKOrEJg7H/+gbXR0t27dIiIinGZLrHA2GxuCkJVY\nGYYwadUjAVEJCk4gZ3BmMib3QI97ca8U6w6AGWZ5mgoGg8Fg3PGwxdQMAPDx8REEoaSk5JNP\nPjGbze+8805ycnKLtb527dqCggKe58VUDZRSs9kcFRVVXV0NQBAEQojZbH711VcTEhLi4+Ob\noUlTKUp/B+Hqh+Lk4U5cZRgTsZnhE4+aXFSfxJG50BcAFISg6Bf0/Qke9pOehJC4NodPn+1m\ntviCEI7URfmvCvDSYujQupoaSqlKpbJarfJwdwCqq6udDub1R/985H+P77dj+1EcrRYq/Q7n\nH43VV3kDgPna7b0E5D7cp0SjtxIzGAwG4/aHOXYMACCE+Pv7l5aWpqamvv322xERES3ZeklJ\nidFotFqtKpVK9OGqq6vbtWvn4eFhNBoBKJVKs9lcWFg4cODAKVOmvPDCC+Hh4U1qklohmEAB\nTgXBfO2OCYfBOgIIkucnIPcLnPoC1AYC6MKgDoVgRt0FXPgB8W84NhUYNkKjnllRHiHYFD66\n476cCc/srK2tPX78uNlsdvTeANhstrKyssDAQMdLHLhJmDQJkwDg0EE8+mjSusIDCSa7Yjz4\nzuj8Ht67nofCYDAYjNseNhXLuEqHDh2Sk5Nb2KsD0KlTJ2nISoz6xvO8t7e3Wq222WwKhbfJ\n1EcQhthsISUlJYsWLXr66afFFA43jKAKNnH+IJQK1vphOo6AU0Lco2q3MZZe8fcoQAXYLKAW\nQAAVoC+EVQ9OBSqgLs9Faw94+n8c1S4ypqPCNyoVEdngvEtKSsxms1qtVqvV4sZYOw+vqKio\n4dy4gYFQKB79xT7FmwKK5Vi+Ezs7oMN1PRYGg8Fg3O4wx47R+owfP75NmzaUUpPJZLPZ1Gq1\nh4fHI488MmLECEI6m0w/U7oE+ArIonSMwWDYu3dvdnZ2U1rMz8/P9XjGBo3ktVGqgDYSSi8o\nPa/OxtqlkSVXArNIUAp9PmxGEB5eca4b7AF8ACwBXgMCAMiTUojL7KTtsaKPW1VVdf78+Qa6\nERGBQYP2dNPbiYMQlIpUH/g0/CAYDAaDcWfBHDtG66PVatetWxcXF6fVajUajYeHR3x8/MSJ\nEz/++GO1+n2gK2AAqoEwYLbVGlpTU5Ofn9+UFsvKyi6rHjgSMK9K3d3KedqIWuDUEAzwbIeu\nc+oLSV6d3dys3dyprQ6mUni1RZvrSNGh0+kAiMOT4gYRKaCduDGZ47jS0lKr1W3APEKE+e/+\ncZ+aUMJREErEeCgqqGxOI7kw7jIyMzPF5Q1yVCpVZmamXUnxN0Z1dbVdYbtiZrM55NosxgDO\nnj07YsQIb29vpzWLZGdni8t209PTx4wZI7+UlpZGCElLS5MLx4wZk56eLuk2ZMgQxzqJWwAM\nGDBg+/bt7h4QwwHHJ+n4Z5XbleNVp8tL7OSOhrR69WqxTrsKpbbk8gYNVUIyvEYq32I47Rea\nyWiZY8e4JWjbtu327dvnzJkzZcqUt99+e9WqVQEBAYA3x3UH9EAtYAYqAD+guyAI0U0Lsid6\nVJXqpENBH+8KX3fY790876dp4ntI/i80wSDXBlux+9Kw8/M4BTq/hn6roAluvAJhYWGenp4W\ni8VoNAqCoNVqpYRpPM9rtVoxxJ3Z7CLd2RV+8z1a6mUCgUAAAgJCQNqhHQ/7eDGMu5DU1NSM\njAx6LRkZGampqU7L5+bmdu7cWV5YulRZWblr167HH3/cLhaPXq9/6KGHUlJSCgoKMjIyJk6c\n6PRNOW/evFmzZgEYMmTI1q1b5csMsrOz4+Pjt27dKkmsVmt2dvYjjzwiSbp06fLNN984Vmun\nqt3p7Nmz586d26gnxZDh3mAyMzMnTpwo2ZWbP7pTnBrSmjVrnnjiiRUrVthVKG9LLndjqHZI\nhtcsyjcXmZmZTzzxxA8//OCoQ/MYLW1BNmzYkJSUNGLEiJZslHH7YjDQ4OBLQAFQDBgAE2AA\npms0GpvN1pSac3Nzs7Ozs7Ozt27dKh4cOHCg/trR92k6T9MVdDnq/6Xj6rHjv1WhN6aDxWIp\nKCg4e/ZsYWGhxWKpqanZtWvX9u3b9+7d+7///W/Hjh179+6Vtgm7YjQdraVajnKEEkIJKLRU\nu4PuuDGVGJTSESNGJCUlJSUlFRQUtLYuTQWAUxNy/OYXJT/88IOr72dXb41FixY99dRT0umq\nVauUSqXdvTk5OYmJidJpnz59duy4aqKxsbEZGRlxcXGSZNu2bf369ZM3rdfre/bsmZ+f76YX\nTl9nCQkJOTk5TnvEcIrTxygJV61apdPpVq1aJb9qJ3TlV8AhuaJ0SaPRiF6dVKFoRUqlUt6W\nJHdjqHLsDK8xyrcMSqVy+fLlch3kn5qmGy0bsWPcumg0GDhQAHyBQEAJ8AAPPK1WRw8bNmzR\nokVOMzTYkZeXt3jx4oULF27btk0SxsXFeXh44MoXjUqliou7skIuIAngQenVcTunvwalYTzt\nDUZyVigU4eHhsbGxYWFhCoXC09OzTZs2SqXSaDSazWaFQhETE+NmigGABZZTOOUHPx/4KKHk\nwfPgn8JTfdH3xlRi3GGEh4f/8ssvjS9/+vTpdu3aOb0kvTPs5OvXr588ebJ0Onr0aMdh5h9/\n/HH8+PHS6bBhw7KyssTjCxcu6PX6sWPH1tbW5uXVbz/KysoaNmyYvAatVrtw4cKpU6c2vi8i\n48aNa625tjuG0tJSP7/6BNmpqalpaWmjR4+WFxg9enRaWpqrYWA7nBqS0WiUT9CPHj3aYrEA\nsFgs8rYkuRtDlWNneI1RnhCyevXqbt26SRFGXc2Z2n05S6eEkKysrO7du6tUKldTvRaLZdy4\ncdJpr169AgICpNOmGy1z7BitT3V1dVpa2kcffbRu3Tq7VWWffhqqUJgACliBWiAPCLGaEvbv\n3fXBBx+8++677mvevn3PAw9kTJ/e/Y03HhwxIv+llz4R5WIei/bt24eGhsbExHTr1s3H58pW\ng5AU+CeCiFObCoCAcPaTs7ji7RGCuD83+QHUExYW1qVLl/DwcJVKZbPZTp8+ffLkSfGLzClK\nKEMRaoY5DnHt0T4c4SEIGYVRzaUP43Zn2bJl06dPT05Onj17thgtUpQ7+mciubm5OTk5HTp0\nUCqVjVx+dOjQoSNHjiQkJLi55bfffuvTp490Onz4cMnd3Lp1a3JyMiEkOTlZ2hH1yy+/DB8+\n3K6Sfv36xcTEfP/99w2qZHfX7t27r+sWhoTVas3Ly5s/f/6IESNEiZ2nJeHUoW88lFJ5tqEV\nK1Z06OBuR38jDdXO8Bqp/Lx585YsWSJG2nI1F+yel156afHixQaDwdUtlFLRcTQYDHv27Jkw\nYcLMmTOlq81gtE0Z7rte2FQsw5GCgoL+/fuHh4cHBARERUU98cQTJpNJumoy0ZiYKoWimOMO\nc2QfcIiQQl+P1FA/PtDfu3PnzlVVVXYVVlZWSsedO39LSD4hJYQUEHJZo9m+e/fhhnWqOUO3\n/4lmRtLV0fTXXvRHf7qco+mkfu41DVcP0hX0twn04hpKG5gzbSQ2m03c85udnb1ly5bs7OyD\nBw+6Kf89/T6OxgXQgGAaHESDhtFhdbSuWTS5a7mTpmIppTabbefOnbNnz+7fv79are7Xr9+F\nCxcci4nvgvvvv3/QoEHHjx+32Wyupqjs3hoajaZLly67du0ym82ubgkPDy8qKpJLYmJiiouL\nKaWTJk364osvKKWff/75pEmTKKWFhYWxsbFOW6ytrU1KShJvdNTE6evs4sWLERERTp4LwwVO\n/YTt27dLV2+sBkd/w2lVer1+zpw53t7ee/bscVqm8YZKHQzPUQFH3QAcPnz1HeFqLthpbdLB\nunXrnN7i2BeJXbt2SfKmGy1z7BitzMyZM4ODgzt37tyjR4+4uLjIyEi75d6TJ1NfX4ufd563\n9hhBCcFRno9T8FDyJCAg4OTJk1LJNWvWDBgwoGPHjn369Fm+fHlVVY1KdRAoJmQ/sA84AeRN\nnfpzAwpZ9TR/LT3zDb2YSXO/pGvb0R/9aDpPlxO6nNgvsFuhpD8F0DVt6Xn7Jeo3RlVVlejS\nSWRnZ+v1elflBSr8TH+eQCcMoUNm0VlFtMhVSUYjucMcOzm1tbUZGRnDhw//+Wf7T4HTt6zT\nd5Lj+2zv3r3ub9FqtQaDQS55/vnnly5dSimNjo4+ceIEpfT48ePR0dGU0iVLlkybNs1Vi5s2\nbRozZowrTRy7oNfrNRqNo5zhCvlj1Ov1J06cmDFjRu/evR2vNqYGN3LHYt9//31kZOTAgQPP\nnDnjqkzjDZU6GJ4rxfR6vdwtky9LdaOAG8dOPrjgpl1KaUVFxfbt2/v37//ee+/J9Wmi0bKp\nWEYrc/jwYQA6nY4Q4ufnZzKZjh49Ki8wezbuv9fso6Y6tRdIMcg8FX+WEFgFKuYcE/n999//\n+te/Hjt2zGAwnDp1avbs2Vu27LbZtICJ1v82MgOKfXtz3WljLMG2P2HvCzj4Fg68giN/h7kK\nHrHQhDpPSgEFPNvAUoVzS5vjYUD8ioFsxQal9MSJE67KE5BH8ehyLM9C1lzMDUVos6jBuAOg\nlBJC5PtPPTw8xo0b98knnzz11FPyFUI2m02KtiNHWtLknu7du7u/RUyaJ5eIy+zOnj1rsVg6\nduwIoFOnTmaz+ezZs1lZWY7zsBKDBg3y8/P78ccfG9RKRB5IiHG9aLXajh07zpo1KycnR5Qo\nlUqnc5HiQrQbbqiuru6RRx6ZNWvWnDlzNm/efHXFc+NwZah2hudK+WPHjslP3a9sdor0vS2i\n0Wjcl5ea8PX17devX1pa2rfffiu/2kSjZRbPaGVCQ0PFLa4ALBYLISQ4+Jq4IdGRptWvP54+\nfcLfHp3sq3uY534wmmGxglJYrdZx48aJH6pff/21rKwsMjIyKCgoKiqqqqpq69b/qtXnAR2g\nAXggAKi9nLei9n+zT58+ZfdRrOfkIlQchMIDmlBQK4wlAEAItMHgnX1WCQcK8Gro80GFpj8N\nDw8P6Rcernz+q6qqGghox2A4IJqQGApbTlxcXF1dnZeXV3l5uSgpKSkJCgoC4BhCTIrC4wql\nUrlu3Tr3twQHB0ttiQwaNGjnzp2bN29OSUmRhCkpKRs3bty9e7dc6MiHH364YMGCsrIy94qJ\nVFRUSAv/GTeGv7+/t7e3eOx00Zi0EO2Gm3j77bf79u177ty5Z555Ri63c8Uk62qkodoZnqsV\nbx999JF8hV9jFMCVRJfi8f79++V3yU+d6mZX7b59++Th6JtutMyxY7QyY8aM8fPzy8vLKyoq\nKiwsjIyMlG+IEwQs+Sx/4t9Gr/y9733tDoX4FHFEFjmY0kOHDn388ccAqqqqpB86YvDJqqqq\n7t3/A+QBAUA4YODIN1X6nAfGLRgyuH/fvn2//vpre23K94MACk9wCqiDAcCmBwSAs49mB4AA\nvBaCBTYTfOJBmuHTJO7VlZB+bkrL3hmMxqNQKFavXm0nFN80AwYMmD9/fnV1dVFR0axZs4YO\nHYpr33yNfFs35pbu3bufOnVKLtFqtYmJiQsXLrRz7D766KMePXqo1Wo3LXp7e8+bN2/GjBkN\ndB4AcPLkSXFEkHEDWK3WkydPvv76688++6woEfeQyt0j8Y/uuNv0usjKypoyZYqj3JV1NdJQ\n7QzPTvnq6uqdO3c+8sgj3bp1c7XH1k1DXbt2/eijj8Ss36+99pr8rqlTp+bk5NhsNle6ude/\nGYy2KfO41wtbY8dwypo1a/70pz/17Nnz6aef/uOPPyS52Uzj4/WEWAAbYFEpLj52bwLn4GAF\nBQVVV1cvW7YsIiKiTZs2iYmJcXFxYWFhn3zyyYoVK9TqaJ57iiPvcty3HPe5UjEt2MenbYR3\ncHBw27Ztf/3112tU+W0i/TGAZiXRrCT63250pQ/9OZj+FEh/CqQrdLI9E1fW2GXo6M/B9L+J\ntGxfcz2NrVu3iqvrNm/eLK20O378eHPVz3DPnbTG7qeffnI6WrBq1aqCgoLHHnvM29vb19f3\nySefrKioEG8RlytJxRzrdHxrNHjL3/72t7lz59oJP/vsMwC5ubmSJDc3F8DixYsbbJFSOmnS\nJDRijd28efPefvttRznDFXJT4Xk+NjZ2zpw5dnFDpb+40z+6K7/Czd/L6dylXVtO9zG4sTrq\nwvCkG728vO677z5xSffQoUNdKe+qob179yYmJqpUqs6dO2dmZhJCpBq+/PLL2NhYjuPc6OZG\n/6YbLaFu97A0Lxs3bpw5c2Z4ePjatWtbrFHG7ctf/0o//JBSSgmxAApKOR9dtsky2OiwmiIm\nJuaDDz7YuHHj5s2bDQaDRqO5//77v/nmG57nx48fv2ljnV6/iCIQAGD11e2OjZwOj9CLFy9O\nmTLlvffeu1rRhZXI+T9YasFrYDNAG4Eub6H8AKy1IDxOLcI1nxeCkBREjkLkSHi0aa5e//HH\nH5WVlXbCqKioxsRtYjSdkSNHFhYWAli7dm14eHhrq3MnkJOTM3HiRHE1bQuTmJi4fPnyrl27\ntnzTjFanVQxP9PCaUkPTjVbRlOYZjJtKdraJUjUhJkIIqAVQ1xoT2oVrTuYb7Urm5eVNnTp1\n6tSpaWlp58+fj4yM7Nu3rzgtu3Tp0gEPnDp+Kshbm2+xCTUGr1pTr4q6h/w8DsFxBVLMWNj0\nOLsExlIE9ETn1xD4AKLHAkDlUZz6DLCCAJSAACCIHoP21x001T0xMTFyx47jOEEQmIfBuH3p\n1q1bYGDgli1bBg4c2JLtbtq0KTg4mHl1dy2tZXhNoVmMlq2xY9y66HQAKKVqQdAIVEvBEUJf\nfWu+FBNcgl4J5eDl5TVhwoT+/ftLu4q8vb2Jqqe/r0rBo9Yg2IQai1V9vsDr8OHDFoulZ8+e\n19ZEEPdnDN6OoX+g3yoEPnD1irUaKj+AByUAAQg4Nfx7NHuvfXx85NNngiAEBwfrdLpmb4jB\naDHeeeedlk/bKk5ptXCjjFuKVjG8ptAsRstG7Bi3Lk8+qd6xA5ReXVVH4Tvq0edyzxZ++OGH\ndsPdVqu1trb21KlTjr914uLo8ePEVB/2xBMwUXrearUKgiAPmHINnMPyba8O0IaA2sDxsFkg\nWOAVB99uTe6lPWVlZRzHqdVqQRCaPqrPYNwKJCcnb926tYUbbfkWGbcaLW94TfzGbhZt2Ygd\n45ZEsODkp8mqx8m1oeMo1W7frn3ttdc8PT3t7qCUCoIQFhZmJ6+uru7ceavVeonSSKAN4AXs\nBTYSQhQKxfr16xurkjoAnV6BOgAU4JXwjEbXvzmPgdI0KioqLBYLpVSpVKrVao7jxOQ2DAaD\nwWA0CBuxY9ySnPoMJxYWXOhBQQgoiLgiFYKAffswZkzw008/vWjRIvkdlNLIyEi7qdXs7Ow3\n33zz3LlzghALPAr4A4eA7wCjuHOoqqrqOrSKexp+96BkBzgewSnw6dwcXb2GvLy80tJSQRDM\nZrMUu06tVlsslgYjijEYDAaDwRw7xi1J3k8QTG1jLYRQgYqh6yilhBCIuwjeeuutL7/8Up65\nmRAydOhQefTzioqKN95449y5c97e3pcv/w/YY9eIyWTq3Pk6nTO/7vDr3nCxG8JsNhcVFYlD\niTabTRAEADbOptfrT548GR8fz2LoMxgMBsM97D3BuPUQzLBUglOFB1wO8y8lhFJKBIEQAi8v\n9O4NAMHBwYMGDZLnfqGUfv3118ePH5ckhw8fvnTpUmBgoEajcXSJKKUmk2nr1q3/+c9/WqRX\nDWMwGARB4Hleq9USQk57np6XMO+5Xs9N7z79e933FVUVra0gg8FgMG51mGPHuPXgVPBLhM3I\nCfq3Ur8J9b2sUZo0aurri1Gj0OPKPtSgoCC7ZaqXL19++eWXFyxYsGPHDgAKhcJkMpWWlhYX\nFwOwSxrD8zzHcUVFRXPnzpWP/LUi4oo6MblnHfspJQAAIABJREFUqap0YaeF+/321ypqL2ou\nLotYtpRrnnS0DAaDwbiDYVOxjFuS+DdRcwb6i8+lfBqgK/7vyWf1qq733Ye//AVSqJNjx45J\nI3bi7lFBELZt27Z///5vvvnmxRdf1Ov1VVVVJpNJ2lsqPxCnO5VKZVlZ2ZkzZ657TvYmoNFo\nAgMDL126ZDQaDwYcLFGX+Fv8NdBQSotVxb96/PoKXmltHRkMBoNxS8McO8YtiW8XJK9H0QbO\nWjMu+Z5xAU6iNfr7+4sHonsnemxqtTooKKiiokJMWOTp6alSqUwmk10gYkqp2WzmOE7cfBoY\nGHjTe9Q4YmNjdTpdZWWlzdsmEEEhKAQIAFREVaFgU7EMBoPBaADm2DFuVdQBaPO4m+ujRo3a\ntWtXXV2duMlApKamJjc3V6vVGo1GhUIRGBio0+kqKiouXLhAr+QBFEuKI3w8zw8ZMuTChQuZ\nmZmenp6DBw8ODg6+LjUrKrB3LwQBPXrgOm91AsdxYWFhYWFhQzAkXUiv1db6Wn2tCivlaS/0\namrtDAaDwbjTYY4d43bl2WefvXz58scff1xdXS25a4IgCIJQU1MDgBCi1+tVKlV1dbXk/NmF\n/FWr1Vardfz48TU1NTzPR0ZGfvXVV/fcc08jddizBy+/jPx8AAgOxvz5eOih5uldP/R7nHt8\nJVaW8WVKKDuj8+t4vXmqZjAYDMadS4vGtd+4cePMmTPDw8PXrl3bYo0y7myGDx++bds2cT/p\nDRuzh4eHh4cHALPZnJKSsmrVKrsCRhhVUHHXbjay2TBoEI4cQUAAeB6XLiE0FKtWoWPHG9PC\nCbux+wiOBCDgQTzoBa9mq5fhmpEjRxYWFgJYu3YtS9HLYDBuO9iIHeP2RqFQiJOqbrw6juPk\n07WO1NXViYvwCCH79++vq6sT/TwAR3DkXbybgxwNNGMx9mW8rEF9tom9e3HwIAwG1NWB42A2\no7ISffvi2Wcxbx6aJeTc/bj/ftzfDBUxGAwG4+6AhTth3N4MGDCgwYE6916dCM/zSqXSZrNV\nVVVptVpRWI7yqZi6CZtqUVuEos/w2T/xT/ESpfjoI9TWwmKByQS9HlYrBAGlpVi4EN9+28Ru\nMRgMBoNxI7ARO8btzdSpUxcvXpybm9uUqVgAFotFPOB53mg06nQ6ALux+xzOBSFIBx2AAhSs\nOPs/SzotuUQCA7FvHyiFY5smEz77DJMn37AuDAaDwWDcIGzEjnF7o9FopkyZIs2c3jCiXyim\no8jNzRUleZfyrGarEleStB6LP/no/E8XISMDn36KoiIQgmvDHtdz8mQT1WEwGAwG40Zgjh3j\ntue5554LCwurTyXbNJRKJSHklVdeuf/++/38/F5v/3pZ+7JzC86ZrWY99FXfjKUF4WFhJDoa\nQUGw2SAIztfSmUw4dKiJujAYTYUQ0quXfZScXr16ySN7u7px7dq1w4cPd3p12LBh69atc5Sv\nXbtWoXAyC+RKDiA7Ozs5ORlAenr6mDFj5JfS0tIIIWlpaXLhmDFj0tPTJSWHDBniVHk3ABgw\nYMD27dud6sNoAZvJzMxUqVSEEJVKlZmZ6bR8Y2xGQqrQfZ03G1f9agV7oy3Ihg0bkpKSRowY\n0ZKNMu5s9Hr9O++8c73B5+TIv7C8vLyUSqWUWJYQwvEcH8B7f+kdQSP8R+z0D7ImJVHxn5cX\nVSgox1HA/p9SSf/979Z+NIwbYsSIEUlJSUlJSQUFBa2tS1MB4OfnJ+9Ifn6+GNlbKuDqRqvV\nGhUVVVRUZHepoKAgKirKarXayX/77bcBAwY4VuhKLjJw4MBNmzZRSsvKygICAuTVPvPMM/Hx\n8ZMnT5YkFovF39+/vLxcUvLVV1/9+uuvHZV3c0op3bRp06BBg5zqw7jZNrNq1SqtVvvTTz+J\nxzqdbtWqVXblG2kzInaVuKrzZuOmXy1vb8yxY9zevPvuu+Hh4f7+/lxD21AbM54nlpH/vud5\nXqlSJo1KWk1XPz/DFBhIu3WjSUk0Pp4GBtJHhlgDAykh9o4dx9GBA6nN1tpPh3H93GGO3YQJ\nE7744gtJ8vnnnz/xxBONeUlTSmfOnPnBBx/YXZo/f/5bb71lJzx27Nh9991XUlJiV6EruUhO\nTk5iYqJ02qdPnx07dkinsbGxGRkZcXFxkmTbtm39+vWTK6nX63v27Jmfn++ovKtTkYSEhJyc\nHEc542bbjFKpXL58uXRp1apVSqVSXvi6bMapG9cqvp37frWwvbGpWMZtDKV03bp1giBER0eL\n2x1cQQiJiIiQtru6KuP0lArUu9p7FEYldVUZDDhyBDk5yD0l1FWbcnMKBsT/9uCACruKKcWu\nXWilCQEG4yqjRo1as2aNdLpmzZpRo0Y18t7JkycvXbrUTrh06dLJ1+4MKigo+POf/5yenh4U\nFNQYucSPP/44fvx46XTYsGFZWVni8YULF/R6/dixY2tra/Py8kRhVlbWsGHD5DVotdqFCxdO\nnTq1kT2SGDduXGtN2N363FSbsVgs48aNky716tUrICBAOr1em0lNTU1LSxs9erS8zOjRo9PS\n0lJTU8VTQsjq1au7desmze26mjN19QoghGRlZXXv3l2lUrma6nXfr5a2txZzISkbsWM0NxaL\npVu3bhEREUlJSV27dlWpVG5MXaVS+fn5Sd8XDp9hf7V6Os//nZBJgIdUhuM4tVr99ttvb95M\ndbqrg3MENNCrNDooP9jn8nOP/Lg2o0itvmbEjhDar5/ZxkbtbjfusBG7qqoqLy+vmpoaSml1\ndbWXl1d1dTUaN/pCKU1JSdmzZ48k37lzp92kUmVl5QMPPHDgwAG7G13J5aSkpGzbtk06PXTo\nUI8ePcTj7777LjU1lVI6bty47777ThR27979yJEjjkpOnz592bJljnI3TW/ZsuXhhx922ve7\nnBawGRG9Xr979+5+/fp98sknouQGbKYxPgyApKSkffv2WSwWeu14nt3YnivLAdC+ffvff//d\narU2OBzo2C/a4vbGHDvG7c3kyZODgoI6duzYpUsXf39/hULhaspVqVQqlUoPDw+O43j7vawh\nhPzBcbVAHXAZyAK8pbtSUlLKysr697efclXw1nvanYgKLOoceargt2Va7VWXDqCE0MhIw759\nB4xGY2s/JMZ1cIc5dpTShx566Mcff6SUrly5Uny7NP4lvXz58ueff16SP/PMMytWrJBOjUbj\ngw8+uGHDBrsbXcntCA8Pt1uPFRMTU1xcTCmdNGmSOBv4+eefT5o0iVJaWFgYGxvrVMna2tqk\npCTxRse2nDZ98eLFiIgIp32/y7nZNiMVlti1axe9UZtx/FvLkYSHDx+WyiiVSrlbJp8zdePY\nrVu3zuktDfZLpIXtjU3FMm5v3nrrrW7duhkMhvLy8uDg4IceesjLy0uj0TiWtFgsFotFr9dT\nh88/8CmlXQVBA6gAH+ABnp+kVCq1Wu1bb721YcMGf3//8+ftQ9ZZbfzlSl8BnCCQw8c04jC/\nvIy3t9VgqLtw4ULzdpnBuC6kmTXHOTVXixOkhQ2PPvrohg0bjEYjgLq6ui1btsinvZ588smn\nnnrqwQcftLvdldyOiooKX19fueSRRx759ddfAWRnZ6ekpABISUnJzs4G8MsvvwwdOtRpPR4e\nHu+///706dPdNycnICCgrKys8eXvNm6ezYhQSisqKrZv396/f/+tW7eiCTZjV62IXq+XyxMS\nEqRji8Ui12f06NFSEFM39OvXrzG3OPZLpKXtrcVcSMpG7Bg3h7q6uo0bN/7888+5ubk5OTkJ\nCQleXk7SqrrePKEBygAbUAfUAWbAGhj4c0JCQkRExOrVq8VWevRwsvuVEIEjtuiggg0/7IuK\noioV5TjKcQIhlOeFKVMKdu7cuXfv3tZ9Pozr4s4bsbt48aK/v7/BYPD39xf3GUjf/G3atJEG\nuiRKS0vbtGkjnb744ovp6emU0m+//fb//b//Z1d/U143SqVS/KElsW7duvHjx585c0YMYCQS\nGhp65syZsWPHZmVlOfZO4tlnn125cqWj3OlrzmAw6HQ6RzmjBWxGfpqXl9e+fXvq2pbsGrKz\nGbvhN4l9+/bBxRCjY51OS9bV1cnldhMvbiqx65dIC9sbG7Fj3PbodLrBgwc/+uij7dq1S0xM\nXLRoUXx8vKMbR11+d8QCHEAB8RYBIHp9bXFxsVqt7tSp065du7766quuXfc7uZUSgFQb/TM2\nJbVvD50OajXlOHAcjY428TxefLHLtGmdPvgANTXN2WUGo/FERka2adPmvffei42NjYiIkF9K\nTk7OyMiwK5+ZmSmOlolMmTLlu+++A/Dtt98+++yz8pKObxSnQklu11BwcHB5eblcMmjQoJ07\nd27evFmuQEpKysaNG3fv3i0XOvLhhx8uWLCgkeMiFRUVfn5+jSl5d3LzbEapVMq3Eezbt+/8\n+fNwbUt2DdnZTEZGxsSJEx33JXz00UcO622cK5CZmalUKqVLZrNZPN6//5ovfPmp/JYG+yXS\n0vbWnF5iQ7ARO0YLUFxc/OSTT6pUqgYDoFzBDzgHWAALYAZsgBEYpFQqg4KC4uLiIiIigoOD\nQ0PbqdWFdsvsOCIoFEJAAE1IoL//TocNo3FxQkSEMTn5clJSpaenxdvbEhhoCwujM2a09nNh\nNI47b8SOUjpnzhyVSjV37lw7+fHjx4OCgv71r39dunTJbDbn5eV9/PHHISEhJ06ckNfTu3fv\n7OzsPn36NL7FxsiHDRu2ZcsWO+HQoUM7deokj063ePHi9u3bjxw5ssE6169fP2HCBDRixC47\nO3vgwIEuOnFXc7Ntxs3eBVeayHG0GXklVVVVO3bsGDJkyIIFCzp27Oi0HjcK9OjR47333qup\nqTl27Ni9994L2YhdYmLiwYMH3WyecN+vFrY35tgx7igOHDgQHR3t+HOqId4FSgADYAbqgJUA\np1AoxJ0WarU6MTExISHBx2eEh0ep3dZXjqMBATQ+ntbWUquVnjlD9+ypTUysEydkFQoaFESj\nomiHDtQhbCfjVuSOdOwOHjwI4ODBg3ZySunx48dTU1NDQkKUSmVISEhqaurx48ft6vnqq69C\nQ0P/3Yig29fl2P3tb3+T/AaJzz77DEBubq4kEVP8LV68uDF1Tpo0qTGO3bx5895++22nqt7l\ntIDNiJsPALiaSHVsUcKpzUgVenl53XfffRkZGZTSoUOHuqrHlQJ79+5NTExUqVSdO3fOzMwk\nhEg1fPnll7GxsRzHudHZTb9a2N4IbULe9Otl48aNM2fODA8PX7t2bYs1yrh7KCoq6tWrV2Fh\nYeOtWhzVEwQVMBl4GFAAW4CvgBoxQDGlVKFQtGvXTqfTXbp0yWSaW139tNUKQQAASsFx8PXF\nqFFYsqS+zhdewOLFsNkgzgYTgpAQaDRYvRqJiYDFgqIiBAfD2Q4PRqszcuTIwsJCAGvXrg0P\nD29tde5kcnJyJk6cePjw4ZZvOjExcfny5V27dm35phlNoVVsRnwRNKWGFrY3tsaOceewfPly\ncVWvPLCkTqfz9PT08fFxnJkVi/E8T4iJkM+BkcBQ4EOgBrIfeYLoxAFms9nHp9LbG56eIKR+\nAywh6NULc+derXb9elB6tYAgoLwcvr5o1w6mZctWJyR82rPn2vh48z//ab/PlsG4m+jWrVtg\nYOCWLVtauN1NmzYFBwczr+52pLVspim0vL0xx45x53DgwAFBEKRU3wBEJ8/Hx6dv376O5cXR\nOFc/xeSVVFVVXbx4UaPRTJoU0KYN9Pp6143joNNBoYAUJn3fPpSWQhDqfTbxf4UC//d/sO7c\nMGbatGkXL86prp5WUDD+rbf0y5c37xNgMG4v3nnnnbnyX0Utgjgv1sKNMpqLVrGZptDy9sYc\nO8adQ0BAgDR/KrllQUFBU6ZMkbIS2WE2m6W843K5OLwnCnmer6ysVKlU/fv3Hz48sUsX2GwA\nQAh4Hlotjh7FsWMA8I9/4OGHcW0EJQBISsK4cfj600/3EML7+wdERMDf/zdCli5e3KwPgMG4\nzUhOTpaH+2oZtm7dmpyc3MKNMpqLlreZJs7Dtry9MceOcecwbdo0MUim6NVxHNe9e/f9+/fP\nnj27srIS7kLZ2SMIgvRh1ul0ZrO5uLh4/fr1jz32+rp1NRwHlap+jVx1NWw21NTg6IGSBe9Z\nysvt51cVCnToAAB7q6uthHhoNBylHh4eZp4/UFvbTF1nMBgMBgNgjh3jTqJTp06ffPJJSEiI\nVqv18fFJSkpatmyZv78/x3EdO3bElXG464IQUnvF/dJoNBUVvlVVdSoVtVrrt0dYLPD0RBef\n1Rn/WF1RpXCsQaVCu3YA4BEeDoDabIRSarEA8IyLa0J3GQwGg8Gwx8l7iMG4fXnmmWcefPDB\nffv2eXh49O7d29u7PuXrggULBg0aVHM9YYLFWV2FQmGxWJRKpU2wVaOa98+zVQvUTAWBmEwA\noFDgry+V++e/faF4NqVORgQpxeOPA8DA4cM3/fbbpfJylUJhtlp9fX2HPP1007vMYDAYDIYE\nc+wYdxpRUVFRUVF2wqSkpI0bNz7//PMHDx6Udrm6h+M4QohY2GqzUp4aPAywnqQgNuGqA0cp\nliyhjzyvjAiq5jhBEOwHBTkOSiVgrR3Z/oRiourbX3GhHKFRXcY9/mT/gQOb1lcGg8FgMK6B\nTcUy7hZ69eqVmprq6enZyPKCIKhUKkEQeJ6nhEIFro5DZWeorcRDL+6cUCrB8zh+2vvfmx9P\nSdwd7FPhWI/VigULsGPZlIFPvP/8orMHTtclhNZ9NEk3YcJ4V+m0GQwGg8G4MZhjx7iLELdW\nNHKlHaXUZDJptdqodlHa4VplhJKoiGcfDz6AA28jHFUqBUGwWq2WmjrTrtOdH+y89v8e+zw6\nSEw7BgCEQK2Gtzf+s65u5Ev/PXDGUm1AeS1du8/yxqe71Ob8m9hVBoPBYNyVMMeOcRcxcODA\ngIAAXJlmlUe8c4ogCEajMS83z7DeYC2xWi9ba7but12uEap1gg0mk2CzgVKDxWL6/XT+0epO\nL438fu/H4xa8uM7HhwYEICwMbdtCq0VZWWWNnlPwRKMkSgWhFAfPmY78sael+s1gMBiMuwXm\n2DHuIuLj4//1r38FBQVxHCf6du5H7yilNpvNZrNRMxUqBUFPhdrZ1OgHgVIKShUAKBUEobis\n/Kt+rx7NrPtr8Lj/PvHXkTExhOPg4wOLBXo9VOoTQI3oQ3IElFKLjdTQwJbpNYPBYDDuHphj\nx7i7GDp06KlTp15++WVvb2+VSuXt7d3Y4HYUoAnASMAEHAbOAUagDsgA/kxpTmVl5XMz3jxw\n+ExEBN55B+HhqK2F2YwePfDgg9s5nrcJECi12SgFfHx8uyb1u8l9ZTAYDMZdB3PsGHcd3t7e\n9913n1arjYyMNBqNDU7IXkEHPAVIw2xWwACYgaXAPrGSioqKrKwsAI8+ig0b8M03SE/HmjV4\n7bU/RUVFE443W2ER4KHTLfjw08BANmLHYDAYjGaGhTth3I3YbLaamprS0lKLxQJAmpO1icnC\nnOAHpAM9AE/AAwgGOEABCMDXwIuEbON5XhCE8vJy8YbgYDz0UP3N99xzz7p167799tsjR46E\nhYW99tpr8fHxN7+XDAaDwbjrYI4d427EarWaTCar1SomlhUEQafThYaGlpSU1NXVOcsMOAW4\nB6gAeMAX0AAUMAOXgWDgTUHIFgRBoVCEhIQ4bTE+Pv4f//gHAAgmcOqb2z0Gg8Fg3K2wqVjG\n3UhxcbFGo9FoNNLmCZPJVFFR8eCDD/I87+yOeIAHaoE8oAKggAnIBy4BeiAK8AcgCMLSpUsP\nHDgAYP9+TJyIAQMwZQr+9z8AAnK/wIYHsL4Hdo5D+b4W6yyDwWAw7h7YiB3jtqS2FkeOgOPQ\npQt0uuu+3dPTU6PRBAcHW63WwsJCMdUYpbS0tNTX17e0tNThjssAAA4QgArAD9ADlQAADVBK\nSDWl4Dju4sWL06ZN697972lpgwwGQgj27sWuXVjx4dpE87swV0CwofY0yvdh8Fboopv0FBgM\nBoPBuBY2Yse4/dizB488gvHjkZqK4cORk3PdNaSkpISEhBQXF1dWVtbV1XEcFx4ertVqd+/e\nXVHhJHsEkAmUAlFAGOAP1AEAIoEYwAykAzYAguBdW/va779/+tVX/WtrwXGUEFgsuHABP64w\nwFgCwQwIgA11eTg8p0lPgcFoCOKASqXKzMyUl8nMzFSpVK6uutpXJMqle+1udCV31MfpAHl2\ndnZycjKA9PT0MWPGyC+lpaURQtLS0uTCMWPGpKenS00MGTKkMY9CDoABAwZs377daWfvKlrL\nZlavXu0or66udvxLOUWymUYq2WK46u/NtTfagmzYsCEpKWnEiBEt2SjjDqOujvbtSwMDaYcO\ntH17GhhIBw+mVut117N58+bBgwcHBwcrFIqIiIiOHTuq1e6Xvg0GfgT2A+nACOBd4H9AIVAI\n5BAyG/AAvgSKgHJAACjH2XQ6qlRShYI+PXgVTefpCg3N0NEMHU3n6Jp2N+HxMJrKiBEjkpKS\nkpKSCgoKWluXpuL4Db9q1SqlUik/1el0q1atcnrqtAZJLi/cmGM7zp8/HxISkpmZ6Xhp4MCB\nmzZtopSWlZUFBARYZR/vZ555Jj4+fvLkyZLEYrH4+/uXl5dLir366qtff/21+0fh2K9NmzYN\nGjTIaWfvKlrFZlavXq3T6X7++Wc7+b59+zp37twYtSWbaaSSLcOqVau0Wu1PP/3kqMNNtTfm\n2DFuMw4coLGxtH17mpREk5Jo27a0bVt66tQN1vbZZ5+FhoZ26dIlPDy8sQHt6vEDthFSClwQ\n3TtCPgJOAeeB04AVEABBq6137OY9s4Qu5+gPapqhpT+oaLqCrm5DbaZmfTaMZuDOduzkQqdv\nOzuhm5e0UqmU3yu9/l3J7UhOTv7kk08c5Tk5OYmJidJpnz59duzYIZ3GxsZmZGTExcVJkm3b\ntvXr10+umF6v79mzZ35+vtNeu+lXQkJCTk6Ok97eTbSKzWg0mhUrVjjKf/jhh8Y4DHY20xgl\nWwalUrl8+XK5DvLPws2zNzYVy7jNUCgAQNq3KgggpF54A4wcOTIyMrK4uLi2tpY62QzrhnuB\nOEJKFYoKf3+jWq0DhgMKwAwYAItYyGSCzQYfH+vYp7Tg1aBW2CwAwKkQeC841Q3qzWDcEKWl\npX5+fuJxampqWlra6NGj5QVGjx6dlpaWmpraYFVKpVJ+b58+fTQaDQCLxSKXjx49WgwqJGfl\nypUFBQXTp093rPbHH38cP368dDps2DAxNiSACxcu6PX6sWPH1tbW5uXlicKsrKxhw4bJa9Bq\ntQsXLpw6dWqDXbBj3LhxrTVhdyvTAjZjNBrlc+6SzZw+fbpdu3YNVmtnM41RkhCyevXqbt26\nKa68PNysH5DXI50SQrKysrp3765SqVxN9VoslnHjxkmnvXr1EnNaitw8e2OOHeM2o2NHtG+P\nykqUl6O8HDU16NIFMTE3WFtkZOTnn38+YMCA60hBUY8PwAuCWRB0lZVBZrMXpXGANxAICEAJ\nYAMEoEapPKPR/GnU83NmbexqUQSA00DhCe+O6PzGDSrNYFw/Vqs1Ly9v/vz5I0aMECV2HpjE\n6NGjzWZzgxUmJyfPnz//4sWLgiDk5+fPmjXrISlyo1tsNtvMmTPfeOMNpwn9fvvttz59+kin\nw4cP/+WXX8TjrVu3JicnE0KSk5Ozs7NF4S+//DJ8+HC7Svr16xcTE/P99983Rh/5Xbt3776u\nW+5sWsxmKKXy1ZYrVqzo0KEDgNzc3JycnA4dOiiVSjfr5OxsppFKzps3b8mSJUajEUBmZubE\niRMzMjIopRkZGRMnTmyMy/XSSy8tXrzYYDC4uoVSKjqOBoNhz549EyZMmDlzpnT1JtrbzRgG\ndAWbimU0C0eP0hEjaLt2tF07OmYMPXOmGeqsra0dMmTI9fh2HYETQAlgFFfUiXOvgA2wANVA\nLrAASFGpvNtG+kQEaiKCPb58bwo99g+a+xXV3/bTfHcqd9hUrCPbt2+Xrt5YDSK///67nWTX\nrl1Oq7WTpKenR0ZGmkzO1yGEh4cXFRXJJTExMcXFxZTSSZMmffHFF5TSzz//fNKkSZTSwsLC\n2NhYp23V1tYmJSWJNzrq4LTvFy9ejIiIcPs87nxaxWYk9Hr9nDlzvL299+zZQym9//77Bw0a\ndPz4cZvN5mYu1c5mHP/WciTh4cOHpTJu1g+4shwA69atc3qLmwci7+/Nszfm2DFuD8rK6OrV\nNCODnj5NKaVWKz1zhp4/T2225mzlL3/5S0NbKOQ8D1SKSWRl/wSgFrgMvCAW4jl0DENSLAK9\nkPqAkladaE6NGc3NHebYScd6vf7EiRMzZszo3bu349XG1GAnHzx48Lvvvmu1Wm02W2Fh4Ztv\nvjlgwACnt9hJ+vTps3DhQlctarVag8Eglzz//PNLly6llEZHR584cYJSevz48ejoaErpkiVL\npk2b5qqtTZs2jRkzxqkOTvul1+s1Go0rxe4SWsVmRL7//vvIyMiBAweecfFj3ZXzZGczrhTQ\n6/Vyt0wQBDe3yEu6kldWVjq95EhFRcX27dv79+//3nvvyfW5SfbGHDvGbcD+/bR3bxoRQcPC\naOfOdOnSm9jWnj17NBpN44buugKlV5w5uWN3EbgMfCwWIkBMIJJiEeSFx3qBbuzXsBKM1uNO\ndexELl++rNVqxWO7UQoJN8MV8pqlekSqq6t1Op3TW+SS06dPe3h4VFdXu9JZqVTq9Xq5ZN26\ndePHjz9z5kxYWJgkDA0NPXPmzNixY7Oysly1RSl99tlnV65c6Sh32i+DwSB24W6mVWxGnDBp\n06bNv//97+tVjzrYjCsl9+3b16C75iiRX6qrq5PLjUZjIysRycvLa9++vXR68+yNrbFj3AbM\nmoUzZ+DtjYAAlJXh/fdxZeV087N79+51ZcNeAAAgAElEQVT/z96dxzVx54/jf01CQgBBLgnG\nowUVPAGllq2KgKyugKJ4gB8Fa8Xu1q27/fTY9ccWrLJU7X6K1X7X1mq90YoowVjFRcQL1K7a\ngnahSr0NCsoVIIQczO+P6U7HSTIkHAnG1/PRPzLveV9hXu28Ou+ZCUmSBu/+YeLxeABB/83k\nCAD6ejtVAgAielulgSdNYGcHocMBGv7TU1NHqCPu7u4uLi7UZ4M3BtE3G3XYlVarZbYtLCyk\nbngXCASsd9oJBAJ6UyaTxcbGOjs7G+vWy8uL/sFlSmRkZHFx8alTpyIiIujCiIiIkydPXrx4\nkVmo79NPP12/fn1tbW2HXwcA6uvr6acEEM0CMZOamjpp0qQ7d+4sXbqUWV///YjMWKKxYsbY\nHW+ZmZlGfliIK2gFAgF9Z97Vq1eZrZibBufG6vbKlSt3796lN3sw3noiWzQGr9ihTnj6lPT3\nJ1966Zf3m4waRQ4YQH77bU8N9+mnn/bp00cgEHSY2wHEAcgBGv57dx11xU4FUAXwEOCX/0Lx\nCOjnAkO94e1pROsuIHPFPTV11B1s9YqdRqP56aef3nvvvdTUVLrQ8u+xi4mJ2bNnD8ecY2Ji\nioqKWIXR0dHDhw9nvp1u69atw4YNi42N5fjKlOPHjy9cuBBMuGJ3+vTpKVOmcMztRWCVmPH3\n96fvhmQy8Z2I+jHDrNzY2Hj+/Pnp06evX7/e39/f4CQ5Bho3btzatWubmprKy8tfffVVYFyx\nCwgIKC0t1Wq1xubGPf+eizdM7FBv19r6SzJHJXZ+fuTgweR/7+XtfiUlJUOGDHFwcODxeNSC\nrPEMz4sgLgE8AWj874vrNABPAX4G+BxAwOPxCAIkrnDkPbiZSZD7gdxHkJfe6Kmpo+5gY4kd\njc/n+/j4rFmzRvfsfanUIhpVR38Ni+MkzWxr8N5zgx16e3vfu3ePY86rV69OT09nFW7evBkA\nKisr6ZLKykoA2Lp1q8GJsbz++uumJHYZGRnMDObFZJWYoV56YvCqE0cs0QzGDN3Q2dk5JCSE\neuI1Ojra2CSNDXT58uWAgAChUDhixAipVEoQBN3Dli1bfHx8eDwex9w45t9z8UaQ5r27q0tO\nnjyZkpIikUhkMpnFBkU2IDkZDh0CHg9EItBqYdw4yM0FJ6eeGu7TTz9dtWpVW1sbj8ejbgpp\nbW3V6XSsanZ2dv7+C2/dWtTW9jKAVigscnXNr6sT6HR3AH6k/s2ys7P7+yK3ldNqgWwHggeu\ngfC7S8AzsJqAeonY2NiqqioAkMlkEonE2tN54ZSVlSUmJl6/ft3yQwcEBOzbt2/MmDGWHxp1\nhVVihsrwutJDz8Ub3mOHertvvoHz50Gng+ZmqKuD0aPhn//swawOAMaNG+fm5jZo0KDhw4f7\n+fn17dtXIBCwbs6ws7NzcHCYOtW9tXXa/fsDfv7Zuapq4fTp/Vxcip2c7vD5fD6fTxDEiBEj\nPtjyM0w6CEHrIeI4TL+MWR1CHAIDAz09PYuKiiw8bmFhoZeXF2Z1zyNrxUxX9Gi8dfaF/QhZ\nhEIBn3wCjY0wdCiQJDx6BI2NMGBAzw5KkiSfz3d0dKRefWJnZ9e/f/8JEybk5+fX19cDgEAg\ncHBw8PDw+O1vfwsAAwc6ATgBgJOTE0EQQ4YMUSgUGo1GoVDMmTOHb+8Mg+b07IwRsiEfffTR\n6tWrp0yZYslBMzIyVq9ebckRUTeySsx0RY/GGyZ2qFe7eRMaG8HV9ZcfDXN1hYYGqKyEwMAe\nHDQwMFAsFt+9e9fd3V2n0ymVyqlTp2ZlZQHAtm3btmzZolAonJ2d33rrLdZvGUVGRubn5z96\n9MjR0VGlUonF4qlTp/bgRBGyReHh4WfOnLHwoJYfEXUjy8dMF9dhe3S2mNihXs3TE+zsoK0N\nqNcjtLWBvT14evb0oJ6ffPJJamrqo0ePeDzehAkTPv74Y2rXm2+++frrr1dVVfXv31//VcZx\ncXGPHz/etm1bS0uLj4/P+++/P2HChJ6dK0IIIcSAiR3q1V56CcLC4NtvQS4HkgSCgBkzwAJ3\ntE+ZMqWgoKC8vNzR0XH48OH0D0UDgFAofPnllw22Igjij3/847Jly2pqary9vZmtEEIIIQvA\nEw/q1QgC/u//YOBAOHkSACAqCv70JzDjB127oE+fPtRbi8wlFAoHDhzY7fNBCCGEOoSJHert\n3Nxg9WrA25oRQgihDuHrThBCCCGEbAQmdgghhBBCNgITO4QQQgghG4GJHUIIIYSQjcDEDiGE\nEELIRmBihxBCCCFkIzCxQwghhBCyEZjYIYQQQgjZCEzsEEIIIYRsBCZ2CCFkm6RSqVAoJJ4l\nFAqlUimrJsH4nT66FaumsXKm06dPh4eHA8D+/fvnzZvH3JWVlUUQRFZWFrNw3rx5+/fvp+cw\nffp0/T4JTgAQFhZ27tw5k/8qiEvnYub27dszZ850cXFh1szLy9MPGIVCoX8EDaJjyeDcOILQ\nkmQyGf2b4L0oDkkLKigoCA4OnjlzpiUHRQgh082cOTM4ODg4OFgul1t7Ll0lEAhyc3NZhbm5\nuQKBgFVInwtyc3MdHR2pVqZ8ZpkyZUphYSFJkrW1tR4eHlqtlt61dOnSkSNHJicn0yUajcbd\n3b2uro6ew/vvv79t2zZjczO4SZJkYWFhZGSk8T8DMkMnYqalpWXIkCGZmZkKhYKOjby8PEdH\nx8OHD5PPBsyVK1dGjBhhykzoWKLnwIw6jiC0mJKSkrCwMPrv0HviEBM7hBD6lS0ldgDQ3t5u\nsNxYCeu8Tp/RjZUzlZWVBQQE0JsTJ048f/48venj45Odne3r60uXnD17NjQ0lDkHpVL5yiuv\nPHz4kGO2Bq9HjBo1qqysTL8cmasTMfP5558vXryYLqdiQyQSHThwgFVIkuQ333xjSg7AiiWD\naZx1c7vy8vKQkJCamhrmX6aXxCEuxSKEkG2SSCQnTpwwq4lGo4mLi6M3Z8yYodFoOMqZcnJy\nFixYQG/GxMTk5+dTn+/du6dUKufPn9/c3Hz//n2qMD8/PyYmhtmDg4PDhg0bli9fbtacASA+\nPr43LMzZgE7EzPHjx5OTk+nNuLg4tVqtUqmYa/FxcXFUwPz8889Dhw7tsE9WLCUkJGRlZTEj\nkOozKysrISGB2iQIIi8vLzAwkF4bNXbzAGv9l94kCCI/Pz8oKEgoFHIv9crl8iVLluzfv79f\nv37M8l4Sh5jYIYSQbdqzZ8+KFSvCw8NXrVolk8nkcjlVTpKksSYCgYB5Zrp27Rp3OVNJScnE\niRPpzRkzZtApwpkzZ8LDwwmCCA8PP336NFV44sSJGTNmsDoJDQ196aWX9u7da/rXpFpdvHjR\nrCbIoE7EzLVr13788cdRo0YJBAI6HyJJks/n03UOHDjg5+cHAJWVlWVlZX5+fszK+lixxPr/\nChqVRNKbGRkZO3bsUKlUACCVShMTE7Ozs0mSzM7OTkxMNCXleuedd7Zu3dra2srRpLGxMT4+\nfsuWLb6+vqxdvSUOLXl5EJdiEUK9nC0txZIkqdPpiouLV61aNXnyZHt7+9DQ0Hv37ulXA717\n7NRq9cWLFwMDAx0cHDjKmSQSyaNHj5glL7300uPHj0mSfP3117/88kuSJL/44ovXX3+dJMmq\nqiofHx+Dc2hubg4ODqYakqYtxT548GDAgAGm/UlQB8yNGZFINHr06AsXLqjVav3lUaVSuWbN\nGhcXl0uXLpEk+dprr0VGRlZUVOh0Oo61VFYs6ceAfhoDANevX6frcNw8YCyiAODo0aMGm9BU\nKtXUqVMLCgoMzq2XxCEmdggh9CsbS+yYmpubs7OzZ8yYQd3SzsQ8OVHnMz6fP3z48E8//XTw\n4MHc5TQHB4fW1lZmyVtvvbV7926SJAcPHvzTTz+RJFlRUUE13LFjx9tvv21sDoWFhfPmzdMv\n19+kKJVKkUjU4V8AmcuUmAGAy5cv0+XMfGjv3r0DBw6cMmXKrVu3DPZvMHki9WLJ2EUopVLJ\nnAbz7kD9JsyaxsobGhoM7qLNnz9/7969xur0kjjExA4hhH5lM4lde3s7ADCfS6XcunXLycmJ\neTbSarU8Hs9gJ3l5ebNnzzaxXCAQKJVKZsnRo0cXLFhw69at/v3704Xe3t63bt2aP39+fn4+\nszLrJPrmm28ePHhQv9zgOb61tdXR0dHgV0Cm61zMAIBGo2HWB4Dm5ubp06e//PLL27dv5x7U\n4AFlxZLBZ3VJkrxy5UqH6Zp+CXNXS0sLs1ylUnXYCcfiZy+JQ7zHDiGEbBBJkgDQ1tbGKvf1\n9W1paXF2dq6rq6NKampq6HvAhUIh81Vz1DOMHOVMXl5edJ+UyMjI4uLiU6dORURE0IUREREn\nT568ePEis1Dfp59+un79+traWlO+bH19vZubmyk1EYfOxYxAIDh69ChdWSqVCgSC1NTUSZMm\n3blzZ+nSpcyu9F+OKBAI9GfCiiVjd7xlZmYy7+RjYt0VyhxIIBDQd+ZdvXqV2Yq5aXBu+lkU\nMG5A7C1xaMksEq/YIYR6OZu5YqfT6ezs7Pbt28cqpxa/ZsyY8cEHHzQ2NlZVVSUnJ7/xxhv0\nXgcHh0OHDlVXV//9738fOXKkWq3mKGeKiYkpKipiFUZHRw8fPpz5drqtW7cOGzYsNjaWVVP/\nfHT8+PGFCxeCCVfsTp8+PWXKFI6/BjJFp2NG/x2H/v7+9F2SrK5MeSGifiwxKzc2Np4/f376\n9Onr16/39/enKrACg2OgcePGrV27tqmpqby8/NVXXwXGFbuAgIDS0lKtVmv6u1SY4/aSOMTE\nDiGEfmUziR1JkocOHdK/5ECtasnl8rlz57q4uLi6uiYlJdXX19OtqLO4s7PzrFmzHjx40GE5\nbfXq1enp6azCzZs3A0BlZSVdUllZCQBbt25l1TSYsb3++uumJHYZGRmpqalG/xDIZF2JGbom\nSZIikcjYhST9yvoMxhLd0NnZOSQkhHriNTo6mtqrHxjGBrp8+XJAQIBQKBwxYoRUKiUIgu5h\ny5YtPj4+PB6PY24szHF7SRwSpPE142538uTJlJQUiUQik8ksNihCCJkuNja2qqoKAGQymUQi\nsfZ0nidlZWWJiYnXr1+3/NABAQH79u0bM2aM5YdGPcEqsURleF3poZfEId5jhxBCqBsEBgZ6\nenoWFRVZeNzCwkIvLy+rn01RN7JWLHVF74lDTOwQQgh1j48++ig9Pd3Cg1LrXxYeFPU0q8RS\nV/SeOLSz9gQQQgjZiPDw8DNnzlh4UMuPiCzA8rHUxXXY3hOHeMUOIYQQQshGYGKHEEIIIWQj\nMLFDCCGEELIRmNghhBBCCNkITOwQQgghhGwEPhWLkMkarkHdVbDrA+IpYO9h7dkghBBCbJjY\nIWSanzZC+XpQ1wMQYO8JYXng/oq154QQQgg9A5diETKB4if4cQ201kC7GtrboFVOFk5RNjy0\n9rQQQgihZ2Bih5AJHp8CdSMQv76+ktA2OR4f3H5ARJ6aAoqfrDg1hBBCiIaJHUIm0CoB9F9K\nTvJ0bUT1acgPhvqrVpgVQggh9CxM7BAyQb9J7BIqzSMAAECnhLNzQdts2TkhhBAJqhpo+BGa\nb4FGYe3JoF4BH55AyARaBQDx60U7Kp8jGR/aquFxEQyMtdL8EEIvmNYquLMHHp0E5UMgtUAQ\nQAjBfRxIYmDwXOAJrT0/ZDWY2CFkChIIHpA6euuXlI5CEAAEtMqtMTGE0IuGhDt74KeN0FoF\nAMB3BJ4dkCTomqHqBNSchTt7Ydyn0HekteeJrAMTO4RM0C8UhK6grgMSfrluR128ozI8ggf2\nHuDsZ80ZIoReEP9ZCz9/DVolOIgNXJnTKKD+MlxYBK9shn4TrDE/ZGV4jx1CJrBzgvFfgtAD\nCD4QdmBn3+7kCTw+8PjAsweRF/T/HXhHWHuWCD1DKpUKhULiWUKhUCqVsmoSBPMSNKjVarFY\nzCzJy8ujutJvTo9isGfK6dOnw8PDAWD//v3z5s1j7srKyiIIIisri1k4b968/fv303ObPn26\nfp8EJwAICws7d+6c0b/Oc+ruPri1HdrV4DTQ8HqrwAVEElA+hB/eh5b75nZPEMT48eNZhePH\nj6cjhBUqzIYymWzGjBkG98bExBw9epT6zBEwZsWSfivuhpYkk8ns7H65cGaFOCQtqKCgIDg4\neObMmZYcFKFu01ZP3jtE3j1IqhtJkiQVlWT5P8jSD8l72aRObe3Joe4xc+bM4ODg4OBguVxu\n7bl0lUAgyM3NZRXm5uYKBAJWIX0uqK+vLykpmTNnDvPskJeX5+joePjwYaq5o6Mj3S1zk7WL\nacqUKYWFhSRJ1tbWenh4aLVaetfSpUtHjhyZnJxMl2g0Gnd397q6Onpu77///rZt24zN2eAm\nSZKFhYWRkZH6k3mOtdaQ+ePIQ/3I/HFkfjDXP8dGk4f6kVf/19wRAMDNzY0Z/A8fPnR3d6f/\nvMbSBgDQarWDBg169OgRa5dcLh80aBB10HNzcx0cHA4dOkR2OZb0W3E3tJiSkpKwsDD6D2X5\nOMTEDiGEfmVLiR0AtLe3Gyw3VmLwf/tFItGBAwfoTWZqyModDWaNZWVlAQEB9ObEiRPPnz9P\nb/r4+GRnZ/v6+tIlZ8+eDQ0NZc5NqVS+8sorDx8+5PgWBhOOUaNGlZWV6Zc/r37aRB4Wk9+O\n6iCro/6RDiJlfqTiplkjAMDChQu//PJLuuSLL75YtGiRKYkdSZIpKSn/+Mc/WLvWrVv34Ycf\nUp8FAsG+ffvoXV2MJYNpnHVzu/Ly8pCQkJqaGuYfysJxiEuxCCFkmyQSyYkTJ8xqQp8bmIUq\nlYq5fhoXF6fRaKjPGo0mLi6O3jVjxgx6Fy0nJ2fBggX0ZkxMTH5+PvX53r17SqVy/vz5zc3N\n9+//sm6Yn58fExPD7MHBwWHDhg3Lly8367sAQHx8fG9YmOs21YUA7cAXmVRZ6AqaRqg+Y+4g\ns2bNOnLkCL155MiRWbNmmdg2OTl59+7drMLdu3cnJydTnzUaTXx8PL1r/PjxHh4e9C5zYykh\nISErK4vZCgDi4uKysrISEhKoTYIg8vLyAgMD6bVRYwu+rFVm5upzfn5+UFCQUCjkXuqVy+VL\nlizZv39/v379mOUWjkNM7BBCyDbt2bNnxYoV4eHhq1atkslkcvkvD26z8rYOkSTJ5/PpzQMH\nDvj5/fKokEAgYJ6xrl27pt+8pKRk4sSJ9OaMGTPodPPMmTPh4eEEQYSHh58+fZoqPHHihP6t\nWqGhoS+99NLevXvNmnloaOjFixfNatJ7aZWgqASeaVkdAPCEAO3QdMPccaZPn15SUtLc3AwA\nTU1NFy5cMHiPo0FDhgzx8vL67rvv6JKSkpIBAwb4+PhQmyRJUglWa2vrpUuXFi5cmJKSQu3q\nRCyxckFaXFycWq2mNzMyMnbs2KFSqQBAKpUmJiZmZ2eTJJmdnZ2YmGhKyvXOO+9s3bq1tbWV\no0ljY2N8fPyWLVt8fX1Zuywch5jYIYSQbYqMjKysrPz4448JgsjMzBwyZMjkyZPpC2Od0Nra\nmp6e/oc//GHPnj1UCX2e02g0ly5dSk5OdnBwYLW6ceMGnQgCwJgxY2pra6urq4FxIzyd2D16\n9KixsXHUqFH6o69fv37Tpk1UQxMNGzbsxx9/NO9L9lrqWiB1QJjzLguyHdqemDuOi4vLa6+9\nRiXfJ06cmDBhgrOzs+nNly1btmvXLnpzx44db775JqsOQRCOjo6vvfba+fPn6Wc1OhFL+t2y\nHqCh7Nq1Kzg4mEoomRf5WNf2OGzYsOHVV1/l8/nGmrS1tc2fP3/16tVjx47Vb27hOMTEDiGE\nbBaPx5s4ceKaNWvOnj1bW1u7YsWKt99+Ozc3txNdZWVl+fn5nT179ocffggJCaEK6fOcg4PD\nG2+8kZSUxFqEAoD6+npXV1dmSVRU1L/+9S8AOH36dEREBABERERQid2JEyeio6MNTsDJyemT\nTz5ZsWKF6XP28PCora01vX6vxncEggBoN6MJQQDfqRND0aux+uuw+skWxdHRkfowZ86cgoIC\n6vJYS0tLUVGR/kU1kiTr6+vPnTs3efLkM2fOUIWdiyVWtxSlUsksZ/5/AusiH/O+Ag6hoaHc\nTZKSkhYvXjx16lSDzS0ch5jYIYSQDSJJkiAInU5Hlzg5OcXHx2/atGnx4sXM6xk6nY7H4zoX\ntLS0REVFpaWlrVmz5tSpU6yVJmrZS6vVVlRUDB06dNy4cazm1OOQzBLqNrvbt29rNBp/f38A\nGD58uFqtvn37dn5+vrFXZgBAZGSkm5tbTk6OCX8AAACCILi/2vPE3g34faBd3XHNX5BAAjj0\n78RQsbGxx48fV6lU+fn5sbHP/KCOWCzWv2haW1vr5eVFfRaJRDExMdRiZU5OzuzZs4XCX1/L\nQgeeq6traGhoVlbWzp076b3mxhJr9ZZWXl7O3DT2ihYOrNRQJOpgBTwnJycpKYl1vZB5l54l\n49BWIh4hhBADdf5ra2tjlfv6+ra0tDg7O9fV1VElNTU1+pdGmFJTUydNmnTnzp2lS5eydgmF\nQuYr6L755puZM2ey6nh5edFjUSIjI4uLi0+dOkVdrqNEREScPHny4sWLzEJ9n3766fr16028\n/lFfX+/m5mZKzecBD8Rh0K759bcNuWmawc4RPF/rxEgDBw58+eWX165d6+PjM2DAAOau8PDw\n7OxsVn2pVMo8avRq7M6dO1nrsKxU7MqVK3fv3qU+dyKWjN3xlpmZybwrlGMCUqlUIBDQu+g7\n865evcpsxdxkNqHpP5oKjJtZLRyHmNghhJBtsrOzy8vLYxVSp6WwsLB169YpFIpHjx6lpaUZ\nW/2k5OfnL1u2zOCu7Ozs3//+94cPH66pqcnIyLh+/XpSUhKrTlBQ0M2bN5klDg4OAQEBGzZs\nYCV2mZmZ48aNs7e355iMi4tLRkbGn//8Z446tBs3blBXBG2EZAYInEFd13FNAFDXQR8f6Dex\n45qGzJo165NPPtF/HnblypUZGRmbN2+uqanRaDQPHjzYuHFjamrqypUr6ToBAQEKheLMmTM6\nnW7kyGd+2YyZitHPMdC7zI0lavWW7lChUBQXF0dFRQUGBg4dOtTg9+KYwJgxYzIzM5ubmysq\nKj744ANmq+XLl5eVlel0OlYTE1k6DnvuTSr68D12CKFezpbeY3fo0CH9SwvU28LkcvncuXNd\nXFxcXV2TkpLq6+tZbeHZ99hxnDuo9405OzvPmjXrwYMH+tNYvXp1eno6q3Dz5s0AUFlZSZdU\nVlYCwNatWzlmQnv99dfBhPfYZWRkpKam6pc/t9rJC6+Th/p1/Co76SAy72XyXo65A9B/xtLS\nUgAoLS1llZMkWVFRkZCQIBaLBQKBWCxOSEioqKhg9fPVV195e3tv375dfwgqYOhQ1N9lbizR\nHTo7O4eEhFBPvEZHR+vPnHsCly9fDggIEAqFI0aMkEqlBEHQPWzZssXHx4fH4xl86bdBzHEt\nHIcEaeZz711x8uTJlJQUiUQik8ksNihCCJkuNja2qqoKAGQymUQisfZ0bEFZWVliYuL169ct\nP3RAQMC+ffvGjBlj+aF7SusjuPA/0PAfsPcEO4MPRpCgegKkDnwSIWg9gNm3l/VmVoklKsPr\nSg8WjkNcikUIIdSDAgMDPT09i4qKLDxuYWGhl5eXTWV1AODQH17dBu7BoK4HpRy0yl8zN1IH\nmkZQ3geeHfgshjFrbCyrA+vFUldYPg4xsUMIIdSzPvroo/T0dAsPSq1/WXhQS3AeBhMPwPB3\nQdQPtApovgfK+9ByH1rl0K4F1yAI3ghBa039gYrnjVViqSssH4fmvOoQoeff06dPv//+ex6P\nN378+L59+1p7Ogi9EMLDw+nXlVmM5Ue0HKErjFwJQ5ZBzTlQlIPqCfDtQeQNHiHg8QrwhB33\n8NyyfCx1cR3W8nGIiR16gRQVFa1cufLx48cEQQwcOPCzzz6j37OKEELPGXsPGBQHYOA3tdCL\nDJdi0Yuiubk5JSXl7t27zs7OTk5OlZWVK1eu1Gq11p4XQggh1G0wsUMvioqKipqaGnd3d0dH\nRycnp759+z58+PD27dvWnhdCCCHUbTCxQy8KR0dHHo9H/8IS9TNKxn73ECGEEHoedZDYyeXy\nZcuWeXt729vbBwUF7du3T79OJ36FDSHLGzZs2MiRIxsbG58+ffrkyZOWlpbg4OCBAwdae14I\nIYRQt+FK7G7dujVu3Ljt27dXV1er1WrqxYCrV6+21NwQ6k5CoXDjxo2RkZH29vYODg4xMTGf\nfvop/m8JQgghW8L1VGxqaqpIJMrNzQ0LCwOA48eP/+Uvf1mzZs2oUaPmz59vqRki1G2GDBly\n4MCBJ0+e8Pl8d3d3a08HIYQQ6mZcV+xOnTq1b9++uLg4d3d3d3f3xMTEs2fP9u3b9/e///2D\nBw8sNkWEule/fv0wq0MIIWSTuBK7hoaG4OBgZomfn9+2bdsaGhqWLVtmyR+ZRQghhBBCHeJK\n7Ly8vPTfmDx//vxFixYVFBR8/PHHPTgvhBBCCCFkJq7ELjo6euHChZs2bbp582Zraytdvnnz\n5qFDh6alpS1evLiioqLnJ4kQQgghtpoaOHcODh2Cb7+Ff/8b2tqsPSHUCxAcK6rV1dUhISH3\n7t2jNpk1b9++HRUVdfPmTf1dHE6ePJmSkiKRSGQyWRfmjBBCPSU2NraqqgoAZDKZRCKx9nQQ\nMuz8edi2DS5fBqUSNBrg8UAgAA8PiI6Gt94CjNwXGdcVO7FY/P3333/44YdjxoxxdnZm7vL1\n9f3hhx8+//zz0NBQFxeXHp4kQgghhAAAVCpISYE33oD8fGhpAUdH8PAAV1cQCEAuhy1bYMYM\nOHrU2rNE1tPBC4rd3d0zMjKuXbumUChYuxwdHf/0pz+dO3eusbGxx6aHEEIIoV+o1fDHP8Ku\nXdDWBoMGgZcXODmBvT2IRODiAhqJ+g0AACAASURBVAMGgLc3PHwI778P33xj7bkiK8GfFEMI\nIdtEEMT48eNZhePHj6ffy23sBd0EQchkshkzZhjcGxMTc9TQFSGZTGZn98y7UdVqtVgs5pjh\n6dOnw8PDqc9SqVQoFBLPEgqFUqmUe6od7jWlkOmrr74aPHgwn8/nrmaKsLCwc+fOdb0f2oYN\n8K9/gYMDiMXAM3QCFwph4EBoaICMDPj+e7P75z4KTMwoYqH+dHl5eVRXrOb0EAa7pTFjQ39u\n3G0thhnz3X6sO4+0oIKCguDg4JkzZ1pyUIQQMt3MmTODg4ODg4Plcrm159JVAODm5sb8Ig8f\nPqRe4khXMNZQq9UOGjTo0aNHrF1yuXzQoEFarZZVXlJSQr3Kntqsr68vKSmZM2cO91lmypQp\nhYWF1GeBQJCbm8uqkJubKxAIuKfa4V5z9enT59KlS+3t7Z1oyxq0sLAwMjKyc/3ou3mTHDGC\n7N+fDA7u4J+gINLDg1ywgNTpzBuC+ygwGfzb3r17VywWS6XSvLw8R0fHw4cPU80dHR2pbnNz\ncx0cHA4dOsQq18eMDf3K3G0tgxXz3XusuwITO4QQ+pWNJXYLFy788ssv6ZIvvvhi0aJFpiR2\nJEmmpKT84x//YO1at27dhx9+yCosLy8PCQmpqalh9tzh5YOysrKAgADmoAZzKVNSt+5N7ABA\nZ25CZHzQUaNGlZWVdbo3pr//nezXjwwM7DixCw4mfXxIX1/y6lWzJ89xFLhLSJIMDw/ftGkT\nSZIikejAgQN0OZ0aCgSCffv26ZezsGLDYBpn3dxOP+bJbj3WXYFLsQghZLNmzZp15MgRevPI\nkSOzZs0ysW1ycvLu3btZhbt3705OTmaWyOXyJUuW7N+/v1+/fnQhfY7h6D8nJ2fBggX0pkQi\nOXHihIlz6wrmGqJUKh05cqS9vX1AQMD3339P7+Xz+dSHb7/9lrk0yVwB5GhOVaaqxcfHd9ei\nYWEhEATYcf0U6K9cXKC5GcxdG+zKUTh48KBcLl+xYgUAqFSqefPm0bvi4uI0Gg0AaDSa+Ph4\nunz8+PEeHh76XbFiIyEhISsrKy4ujlknLi4uKysrISGB2iQIIi8vLzAwkF4bNbbmy1qLZ8ZD\nfn5+UFCQUCjkXuo1GPPQrce6SyyZReIVO4RQL2djV+waGxudnZ2bmppIklQoFM7OztSTcHQF\nYw2pDxEREZcuXaLLi4uLWYtNDQ0NEyZM+P777411yHGWiYiIOHv2LL1ZWFjo6+sbFhaWlpZ2\n5MiRhw8fmtIPmH/Fjvn1Q0JCysvLm5qa0tLSXn31Vf1WLi4uu3btojeZV5hMaU6SZFFR0e9+\n9zuD0zNLYyPp708OHmzS5Trqn379yHfeMW8U7qPAxPqaWq3W19f366+/Nlj5m2++8fPzY5Yo\nlcqLFy+GhoZSV/hYWLFhSq4CAMHBwVeuXNFoNKTe+i/z2h6rN2Y8DBs27LvvvtNqtRyXAzli\nvruOdRdhYocQQr+yscSOJMlp06bl5OSQJHnw4EHqrGN6Yrdv37633nqLLl+6dClzfU2lUk2d\nOrWgoEC/IUcJTSKRsO7h0+l0xcXFq1atmjx5sr29fWho6L179+h+OC5PmHXxgvn1S0tLqc9N\nTU0ODg76c+bxeEqlkjlDs5qTJPngwYMBAwYY+yOY7t49cuhQ0sfHjMTO05NMSjJ7II6jwMT6\nmvv37x84cGBbWxurmlKpXLNmjYuLC/P/EJgH6MKFC/qds2JDPxXTP8QAcP36dboO62ZBVkZu\n8IsAwNGjRw02oXHHfHcd6y7iWort8NEhs6ohhBCyPHo1Vn8d1sHBwWATR0dH6sOcOXMKCgpU\nKhUAtLS0FBUVMZfDkpKSFi9ePHXq1M5NrL6+3tXVlVnC4/EmTpy4Zs2as2fP1tbWrlix4u23\n387NzaX2GjyH0W2N7WU+rak/h9GjR1Mf+vTpw/yBJZpOp3NwcKiurj58+PCf/vSn2NhYs5oD\ngIeHR21trYl/Ew59+wKfD+3tZjQhCHj2D2wS7qNgzObNm9977z2hUMgszMrK8vPzO3v27A8/\n/BASEkKXkyRZX19/7ty5yZMn6/9yKRiKDSb6+CqVSmb5qFGj6M8ajYYZq/RaMLfQ0FDuJtwx\n313HuovMu8cOcziEEHq+xMbGHj9+XKVS5efns/ISsVhcXV3Nql9bW+vl5UV9FolEMTEx1G1D\nOTk5s2fPZp65c3JykpKSWJmT6acJ6tFa6jNJkgRBUNfDKE5OTvHx8Zs2bVq8eLF5X/hZBrNA\nWofvNCkuLh47duy0adN27Njh5uaWmJhoVnMAIAiCZ/DFJGZycQFPT1CpTK1P/S0HDzZjCO6j\nwDyyOp2O+aVu3bpVWlq6bNkyuqSlpSUqKiotLW3NmjWnTp3y9fWlyulOXF1dQ0NDs7Kydu7c\nqT8TZmwAgEAgMHjvWnl5OXOzEykKKzUUiUTc9bljvruOdRdZfwYIIYR6zsCBA19++eW1a9f6\n+PgMGDCAuSs8PDw7O5tVXyqVRkRE0JvLli3btWsXAOzcufPNN99k1jR4hcxg/mSQl5dXXV0d\n3RUAtOn91qmvr29LS4uJHfaEhQsXfv3112VlZceOHUtPT582bZq5PdTX17u5uXV9JgQBU6aA\nTgeMvIuLQgF9+sCkSWYMwX0UnJ2d6eNVU1PDfG5AJpPFxsYyf6EqNTV10qRJd+7cWbp0KbMr\nVop25cqVu3fv6s+EGRsAkJ2dnZiYqJ/bZWZmGsutWQNJpVKBQEDvUqvV1OerV68yWzE3mU1o\n3DHfXce6izCxQwghGzdr1qxPPvlE/3nYlStXZmRkbN68uaamRqPRPHjwYOPGjampqStXrqTr\nBAQEKBSKM2fO6HS6kSNHduOsgoKC6B8cBwA7O7u8vDxWHYMnV0t6/Pjx/fv3AUChUJw6dWrO\nnDkEQTQ1NXE0EYlEt2/fpjdv3Ljh7+/fLZOZNw88PODJk45rtrdDYyMEBIDeC6o7wHEUwsLC\n1q1bp1AoHj16lJaWFh0dTVc4depUVFQUs0l+fj7zAh6NmaJJpdLExET9/7UAvdigHoClGyoU\niuLi4qioqMDAwKFDhxr8IhwDjRkzJjMzs7m5uaKi4oMPPmC2Wr58eVlZmU6n45gbh2481l3C\ncf+d/l6D9bk7YcKHJxBCvZztPTxBkmRpaSkw7vRn/ke7oqIiISFBLBYLBAKxWJyQkFBRUcHq\n56uvvvL29t6+fbvpI3KU0FavXp2enk5vHjp0SD+Ho2+BN9YPdO2p2A7LqTvoAcDFxWXatGmF\nhYXx8fHUcxLGmqekpDg5OdGbGRkZqampRv4GZvvb30hvb3LYMK5nJsaNI8Vi0t+fLC42u3+O\noyCXy+fOnevi4uLq6pqUlFRfX0+38vb2Zj1gYXBNk9pF/0kNvgyZwooNVkNnZ+eQkJDs7GyS\nJKOjo6m9+sfa2ECXL18OCAgQCoUjRoyQSqUEQdA9bNmyxcfHh8fjccyNhTlu9x7rTiNI45fN\nqW/LXWKs0KCTJ0+mpKRIJBKZTGZKfYQQsrDY2NiqqioAkMlkEonE2tOxZWVlZYmJidevX7f2\nRHpWQEDAvn37xowZ0y29KZWwZAmcPQuOjuDuDvo3lWm18OgR9OkDH3wAy5d3y5hWYJXYMD2Z\nMaZ7j3Wn4VIsQgghKwgMDPT09CwqKrL2RHpQYWGhl5dXN57pHR1h2zaIiwOShPv3oa4OVCrQ\n6UCrhZYWqK4GuRw8PeGjj57jrA6ez9jo9mPdaZjYIYQQso6PPvooPT3d2rPoQdTaXPf22bcv\nfPEFfP45jB8PfD7U10NVFTx+DC0t4OICCxZAbi507UniXuG5i42eONad08FPk+g/PIxvPEEI\nIdQtwsPDDb7GzGb00Lfj8SA2FmJj4cYN+M9/4OlTsLcHb28YPx7c3XtiQCuwfGx0cR2290Sy\nab85hxBCCKFext8fesNTmKhX4Ursupi9IoQQQgghS8J77BBCCCGEbAQmdgghhBBCNgITO4QQ\nQgghG8F1j53pD8Di3XgIIYQQQlaHV+wQQgghhGwEV2LH8Utkly5dGjt2LAAQBPHmm29aarYI\nIYQQQsgos6/Y1dXV/eEPf3jttdd++OGHoKCgCxcubN26tSdmhhBCCCGEzGJGYkeS5K5du4YP\nH75161ZnZ+eNGzdeuXLlN7/5Tc9NDiGEEEIImc7UX5748ccf//jHP54/fx4AEhISPvvss/79\n+/fkxBBCCCFkVFtb23fffVdRUVFTUyMSicRi8auvvjp8+HBrzwtZWceJXUtLy+rVqzdu3KjV\nav38/DZv3vzb3/7WAjNDCCGEkL7W1tasrKyvv/66urq6tbW1vb0dAOzs7Pr06RMQEPDuu+9O\nmjTJ2nNEVtNBYpebm/u///u/Dx48EIlEaWlpK1eutLe3t8zMEEIIIcRSVVX1zjvvXLx4UaPR\n9O3b183Njc/nA4BarW5qajp79uy1a9d+//vfv/fee1Q5etFwJXbR0dH5+fkAEBUV9c9//tPX\n19dSs0IIIYQQW11dXXJy8pUrV9zc3FxcXJi7hEKhh4eHm5vb48eP//nPf6rV6g8//NBa80RW\nxPXwBJXVUR+GDBlCGGeRqSKEEDKDVCoVCoWs/1wLhUKpVMqqafA/4zKZzM7ODgAUCoUp/82n\n6+s7ffp0eHi4KbMy1jlVzr3XlEKmr776avDgwd1yWSssLOzcuXNd76dDq1at+v777z09PVlZ\nHY3H40kkEq1Wu3PnzuPHj5vbfydiJi8vj2pisJparRaLxayGLAYPATNm9OdmcCzLY8a8xWKg\nQ/iCYoQQsk0JCQnZ2dmst5BmZ2cnJCR02PbChQsbNmzQ6XQAUFlZOWLECGYn3PX1ZWRkpKWl\ndX1WZunwJ5E++OCDnJwcrVbbuf6ZieOqVavS09M714/pLl68eOLECZFI5OTkxF3T29u7oaFh\n48aNarXarCHMPTpHjhxZtGjRgQMHqGqJiYl0vtXQ0HDhwoX/+Z//qampYTZh9nz37l2xWHz4\n8GH9npkxAwBSqTQxMZGeG2ssq2DFvGViwCQcbyHudgUFBcHBwTNnzrTkoAghZLqZM2cGBwcH\nBwfL5XJrz6WrAKC9vd1gOXdJeXl5SEgIdT4mSfKbb77h/u82qz5LWVlZQECAibMydlYyZa+5\nAECn03WiocFBR40aVVZW1uneTPHXv/7V09Nz7NixwSYYOnToSy+9dObMGbOGMDdmRCIRldVR\ncnNzBQIBXaHDTCM8PHzTpk365ayYyc3NdXR0zM3NZdYxWGgxBmPeAjFgCrxihxBCtkkikZw4\nccLcVnK5fMmSJfv37+/Xrx9V8vPPPw8dOtT0+iw5OTkLFizo4qw6gb6iRhCEVCodOXKkvb19\nQEDA999/T+/l8/nUh2+//Za5BMlc6eNozlyYjo+P79ELSO3t7WfOnLGzs+PxTDpxOzs7K5XK\nCxcumDWKuUdHpVLNmzeP3oyLi9NoNNRnOs8w1vbgwYNyuXzFihX6u1gxk5CQkJWVFRcXx6wT\nFxeXlZVFX0okCCIvLy8wMJBeG6WXblnrtqw1emac5OfnBwUFCYVC7qVeYzHf0zFgIkzsEELI\nNu3Zs2fFihXh4eGrVq2SyWRyuZwq5zjXNjY2xsfHb9myhfm0XGVlZVlZmZ+fn0AgYJ3wDNZn\nKSkpmThxYldm1XWffPLJ4cOHa2trZ8+evXz5cnj2+h8ALFq0aNu2bXQuwlp87LA5AISGhl68\neLHnvkJDQ0NjY6PpL6bg8XgkSdJ/XhOZe3RIkmTeIXfgwAE/Pz9TBtLpdCkpKStXrjSYp7Ji\nRqPRsLI6SlxcHHOtOSMjY8eOHSqVCp5dujV93fadd97ZunVra2srRxOOmO/pGDAV9wW9O3fu\nxMbGOjs70yUVFRVTp051cHBwd3ePj49/8OCB6ZcHcSkWIdTL2dJSLEmSOp2uuLh41apVkydP\ntre3Dw0NvXfvnn416lygUqmmTp1aUFDAKn/ttdciIyMrKip0Oh1z/ctYfRaJRPLo0SMTZ8V9\ntjLrXAaMxKu0tJT63NTU5ODgoD9bHo+nVCqZMzSrOUmSDx48GDBggP40usu9e/eGDh3q4+Nj\nyjosxdPTMykpydyBzIoZmlKpXLNmjYuLy6VLl7hrUvbv3z9w4MC2tjaDc2DFDKsHg4ceAK5f\nv07XEQgEzFVa1hqxwekBwNGjRw02oXHHfE/HgIkI0vi/KkqlcsyYMbdv36b/jWpubvb396+q\nqqLr+Pv7l5aWikQijn/faCdPnkxJSZFIJDKZzJT6CCFkYbGxsdR/4mQymUQisfZ0ulNLS8ux\nY8f27t37xhtvzJkzh7mLIAiSJOPj42NjYxMTE1nlrH6kUmlCQoJarTaxvqOjY11dnbHTBGtW\nBnuge+bYy9yk6tCVCYLQarX0hSVmOau36urq4uLiM2fO3Llz59ixY2Y1b21tdXd3b21tNfg1\nu66lpeWVV15RKpXGlrz13b9/PzExccOGDV0ZlDtmqM9ZWVkpKSl+fn7btm3Tv45l8KhNmjRp\n7ty57777rsFxWTFj7Li3trY6OjrSh6O9vZ25rspqYuzAMcsbGhr69u3LMXPumO/pGDAR11Js\ndnb2/fv3P/744/v371Mlu3fvrqqqSkhIuHv3bnNz86FDhx49epSTk2ORqSKEEDIVlQYxH1N1\ncnKKj4/ftGnT4sWLmZmQTqejlsNycnKSkpJYrzXRf2kIfR+VifW1Wi195uOeVRe/L01/b4fv\nNCkuLh47duy0adN27Njh5ubGPHOb0hwACIIw8e63znFycnr55ZdNTxo0Gg2Px+O4OVJfJ2Km\npaUlKioqLS1tzZo1p06dMvF9t7du3SotLV22bJmxCsyYAQCBQGBwVbS8vJy52eELbvQplUrm\nZodXqbhjvqdjwERcM9i/f39aWtrf/va3QYMGUSVHjx6VSCQ7d+586aWXnJyc5s6du27dur17\n91pkqgghhExFnRfb2tpY5b6+vi0tLc7OznV1dVRJTU0NdRFIf02HKmTdVyeVSgUCAUd91ohe\nXl70WNyz6p5v3ikLFy78+uuvy8rKjh07lp6ePm3aNHN7qK+vd3Nz64m50aZMmUIQhIlvMGls\nbHR2dma9Co5bJ2ImNTV10qRJd+7cWbp0qekDyWQy6i4vYxWYMQMAxu54y8zMNJZzs3JBOmip\nXfTf8OrVq8xWzE1mExp3zFsgBkzBldhVVlYmJSXRm/X19UVFRQsXLnRwcKALZ8+effPmzR6c\nIEIIoU6xs7PLy8tjFVKnq7CwsHXr1ikUikePHqWlpUVHR3P0wzyt0vekmz6NoKAg5mmCY1am\n99ntHj9+TK1NKRSKU6dOUYvCTU1NHE1EIhF1qxLlxo0b/v7+PTrJhQsXent7s14LZ5BGo2lp\naYmIiBg5cqRZQ5gbM/n5+RwX3ow5depUVFQURwVWzFAPwNJBqFAoiouLo6KiAgMDjV2S5Aja\nMWPGZGZmNjc3V1RUfPDBB8xWy5cvLysr0+l0nYhzsEgMmITj/juhUKhWq+nNnTt3AsD58+eZ\ndVQqlVAo5OiECR+eQAj1crb08MShQ4f0syXqpnK5XD537lwXFxdXV9ekpKT6+nqDPdDnCOpG\ncrq5sRENnlNWr16dnp5uyqyM9UAyrouYPi4wborvsJz+gi4uLtOmTSssLIyPj6eekzDWPCUl\nhXpRMLWZkZGRmppqcHrd6Isvvhg8ePCgQYM4npkICgry9PQMCgq6ceOGuf2bGzMG1y5ZfeqX\neHt7G3wgg8aKGQp9jJydnUNCQqgnXqOjo42NYixoL1++HBAQIBQKR4wYIZVKqZvkqB62bNni\n4+PD4/G449zYt7NMDHSI6+GJPn363L5928vLi9qMjo4uLi6uq6tj/mhMRUXFpEmTamtrjXXC\nhA9PIIR6ORt+eMJaysrKEhMTr1+/bu2J9KyAgIB9+/aNGTOmR0dpb2//y1/+kpOTo9FoxGKx\n/m+4KZXKJ0+eeHl5ffbZZ51YUO4lrBIzxh7RMJ1lYqBDXEuxQ4YMod9t+Pjx45MnT06ZMoUV\nRpmZmWPHju3BCSKEEHqeBQYGenp6FhUVWXsiPaiwsNDLy8sCZ3Qej/d///d/7777rpubW1VV\nVVVVVUNDQ3Nzc3Nzc21t7b179xoaGkaMGPH1118/v1kdPJ8xY7EY6BBXYjd9+vT333//3//+\n95MnT1asWKHVaumnnRsbG4uLixcsWLB9+3azbplECCH0ovnoo496y89o9gxqDc4yY/F4vHff\nfVcqlb7xxhvUW9NaW1tVKpVAIHjllVdSU1OPHTv22muvWWYyPee5ixlLxgA3rguP1dXVo0eP\nfvr0KbXp7e1969YtR0dHYDzcm5CQcODAARMHw6VYhFAvh0ux6DnS1tZ2+/btJ0+e2Nvb9+/f\nf/DgwdaeEbI+9vI8k1gsPn369Ntvv/3dd9/5+Phs376dyuoAQCQS+fv7Jycnv/322xaZJ0II\nIYSeYW9vP2LEiBEjRlh7IqgX4UrsAGD06NFnz57VL7f6i5URQgghhBCL9V+RjBBCCCGEuoVJ\niR33z3R04kc8EEIIIYRQt+vqFbsrV66Y8iN6CCGEEEKop3HdY8e8FMdxWa5X/IAGQgghhNAL\nr4OHJ7jx+Xw/P7/PP/+8u2bTc7QtNc0Vu8iWh2Qfnz4jXhc6ult7RgghhBBC3YwrsaNfcdf1\n39mwLm3zw7ZTsU6qe0C2k4Sd6mFO+5Q8kbOXteeFEEIIIdSdTLrH7rnO6gBAee1zoeqult9X\nI+zfzncSqW62/PiVtSeFEEIIIdTNXojXnRDNNwloJ3kiAGjnORKkjmiutPakEEIIIYS6WZfu\nsXtuOEjIOgJIHRB8aNcCEKTI29pzQgghhDpPqVQ2NDS0tLSo1Woej2dvb+/i4uLq6mpn92Kc\n2ZERL8Thtx+2WFf9L4GmGoAHRLtGIHbwT7L2pBBCCKHOaGtre/DgQW1trVarBQCSJKk3Vzx+\n/Jj60dj+/fvjK2ZfWC9EYif0/o3qN9tbK77kt97XOg4VjHzb0WuMtSeFEEIIma2pqenmzZut\nra18Pl8oFDITOJ1O19raevfuXYVCMWzYMHzL7IvphUjsAEA0MFw0MNzas0AIIYQ6r7W19caN\nGyqVSigU8njsu+T5fD6fz9doNE+fPiUIws/PD6/bvYBeiIcnEELoBSSVSqkrOkxCoVAqlbJq\nUqd/hULBqsyswEJdDTJWznL69Onw8HBTZmUsEaHKufd2RYc9GKwQFhZ27ty5Lg5tllu3brW1\ntdnb2+tndTSBQMDn82trax8/fmxu/wRBjB8/nlU4fvx4+utzHAKZTDZjxgyDe2NiYo4ePUp9\nzsvLowKAFYqdiCUKM6IMhrflyWQy+k5HywcJJnYIIWSbEhISsrOzyWdlZ2cnJCQYrF9ZWTli\nxAhmZXoXs/Du3btisfjw4cMc5SwZGRlpaWmdm1Vvw8xsVq1alZ6ebrGha2trm5qa+Hx+h2mo\nQCAgSbKqqoq6Cc8st27dqqqqojflcvnt27dNaRgTE3Pt2jX9bLKqqur69evR0dEAcOTIkUWL\nFh04cIA66ImJiXQe1olYAgCpVJqYmEhHFKtPq7hw4cKGDRt0Oh21aeEgAUzsEELIVmk0mtmz\nZ7MK4+LiNBqNwfqVlZVDhw7tsNslS5b87W9/0+/ZWPm1a9eePn0aGRnZuVn1ZpGRkY8fP752\n7ZplhqutrdXpdCY+9GpnZ9fW1tbY2GjuKFFRUTKZjN6UyWRRUVGmNOTz+YmJiXv37mWV79mz\nZ/HixdTltwULFuzYsWPOnDkAEBcXl5WVZTChNzGWqKwuKysrLi6OKqH6tGJuV1FR8d577+Xk\n5NAlFg4SwMQOIYRslUQiOXHihOn1f/755w4Tu4MHD8rl8hUrVphYDgA5OTkLFizo9Kw67dtv\nv2Wu+bIW6aj1Oz6f7+Pj88UXX9DlrIthBjeZ69Tx8fGWySFIkmxsbDR90ZnP57e3tysUCnMH\nmjVr1pEjR+jNI0eOzJo1y8S2ycnJu3fvZhXu3r07OTmZ+qxSqebNm0fvMpjQmx5LCQkJzKyO\n7pOZLxIEkZeXFxgYSCfE9NKt/lowsx/m6nN+fn5QUJBQKORe6pXL5UuWLNm/f3+/fv2Y5RYL\nEgomdgghZJv27NmzYsWK8PDwVatWyWQyuVxOlRv7MaHKysqysjI/Pz+BQGDwBKbT6VJSUlau\nXMm6wctYOaWkpGTixImdnlWnLVq0aNu2bQZXe+n1u7a2tr1792ZmZprYJzVJ5jp1aGjoxYsX\nu3fmBmk0mvb2dnMfdFWr1eYONH369JKSkubmZgBoamq6cOHC9OnTTWw7ZMgQLy+v7777ji4p\nKSkZMGCAj48PtUmSJPMrHDhwwM/Pj9mDWbGk0WhYWR0lLi6O+cUzMjJ27NihUqng2aVb09dt\n33nnna1bt7a2tnI0aWxsjI+P37Jli6+vL2uXxYKEgokdQgjZpsjIyMrKyo8//pggiMzMzCFD\nhkyePPn+/fvG6ldWVlK3wLe1tRk8gR08eFCtViclsd8DaqyccuPGDebJu8NZ6d9Ez/0Yh7GL\nWM3NzfHx8fTmrFmz6ItD9JUeOzu7SZMmbd682djfpEPDhg378ccfO93cdPo3PnaI6NTvvLu4\nuLz22mvUVdUTJ05MmDDB2dnZ9ObLli3btWsXvbljx44333xTv1pra2t6evof/vCHPXv2MMvN\niiUWY1Gxa9eu4OBg6ood8yIfx1owy4YNG1599VU+n2+sSVtb2/z581evXj127Fj95hYLEkpn\njnqnnTx5MiUlRSKRMNfvEUKo94iNjaXuHJfJZBKJxNrT6U4tLS3Hjh3bu3fvG2+8Qd3kRDOY\nAUil0oSEBOaVj0mTJs2dO/fdd99l1TRWTnF0dKyrqxOJRKbMylguQpVz7GVuMutUV1cXFxef\nOXPmzp07x44do3ax+mlsVGboLgAAIABJREFUbHR1dTW4i97U/0BpbW11d3dvbW01+O26UXt7\n++XLl3U6nVAoNLFJW1ubt7f3kCFDTB+F+nZffPHFxYsX9+7dm5iYOHHixOXLl9Pf2tHRUalU\n6jd0cnJqaWkBAJVKNWrUqP/85z8ikailpWX06NE3btxgzTkrKyslJcXPz2/btm2s61tmxZKx\neGhtbXV0dKSPV3t7O3NdldXE2JFlljc0NPTt21d/Fy0+Pj42NjYxMdFgHYsFCQWv2CGEkA2i\n0iD60TwAcHJyio+P37Rp0+LFi4ln32prcNmLdf/TrVu3SktLly1bxqpmrJym1WrpMxz3rMz8\nis/Qf5i3uLh47Nix06ZN27Fjh5ubG/Okqz9Dg+UGMxgWgiA43jzSjXg8npOTk+nXYqiaTk5O\nnRgrNjb2+PHjKpUqPz8/NjaWuUssFldXV7Pq19bWenl5UZ9FIlFMTAx1rTcnJ2f27NnMrK6l\npSUqKiotLW3NmjWnTp1iZXVmxRIACAQCg6ui5eXlzE3Tb0yksQ69sf8toeXk5CQlJbGuFzKz\nScsECQUTO4QQskHU+a+trY1V7uvr29LS4uzsXFdXR5XU1NRQ93rrP14gEAjoTZlMFhsbq78q\nZ6yc5uXlRY/FPStzvl/HFi5c+PXXX5eVlR07diw9PX3atGn0LlZCwHzNmEAgoC9SXr16tcNR\n6uvr3dzcum/WXNzd3QGAmRZz0Gg0AoHA1dW1EwMNHDjw5ZdfXrt2rY+Pz4ABA5i7wsPDs7Oz\nWfWlUmlERAS9Sa/G7ty5k7UOm5qaOmnSpDt37ixdulR/XLNiCQCM3fGWmZlp7GZE1qFnBjnH\noWdusv69oJB6gHHl2JJBApjYIYSQrbKzs8vLy2MVUqelsLCwdevWKRSKR48epaWlUe8YY54m\n6XvM6YanTp0y+NoLY+W0oKCgmzdvmjIrM79fBx4/fkzduqdQKE6dOkWt8zY1NQHjm2q12nPn\nzv31r3+lW40ZMyYzM7O5ubmiouKDDz7Q71YkEjHf63bjxg1/f//unbkxYrHYwcHBlFfTtbe3\nkyTZr1+/Di81GTNr1qxPPvlE/3nYlStXZmRkbN68uaamRqPRPHjwYOPGjampqStXrqTrBAQE\nKBSKM2fO6HS6kSNHMpvn5+dzXJAzN5ZYLzdRKBTFxcVRUVGBgYHGnu/mCHKOQ798+fKysjKd\nTqf/74UpLBkkAIbSzJ5TUFAQHBw8c+ZMSw6KEEKmmzlzZnBwcHBwsFwut/ZcuurQoUP62ZJA\nIMjNzZXL5XPnznVxcXF1dU1KSqqvr6ea5ObmUk2oaszevL297927pz+KsXLa6tWr09PTTZkV\nyXjmlAWefSLV4F4W+ru4uLhMmzatsLAwPj7ewcGBudfOzm7MmDH5+fl0D5cvXw4ICBAKhSNG\njJBKpdSdUswhUlJSqPVNajMjIyM1NZXj63evp0+fXrx4sbi4+LJx//73v8+fP//DDz+o1Wpz\n+6e/V2lpKQCUlpayykmSrKioSEhIEIvFAoFALBYnJCRUVFSw+vnqq6+8vb23b9/OKjeYaNJ7\nzY0lCn2gnZ2dQ0JCqCdeo6Oj9WfOqs8Kco5Dv2XLFh8fHx6Pp//vhTHMcS0cJPjwBEII/cqG\nH56wlrKyssTExOvXr1t7Ij0iICBg3759Y8aMsdiIVVVV9+/f12q11E+HsfZqNBqdTufg4DB8\n+PDO3WDXm1klljr3cDGThYMEl2IRQgj1oMDAQE9Pz6KiImtPpPsVFhZ6eXlZMqsDAIlEMmzY\nMEdHR51Op1Kp1Gq1RqPRaDRqtZp6VZunp+fo0aNtL6uD5zOWLB8kmNghhBDqWR999JGFfy7T\nMqglNsuP6+HhERgY6Ovr6+bmZmdnx+PxeDyevb19//79hw8fPnz4cHt7e8vPyjKeu1iyfJCY\n9JNzCCGEUKeFh4efOXPG2rPoflb8Unw+39vb29vbGwDUajWfzzf3RymeU5aPpS6uw1o+SDCx\nQwghhJ5jpr+yGL0IcCkWIYQQQshGYGKHEEIIIWQjMLFDCCGEELIRmNghhBBCCNkITOwQQggh\nhGwEJnYIIYQQQjYCEzuEEEIIIRuBiR1CCCGEkI3AxA4hhBB6nmm11p4B6kXwlycQQgih541a\nDfn5cPo0/PAD1NaCUAgSCUyYANHREBRk7ckha8LEDiGEEHquXLwIf/87/PgjKJUgEICdHZAk\nPHwI//437NoFM2ZAaiq4u1t7lsg6MLFDCCGEnh85ObB6NdTUgJsbeHk9s0ung9pa2LcPfvwR\nvvoKfHysNEVkTXiPHUII2SaCIMaPH88qHD9+PEEQdAVjDWUy2YwZMwzujYmJOXr0KLNErVaL\nxWL9msbKaadPnw4PD2eWSKVSoVBIEARBEEKhUCqVcjTvOfQ0WHMICws7d+6cVab0i3Pn4KOP\n4MkTGDgQXFzYe/l88PICDw/44Qf4059AoTC3e0KP/lHgPkYcQUV/1g8MY39wY+XG+mHptQHG\nJJPJ7Ox+ucrWPQFGWlBBQUFwcPDMmTMtOShCCJlu5syZwcHBwcHBcrnc2nPpKgBwc3NjfpGH\nDx+6u7vT/+U3dgoAAK1WO2jQoEePHrF2yeXyQYMGabVaarO+vr6kpGTOnDmsroyVs0yZMqWw\nsJDezM3NdXR0zM3NNbhpMbm5uQ4ODocOHdKfQ2FhYWRkpIXn86uWFjI8nPT0JIOCyOBgrn/8\n/Ulvb3LtWnNH0D9eubm5AoGAucl9jDiCijQSGMxOWJ+NHYjnOsCYSkpKwsLC6G/RLQGGiR1C\nCP3KxhK7hQsXfvnll3TJF198sWjRIlMSO5IkU1JS/vGPf7B2rVu37sMPP2TWNHiZwJTLB2Vl\nZQEBAfSmwbOsVU69AoFg3759zDkwM5tRo0aVlZVZcj6/2r2bHDCA9PfvIKuj/vH2JkePJvVS\nc24GjxddaMox4g4qg4EhEAiYfdJ/cI4D8VwHGK28vDwkJKSmpob5LboeYLgUixBCNmvWrFlH\njhyhN48cOTJr1iwT2yYnJ+/evZtVuHv37uTkZHqTPpewqhkrZ8rJyVmwYAG9mZCQkJWVFRcX\nx6wTFxeXlZWVkJBAbRIEkZeXFxgYSC9dGVuqYy0IMlef8/Pzg4KChEKhsZU4jUYTHx9Pb44f\nP97Dw4PejI+Pt9r6XUEBtLVBnz4mVXZ3h7o6OHOmi2M+ffrUzc2N+mzKMeJmMDA0Gg2zz7i4\nOI1GA5wH4rkOMIpcLl+yZMn+/fv79evHLO+GAOtKVmguvGKHEOrlbOyKXWNjo7Ozc1NTE0mS\nCoXC2dlZoVCAaVfsSJKMiIi4dOkSXV5cXGxsnajDrvRFREScPXvWlJrMOsHBwVeuXNFoNKTx\nJTz93phfediwYd99951Wq+3wao1Sqbx48WJoaOimTZvowqKiot/97ncdTrX7tbWRI0eSAwea\ndLmO+sfTk/z//j+zBmH+3TQazb179957773Fixfr7zWlB45yeDY5465s8EB0OKXeHGANDQ0T\nJkz4/vvv9XvreoBhYocQQr+yscSOJMlp06bl5OSQJHnw4EHqhGF6Yrdv37633nqLLl+6dOmB\nAwe4m5hYTpKkRCJh3sOnf6bUvwYBANevX6frGFvCM9gb/eHo0aMGm+jPnHbhwgW6/MGDBwMG\nDDD2pXqQXE4OG0b6+JiX2C1datYgBi8AnTt3jt7buR70LySZntgZOxAdTqnXBphKpZo6dWpB\nQYHBuXU9wHApFiGEbBm9Gqu/Duvg4GCwiaOjI/Vhzpw5BQUFKpUKAFpaWoqKilgrWV1RX1/v\n6upqbC99llIqlczyUaNG0Z+NLeFxCw0NNaUJSZL19fXnzp2bPHnyGcaCpoeHR21tbYejdD8e\nDwgCODMnA/h8c8dh/uV/+umnP//5z3/961871wOTudNg9mbwQHSo1wZYUlLS4sWLp06darB5\n1wMMEzuEELJlsbGxx48fV6lU+fn5sbGxzF1isbi6uppVv7a21uu/b0cTiUQxMTHUHT85OTmz\nZ88WCoXdNTHq0Vp6UyAQGLy1qLy8nLlp7G0aHFhnbpFIxF2fHsLV1TU0NDQrK2vnzp3MvTye\nNU6dnp4gFIIJicUzWC+6M4eDg4O/v39aWlpZWRlVYuwYUTeidXogY7gPRId6bYDl5OQkJSUx\n3ykDz96l18UAw8QOIYRs2cCBA19++eW1a9f6+PgMGDCAuSs8PDw7O5tVXyqVRkRE0JvLli3b\ntWsXAOzcufPNN9/sxol5eXnV1dXRm9nZ2YmJifqn3szMTL6Ry06sU7VUKhUIBPQutVpNfb56\n9SqzFXOT2cRYt1euXLl79y69WV9fTz9MYFF2dhASAm1tptZXKkEkguDgLg7r7u7u8t8X5hk8\nRlKpNDExUT+QTGfsOHIfiA712gAzeDmTzkG7IcC6so5rLrzHDiHUy9nePXYkSa5Zs0YoFKan\np7PKKyoq+vXr989//rO6ulqtVt+/f/+zzz4Ti8U//fQTs5/f/OY3p0+fnjhxoiljmVhOkmRM\nTExRURGzhHmzeWNj4/nz56dPn75+/Xp/f3+DvXHc2z5u3Li1a9c2NTWVl5e/+uqrwLgFKiAg\noLS0lOPedo5uSZI8ffr0lClTOP4UPejgQXLQIHLYMJNusBOLybFjydpas0Zg/oU1Gs1PP/30\n3nvvpaam0oVdfI+dwU2O99hxHAju4cheHGDcf5yuBxgmdggh9CubTOxKS0sBoLS0lFVOkmRF\nRUVCQoJYLBYIBGKxOCEhoaKigtXPV1995e3tvX37dlPGMrGcJMnVq1fTuSaNutkcAJydnUNC\nQrKzs0mSjI6ONtYbXZ91n/vly5cDAgKEQuGIESOkUilBEHQPW7Zs8fHx4fF4rCamdEuSZEZG\nBjPRsSi1moyKIj09ycDADrK6oUNJiYT8f//P3BGY1334fL6Pj8+aNWt0Oh2zDv3HMfj36URi\nRxr/g3McCO7hyN4dYBzfousBRpBduKXRXCdPnkxJSZFIJDKZzGKDIoSQ6WJjY6uqqgBAJpNJ\nJBJrT8eWlZWVJSYmXr9+3ZKDUifg/7+9e42OqszzPf6vJFUmwaJByKUDeEgUEZCEISdgiyEB\nB6XJzeg0YSRwxGC3PTqrz5lxrTMcUSSyosv20vRSxPEC2AkYgykT1gIG0oncRBZREzOd6Iz0\ncEswTEO4JYRKwj4vqnuze9clFS5VlcfvZ/Gi9rOfW1We7v1zPzuVa+khOTm5rKxs8uTJ12tK\nA/P117J4sbS2SkKCuO3x/dm5c3LmjPzt38r770t/z3sp7Ae7wHjGDgAQBCkpKSNHjqytrQ32\nRAagpqYmNjY2aKlORJKT5bXX5NZbpa1NTp2Svr6/OnvpkrS2yoULkpkpv/nNDznVyQ94gRHs\nAADBsWLFiuLi4mDPYgBc22RBnsTs2bJxo8ybJ2Fh0toqR47I8eNy7JgcPix/+pOMHCm/+pWs\nX38tvw+rjB/mAou4LlMBAGCgMjMzB/TNZNfuGrfJAjxbr8aPl/Xr5cABqauTpiY5cUIiI2XM\nGLn7brn/fhk9OtjzCxU/zAVGsAMAYLCxWOTuu+Xuu4M9D4QctmIBAAAUQbADAABQBMEOAABA\nEQQ7AAAARRDsAAAAFEGwAwAAUATBDgAAQBEEOwAAAEUQ7AAAABTBX54AAGCQ0kT+W+SkSJRI\njMjQYM8HwUewAwBg0GkT+UBkp8hxkV4Ri4hNZKpIlsjDIrZgTw9BQ7ADAGAQ0UQ+EPmNSJuI\niESLRIhoIhdEtovsEvmdyCsiE4M8TQQJwQ4AgEGkRORdkS6ROE935s6JHBRZKPKmyD1BmB2C\njV+eAAA1ORwOm81m+Ws2m83hcJhqWiwW46HT6YyLizNVMAkPD3cfsa6uLjMz09scPA4dGPo0\nTHPIyMjYvXt3UKZ0tcpE3hNxioz2st86VCRB5LjIP4scHWjvFoslLS3NVJiWlqavENNSMTas\nrq7Ozs72eDYrK2vLli3GEvc1Jt5/TEYhu8aMqqurIyL+fOMsCGtMC6AdO3akpqbm5OQEclAA\n8F9OTk5qampqampra2uw53KtrFZrZWWlqbCystJqtZoK9WtBR0fHvn37HnroIR9Xh8OHD8fF\nxTkcDvdTs2fPrqmpMY4VHR2tz8F0GDCVlZVRUVGbN292n0NNTc19990X4Plcg5OaNlXTYjRt\nqqal+vx3l6bFaNr/HugAIjJ8+HDj4j9+/Pgtt9yirwdvC0NEent7x4wZc+LECdOp1tbWMWPG\n9Pb2ug69rTHjj8bHUgnNNWa0b9++jIwM/d0Ffo0R7ADgCpWCnYhcvnzZY7m3En/+sz8zM3P1\n6tXu5Y2NjcnJyfqhx0tsUK67Vqu1rKzMOAdjtJ00aVJjY2Mg53MNVmtanKZN6i/Vuf6N0bQ7\nNO0/BjSAiDzyyCNvvfWWXrJmzZqFCxf6E+w0TVu2bNnLL79sOvXiiy8+88wzxpoe15jpv0M8\n/hdIyK4xXXNz8/Tp00+ePGl8dwFeY2zFAoCaEhIStm/fPqAm+rXBW4WPPvqotbX1qaeecj9V\nUVGxYMEC/bCgoKC0tDQ/P99YJz8/v7S0tKCgwHVosVg++eSTlJQUfd/K22acaQfQuDO4bdu2\nKVOm2Gw2b9twPT098+fP1w/T0tJGjBihH86fPz8UNu/8UyNyWSTSv8rDRM6KfDrQMfLy8qqq\nqvTDqqqqvLw8P9sWFRVt2LDBVLhhw4aioiL90Nsa6+npMa6W7Ozsnp4eU52QXWMura2tjz76\n6MaNG2NiYozlgV5jAYuQGnfsAIQ8le7Y1dTUJCUlZWRkPPvss1VVVcePH/dW0/1a4PHq0Nvb\nm5SU9O6773rsZNasWbt27fLdg/soqamp9fX1PT09ms/NOFNvYriBNG7cuAMHDvT29vZ7q6ar\nq2v//v3p6enGO461tbUPPPBAv1MNAZ2adqemjfHvdp3rX4ym/fOAxhCRs2fP2u328+fPa5p2\n7tw5u91+7tw58e+OnaZps2bN+vzzz/XyvXv3etuIFJ937Orr693HCuU1dubMmXvuuefLL790\n7y3Aa4xgBwBXqBTsNE3r6+vbu3fvc889N3PmzJtuuik9Pf3IkSPu1fwMdhs3bhw9evSlS5c8\njpWQkGB8vsr9Mul+T0FEmpqa9Do+NuN8XHS3bNnisYn7O9J99tlnevmxY8dGjRrlsUmIOapp\n4zQtcSDBboSmLR7QGK4P9v7776+oqNA07aOPPnIlEv+DXVlZ2RNPPKGXP/bYYx9++KHvJi56\nZnI6nfv3709JSYmKijI1Cdk11t3dPWfOnB07dnicW4DXGMEOAK5QLNgZXbhwoby8PDs7++OP\nPzad8jPYzZgx47XXXvPWf1RU1MWLF333oGlaV1eX8ZJpfArQxzR8XHTPnDnT78xdOjo6du/e\nPXPmzJKSEuN8IiMjvTUJJX/StDs07X8MJNiN1LRfDmgM16f35ptvFhYWapq2cOHCNWvWaIZP\n1T1suURHR7teXLx4MSkpybUSLly4MHbsWG//JeD+k3JlpvDw8DvvvPOVV1659dZbTRVCdo39\n7Gc/+93vfuetToDXGM/YAYCCNE2zWCx9fX16yZAhQ+bPn7969erFixcbHyfq6+sLC+v/WnDo\n0KGGhoalS5d6q+D6tUf90Gq1enyuqLm52Xjo7eszfHBdtnWRkf08c6YPMWzYsPT09NLS0nXr\n1hnP+vP2Q8BwkZtFnH7Xd/0sfnwVI+Xm5m7durW7u3vbtm25ubnGU3Fxce3t7ab6p06dio2N\ndb2OjIzMyspy/egrKioefPBBm83fP4ORn5/vdDp7e3tbWlpuv/32qVOnmiqE7BqrqKhYtGiR\n8fuA5K+f0gvkGhsUqxkAMDCu69+lS5dM5UlJSZ2dnXa7/fTp066SkydPmp719qi6ujo3N9du\nt3urEBsbq/cpIuXl5YWFhe7X3VdffdXjd+CJ23Xa4XBYrVb9lNP550zzxRdfGFsZD41NvHVb\nX19/+PBh/bCjo2P48OHe3lQoCRPJEOn5S2Lr1wWRaJGfXMVIo0ePHjt2bElJSWJi4qhRo4yn\nMjMzy8vLTfUdDsesWbP0w6VLl65fv15E1q1b9/jjj/s5qM1mKy0t1Q83bdqUk5NjqhOya8z9\ntpn85X+DEvg1FrB7gxpbsQBCnjJbsX19fREREcav+XBx7XZlZ2c//fTTZ8+ebWtrKyoqWrJk\niama+9UhKyvrgw8+8DFiVlZWbW2taSz9SfOzZ8/u2bNn7ty5L7300vjx4z2O4uPB9qlTp5aU\nlJw/f765uXnatGli2CZLTk5uaGjw8WC77y9Iq6urmz17to/3FUp+r2mJmjbW733YDE3rGtAA\n+ge7cuVKm81WXFxsKm9paYmJiXnjjTfa29udTufRo0dff/31uLi4b775xtjP3XffXVdXN2PG\nDH/GctG/brC9vf2FF16YOHGi0+k0NQnZNeb73QV4jRHsAOAKZYKdpmmbN2/2ePuqsrKytbX1\n4YcfHjp06LBhwxYtWtTR0WFq6x7s4uPjPf7ihe7555/Xc4DOlSNFxG63T58+vby8XNO0efPm\neRtFr296yP3gwYPJyck2m23ChAkOh8Niseg9rF27NjExMSwszOMXMvvuVtO0VatWLV++3Mf7\nCiWXNe1/aVqMH19lN0bTxmpaxUAH0H8iDQ0NItLQ0GAq1zStpaWloKAgLi7OarXGxcUVFBS0\ntLSY+nn77bfj4+Pfe+89f8bSuX5Mdrs9Ly/v2LFj7k1CeY35eHcBXmMWzfv3FV13O3fuXLZs\nWUJCQnV1dcAGBQD/5ebmtrW1iUh1dXVCQkKwpzOYNDY2FhYWNjU1BXJQ19X3WnpITk4uKyub\nPHny9ZrSDXZC5O9F/iAyUmSIpwqayH+L9IkUirwkMuDHy0IZa8wfPGMHALgOUlJSRo4cWVtb\nG+yJDEBNTU1sbOzgSXUi8mORd0RSRTpEWkWMz/j3iZwVOSoSIbJYZKViqU5YY/4h2AEAro8V\nK1YUFxcHexYD4NojC/YsBmqcyIci/0ckRuScyBGRoyJHRVpFekWmiPxGpMTvP1AxyLDG+hUR\nyMEAAArLzMz89NNPAzniNe6RBXi2188wkf8rslRkt0izyH+L3CQSLzJd5H+K+Pv1IoMRa6xf\nBDsAAAajESL5Ivn9V8QPCVuxAAAAiiDYAQAAKIJgBwAAoAiCHQAAgCIIdgAAAIog2AEAACiC\nYAcAAKAIgh0AAIAiCHYAAACK4C9PAAAwKJ2Uk9/INyflZKRExkpsiqTcJDcFe1IIMoIdAACD\nzB7Z8468c1AOdklXj/SESZhVrCNkxDyZ94Q8kSAJwZ4ggoatWAAABo1u6V4my5bIkm2yrVM6\noyV6hIwYJsOsYm2V1rWyNluyt8iWYE8TQcMdOwAABgenOP9B/mGbbIuQiDEyJsxwdyZSIofK\nUKc4j8vxf5Z/viAX/l7+PohTRbBwxw4A1ORwOGw2m+Wv2Ww2h8NhqmmxWNxbudd0Op1xcXGm\nhibh4eHuM6mrq8vMzPQ2N49TCgxvbzYjI2P37t1BmZJvr8lr/yb/FiVRcRIX5ukKbhPbaBl9\nRs6sklVfypcD7d9isaSlpZkK09LS9BViXCqmhtXV1dnZ2R7PZmVlbdny55uI3j7zwb6WjKqr\nqyMi/nzjLAhrSQugHTt2pKam5uTkBHJQAPBfTk5Oampqampqa2trsOdyraxWa2VlpamwsrLS\narWaCvVrQWVlZXR0tKuV8XVHR8e+ffseeughH1eNw4cPx8XFORwO91OzZ8+uqakxzkHv2f0w\nYCorK6OiojZv3uw+h5qamvvuuy/A8+nXf2j/MUGb8GPtx6laqu9/U7QpI7QRC7QFfVrfgIYQ\nkeHDhxsX//Hjx2+55Rb95+5tAYhIb2/vmDFjTpw4YTrV2to6ZsyY3t5ezednbjTo1pLRvn37\nMjIy9A8q8GuJYAcAV6gU7ETk8uXLHsu9lZiyoJ4C/bkdkJmZuXr1avfyxsbG5ORkY5/ul96g\nXI+tVmtZWZlxDsbIO2nSpMbGxkDOp18vaC/EaDEpWkq/wS5VS03UEpO0pC+0LwY0hIg88sgj\nb731ll6yZs2ahQsX+hPsNE1btmzZyy+/bDr14osvPvPMM67Xvj9z3aBbS7rm5ubp06efPHnS\n+EEFeC2xFQsAakpISNi+ffuAmvT09OTn5+uH2dnZPT09YrhEeWv40Ucftba2PvXUU+6nKioq\nFixYoB8WFBSUlpYaRxGR/Pz80tLSgoIC16HFYvnkk09SUlL0/Swf+3fGfow7htu2bZsyZYrN\nZvO2PdfT0zN//nz9MC0tbcSIEfrh/PnzQ2FTz6hGaixiifDv4fihMvSCXNgtA94EzMvLq6qq\n0g+rqqry8vL8bFtUVLRhwwZT4YYNG4qKilyvfX/mLoNxLbm0trY++uijGzdujImJMZYHei0F\nLEJq3LEDEPJUumNXU1OTlJSUkZHx7LPPVlVVHT9+3FtN8XLHrr6+3nSZ8HjV6O3tTUpKevfd\ndz12PmvWrF27dvnuwX0+qamp9fX1PT09mvcNYvfexHBjady4cQcOHOjt7e33Fk5XV9f+/fvT\n09ONd4lqa2sfeOCBfqcaMGe1s+O18bdqt/pzu871L0aL+ZX2qwGNIiJnz5612+3nz5/XNO3c\nuXN2u/3cuXPi3x07TdNmzZr1+eef6+V79+71uBHp8TPXBvNaOnPmzD333PPll1+69xbgtUSw\nA4ArVAp2mqb19fXt3bv3ueeemzlz5k033ZSenn7kyBH3auL2jJ3T6dy/f39KSkpUVJTHmkYb\nN24cPXr0pUuXPM4hISHB+NyV++XT/V6DiDQ1Nel1vG0Qe+xNf7FlyxaPTdzfu+6zzz7Ty48d\nOzZq1CiPTYLiiHbkdu32RC3R/2A3Uhu5SFs0oFFcH+D9999fUVGhadpHH33kSiT+B7uysrIn\nnnhCL3/sscc+/PCD5CKSAAAc8klEQVRD98oeP3Nt0K6l7u7uOXPm7Nixw+PcAryW2IoFAGWF\nhYXNmDFj5cqVu3btOnXq1FNPPfXkk09WVlZ6q69vY0VFRS1ZsmTRokWmTSWP3nzzzX/6p3+y\n2Wwez3Z0dAwbNsxbW/1q1NXVZSyfNGmS/tq0QZyfn+/aIPYtPT3dnyaapnV0dOzevXvmzJmf\nfvqpXj5ixIhTp071O0rA/Eh+FC7hl+Wy/00sYhkmXj95H/TdWPd92KioKI9NoqOjXS8eeuih\nHTt2dHd3i0hnZ2dtba1pq1S8f+YyaNfSokWLFi9ePGfOHI/NA7yWCHYAoCBN0ywWS19fn14y\nZMiQ+fPnr169evHixcbHifr6+sLCrlwL8vPznU5nb29vS0vL7bffPnXqVN8DHTp0qKGhYenS\npd4quH4dUj+0Wq0enzdqbm42Hnr7Wg0fTJfzyMhI3/X1IYYNG5aenl5aWrpu3TrjWePHEnRD\nZehIGdkt3X7W75M+EblVbr2KsXJzc7du3drd3b1t27bc3Fzjqbi4uPb2dlP9U6dOxcbGul5H\nRkZmZWW5fsQVFRUPPvigMaX5/swH71qqqKhYtGiR8bta5K+f0gvkWgqhVQsAuF5c179Lly6Z\nypOSkjo7O+12++nTp10lJ0+e1G/L2Wy20tJSvfKmTZtycnJ8D1RdXZ2bm2u3271ViI2N1ccS\nkfLy8sLCQvfr8auvvurxe8vE7frtcDisVqt+yul0ul5/8cUXxlbGQ2MTb93W19cfPnxYP+zo\n6Bg+fLi3NxV4FrHMltl90udKbP06J+dulpvvlXuvYqzRo0ePHTu2pKQkMTFx1KhRxlOZmZnl\n5eWm+g6HY9asWfrh0qVL169fLyLr1q17/PHHjTV9f+aDdy2574fKX/43KIFfSzd0o9eEZ+wA\nhDhlnrHr6+uLiIgwfrWEi+sJoezs7Keffvrs2bNtbW1FRUVLlizRz7q+Zqy9vf2FF16YOHGi\n0+k0Nne/amRlZX3wwQc+ZpKVlVVbW2uag/4E+tmzZ/fs2TN37tyXXnpp/PjxHkfx8cD71KlT\nS0pKzp8/39zcPG3aNDE8F5WcnNzQ0ODjgXcf3WqaVldXN3v2bB/vK/Aatcbx2vhR2qh+n677\nG+1vRmgj8rS8Xq13QEPoH+DKlSttNltxcbGpvKWlJSYm5o033mhvb3c6nUePHn399dfj4uK+\n+eYbYz933313XV3djBkzTP37/swH71pyZxw3wGuJYAcAVygT7DRN27x5s8fbVJWVla2trQ8/\n/PDQoUOHDRu2aNGijo4OvZUr+dnt9ry8vGPHjpn6dA928fHxHn8hQ/f888/r+cA0iojY7fbp\n06eXl5drmjZv3jxvo+j1TQ+/Hzx4MDk52WazTZgwweFwWCwWvYe1a9cmJiaGhYV5/KJm391q\nmrZq1arly5f7eF9B8f+0/xevxY/TxvlIdVO1qXFa3Hht/F5t70D71z/5hoYGEWloaDCVa5rW\n0tJSUFAQFxdntVrj4uIKCgpaWlpM/bz99tvx8fHvvfee+xA+PvNBvZZMjOMGeC1ZNO/fS3Td\n7dy5c9myZQkJCdXV1QEbFAD8l5ub29bWJiLV1dUJCQnBno4KGhsbCwsLm5qaAjmo66p8LT0k\nJyeXlZVNnjz5ek3puuiSrkfl0V2yK1qib5FbLGJ+eqxXek/IiZvl5qfl6V/KL4MyyRuHteQP\nnrEDANxAKSkpI0eOrK2tDfZEBqCmpiY2NjbUUp2IREv0O/JOvuRroh2Vo6fldLd090lfr/R2\nSme7tLdK60gZuUJWqJfqhLXkH4IdAODGWrFiRXFxcbBnMQCuvbNgz8KzH8mP1sia38pv0yQt\nXMI7pKNN2r6X7zulc6gMXSALKqVysSwO9jRvFNZSv/z6yyQAAFy1zMxM09eV3WjXuHcW4NkO\nVJiE5UpuruR+K9/+Qf7wJ/nTTXJTvMSnSdotckuwZ3djsZb6RbADAGBQGi/jx8v4YM8CoYWt\nWAAAAEUQ7AAAABRBsAMAAFAEwQ4AAEARBDsAAABFEOwAAAAUQbADAABQBMEOAABAEQQ7AAAA\nRRDsAAAAFEGwAwAAUATBDgAAQBEEOwAAAEUQ7ABATQ6Hw2azWf6azWZzOBymmhaLRX/9xz/+\nMScnZ+jQoe41nU5nXFycqaFJeHi4+0zq6uoyMzO9zc3jlAJDn4ZpDhkZGbt37w7KlILr6taM\nx4/R29rw9pmbhOyaMaquro6IiHC9DqE1owXQjh07UlNTc3JyAjkoAPgvJycnNTU1NTW1tbU1\n2HO5VlartbKy0lRYWVlptVpNhfq1oLOz87bbbnv11VfPnTtXWVkZHR3t6qGjo2Pfvn0PPfSQ\nj6vG4cOH4+LiHA6H+6nZs2fX1NQY56D37H4YMJWVlVFRUZs3b3afQ01NzX333Rfg+YSCq1gz\nxo/O249SXxv+VHYJzTVjtG/fvoyMDP1zCJ01Q7ADgCtUCnYicvnyZY/l3kp++9vfLl68WC/X\nr+j+3A7IzMxcvXq1e3ljY2NycrKxT/dLclCu01artayszDgHY3yZNGlSY2NjIOcTCq5izZiy\noMcUqK8NfyprIbxmdM3NzdOnTz958qTxkwmRNUOwA4ArVAp2CQkJW7dudS/3cZGeO3furl27\nfPTpLdiVl5ePGzeur6/P/dTy5ctLSkr0Q4/3hLS/vsaLiMPhSE5ODg8PN54VEVNz03z0QxHZ\nunVrSkqK1Wr1NqKI9PT06IfHjh2Lj4/XD1euXPn88897fLMKu4o1YzrldDpNJca14aMfo5Bd\nMy7Hjx+fNm3aoUOHTL2FyJoh2AHAFSoFu5qamqSkpIyMjGeffbaqqur48ePeauoXp4SEhDff\nfHPixIkREREer20eL8O9vb1JSUnvvvuux85nzZplDIv+PAIkIqmpqfX19a7g5WP/zsdFety4\ncQcOHOjt7e331k5XV9f+/fvT09ONdxxra2sfeOCBfqeqmKtYM6Z1Ul9fb/yhmNaGn8EulNfM\nmTNn7rnnni+//NK9txBZMwQ7ALhCpWCnaVpfX9/evXufe+65mTNn3nTTTenp6UeOHHGvpl+c\nIiMj77rrrs8++8zpdHq8tnm8xG7cuHH06NGXLl3yOIeEhIQTJ05468HjY98i0tTUpNfxsX/n\n4yK9ZcsWj03c37vus88+08uPHTs2atQoj03UNtA1o68Tp9O5f//+lJSUqKgovZppbfgZ7EJ2\nzXR3d8+ZM2fHjh0e5xYia4ZgBwBXKBbsjC5cuFBeXp6dnf3xxx+bThmvbQcPHtTL3a9tHi/D\nM2bMeO2117yNGxUVdfHiRd89aJrW1dVlnIbxSS//9wGN5WfOnOl35i4dHR27d++eOXOmcfuv\nq6srMjLSW5MfCH/WjPaXdRIeHn7nnXe+8sort956q37KtDb8DHYhu2Z+9rOf/e53v/NWJ0TW\nDMEOAK5QJthdvnxZRHp7e03lhw4dGjJkiGmzLCwszPXa9NiZ5v0qqPvuu++GDBly7tw5bzOx\nWq1dXV3GQ487XMYtvH4H9Vizs7PTWN7d3e1nJy5Hjx4dN26cfnjx4sXo6Ghvb0pJV7dmTD75\n5JMHH3zQ9dp9bfgZ7EJ5zXjkOhsia4bvsQMABWmaJiKXLl0ylSclJXV2dtrt9tOnT7tKTp48\nGRMT43pttVq3bNmiV3Y4HK7Hz32orq7Ozc212+3eKsTGxupjiUh5eXlhYaH7l5C9+uqrHr8D\nzzUrY33jrKxWq+tpfRH54osvjK2Mhx7fiKnb+vr6w4cP64cdHR3Dhw/39qaUdHVrxmazlZaW\n6pU3bdqUk5Pjeu2+Nnz8KI1Cds24pyj5y+cmobNmApkiuWMHIMQpc8eur68vIiLC+HUeLq5d\ns+zs7Keffvrs2bNtbW1FRUVLlizRz/r+mjH3q0ZWVtYHH3zgYyZZWVm1tbWmOeg9nz17ds+e\nPXPnzn3ppZfGjx/vcRQfs5o6dWpJScn58+ebm5unTZsmhrsvycnJDQ0NPh6E9/1m6+rqZs+e\n7eN9qeeq14zr6wDb29tfeOGFiRMnOp1O1yn3teHn99iF7JpxZxw3RNYMwQ4ArlAm2Gmatnnz\nZo+3qSorK1tbWx9++OGhQ4cOGzZs0aJFHR0deitvXxLh4h7s4uPjPT5cr3v++eeLi4tNhfoo\ndrt9+vTp5eXlmqbNmzfP2yjeZnXw4MHk5GSbzTZhwgSHw2GxWPQe1q5dm5iYGBYW5uOrK3y8\n2VWrVi1fvtzH+1LStawZu92el5d37Ngxvdzj2vC9wFxCec2YGMcNkTVj0bzvGV93O3fuXLZs\nWUJCQnV1dcAGBQD/5ebmtrW1iUh1dXVCQkKwp6OCxsbGwsLCpqamQA7qulpfSw/JycllZWWT\nJ0++XlOC/1gz14Jn7AAAN1BKSsrIkSNra2uDPZEBqKmpiY2NDfoV+geLNXMtCHYAgBtrxYoV\nxcXFwZ7FALj21II9ix801sxViwj2BAAAisvMzPz0008DOeI17qkFeLZwx5q5atyxAwAAUATB\nDgAAQBEEOwAAAEUQ7AAAABRBsAMAAFAEwQ4AAEARBDsAAABFEOwAAAAUQbADAABQBMEOAABA\nEQQ7AAAARRDsAAAAFEGwAwAAUATBDgAAQBEEOwBQk8ViSUtLMxWmpaVZLBa9greG1dXV2dnZ\nHs9mZWVt2bLF9fqTTz6x2WwWi8VmszkcDmM1h8Ph7ZSurq4uMzPTYyvfDW80b5PPyMjYvXt3\nUKYE+IlgBwDKOnToUFtbm37Y2tr6xz/+0Z+GWVlZX3/99ffff28qb2tra2pqmjdvnohUVVUt\nXLjwww8/1DStvLy8sLBQz0AOh6OwsLC8vNz9lNGqVaueffZZ/dDYynfDG8rhcCxcuHDTpk3u\nc3juueeKi4sDPB9gYLQA2rFjR2pqak5OTiAHBQD/5eTkpKampqamtra2Bnsu10pEHnnkkbfe\neksvWbNmzcKFC/X/5/d2CXCVL1u27OWXXzadevHFF5955hnX68jISFeqc6msrLRara7XVqu1\nsrLS4yldY2NjcnKysU50dLSxlbfCG81qtZaVlRnnYJz8pEmTGhsbAzkfYEC4YwcAysrLy6uq\nqtIPq6qq8vLy/GxbVFS0YcMGU+GGDRuKiopcr7u7u//u7/5OP5Wfn9/T0+N63dPTk5+fr5/K\nzs7WT+kqKioWLFigHxYUFJSWlhpbufosLS0tKChwHVoslk8++SQlJSUiIsJV4m3P1LTLbNx9\n3rZt25QpU2w2m7et3p6envnz5+uHaWlpI0aM0A/nz58frA1iwB8EOwBQ1ty5c/ft23fhwgUR\nOX/+/GeffTZ37lw/2952222xsbEHDhzQS/bt2zdq1KjExETXoaZp4eHh+tkPP/zwjjvucL22\nWq3G9PP111+7979v374ZM2boh6YsqMvPz3c6nfrhqlWr3n///e7ubvF7w9fkV7/61b/+679e\nvHjRWxNN01zB8eLFi59//vkjjzyybNky/Wx6evr+/fv7HQUIFoIdAChr6NChP/nJT7Zv3y4i\n27dvv+eee+x2u//Nly5dun79ev3w/ffff/zxx92rXbx4sbi4+Be/+MUHH3zgKtEzU09Pz+ef\nf15UVBQVFWVq9e233+pB0J3lr+nl69evT01NdQUv400+0709H1577bVp06aFh4f7bmKxWKKj\no3/yk5/s2bPH+Dso48aN+/d///d+RwGChWAHACrTd2Pd92Hdw5ZLdHS068VDDz20Y8cO1+2x\nzs7O2tpa95tqpaWld9xxx65du7766qvp06e7CvXMFBUVtWTJkkWLFsXExJgadnR0DBs2zNu0\n9QeGurq6jOWTJk3SX5tu8hn3gn1IT0/3p4mmaR0dHbt37545c+ann36ql48YMeLUqVP9jgIE\nC8EOAFSWm5u7devW7u7ubdu25ebmGk/FxcW1t7eb6p86dSo2Ntb1OjIyMisry7VZWVFR8eCD\nD9psNr1mZ2fnT3/602effXblypW///3vk5KSjP24tlB7e3tbWlpuv/32qVOnmgbq7e3VNE0/\nNO3e6pqbm42H3r6ixQdTNIyMjPRdXx9i2LBh6enppaWl69atM54NC+PSidDF6gQAlY0ePXrs\n2LElJSWJiYmjRo0ynsrMzCwvLzfVdzgcs2bN0g/13dh169aZ9mGXL19+7733/td//ddjjz1m\n6sRms5WWluqHmzZtysnJMdWJjY09ffq0fujtibdXX33V+CSfkSkLOhwOq9Wqn9KfzPviiy+M\nrYyHxibeuq2vrz98+LB+2NHRMXz4cI/zAUJCIH8Fl687ARDiFPu6E9eLlStX2my24uJiU3lL\nS0tMTMwbb7zR3t7udDqPHj36+uuvx8XFffPNN8Z+7r777rq6uhkzZpj6Hz9+/Pfff+9x6MrK\nyqioqM2bN7e3t7/wwgsTJ050Op2mOllZWbW1taZW+pebnD17ds+ePXPnzn3ppZfGjx9vmrl7\nfdMXo0ydOrWkpOT8+fPNzc3Tpk0Twze8JCcnNzQ09Pb2evsuFR/dappWV1c3e/Zsj+8aCAUE\nOwC4Qslg19DQICINDQ2mck3TWlpaCgoK4uLirFZrXFxcQUFBS0uLqZ+33347Pj7+vffeM5V7\n3NPUz7q+/s1ut+fl5R07dsx9es8//7yeNU2tRMRut0+fPt31G6/z5s1zn7mpvumb8w4ePJic\nnGyz2SZMmOBwOCwWi97D2rVrExMTw8LCTE386VbTtFWrVi1fvtxjKyAUWDTDIw432s6dO5ct\nW5aQkFBdXR2wQQHAf7m5ua4/1VBdXZ2QkBDs6aissbGxsLCwqakpkIO6Et619JCcnFxWVjZ5\n8uTrNSXg+uIZOwBAEKSkpIwcObK2tjbYExmAmpqa2NhYUh1CGcEOABAcK1asGFx/etW1Dxvs\nWQC+RAR7AgCAH6jMzEzjV8QFwDXuwwZ4tsBV4I4dAACAIgh2AAAAiiDYAQAAKIJgBwAAoAiC\nHQAAgCIIdgAAAIog2AEAACiCYAcAAKAIgh0AAIAiCHYAAACKINgBAAAogmAHAACgCIIdAACA\nIgh2AAAAiiDYAQAAKIJgBwAAoAiCHQAAgCIIdgAAAIog2AEAACiCYAcAAKAIgh0AAIAiCHYA\nAACKINgBAAAogmAHAACgCIIdAACAIgh2AAAAiiDYAQAAKIJgBwAAoAiCHQAAgCIIdgAAAIog\n2AEAACiCYAcAAKAIgh0AAIAiCHYAAACKINgBAAAogmAHAACgCIIdAACAIgh2AAAAiiDYAQAA\nKIJgBwAAoAiCHQAAgCIIdgAAAIog2AEAACiCYAcAAKAIgh0AAIAiCHYAAACKINgBAAAogmAH\nAACgCIIdAACAIgh2AAAAiiDYAQAAKIJgBwAAoAiCHQAAgCIIdgAAAIog2AEAACiCYAcAAKAI\ngh0AAIAiCHYAAACKINgBAAAogmAHAACgCIIdAACAIgh2AAAAiiDYAQAAKIJgBwAAoAiCHQAA\ngCIIdgAAAIog2AEAACiCYAcAAKAIgh0AAIAiCHYAAACKINgBAAAogmAHAACgCIIdAACAIgh2\nAAAAiiDYAQAAKIJgBwAAoAiCHQAAgCIIdgAAAIog2AEAACiCYAcAAKAIgh0AAIAiCHYAAACK\nINgBAAAogmAHAACgCIIdAACAIgh2AAAAiiDYAQAAKIJgBwAAoAiCHQAAgCIIdgAAAIog2AEA\nACiCYAcAAKAIgh0AAIAiCHYAAACKINgBAAAogmAHAACgCIIdAACAIgh2AAAAiiDYAQAAKIJg\nBwAAoAiCHQAAgCIIdgAAAIog2AEAACiCYAcAAKAIgh0AAIAiCHYAAACKINgBAAAogmAHAACg\nCIIdAACAIgh2AAAAiiDYAQAAKIJgBwAAoAiCHQAAgCIIdgAAAIog2AEAACiCYAcAAKAIgh0A\nAIAiCHYAAACKINgBAAAogmAHAACgCIIdAACAIgh2AAAAiiDYAQAAKIJgBwAAoAiCHQAAgCII\ndgAAAIog2AEAACiCYAcAAKAIgh0AAIAiCHYAAACKINgBAAAogmAHAACgCIIdAACAIgh2AAAA\niiDYAQAAKIJgBwAAoAiCHQAAgCIIdgAAAIog2AEAACiCYAcAAKAIgh0AAIAiCHYAAACKINgB\nAAAogmAHAACgCIIdAACAIgh2AAAAiiDYAQAAKIJgBwAAoAiCHQAAgCIIdgAAAIog2AEAACiC\nYAcAAKAIgh0AAIAiCHYAAACKINgBAAAogmAHAACgCIIdAACAIgh2AAAAiiDYAQAAKIJgBwAA\noAiCHQAAgCIIdgAAAIog2AEAACiCYAcAAKAIgh0AAIAiCHYAAACKINgBAAAogmAHAACgCIId\nAACAIgh2AAAAiiDYAQAAKIJgBwAAoAiCHQAAgCIIdgAAAIog2AEAACiCYAcAAKAIgh0AAIAi\nCHYAAACKINgBAAAogmAHAACgCIIdAACAIgh2AAAAiiDYAQAAKIJgBwAAoIiIwA/Z3t6el5cX\n+HEBoF/t7e3BngIAXL0gBLu+vr7W1tbAjwsAAKA2tmIBAAAUYdE0LWCDtbe3NzU1BWw4ALhq\nM2bMiIqKCvYsAGBgAhrsAAAAcOOwFQsAAKAIgh0AAIAiCHYAAACKINgBAAAogmAHAACgCIId\nAACAIgh2CBCLxWKxWEyFJ0+e/Jd/+ZcpU6bY7fbIyMjExMQ5c+asX7++q6vLvW1WVpY/3Vp8\n6neeP//5z/2pBgBACCLYIWj+8Ic/3HXXXRUVFStWrPjuu+/OnTtXU1Nz7733Ll26dPLkyYcP\nHzbV37p1a2VlpZ+da174btXV1VVVVXUV7wUAgFDAFxQjQFy3wYzrLSMjY/fu3QcOHJg2bZqx\nZklJyTPPPJObm6tnLP0W2pgxY1paWoYMGeKjW/cSf5w+ffrAgQO//vWv6+rqrqI5AAChgGCH\nQDBtbrpWXWRk5KVLl06ePBkTE2M8e+zYsVtvvdVut587d87Y/Kc//em2bduefvrpX//616ae\nrz3YeZwhAACDC8EOAeKetyZMmPDNN988/PDDb7zxRnx8fL9tv/vuu7vuuqu3t7ehoWHSpEne\nur26YHe9mgMAEEQ8Y4egee+99+Lj4z/++OMxY8ZkZWVt2LDhzJkzPurfdttty5Yt6+3t/eUv\nf0nwAgDAHXfsECAe74R1dXVVVFRUVlbW1tZeuHDBarXOnj07Nzd38eLFN998s3vbS5cuTZ48\n+T//8z/XrVv36KOPeuzWx++0+rPauWMHABi8CHYIEN+Bqaen5/PPP6+qqtq0aVNbW1tCQsLv\nf//7O++8073tzp0777///piYmG+//Xb48OFsxQIAoCPYIUD8DEwXL15csmRJeXn5nDlzduzY\n4bFtQUHBRx999Itf/GLt2rUEOwAAdAQ7BIh7YIqKiuru7j5z5syPfvQjY80jR46MHTs2KipK\n/5piU9u2trY777yzs7Nz//7906dPF4IdAAAiwi9PIIjGjx8vIk1NTabykSNHikhkZKS3hgkJ\nCcXFxZcvX37iiSdu6AwBABhcCHYIkPDwcFNJUVGRiLzxxhum8u3bt4vInDlzfPT2j//4j1Om\nTPnqq6+u6xwBABjcCHYIkDvuuENE6uvr9ZInn3xy/vz55eXljz/++Lfffut0Ok+cOLFmzZrH\nHnvs9ttvf+WVV3z0Fh4e/tZbb/FHXQEAMOIZOwRITU3Nk08+eejQob6+PuOqKysre//997/6\n6qsLFy78+Mc/Hjdu3MKFCxcsWBAVFaXX8fbc289//vN33nlH/P66E4+dmPCMHQBg8CLYAQAA\nKIKtWAAAAEUQ7AAAABRBsAMAAFAEwQ4AAEARBDsAAABFEOwAAAAUQbADAABQBMEOAABAEQQ7\nAAAARRDsAAAAFEGwAwAAUATBDgAAQBEEOwAAAEX8f5IBF4kezxx/AAAAAElFTkSuQmCC",
      "text/plain": [
       "plot without title"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "patient.tSNE <- ggplot(meta, aes(x=tSNE_1, y=tSNE_2, color=Sample_Label)) + \n",
    "            geom_point(alpha = 0.8, size = 0.8) +  \n",
    "            labs(x = \"tSNE 1\", y = \"tSNE 2\", title = \"7,745 cells (Hovestadt et al.)\") +\n",
    "            scale_colour_manual(values = col.pal.2, name = \"Patient (Methylation Subgroup)\") + \n",
    "            theme_bw() + \n",
    "            theme(axis.text.x = element_blank(), axis.text.y = element_blank(), axis.ticks = element_blank(),\n",
    "              panel.border = element_rect(linetype = \"solid\", fill = NA, size = 1),\n",
    "              panel.grid.major = element_blank(), panel.grid.minor = element_blank(),\n",
    "              panel.background = element_blank()) + \n",
    "                guides(colour = guide_legend(override.aes = list(size=4)))\n",
    "\n",
    "patient.tSNE\n",
    "\n",
    "#pdf(\"~/Desktop/Branavan_Hovestadt_Patient_tSNE.pdf\", width = 9, height = 6)\n",
    "#patient.tSNE\n",
    "#dev.off()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "----\n",
    "### 5.2 Wnt Subgroup panels\n",
    "----\n",
    "\n",
    "- tSNE with GSVA score\n",
    "- tSNE with classification (binary)\n",
    "- barplot with classification"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "setwd(\"~/Desktop/Samwise/projects/SinghLab_scMedullo/Brananvan_Wnt/PublicData_Analysis_April2020/figures/\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "data": {},
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {},
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<strong>pdf:</strong> 2"
      ],
      "text/latex": [
       "\\textbf{pdf:} 2"
      ],
      "text/markdown": [
       "**pdf:** 2"
      ],
      "text/plain": [
       "pdf \n",
       "  2 "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "pdf(\"Branavan_Hovestadt_WntSubgroup_tSNE.pdf\", width = 6.5, height = 6)\n",
    "\n",
    "#### \n",
    "\n",
    "i <- meta$Wnt_Subgroup_GSVA\n",
    "name <- \"Wnt Subgroup \\nGSVA Score\"\n",
    "j <- meta$Wnt_Subgroup_Classification\n",
    "name2 <- \"Wnt Subgroup\\nClassification\"\n",
    "\n",
    "\n",
    "sp2 <- ggplot(meta, aes(x=tSNE_1, y=tSNE_2, color=i)) + \n",
    "        geom_point(alpha = 0.6, size = 1.5) +  \n",
    "       \n",
    "         scale_colour_gradientn(colours = rev(brewer.pal(10, \"Spectral\")), \n",
    "                                #limits = c(-0.6,0.6), \n",
    "                                name = name) +\n",
    "        labs(x = \"tSNE 1\", y = \"tSNE 2\", title = \"7,745 cells (Hovestadt et al.)\") +\n",
    "        theme_bw() + \n",
    "        theme(axis.text.x = element_blank(), axis.text.y = element_blank(), axis.ticks = element_blank(),\n",
    "              panel.border = element_rect(linetype = \"solid\", fill = NA, size = 1),\n",
    "              panel.grid.major = element_blank(), panel.grid.minor = element_blank(),\n",
    "              panel.background = element_blank())\n",
    "sp2\n",
    "\n",
    "#### Cell Classification\n",
    "\n",
    "sp3 <- ggplot(meta, aes(x=tSNE_1, y=tSNE_2, color=j)) + \n",
    "        geom_point(alpha = 0.45, size = 1.5) +  \n",
    "        #scale_colour_gradientn(colours = c(\"darkblue\", \"white\", \"darkred\"), limits = c(-1,1), name = \"Wnt Subgroup Score\") \n",
    "         scale_colour_manual(values = c(\"grey\", '#9E0142'), name = name2) +\n",
    "        labs(x = \"tSNE 1\", y = \"tSNE 2\", title = \"7,745 cells (Hovestadt et al.)\") +\n",
    "        theme_bw() + \n",
    "        theme(axis.text.x = element_blank(), axis.text.y = element_blank(), axis.ticks = element_blank(),\n",
    "              panel.border = element_rect(linetype = \"solid\", fill = NA, size = 1),\n",
    "              panel.grid.major = element_blank(), panel.grid.minor = element_blank(),\n",
    "              panel.background = element_blank())\n",
    "\n",
    "sp3\n",
    "\n",
    "#### Stacked Barplot\n",
    "\n",
    "counts <- data.matrix(table(meta$Sample_Label, \n",
    "                            meta$Wnt_Subgroup_Classification)\n",
    "                     )\n",
    "\n",
    "\n",
    "par(mar=c(10,5,5,5))\n",
    "barplot(t(counts),\n",
    "        las = 2,\n",
    "        col = c(\"grey\", '#9E0142'),\n",
    "        horiz = F,\n",
    "        ylab = \"Cell Count\",\n",
    "        border = \"black\",\n",
    "        main = name2\n",
    "       )\n",
    "\n",
    "dev.off()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---\n",
    "### 5.3 KEGG Wnt pathway panels\n",
    "---"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [
    {
     "data": {},
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {},
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<strong>pdf:</strong> 2"
      ],
      "text/latex": [
       "\\textbf{pdf:} 2"
      ],
      "text/markdown": [
       "**pdf:** 2"
      ],
      "text/plain": [
       "pdf \n",
       "  2 "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "pdf(\"Branavan_Hovestadt_KEGGWnt_tSNE.pdf\", width = 6.5, height = 6)\n",
    "\n",
    "#### \n",
    "\n",
    "i <- meta$KEGG_Wnt_pathway_GSVA\n",
    "name <- \"KEGG Wnt\\nPathway\\nGSVA Score\"\n",
    "\n",
    "j <- meta$KEGG_Wnt_pathway_Classification\n",
    "name2 <- \"KEGG Wnt\\nPathway\\nClassification\"\n",
    "\n",
    "sp2 <- ggplot(meta, aes(x=tSNE_1, y=tSNE_2, color=i)) + \n",
    "        geom_point(alpha = 0.6, size = 1.5) +  \n",
    "       \n",
    "         scale_colour_gradientn(colours = rev(brewer.pal(10, \"Spectral\")), \n",
    "                                #limits = c(-0.6,0.6), \n",
    "                                name = name) +\n",
    "        labs(x = \"tSNE 1\", y = \"tSNE 2\", title = \"7,745 cells (Hovestadt et al.)\") +\n",
    "        theme_bw() + \n",
    "        theme(axis.text.x = element_blank(), axis.text.y = element_blank(), axis.ticks = element_blank(),\n",
    "              panel.border = element_rect(linetype = \"solid\", fill = NA, size = 1),\n",
    "              panel.grid.major = element_blank(), panel.grid.minor = element_blank(),\n",
    "              panel.background = element_blank())\n",
    "sp2\n",
    "\n",
    "#### Cell Classification\n",
    "\n",
    "sp3 <- ggplot(meta, aes(x=tSNE_1, y=tSNE_2, color=j)) + \n",
    "        geom_point(alpha = 0.45, size = 1.5) +  \n",
    "        #scale_colour_gradientn(colours = c(\"darkblue\", \"white\", \"darkred\"), limits = c(-1,1), name = \"Wnt Subgroup Score\") \n",
    "         scale_colour_manual(values = c(\"grey\", '#9E0142'), name = name2) +\n",
    "        labs(x = \"tSNE 1\", y = \"tSNE 2\", title = \"7,745 cells (Hovestadt et al.)\") +\n",
    "        theme_bw() + \n",
    "        theme(axis.text.x = element_blank(), axis.text.y = element_blank(), axis.ticks = element_blank(),\n",
    "              panel.border = element_rect(linetype = \"solid\", fill = NA, size = 1),\n",
    "              panel.grid.major = element_blank(), panel.grid.minor = element_blank(),\n",
    "              panel.background = element_blank())\n",
    "\n",
    "sp3\n",
    "\n",
    "#### Stacked Barplot\n",
    "\n",
    "counts <- data.matrix(table(meta$Sample_Label, \n",
    "                            meta$KEGG_Wnt_pathway_Classification)\n",
    "                     )\n",
    "\n",
    "\n",
    "par(mar=c(10,5,5,5))\n",
    "barplot(t(counts),\n",
    "        las = 2,\n",
    "        col = c(\"grey\", '#9E0142'),\n",
    "        horiz = F,\n",
    "        ylab = \"Cell Count\",\n",
    "        border = \"black\",\n",
    "        main = name2\n",
    "       )\n",
    "\n",
    "dev.off()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---\n",
    "### 5.3 Canonical Wnt Pathway\n",
    "---"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [
    {
     "data": {},
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {},
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<strong>pdf:</strong> 2"
      ],
      "text/latex": [
       "\\textbf{pdf:} 2"
      ],
      "text/markdown": [
       "**pdf:** 2"
      ],
      "text/plain": [
       "pdf \n",
       "  2 "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "pdf(\"Branavan_Hovestadt_CanonicalWnt_tSNE.pdf\", width = 6.5, height = 6)\n",
    "\n",
    "#### \n",
    "\n",
    "i <- meta$Canonical_Wnt_pathway_GSVA\n",
    "name <- \"Canonical Wnt\\npathway\\nGSVA Score\"\n",
    "\n",
    "j <- meta$Canonical_Wnt_pathway_Classification\n",
    "name2 <- \"Canonical Wnt\\nPathway\\nClassification\"\n",
    "\n",
    "sp2 <- ggplot(meta, aes(x=tSNE_1, y=tSNE_2, color=i)) + \n",
    "        geom_point(alpha = 0.6, size = 1.5) +  \n",
    "       \n",
    "         scale_colour_gradientn(colours = rev(brewer.pal(10, \"Spectral\")), \n",
    "                                #limits = c(-0.6,0.6), \n",
    "                                name = name) +\n",
    "        labs(x = \"tSNE 1\", y = \"tSNE 2\", title = \"7,745 cells (Hovestadt et al.)\") +\n",
    "        theme_bw() + \n",
    "        theme(axis.text.x = element_blank(), axis.text.y = element_blank(), axis.ticks = element_blank(),\n",
    "              panel.border = element_rect(linetype = \"solid\", fill = NA, size = 1),\n",
    "              panel.grid.major = element_blank(), panel.grid.minor = element_blank(),\n",
    "              panel.background = element_blank())\n",
    "sp2\n",
    "\n",
    "#### Cell Classification\n",
    "\n",
    "sp3 <- ggplot(meta, aes(x=tSNE_1, y=tSNE_2, color=j)) + \n",
    "        geom_point(alpha = 0.45, size = 1.5) +  \n",
    "        #scale_colour_gradientn(colours = c(\"darkblue\", \"white\", \"darkred\"), limits = c(-1,1), name = \"Wnt Subgroup Score\") \n",
    "         scale_colour_manual(values = c(\"grey\", '#9E0142'), name = name2) +\n",
    "        labs(x = \"tSNE 1\", y = \"tSNE 2\", title = \"7,745 cells (Hovestadt et al.)\") +\n",
    "        theme_bw() + \n",
    "        theme(axis.text.x = element_blank(), axis.text.y = element_blank(), axis.ticks = element_blank(),\n",
    "              panel.border = element_rect(linetype = \"solid\", fill = NA, size = 1),\n",
    "              panel.grid.major = element_blank(), panel.grid.minor = element_blank(),\n",
    "              panel.background = element_blank())\n",
    "\n",
    "sp3\n",
    "\n",
    "#### Stacked Barplot\n",
    "\n",
    "counts <- data.matrix(table(meta$Sample_Label, \n",
    "                            meta$Canonical_Wnt_pathway_Classification)\n",
    "                     )\n",
    "\n",
    "\n",
    "par(mar=c(10,5,5,5))\n",
    "barplot(t(counts),\n",
    "        las = 2,\n",
    "        col = c(\"grey\", '#9E0142'),\n",
    "        horiz = F,\n",
    "        ylab = \"Cell Count\",\n",
    "        border = \"black\",\n",
    "        main = name2\n",
    "       )\n",
    "\n",
    "dev.off()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "----"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "R",
   "language": "R",
   "name": "ir"
  },
  "language_info": {
   "codemirror_mode": "r",
   "file_extension": ".r",
   "mimetype": "text/x-r-source",
   "name": "R",
   "pygments_lexer": "r",
   "version": "3.6.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}