02CAGEfightR_for_downstream_analysis.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "1e783bcf",
   "metadata": {},
   "outputs": [],
   "source": [
    "library(CAGEfightR)\n",
    "library(GenomicFeatures)\n",
    "library(diptest)\n",
    "setwd(\"./soybean_data/stripe-seq_data/\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "7747a420",
   "metadata": {},
   "outputs": [],
   "source": [
    "bw_plus <- BigWigFileList(c(\"W82.Flower.plus.rd.bw\",\"W82.Leaf.plus.rd.bw\",\n",
    "                            \"W82.Nodule.plus.rd.bw\",\"W82.Pod.plus.rd.bw\",\n",
    "                            \"W82.Root.plus.rd.bw\",\"W82.Seed.plus.rd.bw\",\n",
    "                            \"W82.Shoot.plus.rd.bw\",\"W82.Stemtip.plus.rd.bw\"))\n",
    "\n",
    "bw_minus <- BigWigFileList(c(\"W82.Flower.minus.rd.bw\",\"W82.Leaf.minus.rd.bw\",\n",
    "                             \"W82.Nodule.minus.rd.bw\",\"W82.Pod.minus.rd.bw\",\n",
    "                             \"W82.Root.minus.rd.bw\",\"W82.Seed.minus.rd.bw\",\n",
    "                             \"W82.Shoot.minus.rd.bw\",\"W82.Stemtip.minus.rd.bw\"))\n",
    "\n",
    "names(bw_plus) <- c(\"Flower\",\"Leaf\",\n",
    "                    \"Nodule\",\"Pod\",\n",
    "                    \"Root\",\"Seed\",\n",
    "                    \"Shoot\",\"Stemtip\")\n",
    "\n",
    "names(bw_minus) <-  c(\"Flower\",\"Leaf\",\n",
    "                      \"Nodule\",\"Pod\",\n",
    "                      \"Root\",\"Seed\",\n",
    "                      \"Shoot\",\"Stemtip\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "a9f6d39e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Seqinfo object with 20 sequences from an unspecified genome:\n",
       "  seqnames seqlengths isCircular genome\n",
       "  Chr01      56831624       <NA>   <NA>\n",
       "  Chr02      48577505       <NA>   <NA>\n",
       "  Chr03      45779781       <NA>   <NA>\n",
       "  Chr04      52389146       <NA>   <NA>\n",
       "  Chr05      42234498       <NA>   <NA>\n",
       "  ...             ...        ...    ...\n",
       "  Chr16      37887014       <NA>   <NA>\n",
       "  Chr17      41641366       <NA>   <NA>\n",
       "  Chr18      58018742       <NA>   <NA>\n",
       "  Chr19      50746916       <NA>   <NA>\n",
       "  Chr20      47904181       <NA>   <NA>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "Gm <- seqinfo(bw_plus[[1]])\n",
    "Gm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "e0115ec5",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Checking design...\n",
      "\n",
      "Checking supplied genome compatibility...\n",
      "\n",
      "Iterating over 1 genomic tiles in 8 samples using 102 worker(s)...\n",
      "\n",
      "Importing CTSSs from plus strand...\n",
      "\n",
      "Importing CTSSs from minus strand...\n",
      "\n",
      "Merging strands...\n",
      "\n",
      "Formatting output...\n",
      "\n",
      "### CTSS summary ###\n",
      "\n",
      "Number of samples: 8\n",
      "\n",
      "Number of CTSSs: 2.318 millions\n",
      "\n",
      "Sparsity: 78.38 %\n",
      "\n",
      "Type of rowRanges: StitchedGPos\n",
      "\n",
      "Final object size: 61.90 MB\n",
      "\n",
      " - Running calcTPM and calcPooled:\n",
      "\n",
      "Calculating library sizes...\n",
      "\n",
      "Calculating TPM...\n",
      "\n",
      "\n",
      " - Running clusterUnidirectionally:\n",
      "\n",
      "Splitting by strand...\n",
      "\n",
      "Slice-reduce to find clusters...\n",
      "\n",
      "Calculating statistics...\n",
      "\n",
      "Preparing output...\n",
      "\n",
      "Tag clustering summary:\n",
      "\n",
      "  Width  Count Percent\n",
      "  Total 656075 1e+02 %\n",
      "    >=1 492084 8e+01 %\n",
      "   >=10 148200 2e+01 %\n",
      "  >=100  15781 2e+00 %\n",
      " >=1000     10 2e-03 %\n",
      "\n",
      "\n",
      " - Running quantifyClusters:\n",
      "\n",
      "Finding overlaps...\n",
      "\n",
      "Aggregating within clusters...\n",
      "\n"
     ]
    }
   ],
   "source": [
    "CTSSs <- quantifyCTSSs(plusStrand=bw_plus,\n",
    "                       minusStrand=bw_minus,\n",
    "                       genome=Gm)\n",
    "TSSs <- quickTSSs(CTSSs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "a8d0771e",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Import genomic features from the file as a GRanges object ... \n",
      "OK\n",
      "\n",
      "Prepare the 'metadata' data frame ... \n",
      "OK\n",
      "\n",
      "Make the TxDb object ... \n",
      "OK\n",
      "\n"
     ]
    }
   ],
   "source": [
    "\n",
    "txdb <- makeTxDbFromGFF(file=\"../genome_file/Gmax_v2.0_changed.rmscaff.gff3\",\n",
    "                         #file=\"Gmax_merged.gff3\",\n",
    "                        format=\"gff3\",\n",
    "                        dataSource=\"gff for Gm\",\n",
    "                       organism=\"Glycine max\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "7e1b9fab",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Finding hierachical overlaps...\n",
      "\n",
      "### Overlap summary: ###\n",
      "\n",
      "      txType  count percentage\n",
      "1   promoter 673537       29.1\n",
      "2   proximal 104049        4.5\n",
      "3    fiveUTR 412753       17.8\n",
      "4   threeUTR  25428        1.1\n",
      "5        CDS 839307       36.2\n",
      "6       exon      0        0.0\n",
      "7     intron 132715        5.7\n",
      "8  antisense  49923        2.2\n",
      "9 intergenic  80621        3.5\n",
      "\n",
      "Extracting transcripts...\n",
      "\n",
      "Finding hierachical overlaps...\n",
      "\n",
      "### Overlap Summary: ###\n",
      "\n",
      "Features overlapping transcripts: 94.37 %\n",
      "\n",
      "Number of unique transcripts: 77558\n",
      "\n"
     ]
    }
   ],
   "source": [
    "# Annotate both TSSs and enhancers\n",
    "CTSSs <- assignTxType(CTSSs, txModels=txdb,\n",
    "                     outputColumn=\"txType\")\n",
    "#####\n",
    "CTSSs <- assignTxID(CTSSs,\n",
    "                   txModels=txdb,\n",
    "                   outputColumn=\"txID\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "50ebb115",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Calculating library sizes...\n",
      "\n",
      "Calculating TPM...\n",
      "\n"
     ]
    }
   ],
   "source": [
    "CTSSs <- calcTPM(CTSSs,\n",
    "                 inputAssay=\"counts\",\n",
    "                 outputAssay=\"TPM\",\n",
    "                 outputColumn=\"subsetTags\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "f23665a9",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "class: RangedSummarizedExperiment \n",
       "dim: 2318333 8 \n",
       "metadata(0):\n",
       "assays(2): counts TPM\n",
       "rownames: NULL\n",
       "rowData names(2): txType txID\n",
       "colnames(8): Flower Leaf ... Shoot Stemtip\n",
       "colData names(1): subsetTags"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "CTSSs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "33d93338",
   "metadata": {},
   "outputs": [],
   "source": [
    "###export a table\n",
    "result <- data.frame(as.data.frame(CTSSs@rowRanges),\n",
    "                     as.data.frame(as.matrix(CTSSs@assays@data@listData$counts)),\n",
    "                    as.data.frame(as.matrix(CTSSs@assays@data@listData$TPM)))\n",
    "\n",
    "write.table(result,\n",
    "            file= \"../result_file/CTSS_raw.txt\",\n",
    "            col.names=T,row.names=F,\n",
    "            sep=\"\\t\",quote=F)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "ecf9c5a1",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "StitchedGPos object with 2318333 positions and 3 metadata columns:\n",
       "            seqnames       pos strand |   txType                   txID\n",
       "               <Rle> <integer>  <Rle> | <factor>            <character>\n",
       "        [1]    Chr01    116022      + | promoter Glyma.01G000600.1.Wm..\n",
       "        [2]    Chr01    116026      + | promoter Glyma.01G000600.1.Wm..\n",
       "        [3]    Chr01    116072      + | promoter Glyma.01G000600.1.Wm..\n",
       "        [4]    Chr01    116088      + | promoter Glyma.01G000600.1.Wm..\n",
       "        [5]    Chr01    116089      + | promoter Glyma.01G000600.1.Wm..\n",
       "        ...      ...       ...    ... .      ...                    ...\n",
       "  [2318329]    Chr20  47901250      - | promoter Glyma.20G250200.2.Wm..\n",
       "  [2318330]    Chr20  47901252      - | promoter Glyma.20G250200.2.Wm..\n",
       "  [2318331]    Chr20  47901287      - | promoter Glyma.20G250200.2.Wm..\n",
       "  [2318332]    Chr20  47901289      - | promoter Glyma.20G250200.2.Wm..\n",
       "  [2318333]    Chr20  47901358      - | promoter Glyma.20G250200.2.Wm..\n",
       "                score\n",
       "            <numeric>\n",
       "        [1]  0.953222\n",
       "        [2]  0.299031\n",
       "        [3]  0.173313\n",
       "        [4]  1.885623\n",
       "        [5]  8.781491\n",
       "        ...       ...\n",
       "  [2318329] 1.4762020\n",
       "  [2318330] 0.2990306\n",
       "  [2318331] 0.0866565\n",
       "  [2318332] 1.2703863\n",
       "  [2318333] 0.5199391\n",
       "  -------\n",
       "  seqinfo: 20 sequences from an unspecified genome"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "CTSSs <- calcPooled(CTSSs, inputAssay=\"TPM\")\n",
    "rowRanges(CTSSs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "01a193ed",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\n",
       "      0       1       2       3       4       5       6       7       8 \n",
       "1825475  327548   63415   30885   19682   14658   12443   12008   12219 "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "CTSSs <- calcSupport(CTSSs,inputAssay=\"TPM\", \n",
    "                            outputColumn=\"support\", \n",
    "                            unexpressed=1)\n",
    "table(rowRanges(CTSSs)$support)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "320ca12d",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Warning message in calcPooled(supportedCTSSs):\n",
      "\"object already has a column named score in rowData: It will be overwritten!\"\n"
     ]
    }
   ],
   "source": [
    "## ----subset, tidy=FALSE-------------------------------------------------------\n",
    "supportedCTSSs <- subset(CTSSs, support > 0)\n",
    "#supportedCTSSs <- calcTPM(supportedCTSSs, totalTags=\"totalTags\")\n",
    "supportedCTSSs <- calcPooled(supportedCTSSs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "0a7430c7",
   "metadata": {},
   "outputs": [],
   "source": [
    "result <-data.frame(as.data.frame(supportedCTSSs@rowRanges),\n",
    "                    as.data.frame(as.matrix(assay(supportedCTSSs,\"counts\"))),\n",
    "                    as.data.frame(as.matrix(assay(supportedCTSSs,\"TPM\"))))\n",
    "write.table(result,\n",
    "            file=\"../result_file/supportTSSs.txt\",\n",
    "            col.names=T,row.names=F,\n",
    "            sep=\"\\t\",quote=F)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "e81664d0",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Splitting by strand...\n",
      "\n",
      "Slice-reduce to find clusters...\n",
      "\n",
      "Calculating statistics...\n",
      "\n",
      "Preparing output...\n",
      "\n",
      "Tag clustering summary:\n",
      "\n",
      "  Width  Count Percent\n",
      "  Total 193579   100 %\n",
      "    >=1 151172    78 %\n",
      "   >=10  40448    21 %\n",
      "  >=100   1959     1 %\n",
      " >=1000      0     0 %\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "GRanges object with 193579 ranges and 2 metadata columns:\n",
       "                            seqnames            ranges strand |     score\n",
       "                               <Rle>         <IRanges>  <Rle> | <numeric>\n",
       "      Chr01:116089-116110;+    Chr01     116089-116110      + |  20.03456\n",
       "      Chr01:116131-116134;+    Chr01     116131-116134      + |  17.41587\n",
       "      Chr01:116213-116213;+    Chr01            116213      + |   1.73737\n",
       "      Chr01:116245-116245;+    Chr01            116245      + |   1.19612\n",
       "      Chr01:117419-117419;+    Chr01            117419      + |   1.38735\n",
       "                        ...      ...               ...    ... .       ...\n",
       "  Chr20:47900986-47900986;-    Chr20          47900986      - |   2.44671\n",
       "  Chr20:47901093-47901110;-    Chr20 47901093-47901110      - |  12.71380\n",
       "  Chr20:47901180-47901201;-    Chr20 47901180-47901201      - |  26.73318\n",
       "  Chr20:47901226-47901250;-    Chr20 47901226-47901250      - |   9.13046\n",
       "  Chr20:47901289-47901289;-    Chr20          47901289      - |   1.27039\n",
       "                                thick\n",
       "                            <IRanges>\n",
       "      Chr01:116089-116110;+    116089\n",
       "      Chr01:116131-116134;+    116131\n",
       "      Chr01:116213-116213;+    116213\n",
       "      Chr01:116245-116245;+    116245\n",
       "      Chr01:117419-117419;+    117419\n",
       "                        ...       ...\n",
       "  Chr20:47900986-47900986;-  47900986\n",
       "  Chr20:47901093-47901110;-  47901110\n",
       "  Chr20:47901180-47901201;-  47901187\n",
       "  Chr20:47901226-47901250;-  47901238\n",
       "  Chr20:47901289-47901289;-  47901289\n",
       "  -------\n",
       "  seqinfo: 20 sequences from an unspecified genome"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#change !!!\n",
    "prefiltered_TCs <- clusterUnidirectionally(supportedCTSSs, \n",
    "                                           pooledCutoff=0, \n",
    "                                           mergeDist=20)\n",
    "prefiltered_TCs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "977b3d5b",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Finding overlaps...\n",
      "\n",
      "Aggregating within clusters...\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "class: RangedSummarizedExperiment \n",
       "dim: 193579 8 \n",
       "metadata(0):\n",
       "assays(1): counts\n",
       "rownames(193579): Chr01:116089-116110;+ Chr01:116131-116134;+ ...\n",
       "  Chr20:47901226-47901250;- Chr20:47901289-47901289;-\n",
       "rowData names(2): score thick\n",
       "colnames(8): Flower Leaf ... Shoot Stemtip\n",
       "colData names(1): subsetTags"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#potential change the sample number.\n",
    "sup_TSSs <- quantifyClusters(supportedCTSSs,\n",
    "                                      clusters=prefiltered_TCs,\n",
    "                                      inputAssay=\"counts\")\n",
    "sup_TSSs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "3defcdc6",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Calculating library sizes...\n",
      "\n",
      "Warning message in calcTotalTags(object = object, inputAssay = inputAssay, outputColumn = outputColumn):\n",
      "\"object already has a column named subsetTags in colData: It will be overwritten!\"\n",
      "Calculating TPM...\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "class: RangedSummarizedExperiment \n",
       "dim: 193579 8 \n",
       "metadata(0):\n",
       "assays(2): counts TPM\n",
       "rownames(193579): Chr01:116089-116110;+ Chr01:116131-116134;+ ...\n",
       "  Chr20:47901226-47901250;- Chr20:47901289-47901289;-\n",
       "rowData names(2): score thick\n",
       "colnames(8): Flower Leaf ... Shoot Stemtip\n",
       "colData names(1): subsetTags"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "## ----supportOnTPM, tidy=FALSE-------------------------------------------------\n",
    "# Calculate TPM using pre-calculated total tags:\n",
    "Unidirectional <- calcTPM(sup_TSSs,\n",
    "                          inputAssay=\"counts\", \n",
    "                          outputAssay=\"TPM\", \n",
    "                          outputColumn=\"subsetTags\")\n",
    "Unidirectional"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "88d2e6ea",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Calculating support...\n",
      "\n",
      "Subsetting...\n",
      "\n",
      "Removed 0 out of 193579 regions (0%)\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "class: RangedSummarizedExperiment \n",
       "dim: 193579 8 \n",
       "metadata(0):\n",
       "assays(2): counts TPM\n",
       "rownames(193579): Chr01:116089-116110;+ Chr01:116131-116134;+ ...\n",
       "  Chr20:47901226-47901250;- Chr20:47901289-47901289;-\n",
       "rowData names(3): score thick support\n",
       "colnames(8): Flower Leaf ... Shoot Stemtip\n",
       "colData names(1): subsetTags"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Only TSSs expressed at more than 1 TPM in more than 2 samples\n",
    "#Unidirectional <- subsetBySupport(Unidirectional,\n",
    "#                                  inputAssay=\"TPM\",\n",
    "#                                  unexpressed=0,\n",
    "#                                  minSamples=2)\n",
    "#TSSs expressed at more than 0 TPM as least in 1 samples\n",
    "Unidirectional <- subsetBySupport(Unidirectional,\n",
    "                                     inputAssay=\"TPM\",\n",
    "                                     unexpressed=1,\n",
    "                                     minSamples=0)\n",
    "Unidirectional"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "9c5255a2",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Extracting transcripts...\n",
      "\n",
      "Finding hierachical overlaps...\n",
      "\n",
      "### Overlap Summary: ###\n",
      "\n",
      "Features overlapping transcripts: 93.94 %\n",
      "\n",
      "Number of unique transcripts: 68290\n",
      "\n"
     ]
    }
   ],
   "source": [
    "## ----assignTxID, tidy=FALSE---------------------------------------------------\n",
    "Unidirectional <- assignTxID(Unidirectional,\n",
    "                            txModels=txdb,\n",
    "                            outputColumn=\"txID\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "id": "275cfc01",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "GRanges object with 2 ranges and 4 metadata columns:\n",
       "                        seqnames    ranges strand |     score     thick\n",
       "                           <Rle> <IRanges>  <Rle> | <numeric> <IRanges>\n",
       "  Chr01:117419-117419;+    Chr01    117419      + |   1.38735    117419\n",
       "  Chr01:118130-118130;+    Chr01    118130      + |   1.09059    118130\n",
       "                          support                   txID\n",
       "                        <integer>            <character>\n",
       "  Chr01:117419-117419;+         1 Glyma.01G000600.1.Wm..\n",
       "  Chr01:118130-118130;+         1 Glyma.01G000600.1.Wm..\n",
       "  -------\n",
       "  seqinfo: 20 sequences from an unspecified genome"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "## ----multipleTxs, tidy=FALSE--------------------------------------------------\n",
    "rowRanges(Unidirectional)[5:6]\n",
    "#export(rowRanges(Unidirectional),\"TC.bed\",\"BED\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "792a5661",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Finding hierachical overlaps...\n",
      "\n",
      "### Overlap summary: ###\n",
      "\n",
      "      txType count percentage\n",
      "1   promoter 54851       28.3\n",
      "2   proximal  9585        5.0\n",
      "3    fiveUTR 33092       17.1\n",
      "4   threeUTR  1786        0.9\n",
      "5        CDS 69569       35.9\n",
      "6       exon     0        0.0\n",
      "7     intron 12957        6.7\n",
      "8  antisense  4832        2.5\n",
      "9 intergenic  6907        3.6\n",
      "\n"
     ]
    }
   ],
   "source": [
    "## ----assignTxType, tidy=FALSE-------------------------------------------------\n",
    "Unidirectional <- assignTxType(Unidirectional,\n",
    "                              txModels=txdb,\n",
    "                              outputColumn=\"txType\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "52b21636",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Finding hierachical overlaps with swapped ranges...\n",
      "\n",
      "### Overlap summary: ###\n",
      "\n",
      "      txType count percentage\n",
      "1   promoter 52878       27.3\n",
      "2   proximal  9870        5.1\n",
      "3    fiveUTR 34187       17.7\n",
      "4   threeUTR  1784        0.9\n",
      "5        CDS 69942       36.1\n",
      "6       exon     0        0.0\n",
      "7     intron 13175        6.8\n",
      "8  antisense  4824        2.5\n",
      "9 intergenic  6919        3.6\n",
      "\n"
     ]
    }
   ],
   "source": [
    "## ----swappedTxType, tidy=FALSE------------------------------------------------\n",
    "Unidirectional <- assignTxType(Unidirectional,\n",
    "                              txModels=txdb,\n",
    "                              outputColumn=\"peakTxType\",\n",
    "                              swap=\"thick\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "id": "deae20f8",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Splitting by strand...\n",
      "\n",
      "Applying function to each cluster...\n",
      "\n",
      "Preparing output output...\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "class: RangedSummarizedExperiment \n",
       "dim: 193579 8 \n",
       "metadata(0):\n",
       "assays(2): counts TPM\n",
       "rownames(193579): Chr01:116089-116110;+ Chr01:116131-116134;+ ...\n",
       "  Chr20:47901226-47901250;- Chr20:47901289-47901289;-\n",
       "rowData names(7): score thick ... peakTxType IQR\n",
       "colnames(8): Flower Leaf ... Shoot Stemtip\n",
       "colData names(1): subsetTags"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Splitting by strand...\n",
      "\n",
      "Applying function to each cluster...\n",
      "\n",
      "Preparing output output...\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "class: RangedSummarizedExperiment \n",
       "dim: 193579 8 \n",
       "metadata(0):\n",
       "assays(2): counts TPM\n",
       "rownames(193579): Chr01:116089-116110;+ Chr01:116131-116134;+ ...\n",
       "  Chr20:47901226-47901250;- Chr20:47901289-47901289;-\n",
       "rowData names(8): score thick ... IQR Entropy\n",
       "colnames(8): Flower Leaf ... Shoot Stemtip\n",
       "colData names(1): subsetTags"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Calculate shape\n",
    "Unidirectional <- calcShape(Unidirectional,\n",
    "                           pooled=supportedCTSSs,\n",
    "                           outputColumn = \"IQR\",\n",
    "                           shapeFunction = shapeIQR,\n",
    "                           lower=0.25, upper=0.75)\n",
    "Unidirectional\n",
    "Unidirectional <- calcShape(Unidirectional,\n",
    "                        pooled=supportedCTSSs,\n",
    "                        outputColumn = \"Entropy\",\n",
    "                        shapeFunction = shapeEntropy\n",
    "                        )\n",
    "Unidirectional"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "id": "8fbeac6b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA0gAAANICAIAAAByhViMAAAACXBIWXMAABJ0AAASdAHeZh94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\nSLzPgQhbKpEAAAAASUVORK5CYII=",
      "text/plain": [
       "Plot with title \"Histogram of rowRanges(Unidirectional)$IQR\""
      ]
     },
     "metadata": {
      "image/png": {
       "height": 420,
       "width": 420
      }
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA0gAAANICAIAAAByhViMAAAACXBIWXMAABJ0AAASdAHeZh94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\nAAAUQbADAABQxP8AkcNFGVlJzngAAAAASUVORK5CYII=",
      "text/plain": [
       "Plot with title \"Histogram of rowRanges(Unidirectional)$Entropy\""
      ]
     },
     "metadata": {
      "image/png": {
       "height": 420,
       "width": 420
      }
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "## ----histIQR, tidy=FALSE------------------------------------------------------\n",
    "hist(rowRanges(Unidirectional)$IQR,\n",
    "     breaks=max(rowRanges(Unidirectional)$IQR),\n",
    "     xlim=c(0,50), \n",
    "     xlab = \"IQR\", \n",
    "     col=\"red\")\n",
    "\n",
    "hist(rowRanges(Unidirectional)$Entropy,\n",
    "     breaks=max(rowRanges(Unidirectional)$Entropy),\n",
    "     xlim=c(0,20), \n",
    "     xlab = \"Entropy\", \n",
    "     col=\"red\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "id": "a8117f1a",
   "metadata": {},
   "outputs": [],
   "source": [
    "## ----customShape, tidy=FALSE--------------------------------------------------\n",
    "# Write a function that quantifies the lean of a TSS\n",
    "\n",
    "TSRshapeIndex <- function(x) {\n",
    "  x <- as.vector(x)\n",
    "  tagcount <- sum(x)  \n",
    "  SI <- suppressWarnings(2+sum(ifelse(x > 0, (x/tagcount) * log2(x/tagcount), 0)))  \n",
    "  return(SI)\n",
    "}\n",
    "\n",
    "TSRshapeIndex <- function(x) {\n",
    "  x <- as.vector(x)\n",
    "  tagcount <- sum(x) \n",
    "  SI <- suppressWarnings(2+sum(ifelse(x > 0, (x/tagcount) * log2(x/tagcount), 0))) \n",
    "  return(SI)\n",
    "}\n",
    "\n",
    "TSRmshapeIndex <- function(x) {\n",
    "  x <- as.vector(x)\n",
    "  tagcount <- sum(x)\n",
    "  tsscount <- length(x)\n",
    "  mSI <- suppressWarnings(1+sum(ifelse((tsscount != 1 & x != 0), (x/tagcount) * log((x/tagcount))/log(tsscount), 0)))\n",
    "  return(mSI)\n",
    "}\n",
    "\n",
    "TSRshapiro <- function(x) {\n",
    "  x <- as.vector(x)\n",
    "  x[1]<-x[1]+0.0000001 \n",
    "  shapiro <-suppressWarnings(ifelse(length(x) >=3, shapiro.test(x)$p.value,0))\n",
    "  return(shapiro)\n",
    "}\n",
    "\n",
    "\n",
    "TSRdip <- function(x) {\n",
    "  x <- as.vector(x)\n",
    "  dip <-suppressWarnings(ifelse(length(x) >=3, dip.test(x)$p.value,1))\n",
    "  return(dip)\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "id": "4b2f74e7",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Splitting by strand...\n",
      "\n",
      "Applying function to each cluster...\n",
      "\n",
      "Preparing output output...\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "class: RangedSummarizedExperiment \n",
       "dim: 193579 8 \n",
       "metadata(0):\n",
       "assays(2): counts TPM\n",
       "rownames(193579): Chr01:116089-116110;+ Chr01:116131-116134;+ ...\n",
       "  Chr20:47901226-47901250;- Chr20:47901289-47901289;-\n",
       "rowData names(9): score thick ... Entropy SI\n",
       "colnames(8): Flower Leaf ... Shoot Stemtip\n",
       "colData names(1): subsetTags"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "Unidirectional <- calcShape(Unidirectional,\n",
    "                    pooled=supportedCTSSs,\n",
    "                    outputColumn = \"SI\",\n",
    "                    shapeFunction = TSRshapeIndex)\n",
    "Unidirectional"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "id": "ef82c8fa",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Splitting by strand...\n",
      "\n",
      "Applying function to each cluster...\n",
      "\n",
      "Preparing output output...\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "class: RangedSummarizedExperiment \n",
       "dim: 193579 8 \n",
       "metadata(0):\n",
       "assays(2): counts TPM\n",
       "rownames(193579): Chr01:116089-116110;+ Chr01:116131-116134;+ ...\n",
       "  Chr20:47901226-47901250;- Chr20:47901289-47901289;-\n",
       "rowData names(10): score thick ... SI mSI\n",
       "colnames(8): Flower Leaf ... Shoot Stemtip\n",
       "colData names(1): subsetTags"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "Unidirectional <- calcShape(Unidirectional,\n",
    "                        pooled=supportedCTSSs,\n",
    "                        outputColumn = \"mSI\",\n",
    "                        shapeFunction = TSRmshapeIndex)\n",
    "Unidirectional"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "id": "38c1934c",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Splitting by strand...\n",
      "\n",
      "Applying function to each cluster...\n",
      "\n",
      "Preparing output output...\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "class: RangedSummarizedExperiment \n",
       "dim: 193579 8 \n",
       "metadata(0):\n",
       "assays(2): counts TPM\n",
       "rownames(193579): Chr01:116089-116110;+ Chr01:116131-116134;+ ...\n",
       "  Chr20:47901226-47901250;- Chr20:47901289-47901289;-\n",
       "rowData names(11): score thick ... mSI shapiro\n",
       "colnames(8): Flower Leaf ... Shoot Stemtip\n",
       "colData names(1): subsetTags"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "Unidirectional <- calcShape(Unidirectional,\n",
    "                        pooled=supportedCTSSs,\n",
    "                        outputColumn = \"shapiro\",\n",
    "                        shapeFunction = TSRshapiro)\n",
    "Unidirectional"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "id": "aacdae13",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Splitting by strand...\n",
      "\n",
      "Applying function to each cluster...\n",
      "\n",
      "Preparing output output...\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "class: RangedSummarizedExperiment \n",
       "dim: 193579 8 \n",
       "metadata(0):\n",
       "assays(2): counts TPM\n",
       "rownames(193579): Chr01:116089-116110;+ Chr01:116131-116134;+ ...\n",
       "  Chr20:47901226-47901250;- Chr20:47901289-47901289;-\n",
       "rowData names(12): score thick ... shapiro Dip\n",
       "colnames(8): Flower Leaf ... Shoot Stemtip\n",
       "colData names(1): subsetTags"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "Unidirectional <- calcShape(Unidirectional,\n",
    "                        pooled=supportedCTSSs,\n",
    "                        outputColumn = \"Dip\",\n",
    "                        shapeFunction = TSRdip)\n",
    "Unidirectional"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "id": "477fb190",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA0gAAANICAIAAAByhViMAAAACXBIWXMAABJ0AAASdAHeZh94\nAAAgAElEQVR4nOzdeVyU5f7/8QuQRcWGRSgwm0ByIcssQ0lxK8XsuJyDZikO2LG0OpVGlmmm\n1emo5dLJPOlJcsFoQyqzE2aSJkJkq2maS4Ai5QqjIjv374/7e+7fnAHmHkzmnvv29fzDx3DN\nxcznnrmZeXtvHw9JkgQAAAD0z1PrAgAAAHBpEOwAAAAMgmAHAABgEAQ7AAAAgyDYAQAAGATB\nDgAAwCAIdgAAAAZBsAMAADAIgh0AAIBBEOwAAAAMgmAHAABgEAQ7AAAAgyDYAQAAGATBDgAA\nwCAIdgAAAAZBsAMAADAIgh0AAIBBEOwAAAAMgmAHAABgEAQ7AAAAgyDYAQAAGATBDgAAwCAI\ndgAAAAZBsAMAADAIgh0AAIBBEOwAAAAMgmAHAABgEAQ7AAAAgyDYAQAAGATBDgAAwCAIdgAA\nAAZBsAMAADAIgh0AAIBBEOwAAAAMgmAHAABgEAQ7AAAAgyDYAQAAGATBDgAAXTp27NiKFSt+\n+uknrQuBGyHYAQCgP7W1tVu2bHnwwQfffffd+vp6rcuBuyDY4VLKycnx+K+cnJyGE/z8/OR7\n582bJ49kZWUpv1JbW3txz3vkyJFTp05ddNmXoXXr1t16663+/v7+/v7dunV77bXXtK6ocbZr\nlC0vL6927dp17tx53LhxO3fu1LpMV3j//fc9PDzatWt39uxZ8b9/OD/88IPd5PPnzyv3Xtyb\n26w/zP379yuTDx06dBGP0HLsPhzcpCo7Das6e/Zsu3btPDw8MjMzG86vr6+fP39+WFjYpEmT\nhBAvvvhiaGjo3Llza2pq7GbW1NSsXLly2LBhV199ta+vr7+/f+fOnS0Wy/bt2+1mrl+/Xqmh\nuLi4ZRYUrkCwg75VVlb+/e9/79atG59EznvzzTeTkpK++eab8vLy8vLy/fv3nz59Wuuimqe+\nvv78+fMHDx5877334uLili9frnVFLevChQspKSlCCIvFcsUVV2hdjj7o+sPhiiuumDhxohDi\n8ccfr6iosLv3gQcemDVrlm1gPX369PPPPz9u3DjbaSUlJTfddNPUqVM3b9587Nix6urq8vLy\ngwcPpqWlDRw48MEHH5QkyQXLAhcj2EHHzpw5c/3118+ZM+fChQta16InaWlp8g1/f//ExMSR\nI0cOHDhQ04qc0rp167Zt27Zt27ZNmzY+Pj7KuCRJ06ZN++WXXzSsraUtWbLk6NGjQoj7779f\n61r0wQAfDvJ7XVRU9PLLL9uOb9++PTU1VQjh5eU1cuRIIcTNN9/s6ekphPjggw82bNggT5Mk\nafTo0T///LPyi76+vraPs2LFioULF7bwQkADBDtobNCgQUf/q1WrVs363bNnz/76668tVJiB\nKRswUlJS0tLSPvroowEDBmhbkjM+++yz8+fPnz9/vry8vKqq6vTp01OnTpXvqq2tXbVqlbbl\ntZyqqip5d+pNN9100003ueZJ/8gf5qV6hD+iqQ8Hbatqlp49e954441CiNdee62qqkoZz8jI\nkG/MnTv3kUceEUJMnDjx6aeflgeVXbfZ2dm7du2Sb0+fPr24uLiysrKysjIzM9NkMsnj8+fP\nt31kGINbr9a4HPj6+l599dVaV3F5qayslG9cddVV2lbyRwQFBb366qsbNmw4efKkEGL37t1a\nV9RS0tPTjx8/LoQYO3asy570j/9huueftntW1ZQxY8bs3r375MmT77//fmJiojx47Ngx+Ua/\nfv3q6urk29OnT7/99ts7deqkLN23334r37j22muXLFki3/b19f3zn//822+/Pfzww0KIs2fP\n7tq1q1+/fi5bIrgAW+ygsaaOZS4pKZk2bdr111/v7+/fqlWr4ODguLi4ZcuWKUcHJyYmRkRE\nKPN79uzp4eHxyiuvKCN1dXVvvfVWfHx8hw4dfH19O3TocPfddzc8ZFj2448/jhkzpn379m3a\ntOnVq9fatWuFEP7+/nJhhYWF8rTXXntNHklMTNy9e3dMTIz8yPLmotra2mXLlvXt2zcwMLBV\nq1Zt2rTp3LnzAw88YHuIzzPPPCM/wjPPPHP69OnHHnvsmmuu8fPz69q165IlS+RDXlavXn3z\nzTe3bt36yiuvTExMPHLkiDOvpDPLO2zYMNsjox988EEPD4+oqKimHtPx8jr5pIMHD5YfRPl2\nEUK8/fbb8qCnp6ftoUJ33HGH3ek1Dnh7e19zzTXybavVqow780Zs2rRJfiL5W+3NN9+MiYlp\n27ZtYGDgiBEj8vPzGz6dkyuJbM+ePYmJieHh4T4+PlddddXo0aM/++wzuwdUXcllq1evlm/c\nddddqq+JA81aZAcnGezatWv06NHBwcFt2rTp3bv322+/3ejTNfoIqmuUM6+bEOLUqVOzZ8++\n4YYb/P39TSbTjTfemJKSoiQeBx8ODpbLyU+M5q45zqyNTRk+fLh844033lAGle1t//nPf5TB\n4ODgQYMGXXPNNfI+WSGEl5eXfOPYsWNffvml7cPee++9q/9L+QuCcUjApbNjxw5l1dqxY0fD\nCcpBHnPnzpVHPv30U+VXampq5MF9+/a1b9++0TV21KhR1dXVkiRNmDCh4b1Lly6VH+H3339v\navfiQw89VFtba1tVZmam7TFbyrS2bdvKtwsKCuSZy5Ytk0fuuOOOK6+8UpmclZVVV1d3zz33\nNPqMgYGBBw8elB9h9uzZ8uA999zTcMvBI488YrFY7AbDwsJOnz7t+JV3cnnj4+MbTujUqVNT\nD+tgeZ1/0qVLl8qD8fHxyiM/8MADyuTMzEx5sLy8XFlDvvvuO9U1qqKiIiAgQL537Nix8qCT\nb8THH38sD8bGxjac7+3tvW3btotbSSRJeuutt7y9vRsWMH36dGWOMyu5JEknTpyQv6pNJlN9\nfb3y67Z/ON9//73dK3Pu3Dnl3mXLll3EIjf6hykvWsM9mCNGjFBuK69wo4/geI1y5nWTJOm7\n775rdGNzu3btvvzyS8nhh0NTy+X8J0azXkYn18amqqqvr5dXMG9v7/Pnz8uDb775pjK5b9++\nwuZzz+5VUqZ5eHjccccdr7/++oEDBxrOlCmH3gohjh492tQ0uD+CHS4l269hHx8f3waUex0H\nu2HDhskjt95664svvvjKK69YLBYPDw95MDU1VZKkf/7zn3feeafyu3fccUdCQsKmTZskSaqr\nq7P9jA4JCbn11luVL2AhxJNPPqnUfOTIEX9/f6XmUaNGjRs3TokLsobBzlZAQEB1dfW7776r\nfIYOHz78wQcfHDp0qFLz1KlT5UdQgp1syJAh999/v90XfPv27e+7777evXsrI88//7yDl935\n5X3ppZcmTJjQpk0beTwmJmbChAkpKSlNPbKD5XX+SQ8fPiyPtG7durKyUh7s3LmzMnPatGny\n4KZNm+SRjh07NlyjlGBXX19fXl7+yy+/3H333cq969evl+918o1Qvp7luyIiIiZPnhwXF6c8\n4MCBAy9uJfnuu++UCNilS5e//vWvgwcPVqatXr3a+ZVckiTlkKnbbrvN9q35I8HOmUVu9A/z\n0KFDrVu3lge9vb1Hjhx57733BgUF2b4OTgY7W/Ia5eTrVlZWpvyPyNvbe8CAAXfddZefn588\nEhYWdv78eQcfDo1W1axPjGa9jE6ujU0FO0mSYmJi5PFPP/1UHrlw4UKnTp1sX8Dw8PBp06bt\n27fPbjVQ9t7a6tix49SpU3fu3Gk3mWBnGAQ7XEq2X8OOOQ52ypfozz//rDz4888/P3jw4GnT\npsn/uZckqaCgQPld2+8220+o2bNn19XVSZJUVlb2pz/9SR709PRUHvmpp56SB728vHJycuTB\nX3/9tWPHjsqDNBrsevXqVVBQcOLEiU8++USSpMWLF994443e3t5z5sxRKnnwwQflyf369ZNH\nbIPd4sWL5cGtW7cqg2FhYb/99pskSXV1dcpn+ogRIxy87M1aXkmSOnToII+//vrrjt9QB8vb\nrCeNjo6WB7ds2SJJUklJie3K0LNnT3mafCS4EOJvf/ubPOLkGjVw4EC5AOffCOXrWQgxdOjQ\nqqoqeXzMmDHyYNu2bZVfb9ZKMnr0aHnkrrvuUlZp5WJy1113nbzhzcmVfM6cOfK0yZMn2741\nfyTYObPIjf5hTps2TXkdlJxdVFRkuzvP+WBnt0Y5+bopJ3K2bt06Ly9PnrZ7924l261YsUJq\n+sOh0aqatTI362V0cm10EOzuu+8+eVz5zJQkad++fbb7mpUip0+frvwhSJJUWVkpXzOlUcOH\nDz9+/LgymWBnGAQ7XEqXKtgpySM8PDwlJSUrK+vs2bMNn66pz27lP/p9+vSxnX/q1Cnl8JSn\nn35aHrz55pvlkXHjxtlOfv3115UHbzTYbd26tWFJ1dXVygf9uXPnlKTSo0cPeVAJdu3atVOW\nV77krGzmzJnKoym/HhcX5+Blb9byShcb7OyWt1lPOnPmTHlkxowZkiS99dZb8o/yl7Gnp2dZ\nWZlksxnvs88+k39RdY3y9va+//77z507Z1e56hth+/Ws5ANJkmyPGFMewfmVpKqqSkkYyiYW\nSZJqa2uVbV3y7jAnV3Jlr+Ls2bNtx7OyspSndhzsli9ffhGL3OgfZvfu3eWRxMRE26ezXU+c\nD3a2a5Tzr5u881EIMWnSJNsaVq1alZ6evmfPHvm5mhXsmrUyN+tllKmujQ6CnXK665QpU2zH\nz507N3/+fLtNd6LBeiJJ0q5duyZPntzozuuePXteuHBBnkawMwxOnkBLcXyMnWN//etf5Rsl\nJSWLFy8eNmxYYGDgbbfdtnjx4rKyMtVf//777+Ubtof+CCGCg4Nvu+02+fY333wj3zh48KB8\no1evXraTlZlNUb5gbJWVlWVkZPztb3+LiYkJCgpSvsYaNvyJjIxUjlWy3emjbNwSQigbdaqr\nqx1U0qzlvWh2y9usJ1XmyAfCb9u2Tf5xypQpQoj6+vqcnJyCgoIDBw4IIUwmU6PX1evSpUv3\n7t2V10QI0a9fv6NHj/773/+2HZQ5/0YIIbp27arctt2xqJzE4PxKcvjwYeWk4zvvvFM5Tr9V\nq1bKZWblS4s5uZKfOXNGvmF3XWLl0HghhNTgMrO2i2k70/lFbpRyARG710HZtNwstmuU86+b\ncmG2Hj162D7aX//613vvvff666+/iIuYXPRfkJMvY7PWRjtKsjxx4oTtuL+//8yZM1esWCGE\niIuLU3LbK6+8Ynfpvl69er3xxhslJSU//PDDokWL5JOZlAVfs2aNag3QF4Id3NGzzz47Y8YM\n28Oo6+rq8vLynnjiiaioKCUWNKq+vl75XgwPD7e7VxmR50g22zbsDnRrmBVsmUwmu5B64cKF\nyZMnh4eHT5gwYfny5bt27aqvrw8LC5PvVU5VU9iGOdt727Vr1+h4U5q1vBfNbnmb+6R9+vQJ\nCQkRQuzevfv48eNffPGFEOLqq69WTqH48ssvN2/eLN++8847Gz2CftWqVT/99NOpU6fkHgxC\niJycHIvFYvc11tw3QvxvZrK79LFo5kpiu/G1KfIcJ1dyJdPbrjBCCOVASWFz/RqF7cXJlC1h\nthwvcqMkSVJearvXwXaldZLdGuX866bMVLbk/UF/5C9I9WW8iLXRjvK+nz9/vqk5f/nLX77/\n/nv5TSkvL9+zZ48QorKy8tdff83JyZEToYeHR48ePVJSUrZu3fr1118rlSsHtsIwCHZwR15e\nXi+99FJJScmKFStGjhxpe5T66dOnJ0yYoFy9qSFPT0/lW8fuWC5hcwko+b/XHh4eyhdkaWlp\nozMbZfctK4R49NFHU1NTa2trAwICZsyY8dlnn5WWlioHSDX8BFf+02ynudsbmrW8F81ueZv7\npJ6envKlOiRJWrt2rdxOdNCgQdHR0fKpkdu3b1f2LY4aNcpBJb6+vosWLVJOm/jss8+SkpJs\nJzT3jWhqUNGslcQ2b23fvv1cY+69917h9EoeGBgoD9pdSNY2WjW8II7t1h3lEZxf5EZ5eHgo\nQVbZjij7/fffm/toDnKq49dNSSR2NTj4THDsj/wFqb6MF7E22lHynPw+njt3bsmSJSkpKRaL\nxXaRr7rqKuWcd7lVYNu2bTt16hQXF2d7wICsV69eQ4YMkW83XGToHcEO7qiuru7gwYPffvvt\nlClTPvroo9OnT//444+TJ0+W7y0pKVFOtFTYbmlQ9mbaHg0jhDh9+nReXp58WznntEuXLvIN\nu0s9NXr1LIVd/KqoqFi3bp18e+3atS+99NKQIUOUfu3ior5Hndes5b04DeNmc59U7n0khFAO\nfpf3tw4aNEgI8e2338pnkHh7e9uez9iUlStXKseoZWRkKFd6a6E3wvmVpFOnTsprtX//fv//\nqq2t/eSTT4qKinx9feUJTq7k8pZO0WCbltlsVvaxyqfQ2t5ruxnm2muvvYhFbpSy53Hnzp22\n43YvizPs1ijnXzdlxVNWM9k999zToUOHIUOGNOxB4mAzpKyF/oIuydqobCkMDg4WQvj5+c2e\nPXvJkiVpaWkffvih7Uwl31999dVt27ZVFio1NdXuPySVlZXKxVBsLzoDYyDYwe1s3bpVvobn\nsGHDsrOzhRCenp433nij7ZXP5P/u2+77kC9yK19xVP6fvRDiq6++euaZZ+QDWc6ePZucnCxf\nxrZVq1bKyWLKJUA/+ugj5aD+bdu22V7ruCG77W2nTp1qeHBSRUXFO++8I9+2uxTqpdWs5b04\nDbcvNvdJhw4dKu96Uza0yJFO/re2tlbeMjFgwADloCIHAgICbC/ZmpKSIvdmaKE3wvmVpE2b\nNsqR+AsXLlSuvfzPf/7znnvu6d69+5VXXnnu3DnnV3Kz2Sz/aLdZztfXVzkSccuWLffcc8+3\n335bVlZWWFi4dOnSZ599Vr4rPDxcOePhj1Myd0ZGhnK8//fff+/4j6VRdmuUk6+bsPkfwqZN\nm5S3NS8vb+PGjSUlJZ9//rm8AjT14dCoFvoLuiRro7xii/+mam9vb+Vc3alTp8rXQ66vr1++\nfLl8plHnzp2vu+46IYRyRcyjR48OGTJE3lhotVp37tw5fPhw5eSSRq9tCX1z0UkauDxckgsU\nV1dXd+vWTR7x9/d/6KGHFi9ePGPGDOXAFOUaARUVFcpGC5PJdN11173wwguSJFVWViqnMQoh\nQkJCYmJibA+Hsr30wG+//WZ7hFB0dPQNN9xg9z/phmfFms1m2+WqqalR9tSEhoa+9NJLixYt\nUjbzCCEiIyPlmcpZsX379rV9BGXmxx9/rAwqk3v37u3gZW/W8koXdVas3fJexJNKkmS7Ke6a\na66RB5XzEmTKtTlkjteoSZMmKffee++9UnPeCNttM7aPKR//J1NOtm3WSvLll18qqaVDhw5/\n+9vf7rnnHmWyfNaz8yu58go0PDN6586dqpt8bF/PZi1yo+dp/vbbb8qOXU9Pz/79+w8aNMju\nYFMnz4ptuEY587pJ/3sdOyFEVFRUr169lM+BDh06yBfyberDodGqmrUyO/8yOr82OjgrVjlP\nJTc3Vx7ZvXu33bWybY/U/PDDD+VpFRUVts/VqIiICOUd56xYwyDY4VK6VJ0n9u/f31SjG7PZ\nfPjwYeUB5e09CuWKAEeOHOnTp0+jjzBt2jTbK/hLkvTRRx/Z7Rjy9PR85plnlB8LCwvlmQ6+\nlpYvX97wuZSl8Pb2li8r0BLBrrnLe0mCXXOfVJKkf/3rX8qEpKQkZdz2anBFRUW2v+J4jSor\nK7P9XfkiKU6+Ec1KOVJzVhJJkl577bVGI9fo0aObu5JXVlbKfzXt2rWzvUSZ7K233mr03AiZ\n3bvwx4OdJEmbN29ueNbCU089Je8oFH8g2Dn5ukmS9O233za6AzEgIMD2+iONfjg0tVzOr8zN\nehmdXBubqqqurk4++tDHx0e5LokkSRkZGQ3fBT8/P/kafoqCgoIbbrih0YUSQnTq1Mn2msYE\nO8Mg2OFSulTBTpIkq9X64osv9u7d22QyeXl5XXHFFb169XruuefkC54pjh49mpCQcMUVV7Ru\n3bpr167/+te/lLtqamrWrl0bHx8fFhbm4+MTHh4+bty4RquSJCk/P/+uu+4KCAjw9/cfNGjQ\n9u3bf/rpJ6Uw+YrBktrX0jvvvBMTE+Pn5+fr69u1a9dZs2adPXtW+RB/9913pRYLds1a3ksV\n7Jr1pJIk2TbHVBoJSJKk7DO66aab7H5FdY1SzqUVQkRFRVVUVEjOvRHNDXaS0yuJ7JtvvklM\nTLzmmmt8fX1DQ0Pj4uLWr19v18vOyZVc6VGxe/fuhq/Ar7/++vDDD3fv3r1du3aenp7+/v6d\nOnWyWCzKhZQVlyTYSZK0d+/eMWPGBAUFtWnTpm/fvhs2bJAk6ZIEOydfN0mSTpw48dRTT0VH\nR7du3drX17dLly6PPvpocXGx7ZxGPxwcLJeTK3Nz1xxn1samqvrxxx/lwWHDhtmVUVhY+OST\nT8q5LTo6+tlnn7X9r4WiqqrqzTffHD58uLxQrVu37tix45133vnaa6/JfywKgp1heEhqR5UC\nxnbs2DEvL68rr7zS7qCfXbt2yVfn8vDwqKiocPIKfDAkbVeSDRs2yF0NXn755SeeeKIlngLu\naenSpY8//rgQYu3atQ27SAshPv/88yFDhixdulRpCgJw8gQud4sXLw4LC/P19Y2IiLD9P/cn\nn3wi34iKiiLVXea0XUlGjhwZGhoqGpyzCcOTz3sNDg62bYsMONbsK3QDBtO/f/+lS5fW1NQU\nFhb+5S9/kRuK//zzz8plDhptpI3LirYribe392OPPTZ79uwdO3YUFhZewsuXwJ0VFBTIhyJM\nmzatqSMp77jjDna7wQ67YgExbty49957r9G7evfu/cUXX1yqa9xDv7RdSeQzHI8ePTpz5sz5\n8+e33BPBfTz99NMLFiwIDQ09dOjQRfT2wGWLYAcISZLS09PffvvtH3/88eTJk3V1dQEBAd27\ndx8zZsz9999vd2UBXJ40X0nWr18/ceLE4ODgI0eO2PZpgCFduHChY8eOZ86cWb16dXJystbl\nQE8IdgAAAAbByRMAAAAGQbADAAAwCIIdAACAQRDsAAAADIJgBwAAYBAEOwAAAIMg2AEAABgE\nwQ4AAMAgCHYAAAAGQbADAAAwCIIdAACAQRDsAAAADIJgBwAAYBAEOwAAAIMg2AEAABgEwQ4A\nAMAgCHYAAAAGQbADAAAwCIIdAACAQRDsAAAADIJgBwAAYBAEOwAAAIMg2AEAABgEwQ4AAMAg\nCHYAAAAGQbADAAAwCIIdAACAQRDsAAAADIJgBwAAYBAEOwAAAIMg2AEAABgEwQ4AAMAgCHYA\nAAAGQbADAAAwCIIdAACAQRDsAAAADIJgBwAAYBAEOwAAAIMg2AEAABgEwQ4AAMAgCHYAAAAG\nQbADAAAwCIIdAACAQRDsAAAADIJgBwAAYBAEOwAAAIMg2AEAABgEwQ4AAMAgCHYAAAAGQbAD\nAAAwCIIdAACAQRDsAAAADIJgBwAAYBAEOwAAAIMg2AEAABgEwQ4AAMAgWmldwMUoLS21Wq0V\nFRU+Pj7BwcEBAQFaVwQAAP6/ffv2HTt2TOsq7F1//fVhYWFaV9GyPCRJ0roGZ2VlZaWmpm7b\ntu3UqVO248HBwQMGDHj44YcHDx6sVW0AAEARHR1dtG+fr9Zl2CoXYuJf/7pq1SqtC2lZ+thi\nV1NTM378+IyMDCFEZGRkjx49TCaTr69vVVWV1WotKCjIzMzMzMy0WCyrVq3y9vbWul4AAC5r\ntbW1rwhxv9Zl2JokRF1dndZVtDh9BLv58+dnZGSMGzdu4cKFZrO54YSioqKZM2euW7eue/fu\nM2bMcH2FAAAAmtPHyRNr1qyJiYlJT09vNNUJIcxmc3p6emxsbGpqqotrAwAAcBP6CHYlJSX9\n+/f39HRUrYeHR1xcXFFRkcuqAgAAcCv6CHZmszk/P9/xHEmScnJyDH+2CwAAQFP0EeySk5N3\n7NhhsVgKCgoanVBcXGyxWHJzcxMTE11cGwAAgJvQx8kTKSkpe/fuTUtLS0tLM5vNUVFRAQEB\nvr6+1dXVZWVlhYWFhw4dEkKMGjVq9uzZWhcLAACgDX0EOx8fn/Xr10+cOHHlypV5eXlbt261\nvTc0NDQhISEpKWnEiBFaVQgAAKA5fQQ7WXx8fHx8vBDCarVardby8nI/P7/AwEA6TwAAAAi9\nHGNnp76+vr6+Xr6hdS0AAADuQk/BLisra+zYsSEhIUFBQREREdHR0VFRUYGBge3bt09ISMjO\nzta6QAAAAC3pY1csLcUAAABU6SPY0VIMAABAlT52xdJSDAAAQJU+gh0txQAAAFTpI9jRUgwA\nAECVPoIdLcUAAABU6ePkCVqKAQAAqNJHsKOlGAAAgCp9BDsZLcUAAAAc0McxdnZoKQYAANCQ\nnoIdLcUAAAAc0MeuWFqKAQAAqNJHsKOlGAAAgCp97IqlpRgAAIAqfQQ7WooBAACo0kewo6UY\nAACAKn0EO1qKAQAAqNLHyRO0FAMAAFClj2BHSzEAAABV+gh2spZoKXbkyJEhQ4bU1tY6mFNb\nW3v69Gmr1erl5XXRTwQAANDS9BTsFJewpVh4ePiCBQscB7tffvllzpw5dXV1BDsAAODO9BTs\nsrKyUlNTt23bdurUKdvx4ODgAQMGPPzww4MHD27uY7Zq1erPf/6z4zm5ublz5sxp7iMDAAC4\nmD6CHS3FAAAAVOkj2NFSDAAAQJU+rmNHSzEAAABV+gh2tBQDAABQpY9gR0sxAAAAVfoIdrQU\nAwAAUKWPkydoKQYAAKBKH8GOlmIAAACq9BHsZC3RUgwAAMAw9BTshBCSJNXV1ZlMJpPJ1PDe\n8+fP19bWkvMAAMDlSR8nTwghfvnllxEjRrRu3drHx+emm25as2aNJEl2c/70pz8FBgZqUh4A\nAIDm9BHsDh8+3KdPn02bNnXo0CE6OnrPnj2TJk0aPXp0eXm51qUBAAC4C30Eu830CgEAACAA\nSURBVGeffbasrGzNmjWHDx/es2fP3r17hwwZsnHjxqFDh54/f17r6gAAANyCPoLdli1bhg8f\nnpSUJP/YpUuXrKysKVOm5Obmjh49urq6WtvyAAAA3IE+gl1paWmXLl1sRzw9PV9//fVJkyZt\n3bo1KSmp4fF2AAAAlxt9nBV75ZVX/vjjj3aDHh4e//73v48fP/7OO++Eh4cvXrxYk9oAAADc\nhD622A0fPjw7O3vp0qX19fW2461atXr//fd79+69ZMmSBx54oLKyUqsKAQAANKePYDdv3rzw\n8PDHH388LCxszZo1tne1adPm008/jY2NfeONN/Lz8zUqEAAAQHv6CHbh4eG7du2aPHlyXV2d\n3UY7IURgYGB2dvb06dN9fHw0KQ8AAMAd6CPYCSHCw8PfeOONU6dOKefG2vLz81uyZElRUdEH\nH3zg+toAAADcgW6CncLLy6upu6666qrRo0e7shgAAAD3ob9gBwAAgEYR7AAAAAxCH9exW7Ro\nkfOTn3jiiZarBAAAwG3pI9gtWLDg9OnTTk4m2AEAgMuTPoLd/v37x4wZs3379iFDhjz00ENa\nlwMAAOCO9BHs2rdvn5WVNXjw4M8//3zWrFkDBw7UuiIAAAC3o5uTJ/z8/DZs2GAymaZOnVpb\nW6t1OQAAAG5HN8FOCBEWFrZgwYJWrVp98cUXWtcCAADgdvSxK1YxZcqUKVOmaF0FAACAO9LT\nFjsAAAA4QLADAAAwCIIdAACAQRDsAAAADIJgBwAAYBAEOwAAAIMg2AEAABgEwQ4AAMAgCHYA\nAAAGQbADAAAwCIIdAACAQRDsAAAADIJgBwAAYBAEOwAAAIMg2AEAABgEwQ4AAMAgCHYAAAAG\nQbADAAAwCIIdAACAQRDsAAAADIJgBwAAYBAEOwAAAIMg2AEAABgEwQ4AAMAgCHYAAAAG0Urr\nAi5GaWmp1WqtqKjw8fEJDg4OCAjQuiIAAADt6WmLXVZW1tixY0NCQoKCgiIiIqKjo6OiogID\nA9u3b5+QkJCdna11gQAAAFrSxxa7mpqa8ePHZ2RkCCEiIyN79OhhMpl8fX2rqqqsVmtBQUFm\nZmZmZqbFYlm1apW3t7fW9QIAAGhAH8Fu/vz5GRkZ48aNW7hwodlsbjihqKho5syZ69at6969\n+4wZM1xfIQAAgOb0sSt2zZo1MTEx6enpjaY6IYTZbE5PT4+NjU1NTXVxbQAAAG5CH8GupKSk\nf//+np6OqvXw8IiLiysqKnJZVQAAAG5FH8HObDbn5+c7niNJUk5OTlhYmGtKAgAAcDf6CHbJ\nyck7duywWCwFBQWNTiguLrZYLLm5uYmJiS6uDQAAwE3o4+SJlJSUvXv3pqWlpaWlmc3mqKio\ngIAAX1/f6urqsrKywsLCQ4cOCSFGjRo1e/ZsrYsFAADQhj6CnY+Pz/r16ydOnLhy5cq8vLyt\nW7fa3hsaGpqQkJCUlDRixAitKgQAANCcPoKdLD4+Pj4+XghhtVqtVmt5ebmfn19gYCCdJwAA\nAIRejrGzU19fX19fL9/QuhYAAAB3oadgR0sxAAAAB/SxK5aWYgAAAKr0EexoKQYAAKBKH7ti\naSkGAACgSh/BjpZiAAAAqvQR7GgpBgAAoEofwY6WYgAAAKr0cfIELcUAAABU6SPY0VIMAABA\nlT6CnYyWYgAAAA7o4xg7O7QUAwAAaEhPwY6WYgAAAA7oY1csLcUAAABU6SPY0VIMAABAlT52\nxdJSDAAAQJU+gh0txQAAAFTpI9jRUgwAAECVPoIdLcUAAABU6ePkCVqKAQAAqNJHsKOlGAAA\ngCp9BDtZS7QUO3v27LPPPltRUeFgzvHjxy/68QEAAFxGT8FOcQlbilVVVZ08ebKmpsbBnHPn\nzgkhJEn6g88FAADQovQU7LKyslJTU7dt23bq1Cnb8eDg4AEDBjz88MODBw9u7mOGhIS89dZb\njufk5uZmZ2d7eHg098EBAABcSR/BjpZiAAAAqvQR7GgpBgAAoEof17GjpRgAAIAqfQQ7WooB\nAACo0kewo6UYAACAKn0EO1qKAQAAqNLHyRO0FAMAAFClj2BHSzEAAABV+gh2spZoKQYAAGAY\n+jjGzs4lbCkGAABgGHoKdllZWWPHjg0JCQkKCoqIiIiOjo6KigoMDGzfvn1CQkJ2drbWBQIA\nAGhJH7tiaSkGAACgSh/BjpZiAAAAqvSxK5aWYgAAAKr0EexoKQYAAKBKH8GOlmIAAACq9BHs\naCkGAACgSh8nT9BSDAAAQJU+gh0txQAAAFTpI9jJaCkGAADggJ6CncJkMplMJq2rAAAAcC/6\nOHkCAAAAqgh2AAAABqGPXbGLFi1yfvITTzzRcpUAAAC4LX0EuwULFpw+fdrJyQQ7AABwedJH\nsNu/f/+YMWO2b98+ZMiQhx56SOtyAAAA3JE+gl379u2zsrIGDx78+eefz5o1a+DAgVpXBAAA\n4HZ0c/KEn5/fhg0bTCbT1KlTa2trtS4HAADA7egm2AkhwsLCFixY0KpVqy+++ELrWgAAANyO\nPnbFKqZMmTJlyhStqwAAAHBHetpiBwAAAAcIdgAAAAZBsAMAADAIgh0AAIBBEOwAAAAMgmAH\nAABgEAQ7AAAAgyDYAQAAGATBDgAAwCAIdgAAAAZBsAMAADAIgh0AAIBBEOwAAAAMgmAHAABg\nEAQ7AAAAgyDYAQAAGATBDgAAwCAIdgAAAAZBsAMAADAIgh0AAIBBEOwAAAAMgmAHAABgEAQ7\nAAAAgyDYAQAAGATBDgAAwCAIdgAAAAZBsAMAADAIgh0AAIBBEOwAAAAMgmAHAABgEAQ7AAAA\ng2ildQEXo7S01Gq1VlRU+Pj4BAcHBwQEaF0RAACA9vS0xS4rK2vs2LEhISFBQUERERHR0dFR\nUVGBgYHt27dPSEjIzs7WukAAAAAt6WOLXU1Nzfjx4zMyMoQQkZGRPXr0MJlMvr6+VVVVVqu1\noKAgMzMzMzPTYrGsWrXK29tb63oBAAA0oI9gN3/+/IyMjHHjxi1cuNBsNjecUFRUNHPmzHXr\n1nXv3n3GjBmurxAAAEBz+tgVu2bNmpiYmPT09EZTnRDCbDanp6fHxsampqa6uDYAAAA3oY9g\nV1JS0r9/f09PR9V6eHjExcUVFRW5rCoAAAC3oo9gZzab8/PzHc+RJCknJycsLMw1JQEAALgb\nfQS75OTkHTt2WCyWgoKCRicUFxdbLJbc3NzExEQX1wYAAOAm9HHyREpKyt69e9PS0tLS0sxm\nc1RUVEBAgK+vb3V1dVlZWWFh4aFDh4QQo0aNmj17ttbFAgAAaEMfwc7Hx2f9+vUTJ05cuXJl\nXl7e1q1bbe8NDQ1NSEhISkoaMWKEVhUCAABoTh/BThYfHx8fHy+EsFqtVqu1vLzcz88vMDCQ\nzhMAAABCL8fY2amvr6+vr5dvaF0LAACAu9BTsKOlGAAAgAP62BVLSzEAAABV+gh2tBQDAABQ\npY9dsbQUAwAAUKWPYEdLMQAAAFX6CHa0FAMAAFClj2BHSzEAAABV+jh5gpZiAAAAqvQR7Ggp\nBgAAoEofwU5GSzEAAAAH9HGMnR1aigEAADSkp2BHSzEAAAAH9LErlpZiAAAAqvQR7GgpBgAA\noEofwU5pKdZU8wm5pVhRUVFqamqzgp0kSTt37qysrHQwZ+/evc0rFwAAQAv6CHYlJSUJCQnO\ntBR79dVXm/XIBQUFgwcPrqmpUZ0pSVKzHhkAAMDF9HHyRMu1FIuMjKyurpYc2rlzpxDCw8Pj\n4hcAAACg5ekj2NFSDAAAQJU+dsXSUgwAAECVPoIdLcUAAABU6SPYyWgpBgAA4IA+jrGzQ0sx\nAACAhvQU7GgpBgAA4IA+dsXSUgwAAECVPoIdLcUAAABU6WNXrNJSrNFUJ/7bUiw2NjY1NdXF\ntQEAALgJfQS7kpKS/v37O9NSrKioyGVVAQAAuBV9BLuWaykGAABgGPoIdrQUAwAAUKWPkydo\nKQYAAKBKH8GOlmIAAACq9BHsZLQUAwAAcEBPwU5hMplMJpN8u7CwcP/+/VdfffXVV1+tbVUA\nAADa0sfJE0KICxcuvPrqq8nJyXPmzPntt9+EEPv27bv11lsjIiJiY2M7duzYv3//gwcPal0m\nAACAZvSxxa6srCwuLm7Pnj3yj2vWrNm2bdvtt99+/Pjx+Pj4iIiIffv2bd++vV+/ft9++y2b\n7gAAwOVJH1vsnnnmmT179jzyyCPffPNNenp6ZWXlgAEDSktLt23blpWV9frrr2/btu3dd989\nefLkvHnztC4WAABAG/rYYvfhhx8OGjTo1VdfFULccsst1dXVycnJKSkpcXFxypy77747LS3t\n008/1a5MAAAALelji93x48d79eql/HjTTTcJIbp37243rVu3bidPnnRpZQAAAG5DH8Huqquu\n+vrrr5Uf5ds//fST3bTdu3eHhoa6tDIAAAC3oY9gl5CQsH379gcffHDHjh0rV6588sknIyMj\nly9fvmXLFmXOqlWrNm/ePHz4cA3rBAAA0JA+jrF74YUXcnNzV6xYsWLFCiFE27Zt33nnnWee\neWbo0KG9e/e+9tpr9+7du2fPntDQ0GeffVbrYgEAALShj2DXrl27nTt3pqen79q1KygoaOLE\nidddd9277747efLkzMzM/Px8Dw+P22+/fcWKFVzrBAAAXLb0EeyEEN7e3klJSUlJScpIQEBA\nRkbGuXPniouLQ0NDg4ODNSwPAABAc+rH2E2YMGHz5s11dXUuqOYitGvXrlu3bqQ6AAAA9WD3\n9ttvDxs27Oqrr05JSfnhhx9cUBMAAAAugnqwKygoePHFF4OCgpYsWdKzZ88bb7zx5ZdfPnbs\nmAuKAwAAgPPUg53ZbJ41a9bevXu/++67lJSU06dPP/nkk9dcc82QIUPWrVt3/vx5F1QJAAAA\nVc24jl3Pnj0XLVp09OjR7du3p6Sk/PLLL0lJSVdeeaXFYvnyyy9brkQAAAA4o9kXKD58+HBO\nTs6OHTvkvbFms/n9998fMGDA7bfffuLEiRaoEAAAAE5xNtiVlJQsXbr01ltv7dy58+zZs48c\nOfL444//9NNPP//88++///70009nZ2fbXosEAAAALqZ+HbtVq1alp6dv3769vr6+devW9957\nb1JS0h133OHl5SVPMJlM//jHP9599112yAIAAGhIPdjdf//9Hh4ecXFxSUlJY8eObdeuXaPT\nwsPDb7jhhktdHgAAAJylHuyef/55i8ViNpsdT9uxY8clKgkAAAAXQ/0Yuzlz5oSFhaWnp+fk\n5CiDEydOXLdunSRJLVkbAAAAmkE92Fmt1ttvv33ChAmffvqpPFJeXp6enp6UlHTXXXdVVFS0\ncIUAAABwinqw+8c//pGTk/Poo48+9NBD8kjbtm2PHz/++OOPf/rppy+88EILVwgAAACnqAe7\nDRs2DBky5J///GeHDh2Uwfbt2y9evHjQoEHp6ektWR4AAACcpR7sjh071qNHj0bvuvnmm0tK\nSi51SQAAALgY6sEuLCxs165djd61e/fuq6666lKXBAAAgIuhHuz+8pe/bN++/eWXX7Y7B3bZ\nsmVbtmwZNWpUi9UGAACAZlC/jt2sWbM++OCDJ598cvny5b17977iiivOnj27a9eugoKCTp06\nzZ071wVVAgAAQJV6sAsKCvrqq69mzZr1/vvvv/fee/Jg27Zt77vvvoULF7Zv376FKwQAAIBT\n1IOdECIkJOSNN954/fXXjxw5Ulpa6u/v36lTp1atnPpdAAAAuEYzwlmrVq0iIyNbrhQAAAD8\nEU4Fu++++27VqlWHDx+urq5u2EZs27Ztl74uAAAANJN6sPvkk09GjhxZX1/vgmoAAABw0dQv\nd/L8888HBwf/5z//KS0trWmMC6oEAACAKvUtdj/++OOLL7545513uqAaAAAAXDT1LXZt2rQx\nmUwuKAUAAAB/hHqwu/322zMzM11QCgAAAP4I9V2x8+fP79u373333Td58uSIiAhvb2+7CVyj\nGAAAwB2oB7t+/fqVl5evXr169erVjU5oeAEUAAAAuJ56sOvTp48L6gAAAMAfpB7sPvzwQxfU\nAQAAgD9I/eQJxfnz57/66qusrCwhhNVqZQ8sAACAW3Eq2P3+++/33ntvYGBgbGysfEG7ZcuW\nde3aNScnp4XLAwAAgLPUg93x48f79Onzzjvv3HrrrT179pQH/f39Dx8+PHTo0J9++qmFKwQA\nAIBT1IPdvHnziouLN27cmJubO3LkSHlw2rRpn3/+eW1t7QsvvNDCFQIAAMAp6idPbNy4MSEh\nYcSIEXbjAwcOHDNmzI4dO1qmMEdKS0utVmtFRYWPj09wcHBAQIDrawAAAHA36lvsTp48aTab\nG70rLCzs5MmTl7qkJmVlZY0dOzYkJCQoKCgiIiI6OjoqKiowMLB9+/YJCQnZ2dkuqwQAAMAN\nqW+x69ChQ35+fqN37dy5s0OHDpe6pEbU1NSMHz8+IyNDCBEZGdmjRw+TyeTr61tVVWW1WgsK\nCjIzMzMzMy0Wy6pVqxr2xgAAALgcqAe7MWPGLFq0aP78+TNnzlQGa2tr586dm5+fP3369JYs\n7//Mnz8/IyNj3LhxCxcubHTzYVFR0cyZM9etW9e9e/cZM2a4oCQAAAB346F6ObqysrLbbrtt\n37591157rY+Pz4EDByZMmLBz587CwsLIyMivv/46ODi4pauMjIwMCQnJy8vz9Gxy37EkSX37\n9j1z5sz+/fsv7bPn5ub27du3qqrKx8fn0j4yAACG1Llz5xkHD96vdRm2JgkhkpOb6o9qGOrH\n2AUEBOzcufOBBx44ceLEgQMHhBBvvfXWsWPHxo8fn5ub64JUJ4QoKSnp37+/g1QnhPDw8IiL\niysqKnJBPQAAAG5IfVesECIwMHDlypXLli3bt29fWVmZv79/t27d2rRp09LFKcxmc1PH+Skk\nScrJyQkLC3NNSQAAAO6mGS3FfHx8evToMWDAgFtuucWVqU4IkZycvGPHDovFUlBQ0OiE4uJi\ni8WSm5ubmJjoysIAAADch/oWu9GjRzue8OGHH16iYpqUkpKyd+/etLS0tLQ0s9kcFRUVEBDg\n6+tbXV1dVlZWWFh46NAhIcSoUaNmz57d0sUAAAC4J/Vg99FHHzV1V0hIiGvOJ/Dx8Vm/fv3E\niRNXrlyZl5e3detW23tDQ0MTEhKSkpIaXkUZAADg8qEe7EpLS+1GysvLf/7553nz5lVVVdll\nrBYVHx8fHx8vhLBarVartby83M/PLzAwkM4TAAAAwplg1zA2BQQEdOjQITY29rrrrnvqqadW\nrFjRMrU1qb6+vr6+Xr7h4qcGAABwW804ecKOv79/fHz8Bx98cAmrcYyWYgAAAA44dbmTppw8\nefL8+fOXqhQHaCkGAACgSj3YVVZW2o3U1dWdPXv2nXfeycrK6tevX8sU9j9oKQYAAKBKPdi1\nbt26qbu8vb2ff/75S1pP49asWRMTE5Oent5U8wmz2Zyenl5UVJSamkqwAwAAlyf1YCefiGrL\nw8PDx8ena9eukyZN6tq1a8sU9j9KSkoSEhKcaSn26quvuqAeAAAAN6Qe7LKyslxQh2O0FAMA\nAFB18WfFuhItxQAAAFRdgpZitlqovRgtxQAAAFSpB7tvvvmmvLy8rKxMCOHr62symc6cOVNb\nW9vytf1/tBQDAABQpR7scnJyBgwYcOONNy5cuPDWW2/18vKqq6v74Ycfnnrqqe+//37z5s1R\nUVEuKFTQUgwAAMAh9WD32GOPBQQEbNmyxcfHRx7x8vK65ZZbNm3adPPNNz/33HMff/xxCxdp\nj5ZiAAAADamfPJGdnX333XcrqU7h5+c3bNgwVzbyoqUYAACAA+pb7Hx9fQ8fPtzoXb/++qu/\nv/+lLqkRtBQDAABQpR7sBg8enJaW9uc//9nu1IR///vfH3300dSpU1ustv+PlmIAAACqPCRJ\ncjzjwIEDsbGxZ86c6dOnT69evfz9/UtLS3Nycvbu3du5c+fc3Nzg4OCWrjIyMjIkJCQvL89B\n8wlJkvr27XvmzJn9+/df2mfPzc3t27dvVVVVw/3RAACgoc6dO884ePB+rcuwNUkIkZy8evVq\nrQtpWepb7OT0lpKSsnnz5q+++koe9PX1ve+++15++eWgoKAWrlCIlmwpVlhYGBsbW1VV5WCO\nfG0X1QQMAACgLfVgJ4To0qXLpk2bysrK9u/fb7VaAwMDb7jhhtatW7d0cYqWaynWsWPHlStX\nOg52v/zyy5w5czw8PJr1yAAAAC7mVLD7v6mtWgkhJEmKiYmxWq1+fn4uyzrJycmzZs2yWCzP\nPfdcREREwwnFxcVPP/10bm7unDlzmvXIXl5eI0eOdDznIh4WAADA9ZwKdr///vv06dMzMjKU\nnZLLli1LS0tLTU3t169fC1coBC3FAAAAnKAe7I4fP96nT5+ioqLY2NjKysrvv/9eCOHv73/4\n8OGhQ4fm5+ffcMMNLV0lLcUAAABUqQe7efPmFRcXb9y4ccSIEfPmzZOD3bRp02666aahQ4e+\n8MIL7733XsvXKQQtxQAAABxS7zyxcePGhISEhhvDBg4cOGbMmLy8vJYpzBFaigEAADSkHuxO\nnjzZ6DWBhRBhYWEnT5681CU1iZZiAAAADqjviu3QoUNTlxrZuXNnhw4dLnVJjaClGAAAgCr1\nYDdmzJhFixbNnz9/5syZymBtbe3cuXPz8/OnT5/ekuX9H1qKAQAAqFJvKVZWVnbbbbft27fv\n2muv9fHxOXDgwIQJE3bu3FlYWBgZGfn111/TUgwAANiipZhW1I+xCwgI2Llz5wMPPHDixIkD\nBw4IId56661jx46NHz/eNY1ihRAlJSX9+/d3pqVYUVGRC+oBAABwQ05doDgwMHDlypXLli3b\nt29fWVmZv79/t27d2rRp09LFKVqupRgAAIBhqG+xGzdu3LPPPiuE8PHx6dGjx4ABA2655RZX\npjohRHJy8o4dOywWS0FBQaMTiouLLRZLbm5uYmKiKwsDAABwH+pb7D7++OPS0lIXlOIALcUA\nAABUqQe7K6+88vz58y4oxQFaigEAAKhSD3arV69OSEhITk6eNGlSVFRU27Zt7Sa4rKMXLcUA\nAAAcUA92ycnJkiStXbt27dq1jU5QvWDKJWcymUwmkxDi3LlzRUVFZ86c6dixI9clBgAAlzn1\nYNe9e3cX1KHq4MGDc+fO7dy587x584QQhw8ffuyxx7Kysurq6oQQJpNp/Pjxf//734OCgjQu\nFAAAQCPqwW7Tpk0uqMOxn3/+OTY29uzZsw8//LAQ4tChQ3369Dl9+nSnTp1uueUWSZK+//77\n119/ffPmzV999VVISIjW9QIAAGig8cudbNq0KS8vz8WlODB79uwLFy5s3LjxtddeE0I89dRT\np0+ffuGFFw4cOPDuu+++9957v/zyyyuvvPLrr7/KV2YBAAC4DDUe7EaMGPH000/bjvzwww/b\ntm1zRUWNycvLGzt2rHLS65YtW3r27PnMM88ovSg8PT0fe+yx4cOHb9y4UasiAQAAtKV+gWLZ\ntGnTBg0a1KKlOHDu3Lnw8HDlRw8Pj+uuu67htG7dup05c8aFdQEAALgRZ4Odtrp37/7RRx9V\nVFTIPw4aNCg/P7+ystJ2Tl1d3WeffRYVFaVFgQAAANrTR7B77LHHDh06NGzYsIMHDwohFixY\ncObMmcTExFOnTskTiouLx44d+9NPP91///2aVgoAAKAZ9bNi3cH48eN//vnnf/zjH127dr35\n5pt79uw5cODADRs2bNy4sVOnTpIkHThwQJKkMWPGyKfNAgAAXIb0EeyEEH//+99Hjhy5fPny\nTz/99JtvvpEHa2pq9u/f7+Xl1bt376lTp1osFg8PD23rBAAA0Ipugp0QIiYmJiYmRghx4sSJ\nkpKS8vJyT09Pk8nUqVMnX19frasDAADQWJPB7tdff505c6btj0II2xHFggULWqIyB0JDQ0ND\nQ138pAAAAG6uyWB39OjRhQsX2g02HBFaBDsAAAA01HiwS0tLc3EdAAAA+IMaD3aJiYkurgMA\nAAB/kD6uYwcAAABVBDsAAACDINgBAAAYBMEOAADAIAh2AAAABkGwAwAAMAiCHQAAgEEQ7AAA\nAAyCYAcAAGAQBDsAAACDINgBAAAYBMEOAADAIAh2AAAABkGwAwAAMAiCHQAAgEEQ7AAAAAyC\nYAcAAGAQBDsAAACDINgBAAAYBMEOAADAIAh2AAAABkGwAwAAMAiCHQAAgEEQ7AAAAAyCYAcA\nAGAQBDsAAACDINgBAAAYBMEOAADAIAh2AAAABkGwAwAAMAiCHQAAgEEQ7AAAAAyildYFXIzS\n0lKr1VpRUeHj4xMcHBwQEKB1RQAAANrT0xa7rKyssWPHhoSEBAUFRUREREdHR0VFBQYGtm/f\nPiEhITs7W+sCAQAAtKSPLXY1NTXjx4/PyMgQQkRGRvbo0cNkMvn6+lZVVVmt1oKCgszMzMzM\nTIvFsmrVKm9vb63rBQAA0IA+gt38+fMzMjLGjRu3cOFCs9nccEJRUdHMmTPXrVvXvXv3GTNm\nuL5CAAAAzeljV+yaNWtiYmLS09MbTXVCCLPZnJ6eHhsbm5qa6uLaAAAA3IQ+gl1JSUn//v09\nPR1V6+HhERcXV1RU5LKqAAAA3Io+gp3ZbM7Pz3c8R5KknJycsLAw15QEAADgbvQR7JKTk3fs\n2GGxWAoKChqdUFxcbLFYcnNzExMTXVwbAACAm9DHyRMpKSl79+5NS0tLS0szm81RUVEBAQG+\nvr7V1dVlZWWFhYWHDh0SQowaNWr27NlaFwsAAKANfQQ7Hx+f9evXT5w4ceXKlXl5eVu3brW9\nNzQ0NCEhISkpacSIEVpVCAAAoDl9BDtZfHx8fHy8EMJqtVqt1vLycj8/v8DAQDpPAAAACL0c\nY2envr6+vr5evqF1LQAAAO5CT8GOlmIAAAAO6GNXLC3FAAAAVOkj2NFSkVCQ6QAAG9xJREFU\nDAAAQJU+dsXSUgwAAECVPoIdLcUAAABU6SPY0VIMAABAlT6CHS3FAAAAVOnj5AlaigEAAKjS\nR7CjpRgAAIAqfQQ7GS3FAAAAHNDHMXZ2aCkGAADQkJ6CHS3FAAAAHNDHrlhaigEAAKjSR7Cj\npRgAAIAqfQQ7paVYU80n5JZiRUVFqampzQp2ZWVlc+bMqa6udjDn+PHjzSsXAABAC/oIdiUl\nJQkJCc60FHv11Veb9ch1dXVWq7WystLBnHPnzjXrMQEAADShj2DXci3FgoOD161b53hObm4u\nZ2YAAAD3p4+zYmkpBgAAoEofW+xoKQYAAKBKH8GOlmIAAACq9BHsZLQUAwAAcEAfx9jZoaUY\nAABAQ3oKdrQUAwAAcEAfu2JpKQYAAKBKH8GOlmIAAACq9LErVmkp1miqE/9tKRYbG5uamuri\n2gAAANyEPoJdSUlJ//79nWkpVlRU5LKqAAAA3Io+gl3LtRQDAAAwDH0EO1qKAQAAqNLHyRO0\nFAMAAFClj2BHSzEAAABV+gh2MlqKAQAAOKCnYKcwmUwmk0n5cf/+/VdccUV4eLiGJQEAAGhO\nHydPONatW7dHH31U6yoAAAA0po8tdl999ZXjCWfOnFHm9OnTp+UrAgAAcDv6CHaxsbGOJ3zx\nxRfKHEmSWr4iAAAAt6OPYJeSkvLKK6/U1dUNHTq0Z8+edvcuXLiwS5cuo0eP1qQ2AAAAN6GP\nYLdo0aK77777vvvuy87Ovvnmm+fOnevn56fcu3Dhwu7duy9YsEDDCgEAADSnm5MnYmJivvvu\nu1mzZi1evLhHjx5ffvml1hUBAAC4F90EOyGEj4/Pc889980337Rr127gwIFTp049e/as1kUB\nAAC4Cz0FO9mNN96Yn58/f/78tWvXRkdHb9y4UeuKAAAA3IL+gp0QwsvL66mnnvrxxx8jIiJG\njRqldTkAAABuQZfBTta5c+cvv/zy1VdfHTBgwPXXX691OQAAABrTx1mxTfHw8HjkkUceeeQR\nrQsBAADQno632AEAAMAWwQ4AAMAgCHYAAAAGQbADAAAwCIIdAACAQRDsAAAADIJgBwAAYBAE\nOwAAAIMg2AEAABgEwQ4AAMAgCHYAAAAGQbADAAAwCIIdAACAQRDsAAAADIJgBwAAYBAEOwAA\nAIMg2AEAABgEwQ4AAMAgCHYAAAAGQbADAAAwCIIdAACAQRDsAAAADIJgBwAAYBAEOwAAAIMg\n2AEAABhEK60LAAAAf0h5eXl1dbXWVfyP+vp6rUu4TBHsAADQsbNnz4aEhLhbsINWCHYAAOhY\nZWVldXX1ZiGitK7EVletC7hsEewAANC9q4WI1LoGuANOngAAADAIgh0AAIBBEOwAAAAMgmAH\nAABgEAQ7AAAAg9DlWbGlpaVWq7WiosLHxyc4ODggIEDrigAAALSnpy12WVlZY8eODQkJCQoK\nioiIiI6OjoqKCgwMbN++fUJCQnZ2ttYFAgAAaEkfW+xqamrGjx+fkZEhhIiMjOzRo4fJZPL1\n9a2qqrJarQUFBZmZmZmZmRaLZdWqVd7e3lrXCwAAoAF9BLv58+dnZGSMGzdu4cKFZrO54YSi\noqKZM2euW7eue/fuM2bMcH2FAAAAmtPHrtg1a9bExMSkp6c3muqEEGazOT09PTY2NjU11cW1\nAQAAuAl9BLuSkpL+/ft7ejqq1sPDIy4urqioyGVVAQAAuBV9BDuz2Zyfn+94jiRJOTk5YWFh\nrikJAADA3egj2CUnJ+/YscNisRQUFDQ6obi42GKx5ObmJiYmurg2AAAAN6GPkydSUlL27t2b\nlpaWlpZmNpujoqICAgJ8fX2rq6vLysoKCwsPHTokhBg1atTs2bO1LhYAAEAb+gh2Pj4+69ev\nnzhx4sqVK/Py8rZu3Wp7b2hoaEJCQlJS0ogRI7SqEAAAQHP6CHay+Pj4+Ph4IYTVarVareXl\n5X5+foGBgXSeAAAAEHo5xs5OfX19fX29fEPrWgAAANyFnoIdLcUAAAAc0MeuWFqKAf+vvTsP\nrqo8/Ab+hCUs1ia8lC0uwYCCgNVRK6BYKm2dKlKpWO0gSAGVVqZv6wZqi8X+pIgKLl0NYqkV\nV0ZGixbBbQqKqIODiEBlMWpT8KWGDAay3vv+kV8xhJBglZxzTz6fv7zPvXPny5mTx+892wMA\nTcqMYmdJMQCAJmXGqVhLigEANCkzip0lxQAAmpQZxc6SYgAATcqMYmdJMQCAJmXGzROWFAMA\naFJmFDtLigEANCkzil0tS4oBADQiM66xq8eSYgAA+8ukYmdJMQCARmTGqVhLigEANCkzip0l\nxQAAmpQZxW7vkmIHWnyidkmxoqKiefPmfaZil0qlli9fXlVV1chn1q1b99niAgBEITOKXXFx\n8ahRow5mSbF77rnnM31zUVHROeecs2fPns8XEAAgeplx88ShW1LsmGOO2b17d7pRL7/88ufI\nDgDQTDKj2FlSDACgSZlxKtaSYgAATcqMYmdJMQCAJmVGsatlSTEAgEZkxjV29VhSDABgf5lU\n7CwpBgDQiMw4FWtJMQCAJmVGsbOkGABAkzLjVOzeJcUabHXhP0uKDR48eN68ec2cDQAgJjKj\n2BUXF3/9618/mCXFioqKmi0VAECsZEaxO3RLigEAJEZmFDtLigEANCkzbp6wpBgAQJMyo9hZ\nUgwAoEmZUexqWVIMAKARmVTsQgjpdLqmpiYnJycnJ2f/dz/55JPq6mo9DwBomTLj5okQwsaN\nG0eMGNGhQ4fs7OyTTjpp/vz56XS63mfOO++8Tp06RRIPACBymVHsNm/ePGjQoMWLFx9xxBH9\n+vV7++23x48fP3LkyLKysqijAQDERWYUu5tuumnnzp3z58/fvHnz22+/vW7dum9/+9tPPfXU\n2Wef/cknn0SdDgAgFjKj2C1btuzcc88dN25c7cs+ffosWbJk0qRJr7zyysiRIysrK6ONBwAQ\nB5lR7EpKSvr06VN3pFWrVn/4wx/Gjx///PPPjxs3bv/r7QAAWprMuCu2W7dua9asqTeYlZVV\nWFi4ffv2Rx55JC8vb/bs2ZFkAwCIicw4Ynfuuee+8MILd955ZyqVqjvepk2bxx9/fODAgXPm\nzLniiivKy8ujSggAELnMKHbTp0/Py8u7+uqre/ToMX/+/LpvdezY8W9/+9vgwYPnzp27atWq\niAICAEQvM4pdXl7e66+/ftlll9XU1NQ7aBdC6NSp0wsvvHDVVVdlZ2dHEg8AIA4yo9iFEPLy\n8ubOnbtjx46998bW1b59+zlz5hQVFS1atKj5swEAxEHGFLu9WrdufaC3unfvPnLkyOYMAwAQ\nH5lX7AAAaJBiBwCQEIodAEBCZMYDigFogZYtWzZ37tyoU9TXq1evmTNnRp0CGqbYARBTS5cu\nXfX449+JOkZd74Vwb6dOih2xpdgBEF9fDeHeqDPUtSiE16POAI1wjR0AQEIodgAACaHYAQAk\nhGIHAJAQih0AQEIodgAACaHYAQAkhOfYAcDB2hNCRUVFYWFh1EE+tWvXrqgjECOKHQAcrLUh\nlO/ePWvSpKiDfKoy6gDEimIHAAcrHUJOCJujjlHXOyH0jzoD8eEaOwCAhFDsAAASQrEDAEgI\nxQ4AICEUOwCAhFDsAAASQrEDAEgIxQ4AICEUOwCAhFDsAAASQrEDAEgIxQ4AICHaRB0AgFj4\n61//+vLLL0edYh9///vfu0adATKLYgdACCHMnDnz/61c2TPqGHVtCEGxg89EsQPgf00I4Yao\nM9R1WtQBIOO4xg4AICEUOwCAhFDsAAASQrEDAEgIxQ4AICEUOwCAhFDsAAASQrEDAEgIxQ4A\nICEUOwCAhLCkGEAEtmzZUlJSEnWKfZSVlUUdAfi8FDuA5pZOp48//vjKysqog9T3g6gDAJ+T\nYgfQ3NLpdGVl5eIQTo86SV1HRx0A+PwUO4BoHB5Cp6gz1JUVdQDg83PzBABAQih2AAAJodgB\nACSEYgcAkBBungCSr6ysLFbPFkmlUlFHAJJJsQMSbu3atSeddJIuBbQEih2QcCUlJalU6rU4\nXXpSHcKgqDMAiaTYAS3CKXEqdlVRBwCSKj4THQAAn4tiBwCQEE7FAl+knTt3Pv/887G6U2H9\n+vVRRwBoJhlZ7EpKSkpLS/fs2ZOdnd25c+fc3NyoEwH/a+7cuTdOmXJ41DHqqog6AECzyaRi\nt2TJknnz5r300ks7duyoO965c+ehQ4dOnjx52LBhUWUDalVXV38thFeijlHXb0L4v1FnAGge\nmVHsqqqqRo8evXDhwhBCQUHBiSeemJOT065du4qKitLS0q1btz7xxBNPPPHEpZdeet9997Vt\n2zbqvAAAEciMYjdz5syFCxdefPHFs2bNys/P3/8DRUVF119//QMPPDBgwIDrrruu+RNCJGbP\nnv3www9HnWIf27ZtOzrqDAAtVmYUu/nz55922mkPPfRQq1YN38abn5//0EMPFRUVzZs3T7Hj\nEHnmmWc+/PDDqFPs44EHHsh9661zo45R1/1RBwBoyTKj2BUXF48aNepAra5WVlbWmWeeec89\n93ymb966devAgQOrq6sb+Uztu1lZWZ/pmw/klltumTNnzhfyVV+UqqqqsrKydDoddRD+G+1D\nWBt1hro+CWFLCP8n6hh11d488ZWIUzTgvJhNwZ+E8D8h3B51jLp2hZAVs91pTwgVMYtUewv6\n4BBaRxxkH1Uh/CyEqVHHqKsshLGtY7WRDolYzSoHlJ+fv2rVqsY/k06nV6xY0aNHj8/6zY89\n9ljjxS6dTn/00Udf1KV7EydOHDQoXosJVVRUbN++/eij43UC7d133z322GOjTrGPrVu3HnXU\nUW3axOivZvv27e3bt8/JyYk6yKfKyspKS0vz8vKiDvKpdDq9ZcuWXr16RR1kH5s3by4oKPii\nfjF+IYqLi3Nycg477LCog3yqtLS0vLy8W7duUQf5VHV19QcffHDMMcdEHWQfMZww33///W7d\nurVr1y7qIPvo379/1BEOuayMOE4zc+bMG2+8cezYsTfffHODf04ffvjhDTfc8OCDD06bNu1X\nv/pV8ycEAIhcZhS7ysrKCRMmLFiwIISQn5/fu3fv3Nzcdu3aVVZW7ty587333tu0aVMI4fzz\nz3/00Ufj9vsAAKB5ZEaxq/Xss8/ee++9K1eu3LZtW93xrl27nnnmmePGjRsxYkRU2QAAIpdJ\nxW6v0tLS0tLSsrKy9u3bd+rUycoTAAAhQ4sdAAD7a+wBIgAAZBDFDgAgIRQ7AICEUOwAABJC\nsQMASAjFDgAgIRQ7AICEUOwAABJCsQMASAjFDgAgIRQ7AICEUOwAABJCsQMASAjFDgAgIRQ7\nAICEUOwAABKiTdQBiN60adNuueWWqFMAwKE1aNCglStXRp3i0FLsCEccccTRRx/9xBNPRB0k\n7oYNGzZt2rSzzjor6iCx9qc//Wn58uX3339/1EFiLZVKnXbaaYWFhSeffHLUWWLt7rvvfu+9\n9+68886og8Taxx9/fPbZZz/22GMFBQVRZ4m1m2+++fDDD486xSGn2BFat27drl27U045Jeog\ncde6detevXrZUI1bunTpYYcdZis1LpVKhRD69OljQzWuW7duJSUltlLjPvrooxBC//79+/Xr\nF3WWWOvcuXPUEZqDa+wAABJCsQMASAjFDgAgIRQ7AICEUOwAABJCsQMASAjFDgAgIRQ7AICE\nUOwAABLCyhOE7Ozs7OzsqFNkABvqYNhKByMrK6tt27Y2VJPsTgejbdu2WVlZNlSTWsgmykqn\n01FnIGJVVVXbtm076qijog4Sd0VFRUceeWTr1q2jDhJre/bs2blzZ48ePaIOEndbt27t2bNn\nVlZW1EFibdeuXeXl5V26dIk6SNxt2bLFQrFNKikpCSF06tQp6iCHlmIHAJAQrrEDAEgIxQ4A\nICEUOwCAhFDsAAASQrEDAEgIxQ4AICEUOwCAhFDsAAASQrEDAEgIxQ4AICEUOwCAhFDsAAAS\nQrEDAEgIxQ4AICEUO+q79tprs7KyXn311aiDxNH27dt/9KMfFRQUdOzYccCAAbNnz66uro46\nVOz861//mjRp0lFHHZWdnd29e/exY8du3bo16lCxtmPHjq985St33HFH1EFipKqq6vbbb+/b\nt2+HDh169+49Y8YMf2uNsAs1rkVNSood+3jttdfuuuuuqFPE1I4dOwYNGlRYWHjiiSdOnjy5\nY8eO11577YUXXhh1rngpLi7+2te+VlhYeNxxx02ePPnkk09+8MEHTz311HfffTfqaDFVXl5+\n4YUX/vvf/446SLxMnDhxypQpXbt2nTx5cufOnX/xi1+MGTMm6lAxZRdqXIublNLwHxUVFQMG\nDKjdMVauXBl1nNi58sorQwj33Xdf7ctUKvW9730vhLB06dJog8XKZZddFkKYM2fO3pH58+eH\nEEaMGBFhqtgqLi4ePHhw7R/d7bffHnWcuFi6dGkIYezYsbUva2pqLr744hDCs88+G22wGLIL\nNamlTUqO2PGpGTNmbNiw4eyzz446SEx9/PHHJ5xwwoQJE2pfZmVl1R6ue+ONNyLNFS+LFi3q\n2rXrT3/6070j48aNKygoqK2/EQaLoXvvvbdfv36vvvrqt771raizxMtvf/vbEML06dNrX7Zq\n1WrWrFkhhLlz50aYKobsQgejpU1KbaIOQFy89dZbM2fOvPbaa9P/+blMPQ8//HC9kTfffDOE\n0LNnzwjSxFJNTc2sWbNatWrVqtU+Pxo7dOhQWVlZU1PTpo0551O/+c1vunTpsnDhwl27dj33\n3HNRx4mR5cuX5+fnFxQU7B2pffnSSy9FFyqO7EJNaoGTkiN2hBBCTU3NhAkTevbs+ctf/jLq\nLBlg165db7755tSpU+fMmTNw4MBRo0ZFnSguWrduPXHixPHjx9cd3Lhx44YNG7761a8mbwL9\nnObMmbNu3bpvfvObUQeJl48//rikpKR37971xo855pgdO3aUlJREkiqe7EJNaoGTUgL/SfwX\n7rjjjtWrV7/44ovt27ePOksGuPHGG2tPFeXl5c2fPz87OzvqRPGVSqV+8pOf1NTU1D0PQi2X\nPTSo9iaA3NzceuM5OTkhhNLS0k6dOkUQK5bsQv+FxE9KjtgR/vGPf0yfPv2KK64YOnRo1Fky\nw/DhwxcuXDh9+vTS0tJTTjnF6aEDSafTV1555bJly4YPH/7DH/4w6jhkhqqqqhBCu3bt6o3X\njpSXl0eQiaRoCZOSI3YtXTqdnjhxYufOnW+77baos2SM73znOyGEUaNGDRs2bOjQoZMmTVq/\nfn29Czioqam54oor7r///sGDBz/yyCNZWVlRJyIzdOjQIYRQWVlZb7yioiKEcNhhh0WQiURo\nIZOSYteCLF68+Prrr687snr16sLCwhUrVjz55JNf/vKXowoWNw1uqAbPt5555pmnnnrq66+/\n/sEHH+Tn5zdXwFhofCvt3r37oosuevrpp4cNG/bkk09+6UtfiiJjLBz87kSt2jOtpaWl9cZr\nR2pPyMJn1XImJcWuBdm5c+e6devqjqRSqYULF4YQzj///Hofrn0w0vr16/v27dtsCWOiwQ21\ne/fuRx555PDDD//+979f960uXbqEEEpKSlpasWtwK9X+R0lJyTnnnLNq1aof/OAHf/7zn1t4\niWlkQ9Gg3Nzcrl277r8wwNatW/Py8vwE5b/QoiYlJ49akDFjxtR7jGH79u2HDh06al99+vQJ\nIZx11lmjRo1qmXNogxuqVatWP/7xj3/2s5/V/b9yVVXVmjVr2rZtW/e5DC1Eg1sphFBeXn7e\neeetWrXq6quvfuihh5I9gR6MA20oGjFkyJBNmzb985//3Dvy/vvvb9my5fTTT48wFRmqxU1K\nX9STjkmMqVOnBitPNOSiiy4KIdx22221L1Op1HXXXRdCuOSSS6INFitXXXVVCOHyyy+POkgm\nWbRoUbBsQB1PPvlkCOHSSy9NpVLpdDqVSo0ePTqEsGzZsqijxZRdqBEtbVJyKhYO1uzZs1es\nWDFlypTnnnvu+OOPX7ly5WuvvdavX7977rkn6mhxsW3btt/97nchhLfeeusb3/hGvXcXL16c\n4Ota+AJ997vfPf/88x944IHNmzcPGTJk+fLlr7zyypgxY6yvwGfVAiclxQ4O1pFHHvnGG29M\nmzbt6aeffvHFF4888sgpU6b8/Oc/b5knrBv06quv1t7MuGrVqv3fra6ubvZEZKpHH310xowZ\nf/nLX+666678/Pxf//rX11xzTdShyDwtcFLKSidxoTQAgBbIzRMAAAmh2AEAJIRiBwCQEIod\nAEBCKHYAAAmh2AEAJIRiBwCQEIodAEBCKHYAAAmh2AEAJIRiBwCQEIodAEBCKHYAAAmh2AEA\nJIRiBwCQEIodAEBCKHYAAAmh2AEAJIRiBwCQEIodAEBCKHYAAAmh2AEAJIRiBwCQEIodAEBC\nKHYAAAmh2AEAJIRiBwCQEIodAEBCKHYAAAmh2AEAJIRiBwCQEIodAEBCKHYAB/TMM8+MHDmy\ne/fu2dnZ3bt3Hz58+OOPP173A5s2bcrKyhoyZEhUCQHqUuwAGjZ16tThw4c/9dRTOTk5AwcO\nzMnJWbJkyUUXXTRmzJh0Oh11OoAGKHYADVi9evVtt93Wq1evtWvXbty4cfny5Rs3bly7dm3f\nvn0XLFiwYMGCqAMCNECxA2jAkiVLQgjTp0/v37//3sF+/foVFhaGEBQ7IJ7aRB0AIL5KSkrq\njZx++umzZs0qKCiIJA9A47JcKQKwv7Vr15500kkdO3a86aabLrnkkry8vAY/tmnTpmOPPfaM\nM85YsWJFMycE2J9TsQANOOGEE/74xz9WVVVNmTLliCOOGDBgwOTJkxctWlRWVhZ1NIADUuwA\nGnb55ZevW7du6tSpxx9//Lp1637/+99fcMEFPXr0uPXWW1OpVNTpABqg2AEcUK9evW699dZ3\n3nmnuLh4wYIFl1xySUVFxQ033HD11VdHHQ2gAa6xA2jY7t27O3bsWG9ww4YNQ4YM2blz5wcf\nfNCjRw/X2AGx4ogdQH0VFRUdOnQ444wz9n+rb9++F198cU1NzYYNG5o/GEDjFDuA+tq1a9er\nV681a9asXr16/3fff//9EELXrl2bPRdAExQ7gAZcc8016XR6xIgRixYtqq6urh0sLy+fMWPG\n008/ffLJJ9d9cDFATHhAMUADxo8fv2bNmrvvvvuCCy7Izc0tKChIpVIbN27cs2dPjx49Hn74\n4agDAjTAETuAht11113PP//86NGjc3Nz169f/+677/bu3XvatGnvvPPOcccdF3U6gAa4KxYA\nICEcsQMASAjFDgAgIRQ7AICEUOwAABJCsQMASAjFDgAgIRQ7AICEUOwAABJCsQMASAjFDgAg\nIRQ7AICEUOwAABJCsQMASAjFDgAgIRQ7AICEUOwAABJCsQMASAjFDgAgIRQ7AICEUOwAABJC\nsQMASAjFDgAgIRQ7AICEUOwAABJCsQMASAjFDgAgIRQ7AICEUOwAABJCsQMASAjFDgAgIf4/\nETJnPoZXToAAAAAASUVORK5CYII=",
      "text/plain": [
       "Plot with title \"Histogram of rowRanges(Unidirectional)$SI\""
      ]
     },
     "metadata": {
      "image/png": {
       "height": 420,
       "width": 420
      }
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "hist(rowRanges(Unidirectional)$SI,\n",
    "    # breaks=max(rowRanges(Unidirectional)$SI),\n",
    "     #  xlim=c(0,20), \n",
    "     xlab = \"SI\", \n",
    "     col=\"red\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "id": "eb9964e9",
   "metadata": {},
   "outputs": [],
   "source": [
    "result <- data.frame(as.data.frame(Unidirectional@rowRanges),\n",
    "                   as.data.frame(as.matrix(Unidirectional@assays@data@listData$counts)),\n",
    "                   as.data.frame(as.matrix(Unidirectional@assays@data@listData$TPM)))\n",
    "\n",
    "write.table(as.data.frame(result),\n",
    "            file=\"../result_file/Unidirection_result.txt\",\n",
    "            col.names=T,row.names=F,\n",
    "            sep=\"\\t\",quote=F)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "id": "fd3c623c",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Pre-filtering bidirectional candidate regions...\n",
      "\n",
      "Retaining for analysis: 2%\n",
      "\n",
      "Splitting by strand...\n",
      "\n",
      "Calculating windowed coverage on plus strand...\n",
      "\n",
      "Calculating windowed coverage on minus strand...\n",
      "\n",
      "Calculating balance score...\n",
      "\n",
      "Slice-reduce to find bidirectional clusters...\n",
      "\n",
      "Calculating statistics...\n",
      "\n",
      "Preparing output...\n",
      "\n",
      "# Bidirectional clustering summary:\n",
      "\n",
      "Number of bidirectional clusters: 929\n",
      "\n",
      "Maximum balance score: 1\n",
      "\n",
      "Minimum balance score: 0.707106781186548\n",
      "\n",
      "Maximum width: 1989\n",
      "\n",
      "Minimum width: 603\n",
      "\n"
     ]
    }
   ],
   "source": [
    "BCs <- clusterBidirectionally(supportedCTSSs, balanceThreshold=0)#0.95"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "id": "23e6cae4",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Finding overlaps...\n",
      "\n",
      "Aggregating within clusters...\n",
      "\n",
      "Calculating support...\n",
      "\n",
      "Subsetting...\n",
      "\n",
      "Removed 0 out of 929 regions (0%)\n",
      "\n"
     ]
    }
   ],
   "source": [
    "BCs <- calcBidirectionality(BCs, samples=supportedCTSSs)\n",
    "\n",
    "BCs <- quantifyClusters(supportedCTSSs,\n",
    "                        clusters=BCs,\n",
    "                        inputAssay=\"counts\")\n",
    "\n",
    "BCs <- subsetBySupport(BCs,\n",
    "                       inputAssay=\"counts\",\n",
    "                       unexpressed=0,\n",
    "                       minSamples=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "id": "eea66a94",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Calculating library sizes...\n",
      "\n",
      "Warning message in calcTotalTags(object = object, inputAssay = inputAssay, outputColumn = outputColumn):\n",
      "\"object already has a column named subsetTags in colData: It will be overwritten!\"\n",
      "Calculating TPM...\n",
      "\n",
      "Calculating support...\n",
      "\n",
      "Warning message in calcSupport(object = object, inputAssay = inputAssay, outputColumn = outputColumn, :\n",
      "\"object already has a column named support in rowData: It will be overwritten!\"\n",
      "Subsetting...\n",
      "\n",
      "Removed 0 out of 929 regions (0%)\n",
      "\n",
      "Finding hierachical overlaps...\n",
      "\n",
      "### Overlap summary: ###\n",
      "\n",
      "      txType count percentage\n",
      "1   promoter   501       53.9\n",
      "2   proximal    24        2.6\n",
      "3    fiveUTR   103       11.1\n",
      "4   threeUTR    69        7.4\n",
      "5        CDS   198       21.3\n",
      "6       exon     0        0.0\n",
      "7     intron    16        1.7\n",
      "8  antisense     0        0.0\n",
      "9 intergenic    18        1.9\n",
      "\n",
      "Extracting transcripts...\n",
      "\n",
      "Finding hierachical overlaps...\n",
      "\n",
      "### Overlap Summary: ###\n",
      "\n",
      "Features overlapping transcripts: 98.06 %\n",
      "\n",
      "Number of unique transcripts: 2073\n",
      "\n"
     ]
    }
   ],
   "source": [
    "Bidirectional <- calcTPM(BCs,\n",
    "                         inputAssay=\"counts\", \n",
    "                         outputAssay=\"TPM\", \n",
    "                         outputColumn=\"subsetTags\")\n",
    "\n",
    "Bidirectional <- subsetBySupport(Bidirectional,\n",
    "                                 inputAssay=\"TPM\",\n",
    "                                 unexpressed=1,\n",
    "                                 minSamples=0)\n",
    "\n",
    "Bidirectional <- assignTxType(Bidirectional,\n",
    "                              txModels=txdb,\n",
    "                              outputColumn=\"txType\")\n",
    "\n",
    "Bidirectional <- assignTxID(Bidirectional,\n",
    "                            txModels=txdb,\n",
    "                            outputColumn=\"txID\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "id": "30347995",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Only keep intronic and intergenic enhancers\n",
    "#Bidirectional <- subset(Bidirectional,\n",
    "#                               txType %in% c(\"intron\", \"intergenic\"))\n",
    "result<-data.frame(as.data.frame(Bidirectional@rowRanges),\n",
    "                   as.data.frame(as.matrix(Bidirectional@assays@data@listData$counts)),\n",
    "                   as.data.frame(as.matrix(Bidirectional@assays@data@listData$TPM)))\n",
    "\n",
    "write.table(as.data.frame(result),\n",
    "            file = \"../result_file/Bidirection_result.txt\",\n",
    "            col.names=T,row.names=F,\n",
    "            sep=\"\\t\",quote=F)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "da84b78d",
   "metadata": {},
   "outputs": [],
   "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": "4.3.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}