{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "6af62535-2a26-48eb-8d9f-1e5c9d24e5f7",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "total 12629464\n",
      "drwxr-xr-x 3 username username       4096 Apr 28 15:04  checkpoint-388560\n",
      "-rw-r--r-- 1 username username 1407182615 May 12 18:30 'DNA_rep1.fq.gz?download=1'\n",
      "-rw-r--r-- 1 username username 1293646655 May 12 18:30 'DNA_rep2.fq.gz?download=1'\n",
      "-rw-r--r-- 1 username username 1304107626 May 12 18:30 'DNA_rep3.fq.gz?download=1'\n",
      "-rw-r--r-- 1 username username 1405504005 May 12 18:30 'DNA_rep4.fq.gz?download=1'\n",
      "-rw-r--r-- 1 username username 1715014321 May 12 18:31 'plasmid.fq.gz?download=1'\n",
      "-rw-r--r-- 1 username username 1286362892 May 12 18:30 'RNA_rep1.fq.gz?download=1'\n",
      "-rw-r--r-- 1 username username 1325622556 May 12 18:30 'RNA_rep2.fq.gz?download=1'\n",
      "-rw-r--r-- 1 username username 1969174493 May 12 18:32 'RNA_rep3.fq.gz?download=1'\n",
      "-rw-r--r-- 1 username username 1202971591 May 12 18:31 'RNA_rep4.fq.gz?download=1'\n",
      "-rw-r--r-- 1 username username   22913898 May 12 18:29  TILE_mpra.fa\n",
      "-rw-r--r-- 1 username username      39297 May  8 01:40  train_rnaseek_stability_from_tsv_NO_DOUBLE_BOS.py\n",
      "-rw-r--r-- 1 username username         72 May 12 18:35  Untitled.ipynb\n"
     ]
    }
   ],
   "source": [
    "!ls -l"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "c3043346-ff33-4f0d-bfb8-142d0a151922",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "already exists: RNA_rep1.fq.gz\n",
      "already exists: RNA_rep2.fq.gz\n",
      "already exists: RNA_rep3.fq.gz\n",
      "already exists: RNA_rep4.fq.gz\n",
      "already exists: DNA_rep1.fq.gz\n",
      "already exists: DNA_rep2.fq.gz\n",
      "already exists: DNA_rep3.fq.gz\n",
      "already exists: DNA_rep4.fq.gz\n",
      "already exists: plasmid.fq.gz\n",
      "\n",
      "FASTQs:\n",
      "DNA_rep1.fq.gz\n",
      "DNA_rep2.fq.gz\n",
      "DNA_rep3.fq.gz\n",
      "DNA_rep4.fq.gz\n",
      "RNA_rep1.fq.gz\n",
      "RNA_rep2.fq.gz\n",
      "RNA_rep3.fq.gz\n",
      "RNA_rep4.fq.gz\n",
      "plasmid.fq.gz\n"
     ]
    }
   ],
   "source": [
    "from pathlib import Path\n",
    "import os\n",
    "\n",
    "workdir = Path(\".\").resolve()\n",
    "\n",
    "rename_map = {\n",
    "    \"RNA_rep1.fq.gz?download=1\": \"RNA_rep1.fq.gz\",\n",
    "    \"RNA_rep2.fq.gz?download=1\": \"RNA_rep2.fq.gz\",\n",
    "    \"RNA_rep3.fq.gz?download=1\": \"RNA_rep3.fq.gz\",\n",
    "    \"RNA_rep4.fq.gz?download=1\": \"RNA_rep4.fq.gz\",\n",
    "    \"DNA_rep1.fq.gz?download=1\": \"DNA_rep1.fq.gz\",\n",
    "    \"DNA_rep2.fq.gz?download=1\": \"DNA_rep2.fq.gz\",\n",
    "    \"DNA_rep3.fq.gz?download=1\": \"DNA_rep3.fq.gz\",\n",
    "    \"DNA_rep4.fq.gz?download=1\": \"DNA_rep4.fq.gz\",\n",
    "    \"plasmid.fq.gz?download=1\": \"plasmid.fq.gz\",\n",
    "}\n",
    "\n",
    "for old, new in rename_map.items():\n",
    "    old_path = workdir / old\n",
    "    new_path = workdir / new\n",
    "    if old_path.exists() and not new_path.exists():\n",
    "        old_path.rename(new_path)\n",
    "        print(f\"renamed: {old} -> {new}\")\n",
    "    elif new_path.exists():\n",
    "        print(f\"already exists: {new}\")\n",
    "    else:\n",
    "        print(f\"missing: {old}\")\n",
    "\n",
    "print(\"\\nFASTQs:\")\n",
    "for p in sorted(workdir.glob(\"*.fq.gz\")):\n",
    "    print(p.name)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "2f823d74-0460-4996-980c-cca8789e2a28",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "bowtie2        /opt/conda/bin/bowtie2  exists=True executable=True\n",
      "bowtie2-build  /opt/conda/bin/bowtie2-build  exists=True executable=True\n",
      "samtools       /opt/conda/bin/samtools  exists=True executable=True\n",
      "\n",
      "PATH now starts with:\n",
      "['/opt/conda/bin', '/opt/conda/envs/platform/bin', '/opt/conda/condabin', '/bin', '/usr/local/bin']\n"
     ]
    }
   ],
   "source": [
    "from pathlib import Path\n",
    "import os\n",
    "import shutil\n",
    "import subprocess\n",
    "\n",
    "CONDA_BIN = Path(\"/opt/conda/bin\")\n",
    "\n",
    "# Make Jupyter subprocesses see your conda tools\n",
    "os.environ[\"PATH\"] = f\"{CONDA_BIN}:{os.environ.get('PATH', '')}\"\n",
    "\n",
    "BOWTIE2 = str(CONDA_BIN / \"bowtie2\")\n",
    "BOWTIE2_BUILD = str(CONDA_BIN / \"bowtie2-build\")\n",
    "SAMTOOLS = str(CONDA_BIN / \"samtools\")\n",
    "\n",
    "tools = {\n",
    "    \"bowtie2\": BOWTIE2,\n",
    "    \"bowtie2-build\": BOWTIE2_BUILD,\n",
    "    \"samtools\": SAMTOOLS,\n",
    "}\n",
    "\n",
    "for name, path in tools.items():\n",
    "    p = Path(path)\n",
    "    print(f\"{name:14s} {path}  exists={p.exists()} executable={os.access(p, os.X_OK)}\")\n",
    "\n",
    "missing = [name for name, path in tools.items() if not Path(path).exists()]\n",
    "if missing:\n",
    "    raise FileNotFoundError(f\"Missing tools: {missing}\")\n",
    "\n",
    "print(\"\\nPATH now starts with:\")\n",
    "print(os.environ[\"PATH\"].split(\":\")[:5])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "aab7eab4-a277-4972-9e74-3536cfc46290",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "running:\n",
      "/opt/conda/bin/bowtie2-build TILE_mpra.fa bt2_index/TILE_mpra\n",
      "Settings:\n",
      "  Output files: \"bt2_index/TILE_mpra.*.bt2\"\n",
      "  Line rate: 6 (line is 64 bytes)\n",
      "  Lines per side: 1 (side is 64 bytes)\n",
      "  Offset rate: 4 (one in 16)\n",
      "  FTable chars: 10\n",
      "  Strings: unpacked\n",
      "  Max bucket size: default\n",
      "  Max bucket size, sqrt multiplier: default\n",
      "  Max bucket size, len divisor: 4\n",
      "  Difference-cover sample period: 1024\n",
      "  Endianness: little\n",
      "  Actual local endianness: little\n",
      "  Sanity checking: disabled\n",
      "  Assertions: disabled\n",
      "  Random seed: 0\n",
      "  Sizeofs: void*:8, int:4, long:8, size_t:8\n",
      "Input files DNA, FASTA:\n",
      "  TILE_mpra.fa\n",
      "Reading reference sizes\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Building a SMALL index\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "  Time reading reference sizes: 00:00:00\n",
      "Calculating joined length\n",
      "Writing header\n",
      "Reserving space for joined string\n",
      "Joining reference sequences\n",
      "  Time to join reference sequences: 00:00:00\n",
      "bmax according to bmaxDivN setting: 3871792\n",
      "Using parameters --bmax 2903844 --dcv 1024\n",
      "  Doing ahead-of-time memory usage test\n",
      "  Passed!  Constructing with these parameters: --bmax 2903844 --dcv 1024\n",
      "Constructing suffix-array element generator\n",
      "Building DifferenceCoverSample\n",
      "  Building sPrime\n",
      "  Building sPrimeOrder\n",
      "  V-Sorting samples\n",
      "  V-Sorting samples time: 00:00:00\n",
      "  Allocating rank array\n",
      "  Ranking v-sort output\n",
      "  Ranking v-sort output time: 00:00:00\n",
      "  Invoking Larsson-Sadakane on ranks\n",
      "  Invoking Larsson-Sadakane on ranks time: 00:00:00\n",
      "  Sanity-checking and returning\n",
      "Building samples\n",
      "Reserving space for 12 sample suffixes\n",
      "Generating random suffixes\n",
      "QSorting 12 sample offsets, eliminating duplicates\n",
      "QSorting sample offsets, eliminating duplicates time: 00:00:00\n",
      "Multikey QSorting 12 samples\n",
      "  (Using difference cover)\n",
      "  Multikey QSorting samples time: 00:00:00\n",
      "Calculating bucket sizes\n",
      "Splitting and merging\n",
      "  Splitting and merging time: 00:00:00\n",
      "Split 1, merged 7; iterating...\n",
      "Splitting and merging\n",
      "  Splitting and merging time: 00:00:00\n",
      "Split 1, merged 0; iterating...\n",
      "Splitting and merging\n",
      "  Splitting and merging time: 00:00:00\n",
      "Avg bucket size: 1.9359e+06 (target: 2903843)\n",
      "Converting suffix-array elements to index image\n",
      "Allocating ftab, absorbFtab\n",
      "Entering Ebwt loop\n",
      "Getting block 1 of 8\n",
      "  Reserving size (2903844) for bucket 1\n",
      "  Calculating Z arrays for bucket 1\n",
      "  Entering block accumulator loop for bucket 1:\n",
      "  bucket 1: 10%\n",
      "  bucket 1: 20%\n",
      "  bucket 1: 30%\n",
      "  bucket 1: 40%\n",
      "  bucket 1: 50%\n",
      "  bucket 1: 60%\n",
      "  bucket 1: 70%\n",
      "  bucket 1: 80%\n",
      "  bucket 1: 90%\n",
      "  bucket 1: 100%\n",
      "  Sorting block of length 2046717 for bucket 1\n",
      "  (Using difference cover)\n",
      "  Sorting block time: 00:00:00\n",
      "Returning block of 2046718 for bucket 1\n",
      "Getting block 2 of 8\n",
      "  Reserving size (2903844) for bucket 2\n",
      "  Calculating Z arrays for bucket 2\n",
      "  Entering block accumulator loop for bucket 2:\n",
      "  bucket 2: 10%\n",
      "  bucket 2: 20%\n",
      "  bucket 2: 30%\n",
      "  bucket 2: 40%\n",
      "  bucket 2: 50%\n",
      "  bucket 2: 60%\n",
      "  bucket 2: 70%\n",
      "  bucket 2: 80%\n",
      "  bucket 2: 90%\n",
      "  bucket 2: 100%\n",
      "  Sorting block of length 1502968 for bucket 2\n",
      "  (Using difference cover)\n",
      "  Sorting block time: 00:00:00\n",
      "Returning block of 1502969 for bucket 2\n",
      "Getting block 3 of 8\n",
      "  Reserving size (2903844) for bucket 3\n",
      "  Calculating Z arrays for bucket 3\n",
      "  Entering block accumulator loop for bucket 3:\n",
      "  bucket 3: 10%\n",
      "  bucket 3: 20%\n",
      "  bucket 3: 30%\n",
      "  bucket 3: 40%\n",
      "  bucket 3: 50%\n",
      "  bucket 3: 60%\n",
      "  bucket 3: 70%\n",
      "  bucket 3: 80%\n",
      "  bucket 3: 90%\n",
      "  bucket 3: 100%\n",
      "  Sorting block of length 2645085 for bucket 3\n",
      "  (Using difference cover)\n",
      "  Sorting block time: 00:00:00\n",
      "Returning block of 2645086 for bucket 3\n",
      "Getting block 4 of 8\n",
      "  Reserving size (2903844) for bucket 4\n",
      "  Calculating Z arrays for bucket 4\n",
      "  Entering block accumulator loop for bucket 4:\n",
      "  bucket 4: 10%\n",
      "  bucket 4: 20%\n",
      "  bucket 4: 30%\n",
      "  bucket 4: 40%\n",
      "  bucket 4: 50%\n",
      "  bucket 4: 60%\n",
      "  bucket 4: 70%\n",
      "  bucket 4: 80%\n",
      "  bucket 4: 90%\n",
      "  bucket 4: 100%\n",
      "  Sorting block of length 378013 for bucket 4\n",
      "  (Using difference cover)\n",
      "  Sorting block time: 00:00:00\n",
      "Returning block of 378014 for bucket 4\n",
      "Getting block 5 of 8\n",
      "  Reserving size (2903844) for bucket 5\n",
      "  Calculating Z arrays for bucket 5\n",
      "  Entering block accumulator loop for bucket 5:\n",
      "  bucket 5: 10%\n",
      "  bucket 5: 20%\n",
      "  bucket 5: 30%\n",
      "  bucket 5: 40%\n",
      "  bucket 5: 50%\n",
      "  bucket 5: 60%\n",
      "  bucket 5: 70%\n",
      "  bucket 5: 80%\n",
      "  bucket 5: 90%\n",
      "  bucket 5: 100%\n",
      "  Sorting block of length 2761753 for bucket 5\n",
      "  (Using difference cover)\n",
      "  Sorting block time: 00:00:01\n",
      "Returning block of 2761754 for bucket 5\n",
      "Getting block 6 of 8\n",
      "  Reserving size (2903844) for bucket 6\n",
      "  Calculating Z arrays for bucket 6\n",
      "  Entering block accumulator loop for bucket 6:\n",
      "  bucket 6: 10%\n",
      "  bucket 6: 20%\n",
      "  bucket 6: 30%\n",
      "  bucket 6: 40%\n",
      "  bucket 6: 50%\n",
      "  bucket 6: 60%\n",
      "  bucket 6: 70%\n",
      "  bucket 6: 80%\n",
      "  bucket 6: 90%\n",
      "  bucket 6: 100%\n",
      "  Sorting block of length 2208869 for bucket 6\n",
      "  (Using difference cover)\n",
      "  Sorting block time: 00:00:00\n",
      "Returning block of 2208870 for bucket 6\n",
      "Getting block 7 of 8\n",
      "  Reserving size (2903844) for bucket 7\n",
      "  Calculating Z arrays for bucket 7\n",
      "  Entering block accumulator loop for bucket 7:\n",
      "  bucket 7: 10%\n",
      "  bucket 7: 20%\n",
      "  bucket 7: 30%\n",
      "  bucket 7: 40%\n",
      "  bucket 7: 50%\n",
      "  bucket 7: 60%\n",
      "  bucket 7: 70%\n",
      "  bucket 7: 80%\n",
      "  bucket 7: 90%\n",
      "  bucket 7: 100%\n",
      "  Sorting block of length 2547294 for bucket 7\n",
      "  (Using difference cover)\n",
      "  Sorting block time: 00:00:01\n",
      "Returning block of 2547295 for bucket 7\n",
      "Getting block 8 of 8\n",
      "  Reserving size (2903844) for bucket 8\n",
      "  Calculating Z arrays for bucket 8\n",
      "  Entering block accumulator loop for bucket 8:\n",
      "  bucket 8: 10%\n",
      "  bucket 8: 20%\n",
      "  bucket 8: 30%\n",
      "  bucket 8: 40%\n",
      "  bucket 8: 50%\n",
      "  bucket 8: 60%\n",
      "  bucket 8: 70%\n",
      "  bucket 8: 80%\n",
      "  bucket 8: 90%\n",
      "  bucket 8: 100%\n",
      "  Sorting block of length 1396464 for bucket 8\n",
      "  (Using difference cover)\n",
      "  Sorting block time: 00:00:00\n",
      "Returning block of 1396465 for bucket 8\n",
      "Exited Ebwt loop\n",
      "fchr[A]: 0\n",
      "fchr[C]: 4185231\n",
      "fchr[G]: 7753926\n",
      "fchr[T]: 11625206\n",
      "fchr[$]: 15487170\n",
      "Exiting Ebwt::buildToDisk()\n",
      "Returning from initFromVector\n",
      "Wrote 18059023 bytes to primary EBWT file: bt2_index/TILE_mpra.1.bt2.tmp\n",
      "Wrote 3871800 bytes to secondary EBWT file: bt2_index/TILE_mpra.2.bt2.tmp\n",
      "Re-opening _in1 and _in2 as input streams\n",
      "Returning from Ebwt constructor\n",
      "Headers:\n",
      "    len: 15487170\n",
      "    bwtLen: 15487171\n",
      "    sz: 3871793\n",
      "    bwtSz: 3871793\n",
      "    lineRate: 6\n",
      "    offRate: 4\n",
      "    offMask: 0xfffffff0\n",
      "    ftabChars: 10\n",
      "    eftabLen: 20\n",
      "    eftabSz: 80\n",
      "    ftabLen: 1048577\n",
      "    ftabSz: 4194308\n",
      "    offsLen: 967949\n",
      "    offsSz: 3871796\n",
      "    lineSz: 64\n",
      "    sideSz: 64\n",
      "    sideBwtSz: 48\n",
      "    sideBwtLen: 192\n",
      "    numSides: 80663\n",
      "    numLines: 80663\n",
      "    ebwtTotLen: 5162432\n",
      "    ebwtTotSz: 5162432\n",
      "    color: 0\n",
      "    reverse: 0\n",
      "Total time for call to driver() for forward index: 00:00:04\n",
      "Reading reference sizes\n",
      "  Time reading reference sizes: 00:00:00\n",
      "Calculating joined length\n",
      "Writing header\n",
      "Reserving space for joined string\n",
      "Joining reference sequences\n",
      "  Time to join reference sequences: 00:00:00\n",
      "  Time to reverse reference sequence: 00:00:00\n",
      "bmax according to bmaxDivN setting: 3871792\n",
      "Using parameters --bmax 2903844 --dcv 1024\n",
      "  Doing ahead-of-time memory usage test\n",
      "  Passed!  Constructing with these parameters: --bmax 2903844 --dcv 1024\n",
      "Constructing suffix-array element generator\n",
      "Building DifferenceCoverSample\n",
      "  Building sPrime\n",
      "  Building sPrimeOrder\n",
      "  V-Sorting samples\n",
      "  V-Sorting samples time: 00:00:00\n",
      "  Allocating rank array\n",
      "  Ranking v-sort output\n",
      "  Ranking v-sort output time: 00:00:00\n",
      "  Invoking Larsson-Sadakane on ranks\n",
      "  Invoking Larsson-Sadakane on ranks time: 00:00:00\n",
      "  Sanity-checking and returning\n",
      "Building samples\n",
      "Reserving space for 12 sample suffixes\n",
      "Generating random suffixes\n",
      "QSorting 12 sample offsets, eliminating duplicates\n",
      "QSorting sample offsets, eliminating duplicates time: 00:00:00\n",
      "Multikey QSorting 12 samples\n",
      "  (Using difference cover)\n",
      "  Multikey QSorting samples time: 00:00:00\n",
      "Calculating bucket sizes\n",
      "Splitting and merging\n",
      "  Splitting and merging time: 00:00:00\n",
      "Split 2, merged 7; iterating...\n",
      "Splitting and merging\n",
      "  Splitting and merging time: 00:00:00\n",
      "Avg bucket size: 2.21245e+06 (target: 2903843)\n",
      "Converting suffix-array elements to index image\n",
      "Allocating ftab, absorbFtab\n",
      "Entering Ebwt loop\n",
      "Getting block 1 of 7\n",
      "  Reserving size (2903844) for bucket 1\n",
      "  Calculating Z arrays for bucket 1\n",
      "  Entering block accumulator loop for bucket 1:\n",
      "  bucket 1: 10%\n",
      "  bucket 1: 20%\n",
      "  bucket 1: 30%\n",
      "  bucket 1: 40%\n",
      "  bucket 1: 50%\n",
      "  bucket 1: 60%\n",
      "  bucket 1: 70%\n",
      "  bucket 1: 80%\n",
      "  bucket 1: 90%\n",
      "  bucket 1: 100%\n",
      "  Sorting block of length 1961240 for bucket 1\n",
      "  (Using difference cover)\n",
      "  Sorting block time: 00:00:00\n",
      "Returning block of 1961241 for bucket 1\n",
      "Getting block 2 of 7\n",
      "  Reserving size (2903844) for bucket 2\n",
      "  Calculating Z arrays for bucket 2\n",
      "  Entering block accumulator loop for bucket 2:\n",
      "  bucket 2: 10%\n",
      "  bucket 2: 20%\n",
      "  bucket 2: 30%\n",
      "  bucket 2: 40%\n",
      "  bucket 2: 50%\n",
      "  bucket 2: 60%\n",
      "  bucket 2: 70%\n",
      "  bucket 2: 80%\n",
      "  bucket 2: 90%\n",
      "  bucket 2: 100%\n",
      "  Sorting block of length 1458326 for bucket 2\n",
      "  (Using difference cover)\n",
      "  Sorting block time: 00:00:01\n",
      "Returning block of 1458327 for bucket 2\n",
      "Getting block 3 of 7\n",
      "  Reserving size (2903844) for bucket 3\n",
      "  Calculating Z arrays for bucket 3\n",
      "  Entering block accumulator loop for bucket 3:\n",
      "  bucket 3: 10%\n",
      "  bucket 3: 20%\n",
      "  bucket 3: 30%\n",
      "  bucket 3: 40%\n",
      "  bucket 3: 50%\n",
      "  bucket 3: 60%\n",
      "  bucket 3: 70%\n",
      "  bucket 3: 80%\n",
      "  bucket 3: 90%\n",
      "  bucket 3: 100%\n",
      "  Sorting block of length 2583276 for bucket 3\n",
      "  (Using difference cover)\n",
      "  Sorting block time: 00:00:00\n",
      "Returning block of 2583277 for bucket 3\n",
      "Getting block 4 of 7\n",
      "  Reserving size (2903844) for bucket 4\n",
      "  Calculating Z arrays for bucket 4\n",
      "  Entering block accumulator loop for bucket 4:\n",
      "  bucket 4: 10%\n",
      "  bucket 4: 20%\n",
      "  bucket 4: 30%\n",
      "  bucket 4: 40%\n",
      "  bucket 4: 50%\n",
      "  bucket 4: 60%\n",
      "  bucket 4: 70%\n",
      "  bucket 4: 80%\n",
      "  bucket 4: 90%\n",
      "  bucket 4: 100%\n",
      "  Sorting block of length 1427785 for bucket 4\n",
      "  (Using difference cover)\n",
      "  Sorting block time: 00:00:00\n",
      "Returning block of 1427786 for bucket 4\n",
      "Getting block 5 of 7\n",
      "  Reserving size (2903844) for bucket 5\n",
      "  Calculating Z arrays for bucket 5\n",
      "  Entering block accumulator loop for bucket 5:\n",
      "  bucket 5: 10%\n",
      "  bucket 5: 20%\n",
      "  bucket 5: 30%\n",
      "  bucket 5: 40%\n",
      "  bucket 5: 50%\n",
      "  bucket 5: 60%\n",
      "  bucket 5: 70%\n",
      "  bucket 5: 80%\n",
      "  bucket 5: 90%\n",
      "  bucket 5: 100%\n",
      "  Sorting block of length 2609285 for bucket 5\n",
      "  (Using difference cover)\n",
      "  Sorting block time: 00:00:00\n",
      "Returning block of 2609286 for bucket 5\n",
      "Getting block 6 of 7\n",
      "  Reserving size (2903844) for bucket 6\n",
      "  Calculating Z arrays for bucket 6\n",
      "  Entering block accumulator loop for bucket 6:\n",
      "  bucket 6: 10%\n",
      "  bucket 6: 20%\n",
      "  bucket 6: 30%\n",
      "  bucket 6: 40%\n",
      "  bucket 6: 50%\n",
      "  bucket 6: 60%\n",
      "  bucket 6: 70%\n",
      "  bucket 6: 80%\n",
      "  bucket 6: 90%\n",
      "  bucket 6: 100%\n",
      "  Sorting block of length 2718544 for bucket 6\n",
      "  (Using difference cover)\n",
      "  Sorting block time: 00:00:01\n",
      "Returning block of 2718545 for bucket 6\n",
      "Getting block 7 of 7\n",
      "  Reserving size (2903844) for bucket 7\n",
      "  Calculating Z arrays for bucket 7\n",
      "  Entering block accumulator loop for bucket 7:\n",
      "  bucket 7: 10%\n",
      "  bucket 7: 20%\n",
      "  bucket 7: 30%\n",
      "  bucket 7: 40%\n",
      "  bucket 7: 50%\n",
      "  bucket 7: 60%\n",
      "  bucket 7: 70%\n",
      "  bucket 7: 80%\n",
      "  bucket 7: 90%\n",
      "  bucket 7: 100%\n",
      "  Sorting block of length 2728708 for bucket 7\n",
      "  (Using difference cover)\n",
      "  Sorting block time: 00:00:00\n",
      "Returning block of 2728709 for bucket 7\n",
      "Exited Ebwt loop\n",
      "fchr[A]: 0\n",
      "fchr[C]: 4185231\n",
      "fchr[G]: 7753926\n",
      "fchr[T]: 11625206\n",
      "fchr[$]: 15487170\n",
      "Exiting Ebwt::buildToDisk()\n",
      "Returning from initFromVector\n",
      "Wrote 18059023 bytes to primary EBWT file: bt2_index/TILE_mpra.rev.1.bt2.tmp\n",
      "Wrote 3871800 bytes to secondary EBWT file: bt2_index/TILE_mpra.rev.2.bt2.tmp\n",
      "Re-opening _in1 and _in2 as input streams\n",
      "Returning from Ebwt constructor\n",
      "Headers:\n",
      "    len: 15487170\n",
      "    bwtLen: 15487171\n",
      "    sz: 3871793\n",
      "    bwtSz: 3871793\n",
      "    lineRate: 6\n",
      "    offRate: 4\n",
      "    offMask: 0xfffffff0\n",
      "    ftabChars: 10\n",
      "    eftabLen: 20\n",
      "    eftabSz: 80\n",
      "    ftabLen: 1048577\n",
      "    ftabSz: 4194308\n",
      "    offsLen: 967949\n",
      "    offsSz: 3871796\n",
      "    lineSz: 64\n",
      "    sideSz: 64\n",
      "    sideBwtSz: 48\n",
      "    sideBwtLen: 192\n",
      "    numSides: 80663\n",
      "    numLines: 80663\n",
      "    ebwtTotLen: 5162432\n",
      "    ebwtTotSz: 5162432\n",
      "    color: 0\n",
      "    reverse: 1\n",
      "Total time for backward call to driver() for mirror index: 00:00:04\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Renaming bt2_index/TILE_mpra.3.bt2.tmp to bt2_index/TILE_mpra.3.bt2\n",
      "Renaming bt2_index/TILE_mpra.4.bt2.tmp to bt2_index/TILE_mpra.4.bt2\n",
      "Renaming bt2_index/TILE_mpra.1.bt2.tmp to bt2_index/TILE_mpra.1.bt2\n",
      "Renaming bt2_index/TILE_mpra.2.bt2.tmp to bt2_index/TILE_mpra.2.bt2\n",
      "Renaming bt2_index/TILE_mpra.rev.1.bt2.tmp to bt2_index/TILE_mpra.rev.1.bt2\n",
      "Renaming bt2_index/TILE_mpra.rev.2.bt2.tmp to bt2_index/TILE_mpra.rev.2.bt2\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "CompletedProcess(args=['/opt/conda/bin/bowtie2-build', 'TILE_mpra.fa', 'bt2_index/TILE_mpra'], returncode=0)"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from pathlib import Path\n",
    "import subprocess\n",
    "\n",
    "ref_fasta = Path(\"TILE_mpra.fa\")\n",
    "assert ref_fasta.exists(), \"TILE_mpra.fa not found in current directory.\"\n",
    "\n",
    "Path(\"bt2_index\").mkdir(exist_ok=True)\n",
    "\n",
    "index_prefix = \"bt2_index/TILE_mpra\"\n",
    "\n",
    "cmd = [\n",
    "    BOWTIE2_BUILD,\n",
    "    str(ref_fasta),\n",
    "    index_prefix,\n",
    "]\n",
    "\n",
    "print(\"running:\")\n",
    "print(\" \".join(cmd))\n",
    "\n",
    "subprocess.run(cmd, check=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "241c2493-26a3-46b2-9ab4-e9a7f8380f47",
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "[align] RNA_rep1\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libtinfow.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libtinfow.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "[bam_sort_core] merging from 0 files and 16 in-memory blocks...\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libtinfow.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "[align] RNA_rep2\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libtinfow.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libtinfow.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "[bam_sort_core] merging from 0 files and 16 in-memory blocks...\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libtinfow.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "[align] RNA_rep3\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libtinfow.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libtinfow.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "[bam_sort_core] merging from 0 files and 16 in-memory blocks...\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libtinfow.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "[align] RNA_rep4\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libtinfow.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libtinfow.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "[bam_sort_core] merging from 0 files and 16 in-memory blocks...\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libtinfow.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "[align] DNA_rep1\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libtinfow.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libtinfow.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "[bam_sort_core] merging from 0 files and 16 in-memory blocks...\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libtinfow.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "[align] DNA_rep2\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libtinfow.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libtinfow.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "[bam_sort_core] merging from 0 files and 16 in-memory blocks...\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libtinfow.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "[align] DNA_rep3\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libtinfow.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libtinfow.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "[bam_sort_core] merging from 0 files and 16 in-memory blocks...\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libtinfow.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "[align] DNA_rep4\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libtinfow.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libtinfow.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "[bam_sort_core] merging from 0 files and 16 in-memory blocks...\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libtinfow.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alignment done\n"
     ]
    }
   ],
   "source": [
    "from pathlib import Path\n",
    "import subprocess\n",
    "import os\n",
    "\n",
    "Path(\"align_bam\").mkdir(exist_ok=True)\n",
    "Path(\"align_logs\").mkdir(exist_ok=True)\n",
    "\n",
    "samples = {\n",
    "    \"RNA_rep1\": \"RNA_rep1.fq.gz\",\n",
    "    \"RNA_rep2\": \"RNA_rep2.fq.gz\",\n",
    "    \"RNA_rep3\": \"RNA_rep3.fq.gz\",\n",
    "    \"RNA_rep4\": \"RNA_rep4.fq.gz\",\n",
    "    \"DNA_rep1\": \"DNA_rep1.fq.gz\",\n",
    "    \"DNA_rep2\": \"DNA_rep2.fq.gz\",\n",
    "    \"DNA_rep3\": \"DNA_rep3.fq.gz\",\n",
    "    \"DNA_rep4\": \"DNA_rep4.fq.gz\",\n",
    "}\n",
    "\n",
    "for sample, fq in samples.items():\n",
    "    assert Path(fq).exists(), f\"missing FASTQ: {fq}\"\n",
    "\n",
    "threads = 16\n",
    "index_prefix = \"bt2_index/TILE_mpra\"\n",
    "\n",
    "for sample, fq in samples.items():\n",
    "    bam_out = Path(\"align_bam\") / f\"{sample}.sorted.bam\"\n",
    "    log_out = Path(\"align_logs\") / f\"{sample}.bowtie2.log\"\n",
    "\n",
    "    if bam_out.exists() and bam_out.with_suffix(\".bam.bai\").exists():\n",
    "        print(f\"[skip] {sample}: BAM and index already exist\")\n",
    "        continue\n",
    "\n",
    "    cmd = f\"\"\"\n",
    "    set -euo pipefail\n",
    "\n",
    "    \"{BOWTIE2}\" \\\n",
    "      --very-sensitive-local \\\n",
    "      --local \\\n",
    "      -p {threads} \\\n",
    "      -x \"{index_prefix}\" \\\n",
    "      -U \"{fq}\" \\\n",
    "      2> \"{log_out}\" \\\n",
    "    | \"{SAMTOOLS}\" view -@ {threads} -bS - \\\n",
    "    | \"{SAMTOOLS}\" sort -@ {threads} -o \"{bam_out}\" -\n",
    "\n",
    "    \"{SAMTOOLS}\" index \"{bam_out}\"\n",
    "    \"\"\"\n",
    "\n",
    "    print(f\"\\n[align] {sample}\")\n",
    "    subprocess.run(cmd, shell=True, check=True, executable=\"/bin/bash\")\n",
    "\n",
    "print(\"alignment done\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "041cbc71-87f7-4be2-9931-58f39a0dc58a",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libtinfow.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libtinfow.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libtinfow.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libtinfow.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libtinfow.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libtinfow.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libtinfow.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libtinfow.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libtinfow.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libtinfow.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libtinfow.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libtinfow.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libtinfow.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libtinfow.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libtinfow.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libtinfow.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libtinfow.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libtinfow.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libtinfow.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libtinfow.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libtinfow.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libtinfow.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libtinfow.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libtinfow.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>sample</th>\n",
       "      <th>bam_records</th>\n",
       "      <th>mapped_records</th>\n",
       "      <th>primary_mapped_records</th>\n",
       "      <th>mapped_fraction</th>\n",
       "      <th>primary_mapped_fraction</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>RNA_rep1</td>\n",
       "      <td>17209495</td>\n",
       "      <td>17121799</td>\n",
       "      <td>17121799</td>\n",
       "      <td>0.994904</td>\n",
       "      <td>0.994904</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>RNA_rep2</td>\n",
       "      <td>17708708</td>\n",
       "      <td>17672459</td>\n",
       "      <td>17672459</td>\n",
       "      <td>0.997953</td>\n",
       "      <td>0.997953</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>RNA_rep3</td>\n",
       "      <td>26358824</td>\n",
       "      <td>26260763</td>\n",
       "      <td>26260763</td>\n",
       "      <td>0.996280</td>\n",
       "      <td>0.996280</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>RNA_rep4</td>\n",
       "      <td>15859991</td>\n",
       "      <td>15760819</td>\n",
       "      <td>15760819</td>\n",
       "      <td>0.993747</td>\n",
       "      <td>0.993747</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>DNA_rep1</td>\n",
       "      <td>18680589</td>\n",
       "      <td>18634467</td>\n",
       "      <td>18634467</td>\n",
       "      <td>0.997531</td>\n",
       "      <td>0.997531</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>DNA_rep2</td>\n",
       "      <td>17111454</td>\n",
       "      <td>17071904</td>\n",
       "      <td>17071904</td>\n",
       "      <td>0.997689</td>\n",
       "      <td>0.997689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>DNA_rep3</td>\n",
       "      <td>16853932</td>\n",
       "      <td>16805900</td>\n",
       "      <td>16805900</td>\n",
       "      <td>0.997150</td>\n",
       "      <td>0.997150</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>DNA_rep4</td>\n",
       "      <td>18340853</td>\n",
       "      <td>18275825</td>\n",
       "      <td>18275825</td>\n",
       "      <td>0.996454</td>\n",
       "      <td>0.996454</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     sample  bam_records  mapped_records  primary_mapped_records  \\\n",
       "0  RNA_rep1     17209495        17121799                17121799   \n",
       "1  RNA_rep2     17708708        17672459                17672459   \n",
       "2  RNA_rep3     26358824        26260763                26260763   \n",
       "3  RNA_rep4     15859991        15760819                15760819   \n",
       "4  DNA_rep1     18680589        18634467                18634467   \n",
       "5  DNA_rep2     17111454        17071904                17071904   \n",
       "6  DNA_rep3     16853932        16805900                16805900   \n",
       "7  DNA_rep4     18340853        18275825                18275825   \n",
       "\n",
       "   mapped_fraction  primary_mapped_fraction  \n",
       "0         0.994904                 0.994904  \n",
       "1         0.997953                 0.997953  \n",
       "2         0.996280                 0.996280  \n",
       "3         0.993747                 0.993747  \n",
       "4         0.997531                 0.997531  \n",
       "5         0.997689                 0.997689  \n",
       "6         0.997150                 0.997150  \n",
       "7         0.996454                 0.996454  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import subprocess\n",
    "from pathlib import Path\n",
    "import pandas as pd\n",
    "\n",
    "rows = []\n",
    "\n",
    "for sample in samples:\n",
    "    bam = Path(\"align_bam\") / f\"{sample}.sorted.bam\"\n",
    "    assert bam.exists(), f\"missing BAM: {bam}\"\n",
    "\n",
    "    total = int(subprocess.check_output(\n",
    "        f'\"{SAMTOOLS}\" view -c \"{bam}\"',\n",
    "        shell=True,\n",
    "        text=True,\n",
    "        executable=\"/bin/bash\",\n",
    "    ).strip())\n",
    "\n",
    "    mapped = int(subprocess.check_output(\n",
    "        f'\"{SAMTOOLS}\" view -c -F 4 \"{bam}\"',\n",
    "        shell=True,\n",
    "        text=True,\n",
    "        executable=\"/bin/bash\",\n",
    "    ).strip())\n",
    "\n",
    "    primary_mapped = int(subprocess.check_output(\n",
    "        f'\"{SAMTOOLS}\" view -c -F 260 \"{bam}\"',\n",
    "        shell=True,\n",
    "        text=True,\n",
    "        executable=\"/bin/bash\",\n",
    "    ).strip())\n",
    "\n",
    "    rows.append({\n",
    "        \"sample\": sample,\n",
    "        \"bam_records\": total,\n",
    "        \"mapped_records\": mapped,\n",
    "        \"primary_mapped_records\": primary_mapped,\n",
    "        \"mapped_fraction\": mapped / total if total else None,\n",
    "        \"primary_mapped_fraction\": primary_mapped / total if total else None,\n",
    "    })\n",
    "\n",
    "qc = pd.DataFrame(rows)\n",
    "qc.to_csv(\"alignment_qc_summary.tsv\", sep=\"\\t\", index=False)\n",
    "qc"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "d66271c1-2408-4949-9edf-9d38bf8b48b1",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[count] RNA_rep1\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libtinfow.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[count] RNA_rep2\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libtinfow.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[count] RNA_rep3\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libtinfow.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[count] RNA_rep4\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libtinfow.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[count] DNA_rep1\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libtinfow.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[count] DNA_rep2\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libtinfow.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[count] DNA_rep3\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libtinfow.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[count] DNA_rep4\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libtinfow.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n",
      "/opt/conda/bin/samtools: /opt/conda/bin/../lib/libncursesw.so.6: no version information available (required by /opt/conda/bin/samtools)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "counting done\n"
     ]
    }
   ],
   "source": [
    "from pathlib import Path\n",
    "import subprocess\n",
    "\n",
    "Path(\"counts_per_alignid\").mkdir(exist_ok=True)\n",
    "\n",
    "mapq_min = 10\n",
    "\n",
    "for sample in samples:\n",
    "    bam = Path(\"align_bam\") / f\"{sample}.sorted.bam\"\n",
    "    out = Path(\"counts_per_alignid\") / f\"{sample}.counts.tsv\"\n",
    "\n",
    "    assert bam.exists(), f\"missing BAM: {bam}\"\n",
    "\n",
    "    if out.exists():\n",
    "        print(f\"[skip] {sample}: {out} exists\")\n",
    "        continue\n",
    "\n",
    "    cmd = f\"\"\"\n",
    "    set -euo pipefail\n",
    "\n",
    "    \"{SAMTOOLS}\" view -F 260 -q {mapq_min} \"{bam}\" \\\n",
    "      | awk '{{print $3}}' \\\n",
    "      | sort \\\n",
    "      | uniq -c \\\n",
    "      | awk 'BEGIN{{OFS=\"\\\\t\"}} {{print $2, $1}}' \\\n",
    "      > \"{out}\"\n",
    "    \"\"\"\n",
    "\n",
    "    print(f\"[count] {sample}\")\n",
    "    subprocess.run(cmd, shell=True, check=True, executable=\"/bin/bash\")\n",
    "\n",
    "print(\"counting done\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "3c09bd37-a91d-4f55-a956-8a36278badd8",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "barcode/count rows: (90819, 9)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>alignid</th>\n",
       "      <th>RNA_rep1</th>\n",
       "      <th>RNA_rep2</th>\n",
       "      <th>RNA_rep3</th>\n",
       "      <th>RNA_rep4</th>\n",
       "      <th>DNA_rep1</th>\n",
       "      <th>DNA_rep2</th>\n",
       "      <th>DNA_rep3</th>\n",
       "      <th>DNA_rep4</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>TILE_ID_001-00001|ROTAVIRUS_A|NC_011505.2|1,13...</td>\n",
       "      <td>94</td>\n",
       "      <td>83</td>\n",
       "      <td>70</td>\n",
       "      <td>26</td>\n",
       "      <td>109</td>\n",
       "      <td>102</td>\n",
       "      <td>115</td>\n",
       "      <td>90</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>TILE_ID_001-00001|ROTAVIRUS_A|NC_011505.2|1,13...</td>\n",
       "      <td>171</td>\n",
       "      <td>126</td>\n",
       "      <td>276</td>\n",
       "      <td>118</td>\n",
       "      <td>200</td>\n",
       "      <td>136</td>\n",
       "      <td>112</td>\n",
       "      <td>128</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>TILE_ID_001-00001|ROTAVIRUS_A|NC_011505.2|1,13...</td>\n",
       "      <td>157</td>\n",
       "      <td>126</td>\n",
       "      <td>135</td>\n",
       "      <td>169</td>\n",
       "      <td>184</td>\n",
       "      <td>134</td>\n",
       "      <td>115</td>\n",
       "      <td>169</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>TILE_ID_001-00002|ROTAVIRUS_A|NC_011505.2|66,1...</td>\n",
       "      <td>60</td>\n",
       "      <td>67</td>\n",
       "      <td>102</td>\n",
       "      <td>30</td>\n",
       "      <td>67</td>\n",
       "      <td>59</td>\n",
       "      <td>70</td>\n",
       "      <td>36</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>TILE_ID_001-00002|ROTAVIRUS_A|NC_011505.2|66,1...</td>\n",
       "      <td>131</td>\n",
       "      <td>190</td>\n",
       "      <td>220</td>\n",
       "      <td>66</td>\n",
       "      <td>146</td>\n",
       "      <td>187</td>\n",
       "      <td>130</td>\n",
       "      <td>175</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                             alignid  RNA_rep1  RNA_rep2  \\\n",
       "0  TILE_ID_001-00001|ROTAVIRUS_A|NC_011505.2|1,13...        94        83   \n",
       "1  TILE_ID_001-00001|ROTAVIRUS_A|NC_011505.2|1,13...       171       126   \n",
       "2  TILE_ID_001-00001|ROTAVIRUS_A|NC_011505.2|1,13...       157       126   \n",
       "3  TILE_ID_001-00002|ROTAVIRUS_A|NC_011505.2|66,1...        60        67   \n",
       "4  TILE_ID_001-00002|ROTAVIRUS_A|NC_011505.2|66,1...       131       190   \n",
       "\n",
       "   RNA_rep3  RNA_rep4  DNA_rep1  DNA_rep2  DNA_rep3  DNA_rep4  \n",
       "0        70        26       109       102       115        90  \n",
       "1       276       118       200       136       112       128  \n",
       "2       135       169       184       134       115       169  \n",
       "3       102        30        67        59        70        36  \n",
       "4       220        66       146       187       130       175  "
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from pathlib import Path\n",
    "from functools import reduce\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "\n",
    "samples = {\n",
    "    \"RNA_rep1\": \"RNA_rep1.fq.gz\",\n",
    "    \"RNA_rep2\": \"RNA_rep2.fq.gz\",\n",
    "    \"RNA_rep3\": \"RNA_rep3.fq.gz\",\n",
    "    \"RNA_rep4\": \"RNA_rep4.fq.gz\",\n",
    "    \"DNA_rep1\": \"DNA_rep1.fq.gz\",\n",
    "    \"DNA_rep2\": \"DNA_rep2.fq.gz\",\n",
    "    \"DNA_rep3\": \"DNA_rep3.fq.gz\",\n",
    "    \"DNA_rep4\": \"DNA_rep4.fq.gz\",\n",
    "}\n",
    "\n",
    "count_dir = Path(\"counts_per_alignid\")\n",
    "\n",
    "dfs = []\n",
    "for sample in samples:\n",
    "    p = count_dir / f\"{sample}.counts.tsv\"\n",
    "    assert p.exists(), f\"missing count file: {p}\"\n",
    "    df = pd.read_csv(p, sep=\"\\t\", names=[\"alignid\", sample])\n",
    "    dfs.append(df)\n",
    "\n",
    "count_bc = reduce(\n",
    "    lambda left, right: pd.merge(left, right, on=\"alignid\", how=\"outer\"),\n",
    "    dfs,\n",
    ").fillna(0)\n",
    "\n",
    "for sample in samples:\n",
    "    count_bc[sample] = count_bc[sample].astype(int)\n",
    "\n",
    "count_bc.to_csv(\"countBC_merge_alignid.tsv\", sep=\"\\t\", index=False)\n",
    "\n",
    "print(\"barcode/count rows:\", count_bc.shape)\n",
    "count_bc.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "b7b2602e-44c8-402d-a1cd-c53c0214642f",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "FASTA records: (91101, 9)\n",
      "unique parent elements: 30367\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>alignid</th>\n",
       "      <th>element_id</th>\n",
       "      <th>barcode_id</th>\n",
       "      <th>field_0</th>\n",
       "      <th>field_1</th>\n",
       "      <th>field_2</th>\n",
       "      <th>field_3</th>\n",
       "      <th>n_header_fields</th>\n",
       "      <th>seq</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>TILE_ID_001-00001|ROTAVIRUS_A|NC_011505.2|1,13...</td>\n",
       "      <td>TILE_ID_001-00001|ROTAVIRUS_A|NC_011505.2|1,130</td>\n",
       "      <td>ATCGGAA|0</td>\n",
       "      <td>TILE_ID_001-00001</td>\n",
       "      <td>ROTAVIRUS_A</td>\n",
       "      <td>NC_011505.2</td>\n",
       "      <td>1,130</td>\n",
       "      <td>6</td>\n",
       "      <td>ACTGGCCGCTTCACTGGGCTTTTAAAGCGCTACAGTGATGTCTCTC...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>TILE_ID_001-00001|ROTAVIRUS_A|NC_011505.2|1,13...</td>\n",
       "      <td>TILE_ID_001-00001|ROTAVIRUS_A|NC_011505.2|1,130</td>\n",
       "      <td>CGTACAA|1</td>\n",
       "      <td>TILE_ID_001-00001</td>\n",
       "      <td>ROTAVIRUS_A</td>\n",
       "      <td>NC_011505.2</td>\n",
       "      <td>1,130</td>\n",
       "      <td>6</td>\n",
       "      <td>ACTGGCCGCTTCACTGGGCTTTTAAAGCGCTACAGTGATGTCTCTC...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>TILE_ID_001-00001|ROTAVIRUS_A|NC_011505.2|1,13...</td>\n",
       "      <td>TILE_ID_001-00001|ROTAVIRUS_A|NC_011505.2|1,130</td>\n",
       "      <td>TAAGCGA|2</td>\n",
       "      <td>TILE_ID_001-00001</td>\n",
       "      <td>ROTAVIRUS_A</td>\n",
       "      <td>NC_011505.2</td>\n",
       "      <td>1,130</td>\n",
       "      <td>6</td>\n",
       "      <td>ACTGGCCGCTTCACTGGGCTTTTAAAGCGCTACAGTGATGTCTCTC...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>TILE_ID_001-00002|ROTAVIRUS_A|NC_011505.2|66,1...</td>\n",
       "      <td>TILE_ID_001-00002|ROTAVIRUS_A|NC_011505.2|66,195</td>\n",
       "      <td>ACTGGAA|0</td>\n",
       "      <td>TILE_ID_001-00002</td>\n",
       "      <td>ROTAVIRUS_A</td>\n",
       "      <td>NC_011505.2</td>\n",
       "      <td>66,195</td>\n",
       "      <td>6</td>\n",
       "      <td>ACTGGCCGCTTCACTGAACTATATATAAGAATGAATCGTCTTCAAC...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>TILE_ID_001-00002|ROTAVIRUS_A|NC_011505.2|66,1...</td>\n",
       "      <td>TILE_ID_001-00002|ROTAVIRUS_A|NC_011505.2|66,195</td>\n",
       "      <td>GTCACAA|1</td>\n",
       "      <td>TILE_ID_001-00002</td>\n",
       "      <td>ROTAVIRUS_A</td>\n",
       "      <td>NC_011505.2</td>\n",
       "      <td>66,195</td>\n",
       "      <td>6</td>\n",
       "      <td>ACTGGCCGCTTCACTGAACTATATATAAGAATGAATCGTCTTCAAC...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                             alignid  \\\n",
       "0  TILE_ID_001-00001|ROTAVIRUS_A|NC_011505.2|1,13...   \n",
       "1  TILE_ID_001-00001|ROTAVIRUS_A|NC_011505.2|1,13...   \n",
       "2  TILE_ID_001-00001|ROTAVIRUS_A|NC_011505.2|1,13...   \n",
       "3  TILE_ID_001-00002|ROTAVIRUS_A|NC_011505.2|66,1...   \n",
       "4  TILE_ID_001-00002|ROTAVIRUS_A|NC_011505.2|66,1...   \n",
       "\n",
       "                                         element_id barcode_id  \\\n",
       "0   TILE_ID_001-00001|ROTAVIRUS_A|NC_011505.2|1,130  ATCGGAA|0   \n",
       "1   TILE_ID_001-00001|ROTAVIRUS_A|NC_011505.2|1,130  CGTACAA|1   \n",
       "2   TILE_ID_001-00001|ROTAVIRUS_A|NC_011505.2|1,130  TAAGCGA|2   \n",
       "3  TILE_ID_001-00002|ROTAVIRUS_A|NC_011505.2|66,195  ACTGGAA|0   \n",
       "4  TILE_ID_001-00002|ROTAVIRUS_A|NC_011505.2|66,195  GTCACAA|1   \n",
       "\n",
       "             field_0      field_1      field_2 field_3  n_header_fields  \\\n",
       "0  TILE_ID_001-00001  ROTAVIRUS_A  NC_011505.2   1,130                6   \n",
       "1  TILE_ID_001-00001  ROTAVIRUS_A  NC_011505.2   1,130                6   \n",
       "2  TILE_ID_001-00001  ROTAVIRUS_A  NC_011505.2   1,130                6   \n",
       "3  TILE_ID_001-00002  ROTAVIRUS_A  NC_011505.2  66,195                6   \n",
       "4  TILE_ID_001-00002  ROTAVIRUS_A  NC_011505.2  66,195                6   \n",
       "\n",
       "                                                 seq  \n",
       "0  ACTGGCCGCTTCACTGGGCTTTTAAAGCGCTACAGTGATGTCTCTC...  \n",
       "1  ACTGGCCGCTTCACTGGGCTTTTAAAGCGCTACAGTGATGTCTCTC...  \n",
       "2  ACTGGCCGCTTCACTGGGCTTTTAAAGCGCTACAGTGATGTCTCTC...  \n",
       "3  ACTGGCCGCTTCACTGAACTATATATAAGAATGAATCGTCTTCAAC...  \n",
       "4  ACTGGCCGCTTCACTGAACTATATATAAGAATGAATCGTCTTCAAC...  "
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from pathlib import Path\n",
    "import pandas as pd\n",
    "\n",
    "def read_fasta_records(path):\n",
    "    records = []\n",
    "    header = None\n",
    "    seq_chunks = []\n",
    "\n",
    "    with open(path, \"rt\") as f:\n",
    "        for line in f:\n",
    "            line = line.strip()\n",
    "            if not line:\n",
    "                continue\n",
    "\n",
    "            if line.startswith(\">\"):\n",
    "                if header is not None:\n",
    "                    records.append((header, \"\".join(seq_chunks).upper()))\n",
    "                header = line[1:].split()[0]\n",
    "                seq_chunks = []\n",
    "            else:\n",
    "                seq_chunks.append(line)\n",
    "\n",
    "        if header is not None:\n",
    "            records.append((header, \"\".join(seq_chunks).upper()))\n",
    "\n",
    "    return records\n",
    "\n",
    "\n",
    "def parse_alignid(alignid):\n",
    "    parts = str(alignid).split(\"|\")\n",
    "\n",
    "    # Parent element key: original notebook used '|'.join(alignid.split('|')[:-2])\n",
    "    if len(parts) >= 3:\n",
    "        element_id = \"|\".join(parts[:-2])\n",
    "        barcode_id = \"|\".join(parts[-2:])\n",
    "    else:\n",
    "        element_id = alignid\n",
    "        barcode_id = None\n",
    "\n",
    "    out = {\n",
    "        \"alignid\": alignid,\n",
    "        \"element_id\": element_id,\n",
    "        \"barcode_id\": barcode_id,\n",
    "        \"field_0\": parts[0] if len(parts) > 0 else None,\n",
    "        \"field_1\": parts[1] if len(parts) > 1 else None,\n",
    "        \"field_2\": parts[2] if len(parts) > 2 else None,\n",
    "        \"field_3\": parts[3] if len(parts) > 3 else None,\n",
    "        \"n_header_fields\": len(parts),\n",
    "    }\n",
    "    return out\n",
    "\n",
    "\n",
    "records = read_fasta_records(\"TILE_mpra.fa\")\n",
    "\n",
    "meta_rows = []\n",
    "for alignid, seq in records:\n",
    "    row = parse_alignid(alignid)\n",
    "    row[\"seq\"] = seq\n",
    "    meta_rows.append(row)\n",
    "\n",
    "meta = pd.DataFrame(meta_rows)\n",
    "\n",
    "print(\"FASTA records:\", meta.shape)\n",
    "print(\"unique parent elements:\", meta[\"element_id\"].nunique())\n",
    "\n",
    "meta.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "3046efd0-2f5d-4a8d-a35a-04f0f911c083",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "metadata rows: 91101\n",
      "count rows: 90819\n",
      "merged rows: 91101\n",
      "elements: 30367\n",
      "\n",
      "barcodes per element:\n",
      "count    30367.0\n",
      "mean         3.0\n",
      "std          0.0\n",
      "min          3.0\n",
      "25%          3.0\n",
      "50%          3.0\n",
      "75%          3.0\n",
      "max          3.0\n",
      "Name: alignid, dtype: float64\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>alignid</th>\n",
       "      <th>element_id</th>\n",
       "      <th>barcode_id</th>\n",
       "      <th>field_0</th>\n",
       "      <th>field_1</th>\n",
       "      <th>field_2</th>\n",
       "      <th>field_3</th>\n",
       "      <th>n_header_fields</th>\n",
       "      <th>seq</th>\n",
       "      <th>RNA_rep1</th>\n",
       "      <th>RNA_rep2</th>\n",
       "      <th>RNA_rep3</th>\n",
       "      <th>RNA_rep4</th>\n",
       "      <th>DNA_rep1</th>\n",
       "      <th>DNA_rep2</th>\n",
       "      <th>DNA_rep3</th>\n",
       "      <th>DNA_rep4</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>TILE_ID_001-00001|ROTAVIRUS_A|NC_011505.2|1,13...</td>\n",
       "      <td>TILE_ID_001-00001|ROTAVIRUS_A|NC_011505.2|1,130</td>\n",
       "      <td>ATCGGAA|0</td>\n",
       "      <td>TILE_ID_001-00001</td>\n",
       "      <td>ROTAVIRUS_A</td>\n",
       "      <td>NC_011505.2</td>\n",
       "      <td>1,130</td>\n",
       "      <td>6</td>\n",
       "      <td>ACTGGCCGCTTCACTGGGCTTTTAAAGCGCTACAGTGATGTCTCTC...</td>\n",
       "      <td>94</td>\n",
       "      <td>83</td>\n",
       "      <td>70</td>\n",
       "      <td>26</td>\n",
       "      <td>109</td>\n",
       "      <td>102</td>\n",
       "      <td>115</td>\n",
       "      <td>90</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>TILE_ID_001-00001|ROTAVIRUS_A|NC_011505.2|1,13...</td>\n",
       "      <td>TILE_ID_001-00001|ROTAVIRUS_A|NC_011505.2|1,130</td>\n",
       "      <td>CGTACAA|1</td>\n",
       "      <td>TILE_ID_001-00001</td>\n",
       "      <td>ROTAVIRUS_A</td>\n",
       "      <td>NC_011505.2</td>\n",
       "      <td>1,130</td>\n",
       "      <td>6</td>\n",
       "      <td>ACTGGCCGCTTCACTGGGCTTTTAAAGCGCTACAGTGATGTCTCTC...</td>\n",
       "      <td>171</td>\n",
       "      <td>126</td>\n",
       "      <td>276</td>\n",
       "      <td>118</td>\n",
       "      <td>200</td>\n",
       "      <td>136</td>\n",
       "      <td>112</td>\n",
       "      <td>128</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>TILE_ID_001-00001|ROTAVIRUS_A|NC_011505.2|1,13...</td>\n",
       "      <td>TILE_ID_001-00001|ROTAVIRUS_A|NC_011505.2|1,130</td>\n",
       "      <td>TAAGCGA|2</td>\n",
       "      <td>TILE_ID_001-00001</td>\n",
       "      <td>ROTAVIRUS_A</td>\n",
       "      <td>NC_011505.2</td>\n",
       "      <td>1,130</td>\n",
       "      <td>6</td>\n",
       "      <td>ACTGGCCGCTTCACTGGGCTTTTAAAGCGCTACAGTGATGTCTCTC...</td>\n",
       "      <td>157</td>\n",
       "      <td>126</td>\n",
       "      <td>135</td>\n",
       "      <td>169</td>\n",
       "      <td>184</td>\n",
       "      <td>134</td>\n",
       "      <td>115</td>\n",
       "      <td>169</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>TILE_ID_001-00002|ROTAVIRUS_A|NC_011505.2|66,1...</td>\n",
       "      <td>TILE_ID_001-00002|ROTAVIRUS_A|NC_011505.2|66,195</td>\n",
       "      <td>ACTGGAA|0</td>\n",
       "      <td>TILE_ID_001-00002</td>\n",
       "      <td>ROTAVIRUS_A</td>\n",
       "      <td>NC_011505.2</td>\n",
       "      <td>66,195</td>\n",
       "      <td>6</td>\n",
       "      <td>ACTGGCCGCTTCACTGAACTATATATAAGAATGAATCGTCTTCAAC...</td>\n",
       "      <td>60</td>\n",
       "      <td>67</td>\n",
       "      <td>102</td>\n",
       "      <td>30</td>\n",
       "      <td>67</td>\n",
       "      <td>59</td>\n",
       "      <td>70</td>\n",
       "      <td>36</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>TILE_ID_001-00002|ROTAVIRUS_A|NC_011505.2|66,1...</td>\n",
       "      <td>TILE_ID_001-00002|ROTAVIRUS_A|NC_011505.2|66,195</td>\n",
       "      <td>GTCACAA|1</td>\n",
       "      <td>TILE_ID_001-00002</td>\n",
       "      <td>ROTAVIRUS_A</td>\n",
       "      <td>NC_011505.2</td>\n",
       "      <td>66,195</td>\n",
       "      <td>6</td>\n",
       "      <td>ACTGGCCGCTTCACTGAACTATATATAAGAATGAATCGTCTTCAAC...</td>\n",
       "      <td>18</td>\n",
       "      <td>10</td>\n",
       "      <td>77</td>\n",
       "      <td>34</td>\n",
       "      <td>28</td>\n",
       "      <td>26</td>\n",
       "      <td>45</td>\n",
       "      <td>24</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                             alignid  \\\n",
       "0  TILE_ID_001-00001|ROTAVIRUS_A|NC_011505.2|1,13...   \n",
       "1  TILE_ID_001-00001|ROTAVIRUS_A|NC_011505.2|1,13...   \n",
       "2  TILE_ID_001-00001|ROTAVIRUS_A|NC_011505.2|1,13...   \n",
       "3  TILE_ID_001-00002|ROTAVIRUS_A|NC_011505.2|66,1...   \n",
       "4  TILE_ID_001-00002|ROTAVIRUS_A|NC_011505.2|66,1...   \n",
       "\n",
       "                                         element_id barcode_id  \\\n",
       "0   TILE_ID_001-00001|ROTAVIRUS_A|NC_011505.2|1,130  ATCGGAA|0   \n",
       "1   TILE_ID_001-00001|ROTAVIRUS_A|NC_011505.2|1,130  CGTACAA|1   \n",
       "2   TILE_ID_001-00001|ROTAVIRUS_A|NC_011505.2|1,130  TAAGCGA|2   \n",
       "3  TILE_ID_001-00002|ROTAVIRUS_A|NC_011505.2|66,195  ACTGGAA|0   \n",
       "4  TILE_ID_001-00002|ROTAVIRUS_A|NC_011505.2|66,195  GTCACAA|1   \n",
       "\n",
       "             field_0      field_1      field_2 field_3  n_header_fields  \\\n",
       "0  TILE_ID_001-00001  ROTAVIRUS_A  NC_011505.2   1,130                6   \n",
       "1  TILE_ID_001-00001  ROTAVIRUS_A  NC_011505.2   1,130                6   \n",
       "2  TILE_ID_001-00001  ROTAVIRUS_A  NC_011505.2   1,130                6   \n",
       "3  TILE_ID_001-00002  ROTAVIRUS_A  NC_011505.2  66,195                6   \n",
       "4  TILE_ID_001-00002  ROTAVIRUS_A  NC_011505.2  66,195                6   \n",
       "\n",
       "                                                 seq  RNA_rep1  RNA_rep2  \\\n",
       "0  ACTGGCCGCTTCACTGGGCTTTTAAAGCGCTACAGTGATGTCTCTC...        94        83   \n",
       "1  ACTGGCCGCTTCACTGGGCTTTTAAAGCGCTACAGTGATGTCTCTC...       171       126   \n",
       "2  ACTGGCCGCTTCACTGGGCTTTTAAAGCGCTACAGTGATGTCTCTC...       157       126   \n",
       "3  ACTGGCCGCTTCACTGAACTATATATAAGAATGAATCGTCTTCAAC...        60        67   \n",
       "4  ACTGGCCGCTTCACTGAACTATATATAAGAATGAATCGTCTTCAAC...        18        10   \n",
       "\n",
       "   RNA_rep3  RNA_rep4  DNA_rep1  DNA_rep2  DNA_rep3  DNA_rep4  \n",
       "0        70        26       109       102       115        90  \n",
       "1       276       118       200       136       112       128  \n",
       "2       135       169       184       134       115       169  \n",
       "3       102        30        67        59        70        36  \n",
       "4        77        34        28        26        45        24  "
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "count_bc = pd.read_csv(\"countBC_merge_alignid.tsv\", sep=\"\\t\")\n",
    "\n",
    "bc = meta.merge(count_bc, on=\"alignid\", how=\"left\")\n",
    "\n",
    "for sample in samples:\n",
    "    bc[sample] = bc[sample].fillna(0).astype(int)\n",
    "\n",
    "# Basic diagnostics\n",
    "print(\"metadata rows:\", len(meta))\n",
    "print(\"count rows:\", len(count_bc))\n",
    "print(\"merged rows:\", len(bc))\n",
    "print(\"elements:\", bc[\"element_id\"].nunique())\n",
    "\n",
    "# Check barcode counts per element\n",
    "barcode_per_element = bc.groupby(\"element_id\")[\"alignid\"].nunique()\n",
    "print(\"\\nbarcodes per element:\")\n",
    "print(barcode_per_element.describe())\n",
    "\n",
    "bc.to_csv(\"barcode_level_counts_with_metadata.tsv\", sep=\"\\t\", index=False)\n",
    "\n",
    "bc.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "1073c065-9652-4b10-bdf0-c52d2fb44d1e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "elements with >1 unique sequence: 30367\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>element_id</th>\n",
       "      <th>n_unique_seq</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>TILE_ID_001-00001|ROTAVIRUS_A|NC_011505.2|1,130</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>TILE_ID_001-00002|ROTAVIRUS_A|NC_011505.2|66,195</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>TILE_ID_001-00003|ROTAVIRUS_A|NC_011505.2|131,260</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>TILE_ID_001-00004|ROTAVIRUS_A|NC_011505.2|196,325</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>TILE_ID_001-00005|ROTAVIRUS_A|NC_011505.2|261,390</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>TILE_ID_001-00006|ROTAVIRUS_A|NC_011505.2|326,455</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>TILE_ID_001-00007|ROTAVIRUS_A|NC_011505.2|391,520</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>TILE_ID_001-00008|ROTAVIRUS_A|NC_011505.2|456,585</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>TILE_ID_001-00009|ROTAVIRUS_A|NC_011505.2|515,644</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>TILE_ID_001-00010|ROTAVIRUS_A|NC_011504.2|1,130</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>TILE_ID_001-00011|ROTAVIRUS_A|NC_011504.2|66,195</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>TILE_ID_001-00012|ROTAVIRUS_A|NC_011504.2|131,260</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>TILE_ID_001-00013|ROTAVIRUS_A|NC_011504.2|196,325</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>TILE_ID_001-00014|ROTAVIRUS_A|NC_011504.2|261,390</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>TILE_ID_001-00015|ROTAVIRUS_A|NC_011504.2|326,455</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>TILE_ID_001-00016|ROTAVIRUS_A|NC_011504.2|391,520</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>TILE_ID_001-00017|ROTAVIRUS_A|NC_011504.2|456,585</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>TILE_ID_001-00018|ROTAVIRUS_A|NC_011504.2|521,650</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>TILE_ID_001-00019|ROTAVIRUS_A|NC_011504.2|586,715</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>TILE_ID_001-00020|ROTAVIRUS_A|NC_011504.2|622,751</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                           element_id  n_unique_seq\n",
       "0     TILE_ID_001-00001|ROTAVIRUS_A|NC_011505.2|1,130             3\n",
       "1    TILE_ID_001-00002|ROTAVIRUS_A|NC_011505.2|66,195             3\n",
       "2   TILE_ID_001-00003|ROTAVIRUS_A|NC_011505.2|131,260             3\n",
       "3   TILE_ID_001-00004|ROTAVIRUS_A|NC_011505.2|196,325             3\n",
       "4   TILE_ID_001-00005|ROTAVIRUS_A|NC_011505.2|261,390             3\n",
       "5   TILE_ID_001-00006|ROTAVIRUS_A|NC_011505.2|326,455             3\n",
       "6   TILE_ID_001-00007|ROTAVIRUS_A|NC_011505.2|391,520             3\n",
       "7   TILE_ID_001-00008|ROTAVIRUS_A|NC_011505.2|456,585             3\n",
       "8   TILE_ID_001-00009|ROTAVIRUS_A|NC_011505.2|515,644             3\n",
       "9     TILE_ID_001-00010|ROTAVIRUS_A|NC_011504.2|1,130             3\n",
       "10   TILE_ID_001-00011|ROTAVIRUS_A|NC_011504.2|66,195             3\n",
       "11  TILE_ID_001-00012|ROTAVIRUS_A|NC_011504.2|131,260             3\n",
       "12  TILE_ID_001-00013|ROTAVIRUS_A|NC_011504.2|196,325             3\n",
       "13  TILE_ID_001-00014|ROTAVIRUS_A|NC_011504.2|261,390             3\n",
       "14  TILE_ID_001-00015|ROTAVIRUS_A|NC_011504.2|326,455             3\n",
       "15  TILE_ID_001-00016|ROTAVIRUS_A|NC_011504.2|391,520             3\n",
       "16  TILE_ID_001-00017|ROTAVIRUS_A|NC_011504.2|456,585             3\n",
       "17  TILE_ID_001-00018|ROTAVIRUS_A|NC_011504.2|521,650             3\n",
       "18  TILE_ID_001-00019|ROTAVIRUS_A|NC_011504.2|586,715             3\n",
       "19  TILE_ID_001-00020|ROTAVIRUS_A|NC_011504.2|622,751             3"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>element_id</th>\n",
       "      <th>seq</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>TILE_ID_001-00001|ROTAVIRUS_A|NC_011505.2|1,130</td>\n",
       "      <td>ACTGGCCGCTTCACTGGGCTTTTAAAGCGCTACAGTGATGTCTCTC...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>TILE_ID_001-00002|ROTAVIRUS_A|NC_011505.2|66,195</td>\n",
       "      <td>ACTGGCCGCTTCACTGAACTATATATAAGAATGAATCGTCTTCAAC...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>TILE_ID_001-00003|ROTAVIRUS_A|NC_011505.2|131,260</td>\n",
       "      <td>ACTGGCCGCTTCACTGGTGAACAGTACATTTCACCAGATGCAGAAG...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>TILE_ID_001-00004|ROTAVIRUS_A|NC_011505.2|196,325</td>\n",
       "      <td>ACTGGCCGCTTCACTGGATATTGGACCATCTGATTCTGCTTCAAAC...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>TILE_ID_001-00005|ROTAVIRUS_A|NC_011505.2|261,390</td>\n",
       "      <td>ACTGGCCGCTTCACTGAGTTAAGACAAATGCAGACGCTGGCGTGTC...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                          element_id  \\\n",
       "0    TILE_ID_001-00001|ROTAVIRUS_A|NC_011505.2|1,130   \n",
       "1   TILE_ID_001-00002|ROTAVIRUS_A|NC_011505.2|66,195   \n",
       "2  TILE_ID_001-00003|ROTAVIRUS_A|NC_011505.2|131,260   \n",
       "3  TILE_ID_001-00004|ROTAVIRUS_A|NC_011505.2|196,325   \n",
       "4  TILE_ID_001-00005|ROTAVIRUS_A|NC_011505.2|261,390   \n",
       "\n",
       "                                                 seq  \n",
       "0  ACTGGCCGCTTCACTGGGCTTTTAAAGCGCTACAGTGATGTCTCTC...  \n",
       "1  ACTGGCCGCTTCACTGAACTATATATAAGAATGAATCGTCTTCAAC...  \n",
       "2  ACTGGCCGCTTCACTGGTGAACAGTACATTTCACCAGATGCAGAAG...  \n",
       "3  ACTGGCCGCTTCACTGGATATTGGACCATCTGATTCTGCTTCAAAC...  \n",
       "4  ACTGGCCGCTTCACTGAGTTAAGACAAATGCAGACGCTGGCGTGTC...  "
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "seq_check = (\n",
    "    bc.groupby(\"element_id\")[\"seq\"]\n",
    "    .nunique()\n",
    "    .reset_index(name=\"n_unique_seq\")\n",
    ")\n",
    "\n",
    "n_bad = (seq_check[\"n_unique_seq\"] > 1).sum()\n",
    "print(\"elements with >1 unique sequence:\", n_bad)\n",
    "\n",
    "if n_bad > 0:\n",
    "    display(seq_check[seq_check[\"n_unique_seq\"] > 1].head(20))\n",
    "\n",
    "# Collapse to one sequence per element.\n",
    "# If the library is well-formed, this should be safe.\n",
    "element_seq = (\n",
    "    bc[[\"element_id\", \"seq\"]]\n",
    "    .drop_duplicates()\n",
    "    .sort_values([\"element_id\", \"seq\"])\n",
    "    .groupby(\"element_id\", as_index=False)\n",
    "    .first()\n",
    ")\n",
    "\n",
    "element_seq.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "0607e09c-fe48-463b-b74b-4d3c4d26c96d",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "library sizes:\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>aligned_count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>RNA_rep1</th>\n",
       "      <td>17031527</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RNA_rep2</th>\n",
       "      <td>17580941</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RNA_rep3</th>\n",
       "      <td>25839192</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RNA_rep4</th>\n",
       "      <td>15309884</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DNA_rep1</th>\n",
       "      <td>18613038</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DNA_rep2</th>\n",
       "      <td>17053034</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DNA_rep3</th>\n",
       "      <td>16711933</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DNA_rep4</th>\n",
       "      <td>18192500</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          aligned_count\n",
       "RNA_rep1       17031527\n",
       "RNA_rep2       17580941\n",
       "RNA_rep3       25839192\n",
       "RNA_rep4       15309884\n",
       "DNA_rep1       18613038\n",
       "DNA_rep2       17053034\n",
       "DNA_rep3       16711933\n",
       "DNA_rep4       18192500"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>alignid</th>\n",
       "      <th>element_id</th>\n",
       "      <th>barcode_ratio_mean</th>\n",
       "      <th>barcode_ratio_sd</th>\n",
       "      <th>barcode_rna_total</th>\n",
       "      <th>barcode_dna_total</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>TILE_ID_001-00001|ROTAVIRUS_A|NC_011505.2|1,13...</td>\n",
       "      <td>TILE_ID_001-00001|ROTAVIRUS_A|NC_011505.2|1,130</td>\n",
       "      <td>-0.725349</td>\n",
       "      <td>0.621004</td>\n",
       "      <td>273</td>\n",
       "      <td>416</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>TILE_ID_001-00001|ROTAVIRUS_A|NC_011505.2|1,13...</td>\n",
       "      <td>TILE_ID_001-00001|ROTAVIRUS_A|NC_011505.2|1,130</td>\n",
       "      <td>0.129959</td>\n",
       "      <td>0.355931</td>\n",
       "      <td>691</td>\n",
       "      <td>576</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>TILE_ID_001-00001|ROTAVIRUS_A|NC_011505.2|1,13...</td>\n",
       "      <td>TILE_ID_001-00001|ROTAVIRUS_A|NC_011505.2|1,130</td>\n",
       "      <td>-0.087463</td>\n",
       "      <td>0.248187</td>\n",
       "      <td>587</td>\n",
       "      <td>602</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>TILE_ID_001-00002|ROTAVIRUS_A|NC_011505.2|66,1...</td>\n",
       "      <td>TILE_ID_001-00002|ROTAVIRUS_A|NC_011505.2|66,195</td>\n",
       "      <td>0.001963</td>\n",
       "      <td>0.085016</td>\n",
       "      <td>259</td>\n",
       "      <td>232</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>TILE_ID_001-00002|ROTAVIRUS_A|NC_011505.2|66,1...</td>\n",
       "      <td>TILE_ID_001-00002|ROTAVIRUS_A|NC_011505.2|66,195</td>\n",
       "      <td>-0.145289</td>\n",
       "      <td>0.645646</td>\n",
       "      <td>139</td>\n",
       "      <td>123</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                             alignid  \\\n",
       "0  TILE_ID_001-00001|ROTAVIRUS_A|NC_011505.2|1,13...   \n",
       "1  TILE_ID_001-00001|ROTAVIRUS_A|NC_011505.2|1,13...   \n",
       "2  TILE_ID_001-00001|ROTAVIRUS_A|NC_011505.2|1,13...   \n",
       "3  TILE_ID_001-00002|ROTAVIRUS_A|NC_011505.2|66,1...   \n",
       "4  TILE_ID_001-00002|ROTAVIRUS_A|NC_011505.2|66,1...   \n",
       "\n",
       "                                         element_id  barcode_ratio_mean  \\\n",
       "0   TILE_ID_001-00001|ROTAVIRUS_A|NC_011505.2|1,130           -0.725349   \n",
       "1   TILE_ID_001-00001|ROTAVIRUS_A|NC_011505.2|1,130            0.129959   \n",
       "2   TILE_ID_001-00001|ROTAVIRUS_A|NC_011505.2|1,130           -0.087463   \n",
       "3  TILE_ID_001-00002|ROTAVIRUS_A|NC_011505.2|66,195            0.001963   \n",
       "4  TILE_ID_001-00002|ROTAVIRUS_A|NC_011505.2|66,195           -0.145289   \n",
       "\n",
       "   barcode_ratio_sd  barcode_rna_total  barcode_dna_total  \n",
       "0          0.621004                273                416  \n",
       "1          0.355931                691                576  \n",
       "2          0.248187                587                602  \n",
       "3          0.085016                259                232  \n",
       "4          0.645646                139                123  "
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bc = pd.read_csv(\"barcode_level_counts_with_metadata.tsv\", sep=\"\\t\")\n",
    "\n",
    "rna_cols = [\"RNA_rep1\", \"RNA_rep2\", \"RNA_rep3\", \"RNA_rep4\"]\n",
    "dna_cols = [\"DNA_rep1\", \"DNA_rep2\", \"DNA_rep3\", \"DNA_rep4\"]\n",
    "\n",
    "sample_cols = rna_cols + dna_cols\n",
    "\n",
    "library_sizes = bc[sample_cols].sum(axis=0)\n",
    "print(\"library sizes:\")\n",
    "display(library_sizes.to_frame(\"aligned_count\"))\n",
    "\n",
    "# Counts per million\n",
    "for col in sample_cols:\n",
    "    bc[f\"{col}_cpm\"] = bc[col] / library_sizes[col] * 1_000_000\n",
    "\n",
    "# Pseudocount in CPM space\n",
    "# 0.5 CPM is moderate; if labels look too noisy, try 1.0.\n",
    "pc_cpm = 0.5\n",
    "\n",
    "for i in range(1, 5):\n",
    "    r = f\"RNA_rep{i}_cpm\"\n",
    "    d = f\"DNA_rep{i}_cpm\"\n",
    "    bc[f\"log2_ratio_rep{i}\"] = np.log2((bc[r] + pc_cpm) / (bc[d] + pc_cpm))\n",
    "\n",
    "ratio_cols = [f\"log2_ratio_rep{i}\" for i in range(1, 5)]\n",
    "\n",
    "bc[\"barcode_ratio_mean\"] = bc[ratio_cols].mean(axis=1)\n",
    "bc[\"barcode_ratio_sd\"] = bc[ratio_cols].std(axis=1)\n",
    "bc[\"barcode_rna_total\"] = bc[rna_cols].sum(axis=1)\n",
    "bc[\"barcode_dna_total\"] = bc[dna_cols].sum(axis=1)\n",
    "\n",
    "bc.to_csv(\"barcode_level_ratios.tsv\", sep=\"\\t\", index=False)\n",
    "\n",
    "bc[[\"alignid\", \"element_id\", \"barcode_ratio_mean\", \"barcode_ratio_sd\", \"barcode_rna_total\", \"barcode_dna_total\"]].head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "4584cad1-33ca-4aa1-8fa6-cb3158807ad0",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "total barcodes: 91101\n",
      "passing barcodes: 89612\n",
      "passing fraction: 0.9836555032326758\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "barcode_qc_pass\n",
       "True     89612\n",
       "False     1489\n",
       "Name: count, dtype: int64"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bc = pd.read_csv(\"barcode_level_ratios.tsv\", sep=\"\\t\")\n",
    "\n",
    "rna_cols = [\"RNA_rep1\", \"RNA_rep2\", \"RNA_rep3\", \"RNA_rep4\"]\n",
    "dna_cols = [\"DNA_rep1\", \"DNA_rep2\", \"DNA_rep3\", \"DNA_rep4\"]\n",
    "ratio_cols = [f\"log2_ratio_rep{i}\" for i in range(1, 5)]\n",
    "\n",
    "# Thresholds you can tune\n",
    "min_dna_total_per_barcode = 30\n",
    "min_dna_reps_with_reads = 2\n",
    "min_rna_total_per_barcode = 1\n",
    "max_barcode_rep_sd = 2.0\n",
    "\n",
    "bc[\"barcode_n_dna_reps_positive\"] = (bc[dna_cols] > 0).sum(axis=1)\n",
    "bc[\"barcode_n_rna_reps_positive\"] = (bc[rna_cols] > 0).sum(axis=1)\n",
    "\n",
    "bc[\"barcode_qc_pass\"] = (\n",
    "    (bc[\"barcode_dna_total\"] >= min_dna_total_per_barcode) &\n",
    "    (bc[\"barcode_n_dna_reps_positive\"] >= min_dna_reps_with_reads) &\n",
    "    (bc[\"barcode_rna_total\"] >= min_rna_total_per_barcode) &\n",
    "    (bc[\"barcode_ratio_sd\"] <= max_barcode_rep_sd)\n",
    ")\n",
    "\n",
    "print(\"total barcodes:\", len(bc))\n",
    "print(\"passing barcodes:\", int(bc[\"barcode_qc_pass\"].sum()))\n",
    "print(\"passing fraction:\", bc[\"barcode_qc_pass\"].mean())\n",
    "\n",
    "bc.to_csv(\"barcode_level_ratios_qc.tsv\", sep=\"\\t\", index=False)\n",
    "\n",
    "bc[\"barcode_qc_pass\"].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "bf736bcd-d15a-4fa6-9f4d-53fb33cc42c6",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/tmp/ipykernel_3418/1811304309.py:33: FutureWarning: DataFrameGroupBy.apply operated on the grouping columns. This behavior is deprecated, and in a future version of pandas the grouping columns will be excluded from the operation. Either pass `include_groups=False` to exclude the groupings or explicitly select the grouping columns after groupby to silence this warning.\n",
      "  .apply(summarize_element)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(30102, 16)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>element_id</th>\n",
       "      <th>n_total_barcodes</th>\n",
       "      <th>n_qc_barcodes</th>\n",
       "      <th>rna_total</th>\n",
       "      <th>dna_total</th>\n",
       "      <th>median_barcode_ratio_sd</th>\n",
       "      <th>max_barcode_ratio_sd</th>\n",
       "      <th>element_log2_ratio_rep1</th>\n",
       "      <th>element_log2_ratio_rep2</th>\n",
       "      <th>element_log2_ratio_rep3</th>\n",
       "      <th>element_log2_ratio_rep4</th>\n",
       "      <th>score_mean</th>\n",
       "      <th>score_median</th>\n",
       "      <th>score_sd_across_reps</th>\n",
       "      <th>score_range_across_reps</th>\n",
       "      <th>seq</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>TILE_ID_001-00001|ROTAVIRUS_A|NC_011505.2|1,130</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1551.0</td>\n",
       "      <td>1594.0</td>\n",
       "      <td>0.355931</td>\n",
       "      <td>0.621004</td>\n",
       "      <td>-0.093401</td>\n",
       "      <td>-0.144600</td>\n",
       "      <td>-0.366697</td>\n",
       "      <td>0.123159</td>\n",
       "      <td>-0.120385</td>\n",
       "      <td>-0.119000</td>\n",
       "      <td>0.201078</td>\n",
       "      <td>0.489855</td>\n",
       "      <td>ACTGGCCGCTTCACTGGGCTTTTAAAGCGCTACAGTGATGTCTCTC...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>TILE_ID_001-00002|ROTAVIRUS_A|NC_011505.2|66,195</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1005.0</td>\n",
       "      <td>993.0</td>\n",
       "      <td>0.553352</td>\n",
       "      <td>0.645646</td>\n",
       "      <td>-0.027265</td>\n",
       "      <td>-0.020100</td>\n",
       "      <td>0.122765</td>\n",
       "      <td>-0.011288</td>\n",
       "      <td>0.016028</td>\n",
       "      <td>-0.015694</td>\n",
       "      <td>0.071457</td>\n",
       "      <td>0.150029</td>\n",
       "      <td>ACTGGCCGCTTCACTGAACTATATATAAGAATGAATCGTCTTCAAC...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>TILE_ID_001-00003|ROTAVIRUS_A|NC_011505.2|131,260</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1642.0</td>\n",
       "      <td>1279.0</td>\n",
       "      <td>0.725516</td>\n",
       "      <td>1.000229</td>\n",
       "      <td>0.159878</td>\n",
       "      <td>-1.065968</td>\n",
       "      <td>0.555162</td>\n",
       "      <td>-0.184819</td>\n",
       "      <td>-0.133937</td>\n",
       "      <td>-0.012470</td>\n",
       "      <td>0.691003</td>\n",
       "      <td>1.621130</td>\n",
       "      <td>ACTGGCCGCTTCACTGGTGAACAGTACATTTCACCAGATGCAGAAG...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>TILE_ID_001-00004|ROTAVIRUS_A|NC_011505.2|196,325</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2775.0</td>\n",
       "      <td>2405.0</td>\n",
       "      <td>0.606397</td>\n",
       "      <td>0.619530</td>\n",
       "      <td>-0.314965</td>\n",
       "      <td>0.004966</td>\n",
       "      <td>0.446165</td>\n",
       "      <td>0.151581</td>\n",
       "      <td>0.071937</td>\n",
       "      <td>0.078274</td>\n",
       "      <td>0.316527</td>\n",
       "      <td>0.761130</td>\n",
       "      <td>ACTGGCCGCTTCACTGGATATTGGACCATCTGATTCTGCTTCAAAC...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>TILE_ID_001-00005|ROTAVIRUS_A|NC_011505.2|261,390</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1234.0</td>\n",
       "      <td>971.0</td>\n",
       "      <td>0.502387</td>\n",
       "      <td>1.008844</td>\n",
       "      <td>-0.100975</td>\n",
       "      <td>0.072137</td>\n",
       "      <td>0.795913</td>\n",
       "      <td>-0.144587</td>\n",
       "      <td>0.155622</td>\n",
       "      <td>-0.014419</td>\n",
       "      <td>0.437001</td>\n",
       "      <td>0.940500</td>\n",
       "      <td>ACTGGCCGCTTCACTGAGTTAAGACAAATGCAGACGCTGGCGTGTC...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                          element_id  n_total_barcodes  \\\n",
       "0    TILE_ID_001-00001|ROTAVIRUS_A|NC_011505.2|1,130               3.0   \n",
       "1   TILE_ID_001-00002|ROTAVIRUS_A|NC_011505.2|66,195               3.0   \n",
       "2  TILE_ID_001-00003|ROTAVIRUS_A|NC_011505.2|131,260               3.0   \n",
       "3  TILE_ID_001-00004|ROTAVIRUS_A|NC_011505.2|196,325               3.0   \n",
       "4  TILE_ID_001-00005|ROTAVIRUS_A|NC_011505.2|261,390               3.0   \n",
       "\n",
       "   n_qc_barcodes  rna_total  dna_total  median_barcode_ratio_sd  \\\n",
       "0            3.0     1551.0     1594.0                 0.355931   \n",
       "1            3.0     1005.0      993.0                 0.553352   \n",
       "2            3.0     1642.0     1279.0                 0.725516   \n",
       "3            3.0     2775.0     2405.0                 0.606397   \n",
       "4            3.0     1234.0      971.0                 0.502387   \n",
       "\n",
       "   max_barcode_ratio_sd  element_log2_ratio_rep1  element_log2_ratio_rep2  \\\n",
       "0              0.621004                -0.093401                -0.144600   \n",
       "1              0.645646                -0.027265                -0.020100   \n",
       "2              1.000229                 0.159878                -1.065968   \n",
       "3              0.619530                -0.314965                 0.004966   \n",
       "4              1.008844                -0.100975                 0.072137   \n",
       "\n",
       "   element_log2_ratio_rep3  element_log2_ratio_rep4  score_mean  score_median  \\\n",
       "0                -0.366697                 0.123159   -0.120385     -0.119000   \n",
       "1                 0.122765                -0.011288    0.016028     -0.015694   \n",
       "2                 0.555162                -0.184819   -0.133937     -0.012470   \n",
       "3                 0.446165                 0.151581    0.071937      0.078274   \n",
       "4                 0.795913                -0.144587    0.155622     -0.014419   \n",
       "\n",
       "   score_sd_across_reps  score_range_across_reps  \\\n",
       "0              0.201078                 0.489855   \n",
       "1              0.071457                 0.150029   \n",
       "2              0.691003                 1.621130   \n",
       "3              0.316527                 0.761130   \n",
       "4              0.437001                 0.940500   \n",
       "\n",
       "                                                 seq  \n",
       "0  ACTGGCCGCTTCACTGGGCTTTTAAAGCGCTACAGTGATGTCTCTC...  \n",
       "1  ACTGGCCGCTTCACTGAACTATATATAAGAATGAATCGTCTTCAAC...  \n",
       "2  ACTGGCCGCTTCACTGGTGAACAGTACATTTCACCAGATGCAGAAG...  \n",
       "3  ACTGGCCGCTTCACTGGATATTGGACCATCTGATTCTGCTTCAAAC...  \n",
       "4  ACTGGCCGCTTCACTGAGTTAAGACAAATGCAGACGCTGGCGTGTC...  "
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bc = pd.read_csv(\"barcode_level_ratios_qc.tsv\", sep=\"\\t\")\n",
    "\n",
    "ratio_cols = [f\"log2_ratio_rep{i}\" for i in range(1, 5)]\n",
    "\n",
    "bc_pass = bc[bc[\"barcode_qc_pass\"]].copy()\n",
    "\n",
    "def summarize_element(g):\n",
    "    out = {\n",
    "        \"n_total_barcodes\": g[\"alignid\"].nunique(),\n",
    "        \"n_qc_barcodes\": len(g),\n",
    "        \"rna_total\": g[[\"RNA_rep1\", \"RNA_rep2\", \"RNA_rep3\", \"RNA_rep4\"]].sum().sum(),\n",
    "        \"dna_total\": g[[\"DNA_rep1\", \"DNA_rep2\", \"DNA_rep3\", \"DNA_rep4\"]].sum().sum(),\n",
    "        \"median_barcode_ratio_sd\": g[\"barcode_ratio_sd\"].median(),\n",
    "        \"max_barcode_ratio_sd\": g[\"barcode_ratio_sd\"].max(),\n",
    "    }\n",
    "\n",
    "    # One robust element-level estimate per replicate\n",
    "    for col in ratio_cols:\n",
    "        out[f\"element_{col}\"] = g[col].median()\n",
    "\n",
    "    rep_values = np.array([out[f\"element_{col}\"] for col in ratio_cols], dtype=float)\n",
    "\n",
    "    out[\"score_mean\"] = np.nanmean(rep_values)\n",
    "    out[\"score_median\"] = np.nanmedian(rep_values)\n",
    "    out[\"score_sd_across_reps\"] = np.nanstd(rep_values, ddof=1)\n",
    "    out[\"score_range_across_reps\"] = np.nanmax(rep_values) - np.nanmin(rep_values)\n",
    "\n",
    "    return pd.Series(out)\n",
    "\n",
    "element_readout = (\n",
    "    bc_pass\n",
    "    .groupby(\"element_id\")\n",
    "    .apply(summarize_element)\n",
    "    .reset_index()\n",
    ")\n",
    "\n",
    "# Attach one sequence per element\n",
    "element_seq = (\n",
    "    bc[[\"element_id\", \"seq\"]]\n",
    "    .drop_duplicates()\n",
    "    .sort_values([\"element_id\", \"seq\"])\n",
    "    .groupby(\"element_id\", as_index=False)\n",
    "    .first()\n",
    ")\n",
    "\n",
    "element_readout = element_readout.merge(element_seq, on=\"element_id\", how=\"left\")\n",
    "\n",
    "element_readout.to_csv(\"element_level_readout_all_qc_barcodes.tsv\", sep=\"\\t\", index=False)\n",
    "\n",
    "print(element_readout.shape)\n",
    "element_readout.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "6ed6c2b6-1378-463b-9a67-b961c9bb2723",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "all elements with >=1 passing barcode: 30102\n",
      "consistent elements: 29355\n",
      "consistent fraction: 0.9751843731313534\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>element_id</th>\n",
       "      <th>score_mean</th>\n",
       "      <th>score_sd_across_reps</th>\n",
       "      <th>n_qc_barcodes</th>\n",
       "      <th>dna_total</th>\n",
       "      <th>rna_total</th>\n",
       "      <th>median_barcode_ratio_sd</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>TILE_ID_001-00001|ROTAVIRUS_A|NC_011505.2|1,130</td>\n",
       "      <td>-0.120385</td>\n",
       "      <td>0.201078</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1594.0</td>\n",
       "      <td>1551.0</td>\n",
       "      <td>0.355931</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>TILE_ID_001-00002|ROTAVIRUS_A|NC_011505.2|66,195</td>\n",
       "      <td>0.016028</td>\n",
       "      <td>0.071457</td>\n",
       "      <td>3.0</td>\n",
       "      <td>993.0</td>\n",
       "      <td>1005.0</td>\n",
       "      <td>0.553352</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>TILE_ID_001-00003|ROTAVIRUS_A|NC_011505.2|131,260</td>\n",
       "      <td>-0.133937</td>\n",
       "      <td>0.691003</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1279.0</td>\n",
       "      <td>1642.0</td>\n",
       "      <td>0.725516</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>TILE_ID_001-00004|ROTAVIRUS_A|NC_011505.2|196,325</td>\n",
       "      <td>0.071937</td>\n",
       "      <td>0.316527</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2405.0</td>\n",
       "      <td>2775.0</td>\n",
       "      <td>0.606397</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>TILE_ID_001-00005|ROTAVIRUS_A|NC_011505.2|261,390</td>\n",
       "      <td>0.155622</td>\n",
       "      <td>0.437001</td>\n",
       "      <td>3.0</td>\n",
       "      <td>971.0</td>\n",
       "      <td>1234.0</td>\n",
       "      <td>0.502387</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                          element_id  score_mean  \\\n",
       "0    TILE_ID_001-00001|ROTAVIRUS_A|NC_011505.2|1,130   -0.120385   \n",
       "1   TILE_ID_001-00002|ROTAVIRUS_A|NC_011505.2|66,195    0.016028   \n",
       "2  TILE_ID_001-00003|ROTAVIRUS_A|NC_011505.2|131,260   -0.133937   \n",
       "3  TILE_ID_001-00004|ROTAVIRUS_A|NC_011505.2|196,325    0.071937   \n",
       "4  TILE_ID_001-00005|ROTAVIRUS_A|NC_011505.2|261,390    0.155622   \n",
       "\n",
       "   score_sd_across_reps  n_qc_barcodes  dna_total  rna_total  \\\n",
       "0              0.201078            3.0     1594.0     1551.0   \n",
       "1              0.071457            3.0      993.0     1005.0   \n",
       "2              0.691003            3.0     1279.0     1642.0   \n",
       "3              0.316527            3.0     2405.0     2775.0   \n",
       "4              0.437001            3.0      971.0     1234.0   \n",
       "\n",
       "   median_barcode_ratio_sd  \n",
       "0                 0.355931  \n",
       "1                 0.553352  \n",
       "2                 0.725516  \n",
       "3                 0.606397  \n",
       "4                 0.502387  "
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "element_readout = pd.read_csv(\"element_level_readout_all_qc_barcodes.tsv\", sep=\"\\t\")\n",
    "\n",
    "# Conservative first-pass thresholds\n",
    "min_qc_barcodes_per_element = 2\n",
    "min_dna_total_per_element = 100\n",
    "max_element_rep_sd = 0.75\n",
    "max_element_rep_range = 2.0\n",
    "max_median_barcode_rep_sd = 1.25\n",
    "\n",
    "element_readout[\"consistent_label_pass\"] = (\n",
    "    (element_readout[\"n_qc_barcodes\"] >= min_qc_barcodes_per_element) &\n",
    "    (element_readout[\"dna_total\"] >= min_dna_total_per_element) &\n",
    "    (element_readout[\"score_sd_across_reps\"] <= max_element_rep_sd) &\n",
    "    (element_readout[\"score_range_across_reps\"] <= max_element_rep_range) &\n",
    "    (element_readout[\"median_barcode_ratio_sd\"] <= max_median_barcode_rep_sd)\n",
    ")\n",
    "\n",
    "consistent = element_readout[element_readout[\"consistent_label_pass\"]].copy()\n",
    "\n",
    "print(\"all elements with >=1 passing barcode:\", len(element_readout))\n",
    "print(\"consistent elements:\", len(consistent))\n",
    "print(\"consistent fraction:\", len(consistent) / len(element_readout))\n",
    "\n",
    "consistent.to_csv(\"element_level_readout_consistent.tsv\", sep=\"\\t\", index=False)\n",
    "\n",
    "consistent[\n",
    "    [\n",
    "        \"element_id\",\n",
    "        \"score_mean\",\n",
    "        \"score_sd_across_reps\",\n",
    "        \"n_qc_barcodes\",\n",
    "        \"dna_total\",\n",
    "        \"rna_total\",\n",
    "        \"median_barcode_ratio_sd\",\n",
    "    ]\n",
    "].head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "0a0ea441-93a2-4bca-afde-fa9a4029dac9",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "consistent labels: 29355\n",
      "stabilizing elements: 17030\n",
      "neutral/destabilizing elements: 12325\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>element_id</th>\n",
       "      <th>score_mean</th>\n",
       "      <th>score_sd_across_reps</th>\n",
       "      <th>n_qc_barcodes</th>\n",
       "      <th>seq</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>24813</th>\n",
       "      <td>TILE_ID_138-00443|HUMAN_GAMMAHERPESVIRUS_4_(EP...</td>\n",
       "      <td>1.821754</td>\n",
       "      <td>0.186730</td>\n",
       "      <td>3.0</td>\n",
       "      <td>ACTGGCCGCTTCACTGATTAATGTCCAGTGGGGTAAATGCACCTTG...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8612</th>\n",
       "      <td>TILE_ID_076-00050|BORNA_DISEASE_VIRUS_1_(BODV-...</td>\n",
       "      <td>1.222670</td>\n",
       "      <td>0.216589</td>\n",
       "      <td>3.0</td>\n",
       "      <td>ACTGGCCGCTTCACTGCTGGGGTGAAACCGAGGAGGATTCGGTACA...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28414</th>\n",
       "      <td>TILE_ID_143-00201|HUMAN_BETAHERPESVIRUS_5_(HHV...</td>\n",
       "      <td>1.205050</td>\n",
       "      <td>0.402041</td>\n",
       "      <td>3.0</td>\n",
       "      <td>ACTGGCCGCTTCACTGGGTCGCTGACGTCAAACCATCTCGTGCTCG...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5490</th>\n",
       "      <td>TILE_ID_045-00025|HUMAN_PARVOVIRUS_B19|NC_0008...</td>\n",
       "      <td>1.192586</td>\n",
       "      <td>0.724647</td>\n",
       "      <td>3.0</td>\n",
       "      <td>ACTGGCCGCTTCACTGGGGTTGGCTCTGGGCCAGCGCTTGGGGTTG...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22341</th>\n",
       "      <td>TILE_ID_133-00301|ORF_VIRUS|NC_005336.1|134347...</td>\n",
       "      <td>1.187143</td>\n",
       "      <td>0.286917</td>\n",
       "      <td>3.0</td>\n",
       "      <td>ACTGGCCGCTTCACTGCCGGACCCCGCCTGTGCGGGCGAGAGCGCG...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25998</th>\n",
       "      <td>TILE_ID_140-00299|MOLLUSCUM_CONTAGIOSUM_VIRUS_...</td>\n",
       "      <td>1.185225</td>\n",
       "      <td>0.188261</td>\n",
       "      <td>3.0</td>\n",
       "      <td>ACTGGCCGCTTCACTGGGCGAGTCAGTACATGCCCGGCCTCGCCTC...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19240</th>\n",
       "      <td>TILE_ID_126-00243|HUMAN_CORONAVIRUS_HKU1_(HCOV...</td>\n",
       "      <td>1.093117</td>\n",
       "      <td>0.568881</td>\n",
       "      <td>3.0</td>\n",
       "      <td>ACTGGCCGCTTCACTGTATGTAAGCATTTTAGTATGATGATTTTGA...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8530</th>\n",
       "      <td>TILE_ID_075-00103|HUSAVIRUS_SP.|NC_032480.1|66...</td>\n",
       "      <td>1.056462</td>\n",
       "      <td>0.429773</td>\n",
       "      <td>3.0</td>\n",
       "      <td>ACTGGCCGCTTCACTGGGTGTTGACGAAAGTTAACGCTGTGTATGA...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24812</th>\n",
       "      <td>TILE_ID_138-00442|HUMAN_GAMMAHERPESVIRUS_4_(EP...</td>\n",
       "      <td>1.033406</td>\n",
       "      <td>0.343744</td>\n",
       "      <td>3.0</td>\n",
       "      <td>ACTGGCCGCTTCACTGGCTTGGGAACACCGGGCAGGTCCGTGAGAA...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13815</th>\n",
       "      <td>TILE_ID_108-00035|SIMIAN_FOAMY_VIRUS|NC_001364...</td>\n",
       "      <td>0.956662</td>\n",
       "      <td>0.560202</td>\n",
       "      <td>3.0</td>\n",
       "      <td>ACTGGCCGCTTCACTGCTCAGATGACTAGAGATGAATTAGAAGATA...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27273</th>\n",
       "      <td>TILE_ID_141-00868|NY_014_POXVIRUS|NC_035469.1|...</td>\n",
       "      <td>0.938897</td>\n",
       "      <td>0.277701</td>\n",
       "      <td>3.0</td>\n",
       "      <td>ACTGGCCGCTTCACTGATAATAACCTGTATCATTTTTCTGACATCG...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17793</th>\n",
       "      <td>TILE_ID_123-00116|BREDA_VIRUS|NC_007447.1|7510...</td>\n",
       "      <td>0.916739</td>\n",
       "      <td>0.158169</td>\n",
       "      <td>3.0</td>\n",
       "      <td>ACTGGCCGCTTCACTGTATGTCTTTTTGTTATGGTTAAGTCATGTT...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16849</th>\n",
       "      <td>TILE_ID_120-00285|MARBURG_MARBURGVIRUS|NC_0016...</td>\n",
       "      <td>0.904866</td>\n",
       "      <td>0.344100</td>\n",
       "      <td>3.0</td>\n",
       "      <td>ACTGGCCGCTTCACTGAGAGAGAACTTCATTTAATTCACAAAAACA...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29305</th>\n",
       "      <td>TILE_ID_144-00012|POSITIVE_CONTROL(SL27)|GU937...</td>\n",
       "      <td>0.882973</td>\n",
       "      <td>0.219415</td>\n",
       "      <td>3.0</td>\n",
       "      <td>ACTGGCCGCTTCACTGTCCATTCCTCGTAGGCTGGTCCTGGGGAAC...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27789</th>\n",
       "      <td>TILE_ID_142-00551|COWPOX_VIRUS|NC_003663.2|293...</td>\n",
       "      <td>0.871065</td>\n",
       "      <td>0.316027</td>\n",
       "      <td>3.0</td>\n",
       "      <td>ACTGGCCGCTTCACTGTTATACATCTGGTGGAGGTGGAATGTGGGG...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26276</th>\n",
       "      <td>TILE_ID_140-00585|MOLLUSCUM_CONTAGIOSUM_VIRUS_...</td>\n",
       "      <td>0.870126</td>\n",
       "      <td>0.132511</td>\n",
       "      <td>3.0</td>\n",
       "      <td>ACTGGCCGCTTCACTGACACACGCACCCAAATCGCGTTTGCCAGGC...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>103</th>\n",
       "      <td>TILE_ID_001-00110|ROTAVIRUS_A|NC_011500.2|1366...</td>\n",
       "      <td>0.864396</td>\n",
       "      <td>0.512135</td>\n",
       "      <td>3.0</td>\n",
       "      <td>ACTGGCCGCTTCACTGAAATCAATTAGCAGTAAATGGTATAATGTT...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13371</th>\n",
       "      <td>TILE_ID_105-00132|BAS-CONGO_TIBROVIRUS|NC_0430...</td>\n",
       "      <td>0.861948</td>\n",
       "      <td>0.340317</td>\n",
       "      <td>3.0</td>\n",
       "      <td>ACTGGCCGCTTCACTGCTTTGTAACTGAGAGTTTAACATTCTGGAA...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25211</th>\n",
       "      <td>TILE_ID_139-00360|VARIOLA_VIRUS|NC_001611.1|17...</td>\n",
       "      <td>0.859064</td>\n",
       "      <td>0.217979</td>\n",
       "      <td>3.0</td>\n",
       "      <td>ACTGGCCGCTTCACTGTGTTATATACACTACATATTTTTATGTCAT...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19117</th>\n",
       "      <td>TILE_ID_126-00115|HUMAN_CORONAVIRUS_HKU1_(HCOV...</td>\n",
       "      <td>0.856717</td>\n",
       "      <td>0.394683</td>\n",
       "      <td>3.0</td>\n",
       "      <td>ACTGGCCGCTTCACTGTATTGGTTATTCATTATACACAGTATGGTT...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                              element_id  score_mean  \\\n",
       "24813  TILE_ID_138-00443|HUMAN_GAMMAHERPESVIRUS_4_(EP...    1.821754   \n",
       "8612   TILE_ID_076-00050|BORNA_DISEASE_VIRUS_1_(BODV-...    1.222670   \n",
       "28414  TILE_ID_143-00201|HUMAN_BETAHERPESVIRUS_5_(HHV...    1.205050   \n",
       "5490   TILE_ID_045-00025|HUMAN_PARVOVIRUS_B19|NC_0008...    1.192586   \n",
       "22341  TILE_ID_133-00301|ORF_VIRUS|NC_005336.1|134347...    1.187143   \n",
       "25998  TILE_ID_140-00299|MOLLUSCUM_CONTAGIOSUM_VIRUS_...    1.185225   \n",
       "19240  TILE_ID_126-00243|HUMAN_CORONAVIRUS_HKU1_(HCOV...    1.093117   \n",
       "8530   TILE_ID_075-00103|HUSAVIRUS_SP.|NC_032480.1|66...    1.056462   \n",
       "24812  TILE_ID_138-00442|HUMAN_GAMMAHERPESVIRUS_4_(EP...    1.033406   \n",
       "13815  TILE_ID_108-00035|SIMIAN_FOAMY_VIRUS|NC_001364...    0.956662   \n",
       "27273  TILE_ID_141-00868|NY_014_POXVIRUS|NC_035469.1|...    0.938897   \n",
       "17793  TILE_ID_123-00116|BREDA_VIRUS|NC_007447.1|7510...    0.916739   \n",
       "16849  TILE_ID_120-00285|MARBURG_MARBURGVIRUS|NC_0016...    0.904866   \n",
       "29305  TILE_ID_144-00012|POSITIVE_CONTROL(SL27)|GU937...    0.882973   \n",
       "27789  TILE_ID_142-00551|COWPOX_VIRUS|NC_003663.2|293...    0.871065   \n",
       "26276  TILE_ID_140-00585|MOLLUSCUM_CONTAGIOSUM_VIRUS_...    0.870126   \n",
       "103    TILE_ID_001-00110|ROTAVIRUS_A|NC_011500.2|1366...    0.864396   \n",
       "13371  TILE_ID_105-00132|BAS-CONGO_TIBROVIRUS|NC_0430...    0.861948   \n",
       "25211  TILE_ID_139-00360|VARIOLA_VIRUS|NC_001611.1|17...    0.859064   \n",
       "19117  TILE_ID_126-00115|HUMAN_CORONAVIRUS_HKU1_(HCOV...    0.856717   \n",
       "\n",
       "       score_sd_across_reps  n_qc_barcodes  \\\n",
       "24813              0.186730            3.0   \n",
       "8612               0.216589            3.0   \n",
       "28414              0.402041            3.0   \n",
       "5490               0.724647            3.0   \n",
       "22341              0.286917            3.0   \n",
       "25998              0.188261            3.0   \n",
       "19240              0.568881            3.0   \n",
       "8530               0.429773            3.0   \n",
       "24812              0.343744            3.0   \n",
       "13815              0.560202            3.0   \n",
       "27273              0.277701            3.0   \n",
       "17793              0.158169            3.0   \n",
       "16849              0.344100            3.0   \n",
       "29305              0.219415            3.0   \n",
       "27789              0.316027            3.0   \n",
       "26276              0.132511            3.0   \n",
       "103                0.512135            3.0   \n",
       "13371              0.340317            3.0   \n",
       "25211              0.217979            3.0   \n",
       "19117              0.394683            3.0   \n",
       "\n",
       "                                                     seq  \n",
       "24813  ACTGGCCGCTTCACTGATTAATGTCCAGTGGGGTAAATGCACCTTG...  \n",
       "8612   ACTGGCCGCTTCACTGCTGGGGTGAAACCGAGGAGGATTCGGTACA...  \n",
       "28414  ACTGGCCGCTTCACTGGGTCGCTGACGTCAAACCATCTCGTGCTCG...  \n",
       "5490   ACTGGCCGCTTCACTGGGGTTGGCTCTGGGCCAGCGCTTGGGGTTG...  \n",
       "22341  ACTGGCCGCTTCACTGCCGGACCCCGCCTGTGCGGGCGAGAGCGCG...  \n",
       "25998  ACTGGCCGCTTCACTGGGCGAGTCAGTACATGCCCGGCCTCGCCTC...  \n",
       "19240  ACTGGCCGCTTCACTGTATGTAAGCATTTTAGTATGATGATTTTGA...  \n",
       "8530   ACTGGCCGCTTCACTGGGTGTTGACGAAAGTTAACGCTGTGTATGA...  \n",
       "24812  ACTGGCCGCTTCACTGGCTTGGGAACACCGGGCAGGTCCGTGAGAA...  \n",
       "13815  ACTGGCCGCTTCACTGCTCAGATGACTAGAGATGAATTAGAAGATA...  \n",
       "27273  ACTGGCCGCTTCACTGATAATAACCTGTATCATTTTTCTGACATCG...  \n",
       "17793  ACTGGCCGCTTCACTGTATGTCTTTTTGTTATGGTTAAGTCATGTT...  \n",
       "16849  ACTGGCCGCTTCACTGAGAGAGAACTTCATTTAATTCACAAAAACA...  \n",
       "29305  ACTGGCCGCTTCACTGTCCATTCCTCGTAGGCTGGTCCTGGGGAAC...  \n",
       "27789  ACTGGCCGCTTCACTGTTATACATCTGGTGGAGGTGGAATGTGGGG...  \n",
       "26276  ACTGGCCGCTTCACTGACACACGCACCCAAATCGCGTTTGCCAGGC...  \n",
       "103    ACTGGCCGCTTCACTGAAATCAATTAGCAGTAAATGGTATAATGTT...  \n",
       "13371  ACTGGCCGCTTCACTGCTTTGTAACTGAGAGTTTAACATTCTGGAA...  \n",
       "25211  ACTGGCCGCTTCACTGTGTTATATACACTACATATTTTTATGTCAT...  \n",
       "19117  ACTGGCCGCTTCACTGTATTGGTTATTCATTATACACAGTATGGTT...  "
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "consistent = pd.read_csv(\"element_level_readout_consistent.tsv\", sep=\"\\t\")\n",
    "\n",
    "# Positive score means RNA/DNA is enriched relative to input.\n",
    "# You can tune this threshold.\n",
    "stabilizing_threshold = 0.0\n",
    "\n",
    "stabilizing = consistent[consistent[\"score_mean\"] > stabilizing_threshold].copy()\n",
    "destabilizing_or_neutral = consistent[consistent[\"score_mean\"] <= stabilizing_threshold].copy()\n",
    "\n",
    "print(\"consistent labels:\", len(consistent))\n",
    "print(\"stabilizing elements:\", len(stabilizing))\n",
    "print(\"neutral/destabilizing elements:\", len(destabilizing_or_neutral))\n",
    "\n",
    "stabilizing.to_csv(\"stabilizing_elements_consistent.tsv\", sep=\"\\t\", index=False)\n",
    "\n",
    "stabilizing[\n",
    "    [\n",
    "        \"element_id\",\n",
    "        \"score_mean\",\n",
    "        \"score_sd_across_reps\",\n",
    "        \"n_qc_barcodes\",\n",
    "        \"seq\",\n",
    "    ]\n",
    "].sort_values(\"score_mean\", ascending=False).head(20)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "d76ff9e3-e1a4-48cd-8505-8b7558356f42",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(29355, 9)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>element_id</th>\n",
       "      <th>seq</th>\n",
       "      <th>score</th>\n",
       "      <th>score_median</th>\n",
       "      <th>score_sd_across_reps</th>\n",
       "      <th>n_qc_barcodes</th>\n",
       "      <th>rna_total</th>\n",
       "      <th>dna_total</th>\n",
       "      <th>median_barcode_ratio_sd</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>TILE_ID_001-00001|ROTAVIRUS_A|NC_011505.2|1,130</td>\n",
       "      <td>ACTGGCCGCTTCACTGGGCTTTTAAAGCGCTACAGTGATGTCTCTC...</td>\n",
       "      <td>-0.120385</td>\n",
       "      <td>-0.119000</td>\n",
       "      <td>0.201078</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1551.0</td>\n",
       "      <td>1594.0</td>\n",
       "      <td>0.355931</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>TILE_ID_001-00002|ROTAVIRUS_A|NC_011505.2|66,195</td>\n",
       "      <td>ACTGGCCGCTTCACTGAACTATATATAAGAATGAATCGTCTTCAAC...</td>\n",
       "      <td>0.016028</td>\n",
       "      <td>-0.015694</td>\n",
       "      <td>0.071457</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1005.0</td>\n",
       "      <td>993.0</td>\n",
       "      <td>0.553352</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>TILE_ID_001-00003|ROTAVIRUS_A|NC_011505.2|131,260</td>\n",
       "      <td>ACTGGCCGCTTCACTGGTGAACAGTACATTTCACCAGATGCAGAAG...</td>\n",
       "      <td>-0.133937</td>\n",
       "      <td>-0.012470</td>\n",
       "      <td>0.691003</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1642.0</td>\n",
       "      <td>1279.0</td>\n",
       "      <td>0.725516</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>TILE_ID_001-00004|ROTAVIRUS_A|NC_011505.2|196,325</td>\n",
       "      <td>ACTGGCCGCTTCACTGGATATTGGACCATCTGATTCTGCTTCAAAC...</td>\n",
       "      <td>0.071937</td>\n",
       "      <td>0.078274</td>\n",
       "      <td>0.316527</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2775.0</td>\n",
       "      <td>2405.0</td>\n",
       "      <td>0.606397</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>TILE_ID_001-00005|ROTAVIRUS_A|NC_011505.2|261,390</td>\n",
       "      <td>ACTGGCCGCTTCACTGAGTTAAGACAAATGCAGACGCTGGCGTGTC...</td>\n",
       "      <td>0.155622</td>\n",
       "      <td>-0.014419</td>\n",
       "      <td>0.437001</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1234.0</td>\n",
       "      <td>971.0</td>\n",
       "      <td>0.502387</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                          element_id  \\\n",
       "0    TILE_ID_001-00001|ROTAVIRUS_A|NC_011505.2|1,130   \n",
       "1   TILE_ID_001-00002|ROTAVIRUS_A|NC_011505.2|66,195   \n",
       "2  TILE_ID_001-00003|ROTAVIRUS_A|NC_011505.2|131,260   \n",
       "3  TILE_ID_001-00004|ROTAVIRUS_A|NC_011505.2|196,325   \n",
       "4  TILE_ID_001-00005|ROTAVIRUS_A|NC_011505.2|261,390   \n",
       "\n",
       "                                                 seq     score  score_median  \\\n",
       "0  ACTGGCCGCTTCACTGGGCTTTTAAAGCGCTACAGTGATGTCTCTC... -0.120385     -0.119000   \n",
       "1  ACTGGCCGCTTCACTGAACTATATATAAGAATGAATCGTCTTCAAC...  0.016028     -0.015694   \n",
       "2  ACTGGCCGCTTCACTGGTGAACAGTACATTTCACCAGATGCAGAAG... -0.133937     -0.012470   \n",
       "3  ACTGGCCGCTTCACTGGATATTGGACCATCTGATTCTGCTTCAAAC...  0.071937      0.078274   \n",
       "4  ACTGGCCGCTTCACTGAGTTAAGACAAATGCAGACGCTGGCGTGTC...  0.155622     -0.014419   \n",
       "\n",
       "   score_sd_across_reps  n_qc_barcodes  rna_total  dna_total  \\\n",
       "0              0.201078            3.0     1551.0     1594.0   \n",
       "1              0.071457            3.0     1005.0      993.0   \n",
       "2              0.691003            3.0     1642.0     1279.0   \n",
       "3              0.316527            3.0     2775.0     2405.0   \n",
       "4              0.437001            3.0     1234.0      971.0   \n",
       "\n",
       "   median_barcode_ratio_sd  \n",
       "0                 0.355931  \n",
       "1                 0.553352  \n",
       "2                 0.725516  \n",
       "3                 0.606397  \n",
       "4                 0.502387  "
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "consistent = pd.read_csv(\"element_level_readout_consistent.tsv\", sep=\"\\t\")\n",
    "\n",
    "model_df = consistent.copy()\n",
    "\n",
    "# Clean sequence\n",
    "model_df[\"seq\"] = (\n",
    "    model_df[\"seq\"]\n",
    "    .astype(str)\n",
    "    .str.upper()\n",
    "    .str.replace(\"U\", \"T\", regex=False)\n",
    "    .str.replace(\"[^ACGT]\", \"\", regex=True)\n",
    ")\n",
    "\n",
    "model_ready = model_df[\n",
    "    [\n",
    "        \"element_id\",\n",
    "        \"seq\",\n",
    "        \"score_mean\",\n",
    "        \"score_median\",\n",
    "        \"score_sd_across_reps\",\n",
    "        \"n_qc_barcodes\",\n",
    "        \"rna_total\",\n",
    "        \"dna_total\",\n",
    "        \"median_barcode_ratio_sd\",\n",
    "    ]\n",
    "].rename(columns={\"score_mean\": \"score\"})\n",
    "\n",
    "# Remove empty / malformed sequences\n",
    "model_ready = model_ready[model_ready[\"seq\"].str.len() > 0].copy()\n",
    "\n",
    "model_ready.to_csv(\"training_seq_score_consistent_from_alignment.tsv\", sep=\"\\t\", index=False)\n",
    "\n",
    "print(model_ready.shape)\n",
    "model_ready.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "1db7ad5c-602d-45b7-ae82-464c7e6153bd",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 600x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 600x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 600x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "element_readout = pd.read_csv(\"element_level_readout_all_qc_barcodes.tsv\", sep=\"\\t\")\n",
    "consistent = pd.read_csv(\"element_level_readout_consistent.tsv\", sep=\"\\t\")\n",
    "\n",
    "plt.figure(figsize=(6, 4))\n",
    "plt.hist(element_readout[\"score_sd_across_reps\"].dropna(), bins=80)\n",
    "plt.axvline(0.75, linestyle=\"--\")\n",
    "plt.xlabel(\"Element score SD across 4 replicate RNA/DNA pairs\")\n",
    "plt.ylabel(\"Number of elements\")\n",
    "plt.title(\"Replicate consistency\")\n",
    "plt.show()\n",
    "\n",
    "plt.figure(figsize=(6, 4))\n",
    "plt.hist(element_readout[\"score_mean\"].dropna(), bins=80, alpha=0.5, label=\"all\")\n",
    "plt.hist(consistent[\"score_mean\"].dropna(), bins=80, alpha=0.5, label=\"consistent\")\n",
    "plt.xlabel(\"Mean log2 RNA/DNA score\")\n",
    "plt.ylabel(\"Number of elements\")\n",
    "plt.title(\"Score distribution before/after consistency filtering\")\n",
    "plt.legend()\n",
    "plt.show()\n",
    "\n",
    "plt.figure(figsize=(6, 4))\n",
    "plt.scatter(\n",
    "    element_readout[\"score_mean\"],\n",
    "    element_readout[\"score_sd_across_reps\"],\n",
    "    s=5,\n",
    "    alpha=0.4,\n",
    ")\n",
    "plt.axhline(0.75, linestyle=\"--\")\n",
    "plt.xlabel(\"Mean log2 RNA/DNA score\")\n",
    "plt.ylabel(\"SD across replicate pairs\")\n",
    "plt.title(\"Mean score vs replicate inconsistency\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "cd14ab20-716d-49d7-b3c2-9d7bcec552cd",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{\n",
      "  \"input_file\": \"training_seq_score_consistent_from_alignment.tsv\",\n",
      "  \"output_file\": \"training_seq_score_consistent_from_alignment_zscore_split.tsv\",\n",
      "  \"label_raw_column\": \"score_raw\",\n",
      "  \"label_z_column\": \"score_z\",\n",
      "  \"training_label_column\": \"score\",\n",
      "  \"zscore_mean_train_only\": -0.023867485882891677,\n",
      "  \"zscore_std_train_only\": 0.42136155452796287,\n",
      "  \"n_total\": 29355,\n",
      "  \"n_train\": 23484,\n",
      "  \"n_val\": 2935,\n",
      "  \"n_test\": 2936,\n",
      "  \"group_col\": \"element_id\",\n",
      "  \"random_state\": 42\n",
      "}\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>element_id</th>\n",
       "      <th>seq</th>\n",
       "      <th>score</th>\n",
       "      <th>score_median</th>\n",
       "      <th>score_sd_across_reps</th>\n",
       "      <th>n_qc_barcodes</th>\n",
       "      <th>rna_total</th>\n",
       "      <th>dna_total</th>\n",
       "      <th>median_barcode_ratio_sd</th>\n",
       "      <th>score_raw</th>\n",
       "      <th>split</th>\n",
       "      <th>score_z</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>TILE_ID_001-00001|ROTAVIRUS_A|NC_011505.2|1,130</td>\n",
       "      <td>ACTGGCCGCTTCACTGGGCTTTTAAAGCGCTACAGTGATGTCTCTC...</td>\n",
       "      <td>-0.229060</td>\n",
       "      <td>-0.119000</td>\n",
       "      <td>0.201078</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1551.0</td>\n",
       "      <td>1594.0</td>\n",
       "      <td>0.355931</td>\n",
       "      <td>-0.120385</td>\n",
       "      <td>train</td>\n",
       "      <td>-0.229060</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>TILE_ID_001-00002|ROTAVIRUS_A|NC_011505.2|66,195</td>\n",
       "      <td>ACTGGCCGCTTCACTGAACTATATATAAGAATGAATCGTCTTCAAC...</td>\n",
       "      <td>0.094683</td>\n",
       "      <td>-0.015694</td>\n",
       "      <td>0.071457</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1005.0</td>\n",
       "      <td>993.0</td>\n",
       "      <td>0.553352</td>\n",
       "      <td>0.016028</td>\n",
       "      <td>train</td>\n",
       "      <td>0.094683</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>TILE_ID_001-00003|ROTAVIRUS_A|NC_011505.2|131,260</td>\n",
       "      <td>ACTGGCCGCTTCACTGGTGAACAGTACATTTCACCAGATGCAGAAG...</td>\n",
       "      <td>-0.261223</td>\n",
       "      <td>-0.012470</td>\n",
       "      <td>0.691003</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1642.0</td>\n",
       "      <td>1279.0</td>\n",
       "      <td>0.725516</td>\n",
       "      <td>-0.133937</td>\n",
       "      <td>train</td>\n",
       "      <td>-0.261223</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>TILE_ID_001-00005|ROTAVIRUS_A|NC_011505.2|261,390</td>\n",
       "      <td>ACTGGCCGCTTCACTGAGTTAAGACAAATGCAGACGCTGGCGTGTC...</td>\n",
       "      <td>0.425975</td>\n",
       "      <td>-0.014419</td>\n",
       "      <td>0.437001</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1234.0</td>\n",
       "      <td>971.0</td>\n",
       "      <td>0.502387</td>\n",
       "      <td>0.155622</td>\n",
       "      <td>train</td>\n",
       "      <td>0.425975</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>TILE_ID_001-00006|ROTAVIRUS_A|NC_011505.2|326,455</td>\n",
       "      <td>ACTGGCCGCTTCACTGATGTCGGATGCGATCAAGTGGATTTCTCCT...</td>\n",
       "      <td>0.410825</td>\n",
       "      <td>0.134145</td>\n",
       "      <td>0.174887</td>\n",
       "      <td>3.0</td>\n",
       "      <td>566.0</td>\n",
       "      <td>445.0</td>\n",
       "      <td>0.504621</td>\n",
       "      <td>0.149238</td>\n",
       "      <td>train</td>\n",
       "      <td>0.410825</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                          element_id  \\\n",
       "0    TILE_ID_001-00001|ROTAVIRUS_A|NC_011505.2|1,130   \n",
       "1   TILE_ID_001-00002|ROTAVIRUS_A|NC_011505.2|66,195   \n",
       "2  TILE_ID_001-00003|ROTAVIRUS_A|NC_011505.2|131,260   \n",
       "3  TILE_ID_001-00005|ROTAVIRUS_A|NC_011505.2|261,390   \n",
       "4  TILE_ID_001-00006|ROTAVIRUS_A|NC_011505.2|326,455   \n",
       "\n",
       "                                                 seq     score  score_median  \\\n",
       "0  ACTGGCCGCTTCACTGGGCTTTTAAAGCGCTACAGTGATGTCTCTC... -0.229060     -0.119000   \n",
       "1  ACTGGCCGCTTCACTGAACTATATATAAGAATGAATCGTCTTCAAC...  0.094683     -0.015694   \n",
       "2  ACTGGCCGCTTCACTGGTGAACAGTACATTTCACCAGATGCAGAAG... -0.261223     -0.012470   \n",
       "3  ACTGGCCGCTTCACTGAGTTAAGACAAATGCAGACGCTGGCGTGTC...  0.425975     -0.014419   \n",
       "4  ACTGGCCGCTTCACTGATGTCGGATGCGATCAAGTGGATTTCTCCT...  0.410825      0.134145   \n",
       "\n",
       "   score_sd_across_reps  n_qc_barcodes  rna_total  dna_total  \\\n",
       "0              0.201078            3.0     1551.0     1594.0   \n",
       "1              0.071457            3.0     1005.0      993.0   \n",
       "2              0.691003            3.0     1642.0     1279.0   \n",
       "3              0.437001            3.0     1234.0      971.0   \n",
       "4              0.174887            3.0      566.0      445.0   \n",
       "\n",
       "   median_barcode_ratio_sd  score_raw  split   score_z  \n",
       "0                 0.355931  -0.120385  train -0.229060  \n",
       "1                 0.553352   0.016028  train  0.094683  \n",
       "2                 0.725516  -0.133937  train -0.261223  \n",
       "3                 0.502387   0.155622  train  0.425975  \n",
       "4                 0.504621   0.149238  train  0.410825  "
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import json\n",
    "from pathlib import Path\n",
    "from sklearn.model_selection import GroupShuffleSplit\n",
    "\n",
    "infile = Path(\"training_seq_score_consistent_from_alignment.tsv\")\n",
    "df = pd.read_csv(infile, sep=\"\\t\")\n",
    "\n",
    "assert \"seq\" in df.columns\n",
    "assert \"score\" in df.columns\n",
    "\n",
    "df = df.copy()\n",
    "df[\"score_raw\"] = df[\"score\"].astype(float)\n",
    "\n",
    "# Remove bad rows\n",
    "df = df.dropna(subset=[\"seq\", \"score_raw\"]).copy()\n",
    "df = df[df[\"seq\"].astype(str).str.len() > 0].copy()\n",
    "df = df.reset_index(drop=True)\n",
    "\n",
    "# Prefer grouped split to avoid related sequence leakage.\n",
    "if \"accession\" in df.columns and df[\"accession\"].notna().any():\n",
    "    group_col = \"accession\"\n",
    "elif \"element_id\" in df.columns:\n",
    "    group_col = \"element_id\"\n",
    "else:\n",
    "    group_col = None\n",
    "\n",
    "random_state = 42\n",
    "\n",
    "if group_col is not None:\n",
    "    fallback_groups = pd.Series(df.index.astype(str), index=df.index)\n",
    "    groups = df[group_col].fillna(fallback_groups).astype(str)\n",
    "\n",
    "    gss1 = GroupShuffleSplit(\n",
    "        n_splits=1,\n",
    "        test_size=0.20,\n",
    "        random_state=random_state,\n",
    "    )\n",
    "    train_idx, temp_idx = next(gss1.split(df, groups=groups))\n",
    "\n",
    "    train_df = df.iloc[train_idx].copy()\n",
    "    temp_df = df.iloc[temp_idx].copy().reset_index(drop=True)\n",
    "\n",
    "    fallback_temp_groups = pd.Series(temp_df.index.astype(str), index=temp_df.index)\n",
    "    temp_groups = temp_df[group_col].fillna(fallback_temp_groups).astype(str)\n",
    "\n",
    "    gss2 = GroupShuffleSplit(\n",
    "        n_splits=1,\n",
    "        test_size=0.50,\n",
    "        random_state=random_state,\n",
    "    )\n",
    "    val_rel_idx, test_rel_idx = next(gss2.split(temp_df, groups=temp_groups))\n",
    "\n",
    "    val_df = temp_df.iloc[val_rel_idx].copy()\n",
    "    test_df = temp_df.iloc[test_rel_idx].copy()\n",
    "\n",
    "else:\n",
    "    df_shuf = df.sample(frac=1, random_state=random_state).reset_index(drop=True)\n",
    "    n = len(df_shuf)\n",
    "    n_train = int(n * 0.80)\n",
    "    n_val = int(n * 0.10)\n",
    "\n",
    "    train_df = df_shuf.iloc[:n_train].copy()\n",
    "    val_df = df_shuf.iloc[n_train:n_train+n_val].copy()\n",
    "    test_df = df_shuf.iloc[n_train+n_val:].copy()\n",
    "\n",
    "# Compute z-score stats from TRAIN ONLY\n",
    "label_mean = train_df[\"score_raw\"].mean()\n",
    "label_std = train_df[\"score_raw\"].std(ddof=0)\n",
    "\n",
    "assert label_std > 0, \"label std is zero; cannot z-normalize.\"\n",
    "\n",
    "for split_name, split_df in [\n",
    "    (\"train\", train_df),\n",
    "    (\"val\", val_df),\n",
    "    (\"test\", test_df),\n",
    "]:\n",
    "    split_df[\"split\"] = split_name\n",
    "    split_df[\"score_z\"] = (split_df[\"score_raw\"] - label_mean) / label_std\n",
    "\n",
    "out_df = pd.concat([train_df, val_df, test_df], axis=0).reset_index(drop=True)\n",
    "\n",
    "# For training, use normalized score.\n",
    "# Raw score is preserved in score_raw.\n",
    "out_df[\"score\"] = out_df[\"score_z\"]\n",
    "\n",
    "out_tsv = Path(\"training_seq_score_consistent_from_alignment_zscore_split.tsv\")\n",
    "out_df.to_csv(out_tsv, sep=\"\\t\", index=False)\n",
    "\n",
    "stats = {\n",
    "    \"input_file\": str(infile),\n",
    "    \"output_file\": str(out_tsv),\n",
    "    \"label_raw_column\": \"score_raw\",\n",
    "    \"label_z_column\": \"score_z\",\n",
    "    \"training_label_column\": \"score\",\n",
    "    \"zscore_mean_train_only\": float(label_mean),\n",
    "    \"zscore_std_train_only\": float(label_std),\n",
    "    \"n_total\": int(len(out_df)),\n",
    "    \"n_train\": int((out_df[\"split\"] == \"train\").sum()),\n",
    "    \"n_val\": int((out_df[\"split\"] == \"val\").sum()),\n",
    "    \"n_test\": int((out_df[\"split\"] == \"test\").sum()),\n",
    "    \"group_col\": group_col,\n",
    "    \"random_state\": random_state,\n",
    "}\n",
    "\n",
    "with open(\"label_stats_consistent_from_alignment.json\", \"w\") as f:\n",
    "    json.dump(stats, f, indent=2)\n",
    "\n",
    "print(json.dumps(stats, indent=2))\n",
    "out_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "be5194c8",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Subset: train\n",
      "Elements in model TSV subset: 23,484\n",
      "Elements with replicate readouts: 23,484\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>element_log2_ratio_rep1</th>\n",
       "      <th>element_log2_ratio_rep2</th>\n",
       "      <th>element_log2_ratio_rep3</th>\n",
       "      <th>element_log2_ratio_rep4</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>element_log2_ratio_rep1</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.620345</td>\n",
       "      <td>0.577623</td>\n",
       "      <td>0.562059</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>element_log2_ratio_rep2</th>\n",
       "      <td>0.620345</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.588821</td>\n",
       "      <td>0.568883</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>element_log2_ratio_rep3</th>\n",
       "      <td>0.577623</td>\n",
       "      <td>0.588821</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.579602</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>element_log2_ratio_rep4</th>\n",
       "      <td>0.562059</td>\n",
       "      <td>0.568883</td>\n",
       "      <td>0.579602</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                         element_log2_ratio_rep1  element_log2_ratio_rep2  \\\n",
       "element_log2_ratio_rep1                 1.000000                 0.620345   \n",
       "element_log2_ratio_rep2                 0.620345                 1.000000   \n",
       "element_log2_ratio_rep3                 0.577623                 0.588821   \n",
       "element_log2_ratio_rep4                 0.562059                 0.568883   \n",
       "\n",
       "                         element_log2_ratio_rep3  element_log2_ratio_rep4  \n",
       "element_log2_ratio_rep1                 0.577623                 0.562059  \n",
       "element_log2_ratio_rep2                 0.588821                 0.568883  \n",
       "element_log2_ratio_rep3                 1.000000                 0.579602  \n",
       "element_log2_ratio_rep4                 0.579602                 1.000000  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>element_log2_ratio_rep1</th>\n",
       "      <th>element_log2_ratio_rep2</th>\n",
       "      <th>element_log2_ratio_rep3</th>\n",
       "      <th>element_log2_ratio_rep4</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>element_log2_ratio_rep1</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.323971</td>\n",
       "      <td>0.290921</td>\n",
       "      <td>0.277453</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>element_log2_ratio_rep2</th>\n",
       "      <td>0.323971</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.301407</td>\n",
       "      <td>0.288984</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>element_log2_ratio_rep3</th>\n",
       "      <td>0.290921</td>\n",
       "      <td>0.301407</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.309167</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>element_log2_ratio_rep4</th>\n",
       "      <td>0.277453</td>\n",
       "      <td>0.288984</td>\n",
       "      <td>0.309167</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                         element_log2_ratio_rep1  element_log2_ratio_rep2  \\\n",
       "element_log2_ratio_rep1                 1.000000                 0.323971   \n",
       "element_log2_ratio_rep2                 0.323971                 1.000000   \n",
       "element_log2_ratio_rep3                 0.290921                 0.301407   \n",
       "element_log2_ratio_rep4                 0.277453                 0.288984   \n",
       "\n",
       "                         element_log2_ratio_rep3  element_log2_ratio_rep4  \n",
       "element_log2_ratio_rep1                 0.290921                 0.277453  \n",
       "element_log2_ratio_rep2                 0.301407                 0.288984  \n",
       "element_log2_ratio_rep3                 1.000000                 0.309167  \n",
       "element_log2_ratio_rep4                 0.309167                 1.000000  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAg8AAAGGCAYAAAD8cmIvAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjkuMCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy80BEi2AAAACXBIWXMAAA9hAAAPYQGoP6dpAAB/3UlEQVR4nO3dd1gUxxsH8O/d0cvR4UCRonQREAW7qERAY4lGY1di1BiNMZpo9KeCPfYWY6+JLYmaqDFYMXZREMUORBSVjoC0A+729weyunAgdxwg3vt5nn305mb3Zoa9vXdnZnd5DMMwIIQQQgipJn59F4AQQgghDQsFD4QQQgiRCwUPhBBCCJELBQ+EEEIIkQsFD4QQQgiRCwUPhBBCCJELBQ+EEEIIkQsFD4QQQgiRCwUPhBBCCJELBQ8qaOfOneDxeEhISFDaNhMSEsDj8bBz506lbZO8X0aNGgVbW1tOGo/HQ2hoaL2U531na2uLjz/++J35zp07Bx6Ph3PnzrFpqt7WDbmuv/32G4yNjZGbm1snn1eT43lYWBj09PSQlpYm97pyBQ9lhSxbtLS04OjoiIkTJyIlJUXuDycNz969e7F69er6LgZRAS9evEBoaCiio6Pr/LMvX76M0NBQZGVl1flnV1dDKOP77Oeff1b6yY5EIkFISAi+/vpr6OnpsemLFi3Cn3/+qdTPUobAwEA0a9YMixcvlntdhXoe5s2bh19++QU//fQT2rVrhw0bNqBt27bIz89XZHOkAakseLCxsUFBQQGGDx9e94Ui9aagoACzZs2qlW2/ePECc+fOrbfgYe7cuXXyw9ypUycUFBSgU6dOVeYr39Z1WcYPUW0ED0ePHsXDhw8xduxYTnptBg/Dhw9HQUEBbGxsFFp/3Lhx2LRpE169eiXXegoFD0FBQRg2bBi++OIL7Ny5E5MnT8bjx4/x119/KbI5heTl5dXZZ70PGIZBQUGBzPcKCwshlUrruERcZT1RAoGgXsuh6qRSKQoLC+vs87S0tKCmplZnn/ch4vP50NLSAp9f9eG4vtq6qmMP4dqxYwfat2+PRo0aKbwNeX/bBAIBtLS0wOPxFPq8/v37QywW4/fff5drPaXMeejatSsA4PHjx2zar7/+Cm9vb2hra8PY2BiDBg1CYmIiZ70LFy5gwIABaNKkCTQ1NWFtbY1vv/22wo46atQo6OnpIT4+Hj169IC+vj6GDh0KAIiNjUX//v0hEomgpaWFxo0bY9CgQcjOzmbXLykpwfz589G0aVNoamrC1tYWM2fOhFgs5nxO2RjlxYsX4ePjAy0tLdjb22P37t3VagepVIo1a9bA3d0dWlpaMDMzQ2BgIG7cuKFwWU6cOIFWrVpBW1sbmzZtYsdH9+/fj1mzZqFRo0bQ0dFBTk4OAODatWsIDAyEgYEBdHR00LlzZ1y6dOmdZf/rr7/Qs2dPWFlZQVNTE02bNsX8+fMhkUjYPH5+fvj777/x5MkTduiqbFy2sjkPZ8+eRceOHaGrqwtDQ0P06dMH9+/f5+QJDQ0Fj8dDXFwcRo0aBUNDQxgYGCA4OLhavVny7kdPnz7Fxx9/DD09PTRq1Ajr168HAMTExKBr167Q1dWFjY0N9u7dW+Gz/vvvPwwYMADGxsbQ0dFBmzZt8Pfff3PylP2NfvvtNyxcuBCNGzeGlpYWunXrhri4uArbXL9+Pezt7aGtrQ0fHx9cuHABfn5+8PPze2fdeTweJk6ciD179sDNzQ2ampoICwsDADx//hyff/45LCwsoKmpCTc3N2zfvl1mWQ8cOICZM2dCJBJBV1cXvXv3rvB9rezzy49NP3/+HKNHj2b3JTs7O4wfPx5FRUUAgMzMTHz33Xdwd3eHnp4ehEIhgoKCcOvWLU65WrduDQAIDg5m97e39y9F93UAWLduHdzc3KCjowMjIyO0atWK/XuHhobi+++/BwDY2dmxn102prxjxw507doV5ubm0NTUhKurKzZs2FDpZ508eRKenp7Q0tKCq6srDh06xHlf1pwHWd5u66rK2LlzZ3h4eMjchpOTEwICAqr8nMqOPQCQlZWFyZMnw9raGpqammjWrBmWLFlS4eRl+fLlaNeuHUxMTKCtrQ1vb2/88ccfFT5LLBbj22+/hZmZGfT19dG7d288e/ZMZrlu3ryJoKAgCIVC6OnpoVu3brh69SonT9mxpLzy8wJsbW1x9+5d/Pvvv2zblX3fiouLMXfuXDg4OEBLSwsmJibo0KEDTp06VWW7FRYWIiwsDP7+/px0Ho+HvLw87Nq1i/2sUaNGccp77949DBkyBEZGRujQoQMA4Pbt2xg1ahTs7e2hpaUFkUiEzz//HBkZGVXWrax+1f0tMzc3R4sWLeQ++VdKGBsfHw8AMDExAQAsXLgQs2fPxsCBA/HFF18gLS0N69atQ6dOnXDz5k0YGhoCAH7//Xfk5+dj/PjxMDExQUREBNatW4dnz55ViIJKSkoQEBCADh06YPny5dDR0UFRURECAgIgFovx9ddfQyQS4fnz5zh27BiysrJgYGAAAPjiiy+wa9cufPrpp5g6dSquXbuGxYsX4/79+zh8+DDnc+Li4vDpp59i9OjRGDlyJLZv345Ro0bB29sbbm5uVbbD6NGjsXPnTgQFBeGLL75ASUkJLly4gKtXr6JVq1Zyl+Xhw4cYPHgwxo0bhzFjxsDJyYl9b/78+dDQ0MB3330HsVgMDQ0NnD17FkFBQfD29kZISAj4fD57oLtw4QJ8fHwqLfvOnTuhp6eHKVOmQE9PD2fPnsWcOXOQk5ODZcuWAQD+97//ITs7G8+ePcOqVasAgDOuV97p06cRFBQEe3t7hIaGoqCgAOvWrUP79u0RFRVVYULYwIEDYWdnh8WLFyMqKgpbt26Fubk5lixZUmW7y7MfSSQSBAUFoVOnTli6dCn27NmDiRMnQldXF//73/8wdOhQ9OvXDxs3bsSIESPQtm1b2NnZAQBSUlLQrl075OfnY9KkSTAxMcGuXbvQu3dv/PHHH/jkk084n/Xjjz+Cz+fju+++Q3Z2NpYuXYqhQ4fi2rVrbJ4NGzZg4sSJ6NixI7799lskJCSgb9++MDIyQuPGjausd5mzZ8/it99+w8SJE2FqagpbW1ukpKSgTZs2bHBhZmaGf/75B6NHj0ZOTg4mT57M2cbChQvB4/Ewffp0pKamYvXq1fD390d0dDS0tbWrVQ6gdKjBx8cHWVlZGDt2LJydnfH8+XP88ccfyM/Ph4aGBv777z/8+eefGDBgAOzs7JCSkoJNmzahc+fOuHfvHqysrODi4oJ58+Zhzpw5GDt2LDp27AgAaNeuHVtnRff1LVu2YNKkSfj000/xzTffoLCwELdv38a1a9cwZMgQ9OvXD48ePcK+ffuwatUqmJqaAgDMzMzYv5mbmxt69+4NNTU1HD16FF999RWkUikmTJjA+azY2Fh89tln+PLLLzFy5Ejs2LEDAwYMQFhYGD766KNqt2t5VZVx+PDhGDNmDO7cuYPmzZuz61y/fh2PHj2q1jCTrGNPfn4+OnfujOfPn2PcuHFo0qQJLl++jBkzZiApKYkznLlmzRr07t0bQ4cORVFREfbv348BAwbg2LFj6NmzJ5vviy++wK+//oohQ4agXbt2OHv2LOf9Mnfv3kXHjh0hFAoxbdo0qKurY9OmTfDz88O///4LX19fudpv9erV7LyE//3vfwAACwsLAKU/6IsXL8YXX3wBHx8f5OTk4MaNG4iKiqrybxYZGYmioiK0bNmSk/7LL7+w2yobzmjatCknz4ABA+Dg4IBFixaBYRgAwKlTp/Dff/8hODgYIpEId+/exebNm3H37l1cvXr1nT0N8vyWeXt7yz+swshhx44dDADm9OnTTFpaGpOYmMjs37+fMTExYbS1tZlnz54xCQkJjEAgYBYuXMhZNyYmhlFTU+Ok5+fnV/iMxYsXMzwej3ny5AmbNnLkSAYA88MPP3Dy3rx5kwHA/P7775WWOTo6mgHAfPHFF5z07777jgHAnD17lk2zsbFhADDnz59n01JTUxlNTU1m6tSpVbbN2bNnGQDMpEmTKrwnlUoVLktYWBgnb3h4OAOAsbe357SfVCplHBwcmICAAPbzGKa0je3s7JiPPvqITSv7Oz5+/JiTr7xx48YxOjo6TGFhIZvWs2dPxsbGpkLex48fMwCYHTt2sGmenp6Mubk5k5GRwabdunWL4fP5zIgRI9i0kJAQBgDz+eefc7b5ySefMCYmJhU+qzx596NFixaxaS9fvmS0tbUZHo/H7N+/n01/8OABA4AJCQlh0yZPnswAYC5cuMCmvXr1irGzs2NsbW0ZiUTCMMybv5GLiwsjFovZvGvWrGEAMDExMQzDMIxYLGZMTEyY1q1bM8XFxWy+nTt3MgCYzp07v7PuABg+n8/cvXuXkz569GjG0tKSSU9P56QPGjSIMTAwYNusrKyNGjVicnJy2Hy//fYbA4BZs2YNp/3K/+3Lt9GIESMYPp/PXL9+vUJZy/bLwsJCtq3KPH78mNHU1GTmzZvHpl2/fr3CPlW2neru67L06dOHcXNzqzLPsmXLKnxH3v6c8gICAhh7e3tOWtl3+ODBg2xadnY2Y2lpyXh5ebFpZX+D8PBwNq06bV1ZGbOyshgtLS1m+vTpnPRJkyYxurq6TG5ubiW15pa7/LFn/vz5jK6uLvPo0SNO+g8//MAIBALm6dOnbFr5NioqKmKaN2/OdO3alU0rOx5+9dVXnLxDhgypUNe+ffsyGhoaTHx8PJv24sULRl9fn+nUqRObVnYsKU/WMc/NzU3md8zDw4Pp2bNnhfR32bp1K+f7/TZdXV1m5MiRFdLLyjt48OAK78naz/bt21fhN0pW3eT9LVu0aBEDgElJSXlXNVkKDVv4+/vDzMwM1tbWGDRoEPT09HD48GE0atQIhw4dglQqxcCBA5Gens4uIpEIDg4OCA8PZ7fz9hlNXl4e0tPT0a5dOzAMg5s3b1b43PHjx3Nel/UsnDhxotLu7ePHjwMApkyZwkmfOnUqAFTocnZ1dWXPcoDSSN7JyQn//fdflW1y8OBB8Hg8hISEVHivLEKUtyx2dnaVdjGOHDmS037R0dGIjY3FkCFDkJGRwbZ7Xl4eunXrhvPnz1c5L+Ltbb169Qrp6eno2LEj8vPz8eDBg6qqLlNSUhKio6MxatQoGBsbs+ktWrTARx99xLbF27788kvO644dOyIjI4MdkqlO2auzH33xxRfs/w0NDeHk5ARdXV0MHDiQTXdycoKhoSHn7378+HH4+Piw3YpAac/L2LFjkZCQgHv37nE+Jzg4GBoaGpz6AGC3eePGDWRkZGDMmDGcseyhQ4fCyMioyjq/rXPnznB1dWVfMwyDgwcPolevXmAYhvM9DAgIQHZ2NqKiojjbGDFiBPT19dnXn376KSwtLWX+nSojlUrx559/olevXmxP29vKvgeamprs+L5EIkFGRgb09PTg5ORUoVyy1HRfNzQ0xLNnz3D9+vVq1+1tb+9v2dnZSE9PR+fOnfHff/9xhksBwMrKitMjJRQKMWLECNy8eRPJyckKff67GBgYoE+fPti3bx97FiuRSHDgwAH07dsXurq679yGrGPP77//jo4dO8LIyIizT/n7+0MikeD8+fNs3rfb6OXLl8jOzkbHjh05f9+yfWvSpEmczynfKyaRSHDy5En07dsX9vb2bLqlpSWGDBmCixcvvvMYIQ9DQ0PcvXsXsbGxcq1XNpwgz3e3TPljH8Btw8LCQqSnp6NNmzYAUK3viTy/ZWVlTk9Pr3aZFRq2WL9+PRwdHaGmpgYLCws4OTmxB4PY2FgwDAMHBweZ66qrq7P/f/r0KebMmYMjR47g5cuXnHzlv4RqamoVunHt7OwwZcoUrFy5Env27EHHjh3Ru3dvDBs2jA0snjx5Aj6fj2bNmnHWFYlEMDQ0xJMnTzjpTZo0qVBmIyOjCuUrLz4+HlZWVpwfyvLkLUtZd7ks5d8r29FHjhxZ6TrZ2dmV7th3797FrFmzcPbs2QpfxPJ/i+ooq8vbQy1lXFxccOLECeTl5XEOZOXbvqysL1++hFAorPSz5NmPyuaivM3AwACNGzeu0A1oYGDA2d6TJ09kdo+6uLiw77/dTVxVfcryA6iwP6ipqVUY0qlK+X0hLS0NWVlZ2Lx5MzZv3ixzndTUVM7r8t9XHo+HZs2ayXXteFpaGnJycjhtIEvZ3KCff/4Zjx8/5syrKRv6rEpN9/Xp06fj9OnT8PHxQbNmzdC9e3cMGTIE7du3f+dnA8ClS5cQEhKCK1euVDhpyc7OZo89QOnftvx+5ejoCKB0npBIJKrWZ8prxIgROHDgAC5cuIBOnTrh9OnTSElJqfbVULKOPbGxsbh9+3aF70+Zt/epY8eOYcGCBYiOjubM53q7LcqOh+W78MsfM9LS0pCfn1/psUQqlSIxMfGdw8rVNW/ePPTp0weOjo5o3rw5AgMDMXz4cLRo0aJa65cFbPKQ1d6ZmZmYO3cu9u/fX+H7Wp1jsjy/ZWVllmfSpULBg4+Pj8wzC6D0wMDj8fDPP//InHlfNkYukUjw0UcfITMzE9OnT4ezszN0dXXx/PlzjBo1qsKZw9tnK29bsWIFRo0ahb/++gsnT57EpEmTsHjxYly9epUTbFS3USq7WkCRHaIy1S1LVWPN5d8ra69ly5bB09NT5jqVzU/IyspC586dIRQKMW/ePDRt2hRaWlqIiorC9OnT6+xKDkXaXt79qLLPqI2/e13sS0Dl+8KwYcMq/YGt7oGwNixatAizZ8/G559/jvnz58PY2Bh8Ph+TJ0+u1r5Wk30dKP3BefjwIY4dO4awsDAcPHgQP//8M+bMmYO5c+dW+dnx8fHo1q0bnJ2dsXLlSlhbW0NDQwPHjx/HqlWr6v2qpzIBAQGwsLDAr7/+ik6dOuHXX3+FSCSqMJmvMrKOPVKpFB999BGmTZsmc52yoOjChQvo3bs3OnXqhJ9//hmWlpZQV1fHjh07ZE5CVqbKjq1vB6jv0qlTJ8THx7O/KVu3bsWqVauwceNGTq9leWWB78uXL6s9X6mMrPYeOHAgLl++jO+//x6enp7Q09ODVCpFYGBgtfYzeY4/ZQFF2dyZ6lD6dT9NmzYFwzCws7NjdyZZYmJi8OjRI+zatQsjRoxg0981o1UWd3d3uLu7Y9asWbh8+TLat2+PjRs3YsGCBbCxsYFUKkVsbCx7hgiUTn7LyspS+NrY8po2bYoTJ04gMzOz0t6H2ixLWfQuFAqrfYAoc+7cOWRkZODQoUOca83fvnqmTHUDn7K6PHz4sMJ7Dx48gKmpabW6T99FmfvRu9jY2FRan7L35d0eUDqxqUuXLmx6SUkJEhISFP6BL5u5LpFIqr0vlO+iZRgGcXFxcpXBzMwMQqEQd+7cqTLfH3/8gS5dumDbtm2c9KysLM7Bq7J9rSb7ehldXV189tln+Oyzz1BUVIR+/fph4cKFmDFjRpWXvR09ehRisRhHjhzhnNm9PRz7tri4ODAMw9neo0ePAECu3iVZqvouCgQCDBkyBDt37sSSJUvw559/YsyYMTW6lLpp06bIzc19Z5sfPHgQWlpaOHHiBDQ1Ndn0HTt2cPKVHQ/j4+M5vQrlv2NmZmbQ0dGp9LvH5/NhbW0N4E3vXlZWFjsxH0CFXl2g6vYzNjZGcHAwgoODkZubi06dOiE0NLTK4MHZ2RlA6XHT3d292p8ly8uXL3HmzBnMnTsXc+bMYdPlHUqprsePH8PU1LTSXiVZlH576n79+kEgEGDu3LkVIhyGYdhxobKd+O08DMNgzZo11f6snJwclJSUcNLc3d3B5/PZrrIePXoAQIUbG61cuRIAZM7sVUT//v3BMIzMM5eyOtZmWby9vdG0aVMsX75c5m1Rq7r9qKy/RVFREX7++ecKeXV1davVZWZpaQlPT0/s2rWLcxObO3fu4OTJk2xb1JQy9qPq6tGjByIiInDlyhU2LS8vD5s3b4atrS1n3kF1tGrVCiYmJtiyZQtnP96zZ887h8mqIhAI0L9/fxw8eFDmD7msfWH37t2cm8T88ccfSEpKQlBQULU/l8/no2/fvjh69Cjn8uQyZX8jgUBQ4djw+++/4/nz55y0suCy/E2QarKvA6hwqZuGhgZcXV3BMAyKi4ur/GxZ+1t2dnaFH8YyL1684FxFlZOTg927d8PT07PGQxaVlbHM8OHD8fLlS4wbNw65ubkYNmxYjT5v4MCBuHLlCk6cOFHhvaysLHYfFggE4PF4nLP9hISECrP5y/attWvXctLLHx8FAgG6d++Ov/76izOMlpKSgr1796JDhw7ssGZZYPn2/IuyyyTL09XVldl25fcPPT09NGvWrMLl9OV5e3tDQ0ND5r5f2WdVRtZ+BlRsG2WJjIxE27Zt5VqnVnoeFixYgBkzZrCXnenr6+Px48c4fPgwxo4di++++w7Ozs5o2rQpvvvuOzx//hxCoRAHDx6U66B59uxZTJw4EQMGDICjoyNKSkrwyy+/sAdPAPDw8MDIkSOxefNmtns+IiICu3btQt++fTlnfDXRpUsXDB8+HGvXrkVsbCzbtXThwgV06dIFEydOrNWy8Pl8bN26FUFBQXBzc0NwcDAaNWqE58+fIzw8HEKhEEePHpW5brt27WBkZISRI0di0qRJ4PF4+OWXX2R2b3l7e+PAgQOYMmUKWrduDT09PfTq1UvmdpctW4agoCC0bdsWo0ePZi/VNDAwUNp965WxH1XXDz/8gH379iEoKAiTJk2CsbExdu3ahcePH+PgwYPvvMlPeRoaGggNDcXXX3+Nrl27YuDAgUhISMDOnTvRtGlThW/6ApReJhoeHg5fX1+MGTMGrq6uyMzMRFRUFE6fPo3MzExOfmNjY3To0AHBwcFISUnB6tWr0axZM4wZM0auz120aBFOnjyJzp07Y+zYsXBxcUFSUhJ+//13XLx4EYaGhvj4448xb948BAcHo127doiJicGePXs4k+GA0mOJoaEhNm7cCH19fejq6sLX1xd2dnYK7+sA0L17d4hEIrRv3x4WFha4f/8+fvrpJ/Ts2ZOdNOrt7Q2g9PLkQYMGQV1dHb169UL37t2hoaGBXr16sT/KW7Zsgbm5OZKSkip8lqOjI0aPHo3r16/DwsIC27dvR0pKSqXBhjwqK2NZUOHl5YXmzZvj999/h4uLS4VLCOX1/fff48iRI/j444/ZS/7y8vIQExODP/74AwkJCTA1NUXPnj2xcuVKBAYGYsiQIUhNTcX69evRrFkz3L59m92ep6cnBg8ejJ9//hnZ2dlo164dzpw5I/NeKAsWLMCpU6fQoUMHfPXVV1BTU8OmTZsgFouxdOlSNl/37t3RpEkTjB49Gt9//z0EAgG2b98OMzMzPH36tEL7bdiwAQsWLECzZs1gbm6Orl27wtXVFX5+fvD29oaxsTFu3LiBP/74AxMnTqyyfbS0tNC9e3ecPn0a8+bNq/BZp0+fxsqVK2FlZQU7O7sqLy8VCoXs5eTFxcVo1KgRTp48KbM3uKZSU1Nx+/btCpcZv1O1r8tg3lwSIusyrPIOHjzIdOjQgdHV1WV0dXUZZ2dnZsKECczDhw/ZPPfu3WP8/f0ZPT09xtTUlBkzZgxz69atCpdnjRw5ktHV1a3wGf/99x/z+eefM02bNmW0tLQYY2NjpkuXLszp06c5+YqLi5m5c+cydnZ2jLq6OmNtbc3MmDGDcwkiw5Re3iLrEp3OnTtX67K5kpISZtmyZYyzszOjoaHBmJmZMUFBQUxkZKTSylJ2WVdll6fevHmT6devH2NiYsJoamoyNjY2zMCBA5kzZ86weWRd2nPp0iWmTZs2jLa2NmNlZcVMmzaNOXHiRIVLyHJzc5khQ4YwhoaGDAD2cjJZl2oyDMOcPn2aad++PaOtrc0IhUKmV69ezL179zh5yi5XSktL46TLKqcsNd2POnfuLPPSPVl/g/j4eObTTz9lDA0NGS0tLcbHx4c5duwYJ09lf6PK2mjt2rWMjY0No6mpyfj4+DCXLl1ivL29mcDAwCrrzTCll+9NmDBB5nspKSnMhAkTGGtra0ZdXZ0RiURMt27dmM2bN1co6759+5gZM2Yw5ubmjLa2NtOzZ0/OZa4MU73LBxmGYZ48ecKMGDGCMTMzYzQ1NRl7e3tmwoQJ7GWrhYWFzNSpUxlLS0tGW1ubad++PXPlyhWZ37O//vqLcXV1ZdTU1Cq0XXX2dVk2bdrEdOrUiV2vadOmzPfff89kZ2dz8s2fP59p1KgRw+fzOfvhkSNHmBYtWjBaWlqMra0ts2TJEmb79u0yL5fr2bMnc+LECaZFixaMpqYm4+zsXGG/UPRSzarKWGbp0qUVLk9+l8qOPQxTemnyjBkzmGbNmjEaGhqMqakp065dO2b58uVMUVERm2/btm2Mg4MDW+cdO3bIvIyyoKCAmTRpEmNiYsLo6uoyvXr1YhITE2XWNSoqigkICGD09PQYHR0dpkuXLszly5crlDEyMpLx9fVlNDQ0mCZNmjArV66UeSxJTk5mevbsyejr63MujV6wYAHj4+PDGBoaMtra2oyzszOzcOFCTv0qc+jQIYbH43EuW2WY0ku/O3XqxGhrazMA2Ms2Kzv2MQzDPHv2jPnkk08YQ0NDxsDAgBkwYADz4sWLCm1T2aWa1f0t27BhA6Ojo8O5VLs6eAyj5NlbhJAakUqlMDMzQ79+/bBly5Za/axz586hS5cu+P333/Hpp5/W6meRurdmzRr25mOyZt8T5ZJIJHB1dcXAgQMxf/78+i5OtXh5ecHPz4+98V910SO5CalHhYWFFYaHdu/ejczMzGrdnpqQyjAMg23btqFz584UONQRgUCAefPmYf369XX2SO6aCAsLQ2xsLGbMmCH3uvREG0Lq0dWrV/Htt99iwIABMDExQVRUFLZt24bmzZtjwIAB9V080gDl5eXhyJEjCA8PR0xMTJ0+sJCAvYqnIQgMDFQ4yKHggZB6ZGtrC2tra6xdu5a9zHfEiBH48ccfOXenJKS60tLSMGTIEBgaGmLmzJno3bt3fReJfIBozgMhhBCiROfPn8eyZcsQGRmJpKQkHD58GH379q1ynXPnzmHKlCm4e/curK2tMWvWLPbpm2XWr1+PZcuWITk5GR4eHli3bl2VD4GrTTTngRBCCFGivLw8eHh4YP369dXK//jxY/Ts2RNdunRBdHQ0Jk+ejC+++IJzT42yS+RDQkIQFRUFDw8PBAQEVLh1dV2hngdCCCGklvB4vHf2PEyfPh1///0356ZugwYNQlZWFsLCwgAAvr6+aN26NX766ScApVdlWVtb4+uvv8YPP/xQq3WQheY8vIekUilevHgBfX39Gt0oiBBCGjqGYfDq1StYWVnJfSO2txUWFqKoqKhG5Sh/PNbU1OTcgltRV65cqXDb74CAAPYJo0VFRYiMjORcFcHn8+Hv78+5421douDhPfTixQv2Xu2EEEKAxMREuR84VaawsBB2NnpITq3+A7LK09PTq3BlQkhIiFLulpucnAwLCwtOmoWFBXJyclBQUICXL19CIpHIzFP2bJ26RsHDe6jsFrlPomwh1KNpKYpamiH7sfCk+nJKKn+yK6meEkbxh1ERoCivGDuCjrDHRYW2UVSE5FQJnkTaQqgv/zE155UUNt4JSExMZJ+jAUApvQ4NFQUP76GyrjGhHl+hHZ2U0hKr13cRGjxxCbVhTfEpeFAKZQzh6unzoKcv/3akeH1MFgo5wYOyiEQipKSkcNJSUlIgFAqhra0NgUAAgUAgM09NH7CmKPplIoQQohIkjFThpTa1bdsWZ86c4aSdOnWKfdKlhoYGvL29OXmkUinOnDkj99MwlYV6HgghhKgEKRhIIf8FhvKuk5uby3k66OPHjxEdHQ1jY2M0adIEM2bMwPPnz7F7924AwJdffomffvoJ06ZNw+eff46zZ8/it99+w99//81uY8qUKRg5ciRatWoFHx8frF69Gnl5eQgODpa7PspAwQMhhBCVIIUUivQhyLvWjRs30KVLF/b1lClTAAAjR47Ezp07kZSUxHlEuJ2dHf7++298++23WLNmDRo3boytW7ciICCAzfPZZ58hLS0Nc+bMQXJyMjw9PREWFlZhEmVdofs8vIdycnJgYGCAl4/sac5DDSxId67vIjR42TRhssaKac5DjRTlFmNTp4PIzs5WeL5B2TH1xcPGCk+YtHJ6VqMyfGio54EQQohKkDAMJAqcLyuyzoeOggdCCCEqoa7mPKgCCh4IIYSoBCkYSCh4UAoKHgghhKgE6nlQHpqNRwghhBC5UM8DIYQQlUATJpWHggdCCCEqQfp6UWQ9wkXBAyGEEJUgUXDCpCLrfOhozgMhhBBC5EI9D4QQQlSChCldFFmPcFHwQAghRCXQnAfloeCBEEKISpCCBwl4Cq1HuCh4IIQQohKkTOmiyHqEiyZMEkIIIUQu1PNACCFEJUgUHLZQZJ0PHQUPhBBCVAIFD8pDwQMhhBCVIGV4kDIKTJhUYJ0PHQUPhBBCVAL1PCgPTZgkhBBCiFyo54EQQohKkIAPiQLnzJJaKEtDR8EDIYQQlcAoOOeBoTkPFVDwQAghRCXQnAfloeCBEEKISpAwfEgYBYYt6A6TFdCESUIIIYTIhXoeCCGEqAQpeJAqcM4sBXU9lEfBAyGEEJVAcx6Uh4IHQgghKkHxOQ/U81AeBQ+EEEJUQumwhQK3p6aehwooeCA4f6UAyze8RNRtMZJSJDi4XYS+QXpVrnPucj6+C8nA3UdiWFupY+ZkI4z6TMjJ8/OOLCz/OQvJaRJ4uGpgzUIz+Hhp1WZV6tW1fQm4vDMeueliWDgJ0WOGGxq7G1WavyCnGGfWPsD9M8koyC6GoZU2Aqe5wrGTBQDg/NY43D+dhPTHuVDXEsDawwgffesCU7uq/zYNWfSBONzY9RB5GYUwczREl+lesGxuLDPv3SMJOBFynZMm0ODjm2v92ddF+SW4sPY24sNfoCBbDAMrXXgNdoDHgKa1Wo/6dPtALKJ230d+RiFMHQ3RaZo3RM1NZOa9f+Q/nA6N4KQJNPj46upATlrmf9m4vPYWnkelQVoihbG9AXosaw99S91aqwd5v9HVFkp26NAhdO/eHSYmJuDxeIiOjq7vIr1TXr4UHq6aWLfIrFr5Hz8tRq9hSfBrr42oU03wzRgDjJ2aihPheWyeA3+9wtTQdMyeaowbJ6zRwlUTQYNfIDW9pLaqUa/uhL3AiWX34PelI8b91hEiRyF+GReB3AyxzPwlxVLsHnsVWS8K8NlKb3x91A+9Q1pAaKHN5nlyIwM+g2wxZk8HjNjcBpISBrvHXUNR/ofZhg9PJOLfFbfQZpwrhu39CGaOBjj01XnkZxZWuo6GnhrGnerFLl8c78l5/98V0Ui4nIyghT4YdSgQLYc64uySm4g/96K2q1MvHp14igsrb8JnbHMM2hsAUwdDHJlw7h1tqI7PT/Zhl1F/9+a8n534CgdHn4GRrRD9NnfFkAOBaD3GDQJNQW1XR+mkr+8wKe+iyCTL9evXw9bWFlpaWvD19UVERESlef38/MDj8SosPXu+2Z9HjRpV4f3AwECF2kEZKHh4raioSCnbycvLQ4cOHbBkyRKlbK8uBHXTxfwfTPBJj+qd0W7anQ27JupYHmoKF0cNTPjcEP0/1sPqzdlsntWbsvDFUAMEDxLC1UkDG5aaQUebhx37XtVWNerV5d3/wbu/Nbw+sYZ5U318PMcd6tp83DycKDP/zcOJKMguxuA1rdDEyxhGjXRg29oEIqc3vTfDN/rCq681zJvpQ+QkxCcLPJCdVIAX97JlbrOhi/z1EZr3s0PzPnYwaSqE//+8oaYlwJ0/EypdhwcedE213iwm3J6tF7cy4PaxLaxbmcPAShct+tvDzNEAyXcza7k29SN6zwO4fdIUrn3sYWxvgC7/aw01LTXc++u/KtfTNdVmF51ybXhlfQxs2lui/WRPmDkbwcBaH/adG0HHuOH1IpbNeVBkkceBAwcwZcoUhISEICoqCh4eHggICEBqaqrM/IcOHUJSUhK73LlzBwKBAAMGDODkCwwM5OTbt2+fwm1RUyo7bOHn54fmzZtDTU0Nv/76K9zd3bFu3Tp8//33uHDhAnR1ddG9e3esWrUKpqamnHUA4JdffoG6ujrGjx+PefPmgccrHRMbPnw4ACAhIaFe6lUXrt4oRLeO2py07n46mDInHQBQVMQg8rYY079+02XP5/PQraMOrkRWfgbUUJUUS5F0LxsdRzdj0/h8HuzbmCHx1kuZ6zwMT4a1hxH+XngHD8KToWusCfceVujweTPwBbLHVwtzS3sctA3UlV+JeiYpliLl/kv4fO7MpvH4PNj4WiDpdkal6xUVlGBL0N9gGAYWzkZo/3VzmDY1YN+38jBB/L8v4NbXDnpmWki8kYaXT3LhN9WiVutTHyTFEqTefwnvYFc2jcfnwdrXAslVtGFxQQl29jgChmFg5myMthNbwOR1GzJSBgkXX6DlSGf89dU5pD18CWEjXXgHu6Jpl8a1XidlkyrYiyDvpZorV67EmDFjEBwcDADYuHEj/v77b2zfvh0//PBDhfzGxtyhuf3790NHR6dC8KCpqQmRSCRn6WuHSvc87Nq1CxoaGrh06RJ+/PFHdO3aFV5eXrhx4wbCwsKQkpKCgQMHVlhHTU0NERERWLNmDVauXImtW7fWUw3qR3KaBBZm3C5LCzMBcl5JUVAgRXqmBBIJZOZJSf3wutzzXxZBKmGgZ6LJSdcz0ah02OLls3zcO5UEqZTBsJ990HmcAy7v+g//bo6VmV8qZRC25C6aeBnBwkEoM09DVvBSDEbCVDib1THRQl6G7IDTyEYfASGt0Gd1ewQt8AXDMNg/6ixepeSzebpM94KJvRBbAo5hjc9BHJ5wAd1+8EJj7+oN0TUkBVlFstvQWAv5GQUy1zG0EaJbiA96ruqI7gvagpEy+CP4NHJft2F+ZiGK80sQueM+mrSzRJ+f/WDfpTGOf3cRzyNln0WruqKiIkRGRsLf359N4/P58Pf3x5UrV6q1jW3btmHQoEHQ1eXOKTl37hzMzc3h5OSE8ePHIyOj8qCwtqlszwMAODg4YOnSpQCABQsWwMvLC4sWLWLf3759O6ytrfHo0SM4OjoCAKytrbFq1SrweDw4OTkhJiYGq1atwpgxYxQuh1gshlj85kcmJydH4W2RhoFhAF1jDfQOaQG+gAcrN0PkpBTi0s54dBnvWCH/3wvvIDXuFT7f1a4eSvt+svIwgZWHCef1zv5huP3Hf2g/obSHMHp/HJJiMtBndXsILXXwLCodZ368CV0zbdi0+fB6H+Rl6WEKSw9T9rWohSn29D+OOwfj0OarFii7QtHerxG8hjkBAMycjJB8Kx0xf8Shkbd5fRRbYRKGB4kCD7kqW6f8sVlTUxOamtyThvT0dEgkElhYcPcvCwsLPHjw4J2fFRERgTt37mDbtm2c9MDAQPTr1w92dnaIj4/HzJkzERQUhCtXrkAgqPv5JyodPHh7e7P/v3XrFsLDw6GnV3HcPz4+ng0e2rRpww5RAEDbtm2xYsUKSCQShf+Aixcvxty5cxVatz6IzARISeM+pDYlTQKhPh/a2nwIBDwIBJCZx8L8w9vldIw0wBfwKvQy5GYUVeiNKKNnqgmBGp8zRGFmr4fcdDFKiqVQU3/TKfj3whg8+jcFn+9sBwORtqzNNXjaRprgCXgVJvblZxRWmMdQGYE6H+ZORshKzAUAFBdKcHFdDHqvbA/7jpYAADNHQ6Q9zMKNXx5+cMGDtqGG7DbMLISOSfX2G4E6H2bOb9pQ21ADfDUejO0NOPmM7IRIik5XTsHrkOKP5C6NoqytrTnpISEhCA0NVUbRWNu2bYO7uzt8fHw46YMGDWL/7+7ujhYtWqBp06Y4d+4cunXrptQyVIdKD1u83SWUm5uLXr16ITo6mrPExsaiU6dOtVqOGTNmIDs7m10SE2VPsntftGmlhbMXud2gp8/no4136UFeQ4MH7xaanDxSKYOzF/PR1rvhTbJ6FzV1PixdDfDftTcHU6mUweOr6bD2kH2pZhMvY2Qm5kEqfTOWmv4kF/pmmmzgwDAM/l4Yg/tnkzFqWxsYNdap3YrUI4E6HxYuRnh67U1XOCNl8DQiFZYtZF9mWJ5UwiA9Lhu6pqX7mLRECmkJA165E02egAdIlVb094ZAXQBzFyM8i0hh0xgpg8SIFIiq3YZSpMdlQddU+802XY3xMoF7xp319BX0LRve/ihl+AovAJCYmMg5Vs+YMaPCZ5iamkIgECAlJYWTnpKS8s75Cnl5edi/fz9Gjx79zrrY29vD1NQUcXFxcrSA8nx4p4EKatmyJQ4ePAhbW1uoqVXeLNeuXeO8vnr1KhwcHGrUbSSr66su5eZJEfe4mH2d8LQE0XfEMDbko0ljdcxcmI7nyRLsWld6pjZuhAHWb8/G9PnpCB4kRPilAvx+JBdHf7FktzF5nCGCv0mFt4cmfDy1sGZLFvLyGYwapF/n9asL7UbY4/D/otHIzQCN3A1x5ZfHKCqQwKtv6ZnKoZk3oW+uhY8muwAAWn9mg4h9Cfjnx7vwHWKLzKd5uLAlDr5D7dht/r3wDmKOP8fgNa2hoauGV+mlZ5RaeupQ12p4l8m9i/cwR4TNiYCFqxFEzY0RtTcWxQUlcOtjCwD4Z1YE9My10XGSOwDgyqZ7sGxhDENrPYhfFePGrofIScqD+yf2AABNPXU09jbD+dW3oaYlgNBSF88i03DvWAL8pnjWUy1rl+dQZ5wOuQpzV2NYuBkjeu8jlBSUwLV3aZucnH0VeubaaPe1BwAgYvMdiNxNYGCtD/GrIkTtfoBXSflwe92GANByhAvCfrgMq5bmaNzKHE8uJ+Hx+Rfot7lrvdSxJmra8yAUCiEUVj3nSENDA97e3jhz5gz69u0LAJBKpThz5gwmTpxY5bq///47xGIxhg0b9s4yPXv2DBkZGbC0tHxn3tpAwcNrEyZMwJYtWzB48GBMmzYNxsbGiIuLw/79+7F161Y2OHj69CmmTJmCcePGISoqCuvWrcOKFSvY7WRmZuLp06d48aL0OvKHDx8CAEQi0XszS7a8G7cK0a3/m+vep4aWnkGPGKiPHWsskJQqQeLzN8GFXRN1HP3VElND0rF2axYaW6ph8wpzBHR505PzWR99pGdIELo0E8lpJfB008TxvVawMPswd7nmgVbIyxTj7PpHyE0XQ+QsxPCNPtAzLQ0Ks5MKOMNdBiJtDN/oi7Bld7Gh/3nom2uhzTA7dPj8zRUb1w88AQDs+Jw7yarvfA82KPmQOAVYI/+lGJc33EV+RiHMnAzRb31HdtjiVXI+eG8d98WvinBqXiTyMwqhKVSHhYsRBu/sCpOmbw7uPX9sg4vrYnB85jUU5hRBaKmLDhPc0WKAffmP/yA4BjRBwctCXNsQU3qjLSdD9P7Jj738Mjc5j9uGOUU4O/868jIKoSXUgJmLEQbs8OcMUzTt2hhdZrbCjR33cH5ZFIxs9NFjWXtYeX14k06VZcqUKRg5ciRatWoFHx8frF69Gnl5eezVFyNGjECjRo2wePFiznrbtm1D3759YWLC7SnKzc3F3Llz0b9/f4hEIsTHx2PatGlo1qwZAgIC6qxeb+MxjGretNvPzw+enp5YvXo1mxYbG4vp06cjPDwcYrEYNjY2CAwMxMqVK8Hj8eDn5wc3NzdIpVLs3bsXAoEA48ePx4IFC9gfhp07d7I7yNvkGRvLycmBgYEBXj6yh1BfpUeWamRBuvO7M5EqZZd8mHMs6lIx8+H1EtWlotxibOp0ENnZ2e88669M2TF1U5Q3tPXkP4EpyC3BuJaRcpXhp59+wrJly5CcnAxPT0+sXbsWvr6+AEp/f2xtbbFz5042/8OHD+Hs7IyTJ0/io48+4n5+QQH69u2LmzdvIisrC1ZWVujevTvmz59fYWJmXVHZ4EERsgKO2kDBg3JQ8FBzFDzUHAUPNaPM4GFDVGuFg4fxLa/XqAwfmg+zD5kQQggpR/GnatJJXHkUPBBCCFEJ9FRN5aHgQQ7nzp2r7yIQQggh9Y6CB0IIISqBhi2Uh4IHQgghKkHx+zxQ8FAeBQ+EEEJUgpThQarAsy0UWedDR8EDIYQQlSBVsOdBkcd4f+ioRQghhBAiF+p5IIQQohLefsiVvOsRLgoeCCGEqAQJeJAocM8GRdb50FHwQAghRCVQz4PyUIsQQgghRC7U80AIIUQlSKDYEIRE+UVp8Ch4IIQQohJo2EJ5KHgghBCiEuj21MpDwQMhhBCVwCj4VE2GrraogMIpQgghhMiFeh4IIYSoBBq2UB4KHgghhKgEejCW8lDwQAghRCXQI7mVh4IHQgghKoF6HpSHwilCCCGEyIV6HgghhKgEKfiQKnDOrMg6HzoKHgghhKgECcODRIEhCEXW+dBR8EAIIUQl0JwH5aHggRBCiEpgFHy2BUP3eaiAWoQQQgghcqGeB0IIISpBAp6Cj+SmYYvyKHgghBCiEqSMYvMXpEwtFKaBo+CBEEKISpAqOOdBkXU+dNQihBBCiJKtX78etra20NLSgq+vLyIiIirNu3PnTvB4PM6ipaXFycMwDObMmQNLS0toa2vD398fsbGxtV2NSlHwQAghRCVIwVN4kceBAwcwZcoUhISEICoqCh4eHggICEBqamql6wiFQiQlJbHLkydPOO8vXboUa9euxcaNG3Ht2jXo6uoiICAAhYWFCrVFTVHwQAghRCWU3SRKkUUeK1euxJgxYxAcHAxXV1ds3LgROjo62L59e6Xr8Hg8iEQidrGwsGDfYxgGq1evxqxZs9CnTx+0aNECu3fvxosXL/Dnn38q2hw1QsEDIYQQlVA250GRBQBycnI4i1gsrvAZRUVFiIyMhL+/P5vG5/Ph7++PK1euVFq23Nxc2NjYwNraGn369MHdu3fZ9x4/fozk5GTONg0MDODr61vlNmsTTZh8jy3NcICWWL2+i9FgzTJ9UN9FaPBmprSo7yI0eHR3wprhQ6q0bUmh4B0mXw9bWFtbc9JDQkIQGhrKSUtPT4dEIuH0HACAhYUFHjyQfUxycnLC9u3b0aJFC2RnZ2P58uVo164d7t69i8aNGyM5OZndRvltlr1X1yh4IIQQQqohMTERQqGQfa2pqamU7bZt2xZt27ZlX7dr1w4uLi7YtGkT5s+fr5TPUDYatiCEEKISGAUnSzKvex6EQiFnkRU8mJqaQiAQICUlhZOekpICkUhUrXKqq6vDy8sLcXFxAMCuV5NtKhsFD4QQQlRC2YOxFFmqS0NDA97e3jhz5sybz5VKcebMGU7vQlUkEgliYmJgaWkJALCzs4NIJOJsMycnB9euXav2NpWNhi0IIYSohLq6SdSUKVMwcuRItGrVCj4+Pli9ejXy8vIQHBwMABgxYgQaNWqExYsXAwDmzZuHNm3aoFmzZsjKysKyZcvw5MkTfPHFFwBKr8SYPHkyFixYAAcHB9jZ2WH27NmwsrJC37595a6PMlDwQAghRCXU1SO5P/vsM6SlpWHOnDlITk6Gp6cnwsLC2AmPT58+BZ//JiB5+fIlxowZg+TkZBgZGcHb2xuXL1+Gq6srm2fatGnIy8vD2LFjkZWVhQ4dOiAsLKzCzaTqCo9hGLpr93smJycHBgYGmHElAFp6dLWFouhqi5qjqy1qTiylc7SaEOcWY0Onw8jOzuZMVpRH2TG1z8nPoa6rIff6xXlF+Kv79hqV4UNDezUhhBCVoMjdIsvWI1wUPBBCCFEJdTVsoQooeCCEEKISKHhQHgoeCCGEqAQKHpSH7vNACCGEELlQzwMhhBCVQD0PykPBAyGEEJXAQLErJ+h+BhVR8EAIIUQlUM+D8tCcB0IIIYTIhXoeCCGEqATqeVAeCh4IIYSoBAoelIeCB0IIISqBggfloeCBEEKISmAYHhgFAgFF1vnQ0YRJQgghhMiFeh4IIYSoBHqqpvJQ8EAIIUQl0JwH5aHggRBCiEqgOQ/KQ8EDIYQQlUA9D8pDEyYJIYQQIhfqeSCEEKISaNhCeSh4IIQQohIYBYctKHioiIIHQgghKoEBwCjwfG16JHdFFDwQQghRCVLwwKP7PCgFTZgkhBBCiFyo54EQQohKoAmTykPBAyGEEJUgZXjg0X0elIKCB0IIISqBYRScMEkzJiug4IEQQohKoGEL5aHggQAAru1LwOWd8chNF8PCSYgeM9zQ2N2o0vwFOcU4s/YB7p9JRkF2MQyttBE4zRWOnSwAAOe3xuH+6SSkP86FupYA1h5G+OhbF5ja6dVVlerU+SsFWL7hJaJui5GUIsHB7SL0Daq6rucu5+O7kAzcfSSGtZU6Zk42wqjPhJw8P+/IwvKfs5CcJoGHqwbWLDSDj5dWbValXkXtj8e1XbHIyyiEuaMB/Kd7wMrdWGbemL+e4HhIJCdNoMHHdxF92dd5GYU4t/oOEq6movBVMaxbmsB/uieMbT7M/RAAbh2IxY3dD5GfUQhTR0N0meYFUXMTmXnvHnmMU6HXOWkCDT6+vvop+3p1y99krtvhmxZoNdJZeQUnDQpdbaFExcXFmD59Otzd3aGrqwsrKyuMGDECL168qO+iVelO2AucWHYPfl86YtxvHSFyFOKXcRHIzRDLzF9SLMXusVeR9aIAn630xtdH/dA7pAWEFtpsnic3MuAzyBZj9nTAiM1tIClhsHvcNRTll9RVtepUXr4UHq6aWLfIrFr5Hz8tRq9hSfBrr42oU03wzRgDjJ2aihPheWyeA3+9wtTQdMyeaowbJ6zRwlUTQYNfIDX9w2zD+yee4eyKGLQf54xR+7rC3NEAv311CXmZhZWuo6Gnhgmne7DL+H8C2fcYhsGhb68i63ke+q1qg1H7u0JoqYMDX15AUcGH2YYPTzzF+ZW30GasG4bs/QhmDoY4POE88qtsQ3WMOdmLXT7/+2PO+2+/N+ZkL3wU0hrgAQ7dGtd2dZSurOdBkUVe69evh62tLbS0tODr64uIiIhK827ZsgUdO3aEkZERjIyM4O/vXyH/qFGjwOPxOEtgYGAlW6x9FDy8VlRUVONt5OfnIyoqCrNnz0ZUVBQOHTqEhw8fonfv3kooYe25vPs/ePe3htcn1jBvqo+P57hDXZuPm4cTZea/eTgRBdnFGLymFZp4GcOokQ5sW5tA5PTmrHn4Rl949bWGeTN9iJyE+GSBB7KTCvDiXnZdVatOBXXTxfwfTPBJj+qd0W7anQ27JupYHmoKF0cNTPjcEP0/1sPqzW/aZ/WmLHwx1ADBg4RwddLAhqVm0NHmYce+V7VVjXp1/ZdYePSzRYu+tjBtKkTALC+oawkQ8+eTStfhgQc9Uy120TV50yvz8mkuXtzORPeZXrBsbgwTW30E/M8LJYVS3P9H9r7d0EXteYTmn9jDrY8dTOwN0O1/3lDTUsPdvx5XuZ6uqfabxUSr8vdMtRH/73NYtzKHQeOG13tT9mAsRRZ5HDhwAFOmTEFISAiioqLg4eGBgIAApKamysx/7tw5DB48GOHh4bhy5Qqsra3RvXt3PH/+nJMvMDAQSUlJ7LJv3z6F26KmVDZ48PPzw8SJEzF58mSYmpoiICAAd+7cQVBQEPT09GBhYYHhw4cjPT29wjoTJ06EgYEBTE1NMXv2bDCvZ9MYGBjg1KlTGDhwIJycnNCmTRv89NNPiIyMxNOnT+urqlUqKZYi6V427Nu8OWPm83mwb2OGxFsvZa7zMDwZ1h5G+HvhHSztfBLrP/kX57fEQiqpfFZRYW7pmZ62gbpyK9BAXb1RiG4dtTlp3f10cDWy9AyxqIhB5G0xJw+fz0O3jjq4Eln5WWRDJSmWIvl+Fmx8zdk0Hp8HW19zPL+dWel6RQUl2BD0D34O+AcHJ19BWlzOm20WSQEAappvDnM8Pg8CDT6e3cyohVrUL0mxBKn3X8La14JN4/F5aOJrjqTblde3uKAE23ocw9agozjy7UVkxFce4OdlFCLhYhLc+toptex1pWzCpCKLPFauXIkxY8YgODgYrq6u2LhxI3R0dLB9+3aZ+ffs2YOvvvoKnp6ecHZ2xtatWyGVSnHmzBlOPk1NTYhEInYxMqp8aLm2qWzwAAC7du2ChoYGLl26hB9//BFdu3aFl5cXbty4gbCwMKSkpGDgwIEV1lFTU0NERATWrFmDlStXYuvWrZV+RnZ2Nng8HgwNDWu5NorJf1kEqYSBnokmJ13PRKPSYYuXz/Jx71QSpFIGw372QedxDri86z/8uzlWZn6plEHYkrto4mUECwehzDyqJjlNAgszASfNwkyAnFdSFBRIkZ4pgUQCmXlSUj+8Lvf8l2IwEga65fZDHRNN5KXLDpaMbfXQI7Ql+q1qi48XtgIjZfDrqHPIScl//b4+hJba+HftXRTmFEFSLMXVHQ/xKqUAuZVssyEryCoCI2GgY1yuDY21kJchu75GNvr4KKQ1eq1qj8AFvmCkDA4En8Wr121Y3v2jCVDXUUezrg1vyKKuFBUVITIyEv7+/mwan8+Hv78/rly5Uq1t5Ofno7i4GMbG3Pk+586dg7m5OZycnDB+/HhkZNRfEKzSEyYdHBywdOlSAMCCBQvg5eWFRYsWse9v374d1tbWePToERwdHQEA1tbWWLVqFXg8HpycnBATE4NVq1ZhzJgxFbZfWFiI6dOnY/DgwRAKK//RFIvFEIvf/FDn5ORUmvd9wDCArrEGeoe0AF/Ag5WbIXJSCnFpZzy6jHeskP/vhXeQGvcKn+9qVw+lJR+qRh4maORhwnm9td8pRP/xGJ0muEGgzscnK9rgn9AorOl0DDwBD7a+ZrBvb0HPKnjNysMUVh6m7GvLFqbY3T8MMQfj0e4r9wr57x55DOegJlDTFFR4ryEo7UVQ5GqL0n/LH5s1NTWhqckN1tLT0yGRSGBhYcFJt7CwwIMHD6r1edOnT4eVlRUnAAkMDES/fv1gZ2eH+Ph4zJw5E0FBQbhy5QoEgrr/e6h08ODt7c3+/9atWwgPD4eeXsVxvPj4eDZ4aNOmDXi8Nztf27ZtsWLFCkgkEs4fsLi4GAMHDgTDMNiwYUOV5Vi8eDHmzp1b0+ooRMdIA3wBr0IvQ25GUYXeiDJ6ppoQqPHBF7xpBzN7PeSmi1FSLIWa+psOrb8XxuDRvyn4fGc7GIi0ZW1OJYnMBEhJk3DSUtIkEOrzoa3Nh0DAg0AAmXkszD+8r62OkSZ4Ah7yyu2H+Rli6JpW7+oSgTofFk6GyEp8M+lU5GqE4N+6QfyqGJJiKXSMNbF7WDhErvXX3VtbtA01wBPwkJ9Zrg0zCyvMY6iMQJ0Pc2dDZCXmVnjveVQaXia8Qo8f2yqlvPWhppdqWltbc9JDQkIQGhqqjKKxfvzxR+zfvx/nzp2Dltabv9ugQYPY/7u7u6NFixZo2rQpzp07h27duim1DNWh0sMWurq67P9zc3PRq1cvREdHc5bY2Fh06tRJru2WBQ5PnjzBqVOnqux1AIAZM2YgOzubXRIT624yl5o6H5auBvjv2pu5HVIpg8dX02HtIfsA28TLGJmJeZBK35y/pT/Jhb6ZJhs4MAyDvxfG4P7ZZIza1gZGjXVqtyINTJtWWjh7sYCTdvp8Ptp4lx4sNDR48G6hyckjlTI4ezEfbb0/vEs1Bep8iFwM8STizYQyRsogISIVjVrIvlSzPKmEQVpctsxgQ1NfHTrGmsh8kovkey/h4GeptLK/LwTqApi7GCExIoVNY6QMEiNSYdlC9qWa5UklUqTHZUPXtGKgf+evxzB3MYKZo6GyilznmBosAJCYmMg5Vs+YMaPCZ5iamkIgECAlJYWTnpKSApFIVGX5li9fjh9//BEnT55EixYtqsxrb28PU1NTxMXFvaPWtePDO4VRUMuWLXHw4EHY2tpCTa3yZrl27Rrn9dWrV+Hg4MD2OpQFDrGxsQgPD4eJybu/tLK6vupSuxH2OPy/aDRyM0Ajd0Nc+eUxigok8OpbGmUfmnkT+uZa+GiyCwCg9Wc2iNiXgH9+vAvfIbbIfJqHC1vi4Dv0zSSqvxfeQczx5xi8pjU0dNXw6vUYs5aeOtS1GmaXZ1Vy86SIe1zMvk54WoLoO2IYG/LRpLE6Zi5Mx/NkCXatK+3KHDfCAOu3Z2P6/HQEDxIi/FIBfj+Si6O/vPlRmzzOEMHfpMLbQxM+nlpYsyULefkMRg3Sr/P61YXWwx3w9+wbELkawbK5EW7siUNxgQTufWwAAMdm3YC+uRY6T2oOALi06T6s3I1h1EQPha+KELErFjlJ+fD4xJbd5oOTz6BjpAmhpQ7SYrNxeultOHSxgl07C1lFaPBaDnXEyZAIWLgaQ+RmjKi9j1BcUALX3qXfzROzr0HXXBsdvi79Ybq6+S4s3U1gaK0H8ati3Nj9ADlJ+Wj+CXdCpDi3GLGnEtFpiked10mZatrzIBQK33kyqKGhAW9vb5w5cwZ9+/YFAHby48SJEytdb+nSpVi4cCFOnDiBVq1avbNMz549Q0ZGBiwt6ycQpuDhtQkTJmDLli0YPHgwpk2bBmNjY8TFxWH//v3YunUrGxw8ffoUU6ZMwbhx4xAVFYV169ZhxYoVAEoDh08//RRRUVE4duwYJBIJkpOTAQDGxsbQ0NCot/pVpXmgFfIyxTi7/hFy08UQOQsxfKMP9ExLA5rspALOUI2BSBvDN/oibNldbOh/HvrmWmgzzA4dPm/G5rl+oPTyuh2fcycI9Z3vwQYlH5IbtwrRrf+b+3lMDS3tyRkxUB871lggKVWCxOdvggu7Juo4+qslpoakY+3WLDS2VMPmFeYI6PKmN+yzPvpIz5AgdGkmktNK4OmmieN7rWBh9mF+bV0CGiP/pRgXN9xDXroY5k4GGPhze7bLPScpH2/thijMKUbY/CjkpYuhJVSHhYshhu3yg2nTNwf33PRCnF0Rg7yMQuiZacHt4yZoP9aljmtWd5wCmqDgpRhXNtwpvUmUkyH6/tTpTRsm5wP8N40ozinC6fk3kJ9RCE2hBsxdjPDZjq4wsTfgbPfRiafs9sm7TZkyBSNHjkSrVq3g4+OD1atXIy8vD8HBwQCAESNGoFGjRli8eDEAYMmSJZgzZw727t0LW1tb9ndDT08Penp6yM3Nxdy5c9G/f3+IRCLEx8dj2rRpaNasGQICAuqljjyGUc27dvv5+cHT0xOrV69m02JjYzF9+nSEh4dDLBbDxsYGgYGBWLlyJXg8Hvz8/ODm5gapVIq9e/dCIBBg/PjxWLBgAXg8HhISEmBnJ/sSpvDwcPj5+VWrbDk5OTAwMMCMKwHQ0qNLGxU1y7R6k5NI5WamVN11St5NLP0wg726Is4txoZOh5Gdnf3Os/7KlB1T7XfNhEBH/mE/SX4h/hu5SK4y/PTTT1i2bBmSk5Ph6emJtWvXwtfXF0Dp74+trS127twJALC1tcWTJxXvZ1I2p6KgoAB9+/bFzZs3kZWVBSsrK3Tv3h3z58+vMDGzrqjsXn3u3LkKaQ4ODjh06FCV66mrq2P16tUyJ0Ha2tpCRWMxQgh5/yk4bAEF1im7J5As5X9/EhISqtyWtrY2Tpw4IXcZapPKBg+EEEJUCz1VU3koeCCEEKIS6KmaykPBgxxkDXUQQgghqoaCB0IIIaqB4Sk0f0GhdT5wFDwQQghRCTTnQXkoeCCEEKIa3r5dpLzrEQ4KHgghhKgEmjCpPCr9bAtCCCGEyI96HgghhKgOGoJQCgoeCCGEqAQatlAeCh4IIYSoBpowqTQ054EQQgghcqGeB0IIISqC93pRZD3yNgoeCCGEqAYatlAaCh4IIYSoBgoelIaCB0IIIaqBnm2hNDRhkhBCCCFyoZ4HQgghKoEejKU8FDwQQghRDTTnQWkoeCCEEKIaaM6D0lDwQAghRCXwmNJFkfUIF02YJIQQQohcqOeBEEKIaqA5D0pDwQMhhBDVQHMelIaCB0IIIaqBeh6UhoIHQgghqoGCB6WhCZOEEEIIkQv1PBBCCFEN1POgNBQ8EEIIUQ00YVJpKHgghBCiEugmUcpDcx4IIYQQJVu/fj1sbW2hpaUFX19fREREVJn/999/h7OzM7S0tODu7o7jx49z3mcYBnPmzIGlpSW0tbXh7++P2NjY2qxClSh4IIQQohqYGixyOHDgAKZMmYKQkBBERUXBw8MDAQEBSE1NlZn/8uXLGDx4MEaPHo2bN2+ib9++6Nu3L+7cucPmWbp0KdauXYuNGzfi2rVr0NXVRUBAAAoLC+UrnJJQ8EAIIYQo0cqVKzFmzBgEBwfD1dUVGzduhI6ODrZv3y4z/5o1axAYGIjvv/8eLi4umD9/Plq2bImffvoJQGmvw+rVqzFr1iz06dMHLVq0wO7du/HixQv8+eefdVizNyh4IIQQohJ4eDPvQa7l9fo5OTmcRSwWV/iMoqIiREZGwt/fn03j8/nw9/fHlStXZJbrypUrnPwAEBAQwOZ//PgxkpOTOXkMDAzg6+tb6TZrG02YfI/llGhDXKJe38VosGamtKjvIjR4iyxu13cRGrzpKZ71XYQGTcKXKm9jNbzawtrampMcEhKC0NBQTlp6ejokEgksLCw46RYWFnjw4IHMzScnJ8vMn5yczL5fllZZnrpGwQMhhBBSDYmJiRAKhexrTU3NeixN/aJhC0IIIaqhhhMmhUIhZ5EVPJiamkIgECAlJYWTnpKSApFIJLNYIpGoyvxl/8qzzdpGwQMhhBDVUAdXW2hoaMDb2xtnzpxh06RSKc6cOYO2bdvKXKdt27ac/ABw6tQpNr+dnR1EIhEnT05ODq5du1bpNmsbDVsQQghRCXV1k6gpU6Zg5MiRaNWqFXx8fLB69Wrk5eUhODgYADBixAg0atQIixcvBgB888036Ny5M1asWIGePXti//79uHHjBjZv3lz6+TweJk+ejAULFsDBwQF2dnaYPXs2rKys0LdvX/krpAQUPBBCCFENdfRsi88++wxpaWmYM2cOkpOT4enpibCwMHbC49OnT8Hnv+n4b9euHfbu3YtZs2Zh5syZcHBwwJ9//onmzZuzeaZNm4a8vDyMHTsWWVlZ6NChA8LCwqClpaVAhWqOxzAM3XjzPZOTkwMDAwNMuNAXmnp0tYWi1HmS+i5Cg0dXW9QcXW1RM+LcYqzr8Beys7M5kxXlUXZMtV2wEHwFfmylhYVImPW/GpXhQ0M9D4QQQlQDPVVTaSh4IIQQohLowVjKQ8EDIYQQ1UCP5FYaCh4IIYSoBhq2UBq6zwMhhBBC5EI9D4QQQlQCzXlQHgoeCCGEqAYatlAaCh4IIYSoBgV7Hih4qIiCB0IIIaqBeh6UhiZMEkIIIUQu1PNACCFENVDPg9JQ8EAIIUQl0NUWykPDFoQQQgiRCwUPhBBCCJELDVsQQghRDTTnQWkoeCCEEKISaM6D8lDwQAghRHVQIKAUFDwQQghRDTRsoTQ0YZIQQgghcqGeB0IIISqB5jwoDwUPhBBCVAMNWygNBQ+EEEJUAvU8KA8FD4QQQlQD9TwoDU2YJIQQQohcqOeBEEKIaqCeB6Wh4IEQQohKoDkPykPBAyGEENVAPQ9KQ8EDIYQQ1UDBg9JQ8EAAANEH4nBj10PkZRTCzNEQXaZ7wbK5scy8d48k4ETIdU6aQIOPb671Z18X5ZfgwtrbiA9/gYJsMQysdOE12AEeA5rWaj3qU9T+eFzbFYu8jEKYOxrAf7oHrNxlt2HMX09wPCSSkybQ4OO7iL7s67yMQpxbfQcJV1NR+KoY1i1N4D/dE8Y2erVZjXpx/koBlm94iajbYiSlSHBwuwh9g6qu57nL+fguJAN3H4lhbaWOmZONMOozISfPzzuysPznLCSnSeDhqoE1C83g46VVm1Wpdzf3x+H6rkevv8sG6DbdC5aV7Id3/kpAWMgNTppAg49vI/qxr4vyS3B+TQziwl+gMFsMYSNdtBzcDJ4f8HeZvBtdbaFkoaGhcHZ2hq6uLoyMjODv749r167Vd7Gq9PBEIv5dcQttxrli2N6PYOZogENfnUd+ZmGl62joqWHcqV7s8sXxnpz3/10RjYTLyQha6INRhwLRcqgjzi65ifhzL2q7OvXi/olnOLsiBu3HOWPUvq4wdzTAb19dQt472nDC6R7sMv6fQPY9hmFw6NuryHqeh36r2mDU/q4QWurgwJcXUFRQUhdVqlN5+VJ4uGpi3SKzauV//LQYvYYlwa+9NqJONcE3YwwwdmoqToTnsXkO/PUKU0PTMXuqMW6csEYLV00EDX6B1PQPr/3KPDiRiHMrbqPtOFcM3+cPc0dD/PHVhXfuh+NPf8wuY//pwXn/3PJbSLicjB4LWyP4UAC8hzjgzI/RiGuA3+WyOQ+KLLUlMzMTQ4cOhVAohKGhIUaPHo3c3Nwq83/99ddwcnKCtrY2mjRpgkmTJiE7O5tbVx6vwrJ//36llZuCh9eKioqUsh1HR0f89NNPiImJwcWLF2Fra4vu3bsjLS1NKduvDZG/PkLzfnZo3scOJk2F8P+fN9S0BLjzZ0Kl6/DAg66p1pvFhHs29+JWBtw+toV1K3MYWOmiRX97mDkaIPluZi3Xpn5c/yUWHv1s0aKvLUybChEwywvqWgLE/Pmk0nV44EHPVItd3m7Dl09z8eJ2JrrPLO0BMrHVR8D/vFBSKMX9fxLrokp1KqibLub/YIJPelSvV2XT7mzYNVHH8lBTuDhqYMLnhuj/sR5Wb35zAF29KQtfDDVA8CAhXJ00sGGpGXS0edix71VtVaPe3fjlEdz72cH99X740ayWUK/hd/n5rQy49bJBk9bmMGikC49P7WHuaIDkOw3wu8zUYKklQ4cOxd27d3Hq1CkcO3YM58+fx9ixYyvN/+LFC7x48QLLly/HnTt3sHPnToSFhWH06NEV8u7YsQNJSUns0rdvX6WVW2WDBz8/P0ycOBGTJ0+GqakpAgICcOfOHQQFBUFPTw8WFhYYPnw40tPTK6wzceJEGBgYwNTUFLNnzwbDvNmzhgwZAn9/f9jb28PNzQ0rV65ETk4Obt++XR/VfCdJsRQp91/CxteCTePxebDxtUDS7YxK1ysqKMGWoL+xOfAY/pp8Cenx3KjXysME8f++wKvUAjAMg6fXU/HySS5s2lhUssWGS1IsRfL9LNj4mrNpPD4Ptr7meH678gNsUUEJNgT9g58D/sHByVeQFpfzZptFUgCAmuabryiPz4NAg49nNyv/u6iKqzcK0a2jNietu58OrkaWnmEXFTGIvC3m5OHzeejWUQdXIis/C2/ISr/LFffDJr4WePGO7/KmoOPYFPA3Dk++hPQ47ne5kYcJ4s4l4VXKm+9y5pNc2LRteN/l963n4f79+wgLC8PWrVvh6+uLDh06YN26ddi/fz9evJDds9O8eXMcPHgQvXr1QtOmTdG1a1csXLgQR48eRUkJt1fN0NAQIpGIXbS0lDdkp7LBAwDs2rULGhoauHTpEn788Ud07doVXl5euHHjBsLCwpCSkoKBAwdWWEdNTQ0RERFYs2YNVq5cia1bt8rcflFRETZv3gwDAwN4eHjURZXkVvBSDEbCQMeYu1PpmGghL0P2QdbIRh8BIa3QZ3V7BC3wBcMw2D/qLF6l5LN5ukz3gom9EFsCjmGNz0EcnnAB3X7wQmPv6nVLNyT5r9tQ10STk65joom8dNltaGyrhx6hLdFvVVt8vLAVGCmDX0edQ87rNjS21YfQUhv/rr2LwpwiSIqluLrjIV6lFCC3km2qkuQ0CSzMBJw0CzMBcl5JUVAgRXqmBBIJZOZJSf0why0K2P2Q+13WrXI/1EdgaCt8sqodeiz0ASMF9o4K53yXu/7gCRN7ITYF/I1VrQ/h4FcX4T/DC9Yf4He5rl25cgWGhoZo1aoVm+bv7w8+ny/XcHd2djaEQiHU1LjTGCdMmABTU1P4+Phg+/btnBPdmlLpCZMODg5YunQpAGDBggXw8vLCokWL2Pe3b98Oa2trPHr0CI6OjgAAa2trrFq1CjweD05OToiJicGqVaswZswYdr1jx45h0KBByM/Ph6WlJU6dOgVTU9NKyyEWiyEWi9nXOTk5leZ9H1h5mMDKw4Tzemf/MNz+4z+0n9AcABC9Pw5JMRnos7o9hJY6eBaVjjM/3oSumfYH2fsgr0YeJmj0Vhs28jDB1n6nEP3HY3Sa4AaBOh+frGiDf0KjsKbTMfAEPNj6msG+vQVN/CZKI+u7vKPfCdz64z90eP1dvrmv9Lv8yZp2EFrqIDEqHacX34SemVbD+y7X8GqL8sdmTU1NaGpqylihepKTk2Fubs5JU1NTg7GxMZKTk6u1jfT0dMyfP7/CUMe8efPQtWtX6Ojo4OTJk/jqq6+Qm5uLSZMmKVxeTjmVspUGytvbm/3/rVu3EB4eDj29imOu8fHxbPDQpk0b8Hg89r22bdtixYoVkEgkEAhKz3K6dOmC6OhopKenY8uWLRg4cCCuXbtWYScps3jxYsydO1eZVas2bSNN8AS8CpMj8zMKK5zBVEagzoe5kxGyEksn+RQXSnBxXQx6r2wP+46WAAAzR0OkPczCjV8eNrwDzjvovG7DvAwxJz0/Qwxd0+q3oYWTIbIS30z4E7kaIfi3bhC/KoakWAodY03sHhYOkauRUsvfEInMBEhJk3DSUtIkEOrzoa3Nh0DAg0AAmXkszD/Mw542ux9yv8t5cu6H5m/th8WFElxYdwd9VrZD007c7/L13Y8a3ne5hsGDtbU1JzkkJAShoaEVsv/www9YsmRJlZu8f/++AgXhysnJQc+ePeHq6lqhHLNnz2b/7+Xlhby8PCxbtkxpwYNKD1vo6uqy/8/NzUWvXr0QHR3NWWJjY9GpUye5t9usWTO0adMG27Ztg5qaGrZt21Zp/hkzZiA7O5tdEhPrbkKcQJ0PCxcjPL2WyqYxUgZPI1Jh2cKkijXfkEoYpMdlswcoaYkU0hIGb8VYAACegAdIlVb094ZAnQ+RiyGeRHDbMCEiFY1ayL5ErjyphEHaW234Nk19degYayLzSS6S772Eg5+l0sreULVppYWzFws4aafP56ONd2n7aWjw4N1Ck5NHKmVw9mI+2np/mJdqln6XDfE0ouJ32Uqu73JOxe9yuV8KHp8HRtrw+sB4NVgAIDExkXOsnjFjhszPmTp1Ku7fv1/lYm9vD5FIhNTUVM66JSUlyMzMhEgkqrIur169QmBgIPT19XH48GGoq6tXmd/X1xfPnj3j9HLXxIcZgiugZcuWOHjwIGxtbSuMG72t/DjU1atX4eDgwPY6yCKVSqv8g9W066umvIc5ImxOBCxcjSBqboyovbEoLiiBWx9bAMA/syKgZ66NjpPcAQBXNt2DZQtjGFrrQfyqGDd2PUROUh7cP7EHAGjqqaOxtxnOr74NNS0BhJa6eBaZhnvHEuA3xbOealm7Wg93wN+zb0DkagTL5ka4sScOxQUSuPexAQAcm3UD+uZa6DyptCv40qb7sHI3hlETPRS+KkLErljkJOXD4xNbdpsPTj6DjpEmhJY6SIvNxumlt+HQxQp27RrY2V415OZJEfe4mH2d8LQE0XfEMDbko0ljdcxcmI7nyRLsWlda93EjDLB+ezamz09H8CAhwi8V4PcjuTj6y5vAavI4QwR/kwpvD034eGphzZYs5OUzGDVIv87rV1daDXfEP7Ovw8LVCJbNjRG5p/S73Pz1d/n46+9yp9ff5cub7sHK3RiGTUq/y9d3PXr9XbYDUPZdNsW/q2KgpimA0EoXz26k4d6xJ/Cb+n7O46pSDXsehEIhhEJh1XkBmJmZwczs3XNC2rZti6ysLERGRrI94WfPnoVUKoWvr2+l6+Xk5CAgIACampo4cuRItSZCRkdHw8jISGm/NRQ8vDZhwgRs2bIFgwcPxrRp02BsbIy4uDjs378fW7duZYODp0+fYsqUKRg3bhyioqKwbt06rFixAgCQl5eHhQsXonfv3rC0tER6ejrWr1+P58+fY8CAAfVZvSo5BVgj/6UYlzfcRX5GIcycDNFvfUd22OJVcj7nzEP8qgin5kUiP6MQmkJ1WLgYYfDOrjBp+uZL1fPHNri4LgbHZ15DYU4RhJa66DDBHS0G2Nd19eqES0Bj5L8U4+KGe8hLF8PcyQADf27PtmFOUj6nJ6Ywpxhh86OQly6GllAdFi6GGLbLD6ZvtWFueiHOrohBXkYh9My04PZxE7Qf61LHNasbN24Volv/N7PLp4aWXuU0YqA+dqyxQFKqBInP3wQXdk3UcfRXS0wNScfarVlobKmGzSvMEdDlTW/iZ330kZ4hQejSTCSnlcDTTRPH91rBwuzDPew5v/4uX9pwD/nphTBzMsCnP3cotx++2RHFOUU4MT8K+elvfZd3deHsh72WtMH5tTE4PjPizXd5YnN4fKDf5brk4uKCwMBAjBkzBhs3bkRxcTEmTpyIQYMGwcrKCgDw/PlzdOvWDbt374aPjw9ycnLQvXt35Ofn49dff0VOTg47F8PMzAwCgQBHjx5FSkoK2rRpAy0tLZw6dQqLFi3Cd999p7Sy8xhlTr9sQPz8/ODp6YnVq1ezabGxsZg+fTrCw8MhFothY2ODwMBArFy5EjweD35+fnBzc4NUKsXevXshEAgwfvx4LFiwADweD4WFhRgyZAiuXbuG9PR0mJiYoHXr1pg1axZat25d7bLl5OTAwMAAEy70haZe1V1RpHLqPMm7M5EqLbJ4Py8xbkimp3jWdxEaNHFuMdZ1+Iu9okARZcdUty8XQaAp/7CVRFyIuxtn1qgMlcnMzMTEiRNx9OhR8Pl89O/fH2vXrmXn3yUkJMDOzg7h4eHw8/PDuXPn0KVLF5nbevz4MWxtbREWFoYZM2YgLi4ODMOgWbNmGD9+PMaMGQM+XzmzFVQ2eFCErICjNlDwoBwUPNQcBQ81R8FDzSg1eBhXg+BhU+0EDw3Vh9t/RwghhJRHp8tKQcEDIYQQlaDo3SJr89kWDRUFD3I4d+5cfReBEEIIqXcUPBBCCFENNbxUk7xBwQMhhBCVQMMWykPBAyGEENVAPQ9KQ8EDIYQQlUA9D8qj0s+2IIQQQoj8qOeBEEKIaqBhC6Wh4IEQQohqoOBBaSh4IIQQohJozoPy0JwHQgghhMiFeh4IIYSoBhq2UBoKHgghhKgEHsOAp8CDpBVZ50NHwQMhhBDVQD0PSkPBAyGEEJVAEyaVhyZMEkIIIUQu1PNACCFENdCwhdJQ8EAIIUQl0LCF8lDwQAghRDVQz4PSUPBACCFEJVDPg/LQhElCCCGEyIV6HgghhKgGGrZQGgoeCCGEqAwaglAOCh4IIYSoBoYpXRRZj3BQ8EAIIUQl0IRJ5aEJk4QQQgiRC/U8EEIIUQ00YVJpKHgghBCiEnjS0kWR9QgXBQ+EEEJUA/U8KA3NeSCEEKISyiZMKrLUlszMTAwdOhRCoRCGhoYYPXo0cnNzq1zHz88PPB6Ps3z55ZecPE+fPkXPnj2ho6MDc3NzfP/99ygpKVFauanngRBCCKknQ4cORVJSEk6dOoXi4mIEBwdj7Nix2Lt3b5XrjRkzBvPmzWNf6+josP+XSCTo2bMnRCIRLl++jKSkJIwYMQLq6upYtGiRUspNwQMhhBDV8J7d5+H+/fsICwvD9evX0apVKwDAunXr0KNHDyxfvhxWVlaVrqujowORSCTzvZMnT+LevXs4ffo0LCws4Onpifnz52P69OkIDQ2FhoZGjctOwxaEEEJUQk2HLXJycjiLWCyuUXmuXLkCQ0NDNnAAAH9/f/D5fFy7dq3Kdffs2QNTU1M0b94cM2bMQH5+Pme77u7usLCwYNMCAgKQk5ODu3fv1qjMZajn4T1WwgjAZwT1XYwGS8rw6rsIDd70FM/6LkKDt8Qiur6L0KDl6EixTlkbq+GESWtra05ySEgIQkNDFS5OcnIyzM3NOWlqamowNjZGcnJypesNGTIENjY2sLKywu3btzF9+nQ8fPgQhw4dYrf7duAAgH1d1XblQcEDIYQQUg2JiYkQCoXsa01NTZn5fvjhByxZsqTKbd2/f1/hcowdO5b9v7u7OywtLdGtWzfEx8ejadOmCm9XHhQ8EEIIUQk1vT21UCjkBA+VmTp1KkaNGlVlHnt7e4hEIqSmpnLSS0pKkJmZWel8Bll8fX0BAHFxcWjatClEIhEiIiI4eVJSUgBAru1WhYIHQgghqqGOJkyamZnBzMzsnfnatm2LrKwsREZGwtvbGwBw9uxZSKVSNiCojujoaACApaUlu92FCxciNTWVHRY5deoUhEIhXF1d5apLZWjCJCGEEJXwvt3nwcXFBYGBgRgzZgwiIiJw6dIlTJw4EYMGDWKvtHj+/DmcnZ3ZnoT4+HjMnz8fkZGRSEhIwJEjRzBixAh06tQJLVq0AAB0794drq6uGD58OG7duoUTJ05g1qxZmDBhQqVDLfKi4IEQQohqYGqw1JI9e/bA2dkZ3bp1Q48ePdChQwds3ryZfb+4uBgPHz5kr6bQ0NDA6dOn0b17dzg7O2Pq1Kno378/jh49yq4jEAhw7NgxCAQCtG3bFsOGDcOIESM494WoKRq2IIQQQuqJsbFxlTeEsrW1BfPWsIm1tTX+/fffd27XxsYGx48fV0oZZaHggRBCiEqo6YRJ8gYFD4QQQlSDlCldFFmPcFDwQAghRDXQUzWVhoIHQgghKoEHBYctlF6Sho+utiCEEEKIXKjngRBCiGp4z56q2ZBR8EAIIUQl0NUWykPBAyGEENVAEyaVhoIHQgghKoHHMOApMAShyDofOpowSQghhBC5UM8DIYQQ1SB9vSiyHuGg4IEQQohKoGEL5aHggRBCiGqgCZNKQ3MeCCGEECIX6nkghBCiGugmUUpDwQMhhBCVQDeJUh4KHgghhKgG6nlQGgoeCCGEqASetHRRZD3CRRMmCSGEECIX6nkghBCiGmjYQmkoeCCEEKIa6D4PSkPBAyGEEJVAd5hUHgoeCCGEqAYatlAamjBJCCGEELlQzwMhhBDVwECxJ2RSx0MFFDwQQghRCTTnQXkoeCCEEKIaGCg450HpJWnwKHgghBCiGmjCpNJQ8EAAALcPxCJq933kZxTC1NEQnaZ5Q9TcRGbe+0f+w+nQCE6aQIOPr64O5KRl/peNy2tv4XlUGqQlUhjbG6DHsvbQt9SttXrUp1sHYnFj90O2DbtM86q0De8eeYxTodc5aQINPr6++in7enXL32Su2+GbFmg10ll5BX+P3Nwfh+u7HiEvoxBmjgboNt0Llu7GMvPe+SsBYSE3OGkCDT6+jejHvi7KL8H5NTGIC3+BwmwxhI100XJwM3gOaFqr9agP568UYPmGl4i6LUZSigQHt4vQN0ivynXOXc7HdyEZuPtIDGsrdcycbIRRnwk5eX7ekYXlP2chOU0CD1cNrFloBh8vrdqsCmkAKHioRV9++SU2bdqEVatWYfLkyfVdnEo9OvEUF1beRJeZrSByN0H0noc4MuEchh3uCR1j2QcJDT11DDvUg33N4/E472cnvsLB0Wfg2scevl+6Q0NXDRn/5UCgKajVutSXhyee4vzKW+g60xsid2Pc3BOLwxPOY+ThoCrbcOShwDcJ5dpwzMlenNcJl5Jxat51OHRrrPTyvw8enEjEuRW34f+/lrB0N0bUnlj88dUFfP5XAHQrbUM1jP7z7Tbkvn9u+S08vZ6KHgtbw8BKFwlXUnB68U3omWmjmZ9VLdam7uXlS+HhqongQUJ8Ojr5nfkfPy1Gr2FJGDfCAL+st8DZi/kYOzUVluYCBHQpDfAP/PUKU0PT8fMSc/h6aWHNliwEDX6B+xebwNy0Af58SFFhH6n2erUkMzMTX3/9NY4ePQo+n4/+/ftjzZo10NOTHfglJCTAzs5O5nu//fYbBgwYAKDiMRkA9u3bh0GDBiml3A3wr187ioqKoKGhobTtHT58GFevXoWV1ft/gIre8wBunzSFax97AECX/7VGwsUk3PvrP7QKdq10PV1T7Urfu7I+BjbtLdF+siebZmCtr7Qyv2+i9jxC80/s4dan9Evd7X/eeHwxCXf/eozWwS6VrldVG5Z/L/7f57BuZQ6DxlWfTTZUN355BPd+dnDvawsA+GhWS/x3IQl3/kyA7+eye1p44EHXtPKz4Oe3MuDWywZNWpsDADw+tcftg/8h+U7mBxc8BHXTRVC36vfqbdqdDbsm6lgeagoAcHHUwMWIQqzenM0GD6s3ZeGLoQYIHlTaG7FhqRmOn8nDjn2vMP1rI+VXopa9jxMmhw4diqSkJJw6dQrFxcUIDg7G2LFjsXfvXpn5ra2tkZSUxEnbvHkzli1bhqCgIE76jh07EBj4Jrg2NDRUWrlV9j4Pfn5+mDhxIiZPngxTU1MEBATgzp07CAoKgp6eHiwsLDB8+HCkp6dXWGfixIkwMDCAqakpZs+eDabcjvX8+XN8/fXX2LNnD9TV1eu6anKRFEuQev8lrH0t2DQenwdrXwsk386odL3ighLs7HEEO4L+wrFvLyAjPpt9j5EySLj4AoY2+vjrq3PY2u0wfhtxEvHhz2q1LvWlsjZs4muOpHe04bYex7A16CiOfHuR04bl5WUUIuFiEtz6yj7jaOgkxVKk3M+Cja85m1bahhZ4UUUbFhWUYFPQcWwK+BuHJ19Cehy3DRt5mCDuXBJepRSAYRg8vZ6KzCe5sGlrUckWVcfVG4Xo1pEboHb308HVyEIAQFERg8jbYk4ePp+Hbh11cOV1nganbM6DIkstuH//PsLCwrB161b4+vqiQ4cOWLduHfbv348XL17IXEcgEEAkEnGWw4cPY+DAgRV6KwwNDTn5tLSUN9ykssEDAOzatQsaGhq4dOkSfvzxR3Tt2hVeXl64ceMGwsLCkJKSgoEDB1ZYR01NDREREVizZg1WrlyJrVu3su9LpVIMHz4c33//Pdzc3Oq6SnIryCoCI2EqdK3rGGshP6NA5jqGNkJ0C/FBz1Ud0X1BWzBSBn8En0ZuSj4AID+zEMX5JYjccR9N2lmiz89+sO/SGMe/u4jnkam1Xqe69qYNNTnpOsZayMuQfZA1stHHRyGt0WtVewQu8AUjZXAg+CxevW7D8u4fTYC6jjqadf0whywKXorBSBjomnD3Q10TTeSly25DY1t9BIa2wier2qHHQh8wUmDvqHBOG3b9wRMm9kJsCvgbq1ofwsGvLsJ/hhesvc1qtT4NQXKaBBZm3GFECzMBcl5JUVAgRXqmBBIJZOZJSS2py6J+sK5cuQJDQ0O0atWKTfP39wefz8e1a9eqtY3IyEhER0dj9OjRFd6bMGECTE1N4ePjg+3bt1c40a0JlR62cHBwwNKlSwEACxYsgJeXFxYtWsS+v337dlhbW+PRo0dwdHQEUNpltGrVKvB4PDg5OSEmJgarVq3CmDFjAABLliyBmpoaJk2aVO1yiMViiMVi9nVOTo4yqldrLD1MYelhyr4WtTDFnv7HcedgHNp81YIN0u39GsFrmBMAwMzJCMm30hHzRxwaeZvL2qxKsfIwhdVbbWjZwhS7+4ch5mA82n3lXiH/3SOP4RzUBGof6JwRRVh5mMDKw4Tzeke/E7j1x3/oMKE5AODmvjgkxWTgkzXtILTUQWJU+us5D1qwaUO9DyqnhldblD82a2pqQlNTU9Ya1ZKcnAxzc+7xUE1NDcbGxkhOfve8FQDYtm0bXFxc0K5dO076vHnz0LVrV+jo6ODkyZP46quvkJubK9dvU1VUuufB29ub/f+tW7cQHh4OPT09dnF2Lh1njY+PZ/O1adOGMxGlbdu2iI2NhUQiQWRkJNasWYOdO3fKnKxSmcWLF8PAwIBdrK2tlVC76tE21ABPwEN+JvfsLj+zEDomlY/Hv02gzoeZsxGyEnPZbfLVeDC2N+DkM7ITIjdZ9pl1Q/amDcWc9PzMwgpn0pURqPNh7mzItuHbnkel4WXCKzT/xF4p5X0faRtpgifgVeipycsQVzmn4W0CdT7MnQyRlZgHACgulODCujvwm+qBpp2tYOZoiJaDmsE5oDGu736k9Do0NCIzAVLSJJy0lDQJhPp8aGvzYWosgEAAmXkszBvoeWcNhy2sra05x+rFixfL/JgffvgBPB6vyuXBgwc1rk5BQQH27t0rs9dh9uzZaN++Pby8vDB9+nRMmzYNy5Ytq/FnllHp4EFX983kotzcXPTq1QvR0dGcJTY2Fp06darW9i5cuIDU1FQ0adIEampqUFNTw5MnTzB16lTY2tpWut6MGTOQnZ3NLomJiTWtWrUJ1AUwdzHCs4gUNo2RMkiMSIGohezLDMuTSqRIj8tiJ/gJ1AUwdzXGywRulJ719BX0LXWUV/j3RFkbJlZow1RYytWG2TInUN756zHMXYxg5miorCK/dwTqfFi4GOJpxJthLUbK4GlEKqyq3YYM0uNy2GBDWiKFtIQBr9xRjsfngZHSdfttWmnh7EXu0OTp8/lo413afhoaPHi30OTkkUoZnL2Yj7beDfRSTWkNFgCJiYmcY/WMGTNkfszUqVNx//79Khd7e3uIRCKkpnKHcktKSpCZmQmRSPTO6vzxxx/Iz8/HiBEj3pnX19cXz5494/Ry10QDDR+Vr2XLljh48CBsbW2hplZ5s5Qfh7p69SocHBwgEAgwfPhw+Pv7c94PCAjA8OHDERwcXOk2a9r1VVOeQ51xOuQqzF2NYeFmjOi9j1BSUALX3qVnuidnX4WeuTbafe0BAIjYfAcidxMYWOtD/KoIUbsf4FVSPtzeOjNuOcIFYT9chlVLczRuZY4nl5Pw+PwL9NvctV7qWNtaDnXEyZAIWLgaQ+RmjKi9j1BcUALX3qUTHE/MvgZdc210+LoFAODq5ruwdDeBobUexK+KcWP3A+Qk5aP5J9wJkeLcYsSeSkSnKR51Xqe61mq4I/6ZfR0WrkawbG6MyD2xKC4oQfM+tgCA47MioGeujU6TSod1Lm+6Byt3Yxg2KW3D67seIScpD+6v21BTTx2NvU3x76oYqGkKILTSxbMbabh37An8pn547ZmbJ0Xc42L2dcLTEkTfEcPYkI8mjdUxc2E6nidLsGtd6XDNuBEGWL89G9PnpyN4kBDhlwrw+5FcHP3Fkt3G5HGGCP4mFd4emvDxLL1UMy+fwahBDfPKqZpebSEUCiEUCt+RGzAzM4OZ2bvn1bRt2xZZWVmIjIxke8LPnj0LqVQKX1/fd66/bds29O7du1qfFR0dDSMjI6X91lDw8NqECROwZcsWDB48GNOmTYOxsTHi4uKwf/9+bN26FQJB6Vjz06dPMWXKFIwbNw5RUVFYt24dVqxYAQAwMTGBiQn3LEldXR0ikQhOTk51XqfqcgxogoKXhbi2Iab05jxOhuj9kx90Xne55ybncc7exDlFODv/OvIyCqEl1ICZixEG7PDnDFM07doYXWa2wo0d93B+WRSMbPTRY1l7WHl9mBPVnAKaoOClGFc23Cm9SZSTIfr+1IkdtshJzgf4b4ayxDlFOD3/BvIzCqEp1IC5ixE+29EVJuWGeh6deMpu/0PnHGCN/JdiXNpwD/nphTBzMsCnP3d404ZJ+ZzhQHFOEU7Mj0J+eiE0heqwcDHC4F1dYNr0zcG915I2OL82BsdnRqAwpwhCS110mNgcHgM+vCGgG7cK0a3/mxn6U0NLrxQbMVAfO9ZYIClVgsTnb4ILuybqOPqrJaaGpGPt1iw0tlTD5hXm7GWaAPBZH32kZ0gQujQTyWkl8HTTxPG9VrAwo58OZXBxcUFgYCDGjBmDjRs3ori4GBMnTsSgQYPYy/yfP3+Obt26Yffu3fDx8WHXjYuLw/nz53H8+PEK2z169ChSUlLQpk0baGlp4dSpU1i0aBG+++47pZWdxyhz+mUD4ufnB09PT6xevZpNi42NxfTp0xEeHg6xWAwbGxsEBgZi5cqV4PF48PPzg5ubG6RSKfbu3QuBQIDx48djwYIFlc5xsLW1xeTJk+W6SVROTg4MDAww7nx/aOi935d6vs8EtXlnFxWhxqc2rKklFtH1XYQGLeeVFEaO/yE7O7taZ/0yt/H6mOrv8C3UBPKfeZdIxDgdu6pGZahMZmYmJk6cyLlJ1Nq1a9nLLstuChUeHg4/Pz92vZkzZ+LXX39FQkIC+Hzu2FxYWBhmzJiBuLg4MAyDZs2aYfz48RgzZkyFvIpS2eBBEbICjtpAwYNyUPBQcxQ81BwFDzWj1OCh6WTFg4f41bUSPDRU1PdECCFENdCDsZSGggdCCCEqQtG7RVLwUB4FD3I4d+5cfReBEEIIqXcUPBBCCFENNGyhNBQ8EEIIUQ1SBgoNQdBNxSqg4IEQQohqYKSliyLrEQ4KHgghhKgGGrZQGpV+tgUhhBBC5Ec9D4QQQlQDzXlQGgoeCCGEqAYatlAaCh4IIYSoBgYKBg9KL0mDR3MeCCGEECIX6nkghBCiGmjYQmkoeCCEEKIapFJAkaftSuk+D+VR8EAIIUQ1UM+D0lDwQAghRDVQ8KA0NGGSEEIIIXKhngdCCCGqgW4SpTQUPBBCCFEJDCMFo8BDrhRZ50NHwQMhhBDVwDCK9SLQnIcKKHgghBCiGhgFhy0oeKiAJkwSQgghRC7U80AIIUQ1SKUAT4H5CzTnoQIKHgghhKgGGrZQGgoeCCGEqARGKgWjQM8DXW1REQUPhBBCVAP1PCgNTZgkhBBCiFyo54EQQohqkDIAj3oelIGCB0IIIaqBYaDQI7kpeKiAggdCCCEqgZEyYBToeWAoeKiA5jwQQghRDYxU8aWWLFy4EO3atYOOjg4MDQ2rVw2GwZw5c2BpaQltbW34+/sjNjaWkyczMxNDhw6FUCiEoaEhRo8ejdzcXKWVm4IHQgghpJ4UFRVhwIABGD9+fLXXWbp0KdauXYuNGzfi2rVr0NXVRUBAAAoLC9k8Q4cOxd27d3Hq1CkcO3YM58+fx9ixY5VWbhq2IIQQohLex2GLuXPnAgB27txZ7bKsXr0as2bNQp8+fQAAu3fvhoWFBf78808MGjQI9+/fR1hYGK5fv45WrVoBANatW4cePXpg+fLlsLKyqnG5KXh4D5XtqEV5xfVckoaNr8jEKMIh4VMb1lSODrVhTeTklrafMn7ASxixQkMQJSg9Fufk5HDSNTU1oampWeNyyePx48dITk6Gv78/m2ZgYABfX19cuXIFgwYNwpUrV2BoaMgGDgDg7+8PPp+Pa9eu4ZNPPqlxOSh4eA+9evUKALAj6Eg9l4QQUlPr6rsAH4hXr17BwMBAoXU1NDQgEolwMfm4wp+vp6cHa2trTlpISAhCQ0MV3qYikpOTAQAWFhacdAsLC/a95ORkmJubc95XU1ODsbExm6emKHh4D1lZWSExMRH6+vrg8Xj1XZwKcnJyYG1tjcTERAiFwvouToNEbVhz1IY11xDakGEYvHr1qkZd7VpaWnj8+DGKiopqVI7yx+PKeh1++OEHLFmypMrt3b9/H87OzgqXp75R8PAe4vP5aNy4cX0X452EQuF7e8BpKKgNa47asObe9zZUtMfhbVpaWtDS0lJCad5t6tSpGDVqVJV57O3tFdq2SCQCAKSkpMDS0pJNT0lJgaenJ5snNTWVs15JSQkyMzPZ9WuKggdCCCFEiczMzGBmZlYr27azs4NIJMKZM2fYYCEnJwfXrl1jr9ho27YtsrKyEBkZCW9vbwDA2bNnIZVK4evrq5Ry0KWahBBCSD15+vQpoqOj8fTpU0gkEkRHRyM6OppzTwZnZ2ccPnwYAMDj8TB58mQsWLAAR44cQUxMDEaMGAErKyv07dsXAODi4oLAwECMGTMGERERuHTpEiZOnIhBgwYp5UoLgHoeiAI0NTUREhJS57OMPyTUhjVHbVhz1Ib1b86cOdi1axf72svLCwAQHh4OPz8/AMDDhw+RnZ3N5pk2bRry8vIwduxYZGVloUOHDggLC+MMy+zZswcTJ05Et27dwOfz0b9/f6xdu1Zp5eYxdN9NQgghhMiBhi0IIYQQIhcKHgghhBAiFwoeCCGEECIXCh4IIYQQIhcKHkitOHToELp37w4TExPweDxER0fXd5EalOLiYkyfPh3u7u7Q1dWFlZUVRowYgRcvXtR30RqU0NBQODs7Q1dXF0ZGRvD398e1a9fqu1gN1pdffgkej4fVq1fXd1FIPaPggXDU5Patb8vLy0OHDh3eeYvWD5Ey2jA/Px9RUVGYPXs2oqKicOjQITx8+BC9e/dWQgnff8raDx0dHfHTTz8hJiYGFy9ehK2tLbp37460tDSlbP99pqw2LHP48GFcvXpVafcJIA0cQ1Ra586dmQkTJjDffPMNY2Jiwvj5+TExMTFMYGAgo6ury5ibmzPDhg1j0tLSKqwzYcIERigUMiYmJsysWbMYqVRaYfuPHz9mADA3b96sw1rVrdpuwzIREREMAObJkyd1Ua06VVdtmJ2dzQBgTp8+XRfVqlO12YbPnj1jGjVqxNy5c4exsbFhVq1aVce1I+8b6nkg2LVrFzQ0NHDp0iX8+OOP6Nq1K7y8vHDjxg2EhYUhJSUFAwcOrLCOmpoaIiIisGbNGqxcuRJbt26tpxrUv7pow+zsbPB4PBgaGtZybepHbbdhUVERNm/eDAMDA3h4eNRFlepcbbShVCrF8OHD8f3338PNza2uq0TeV/UdvZD61blzZ8bLy4t9PX/+fKZ79+6cPImJiQwA5uHDh+w6Li4unLOT6dOnMy4uLhW2ryo9D7XZhgzDMAUFBUzLli2ZIUOG1EIN6l9ttuHRo0cZXV1dhsfjMVZWVkxEREQt1qT+1FYbLlq0iPnoo4/YPNTzQBiGeh4IwD44BQBu3bqF8PBw6OnpsUvZY2Pj4+PZfG3atOE8nrZt27aIjY2FRCKpu4K/R2qzDYuLizFw4EAwDIMNGzbUck3qT221YZcuXRAdHY3Lly8jMDAQAwcOrPDEwQ+FstswMjISa9aswc6dOys8jpqoNnq2BYGuri77/9zcXPTq1UvmRMe3H/9KuGqrDcsChydPnuDs2bPv9WOTa6q22lBXVxfNmjVDs2bN0KZNGzg4OGDbtm2YMWNGjcv8vlF2G164cAGpqalo0qQJmyaRSDB16lSsXr0aCQkJNS4zaZgoeCAcLVu2xMGDB2Fraws1tcp3j/KXu129ehUODg4QCAS1XcT3nrLasCxwiI2NRXh4OExMTGq13O+T2twPpVIpxGKx0sr6vlJGGw4fPhz+/v6c9wMCAjB8+HAEBwfXSrlJw0DDFoRjwoQJyMzMxODBg3H9+nXEx8fjxIkTCA4O5nQFP336FFOmTMHDhw+xb98+rFu3Dt988w37fmZmJqKjo3Hv3j0ApU+Fi46ORnJycp3Xqa4pow2Li4vx6aef4saNG9izZw8kEgmSk5ORnJys9Evw3kfKaMO8vDzMnDkTV69exZMnTxAZGYnPP/8cz58/x4ABA+qranVGGW1oYmKC5s2bcxZ1dXWIRCI4OTnVV9XIe4B6HgiHlZUVLl26hOnTp6N79+4Qi8WwsbFBYGAg+Pw3seaIESNQUFAAHx8fCAQCfPPNNxg7diz7/pEjRzhnJoMGDQIAhISEIDQ0tM7qUx+U0YbPnz/HkSNHAACenp6c7b/9qN4PlTLaUCAQ4MGDB9i1axfS09NhYmKC1q1b48KFCypx1YCyvsuEyEKP5CZy8/Pzg6enJ91lrgaoDWuO2rDmqA2JomjYghBCCCFyoeCBEEIIIXKhYQtCCCGEyIV6HgghhBAiFwoeCCGEECIXCh4IIYQQIhcKHgghhBAiFwoeCCGEECIXCh4IIYQQIhcKHgghhBAiFwoeCCGEECIXCh4IIYQQIpf/A8p614j+tvLdAAAAAElFTkSuQmCC",
      "text/plain": [
       "<Figure size 500x400 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAiIAAAGGCAYAAABG0epPAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjkuMCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy80BEi2AAAACXBIWXMAAA9hAAAPYQGoP6dpAAB+y0lEQVR4nO3dd3hUxfrA8e/upvfeIKRAqALBIMVLJ5KggiCKFCkRgYtwlQsq4k+aICgdsaACilcQVEAFFaRaqAKGJiC9pzfSk93z+yPZhSWFZJPsAr6f5zkPZHbm7Mzs2bPvzsw5q1IURUEIIYQQwgLUlq6AEEIIIf65JBARQgghhMVIICKEEEIIi5FARAghhBAWI4GIEEIIISxGAhEhhBBCWIwEIkIIIYSwGAlEhBBCCGExEogIIYQQwmIkEBFVMnToUIKDg6t1n5999hkqlYoLFy5U637F3SM4OJihQ4ca/t65cycqlYqdO3darE53qwsXLqBSqZg7d+4d806dOhWVSmWU9k/u63u9rS+88AKPPPKI2Z6vKufz1157jdatW5tUttKByNGjR3nqqacICgrCzs6OWrVq8cgjj7B48WKTKiD+uWbOnMm3335r6WqIf4Ddu3czdepU0tLSzP7cq1atYuHChWZ/3sq4F+p4t8rOzmbq1KnVHuycP3+epUuX8vrrrxvSrl27xtSpU4mNja3W56oOY8eO5fDhw3z//feVLlupQGT37t20bNmSw4cPM3z4cN577z2ef/551Go1ixYtqvSTi3+2sgKRQYMGkZOTQ1BQkPkrJSyiQ4cO5OTk0KFDhxrZ/+7du5k2bdp9H4i88cYb5OTklJuntL6WQMR02dnZTJs2rdoDkUWLFhESEkLnzp0NadeuXWPatGk1Foh88sknnDp1yqSyfn5+PPHEExUaubudVWUyv/XWW7i6uvLHH3/g5uZm9FhCQkKln9xcCgsL0el02NjYWLoqd4Xy+iMrKwtHR0cL1OomjUaDRqOxaB0E5ObmYmNjg1pd8zO4arUaOzu7Gn+e+52VlRVWVuWf1i3Z13IurpiCggJWrlzJv//97yrtJzs7GwcHhwrnt7a2rtLz9e3bl6effppz584RGhpa4XKVOsOcPXuWJk2alAhCAHx8fIz+VqlUjBkzhpUrV9KgQQPs7OyIiIjg119/LVH26tWrPPfcc/j6+mJra0uTJk1Yvny5UZ78/HwmT55MREQErq6uODo60r59e3bs2GGU79b51IULF1K3bl1sbW3566+/DPOnf//9N88++yyurq54e3szadIkFEXh8uXLPPHEE7i4uODn58e8efOqXIePP/7YUIeHHnqIP/74o0J9nZaWxn//+1+Cg4OxtbWldu3aDB48mKSkJEOehIQEhg0bhq+vL3Z2djRv3pwVK1ZUuj/++usvBgwYgLu7O+3atTOU/eKLL4iIiMDe3h4PDw/69evH5cuX71j3uXPn8vDDD+Pp6Ym9vT0RERF88803RnlUKhVZWVmsWLEClUqFSqUyzGOXtUbkgw8+oEmTJtja2hIQEMDo0aNLfMPt1KkTDzzwAH/99RedO3fGwcGBWrVqMXv27Ar0Onz66ad06dIFHx8fbG1tady4MR9++GGJfMHBwTz++OPs3LmTli1bYm9vT9OmTQ3fitatW0fTpk0Nx/2ff/5ZYh/bt2+nffv2ODo64ubmxhNPPMGJEyeM8uhfozNnzjB06FDc3NxwdXUlJiaG7Oxso7w5OTm8+OKLeHl54ezsTM+ePbl69SoqlYqpU6eW2279XPrq1at54403qFWrFg4ODmRkZACwb98+oqOjcXV1xcHBgY4dO7Jr165S63ry5En69u2Li4sLnp6evPTSS+Tm5lbo+W//Vrlv3z4effRR3N3dcXR0pFmzZkajr0eOHGHo0KGEhoZiZ2eHn58fzz33HMnJyUb1euWVVwAICQkxHG+3Hl+mHus3btxg7Nixhvepj48PjzzyCIcOHQKKjscffviBixcvGp5XPwdf0fPJrRYsWEBQUBD29vZ07NiRY8eOGT1e2hqR293e12XVMTMzE0dHR1566aUS+7hy5QoajYZZs2aV+TzlnXsATp48yVNPPYWHhwd2dna0bNmyxLB+SkoKL7/8Mk2bNsXJyQkXFxe6d+/O4cOHS61Tr169cHR0xMfHh//+97/k5eWVWrevv/7a8Hp7eXnx7LPPcvXqVaM8nTp1olOnTiXK3rqO4sKFC3h7ewMwbdo0Q//p329xcXHExMRQu3ZtbG1t8ff354knnrjj+rfff/+dpKQkIiMjDWk7d+7koYceAiAmJsbwXJ999pmhvg888AAHDx6kQ4cOODg4GKZ1vvvuOx577DECAgKwtbWlbt26TJ8+Ha1WW2bb9O2rzGeZvr7fffddue27XaVGRIKCgtizZw/Hjh3jgQceuGP+X375hTVr1vDiiy9ia2vLBx98QHR0NPv37zeUj4+Pp02bNobAxdvbm59++olhw4aRkZHB2LFjAcjIyGDp0qX079+f4cOHc+PGDZYtW0ZUVBT79+8nPDzc6Lk//fRTcnNzGTFiBLa2tnh4eBgee+aZZ2jUqBFvv/02P/zwAzNmzMDDw4OPPvqILl268M4777By5UpefvllHnroIcMQZmXrsGrVKm7cuMHIkSNRqVTMnj2bJ598knPnzpUbeWZmZtK+fXtOnDjBc889x4MPPkhSUhLff/89V65cwcvLi5ycHDp16sSZM2cYM2YMISEhfP311wwdOpS0tLQSJ4/y+uPpp58mLCyMmTNnoigKUDT6NWnSJPr27cvzzz9PYmIiixcvpkOHDvz555+lBqN6ixYtomfPngwcOJD8/HxWr17N008/zcaNG3nssccA+N///sfzzz9Pq1atGDFiBAB169Ytc59Tp05l2rRpREZGMmrUKE6dOsWHH37IH3/8wa5du4z6MzU1lejoaJ588kn69u3LN998w4QJE2jatCndu3cv8zkAPvzwQ5o0aULPnj2xsrJiw4YNvPDCC+h0OkaPHm2U98yZMwwYMICRI0fy7LPPMnfuXHr06MGSJUt4/fXXeeGFFwCYNWsWffv25dSpU4bRha1bt9K9e3dCQ0OZOnUqOTk5LF68mH/9618cOnSoxIKxvn37EhISwqxZszh06BBLly7Fx8eHd955x5Bn6NChfPXVVwwaNIg2bdrwyy+/GPq7oqZPn46NjQ0vv/wyeXl52NjYsH37drp3705ERARTpkxBrVYbArbffvuNVq1alahrcHAws2bNYu/evbz77rukpqby+eefV6ouW7Zs4fHHH8ff35+XXnoJPz8/Tpw4wcaNGw3H95YtWzh37hwxMTH4+flx/PhxPv74Y44fP87evXtRqVQ8+eST/P3333z55ZcsWLAALy8vAMMHSFWO9X//+9988803jBkzhsaNG5OcnMzvv//OiRMnePDBB/m///s/0tPTuXLlCgsWLADAyckJqPz55PPPP+fGjRuMHj2a3NxcFi1aRJcuXTh69Ci+vr6V6ttblVVHJycnevfuzZo1a5g/f77RKOWXX36JoigMHDjwjvsv7dxz/Phx/vWvf1GrVi1ee+01HB0d+eqrr+jVqxdr166ld+/eAJw7d45vv/2Wp59+mpCQEOLj4/noo4/o2LEjf/31FwEBAUBREN61a1cuXbrEiy++SEBAAP/73//Yvn17ifp89tlnxMTE8NBDDzFr1izi4+NZtGgRu3btuuPrfTtvb28+/PBDRo0aRe/evXnyyScBaNasGQB9+vTh+PHj/Oc//yE4OJiEhAS2bNnCpUuXyl0Uunv3blQqFS1atDCkNWrUiDfffJPJkyczYsQI2rdvD8DDDz9syJOcnEz37t3p168fzz77rOG4+Oyzz3BycmLcuHE4OTmxfft2Jk+eTEZGBnPmzLljOyv6Webq6krdunXZtWsX//3vfyvcjyiV8PPPPysajUbRaDRK27ZtlVdffVXZvHmzkp+fXyIvoADKgQMHDGkXL15U7OzslN69exvShg0bpvj7+ytJSUlG5fv166e4uroq2dnZiqIoSmFhoZKXl2eUJzU1VfH19VWee+45Q9r58+cVQHFxcVESEhKM8k+ZMkUBlBEjRhjSCgsLldq1aysqlUp5++23jfZtb2+vDBkyxChvZerg6emppKSkGNK/++47BVA2bNhQor9uNXnyZAVQ1q1bV+IxnU6nKIqiLFy4UAGUL774wvBYfn6+0rZtW8XJyUnJyMiocH/079/fKP3ChQuKRqNR3nrrLaP0o0ePKlZWVkbpQ4YMUYKCgozy6V+zW+v1wAMPKF26dDFKd3R0NOpfvU8//VQBlPPnzyuKoigJCQmKjY2N0q1bN0Wr1RryvffeewqgLF++3JDWsWNHBVA+//xzQ1peXp7i5+en9OnTp8Rz3e72uiuKokRFRSmhoaFGaUFBQQqg7N6925C2efNmBVDs7e2VixcvGtI/+ugjBVB27NhhSAsPD1d8fHyU5ORkQ9rhw4cVtVqtDB482JCmf41uPb4URVF69+6teHp6Gv4+ePCgAihjx441yjd06FAFUKZMmVJuu3fs2KEASmhoqFEf6HQ6JSwsTImKijIce4pS1E8hISHKI488UqKuPXv2NNr3Cy+8oADK4cOHDWlBQUFGr73++fV9VFhYqISEhChBQUFKamqq0f5ur8ftvvzySwVQfv31V0PanDlzjI4pvcoc66VxdXVVRo8eXW6exx57rMR7RFEqfz6xt7dXrly5Ykjft2+fAij//e9/DWn61+BWd+rr8uqoP6Z/+ukno/RmzZopHTt2LKfV5Z97unbtqjRt2lTJzc01pOl0OuXhhx9WwsLCDGm5ublG73n9fm1tbZU333zTkKY/H3711VeGtKysLKVevXpGbc3Pz1d8fHyUBx54QMnJyTHk3bhxowIokydPNqR17Nix1Dbefs5LTEws9T2WmpqqAMqcOXPK7qQyPPvss0bvb70//vhDAZRPP/20xGP6c9+SJUtKPFba+2TkyJGKg4OD0Wtwe9tM+Szr1q2b0qhRozs10UilpmYeeeQR9uzZQ8+ePTl8+DCzZ88mKiqKWrVqlbpStm3btkRERBj+rlOnDk888QSbN29Gq9WiKApr166lR48eKIpCUlKSYYuKiiI9Pd0wxKnRaAzzijqdjpSUFAoLC2nZsqUhz6369Olj+MZzu+eff97wf41GQ8uWLVEUhWHDhhnS3dzcaNCgAefOnTPKW5k6PPPMM7i7uxv+1kewt+6zNGvXrqV58+aGbwW30g+7/vjjj/j5+dG/f3/DY9bW1rz44otkZmbyyy+/GJUrrz9un4dct24dOp2Ovn37Gr0mfn5+hIWFlTt0DGBvb2/4f2pqKunp6bRv377UPqqIrVu3kp+fz9ixY43WKwwfPhwXFxd++OEHo/xOTk48++yzhr9tbGxo1arVHfv99rqnp6eTlJREx44dOXfuHOnp6UZ5GzduTNu2bQ1/6y9d69KlC3Xq1CmRrn/+69evExsby9ChQ41Gppo1a8YjjzzCjz/+WKJet79G7du3Jzk52TB1smnTJgDDKIzef/7znzu2+VZDhgwx6oPY2FhOnz7NgAEDSE5ONhwLWVlZdO3alV9//RWdTme0j9tHjvR1KK1dZfnzzz85f/48Y8eOLfEN9daph1vrmpubS1JSEm3atAGo0PFW1WPdzc2Nffv2ce3atQq3Ta+y55NevXpRq1Ytw9+tWrWidevWlerXyoqMjCQgIICVK1ca0o4dO8aRI0eM3mPluf3ck5KSwvbt2+nbty83btww9HlycjJRUVGcPn3aME1ia2treM9rtVqSk5NxcnKiQYMGRn30448/4u/vz1NPPWVIc3BwMIy26h04cICEhAReeOEFo3Uyjz32GA0bNixxLqkKe3t7bGxs2LlzJ6mpqZUqm5ycbPTZUVG2trbExMSUWhc9fZ+3b9+e7OxsTp48ecf9VuazzN3d3WgJQUVUamoG4KGHHmLdunXk5+dz+PBh1q9fz4IFC3jqqaeIjY2lcePGhrxhYWElytevX5/s7GwSExNRq9WkpaXx8ccf8/HHH5f6fLcugl2xYgXz5s3j5MmTFBQUGNJDQkJKlCstTe/WDwkoGk6ys7MzDNnemn7rXHNl63D78+hfyDsdlGfPnqVPnz7l5rl48SJhYWElFhI2atTI8PityuuP2x87ffo0iqKU+vrBnRc0bdy4kRkzZhAbG2s0R3unueuy6NvSoEEDo3QbGxtCQ0NLtLV27dolnsvd3Z0jR47c8bl27drFlClT2LNnT4k1GOnp6bi6uhr+Lu04AggMDCw1Xf+6l9UeKHr9Nm/eXGLRcHnHkouLCxcvXkStVpd4LevVq3eHFhsr7ViAogClLOnp6UYnqduPm7p166JWqyt1X5izZ88C3HEKOCUlhWnTprF69eoSC+ZvDxxLU9Vjffbs2QwZMoTAwEAiIiJ49NFHGTx4cIUX6lXmfFLW+fSrr76q0HOZQq1WM3DgQD788EPDwseVK1diZ2fH008/XaF93N6WM2fOoCgKkyZNYtKkSaWWSUhIoFatWuh0OhYtWsQHH3zA+fPnjdY0eHp6Gv5/8eJF6tWrV+J9f/t7rLz3XsOGDfn9998r1KaKsLW15Z133mH8+PH4+vrSpk0bHn/8cQYPHoyfn98dyyvF0+SVUatWrVIXAh8/fpw33niD7du3G7686FXkfVKZzzJFUSp9rq90IKJnY2PDQw89xEMPPUT9+vWJiYnh66+/ZsqUKRXeh/6b1LPPPlvmiU4/1/bFF18wdOhQevXqxSuvvIKPj49hsZT+pHWrWyPA25V2RUZZV2ncejBUtg4V2ae5lNcftz+m0+lQqVT89NNPpbZBP8ddmt9++42ePXvSoUMHPvjgA/z9/bG2tubTTz9l1apVpjegEkzt97Nnz9K1a1caNmzI/PnzCQwMxMbGhh9//JEFCxaU+OZf1vPUxOturmOptGMBYM6cOSXWLOiVdzyA6QFoRfTt25fdu3fzyiuvEB4ejpOTEzqdjujo6BKvV2mqcqzrn799+/asX7+en3/+mTlz5vDOO++wbt26O65Hquz5xFIGDx7MnDlz+Pbbb+nfvz+rVq3i8ccfNwrKy1PWMfXyyy8TFRVVahl9AD1z5kwmTZrEc889x/Tp0/Hw8ECtVjN27NgKvb5VoVKpSn1/3b7Aszxjx46lR48efPvtt2zevJlJkyYxa9Ystm/fbrT+43aenp6VHkWB0s/zaWlpdOzYERcXF958803q1q2LnZ0dhw4dYsKECRXqx8qcf1JTU0t8qb8TkwORW7Vs2RIoGnK+lf7b1K3+/vtvHBwcDEN1zs7OaLVao9XBpfnmm28IDQ1l3bp1Rie2ygQ+VWWuOtStW7fEavjbBQUFceTIEXQ6ndGoiH6YrSr34Khbty6KohASEkL9+vUrVXbt2rXY2dmxefNmbG1tDemffvppibwV/YDSt+XUqVNG3zTz8/M5f/78HY+ditqwYQN5eXl8//33Rt8A7jQ8X1m3tud2J0+exMvLq9KXUAcFBaHT6Th//rzRN+czZ85Uqa76BcQuLi4V7ufTp08bfQs+c+YMOp2uUnds1D/vsWPHynze1NRUtm3bxrRp05g8ebLR89+urGOtKse6nr+/Py+88AIvvPACCQkJPPjgg7z11luGQKSs567s+aSs82l13Nm4vPfiAw88QIsWLVi5ciW1a9fm0qVLVbqBpf49bG1tXaHzfufOnVm2bJlRelpamtGHXVBQEMeOHSvxbfz299it770uXboYPXbq1Cmj86a7u3upUw+3j8De6TxWt25dxo8fz/jx4zl9+jTh4eHMmzePL774oswyDRs2ZOXKlSVGYU0J6nfu3ElycjLr1q0zunfM+fPnK72vijh//jzNmzevVJlKrRHZsWNHqRGQfo7y9uGuPXv2GM3jXb58me+++45u3boZ7hXRp08f1q5dW+oHb2JiouH/+ojs1ufft28fe/bsqUwTqsRcdejTp49h2ut2+ud+9NFHiYuLY82aNYbHCgsLWbx4MU5OTnTs2NHk53/yySfRaDRMmzatxOutKEqJ6apbaTQaVCqV0beGCxculHrjMkdHxwrdYCoyMhIbGxveffddo/osW7aM9PT0Sl8ZUpbSXt/09PRSg6iq8Pf3Jzw8nBUrVhi1/9ixY/z88888+uijld6n/pvlBx98YJRe1TseR0REULduXebOnUtmZmaJx299j+q9//77pdbhTiMEt3rwwQcJCQlh4cKFJY4R/etT2usFlHpjLn1gd/u+qnKsa7XaEsPaPj4+BAQEGE1JOjo6ljr8Xdnzybfffmt0ien+/fvZt29fpfq1LGXVUW/QoEH8/PPPLFy4EE9Pzyo9p4+PD506deKjjz4q8eUVSp73b39dvv766xKX2j766KNcu3bN6DYB2dnZJab8W7ZsiY+PD0uWLDF6jX766SdOnDhhdC6pW7cuJ0+eNKrP4cOHS1y2rr9Px+3HVnZ2donL1uvWrYuzs3OZlxXrtW3bFkVROHjwoFF6WcdxeUo7zvLz80ucK6pDeno6Z8+eNbqSpyIqNSLyn//8h+zsbHr37k3Dhg3Jz89n9+7drFmzhuDg4BKLZB544AGioqKMLt+Fouut9d5++2127NhB69atGT58OI0bNyYlJYVDhw6xdetWUlJSAHj88cdZt24dvXv35rHHHuP8+fMsWbKExo0bl3qCrAnmqsMrr7zCN998w9NPP81zzz1HREQEKSkpfP/99yxZsoTmzZszYsQIPvroI4YOHcrBgwcJDg7mm2++YdeuXSxcuBBnZ2eTn79u3brMmDGDiRMncuHCBXr16oWzszPnz59n/fr1jBgxgpdffrnUso899hjz588nOjqaAQMGkJCQwPvvv0+9evVKrNGIiIhg69atzJ8/n4CAAEJCQkr9rQJvb28mTpzItGnTiI6OpmfPnpw6dYoPPviAhx56qMKL5u6kW7du2NjY0KNHD0aOHElmZiaffPIJPj4+pZ4wq2LOnDl0796dtm3bMmzYMMPlu66urne850dpIiIi6NOnDwsXLiQ5Odlw+e7ff/8NmD49olarWbp0Kd27d6dJkybExMRQq1Ytrl69yo4dO3BxcWHDhg1GZc6fP0/Pnj2Jjo5mz549fPHFFwwYMKBS35LUajUffvghPXr0IDw8nJiYGPz9/Tl58iTHjx9n8+bNuLi40KFDB2bPnk1BQQG1atXi559/LvWbnn7R/P/93//Rr18/rK2t6dGjR5WO9Rs3blC7dm2eeuopmjdvjpOTE1u3buWPP/4wugdRREQEa9asYdy4cTz00EM4OTnRo0ePSp9P6tWrR7t27Rg1ahR5eXmGoODVV1+tcL+Wpaw66g0YMIBXX32V9evXM2rUqCrf+Or999+nXbt2NG3alOHDhxMaGkp8fDx79uzhypUrhvuEPP7447z55pvExMTw8MMPc/ToUVauXFliDY7+Tt+DBw/m4MGD+Pv787///a/Ezbysra155513iImJoWPHjvTv399w+W5wcLDRJafPPfcc8+fPJyoqimHDhpGQkMCSJUto0qSJ0ToLe3t7GjduzJo1a6hfvz4eHh488MADFBYW0rVrV/r27Uvjxo2xsrJi/fr1xMfH069fv3L7p127dnh6erJ161ajkZu6devi5ubGkiVLcHZ2xtHRkdatW5e7BvDhhx/G3d2dIUOG8OKLL6JSqfjf//5XI0sEtm7diqIoPPHEE5UrWJlLbH766SflueeeUxo2bKg4OTkpNjY2Sr169ZT//Oc/Snx8vFFeQBk9erTyxRdfKGFhYYqtra3SokULo0vG9OLj45XRo0crgYGBirW1teLn56d07dpV+fjjjw15dDqdMnPmTCUoKMiwr40bN5Z5uVFpl0zpL21LTEw0Sh8yZIji6OhYIn/Hjh2VJk2aVGsdqMCllIqiKMnJycqYMWOUWrVqKTY2Nkrt2rWVIUOGGF3mHB8fr8TExCheXl6KjY2N0rRp0xKXdZnSH3pr165V2rVrpzg6OiqOjo5Kw4YNldGjRyunTp0y5Cnt8t1ly5YZXvOGDRsqn376aamXFZ48eVLp0KGDYm9vrwCGSwxvv3xX77333lMaNmyoWFtbK76+vsqoUaNKXNp5+2tWXj1L8/333yvNmjVT7OzslODgYOWdd95Rli9fXqI+QUFBymOPPVaivP64v1VZr8HWrVuVf/3rX4q9vb3i4uKi9OjRQ/nrr7+M8pT1GpXWR1lZWcro0aMVDw8PxcnJSenVq5dy6tQpBTC6NL00+ks6v/7661If//PPP5Unn3xS8fT0VGxtbZWgoCClb9++yrZt20rU9a+//lKeeuopxdnZWXF3d1fGjBljdKmkolTsklJFUZTff/9deeSRRxRnZ2fF0dFRadasmbJ48WLD41euXFF69+6tuLm5Ka6ursrTTz+tXLt2rdT32fTp05VatWoparW6RN9V5Fi/XV5envLKK68ozZs3N9SvefPmygcffGCULzMzUxkwYIDi5uamAIbj0JTzybx585TAwEDF1tZWad++vdEl0be+BpXt67LqeKtHH320xCXr5Snv3KMoinL27Fll8ODBip+fn2Jtba3UqlVLefzxx5VvvvnGkCc3N1cZP3684u/vr9jb2yv/+te/lD179pR6ae3FixeVnj17Kg4ODoqXl5fy0ksvKZs2bSr1uFqzZo3SokULxdbWVvHw8FAGDhxodGm03hdffKGEhoYqNjY2Snh4uLJ58+ZSzyW7d+9WIiIiFBsbG8Oxl5SUpIwePVpp2LCh4ujoqLi6uiqtW7c2usS4PC+++KJSr169Eunfffed0rhxY8XKysroUt6yzn2Koii7du1S2rRpo9jb2ysBAQGGW2/c3jdV/Sx75plnlHbt2lWofbdSFe+w2qlUKkaPHs17771XE7sXQlRAbGwsLVq04IsvvqjQzaeqQn/TucTExEovVhN3v969e3P06NEqrzsSFXPu3DkaNmzITz/9RNeuXS1dnTuKi4sjJCSE1atXV3pEpOZ/REIIYRal/djZwoULUavVNfZjcuKf4fr16/zwww8MGjTI0lX5xwgNDWXYsGG8/fbblq5KhSxcuJCmTZtWflqGarpqRghhebNnz+bgwYN07twZKysrfvrpJ3766SdGjBhR4t4mQlTE+fPn2bVrF0uXLsXa2pqRI0daukr/KKX9ztXdqioBkwQiQtwnHn74YbZs2cL06dPJzMykTp06TJ06lf/7v/+zdNXEPeqXX34hJiaGOnXqsGLFigrdiEuIyqqxNSJCCCGEKN2vv/7KnDlzOHjwINevX2f9+vX06tWr3DI7d+5k3LhxHD9+nMDAQN544w3Dr5brvf/++8yZM4e4uDiaN2/O4sWLS/ww5d1G1ogIIYQQZpaVlUXz5s1L3HenLOfPn+exxx6jc+fOxMbGMnbsWJ5//nk2b95syKO/BHvKlCkcOnSI5s2bExUVVeLnD+42MiIihBBCWJBKpbrjiMiECRP44YcfjG7+2a9fP9LS0gw/etm6dWseeughw9WqOp2OwMBA/vOf//Daa6/VaBuqQtaI3KN0Oh3Xrl3D2dm5Rn/LQwgh7naKonDjxg0CAgJK/BBoZeTm5pKfn1+letx+Pra1tTX6uQtT7dmzp8Qt8aOiohg7dixQdLfUgwcPMnHiRMPjarWayMhIs96B3BQSiNyjrl27JldCCCHELS5fvkzt2rVNKpubm0tIkBNxCRX/UbvbOTk5lbgr7pQpU0y6W/Lt4uLi8PX1NUrz9fUlIyODnJwcUlNT0Wq1pebR/wbZ3UoCkXuU/hbuFw8F4+IkS31M1flIL0tX4Z5XUFj6L3OKitMqMqpZFdrsPE4PW1Sln7bIz88nLkHLxYPBuDhX/pyacUNHUMQFLl++jIuLiyG9OkZD7ncSiNyj9MN/Lk5qk940oojGQU4SVaWTQKTqJBCpFtUxTe3krMLJufL70VF8TnZxMQpEqoufnx/x8fFGafHx8bi4uGBvb2/4IdnS8tztl13LJ5gQQghRTKvoTN5qUtu2bdm2bZtR2pYtW2jbti0ANjY2REREGOXR6XRs27bNkOduJSMiQgghRDEdCjoqfzFpZctkZmYa/W7P+fPniY2NxcPDgzp16jBx4kSuXr3K559/DsC///1v3nvvPV599VWee+45tm/fzldffcUPP/xg2Me4ceMYMmQILVu2pFWrVixcuJCsrCxiYmIq3R5zkkBECCGEKKZDhyljG5UtdeDAATp37mz4e9y4cQAMGTKEzz77jOvXr3Pp0iXD4yEhIfzwww/897//ZdGiRdSuXZulS5cSFRVlyPPMM8+QmJjI5MmTiYuLIzw8nE2bNpVYwHq3kfuI3KMyMjJwdXUl9e9QWSNSBa3+fNrSVbjnyWLVqpPFqlWjzc7jZP/ZpKenm7w+Q39OvXaqtsmLVQMaXKlSHf6pZERECCGEKKZVFLQmfD83pYwoIoGIEEIIUcxca0TETRKICCGEEMV0KGglEDErCUSEEEKIYjIiYn6yylEIIYQQFiMjIkIIIUQxWaxqfhKICCGEEMV0xZsp5YRpJBARQgghimlNXKxqShlRRNaICCGEEMJiZERECCGEKKZVijZTygnTSCAihBBCFJM1IuYngYgQQghRTIcKLZX/7R+dCWVEEQlEhBBCiGI6pWgzpZwwjSxWFUIIIYTFyIiIEEIIUUxr4tSMKWVEEQlEhBBCiGISiJifBCJCCCFEMZ2iQqeYsFjVhDKiiAQiQgghRDEZETE/WawqhBBCCIuREREhhBCimBY1WhO+o2troC7/FBKICCGEEMUUE9eIKLJGxGQSiAghhBDFZI2I+UkgIoQQQhTTKmq0iglTM3JnVZPJYlUhhBBCWIyMiAghhBDFdKjQmfAdXYcMiZhKAhEhhBCimKwRMT8JRIQQQohipq8RkRERU0kgIoQQQhQrmpox4RbvMiJiMglERIX8uieHuR+mcuhIHtfjtaxd7kev7k7lltm5O5uXpyRz/O88AgOseX2sO0OfcTHK88Gnacz9II24RC3NG9uw6C1vWrWwq8mmWFTChoPEr91HQWom9iE+1BnVDccGAaXmTd11irg1u8m7nopSqMO2lju+vVvh2bUpAEqhlquf/0r6H2fJj0tD42iLc3gwtWI6YePpbMZWmVfSDwdIXL+HwtRM7EJ8qTUiCof6tUrNm7z5EKk7jpJ3MREA+3p++A3qbJS/IDWTuBXbuRF7Dm1mLo5N6lBrZDS2AR5maY8lJP/wB8nfFvdhsC9+I6LL7MOUnw+RvuMIufo+rOuPz219qM3JJ+HzbWTsO4X2Rg42Pm54PN4Kj+4RZmmPuLfJVTMWsG7dOrp164anpycqlYrY2FhLV+mOsrJ1NG9sy+KZ3hXKf/5SAT2evU6nf9lzaEsdXhruyojxCWzekWXIs+a7G4yfmsSk8R4c2BxIs8a2dO9/jYSkwppqhkWl/PIXVz7Zhv+AdjRa/BwOob6cnrSGgrSsUvNbOdvh1+9hGswbTOMPhuEZ2YwLC34g/eA5AHR5BWSficO//79otDiG0DeeJPdKMmenfWPOZplV2m/Hub5sC7792hO24Hnsg305P+VLCsvow6xjF3Hr0ITQt56l7pyhWHu5cG7KKgqSMwBQFIWLM78mPy6V4P/rS9jC4dj4uHJu0hfocvPN2TSzSf/tOPHLt+D9TAdC5w/HLsSXi1NXldmH2Ucv4tr+AYJnDCJ0dgzWXi5cnLrS0IcA8ct/JvPQWWr/txf13huFR8/WXP/4JzL2nTJXs6qNrvjOqpXdTFngCvD+++8THByMnZ0drVu3Zv/+/WXm7dSpEyqVqsT22GOPGfIMHTq0xOPR0dEm1c1cJBCphPz86jkxZWVl0a5dO955551q2Z85dO/qyPTXPOn9aPmjIHoffZ5OSB1r5k71olF9G0Y/50afx51Y+HG6Ic/Cj9J4fqArMf1caNzAhg9ne+Ngr+LTL2/UVDMsKn79fryim+PVrRn2dbyoMyYata0VyT8fKTW/c7Mg3B9ugH0dL2z93fHt9RD2IT5kHr8MgMbRjvoz++PRoRF2tT1xaliLOi90I/tMHPkJ6aXu816X+N0+PLq1wCMyHLs63tR64VFUttakbI0tNX+d8b3xerQl9qF+2NX2ovaYx0GnkHn4AgD511LIPnWVWi88ikNYAHa1Pak16lF0+YWk/nrcfA0zo+Tv9uLerQXuxX3oP+ox1LbWpJbRh7XH98ajuA9ta3sRUNyHWYfPG/Jkn7yCa5dmODYNxsbXDY+oB7EL8SXn9DUztar66NeImLJV1po1axg3bhxTpkzh0KFDNG/enKioKBISEkrNv27dOq5fv27Yjh07hkaj4emnnzbKFx0dbZTvyy+/NKkvzEUCkXJ06tSJMWPGMHbsWLy8vIiKiuLYsWN0794dJycnfH19GTRoEElJSSXKjBkzBldXV7y8vJg0aRLKLQuZBg0axOTJk4mMjLREs8xi74Fcura3N0rr1smBvQdzAcjPVzh4JM8oj1qtomt7B/YU57mf6Aq0ZJ+JwyU8xJCmUqtwDg8m8+TVO5ZXFIWM2AvkXUnB+YE6ZebTZuWBCjRO99/0lq5AS86Z6zjd3ofNg8muQB9C0SiSotWhcbY37BNAZa0x2qfaWkP2X5ersfZ3B12Blpyz13FsbtyHjs1DyDl1pWL7uK0PARwa1ubG/r8pSM5AURSyjlwg/2oKTi1Cq70NNU1XPLphylZZ8+fPZ/jw4cTExNC4cWOWLFmCg4MDy5cvLzW/h4cHfn5+hm3Lli04ODiUCERsbW2N8rm7u5vUF+YigcgdrFixAhsbG3bt2sXbb79Nly5daNGiBQcOHGDTpk3Ex8fTt2/fEmWsrKzYv38/ixYtYv78+SxdutRCLbCMuEQtvt4aozRfbw0ZN3Tk5OhIStGi1VJqnviE+29qpjAjG3QKVu4ORunWbo4UpGSWWU6blcufT87lUM/ZnJnyFYGjHsHlwZBS8+ryC7n66U48OjZG42BbrfW/G2j1fejmaJRu5eZEQVrZfXiruBXbsfZwwqn4g9iutifW3i7Efb6DwswcdAVaEtbupiDpBgWpFdvnveRmHxqPbFq5OVJYwfbGf74NKw9nHJvfDDL8RkRjG+jN388t4q8+M7k4bRX+I6NxbBJUrfW/n+Tn53Pw4EGjL6RqtZrIyEj27NlToX0sW7aMfv364eho/J7YuXMnPj4+NGjQgFGjRpGcnFytda9uslj1DsLCwpg9ezYAM2bMoEWLFsycOdPw+PLlywkMDOTvv/+mfv36AAQGBrJgwQJUKhUNGjTg6NGjLFiwgOHDh5tcj7y8PPLy8gx/Z2RklJNb3C/U9rY0eu85dDkF3Dh8gSufbMPWzw3nZsYneKVQy7lZ61EUhTpj7u75YEtJ+GYXab8dJ/StQahtik59KisNQROf5srijfw1YB6oVTg1D8E5oi5yf6qSEr/ZRcZvxwl+a7ChDwFSNv5Bzqkr1Pm/Z7D2cSXr+CWuf7QJKw9nnMLvrVERraJCa8IP2OnL3H5utrW1xda25BeDpKQktFotvr6+Rum+vr6cPHnyjs+3f/9+jh07xrJly4zSo6OjefLJJwkJCeHs2bO8/vrrdO/enT179qDRaMrYm2VJIHIHERE3V30fPnyYHTt24ORUcp3E2bNnDYFImzZtUKluHsht27Zl3rx5aLVakw+EWbNmMW3aNJPKWoKft4b4ROMfxo5P1OLirMbeXo1Go0KjodQ8vj7332Fp5eIAahWFqdlG6QVpWVh7lL3uRqVWYVd89YZDXV9yLiUT99Ueo0CkKAj5lvyEDOrP6n9fjoYAaPR9eNuiysK0TKzdyl+7lLh+DwlrdxP65kDsQ4xP/A71/Km/aDjarFyUQi1Wro6cfnk5DvX8q70NlnazD41HPwrTsrByL78Pk9bvIWndLoKnPYtd8M0+1OUVkPDFdgIn9sW5ZRgAdsG+5J6LI/nbvfdeIFK8+LTy5Yoi18DAQKP0KVOmMHXq1OqompFly5bRtGlTWrVqZZTer18/w/+bNm1Ks2bNqFu3Ljt37qRr167VXo/qIFMzd3DrkFdmZiY9evQgNjbWaDt9+jQdOnSo0XpMnDiR9PR0w3b58t09f92mpR3bf88xStv6azZtIorWLtjYqIhoZmuUR6dT2P57Nm0j7r/1DWprDQ71/MgoXiQJoOgUbsRexKlh6ZdNlkpRDOsa4GYQknsthbCZ/YsCnvuU2lqDfT1/Mm9ZJKnoFDKPXMChnD5MWLub+DW/EzKlPw5hpV8qDUWLf61cHcm7lkLOmeu4tK5frfW/G6itNdjX9SfryAVDmqJTyDpyHvsGtcssl7RuN4lf/UbQlAHY39aHilaHUqgDlfEogkqjNlobd6/QKWqTN4DLly8bnasnTpxY6vN4eXmh0WiIj483So+Pj8fPz6/cOmZlZbF69WqGDRt2x/aEhobi5eXFmTNnKtgD5nf/ffWsQQ8++CBr164lODgYK6uyu27fvn1Gf+/du5ewsLAqDYuVNbxnLplZOs6cLzD8feFSIbHH8vBwU1OntjWvv5XE1TgtKxYXfVMaOdiV95enM2F6EjH9XNixK4evv89kw/9ufsscO9KNmJcSiGhuS6twOxZ9kkZWtsLQfvfnPTB8e7fiwvyNOIb54VA/gITv/kCXV4DnI80AOD93AzaeztSK6QTA9TW7cQzzx9bfDV2BlowDZ0nefoyg0VFAURByduZ6ss/EUW/q06DVGdabaJztUVvfncOwVeH9RGsuL/we+3r+ONSvRdL3+9DlFuDetTkAlxZ8h7WHM/5DugDFQcjKX6jzci9sfN0M6z7UdjZo7G0ASPv9L6xcHbD2diX3QgLXlv6MS+sGOLeoa5lG1jDPJ9pwddF32Nfzxz4sgOQN+4v6MLKoD68s+BZrT2d8Bxd9e05cu4vEVb9Qe3xvrH1K9qHGwRaHB4KI/2wrahuroqmZY5dI23EEv+cesVg7TVXVEREXFxdcXFzukBtsbGyIiIhg27Zt9OrVCwCdTse2bdsYM2ZMuWW//vpr8vLyePbZZ+/4PFeuXCE5ORl//7t3hE8CkUoYPXo0n3zyCf379+fVV1/Fw8ODM2fOsHr1apYuXWoINC5dusS4ceMYOXIkhw4dYvHixcybN8+wn5SUFC5dusS1a0WXtp06VXStvX6F893owOFcuva5eSne+KlFVwoN7uvMp4t8uZ6g5fLVm4FKSB1rNnzhz/gpSby7NI3a/lZ8PM+HqM43R5ieecKZpGQtU2enEJdYSHgTW35cFYCv9/15WHp0bExhRjbX/vcbBalZ2If6EPZmX6zdi/okPzEDlfrmt0pdbgGXPthMftIN1DZW2AV6EvJyDzw6Ni7Kn3yD9L2nATgxxniVff23B5RYR3I/cGvfhML0bOJX/UJhahZ2ob6ETO2PdfG0QkFiutG0aPJPB1EKtVx8e63Rfnz6tcdvQEcAClMzub58i2F6wr1zM3yeaW++RpmZa/smFGZkk7DqF8NN4YKmDDAsYC1IMj4OUzcV9eHld4zvT+PdrwM+/Yv6sPbLT5Lw+XauzP8WbWYO1t6u+DzbGfdouaFZecaNG8eQIUNo2bIlrVq1YuHChWRlZRETEwPA4MGDqVWrFrNmzTIqt2zZMnr16oWnp6dRemZmJtOmTaNPnz74+flx9uxZXn31VerVq0dUVJTZ2lVZKuVeHDszk06dOhEeHs7ChQsNaadPn2bChAns2LGDvLw8goKCiI6OZv78+ahUKjp16kSTJk3Q6XSsWrUKjUbDqFGjmDFjhuEE+dlnnxkOtFtVZi4xIyMDV1dXUv8OxcVZZthM1erPp++cSZSroPD+G3kxN1MWR4qbtNl5nOw/m/T09AqNRpRGf0796FAE9k6V/zKUk1nIyAcPVroO7733HnPmzCEuLo7w8HDeffddWrduDRR9BgUHB/PZZ58Z8p86dYqGDRvy888/88gjxiNOOTk59OrViz///JO0tDQCAgLo1q0b06dPL7Eo9m4igUg1Ky14qQkSiFQPCUSqTgKRqpNApGqqMxD58NBDJgciox78o0p1+Ke6P8fAhRBCCBOY/uu78oXQVBKICCGEEMXk13fNTwKRarZz505LV0EIIYS4Z0ggIoQQQhSTqRnzk0BECCGEKGb6fUQkEDGVBCJCCCFEMZ2iQmfCVUymlBFFJBARQgghiulMHBHRyYiIyaTnhBBCCGExMiIihBBCFLv1B+wqW06YRgIRIYQQopgWFVoT7gliShlRRAIRIYQQopiMiJif9JwQQgghLEZGRIQQQohiWkybZtFWf1X+MSQQEUIIIYrJ1Iz5SSAihBBCFJNbvJufBCJCCCFEMcXEX99V5KoZk0kIJ4QQQgiLkRERIYQQophMzZifBCJCCCFEMfnRO/OTQEQIIYQopjXxR+9MKSOKSCAihBBCFJMREfOTEE4IIYQQFiMjIkIIIUQxHWp0JnxHN6WMKCKBiBBCCFFMq6jQmjDNYkoZUUQCESGEEKKYrBExPwlEhBBCiGKKib81o8h9REwmPSeEEEIIi5ERESGEEKKYFhVaE343xpQyoogEIkIIIUQxnWLaeg+dUgOV+YeQQEQIIYQopjNxjYgpZUQR6TkhhBDCQt5//32Cg4Oxs7OjdevW7N+/v8y8n332GSqVymizs7MzyqMoCpMnT8bf3x97e3siIyM5ffp0TTejSiQQEUIIIYrpUJm8VdaaNWsYN24cU6ZM4dChQzRv3pyoqCgSEhLKLOPi4sL169cN28WLF40enz17Nu+++y5Llixh3759ODo6EhUVRW5ubqXrZy4SiAghhBDF9Dc0M2WrrPnz5zN8+HBiYmJo3LgxS5YswcHBgeXLl5dZRqVS4efnZ9h8fX0NjymKwsKFC3njjTd44oknaNasGZ9//jnXrl3j22+/NaU7zEICESGEEKKYfo2IKRtARkaG0ZaXl1fq8+Tn53Pw4EEiIyMNaWq1msjISPbs2VNm/TIzMwkKCiIwMJAnnniC48ePGx47f/48cXFxRvt0dXWldevW5e7T0mSx6j2u85FeaBxsLV2Ne9b+Fl9bugr3vAcPPGPpKtzzVHJXziqpzt7TYeKdVYtrERgYaJQ+ZcoUpk6dWiJ/UlISWq3WaEQDwNfXl5MnT5b6HA0aNGD58uU0a9aM9PR05s6dy8MPP8zx48epXbs2cXFxhn3cvk/9Y3cjCUSEEEKIanL58mVcXFwMf9vaVt8XxbZt29K2bVvD3w8//DCNGjXio48+Yvr06dX2POYmUzNCCCFEMcXEhapK8YiIi4uL0VZWIOLl5YVGoyE+Pt4oPT4+Hj8/vwrV1dramhYtWnDmzBkAQ7mq7NMSJBARQgghiul/9M6UrTJsbGyIiIhg27ZtN59bp2Pbtm1Gox7l0Wq1HD16FH9/fwBCQkLw8/Mz2mdGRgb79u2r8D4tQaZmhBBCiGLmvKHZuHHjGDJkCC1btqRVq1YsXLiQrKwsYmJiABg8eDC1atVi1qxZALz55pu0adOGevXqkZaWxpw5c7h48SLPP/88UHRFzdixY5kxYwZhYWGEhIQwadIkAgIC6NWrV6XrZy4SiAghhBDFTBnd0JerrGeeeYbExEQmT55MXFwc4eHhbNq0ybDY9NKlS6jVNwOc1NRUhg8fTlxcHO7u7kRERLB7924aN25syPPqq6+SlZXFiBEjSEtLo127dmzatKnEjc/uJipFUeQO+fegjIwMXF1dCf9mnFw1UwVy1UzVyVUzVafVySx5VWiz8zjRfzbp6elGC0UrQ39OfeLn57B2tKl0+YKsfL7rtrxKdfinkhERIYQQopipd0k1pYwoIoGIEEIIUcycUzOiiAQiQgghRDEJRMxPAhEhhBCimAQi5icrpIQQQghhMTIiIoQQQhSTERHzk0BECCGEKKZg2hUwch8M00kgIoQQQhSTERHzkzUiQgghhLAYGRERQgghismIiPlJICKEEEIUk0DE/CQQEUIIIYpJIGJ+EogIIYQQxRRFhWJCUGFKGVFEFqsKIYQQwmJkREQIIYQoJr++a34SiAghhBDFZI2I+UkgIoQQQhSTNSLmJ4GIEEIIUUxGRMxPFqsKIYQQwmJkREQIIYQoJlMz5ieBiBBCCFFMMXFqRgIR00kgIoQQQhRTAEUxrZwwjQQiQgghRDEdKlRyHxGzksWqQgghhLAYGRERQgghisliVfOTQEQIIYQoplNUqOQ+ImYlgYgQQghRTFFMXKwqq1VNJoGIEEIIUUymZsxPAhFRYQkbDhK/dh8FqZnYh/hQZ1Q3HBsElJo3ddcp4tbsJu96KkqhDtta7vj2boVn16YAKIVarn7+K+l/nCU/Lg2Noy3O4cHUiumEjaezGVtlPr/uyWHuh6kcOpLH9Xgta5f70au7U7lldu7O5uUpyRz/O4/AAGteH+vO0GdcjPJ88Gkacz9IIy5RS/PGNix6y5tWLexqsikWlfTDARLW7aUwNRP7EF9qjeyGQ/1apeZN232ShK93kXc9FQp12AS4492rDR5dmhryKIpC/MpfSf75T7RZeTg2qk3tF7pjG+BhriaZXfIPf5D07R4KUzOxC/bFf0R0mX2Y8vMh0nYcIfdiIgD2df3xHdTZKH9hWiZxK7aR+ec5tFm5ODYJwn9EFLYBnmZpj7i3yVUzZlZQUMCECRNo2rQpjo6OBAQEMHjwYK5du2bpqpUr5Ze/uPLJNvwHtKPR4udwCPXl9KQ1FKRllZrfytkOv34P02DeYBp/MAzPyGZcWPAD6QfPAaDLKyD7TBz+/f9Fo8UxhL7xJLlXkjk77RtzNsussrJ1NG9sy+KZ3hXKf/5SAT2evU6nf9lzaEsdXhruyojxCWzecbPP13x3g/FTk5g03oMDmwNp1tiW7v2vkZBUWFPNsKjU3/7i2tKt+PVvT/2Fw7AL8eHc5NXlHIf2+Pb9F2FzhlJ/8XA8IptzedEGMg6dNeRJXLuHxI1/UPuF7oTNHYrazppzk79El39/9mH6b8eJW74Fn2c6UHf+cOxCfLkwdRWFZfRh1tGLuLZ/gJAZg6g7OwZrLxcuTF1JQXIGUBTIXZz5FflxadT5v2eot2A41j6uXJi8El1uvjmbVi30IyKmbKZ4//33CQ4Oxs7OjtatW7N///4y837yySe0b98ed3d33N3diYyMLJF/6NChqFQqoy06OtqkupmLBCKVkJ9f9TdVdnY2hw4dYtKkSRw6dIh169Zx6tQpevbsWQ01rDnx6/fjFd0cr27NsK/jRZ0x0ahtrUj++Uip+Z2bBeH+cAPs63hh6++Ob6+HsA/xIfP4ZQA0jnbUn9kfjw6NsKvtiVPDWtR5oRvZZ+LIT0g3Z9PMpntXR6a/5knvR8sfBdH76PN0QupYM3eqF43q2zD6OTf6PO7Ewo9v9s/Cj9J4fqArMf1caNzAhg9ne+Ngr+LTL2/UVDMsKunbfXhEheMR2Ry7Ot7UfuFRVLZWpGw5XGp+p6ZBuLZtiF1g0XHo3bMV9sE+ZP1VdBwqikLi9/vx7dsO1zYNsA/xpc5/e1KQcoP0vafM2TSzSfpuL+7dWuAeGY5dHW8CRj2G2taa1K2xpeYPHN8bz0dbYh/qh21tL2qNeRx0CpmHzwOQfy2FnFNXCRjVHYewAGxrexHw70fR5ReQ9utxM7aseuh/9M6UrbLWrFnDuHHjmDJlCocOHaJ58+ZERUWRkJBQav6dO3fSv39/duzYwZ49ewgMDKRbt25cvXrVKF90dDTXr183bF9++aVJfWEuEoiUo1OnTowZM4axY8fi5eVFVFQUx44do3v37jg5OeHr68ugQYNISkoqUWbMmDG4urri5eXFpEmTUIpXMrm6urJlyxb69u1LgwYNaNOmDe+99x4HDx7k0qVLlmpquXQFWrLPxOESHmJIU6lVOIcHk3nyajkliyiKQkbsBfKupOD8QJ0y82mz8kAFGqf7d1qhMvYeyKVre3ujtG6dHNh7MBeA/HyFg0fyjPKo1Sq6tndgT3Ge+0nRcXgd5+a3H4chZJ+6csfyiqJw4/B58q6m4NSk6DjMj0+jMDUL5/BgQz6Nox0O9WuRXYFj+16jK9CSc/Y6Trf1oVPzivUhFI1mKlodGuei404pKBo5UlnfnOlXqVWorKzIPnF3ntPKo1+saspWWfPnz2f48OHExMTQuHFjlixZgoODA8uXLy81/8qVK3nhhRcIDw+nYcOGLF26FJ1Ox7Zt24zy2dra4ufnZ9jc3d1N6QqzkUDkDlasWIGNjQ27du3i7bffpkuXLrRo0YIDBw6wadMm4uPj6du3b4kyVlZW7N+/n0WLFjF//nyWLl1a5nOkp6ejUqlwc3Or4daYpjAjG3QKVu4ORunWbo4UpGSWWU6blcufT87lUM/ZnJnyFYGjHsHlwZBS8+ryC7n66U48OjZG42BbrfW/V8UlavH11hil+XpryLihIydHR1KKFq2WUvPEJ9x/0wpaw3HoaJRu5eZIYWrp0wpQdBwefXo2R3q/zflpawgY2Q3nFqEAhnJWbiX3WZBa9rF9rzL0oZvxqFxRH1asvfGfb8PKwxmn5kV9aFvbC2tvV+L/tx1tZg66Ai2Ja3dRmJxBYTnnh3+6/Px8Dh48SGRkpCFNrVYTGRnJnj17KrSP7OxsCgoK8PAwXs+0c+dOfHx8aNCgAaNGjSI5Obla617dZLHqHYSFhTF79mwAZsyYQYsWLZg5c6bh8eXLlxMYGMjff/9N/fr1AQgMDGTBggWoVCoaNGjA0aNHWbBgAcOHDy+x/9zcXCZMmED//v1xcXEp8bheXl4eeXl5hr8zMjKqq4k1Rm1vS6P3nkOXU8CNwxe48sk2bP3ccG4WZJRPKdRybtZ6FEWhzpi7ey5T3HvU9rbUX/Q8utx8bhy+wLVlW7H1c8epadCdCwsjid/sIv2344S8NRi1TdHHh8pKQ53Xnubqexs4MXAuqFU4NQ/FKaLePXlNa9HohilXzRT9e/u52dbWFlvbkl+ukpKS0Gq1+Pr6GqX7+vpy8uTJCj3nhAkTCAgIMApmoqOjefLJJwkJCeHs2bO8/vrrdO/enT179qDRaMrZm+VIIHIHERERhv8fPnyYHTt24ORUco7/7NmzhkCkTZs2qFQ3D+S2bdsyb948tFqt0YFQUFBA3759URSFDz/8sNx6zJo1i2nTplW1OSaxcnEAtYrC1Gyj9IK0LKw9yl7voFKrsCu+8sChri85l5KJ+2qPUSBSFIR8S35CBvVn9ZfRkFv4eWuIT9QapcUnanFxVmNvr0ajUaHRUGoeX5/7762tMRyHxqMfhWlZJUZJbqVSqwxXwNiH+pF3OYn4r3fj1DTIUK4wLQtrj5tXaxWmZWEf6lvq/u5lhj5MMx6pKOrD8tcuJa3fQ+K6XYRMexa7YOO+sa/nT72FI9Bm5aIUarFydeTsy8uwr1f6VXV3s6pevhsYGGiUPmXKFKZOnVodVTPy9ttvs3r1anbu3Imd3c3p7H79+hn+37RpU5o1a0bdunXZuXMnXbt2rfZ6VAeZmrkDR8ebJ7jMzEx69OhBbGys0Xb69Gk6dOhQqf3qg5CLFy+yZcuWckdDACZOnEh6erphu3z5skntMYXaWoNDPT8yDl8wpCk6hRuxF3FqWPolf6VSFHQFNz809UFI7rUUwmb2Lwp4hEGblnZs/z3HKG3rr9m0iSg66djYqIhoZmuUR6dT2P57Nm0j7r91NkXHoT83jlwwpCk6hczDF3BoULvC+1EUxbCuwcbXDSt3R27ccmxrs/PI/vsqDpU5tu8RamsN9nX9yby9D4+cL7cPE9ftJuGr3wieMgD7sLKDC42jHVaujuRdSybn7HWcW9evzuqbhVKFDeDy5ctG5+qJEyeW+jxeXl5oNBri4+ON0uPj4/Hz8yu3jnPnzuXtt9/m559/plmzZuXmDQ0NxcvLizNnzpSbz5Luv69NNejBBx9k7dq1BAcHY2VVdtft27fP6O+9e/cSFhZmGA3RByGnT59mx44deHre+Vr7sob3zMW3dysuzN+IY5gfDvUDSPjuD3R5BXg+UvQmOD93AzaeztSK6QTA9TW7cQzzx9bfDV2BlowDZ0nefoyg0VFAURByduZ6ss/EUW/q06DVGdabaJztUVvfnUOIVZGZpePM+QLD3xcuFRJ7LA8PNzV1alvz+ltJXI3TsmJx0bfNkYNdeX95OhOmJxHTz4Udu3L4+vtMNvzP37CPsSPdiHkpgYjmtrQKt2PRJ2lkZSsM7Xd/3ovFq1drLi/4Hod6/jjUDyDxu/3ocgvwiCw6Di/N/x5rT2f8h3QGIP7rXTjU88fG3x2lQEvGgTOk7jhG7VFFU4AqlQrvnq1IWLML2wAPbHzdiPviF6w9nHFt08Bi7axJXk+04cqi77Cv5499WADJG4r60D2yOQBXFnyLlaczfoOLvj0nrt1FwqpfqD2+N9Y+boa1M2o7GzT2NgCk7/oLjYsDNt6u5F5M4PrSzbi0boBzi7qWaWQVVHVExMXF5Y5fLAFsbGyIiIhg27Zt9OrVC8Cw8HTMmDFllps9ezZvvfUWmzdvpmXLlnd8nitXrpCcnIy/v/8d81qKBCKVMHr0aD755BP69+/Pq6++ioeHB2fOnGH16tUsXbrUEGhcunSJcePGMXLkSA4dOsTixYuZN28eUBSEPPXUUxw6dIiNGzei1WqJi4sDwMPDAxsbG4u1rzweHRtTmJHNtf/9RkFqFvahPoS92Rfr4qHt/MQMVOqbb15dbgGXPthMftIN1DZW2AV6EvJyDzw6Ni7Kn3yD9L2nATgxxniFeP23B5RYR3I/OHA4l659bt4vZvzUoqutBvd15tNFvlxP0HL56s1AJaSONRu+8Gf8lCTeXZpGbX8rPp7nQ1Tnm6N0zzzhTFKylqmzU4hLLCS8iS0/rgrA1/v+fGu7t2+MNj2LuJW/UJhaNH0SMq0f1sXTCvmJ6aAyPg6vfLiJguSi49C2tid1xj+Be/vGhjzefdoW5Xvvx6KbcTUOJHRaP8MaiPuNa/smFGZkk7Dql6IbmoX4EjxlgGEBa35SBtzyXk7ZdBClUMvld4zv8ePdrwO+/TsCUJiSyfVlW9CmZ2Ll7oxb56Z4963cKPE/0bhx4xgyZAgtW7akVatWLFy4kKysLGJiYgAYPHgwtWrVYtasWQC88847TJ48mVWrVhEcHGz47HBycsLJyYnMzEymTZtGnz598PPz4+zZs7z66qvUq1ePqKgoi7XzTlSKcg+uJjKTTp06ER4ezsKFCw1pp0+fZsKECezYsYO8vDyCgoKIjo5m/vz5qFQqOnXqRJMmTdDpdKxatQqNRsOoUaOYMWMGKpWKCxcuEBJS+pUjO3bsoFOnThWqW0ZGBq6uroR/M07WVVTB/hZfW7oK97wHDzxj6Src87Q6mSWvCm12Hif6zyY9Pb1CoxGl0Z9TQ1e8jsah8lOb2uxczg2ZWek6vPfee8yZM4e4uDjCw8N59913ad26NVD0GRQcHMxnn30GQHBwMBcvXiyxD/06lJycHHr16sWff/5JWloaAQEBdOvWjenTp5dYFHs3kUCkmpUWvNQECUSqhwQiVSeBSNVJIFI11RqIfPZ/qE0IRHTZuZwb+laV6vBPdX+OPQohhBAmkF/fNT8JRIQQQohi8uu75ieBSDXbuXOnpasghBBC3DMkEBFCCCH0FFXRZko5YRIJRIQQQohiskbE/CQQEUIIIfRuvU1qZcsJk0ggIoQQQhSTxarmJxevCyGEEMJiZERECCGEuJVMs5iVBCJCCCFEMZmaMT8JRIQQQgg9WaxqdrJGRAghhBAWIyMiQgghhIGqeDOlnDCFBCJCCCGEnkzNmJ0EIkIIIYSeBCJmJ4GIEEIIoSe/NWN2slhVCCGEEBYjIyJCCCFEMfnRO/OTQEQIIYTQkzUiZieBiBBCCKEna0TMTgIRIYQQophKKdpMKSdMI4tVhRBCCGExMiIihBBC6MkaEbOTQEQIIYTQkzUiZieBiBBCCKEnIyJmJ4GIEEIIoSeBiNnJYlUhhBBCWIyMiAghhBB6MiJidhKICCGEEHqyWNXsJBARQgghiskNzcxP1ogIIYQQFvL+++8THByMnZ0drVu3Zv/+/eXm//rrr2nYsCF2dnY0bdqUH3/80ehxRVGYPHky/v7+2NvbExkZyenTp2uyCVUmgYgQQgihp1Rhq6Q1a9Ywbtw4pkyZwqFDh2jevDlRUVEkJCSUmn/37t3079+fYcOG8eeff9KrVy969erFsWPHDHlmz57Nu+++y5IlS9i3bx+Ojo5ERUWRm5tb+QqaiQQiQgghhAXMnz+f4cOHExMTQ+PGjVmyZAkODg4sX7681PyLFi0iOjqaV155hUaNGjF9+nQefPBB3nvvPaBoNGThwoW88cYbPPHEEzRr1ozPP/+ca9eu8e2335qxZZUjgYgQQghRTMXNdSKV2orLZ2RkGG15eXmlPk9+fj4HDx4kMjLSkKZWq4mMjGTPnj2lltmzZ49RfoCoqChD/vPnzxMXF2eUx9XVldatW5e5z7uBLFa9xxUUatAVaixdjXvWgweesXQV7nmHWq6xdBXueREH+1q6Cvc0lUZbfTur4lUzgYGBRslTpkxh6tSpJbInJSWh1Wrx9fU1Svf19eXkyZOlPkVcXFyp+ePi4gyP69PKynM3kkBECCGEqCaXL1/GxcXF8Letra0Fa3NvkKkZIYQQQq+Ki1VdXFyMtrICES8vLzQaDfHx8Ubp8fHx+Pn5lVrGz8+v3Pz6fyuzz7uBBCJCCCGEnpmumrGxsSEiIoJt27YZ0nQ6Hdu2baNt27allmnbtq1RfoAtW7YY8oeEhODn52eUJyMjg3379pW5z7uBTM0IIYQQxcx5Q7Nx48YxZMgQWrZsSatWrVi4cCFZWVnExMQAMHjwYGrVqsWsWbMAeOmll+jYsSPz5s3jscceY/Xq1Rw4cICPP/64qA4qFWPHjmXGjBmEhYUREhLCpEmTCAgIoFevXpWvoJlIICKEEELomfG3Zp555hkSExOZPHkycXFxhIeHs2nTJsNi00uXLqFW35y4ePjhh1m1ahVvvPEGr7/+OmFhYXz77bc88MADhjyvvvoqWVlZjBgxgrS0NNq1a8emTZuws7MzoVHmoVIURW5Mew/KyMjA1dWVJqtfQeMgi6FMpZL7MleZXDVTdXLVTNVos/M42nce6enpRgtFK0N/Tg2e8RZqEz60dbm5XHjj/6pUh38qGRERQggh9OTXd81OAhEhhBCimPzonflJICKEEELoVfGGZqLyJBARQggh9GRqxuzkPiJCCCGEsBgZERFCCCGKyRoR85NARAghhNCTqRmzk0BECCGE0DNxREQCEdNJICKEEELoyYiI2cliVSGEEEJYjIyICCGEEHoyImJ2EogIIYQQxeSqGfOTqRkhhBBCWIwEIkIIIYSwGJmaEUIIIfRkjYjZSSAihBBCFJM1IuYngYgQQghxKwkqzEoCESGEEEJPpmbMTharCiGEEMJiZERECCGEKCZrRMxPAhEhhBBCT6ZmzE4CESGEEKKYjIiYnwQiQgghhJ6MiJidLFYVQgghhMXIiIgQQgihJyMiZieBiBBCCFFM1oiYnwQiQgghhJ6MiJidBCJCCCGEngQiZieBiKiwpB8OkLh+D4WpmdiF+FJrRBQO9WuVmjd58yFSdxwl72IiAPb1/PAb1Nkof0FqJnErtnMj9hzazFwcm9Sh1shobAM8zNIeS0j64QAJ6/ZSmJqJfYgvtUZ2K7MP03afJOHrXeRdT4VCHTYB7nj3aoNHl6aGPIqiEL/yV5J//hNtVh6OjWpT+4Xu920f/ronh7kfpnLoSB7X47WsXe5Hr+5O5ZbZuTubl6ckc/zvPAIDrHl9rDtDn3ExyvPBp2nM/SCNuEQtzRvbsOgtb1q1sKvJplhU4sYDJKzbZ3QcOjYIKDVv2u6TxH+12+g49Ond2ug4TNt9kuSf/iT7TBzaGznUf3cYDqG+5mqOuMfJVTMWMHXqVBo2bIijoyPu7u5ERkayb98+S1erXGm/Hef6si349mtP2ILnsQ/25fyULylMyyo1f9axi7h1aELoW89Sd85QrL1cODdlFQXJGUDRB+jFmV+TH5dK8P/1JWzhcGx8XDk36Qt0ufnmbJrZpP72F9eWbsWvf3vqLxyGXYgP5yavpqCMPrRytse3778ImzOU+ouH4xHZnMuLNpBx6KwhT+LaPSRu/IPaL3QnbO5Q1HbWnJv8Jbr8QnM1y6yysnU0b2zL4pneFcp//lIBPZ69Tqd/2XNoSx1eGu7KiPEJbN5xs8/XfHeD8VOTmDTegwObA2nW2Jbu/a+RkHR/9mHqr39xbek2/Pq3o8Gi57C/w3GocSo6DuvPHUKD957HM7IZlxZuJOPgOUMeXW4Bjo1rEzC0s7maUWP0a0RM2WpKSkoKAwcOxMXFBTc3N4YNG0ZmZma5+f/zn//QoEED7O3tqVOnDi+++CLp6enGbVWpSmyrV6+uuYaUQQKRSsjPr54PyPr16/Pee+9x9OhRfv/9d4KDg+nWrRuJiYnVsv+akPjdPjy6tcAjMhy7Ot7UeuFRVLbWpGyNLTV/nfG98Xq0JfahftjV9qL2mMdBp5B5+AIA+ddSyD51lVovPIpDWAB2tT2pNepRdPmFpP563HwNM6Okb/fhERWOR2Rz7Op4U/uFR1HZWpGy5XCp+Z2aBuHatiF2gV7Y+rvj3bMV9sE+ZP11GSgK5hK/349v33a4tmmAfYgvdf7bk4KUG6TvPWXOpplN966OTH/Nk96Plj8KovfR5+mE1LFm7lQvGtW3YfRzbvR53ImFH988IS/8KI3nB7oS08+Fxg1s+HC2Nw72Kj798kZNNcOiEr/dj2dUOJ6PFB+Ho7ujLuc4dG4WhNvDDW4eh0+0wj7k5nEI4NGlKX792+MUHmymVtQgpQpbDRk4cCDHjx9ny5YtbNy4kV9//ZURI0aUmf/atWtcu3aNuXPncuzYMT777DM2bdrEsGHDSuT99NNPuX79umHr1atXzTWkDBKIlKNTp06MGTOGsWPH4uXlRVRUFMeOHaN79+44OTnh6+vLoEGDSEpKKlFmzJgxuLq64uXlxaRJk1CUm0fpgAEDiIyMJDQ0lCZNmjB//nwyMjI4cuSIJZp5R7oCLTlnruMUHmJIU6lVODcPJvvk1YrtI68ARatD42xv2CeAylpjtE+1tYbsW05w9wtdgZbsM9dxbn5bH4aHkH3qyh3LK4rCjcPnybuaglOTOgDkx6dRmJqF8y0nf42jHQ71a1X4dbnf7T2QS9f29kZp3To5sPdgLgD5+QoHj+QZ5VGrVXRt78Ce4jz3E/1xeGvAoFKrcAoPIasCx4yiKNyIPU/elRScHqhTgzW1nLttROTEiRNs2rSJpUuX0rp1a9q1a8fixYtZvXo1165dK7XMAw88wNq1a+nRowd169alS5cuvPXWW2zYsIHCQuORPjc3N/z8/AybnZ35pyQlELmDFStWYGNjw65du3j77bfp0qULLVq04MCBA2zatIn4+Hj69u1booyVlRX79+9n0aJFzJ8/n6VLl5a6//z8fD7++GNcXV1p3ry5OZpUadqMbNApWLk5GqVbuTlRkFb28OCt4lZsx9rDCafiD2K72p5Ye7sQ9/kOCjNz0BVoSVi7m4KkGxSkVmyf9xJDH7rf3oeOFKaWPiQOoM3K5ejTsznS+23OT1tDwMhuOLcIBTCUK/m6ON6XfWiKuEQtvt4aozRfbw0ZN3Tk5OhIStGi1VJqnviE+29qRn8cWt92zFhX4Dg88tQcDvd6h3PTvqLWyG44twgpM7+oPnv27MHNzY2WLVsa0iIjI1Gr1ZWa0k9PT8fFxQUrK+OloaNHj8bLy4tWrVqxfPlyoy/N5iKLVe8gLCyM2bNnAzBjxgxatGjBzJkzDY8vX76cwMBA/v77b+rXrw9AYGAgCxYsQKVS0aBBA44ePcqCBQsYPny4odzGjRvp168f2dnZ+Pv7s2XLFry8vMqsR15eHnl5eYa/MzIyqrupNSbhm12k/Xac0LcGobYpOuRUVhqCJj7NlcUb+WvAPFCrcGoegnNEXVl9fgu1vS31Fz2PLjefG4cvcG3ZVmz93HFqGmTpqol/ELW9LQ3eHYY2t4DM2AtcXbYVGz83nJvdh8dhFa+auf3cbGtri62trcnViYuLw8fHxyjNysoKDw8P4uLiKrSPpKQkpk+fXmI6580336RLly44ODjw888/88ILL5CZmcmLL75ocn1NIYHIHURERBj+f/jwYXbs2IGTU8n56bNnzxoCkTZt2qBSqQyPtW3blnnz5qHVatFoir55de7cmdjYWJKSkvjkk0/o27cv+/btK3HA6c2aNYtp06ZVZ9MqTOPiAGpViYWphWmZWLuVP1efuH4PCWt3E/rmQOxDjFfRO9Tzp/6i4WizclEKtVi5OnL65eU41POv9jZYmqEPU2/vw6wSoyS3UqlVhitg7EP9yLucRPzXu3FqGmQoV5iWhbWHs9E+7eWKBQD8vDXEJ2qN0uITtbg4q7G3V6PRqNBoKDWPr8/9d3rUH4e3L0wtqMRx6BDqS+6VJBK+3i2ByO3lKPoieqspU6YwderUEtlfe+013nnnnXJ3eeLECRMqYiwjI4PHHnuMxo0bl6jHpEmTDP9v0aIFWVlZzJkzx+yBiEzN3IGj4803Z2ZmJj169CA2NtZoO336NB06dKj0fuvVq0ebNm1YtmwZVlZWLFu2rMz8EydOJD093bBdvmy+dRRqaw329fzJPHzekKboFDKPXMChYemXngIkrN1N/JrfCZnSH4ew0i8NhKJ1DVaujuRdSyHnzHVcWtev1vrfDdTWGhzq+XPjyAVDmlK8eNehQe0K70dRFJSCoikDG183rNwduXH45j612Xlk/3213Nfln6RNSzu2/55jlLb112zaRBTNg9vYqIhoZmuUR6dT2P57Nm0j7r/Ld/XHYeYtx4z+OHSszDGjUwzrvO43qipsAJcvXzY6V0+cOLHU5xk/fjwnTpwodwsNDcXPz4+EhASjsoWFhaSkpODn51duW27cuEF0dDTOzs6sX78ea2vrcvO3bt2aK1euGI2+m8P9F/LXoAcffJC1a9cSHBxcYp7tVrfP2+3du5ewsDDDaEhpdDpduS9+VYf3qsr7idZcXvg99vX8cahfi6Tv96HLLcC9a9G6lksLvsPawxn/IV2A4iBk5S/UebkXNr5uhjULajsbNPY2AKT9/hdWrg5Ye7uSeyGBa0t/xqV1A5xb1LVMI2uYV6/WXF7wPQ71/HGoH0Did/vR5RbgEdkMgEvzv8fa0xn/IUWXQMZ/vQuHev7Y+LujFGjJOHCG1B3HqD0qGii69M67ZysS1uzCNsADG1834r74BWsPZ1zbNLBYO2tSZpaOM+cLDH9fuFRI7LE8PNzU1KltzetvJXE1TsuKxUUjQiMHu/L+8nQmTE8ipp8LO3bl8PX3mWz4381Rt7Ej3Yh5KYGI5ra0Crdj0SdpZGUrDO3nXOL57wfevVpxacEGHMJKPw4vzis6DvWX4sZ/tRuHMH9s/N2KjsM/zpKy4xiBL0Qb9ll4I4f8xAwKk4uuNMq7kgyAtbsj1u4Vu8LprlHFEREXFxdcXFzKzwt4e3vj7X3ny9Dbtm1LWloaBw8eNIzQb9++HZ1OR+vWrcssl5GRQVRUFLa2tnz//fcVWoQaGxuLu7u72T9rJBCphNGjR/PJJ5/Qv39/Xn31VTw8PDhz5gyrV69m6dKlhkDj0qVLjBs3jpEjR3Lo0CEWL17MvHnzAMjKyuKtt96iZ8+e+Pv7k5SUxPvvv8/Vq1d5+umnLdm8crm1b0Jhejbxq36hMDULu1BfQqb2N5xkChLTjaajkn86iFKo5eLba43249OvPX4DOgJQmJrJ9eVbiqcnnHDv3AyfZ9qbr1Fm5t6+Mdr0LOJWFvWhfagvIdP6GfowPzEdbulDXW4BVz7cREHyDdQ2VtjW9qTO+Cdwb9/YkMe7T9uifO/9iDYrF8fGgYRO62dYi3O/OXA4l659bl4pMH5q0RVrg/s68+kiX64naLl89WagElLHmg1f+DN+ShLvLk2jtr8VH8/zIarzzZHOZ55wJilZy9TZKcQlFhLexJYfVwXg631/9qF7h8YUpmdz/YtfDcdh6JvP3HIcZoD6luMwL5/LHxgfh0Hje+Le4eZxmL7vNJcXbjT8fXH2twD49m+H/8DKjRYLY40aNSI6Oprhw4ezZMkSCgoKGDNmDP369SMgoGik+erVq3Tt2pXPP/+cVq1akZGRQbdu3cjOzuaLL74gIyPDsHbF29sbjUbDhg0biI+Pp02bNtjZ2bFlyxZmzpzJyy+/bPY2qhRLLJG9R3Tq1Inw8HAWLlxoSDt9+jQTJkxgx44d5OXlERQURHR0NPPnz0elUtGpUyeaNGmCTqdj1apVaDQaRo0axYwZM1CpVOTm5jJgwAD27dtHUlISnp6ePPTQQ7zxxhs89NBDFa5bRkYGrq6uNFn9ChoHy42U3OtU8ktVVXao5RpLV+GeF3Gw750ziTJps/M42nee4coQUxjOqf+eica28tNy2rxcji95vUp1KEtKSgpjxoxhw4YNqNVq+vTpw7vvvmtYr3jhwgVCQkLYsWMHnTp1YufOnXTuXPrN5c6fP09wcDCbNm1i4sSJnDlzBkVRqFevHqNGjWL48OGo1eZdtSGBSDUrLXipCRKIVA8JRKpOApGqk0Ckaqo1EBlZhUDko5oJRO539+fYoxBCCGEq+X5iVhKICCGEEMVMvUuqDK6aTgKRarZz505LV0EIIYS4Z0ggIoQQQuhV8fJdUXkSiAghhBDFZGrG/CQQEUIIIfRkRMTsJBARQgghismIiPnJb80IIYQQwmJkREQIIYTQk6kZs5NARAghhNCTQMTsJBARQgghiskaEfOTNSJCCCGEsBgZERFCCCH0ZGrG7CQQEUIIIYqpFAWVCT9Kb0oZUUQCESGEEEJPRkTMTgIRIYQQopgsVjU/WawqhBBCCIuREREhhBBCT6ZmzE4CESGEEKKYTM2YnwQiQgghhJ6MiJidBCJCCCFEMRkRMT9ZrCqEEEIIi5ERESGEEEJPpmbMTgIRIYQQ4hYyzWJeEogIIYQQeopStJlSTphEAhEhhBCimCxWNT9ZrCqEEEIIi5ERESGEEEJPFquanQQiQgghRDGVrmgzpZwwjQQiQgghhJ6MiJidrBERQgghiukXq5qy1ZSUlBQGDhyIi4sLbm5uDBs2jMzMzHLLdOrUCZVKZbT9+9//Nspz6dIlHnvsMRwcHPDx8eGVV16hsLCw5hpSBhkREUIIIe5iAwcO5Pr162zZsoWCggJiYmIYMWIEq1atKrfc8OHDefPNNw1/Ozg4GP6v1Wp57LHH8PPzY/fu3Vy/fp3BgwdjbW3NzJkza6wtpZFARAghhNC7y+4jcuLECTZt2sQff/xBy5YtAVi8eDGPPvooc+fOJSAgoMyyDg4O+Pn5lfrYzz//zF9//cXWrVvx9fUlPDyc6dOnM2HCBKZOnYqNjU2NtKc0MjUjhBBCFKvq1ExGRobRlpeXV6X67NmzBzc3N0MQAhAZGYlarWbfvn3lll25ciVeXl488MADTJw4kezsbKP9Nm3aFF9fX0NaVFQUGRkZHD9+vEp1riwZEbnHaRUVKCpLV+OepZK+q7KIg30tXYV73sGIryxdhXtaxg0d7tW1syouVg0MDDRKnjJlClOnTjW5OnFxcfj4+BilWVlZ4eHhQVxcXJnlBgwYQFBQEAEBARw5coQJEyZw6tQp1q1bZ9jvrUEIYPi7vP3WBAlEhBBCiGpy+fJlXFxcDH/b2tqWmu+1117jnXfeKXdfJ06cMLkeI0aMMPy/adOm+Pv707VrV86ePUvdunVN3m9NkEBECCGEKFbVW7y7uLgYBSJlGT9+PEOHDi03T2hoKH5+fiQkJBilFxYWkpKSUub6j9K0bt0agDNnzlC3bl38/PzYv3+/UZ74+HiASu23OkggIoQQQuiZabGqt7c33t7ed8zXtm1b0tLSOHjwIBEREQBs374dnU5nCC4qIjY2FgB/f3/Dft966y0SEhIMUz9btmzBxcWFxo0bV6otVSWLVYUQQohid9t9RBo1akR0dDTDhw9n//797Nq1izFjxtCvXz/DFTNXr16lYcOGhhGOs2fPMn36dA4ePMiFCxf4/vvvGTx4MB06dKBZs2YAdOvWjcaNGzNo0CAOHz7M5s2beeONNxg9enSZ00k1RQIRIYQQQk+pwlZDVq5cScOGDenatSuPPvoo7dq14+OPPzY8XlBQwKlTpwxXxdjY2LB161a6detGw4YNGT9+PH369GHDhg2GMhqNho0bN6LRaGjbti3PPvssgwcPNrrviLnI1IwQQghxF/Pw8Cj35mXBwcEot0wNBQYG8ssvv9xxv0FBQfz444/VUseqkEBECCGEKFbVxaqi8iQQEUIIIfR0StFmSjlhEglEhBBCCD359V2zk0BECCGEKKbCxKmZaq/JP4dcNSOEEEIIi5ERESGEEELvLvv13X8CCUSEEEKIYnLVjPlJICKEEELoyWJVs5NARAghhCimUhRUJkyzmFJGFJHFqkIIIYSwGBkREUIIIfR0xZsp5YRJJBARQgghisnUjPlJICKEEELoyWJVs5M1IkIIIYSwGBkREUIIIfTkhmZmJ4GIEEIIUUxuaGZ+EogIIYQQejIiYnYSiAghhBDFVLqizZRywjSyWFUIIYQQFiMjIkIIIYSeTM2YnQQiQgghhJ7cR8TsJBARQgghismdVc1PAhEhhBBCT6ZmzE4WqwohhBDCYmRERAghhNBTMO2XdGVAxGQSiAghhBDFZI2I+UkgIoQQQugpmLhGpNpr8o8hgYgQQgihJ4tVzU4CEVFhyT/8QfK3eyhMzcQu2Be/EdE41K9Vat6Unw+RvuMIuRcTAbCv64/PoM5G+bU5+SR8vo2MfafQ3sjBxscNj8db4dE9wiztsYTkH/4g6ZY+9L9DH6bd1oe+t/VhYVomcSu2kfnnObRZuTg2CcJ/RBS2AZ5maY8lJG48QMK6fRSmZmIf4kutkd1wbBBQat603SeJ/2o3eddToVCHTYA7Pr1b49GlqVGe5J/+JPtMHNobOdR/dxgOob7mao7Z/bonh7kfpnLoSB7X47WsXe5Hr+5O5ZbZuTubl6ckc/zvPAIDrHl9rDtDn3ExyvPBp2nM/SCNuEQtzRvbsOgtb1q1sKvJpoj7hFw1Y2H//ve/UalULFy40NJVKVf6b8eJX74F72c6EDp/OHYhvlycuorCtKxS82cfvYhr+wcInjGI0NkxWHu5cHHqSgqSMwx54pf/TOahs9T+by/qvTcKj56tuf7xT2TsO2WuZplV+m/HiVu+BZ9nOlC3uA8vlNOHWcV9GDJjEHWL+/DCLX2oKAoXZ35Fflwadf7vGeotGI61jysXJq9El5tvzqaZTeqvf3Ft6Tb8+rejwaLnsA/x4dzk1RSU0YcaJ3t8+/6L+nOH0OC95/GMbMalhRvJOHjOkEeXW4Bj49oEDO1srmZYVFa2juaNbVk807tC+c9fKqDHs9fp9C97Dm2pw0vDXRkxPoHNO272+ZrvbjB+ahKTxntwYHMgzRrb0r3/NRKSCmuqGTVHV4WthqSkpDBw4EBcXFxwc3Nj2LBhZGZmlpn/woULqFSqUrevv/7akK+0x1evXl1zDSmDBCKVkJ9fvSf39evXs3fvXgICSv82dzdJ/m4v7t1a4B4Zjl0db/xHPYba1prUrbGl5q89vjcej7bEPtQP29peBIx5HHQKWYfPG/Jkn7yCa5dmODYNxsbXDY+oB7EL8SXn9DUztcq8km7rw4A79GHg+N543tKHtYr7MLO4D/OvpZBz6ioBo7rjEBZQ1M//fhRdfgFpvx43Y8vMJ/Hb/XhGheP5SHPs6nhTe3R31LZWpGw5XGp+52ZBuD3cALtAL2z93fF+ohX2IT5k/XXZkMejS1P8+rfHKTzYTK2wrO5dHZn+mie9Hy1/FETvo8/TCaljzdypXjSqb8Po59zo87gTCz9ON+RZ+FEazw90JaafC40b2PDhbG8c7FV8+uWNmmpGjdEvVjVlqykDBw7k+PHjbNmyhY0bN/Lrr78yYsSIMvMHBgZy/fp1o23atGk4OTnRvXt3o7yffvqpUb5evXrVWDvKIoFIOTp16sSYMWMYO3YsXl5eREVFcezYMbp3746TkxO+vr4MGjSIpKSkEmXGjBmDq6srXl5eTJo0CeW2g/Tq1av85z//YeXKlVhbW5u7aZWiK9CSc/Y6js1DDGkqtQrH5iHknLpSsX3kFaBodWic7Q1pDg1rc2P/3xQkZ6AoCllHLpB/NQWnFqHV3gZL0/eh02196NQ8hGwT+1ApKPq2qbK+OcOqUqtQWVmRfeJSNdb+7qAr0JJ95rpRwKBSq3AKDyHr5NU7llcUhRux58m7koLTA3VqsKb3l70Hcuna3t4orVsnB/YezAUgP1/h4JE8ozxqtYqu7R3YU5znnqJfI2LKVgNOnDjBpk2bWLp0Ka1bt6Zdu3YsXryY1atXc+1a6V/aNBoNfn5+Rtv69evp27cvTk7GAaibm5tRPjs780+nSSByBytWrMDGxoZdu3bx9ttv06VLF1q0aMGBAwfYtGkT8fHx9O3bt0QZKysr9u/fz6JFi5g/fz5Lly41PK7T6Rg0aBCvvPIKTZo0MXeTKk2bkQ06BSs34wPYys2RwtSyhwdvFf/5Nqw8nHFsfjPI8BsRjW2gN38/t4i/+szk4rRV+I+MxrFJULXW/25QnX3oVNyHtrW9sPZ2Jf5/29Fm5qAr0JK4dheFyRkUplRsn/cSfR9auzkapVu7OVKYWvrUDIA2K5cjT83hcK93ODftK2qN7IZzi5Ay8wtjcYlafL01Rmm+3hoybujIydGRlKJFq6XUPPEJ9+DUzF1mz549uLm50bJlS0NaZGQkarWaffv2VWgfBw8eJDY2lmHDhpV4bPTo0Xh5edGqVSuWL19e4kuzOchi1TsICwtj9uzZAMyYMYMWLVowc+ZMw+PLly8nMDCQv//+m/r16wNFw2ILFixApVLRoEEDjh49yoIFCxg+fDgA77zzDlZWVrz44osVrkdeXh55eXmGvzMyMsrJfXdJ/GYXGb8dJ/itwahtbh5yKRv/IOfUFer83zNY+7iSdfwS1z/aVPRhG37/jYpUReI3u0j/7Tght/ShykpDndee5up7GzgxcC6oVTg1D8Upop6s4L+F2t6WBu8OQ5tbQGbsBa4u24qNnxvOze6/gFdUgypeNXP7udnW1hZbW1uTqxMXF4ePj49RmpWVFR4eHsTFxVVoH8uWLaNRo0Y8/PDDRulvvvkmXbp0wcHBgZ9//pkXXniBzMzMSn02VQcJRO4gIuLmFRyHDx9mx44dJYa2AM6ePWsIRNq0aYNKpTI81rZtW+bNm4dWqyU2NpZFixZx6NAhozx3MmvWLKZNm1aFlphO4+IAahWFacbfsgvTsrByL3+eOWn9HpLW7SJ42rPYBd+8EkGXV0DCF9sJnNgX55ZhANgF+5J7Lo7kb/fed4FIVfswcd0uQm7rQwD7ev7UWzgCbVYuSqEWK1dHzr68DPt6d/+6o8rS9+HtC1ML0rKwcncso1TR9I1tgAcADqG+5F5JIuHr3RKIVJCft4b4RK1RWnyiFhdnNfb2ajQaFRoNpebx9bkHP2KqGIgEBgYaJU+ZMoWpU6eWyP7aa6/xzjvvlLvLEydOVL4et8nJyWHVqlVMmjSpxGO3prVo0YKsrCzmzJlj9kBEpmbuwNHx5gkuMzOTHj16EBsba7SdPn2aDh06VGh/v/32GwkJCdSpUwcrKyusrKy4ePEi48ePJzg4uMxyEydOJD093bBdvny5zLzVTW2twb6uP1lHLhjSFJ1C1pHz2DeoXWa5pHW7SfzqN4KmDMA+zPiDUdHqUAp1cFswptKoLTI0WNP0fZh5Wx9mHjmPQzl9mLhuNwlf/UZwKX14K42jHVaujuRdSybn7HWcW9evzurfFdTWGhzq+ZN5+IIhTdEpZB6+gGPD0i+BLpVOQVegvXM+AUCblnZs/z3HKG3rr9m0iShaS2BjoyKima1RHp1OYfvv2bSNuAcv363iVTOXL182OldPnDix1KcZP348J06cKHcLDQ3Fz8+PhIQEo7KFhYWkpKTg5+d3x+Z88803ZGdnM3jw4Dvmbd26NVeuXDEafTeHezBctZwHH3yQtWvXEhwcjJVV2V13+7zd3r17CQsLQ6PRMGjQICIjI40ej4qKYtCgQcTExJS5z6oO71WV5xNtuLroO+zr+WMfFkDyhv3ocgtwj2wOwJUF32Lt6Yzv4K4AJK7dReKqX6g9vjfWPm4UFK+DUNvZoLG3QeNgi8MDQcR/thW1jVXR1MyxS6TtOILfc49YrJ01yeuJNly5Qx9aeTrjd0sfJpTThwDpu/5C4+KAjbcruRcTuL50My6tG+Dcoq5lGlnDvHu14tKCDTiE+eNQP4DE74r60COyGQAX532Ptaez4VLc+K924xDmj42/G0qBlow/zpKy4xiBL0Qb9ll4I4f8xAwKk4uu8Mi7kgyAtbsj1ncYrboXZWbpOHO+wPD3hUuFxB7Lw8NNTZ3a1rz+VhJX47SsWFw0+jZysCvvL09nwvQkYvq5sGNXDl9/n8mG//kb9jF2pBsxLyUQ0dyWVuF2LPokjaxshaH9nM3evqqq6i3eXVxccHFxuUNu8Pb2xtv7zpdQt23blrS0NA4ePGgYod++fTs6nY7WrVvfsfyyZcvo2bNnhZ4rNjYWd3d3s3/WSCBSCaNHj+aTTz6hf//+vPrqq3h4eHDmzBlWr17N0qVL0WiKFmtdunSJcePGMXLkSA4dOsTixYuZN28eAJ6ennh6Gt9sytraGj8/Pxo0aGD2NlWUa/smFGZkk7Dql6KbcYX4EjRlgGHxZUFSBir1zdGN1E0HUQq1XH7nG6P9ePfrgE//jgDUfvlJEj7fzpX536LNzMHa2xWfZzvjHn1/3tCstD4MvqUP85My4JY+TCmnD32L+7AwJZPry7agTc/Eyt0Zt85N8e5bsdG5e5F7h8YUpmdz/YtfKUzNwj7Ul9A3nzEEDPmJxn2oy8vn8gebKEi+gdrGCtvangSN74l7h8aGPOn7TnN54UbD3xdnfwuAb/92+A+8//rywOFcuva5ebXF+KlFV/0N7uvMp4t8uZ6g5fLVm4FKSB1rNnzhz/gpSby7NI3a/lZ8PM+HqM43R4ufecKZpGQtU2enEJdYSHgTW35cFYCvt3zEVFWjRo2Ijo5m+PDhLFmyhIKCAsaMGUO/fv0Mt364evUqXbt25fPPP6dVq1aGsmfOnOHXX3/lxx9/LLHfDRs2EB8fT5s2bbCzs2PLli3MnDmTl19+2Wxt01Mp9+M4eDXp1KkT4eHhRjcbO336NBMmTGDHjh3k5eURFBREdHQ08+fPR6VS0alTJ5o0aYJOp2PVqlVoNBpGjRrFjBkzylwTEhwczNixYxk7dmyF65aRkYGrqysNv3wVjYPlRkrudRVfpSPKYqWRaY6qOhjxlaWrcE/LuKHDvf450tPTKzQaUeo+is+pkWH/xUpT+XNqoTaPracXVKkOZUlJSWHMmDFs2LABtVpNnz59ePfddw3rFS9cuEBISAg7duygU6dOhnKvv/46X3zxBRcuXECtNl6JsWnTJiZOnMiZM2dQFIV69eoxatQohg8fXiJvTZNApJqVFrzUBAlEqocEIlUngUjVSSBSNdUaiNQda3ogcnZhjQQi9zsZNxNCCCH05EfvzE4CESGEEMLA1LukSiBiKglEqtnOnTstXQUhhBDiniGBiBBCCKEnUzNmJ4GIEEIIoadTMGmaRSeBiKkkEBFCCCH0FF3RZko5YRIJRIQQQgg9mZoxO/mtGSGEEEJYjIyICCGEEHqyRsTsJBARQggh9GRqxuwkEBFCCCH0FEwMRKq9Jv8YskZECCGEEBYjIyJCCCGEnkzNmJ0EIkIIIYSeTgeYcE8QndxHxFQSiAghhBB6MiJidhKICCGEEHoSiJidLFYVQgghhMXIiIgQQgihJzc0MzsJRIQQQohiiqJDMeEH7EwpI4pIICKEEELoKYppoxuyRsRkEogIIYQQeoqJUzMSiJhMFqsKIYQQwmJkREQIIYTQ0+lAZcJ6D1kjYjIJRIQQQgg9mZoxOwlEhBBCiGKKTodiwoiIXDVjOglEhBBCCD0ZETE7WawqhBBCCIuREREhhBBCT6eASkZEzEkCESGEEEJPUQBTrpqRQMRUEogIIYQQxRSdgmLCiIgigYjJZI2IEEIIoafoTN9qyFtvvcXDDz+Mg4MDbm5uFWuGojB58mT8/f2xt7cnMjKS06dPG+VJSUlh4MCBuLi44ObmxrBhw8jMzKyBFpRPAhEhhBDiLpafn8/TTz/NqFGjKlxm9uzZvPvuuyxZsoR9+/bh6OhIVFQUubm5hjwDBw7k+PHjbNmyhY0bN/Lrr78yYsSImmhCuWRqRgghhCh2N07NTJs2DYDPPvuswnVZuHAhb7zxBk888QQAn3/+Ob6+vnz77bf069ePEydOsGnTJv744w9atmwJwOLFi3n00UeZO3cuAQEBNdKW0kggco/SH/Ta7DwL1+TeprJ0Be4DKo3W0lW452XckJthVUVGZlH/VUcwUKjkmTTNUkhBUV0yMozSbW1tsbW1rXK9KuP8+fPExcURGRlpSHN1daV169bs2bOHfv36sWfPHtzc3AxBCEBkZCRqtZp9+/bRu3dvs9VXApF71I0bNwA4PWyRhWsihKgqd0tX4D5x48YNXF1dTSprY2ODn58fv8f9aPLzOzk5ERgYaJQ2ZcoUpk6davI+TREXFweAr6+vUbqvr6/hsbi4OHx8fIwet7KywsPDw5DHXCQQuUcFBARw+fJlnJ2dUanuzu/1GRkZBAYGcvnyZVxcXCxdnXuS9GHVSR9W3d3eh4qicOPGjSpNJ9jZ2XH+/Hny8/OrVI/bz8dljYa89tprvPPOO+Xu78SJEzRs2NDk+twrJBC5R6nVamrXrm3palSIi4vLXXnyupdIH1ad9GHV3c19aOpIyK3s7Oyws7Orhtrc2fjx4xk6dGi5eUJDQ03at5+fHwDx8fH4+/sb0uPj4wkPDzfkSUhIMCpXWFhISkqKoby5SCAihBBCmJm3tzfe3t41su+QkBD8/PzYtm2bIfDIyMhg3759hitv2rZtS1paGgcPHiQiIgKA7du3o9PpaN26dY3Uqyxy+a4QQghxF7t06RKxsbFcunQJrVZLbGwssbGxRvf8aNiwIevXrwdApVIxduxYZsyYwffff8/Ro0cZPHgwAQEB9OrVC4BGjRoRHR3N8OHD2b9/P7t27WLMmDH069fPrFfMgIyIiBpka2vLlClTzL5i/H4ifVh10odVJ31oWZMnT2bFihWGv1u0aAHAjh076NSpEwCnTp0iPT3dkOfVV18lKyuLESNGkJaWRrt27di0aZPR1NPKlSsZM2YMXbt2Ra1W06dPH959913zNOoWKkXuSyuEEEIIC5GpGSGEEEJYjAQiQgghhLAYCUSEEEIIYTESiAghhBDCYiQQERazbt06unXrhqenJyqVitjYWEtX6Z5SUFDAhAkTaNq0KY6OjgQEBDB48GCuXbtm6ardU6ZOnUrDhg1xdHTE3d2dyMhI9u3bZ+lq3bP+/e9/o1KpWLhwoaWrIu4REoiISqvKLZBvlZWVRbt27e54m+P7UXX0YXZ2NocOHWLSpEkcOnSIdevWcerUKXr27FkNNbz7VddxWL9+fd577z2OHj3K77//TnBwMN26dSMxMbFa9n83q64+1Fu/fj179+41+30oxD1OEeIOOnbsqIwePVp56aWXFE9PT6VTp07K0aNHlejoaMXR0VHx8fFRnn32WSUxMbFEmdGjRysuLi6Kp6en8sYbbyg6na7E/s+fP68Ayp9//mnGVplXTfeh3v79+xVAuXjxojmaZVbm6sP09HQFULZu3WqOZplVTfbhlStXlFq1ainHjh1TgoKClAULFpi5deJeJSMiokJWrFiBjY0Nu3bt4u2336ZLly60aNGCAwcOsGnTJuLj4+nbt2+JMlZWVuzfv59FixYxf/58li5daqEWWJ45+jA9PR2VSoWbm1sNt8YyaroP8/Pz+fjjj3F1daV58+bmaJLZ1UQf6nQ6Bg0axCuvvEKTJk3M3SRxr7N0JCTufh07dlRatGhh+Hv69OlKt27djPJcvnxZAZRTp04ZyjRq1MjoW9OECROURo0aldj/P2VEpCb7UFEUJScnR3nwwQeVAQMG1EALLK8m+3DDhg2Ko6OjolKplICAAGX//v012BLLqak+nDlzpvLII48Y8siIiKgMGRERFaL/USSAw4cPs2PHDpycnAyb/qeqz549a8jXpk0bo5/Ebtu2LadPn0ar1Zqv4neRmuzDgoIC+vbti6IofPjhhzXcEsupqT7s3LkzsbGx7N69m+joaPr27Vvil0nvF9XdhwcPHmTRokV89tlnRnmEqCj5rRlRIY6Ojob/Z2Zm0qNHj1IXmd76k9PCWE31oT4IuXjxItu3b79rf6a9OtRUHzo6OlKvXj3q1atHmzZtCAsLY9myZUycOLHKdb7bVHcf/vbbbyQkJFCnTh1DmlarZfz48SxcuJALFy5Uuc7i/iaBiKi0Bx98kLVr1xIcHIyVVdmH0O2XQO7du5ewsDA0Gk1NV/GuV119qA9CTp8+zY4dO/D09KzRet9NavI41Ol05OXlVVtd71bV0YeDBg0iMjLS6PGoqCgGDRpETExMjdRb3F9kakZU2ujRo0lJSaF///788ccfnD17ls2bNxMTE2M03H3p0iXGjRvHqVOn+PLLL1m8eDEvvfSS4fGUlBRiY2P566+/gKJfj4yNjSUuLs7sbTK36ujDgoICnnrqKQ4cOMDKlSvRarXExcURFxdX7Zdl3o2qow+zsrJ4/fXX2bt3LxcvXuTgwYM899xzXL16laefftpSTTOb6uhDT09PHnjgAaPN2toaPz8/GjRoYKmmiXuIjIiISgsICGDXrl1MmDCBbt26kZeXR1BQENHR0ajVN2PbwYMHk5OTQ6tWrdBoNLz00kuMGDHC8Pj3339v9I2pX79+AEyZMoWpU6earT2WUB19ePXqVb7//nsAwsPDjfZ/68+D36+qow81Gg0nT55kxYoVJCUl4enpyUMPPcRvv/32j7j6o7rey0JUhUpRFMXSlRD3n06dOhEeHi53V6wC6cOqkz6sOulDUdNkakYIIYQQFiOBiBBCCCEsRqZmhBBCCGExMiIihBBCCIuRQEQIIYQQFiOBiBBCCCEsRgIRIYQQQliMBCJCCCGEsBgJRIQQQghhMRKICCGEEMJiJBARQgghhMVIICKEEEIIi/l/pBtfsX34TqkAAAAASUVORK5CYII=",
      "text/plain": [
       "<Figure size 500x400 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1300x800 with 6 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>split</th>\n",
       "      <th>replicate_pair</th>\n",
       "      <th>n</th>\n",
       "      <th>pearson</th>\n",
       "      <th>spearman</th>\n",
       "      <th>mse_between_reps</th>\n",
       "      <th>mae_between_reps</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>train</td>\n",
       "      <td>element_log2_ratio_rep1 vs element_log2_ratio_...</td>\n",
       "      <td>23484</td>\n",
       "      <td>0.620345</td>\n",
       "      <td>0.323971</td>\n",
       "      <td>0.186555</td>\n",
       "      <td>0.336141</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>train</td>\n",
       "      <td>element_log2_ratio_rep1 vs element_log2_ratio_...</td>\n",
       "      <td>23484</td>\n",
       "      <td>0.577623</td>\n",
       "      <td>0.290921</td>\n",
       "      <td>0.214509</td>\n",
       "      <td>0.360906</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>train</td>\n",
       "      <td>element_log2_ratio_rep1 vs element_log2_ratio_...</td>\n",
       "      <td>23484</td>\n",
       "      <td>0.562059</td>\n",
       "      <td>0.277453</td>\n",
       "      <td>0.231611</td>\n",
       "      <td>0.376148</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>train</td>\n",
       "      <td>element_log2_ratio_rep2 vs element_log2_ratio_...</td>\n",
       "      <td>23484</td>\n",
       "      <td>0.588821</td>\n",
       "      <td>0.301407</td>\n",
       "      <td>0.208986</td>\n",
       "      <td>0.356526</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>train</td>\n",
       "      <td>element_log2_ratio_rep2 vs element_log2_ratio_...</td>\n",
       "      <td>23484</td>\n",
       "      <td>0.568883</td>\n",
       "      <td>0.288984</td>\n",
       "      <td>0.228450</td>\n",
       "      <td>0.372049</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>train</td>\n",
       "      <td>element_log2_ratio_rep3 vs element_log2_ratio_...</td>\n",
       "      <td>23484</td>\n",
       "      <td>0.579602</td>\n",
       "      <td>0.309167</td>\n",
       "      <td>0.228948</td>\n",
       "      <td>0.373327</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   split                                     replicate_pair      n   pearson  \\\n",
       "0  train  element_log2_ratio_rep1 vs element_log2_ratio_...  23484  0.620345   \n",
       "1  train  element_log2_ratio_rep1 vs element_log2_ratio_...  23484  0.577623   \n",
       "2  train  element_log2_ratio_rep1 vs element_log2_ratio_...  23484  0.562059   \n",
       "3  train  element_log2_ratio_rep2 vs element_log2_ratio_...  23484  0.588821   \n",
       "4  train  element_log2_ratio_rep2 vs element_log2_ratio_...  23484  0.568883   \n",
       "5  train  element_log2_ratio_rep3 vs element_log2_ratio_...  23484  0.579602   \n",
       "\n",
       "   spearman  mse_between_reps  mae_between_reps  \n",
       "0  0.323971          0.186555          0.336141  \n",
       "1  0.290921          0.214509          0.360906  \n",
       "2  0.277453          0.231611          0.376148  \n",
       "3  0.301407          0.208986          0.356526  \n",
       "4  0.288984          0.228450          0.372049  \n",
       "5  0.309167          0.228948          0.373327  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot replicate correlations for the exact dataset used to train the model.\n",
    "# Run this after creating training_seq_score_consistent_from_alignment_zscore_split.tsv.\n",
    "\n",
    "from pathlib import Path\n",
    "from itertools import combinations\n",
    "\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "from scipy.stats import pearsonr, spearmanr\n",
    "\n",
    "MODEL_TSV = Path(\"training_seq_score_consistent_from_alignment_zscore_split.tsv\")\n",
    "READOUT_TSV = Path(\"element_level_readout_consistent.tsv\")\n",
    "\n",
    "# Change this to \"val\", \"test\", or None if you want a different subset.\n",
    "SPLIT_TO_PLOT = \"train\"\n",
    "\n",
    "assert MODEL_TSV.exists(), f\"Missing {MODEL_TSV}. Run the split/z-score cell first.\"\n",
    "assert READOUT_TSV.exists(), f\"Missing {READOUT_TSV}. Run the consistent-label readout cell first.\"\n",
    "\n",
    "model_df = pd.read_csv(MODEL_TSV, sep=\"\\t\")\n",
    "readout_df = pd.read_csv(READOUT_TSV, sep=\"\\t\")\n",
    "\n",
    "required_model_cols = {\"element_id\", \"split\"}\n",
    "missing_model_cols = required_model_cols - set(model_df.columns)\n",
    "assert not missing_model_cols, f\"{MODEL_TSV} missing columns: {missing_model_cols}\"\n",
    "\n",
    "rep_cols = [f\"element_log2_ratio_rep{i}\" for i in range(1, 5)]\n",
    "missing_rep_cols = [c for c in rep_cols if c not in readout_df.columns]\n",
    "assert not missing_rep_cols, f\"{READOUT_TSV} missing replicate columns: {missing_rep_cols}\"\n",
    "\n",
    "if SPLIT_TO_PLOT is None:\n",
    "    subset_ids = set(model_df[\"element_id\"].astype(str))\n",
    "    subset_name = \"all splits\"\n",
    "else:\n",
    "    subset_ids = set(\n",
    "        model_df.loc[\n",
    "            model_df[\"split\"].astype(str).str.lower() == SPLIT_TO_PLOT.lower(),\n",
    "            \"element_id\",\n",
    "        ].astype(str)\n",
    "    )\n",
    "    subset_name = SPLIT_TO_PLOT\n",
    "\n",
    "plot_df = readout_df[readout_df[\"element_id\"].astype(str).isin(subset_ids)].copy()\n",
    "plot_df = plot_df.dropna(subset=rep_cols)\n",
    "\n",
    "print(f\"Subset: {subset_name}\")\n",
    "print(f\"Elements in model TSV subset: {len(subset_ids):,}\")\n",
    "print(f\"Elements with replicate readouts: {len(plot_df):,}\")\n",
    "\n",
    "assert len(plot_df) >= 3, \"Too few elements to compute replicate correlations.\"\n",
    "\n",
    "# Pearson and Spearman correlation matrices\n",
    "pearson_corr = plot_df[rep_cols].corr(method=\"pearson\")\n",
    "spearman_corr = plot_df[rep_cols].corr(method=\"spearman\")\n",
    "\n",
    "display(pd.DataFrame(pearson_corr))\n",
    "display(pd.DataFrame(spearman_corr))\n",
    "\n",
    "# Heatmap helper\n",
    "def plot_corr_heatmap(corr, title):\n",
    "    fig, ax = plt.subplots(figsize=(5, 4))\n",
    "    im = ax.imshow(corr.values, vmin=-1, vmax=1)\n",
    "    fig.colorbar(im, ax=ax, fraction=0.046, pad=0.04)\n",
    "\n",
    "    ax.set_xticks(range(len(rep_cols)))\n",
    "    ax.set_yticks(range(len(rep_cols)))\n",
    "    ax.set_xticklabels([f\"rep{i}\" for i in range(1, 5)])\n",
    "    ax.set_yticklabels([f\"rep{i}\" for i in range(1, 5)])\n",
    "    ax.set_title(title)\n",
    "\n",
    "    for i in range(len(rep_cols)):\n",
    "        for j in range(len(rep_cols)):\n",
    "            ax.text(j, i, f\"{corr.iloc[i, j]:.2f}\", ha=\"center\", va=\"center\")\n",
    "\n",
    "    plt.tight_layout()\n",
    "    plt.show()\n",
    "\n",
    "plot_corr_heatmap(\n",
    "    pearson_corr,\n",
    "    f\"Pearson correlation among replicate stability readouts ({subset_name})\",\n",
    ")\n",
    "\n",
    "plot_corr_heatmap(\n",
    "    spearman_corr,\n",
    "    f\"Spearman correlation among replicate stability readouts ({subset_name})\",\n",
    ")\n",
    "\n",
    "# Pairwise scatter plots\n",
    "pairs = list(combinations(rep_cols, 2))\n",
    "n_pairs = len(pairs)\n",
    "\n",
    "fig, axes = plt.subplots(2, 3, figsize=(13, 8))\n",
    "axes = axes.flatten()\n",
    "\n",
    "for ax, (xcol, ycol) in zip(axes, pairs):\n",
    "    x = plot_df[xcol].to_numpy(dtype=float)\n",
    "    y = plot_df[ycol].to_numpy(dtype=float)\n",
    "\n",
    "    pr = pearsonr(x, y)[0] if np.std(x) > 1e-8 and np.std(y) > 1e-8 else np.nan\n",
    "    sr = spearmanr(x, y)[0] if np.std(x) > 1e-8 and np.std(y) > 1e-8 else np.nan\n",
    "\n",
    "    ax.scatter(x, y, s=8, alpha=0.35)\n",
    "    ax.axline((0, 0), slope=1, linestyle=\"--\", linewidth=1)\n",
    "    ax.set_xlabel(xcol.replace(\"element_log2_ratio_\", \"\"))\n",
    "    ax.set_ylabel(ycol.replace(\"element_log2_ratio_\", \"\"))\n",
    "    ax.set_title(f\"Pearson={pr:.2f}, Spearman={sr:.2f}\")\n",
    "\n",
    "for ax in axes[n_pairs:]:\n",
    "    ax.axis(\"off\")\n",
    "\n",
    "plt.suptitle(f\"Pairwise replicate agreement for model {subset_name} set\", y=1.02)\n",
    "plt.tight_layout()\n",
    "plt.show()\n",
    "\n",
    "# Optional: save a compact numeric summary\n",
    "summary_rows = []\n",
    "for xcol, ycol in pairs:\n",
    "    x = plot_df[xcol].to_numpy(dtype=float)\n",
    "    y = plot_df[ycol].to_numpy(dtype=float)\n",
    "    summary_rows.append({\n",
    "        \"split\": subset_name,\n",
    "        \"replicate_pair\": f\"{xcol} vs {ycol}\",\n",
    "        \"n\": len(plot_df),\n",
    "        \"pearson\": pearsonr(x, y)[0] if np.std(x) > 1e-8 and np.std(y) > 1e-8 else np.nan,\n",
    "        \"spearman\": spearmanr(x, y)[0] if np.std(x) > 1e-8 and np.std(y) > 1e-8 else np.nan,\n",
    "        \"mse_between_reps\": float(np.mean((x - y) ** 2)),\n",
    "        \"mae_between_reps\": float(np.mean(np.abs(x - y))),\n",
    "    })\n",
    "\n",
    "replicate_corr_summary = pd.DataFrame(summary_rows)\n",
    "replicate_corr_summary.to_csv(f\"replicate_correlation_{subset_name}.tsv\", sep=\"\\t\", index=False)\n",
    "display(replicate_corr_summary)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "67c130c4-a381-4649-a9e4-9efb722bf1cf",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Base rows: 29355\n",
      "Columns after merge:\n",
      "['element_id', 'seq', 'score', 'score_median', 'score_sd_across_reps_x', 'n_qc_barcodes_x', 'rna_total_x', 'dna_total_x', 'median_barcode_ratio_sd_x', 'score_raw', 'element_log2_ratio_rep1', 'element_log2_ratio_rep2', 'element_log2_ratio_rep3', 'element_log2_ratio_rep4', 'score_sd_across_reps_y', 'score_range_across_reps', 'median_barcode_ratio_sd_y', 'n_qc_barcodes_y', 'rna_total_y', 'dna_total_y']\n",
      "No split column found. Creating train/val/test split.\n",
      "split\n",
      "train    23484\n",
      "test      2936\n",
      "val       2935\n",
      "Name: count, dtype: int64\n",
      "Exported: training_seq_score_extreme_weighted.tsv\n",
      "Stats: training_seq_score_extreme_weighted_stats.json\n",
      "{\n",
      "  \"input_base_tsv\": \"training_seq_score_consistent_from_alignment.tsv\",\n",
      "  \"input_replicate_tsv\": \"element_level_readout_consistent.tsv\",\n",
      "  \"output_tsv\": \"training_seq_score_extreme_weighted.tsv\",\n",
      "  \"n_total\": 29355,\n",
      "  \"split_counts\": {\n",
      "    \"train\": 23484,\n",
      "    \"test\": 2936,\n",
      "    \"val\": 2935\n",
      "  },\n",
      "  \"score_raw_train_mean\": -0.023867485882891677,\n",
      "  \"score_raw_train_std\": 0.42136155452796287,\n",
      "  \"q10_z\": -0.7895845387373244,\n",
      "  \"q30_z\": -0.13768336498012324,\n",
      "  \"q70_z\": 0.43175982929251244,\n",
      "  \"q90_z\": 0.8113083814767235,\n",
      "  \"extreme_alpha\": 0.75,\n",
      "  \"extreme_max_weight\": 3.0,\n",
      "  \"reliability_max_weight\": 2.0,\n",
      "  \"final_min_weight\": 0.5,\n",
      "  \"final_max_weight\": 4.0,\n",
      "  \"train_sample_weight_mean\": 1.0015832504705322,\n",
      "  \"train_sample_weight_min\": 0.5,\n",
      "  \"train_sample_weight_max\": 2.8585794716672672\n",
      "}\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>element_id</th>\n",
       "      <th>seq</th>\n",
       "      <th>score</th>\n",
       "      <th>score_z</th>\n",
       "      <th>score_raw</th>\n",
       "      <th>split</th>\n",
       "      <th>sample_weight</th>\n",
       "      <th>extreme_weight</th>\n",
       "      <th>reliability_weight</th>\n",
       "      <th>stability_bin_5</th>\n",
       "      <th>is_extreme_bottom10</th>\n",
       "      <th>is_extreme_top10</th>\n",
       "      <th>is_extreme_abs20</th>\n",
       "      <th>rep_sd</th>\n",
       "      <th>rep_range</th>\n",
       "      <th>element_log2_ratio_rep1</th>\n",
       "      <th>element_log2_ratio_rep2</th>\n",
       "      <th>element_log2_ratio_rep3</th>\n",
       "      <th>element_log2_ratio_rep4</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>TILE_ID_001-00001|ROTAVIRUS_A|NC_011505.2|1,130</td>\n",
       "      <td>ACTGGCCGCTTCACTGGGCTTTTAAAGCGCTACAGTGATGTCTCTC...</td>\n",
       "      <td>-0.229060</td>\n",
       "      <td>-0.229060</td>\n",
       "      <td>-0.120385</td>\n",
       "      <td>train</td>\n",
       "      <td>0.971187</td>\n",
       "      <td>1.171795</td>\n",
       "      <td>1.159740</td>\n",
       "      <td>1</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>0.201078</td>\n",
       "      <td>0.489855</td>\n",
       "      <td>-0.093401</td>\n",
       "      <td>-0.144600</td>\n",
       "      <td>-0.366697</td>\n",
       "      <td>0.123159</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>TILE_ID_001-00002|ROTAVIRUS_A|NC_011505.2|66,195</td>\n",
       "      <td>ACTGGCCGCTTCACTGAACTATATATAAGAATGAATCGTCTTCAAC...</td>\n",
       "      <td>0.094683</td>\n",
       "      <td>0.094683</td>\n",
       "      <td>0.016028</td>\n",
       "      <td>train</td>\n",
       "      <td>1.010578</td>\n",
       "      <td>1.071012</td>\n",
       "      <td>1.320338</td>\n",
       "      <td>2</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>0.071457</td>\n",
       "      <td>0.150029</td>\n",
       "      <td>-0.027265</td>\n",
       "      <td>-0.020100</td>\n",
       "      <td>0.122765</td>\n",
       "      <td>-0.011288</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>TILE_ID_001-00003|ROTAVIRUS_A|NC_011505.2|131,260</td>\n",
       "      <td>ACTGGCCGCTTCACTGGTGAACAGTACATTTCACCAGATGCAGAAG...</td>\n",
       "      <td>-0.261223</td>\n",
       "      <td>-0.261223</td>\n",
       "      <td>-0.133937</td>\n",
       "      <td>train</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>1.195917</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>1</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>0.691003</td>\n",
       "      <td>1.621130</td>\n",
       "      <td>0.159878</td>\n",
       "      <td>-1.065968</td>\n",
       "      <td>0.555162</td>\n",
       "      <td>-0.184819</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>TILE_ID_001-00004|ROTAVIRUS_A|NC_011505.2|196,325</td>\n",
       "      <td>ACTGGCCGCTTCACTGGATATTGGACCATCTGATTCTGCTTCAAAC...</td>\n",
       "      <td>0.227368</td>\n",
       "      <td>0.227368</td>\n",
       "      <td>0.071937</td>\n",
       "      <td>test</td>\n",
       "      <td>0.804590</td>\n",
       "      <td>1.170526</td>\n",
       "      <td>0.961841</td>\n",
       "      <td>2</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>0.316527</td>\n",
       "      <td>0.761130</td>\n",
       "      <td>-0.314965</td>\n",
       "      <td>0.004966</td>\n",
       "      <td>0.446165</td>\n",
       "      <td>0.151581</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>TILE_ID_001-00005|ROTAVIRUS_A|NC_011505.2|261,390</td>\n",
       "      <td>ACTGGCCGCTTCACTGAGTTAAGACAAATGCAGACGCTGGCGTGTC...</td>\n",
       "      <td>0.425975</td>\n",
       "      <td>0.425975</td>\n",
       "      <td>0.155622</td>\n",
       "      <td>train</td>\n",
       "      <td>0.720262</td>\n",
       "      <td>1.319481</td>\n",
       "      <td>0.763830</td>\n",
       "      <td>2</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>0.437001</td>\n",
       "      <td>0.940500</td>\n",
       "      <td>-0.100975</td>\n",
       "      <td>0.072137</td>\n",
       "      <td>0.795913</td>\n",
       "      <td>-0.144587</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                          element_id  \\\n",
       "0    TILE_ID_001-00001|ROTAVIRUS_A|NC_011505.2|1,130   \n",
       "1   TILE_ID_001-00002|ROTAVIRUS_A|NC_011505.2|66,195   \n",
       "2  TILE_ID_001-00003|ROTAVIRUS_A|NC_011505.2|131,260   \n",
       "3  TILE_ID_001-00004|ROTAVIRUS_A|NC_011505.2|196,325   \n",
       "4  TILE_ID_001-00005|ROTAVIRUS_A|NC_011505.2|261,390   \n",
       "\n",
       "                                                 seq     score   score_z  \\\n",
       "0  ACTGGCCGCTTCACTGGGCTTTTAAAGCGCTACAGTGATGTCTCTC... -0.229060 -0.229060   \n",
       "1  ACTGGCCGCTTCACTGAACTATATATAAGAATGAATCGTCTTCAAC...  0.094683  0.094683   \n",
       "2  ACTGGCCGCTTCACTGGTGAACAGTACATTTCACCAGATGCAGAAG... -0.261223 -0.261223   \n",
       "3  ACTGGCCGCTTCACTGGATATTGGACCATCTGATTCTGCTTCAAAC...  0.227368  0.227368   \n",
       "4  ACTGGCCGCTTCACTGAGTTAAGACAAATGCAGACGCTGGCGTGTC...  0.425975  0.425975   \n",
       "\n",
       "   score_raw  split  sample_weight  extreme_weight  reliability_weight  \\\n",
       "0  -0.120385  train       0.971187        1.171795            1.159740   \n",
       "1   0.016028  train       1.010578        1.071012            1.320338   \n",
       "2  -0.133937  train       0.500000        1.195917            0.500000   \n",
       "3   0.071937   test       0.804590        1.170526            0.961841   \n",
       "4   0.155622  train       0.720262        1.319481            0.763830   \n",
       "\n",
       "   stability_bin_5  is_extreme_bottom10  is_extreme_top10  is_extreme_abs20  \\\n",
       "0                1                False             False             False   \n",
       "1                2                False             False             False   \n",
       "2                1                False             False             False   \n",
       "3                2                False             False             False   \n",
       "4                2                False             False             False   \n",
       "\n",
       "     rep_sd  rep_range  element_log2_ratio_rep1  element_log2_ratio_rep2  \\\n",
       "0  0.201078   0.489855                -0.093401                -0.144600   \n",
       "1  0.071457   0.150029                -0.027265                -0.020100   \n",
       "2  0.691003   1.621130                 0.159878                -1.065968   \n",
       "3  0.316527   0.761130                -0.314965                 0.004966   \n",
       "4  0.437001   0.940500                -0.100975                 0.072137   \n",
       "\n",
       "   element_log2_ratio_rep3  element_log2_ratio_rep4  \n",
       "0                -0.366697                 0.123159  \n",
       "1                 0.122765                -0.011288  \n",
       "2                 0.555162                -0.184819  \n",
       "3                 0.446165                 0.151581  \n",
       "4                 0.795913                -0.144587  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>split</th>\n",
       "      <th>stability_bin_5</th>\n",
       "      <th>n</th>\n",
       "      <th>score_mean</th>\n",
       "      <th>weight_mean</th>\n",
       "      <th>weight_min</th>\n",
       "      <th>weight_max</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>test</td>\n",
       "      <td>0</td>\n",
       "      <td>308</td>\n",
       "      <td>-2.169028</td>\n",
       "      <td>1.440809</td>\n",
       "      <td>0.618144</td>\n",
       "      <td>2.697390</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>test</td>\n",
       "      <td>1</td>\n",
       "      <td>593</td>\n",
       "      <td>-0.380038</td>\n",
       "      <td>0.914192</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>1.497280</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>test</td>\n",
       "      <td>2</td>\n",
       "      <td>1151</td>\n",
       "      <td>0.159283</td>\n",
       "      <td>0.844689</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>1.257056</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>test</td>\n",
       "      <td>3</td>\n",
       "      <td>592</td>\n",
       "      <td>0.605588</td>\n",
       "      <td>1.056302</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>1.522272</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>test</td>\n",
       "      <td>4</td>\n",
       "      <td>292</td>\n",
       "      <td>1.081452</td>\n",
       "      <td>1.267277</td>\n",
       "      <td>0.576767</td>\n",
       "      <td>2.166564</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>train</td>\n",
       "      <td>0</td>\n",
       "      <td>2349</td>\n",
       "      <td>-2.133106</td>\n",
       "      <td>1.444109</td>\n",
       "      <td>0.570292</td>\n",
       "      <td>2.858579</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>train</td>\n",
       "      <td>1</td>\n",
       "      <td>4696</td>\n",
       "      <td>-0.392609</td>\n",
       "      <td>0.909599</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>1.518908</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>train</td>\n",
       "      <td>2</td>\n",
       "      <td>9394</td>\n",
       "      <td>0.160702</td>\n",
       "      <td>0.837223</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>1.262303</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>train</td>\n",
       "      <td>3</td>\n",
       "      <td>4696</td>\n",
       "      <td>0.597637</td>\n",
       "      <td>1.063832</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>1.526679</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>train</td>\n",
       "      <td>4</td>\n",
       "      <td>2349</td>\n",
       "      <td>1.080552</td>\n",
       "      <td>1.275805</td>\n",
       "      <td>0.576146</td>\n",
       "      <td>2.534974</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>val</td>\n",
       "      <td>0</td>\n",
       "      <td>298</td>\n",
       "      <td>-2.050681</td>\n",
       "      <td>1.447938</td>\n",
       "      <td>0.595173</td>\n",
       "      <td>2.858579</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>val</td>\n",
       "      <td>1</td>\n",
       "      <td>624</td>\n",
       "      <td>-0.386651</td>\n",
       "      <td>0.906535</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>1.472354</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>val</td>\n",
       "      <td>2</td>\n",
       "      <td>1156</td>\n",
       "      <td>0.162156</td>\n",
       "      <td>0.837271</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>1.249064</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>val</td>\n",
       "      <td>3</td>\n",
       "      <td>566</td>\n",
       "      <td>0.603415</td>\n",
       "      <td>1.057468</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>1.494906</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>val</td>\n",
       "      <td>4</td>\n",
       "      <td>291</td>\n",
       "      <td>1.084504</td>\n",
       "      <td>1.276357</td>\n",
       "      <td>0.581203</td>\n",
       "      <td>2.039716</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    split  stability_bin_5     n  score_mean  weight_mean  weight_min  \\\n",
       "0    test                0   308   -2.169028     1.440809    0.618144   \n",
       "1    test                1   593   -0.380038     0.914192    0.500000   \n",
       "2    test                2  1151    0.159283     0.844689    0.500000   \n",
       "3    test                3   592    0.605588     1.056302    0.500000   \n",
       "4    test                4   292    1.081452     1.267277    0.576767   \n",
       "5   train                0  2349   -2.133106     1.444109    0.570292   \n",
       "6   train                1  4696   -0.392609     0.909599    0.500000   \n",
       "7   train                2  9394    0.160702     0.837223    0.500000   \n",
       "8   train                3  4696    0.597637     1.063832    0.500000   \n",
       "9   train                4  2349    1.080552     1.275805    0.576146   \n",
       "10    val                0   298   -2.050681     1.447938    0.595173   \n",
       "11    val                1   624   -0.386651     0.906535    0.500000   \n",
       "12    val                2  1156    0.162156     0.837271    0.500000   \n",
       "13    val                3   566    0.603415     1.057468    0.500000   \n",
       "14    val                4   291    1.084504     1.276357    0.581203   \n",
       "\n",
       "    weight_max  \n",
       "0     2.697390  \n",
       "1     1.497280  \n",
       "2     1.257056  \n",
       "3     1.522272  \n",
       "4     2.166564  \n",
       "5     2.858579  \n",
       "6     1.518908  \n",
       "7     1.262303  \n",
       "8     1.526679  \n",
       "9     2.534974  \n",
       "10    2.858579  \n",
       "11    1.472354  \n",
       "12    1.249064  \n",
       "13    1.494906  \n",
       "14    2.039716  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import json\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "from pathlib import Path\n",
    "from sklearn.model_selection import GroupShuffleSplit\n",
    "\n",
    "# ============================================================\n",
    "# Export weighted dataset for extreme-aware stability training\n",
    "# ============================================================\n",
    "\n",
    "# Input files\n",
    "BASE_TSV = Path(\"training_seq_score_consistent_from_alignment.tsv\")\n",
    "REPLICATE_TSV = Path(\"element_level_readout_consistent.tsv\")\n",
    "\n",
    "# Output files\n",
    "OUT_TSV = Path(\"training_seq_score_extreme_weighted.tsv\")\n",
    "OUT_STATS_JSON = Path(\"training_seq_score_extreme_weighted_stats.json\")\n",
    "\n",
    "RANDOM_STATE = 42\n",
    "\n",
    "# Weighting knobs\n",
    "EXTREME_ALPHA = 0.75          # how much to upweight large |z|\n",
    "EXTREME_MAX_WEIGHT = 3.0      # cap for extreme weighting alone\n",
    "RELIABILITY_MAX_WEIGHT = 2.0  # cap for reliability weighting alone\n",
    "FINAL_MAX_WEIGHT = 4.0        # final sample weight cap\n",
    "FINAL_MIN_WEIGHT = 0.5\n",
    "\n",
    "# Columns expected from replicate-level readout\n",
    "REP_COLS = [\n",
    "    \"element_log2_ratio_rep1\",\n",
    "    \"element_log2_ratio_rep2\",\n",
    "    \"element_log2_ratio_rep3\",\n",
    "    \"element_log2_ratio_rep4\",\n",
    "]\n",
    "\n",
    "# -----------------------------\n",
    "# 1. Load base consistent dataset\n",
    "# -----------------------------\n",
    "df = pd.read_csv(BASE_TSV, sep=\"\\t\")\n",
    "\n",
    "required = [\"element_id\", \"seq\", \"score\"]\n",
    "missing = [c for c in required if c not in df.columns]\n",
    "if missing:\n",
    "    raise ValueError(f\"Missing required columns in {BASE_TSV}: {missing}\")\n",
    "\n",
    "df = df.copy()\n",
    "df[\"seq\"] = (\n",
    "    df[\"seq\"]\n",
    "    .astype(str)\n",
    "    .str.upper()\n",
    "    .str.replace(\"U\", \"T\", regex=False)\n",
    "    .str.replace(\"[^ACGT]\", \"\", regex=True)\n",
    ")\n",
    "\n",
    "df[\"score_raw\"] = pd.to_numeric(df[\"score\"], errors=\"coerce\")\n",
    "df = df.dropna(subset=[\"seq\", \"score_raw\"]).copy()\n",
    "df = df[df[\"seq\"].str.len() > 0].copy()\n",
    "\n",
    "print(\"Base rows:\", len(df))\n",
    "\n",
    "# -----------------------------\n",
    "# 2. Merge replicate readouts if available\n",
    "# -----------------------------\n",
    "if REPLICATE_TSV.exists():\n",
    "    rep = pd.read_csv(REPLICATE_TSV, sep=\"\\t\")\n",
    "    keep_cols = [\"element_id\"] + [c for c in REP_COLS if c in rep.columns]\n",
    "\n",
    "    extra_cols = [\n",
    "        \"score_sd_across_reps\",\n",
    "        \"score_range_across_reps\",\n",
    "        \"median_barcode_ratio_sd\",\n",
    "        \"n_qc_barcodes\",\n",
    "        \"rna_total\",\n",
    "        \"dna_total\",\n",
    "    ]\n",
    "    keep_cols += [c for c in extra_cols if c in rep.columns]\n",
    "    keep_cols = list(dict.fromkeys(keep_cols))\n",
    "\n",
    "    df = df.merge(rep[keep_cols], on=\"element_id\", how=\"left\")\n",
    "\n",
    "print(\"Columns after merge:\")\n",
    "print(df.columns.tolist())\n",
    "\n",
    "# -----------------------------\n",
    "# 3. Create or preserve split\n",
    "# -----------------------------\n",
    "if \"split\" in df.columns and df[\"split\"].notna().any():\n",
    "    df[\"split\"] = df[\"split\"].astype(str).str.lower()\n",
    "    print(\"Using existing split column.\")\n",
    "else:\n",
    "    print(\"No split column found. Creating train/val/test split.\")\n",
    "\n",
    "    if \"accession\" in df.columns and df[\"accession\"].notna().any():\n",
    "        group_col = \"accession\"\n",
    "    elif \"virus\" in df.columns and df[\"virus\"].notna().any():\n",
    "        group_col = \"virus\"\n",
    "    else:\n",
    "        group_col = \"element_id\"\n",
    "\n",
    "    df = df.reset_index(drop=True)\n",
    "    groups = df[group_col].fillna(pd.Series(df.index.astype(str), index=df.index)).astype(str)\n",
    "\n",
    "    gss1 = GroupShuffleSplit(n_splits=1, test_size=0.20, random_state=RANDOM_STATE)\n",
    "    train_idx, temp_idx = next(gss1.split(df, groups=groups))\n",
    "\n",
    "    df[\"split\"] = \"train\"\n",
    "    df.loc[temp_idx, \"split\"] = \"temp\"\n",
    "\n",
    "    temp_df = df[df[\"split\"] == \"temp\"].copy().reset_index()\n",
    "    temp_groups = temp_df[group_col].fillna(\n",
    "        pd.Series(temp_df.index.astype(str), index=temp_df.index)\n",
    "    ).astype(str)\n",
    "\n",
    "    gss2 = GroupShuffleSplit(n_splits=1, test_size=0.50, random_state=RANDOM_STATE)\n",
    "    val_rel_idx, test_rel_idx = next(gss2.split(temp_df, groups=temp_groups))\n",
    "\n",
    "    val_abs_idx = temp_df.iloc[val_rel_idx][\"index\"].values\n",
    "    test_abs_idx = temp_df.iloc[test_rel_idx][\"index\"].values\n",
    "\n",
    "    df.loc[val_abs_idx, \"split\"] = \"val\"\n",
    "    df.loc[test_abs_idx, \"split\"] = \"test\"\n",
    "\n",
    "print(df[\"split\"].value_counts())\n",
    "\n",
    "# -----------------------------\n",
    "# 4. Z-normalize using train only\n",
    "# -----------------------------\n",
    "train_mask = df[\"split\"] == \"train\"\n",
    "train_mean = df.loc[train_mask, \"score_raw\"].mean()\n",
    "train_std = df.loc[train_mask, \"score_raw\"].std(ddof=0)\n",
    "\n",
    "if not np.isfinite(train_std) or train_std <= 0:\n",
    "    raise ValueError(\"Invalid train label std.\")\n",
    "\n",
    "df[\"score_z\"] = (df[\"score_raw\"] - train_mean) / train_std\n",
    "df[\"score\"] = df[\"score_z\"]\n",
    "\n",
    "# -----------------------------\n",
    "# 5. Compute replicate consistency if replicate columns exist\n",
    "# -----------------------------\n",
    "available_rep_cols = [c for c in REP_COLS if c in df.columns]\n",
    "\n",
    "if len(available_rep_cols) >= 2:\n",
    "    df[\"rep_mean\"] = df[available_rep_cols].mean(axis=1)\n",
    "    df[\"rep_sd\"] = df[available_rep_cols].std(axis=1)\n",
    "    df[\"rep_range\"] = df[available_rep_cols].max(axis=1) - df[available_rep_cols].min(axis=1)\n",
    "else:\n",
    "    if \"score_sd_across_reps\" in df.columns:\n",
    "        df[\"rep_sd\"] = df[\"score_sd_across_reps\"]\n",
    "    else:\n",
    "        df[\"rep_sd\"] = np.nan\n",
    "\n",
    "    if \"score_range_across_reps\" in df.columns:\n",
    "        df[\"rep_range\"] = df[\"score_range_across_reps\"]\n",
    "    else:\n",
    "        df[\"rep_range\"] = np.nan\n",
    "\n",
    "# -----------------------------\n",
    "# 6. Build reliability weight\n",
    "# -----------------------------\n",
    "# Lower replicate SD = more reliable.\n",
    "# This gently downweights noisy labels, but does not remove them.\n",
    "if df[\"rep_sd\"].notna().any():\n",
    "    rep_sd = df[\"rep_sd\"].fillna(df[\"rep_sd\"].median())\n",
    "\n",
    "    reliability_weight = 1.0 / (rep_sd**2 + 0.25)\n",
    "\n",
    "    # Normalize by train mean so average train reliability weight ≈ 1\n",
    "    train_rel_mean = reliability_weight[train_mask].mean()\n",
    "    reliability_weight = reliability_weight / train_rel_mean\n",
    "\n",
    "    df[\"reliability_weight\"] = reliability_weight.clip(0.5, RELIABILITY_MAX_WEIGHT)\n",
    "else:\n",
    "    df[\"reliability_weight\"] = 1.0\n",
    "\n",
    "# Optional barcode support factor\n",
    "if \"n_qc_barcodes\" in df.columns:\n",
    "    nbc = pd.to_numeric(df[\"n_qc_barcodes\"], errors=\"coerce\").fillna(1)\n",
    "    barcode_weight = np.sqrt(nbc.clip(lower=1) / 3.0)\n",
    "    barcode_weight = barcode_weight.clip(0.75, 1.5)\n",
    "    df[\"reliability_weight\"] = (df[\"reliability_weight\"] * barcode_weight).clip(\n",
    "        0.5, RELIABILITY_MAX_WEIGHT\n",
    "    )\n",
    "\n",
    "# -----------------------------\n",
    "# 7. Build extreme-aware weight\n",
    "# -----------------------------\n",
    "abs_z = df[\"score_z\"].abs()\n",
    "\n",
    "# Smoothly upweight high-magnitude labels.\n",
    "# Center remains weight ~1.\n",
    "df[\"extreme_weight\"] = (1.0 + EXTREME_ALPHA * abs_z).clip(1.0, EXTREME_MAX_WEIGHT)\n",
    "\n",
    "# Final weight: both reliable and extreme labels matter more.\n",
    "df[\"sample_weight\"] = (\n",
    "    df[\"extreme_weight\"] * df[\"reliability_weight\"]\n",
    ").clip(FINAL_MIN_WEIGHT, FINAL_MAX_WEIGHT)\n",
    "\n",
    "# Normalize final sample weight so train mean is exactly 1.\n",
    "train_weight_mean = df.loc[train_mask, \"sample_weight\"].mean()\n",
    "df[\"sample_weight\"] = df[\"sample_weight\"] / train_weight_mean\n",
    "df[\"sample_weight\"] = df[\"sample_weight\"].clip(FINAL_MIN_WEIGHT, FINAL_MAX_WEIGHT)\n",
    "\n",
    "# -----------------------------\n",
    "# 8. Add bins / extreme flags\n",
    "# -----------------------------\n",
    "# Use train quantiles only to define bins, then apply to all splits.\n",
    "train_scores = df.loc[train_mask, \"score_z\"]\n",
    "\n",
    "q10 = train_scores.quantile(0.10)\n",
    "q30 = train_scores.quantile(0.30)\n",
    "q70 = train_scores.quantile(0.70)\n",
    "q90 = train_scores.quantile(0.90)\n",
    "\n",
    "bins = [-np.inf, q10, q30, q70, q90, np.inf]\n",
    "labels = [0, 1, 2, 3, 4]\n",
    "\n",
    "df[\"stability_bin_5\"] = pd.cut(\n",
    "    df[\"score_z\"],\n",
    "    bins=bins,\n",
    "    labels=labels,\n",
    "    include_lowest=True,\n",
    ").astype(int)\n",
    "\n",
    "df[\"is_extreme_bottom10\"] = df[\"score_z\"] <= q10\n",
    "df[\"is_extreme_top10\"] = df[\"score_z\"] >= q90\n",
    "df[\"is_extreme_abs20\"] = df[\"score_z\"].abs() >= train_scores.abs().quantile(0.80)\n",
    "\n",
    "# -----------------------------\n",
    "# 9. Final export\n",
    "# -----------------------------\n",
    "preferred_cols = [\n",
    "    \"element_id\",\n",
    "    \"seq\",\n",
    "    \"score\",\n",
    "    \"score_z\",\n",
    "    \"score_raw\",\n",
    "    \"split\",\n",
    "    \"sample_weight\",\n",
    "    \"extreme_weight\",\n",
    "    \"reliability_weight\",\n",
    "    \"stability_bin_5\",\n",
    "    \"is_extreme_bottom10\",\n",
    "    \"is_extreme_top10\",\n",
    "    \"is_extreme_abs20\",\n",
    "    \"rep_sd\",\n",
    "    \"rep_range\",\n",
    "    \"n_qc_barcodes\",\n",
    "    \"rna_total\",\n",
    "    \"dna_total\",\n",
    "]\n",
    "\n",
    "# Include replicate columns if present.\n",
    "preferred_cols += available_rep_cols\n",
    "\n",
    "final_cols = [c for c in preferred_cols if c in df.columns]\n",
    "out = df[final_cols].copy()\n",
    "\n",
    "out.to_csv(OUT_TSV, sep=\"\\t\", index=False)\n",
    "\n",
    "stats = {\n",
    "    \"input_base_tsv\": str(BASE_TSV),\n",
    "    \"input_replicate_tsv\": str(REPLICATE_TSV) if REPLICATE_TSV.exists() else None,\n",
    "    \"output_tsv\": str(OUT_TSV),\n",
    "    \"n_total\": int(len(out)),\n",
    "    \"split_counts\": out[\"split\"].value_counts().to_dict(),\n",
    "    \"score_raw_train_mean\": float(train_mean),\n",
    "    \"score_raw_train_std\": float(train_std),\n",
    "    \"q10_z\": float(q10),\n",
    "    \"q30_z\": float(q30),\n",
    "    \"q70_z\": float(q70),\n",
    "    \"q90_z\": float(q90),\n",
    "    \"extreme_alpha\": EXTREME_ALPHA,\n",
    "    \"extreme_max_weight\": EXTREME_MAX_WEIGHT,\n",
    "    \"reliability_max_weight\": RELIABILITY_MAX_WEIGHT,\n",
    "    \"final_min_weight\": FINAL_MIN_WEIGHT,\n",
    "    \"final_max_weight\": FINAL_MAX_WEIGHT,\n",
    "    \"train_sample_weight_mean\": float(out.loc[out[\"split\"] == \"train\", \"sample_weight\"].mean()),\n",
    "    \"train_sample_weight_min\": float(out.loc[out[\"split\"] == \"train\", \"sample_weight\"].min()),\n",
    "    \"train_sample_weight_max\": float(out.loc[out[\"split\"] == \"train\", \"sample_weight\"].max()),\n",
    "}\n",
    "\n",
    "with open(OUT_STATS_JSON, \"w\") as f:\n",
    "    json.dump(stats, f, indent=2)\n",
    "\n",
    "print(\"Exported:\", OUT_TSV)\n",
    "print(\"Stats:\", OUT_STATS_JSON)\n",
    "print(json.dumps(stats, indent=2))\n",
    "\n",
    "display(out.head())\n",
    "display(\n",
    "    out.groupby([\"split\", \"stability_bin_5\"])\n",
    "    .agg(\n",
    "        n=(\"score\", \"size\"),\n",
    "        score_mean=(\"score\", \"mean\"),\n",
    "        weight_mean=(\"sample_weight\", \"mean\"),\n",
    "        weight_min=(\"sample_weight\", \"min\"),\n",
    "        weight_max=(\"sample_weight\", \"max\"),\n",
    "    )\n",
    "    .reset_index()\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e30ecaad-0c7d-43e6-8d61-e8abeb9af0f9",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.14.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}