notebooks/pulldown_assay_data_processing.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Pulldown assay data processing: LY6A-Fc, LY6C1-Fc\n",
    "\n",
    "`sequence`: variant DNA sequence  \n",
    "`AA_sequence`: variant 7mer amino acid sequence\n",
    "\n",
    "**Data samples and replicates**: \n",
    "- starter-virus (starter, 3 replicates)\n",
    "    - `starter_virus_1_counts`\n",
    "    - `starter_virus_2_counts`\n",
    "    - `starter_virus_3_counts`  \n",
    "- LY6A-Fc (3 replicates)\n",
    "    - `LY6A_1_counts`\n",
    "    - `LY6A_2_counts`\n",
    "    - `LY6A_3_counts`  \n",
    "- LY6C1-Fc (3 replicates)\n",
    "    - `LY6C1_1_counts`\n",
    "    - `LY6C1_2_counts`\n",
    "    - `LY6C1_3_counts`  \n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Imports"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-01-26T16:42:44.367157Z",
     "start_time": "2021-01-26T16:42:43.989148Z"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import re\n",
    "import os, sys\n",
    "\n",
    "# import matplotlib as mpl\n",
    "# import matplotlib.pyplot as plt\n",
    "# import seaborn as sns\n",
    "\n",
    "from pathlib import Path\n",
    "current_dir = Path(os.getcwd())\n",
    "top_dir = current_dir.parent\n",
    "sys.path.append(str(top_dir))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "%load_ext autoreload\n",
    "%autoreload 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "from utils.data_processing import *\n",
    "from figures.utils.fig_utils import download_data, urls"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Load data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "download_data('../data/pulldown_assay_counts.csv', urls['pulldown_assay_counts.csv'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-01-26T16:43:56.388715Z",
     "start_time": "2021-01-26T16:42:55.034579Z"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "df = pd.read_csv('../data/pulldown_assay_counts.csv')\n",
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-01-26T16:45:03.458183Z",
     "start_time": "2021-01-26T16:45:03.451787Z"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['sequence',\n",
       " 'AA_sequence',\n",
       " 'starter_virus_1_counts',\n",
       " 'starter_virus_2_counts',\n",
       " 'starter_virus_3_counts',\n",
       " 'LY6A_1_counts',\n",
       " 'LY6A_2_counts',\n",
       " 'LY6A_3_counts',\n",
       " 'LY6C1_1_counts',\n",
       " 'LY6C1_2_counts',\n",
       " 'LY6C1_3_counts']"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "list(df.columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Since the dataframe contains read count values, a NaN value corresponds to 0 reads\n",
    "df = df.fillna(0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## RPM\n",
    "First we need to compute the reads per million (RPM) of each column."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['starter_virus_1_counts',\n",
       " 'starter_virus_2_counts',\n",
       " 'starter_virus_3_counts',\n",
       " 'LY6A_1_counts',\n",
       " 'LY6A_2_counts',\n",
       " 'LY6A_3_counts',\n",
       " 'LY6C1_1_counts',\n",
       " 'LY6C1_2_counts',\n",
       " 'LY6C1_3_counts']"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Isolate the columns corresponding to read counts\n",
    "counts_cols = [col for col in df.columns if \"counts\" in col]\n",
    "counts_cols"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['starter_virus_1_RPM',\n",
       " 'starter_virus_2_RPM',\n",
       " 'starter_virus_3_RPM',\n",
       " 'LY6A_1_RPM',\n",
       " 'LY6A_2_RPM',\n",
       " 'LY6A_3_RPM',\n",
       " 'LY6C1_1_RPM',\n",
       " 'LY6C1_2_RPM',\n",
       " 'LY6C1_3_RPM']"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Create a new RPM column for each counts column \n",
    "RPM_cols = [re.sub(\"_counts\", \"_RPM\", col) for col in counts_cols]\n",
    "RPM_cols"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>sequence</th>\n",
       "      <th>AA_sequence</th>\n",
       "      <th>starter_virus_1_RPM</th>\n",
       "      <th>starter_virus_2_RPM</th>\n",
       "      <th>starter_virus_3_RPM</th>\n",
       "      <th>LY6A_1_RPM</th>\n",
       "      <th>LY6A_2_RPM</th>\n",
       "      <th>LY6A_3_RPM</th>\n",
       "      <th>LY6C1_1_RPM</th>\n",
       "      <th>LY6C1_2_RPM</th>\n",
       "      <th>LY6C1_3_RPM</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>AATGCTGTGAGTCGGGCTGGG</td>\n",
       "      <td>NAVSRAG</td>\n",
       "      <td>329.700380</td>\n",
       "      <td>334.384942</td>\n",
       "      <td>336.334333</td>\n",
       "      <td>21.424134</td>\n",
       "      <td>50.797720</td>\n",
       "      <td>36.544493</td>\n",
       "      <td>90.000646</td>\n",
       "      <td>101.929615</td>\n",
       "      <td>110.859467</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>GTGCGGCCTAATGGGTCTGGT</td>\n",
       "      <td>VRPNGSG</td>\n",
       "      <td>326.575902</td>\n",
       "      <td>330.955353</td>\n",
       "      <td>327.792104</td>\n",
       "      <td>38.303754</td>\n",
       "      <td>50.239503</td>\n",
       "      <td>30.219485</td>\n",
       "      <td>133.822949</td>\n",
       "      <td>165.999659</td>\n",
       "      <td>177.375148</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>GGTACGCGGCCTGGGCTTTTG</td>\n",
       "      <td>GTRPGLL</td>\n",
       "      <td>246.562089</td>\n",
       "      <td>235.982115</td>\n",
       "      <td>235.867519</td>\n",
       "      <td>4.544513</td>\n",
       "      <td>23.445101</td>\n",
       "      <td>15.461132</td>\n",
       "      <td>19.319510</td>\n",
       "      <td>9.319279</td>\n",
       "      <td>27.388810</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>TTTGGGGGTTCGCGGACTCCG</td>\n",
       "      <td>FGGSRTP</td>\n",
       "      <td>215.045613</td>\n",
       "      <td>214.745044</td>\n",
       "      <td>217.890590</td>\n",
       "      <td>30.513160</td>\n",
       "      <td>40.749819</td>\n",
       "      <td>17.569468</td>\n",
       "      <td>65.026644</td>\n",
       "      <td>75.136688</td>\n",
       "      <td>110.859467</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>TTGGATAAGAGGAATCTTGTT</td>\n",
       "      <td>LDKRNLV</td>\n",
       "      <td>206.215565</td>\n",
       "      <td>192.188900</td>\n",
       "      <td>207.690913</td>\n",
       "      <td>16.879620</td>\n",
       "      <td>8.931467</td>\n",
       "      <td>16.866689</td>\n",
       "      <td>42.408681</td>\n",
       "      <td>32.617477</td>\n",
       "      <td>48.908589</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1750113</th>\n",
       "      <td>GCTGCTAGGGCTGCTACTGGT</td>\n",
       "      <td>AARAATG</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1750114</th>\n",
       "      <td>GGGGTGCCGGGTTATGCTTGG</td>\n",
       "      <td>GVPGYAW</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1750115</th>\n",
       "      <td>GGTAATTATCTTCATGCTTTG</td>\n",
       "      <td>GNYLHAL</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1750116</th>\n",
       "      <td>CGTGAGCCTAGTCCTAGTAGG</td>\n",
       "      <td>REPSPSR</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1750117</th>\n",
       "      <td>GTTGCTACTGCGCTGCGGCCG</td>\n",
       "      <td>VATALRP</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1750118 rows × 11 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                      sequence AA_sequence  starter_virus_1_RPM  \\\n",
       "0        AATGCTGTGAGTCGGGCTGGG     NAVSRAG           329.700380   \n",
       "1        GTGCGGCCTAATGGGTCTGGT     VRPNGSG           326.575902   \n",
       "2        GGTACGCGGCCTGGGCTTTTG     GTRPGLL           246.562089   \n",
       "3        TTTGGGGGTTCGCGGACTCCG     FGGSRTP           215.045613   \n",
       "4        TTGGATAAGAGGAATCTTGTT     LDKRNLV           206.215565   \n",
       "...                        ...         ...                  ...   \n",
       "1750113  GCTGCTAGGGCTGCTACTGGT     AARAATG             0.000000   \n",
       "1750114  GGGGTGCCGGGTTATGCTTGG     GVPGYAW             0.000000   \n",
       "1750115  GGTAATTATCTTCATGCTTTG     GNYLHAL             0.000000   \n",
       "1750116  CGTGAGCCTAGTCCTAGTAGG     REPSPSR             0.000000   \n",
       "1750117  GTTGCTACTGCGCTGCGGCCG     VATALRP             0.000000   \n",
       "\n",
       "         starter_virus_2_RPM  starter_virus_3_RPM  LY6A_1_RPM  LY6A_2_RPM  \\\n",
       "0                 334.384942           336.334333   21.424134   50.797720   \n",
       "1                 330.955353           327.792104   38.303754   50.239503   \n",
       "2                 235.982115           235.867519    4.544513   23.445101   \n",
       "3                 214.745044           217.890590   30.513160   40.749819   \n",
       "4                 192.188900           207.690913   16.879620    8.931467   \n",
       "...                      ...                  ...         ...         ...   \n",
       "1750113             0.000000             0.000000    0.000000    0.000000   \n",
       "1750114             0.000000             0.000000    0.000000    0.000000   \n",
       "1750115             0.000000             0.000000    0.000000    0.000000   \n",
       "1750116             0.000000             0.000000    0.000000    0.000000   \n",
       "1750117             0.000000             0.000000    0.000000    0.000000   \n",
       "\n",
       "         LY6A_3_RPM  LY6C1_1_RPM  LY6C1_2_RPM  LY6C1_3_RPM  \n",
       "0         36.544493    90.000646   101.929615   110.859467  \n",
       "1         30.219485   133.822949   165.999659   177.375148  \n",
       "2         15.461132    19.319510     9.319279    27.388810  \n",
       "3         17.569468    65.026644    75.136688   110.859467  \n",
       "4         16.866689    42.408681    32.617477    48.908589  \n",
       "...             ...          ...          ...          ...  \n",
       "1750113    0.000000     0.000000     0.000000     0.000000  \n",
       "1750114    0.000000     0.000000     0.000000     0.000000  \n",
       "1750115    0.000000     0.000000     0.000000     0.000000  \n",
       "1750116    0.000000     0.000000     0.000000     0.000000  \n",
       "1750117    0.000000     0.000000     0.000000     0.000000  \n",
       "\n",
       "[1750118 rows x 11 columns]"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Compute reads per million (RPM) and store in a new dataframe with renamed columns\n",
    "df_RPM = df.drop(columns=counts_cols)\n",
    "RPM_values = RPM(df[counts_cols]).rename(columns=dict(zip(counts_cols, RPM_cols)))\n",
    "df_RPM = pd.concat([df_RPM, RPM_values], axis=1)\n",
    "df_RPM"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Aggregating sample replicates\n",
    "Each sample has 3 replicates, labeled `1`, `2` and `3`. We want the mean RPM, as well\n",
    "as coefficient of variation (CV, computed as *standard deviation / mean*), of the 3\n",
    "replicates per sample.\n",
    "\n",
    "**Note:** if there are variants with $RPM=0$ across all replicates, then `mean_RPM` $=0$,\n",
    "which results in 0-division when computing CV, and yields a `NaN` value. In this case,\n",
    "it's not such a big deal because we end up dropping these variants from the SVAE\n",
    "training data anyway."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-01-27T14:03:59.173887Z",
     "start_time": "2021-01-27T14:03:59.168399Z"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['starter_virus', 'LY6A', 'LY6C1'], dtype=object)"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# First get the names of the samples\n",
    "samples = pd.unique([re.sub('_([123])_RPM', '', col) for col in RPM_cols])\n",
    "samples"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>sequence</th>\n",
       "      <th>AA_sequence</th>\n",
       "      <th>starter_virus_mean_RPM</th>\n",
       "      <th>starter_virus_cv_RPM</th>\n",
       "      <th>LY6A_mean_RPM</th>\n",
       "      <th>LY6A_cv_RPM</th>\n",
       "      <th>LY6C1_mean_RPM</th>\n",
       "      <th>LY6C1_cv_RPM</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>AATGCTGTGAGTCGGGCTGGG</td>\n",
       "      <td>NAVSRAG</td>\n",
       "      <td>333.473218</td>\n",
       "      <td>0.010225</td>\n",
       "      <td>36.255449</td>\n",
       "      <td>0.405151</td>\n",
       "      <td>100.929909</td>\n",
       "      <td>0.103689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>GTGCGGCCTAATGGGTCTGGT</td>\n",
       "      <td>VRPNGSG</td>\n",
       "      <td>328.441120</td>\n",
       "      <td>0.006883</td>\n",
       "      <td>39.587581</td>\n",
       "      <td>0.254412</td>\n",
       "      <td>159.065919</td>\n",
       "      <td>0.142009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>GGTACGCGGCCTGGGCTTTTG</td>\n",
       "      <td>GTRPGLL</td>\n",
       "      <td>239.470574</td>\n",
       "      <td>0.025647</td>\n",
       "      <td>14.483582</td>\n",
       "      <td>0.655096</td>\n",
       "      <td>18.675866</td>\n",
       "      <td>0.484687</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>TTTGGGGGTTCGCGGACTCCG</td>\n",
       "      <td>FGGSRTP</td>\n",
       "      <td>215.893749</td>\n",
       "      <td>0.008040</td>\n",
       "      <td>29.610816</td>\n",
       "      <td>0.392306</td>\n",
       "      <td>83.674266</td>\n",
       "      <td>0.287778</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>TTGGATAAGAGGAATCTTGTT</td>\n",
       "      <td>LDKRNLV</td>\n",
       "      <td>202.031793</td>\n",
       "      <td>0.042350</td>\n",
       "      <td>14.225926</td>\n",
       "      <td>0.322309</td>\n",
       "      <td>41.311582</td>\n",
       "      <td>0.198510</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1750113</th>\n",
       "      <td>GCTGCTAGGGCTGCTACTGGT</td>\n",
       "      <td>AARAATG</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1750114</th>\n",
       "      <td>GGGGTGCCGGGTTATGCTTGG</td>\n",
       "      <td>GVPGYAW</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1750115</th>\n",
       "      <td>GGTAATTATCTTCATGCTTTG</td>\n",
       "      <td>GNYLHAL</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1750116</th>\n",
       "      <td>CGTGAGCCTAGTCCTAGTAGG</td>\n",
       "      <td>REPSPSR</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1750117</th>\n",
       "      <td>GTTGCTACTGCGCTGCGGCCG</td>\n",
       "      <td>VATALRP</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1750118 rows × 8 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                      sequence AA_sequence  starter_virus_mean_RPM  \\\n",
       "0        AATGCTGTGAGTCGGGCTGGG     NAVSRAG              333.473218   \n",
       "1        GTGCGGCCTAATGGGTCTGGT     VRPNGSG              328.441120   \n",
       "2        GGTACGCGGCCTGGGCTTTTG     GTRPGLL              239.470574   \n",
       "3        TTTGGGGGTTCGCGGACTCCG     FGGSRTP              215.893749   \n",
       "4        TTGGATAAGAGGAATCTTGTT     LDKRNLV              202.031793   \n",
       "...                        ...         ...                     ...   \n",
       "1750113  GCTGCTAGGGCTGCTACTGGT     AARAATG                0.000000   \n",
       "1750114  GGGGTGCCGGGTTATGCTTGG     GVPGYAW                0.000000   \n",
       "1750115  GGTAATTATCTTCATGCTTTG     GNYLHAL                0.000000   \n",
       "1750116  CGTGAGCCTAGTCCTAGTAGG     REPSPSR                0.000000   \n",
       "1750117  GTTGCTACTGCGCTGCGGCCG     VATALRP                0.000000   \n",
       "\n",
       "         starter_virus_cv_RPM  LY6A_mean_RPM  LY6A_cv_RPM  LY6C1_mean_RPM  \\\n",
       "0                    0.010225      36.255449     0.405151      100.929909   \n",
       "1                    0.006883      39.587581     0.254412      159.065919   \n",
       "2                    0.025647      14.483582     0.655096       18.675866   \n",
       "3                    0.008040      29.610816     0.392306       83.674266   \n",
       "4                    0.042350      14.225926     0.322309       41.311582   \n",
       "...                       ...            ...          ...             ...   \n",
       "1750113                   NaN       0.000000          NaN        0.000000   \n",
       "1750114                   NaN       0.000000          NaN        0.000000   \n",
       "1750115                   NaN       0.000000          NaN        0.000000   \n",
       "1750116                   NaN       0.000000          NaN        0.000000   \n",
       "1750117                   NaN       0.000000          NaN        0.000000   \n",
       "\n",
       "         LY6C1_cv_RPM  \n",
       "0            0.103689  \n",
       "1            0.142009  \n",
       "2            0.484687  \n",
       "3            0.287778  \n",
       "4            0.198510  \n",
       "...               ...  \n",
       "1750113           NaN  \n",
       "1750114           NaN  \n",
       "1750115           NaN  \n",
       "1750116           NaN  \n",
       "1750117           NaN  \n",
       "\n",
       "[1750118 rows x 8 columns]"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Compute mean and CV (coefficient of variation) RPM, and store in a new dataframe\n",
    "df_mean_cv_RPM = df.drop(columns=counts_cols)\n",
    "for sample in samples:\n",
    "    replicate_cols = [col for col in df_RPM.columns if sample in col]\n",
    "    mean_RPM = df_RPM[replicate_cols].mean(axis=1)\n",
    "    cv_RPM = df_RPM[replicate_cols].std(axis=1) / mean_RPM\n",
    "    df_mean_cv_RPM[\"{}_mean_RPM\".format(sample)] = mean_RPM\n",
    "    df_mean_cv_RPM[\"{}_cv_RPM\".format(sample)] = cv_RPM\n",
    "df_mean_cv_RPM\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## (Log2) enrichment\n",
    "Now, as a way to measure the performance of each variant in each assay, we want to\n",
    "compute the assay enrichment (fold change of mean RPM) relative to the `starter_virus` sample. We\n",
    "take the $\\log_2$ of enrichment to keep all of the values on a similar scale.\n",
    "\n",
    "**Note:** like when computing CV above, variants with `mean_RPM` $=0$ can cause some\n",
    "problems: \n",
    "- when `starter_virus_mean_RPM` $=0$, computing enrichment results in\n",
    "0-division\n",
    "- when the enrichment value exists (`starter_virus_mean_RPM` $\\neq 0$) but is 0 (which\n",
    "  happens when `<assay>_mean_RPM` $=0$), then taking the $\\log_2$ results in `log2enr`\n",
    "  $= -\\infty$ (`-np.inf`).\n",
    "\n",
    "Again, this isn't a big deal in this case because those variants will be ultimately\n",
    "dropped anyway."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python3.9/site-packages/pandas/core/arraylike.py:364: RuntimeWarning: divide by zero encountered in log2\n",
      "  result = getattr(ufunc, method)(*inputs, **kwargs)\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>sequence</th>\n",
       "      <th>AA_sequence</th>\n",
       "      <th>LY6A_enr</th>\n",
       "      <th>LY6A_log2enr</th>\n",
       "      <th>LY6C1_enr</th>\n",
       "      <th>LY6C1_log2enr</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>AATGCTGTGAGTCGGGCTGGG</td>\n",
       "      <td>NAVSRAG</td>\n",
       "      <td>0.108721</td>\n",
       "      <td>-3.201301</td>\n",
       "      <td>0.302663</td>\n",
       "      <td>-1.724217</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>GTGCGGCCTAATGGGTCTGGT</td>\n",
       "      <td>VRPNGSG</td>\n",
       "      <td>0.120532</td>\n",
       "      <td>-3.052515</td>\n",
       "      <td>0.484306</td>\n",
       "      <td>-1.046010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>GGTACGCGGCCTGGGCTTTTG</td>\n",
       "      <td>GTRPGLL</td>\n",
       "      <td>0.060482</td>\n",
       "      <td>-4.047358</td>\n",
       "      <td>0.077988</td>\n",
       "      <td>-3.680601</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>TTTGGGGGTTCGCGGACTCCG</td>\n",
       "      <td>FGGSRTP</td>\n",
       "      <td>0.137155</td>\n",
       "      <td>-2.866125</td>\n",
       "      <td>0.387572</td>\n",
       "      <td>-1.367466</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>TTGGATAAGAGGAATCTTGTT</td>\n",
       "      <td>LDKRNLV</td>\n",
       "      <td>0.070414</td>\n",
       "      <td>-3.827988</td>\n",
       "      <td>0.204481</td>\n",
       "      <td>-2.289964</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1750113</th>\n",
       "      <td>GCTGCTAGGGCTGCTACTGGT</td>\n",
       "      <td>AARAATG</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1750114</th>\n",
       "      <td>GGGGTGCCGGGTTATGCTTGG</td>\n",
       "      <td>GVPGYAW</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1750115</th>\n",
       "      <td>GGTAATTATCTTCATGCTTTG</td>\n",
       "      <td>GNYLHAL</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1750116</th>\n",
       "      <td>CGTGAGCCTAGTCCTAGTAGG</td>\n",
       "      <td>REPSPSR</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1750117</th>\n",
       "      <td>GTTGCTACTGCGCTGCGGCCG</td>\n",
       "      <td>VATALRP</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1750118 rows × 6 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                      sequence AA_sequence  LY6A_enr  LY6A_log2enr  LY6C1_enr  \\\n",
       "0        AATGCTGTGAGTCGGGCTGGG     NAVSRAG  0.108721     -3.201301   0.302663   \n",
       "1        GTGCGGCCTAATGGGTCTGGT     VRPNGSG  0.120532     -3.052515   0.484306   \n",
       "2        GGTACGCGGCCTGGGCTTTTG     GTRPGLL  0.060482     -4.047358   0.077988   \n",
       "3        TTTGGGGGTTCGCGGACTCCG     FGGSRTP  0.137155     -2.866125   0.387572   \n",
       "4        TTGGATAAGAGGAATCTTGTT     LDKRNLV  0.070414     -3.827988   0.204481   \n",
       "...                        ...         ...       ...           ...        ...   \n",
       "1750113  GCTGCTAGGGCTGCTACTGGT     AARAATG       NaN           NaN        NaN   \n",
       "1750114  GGGGTGCCGGGTTATGCTTGG     GVPGYAW       NaN           NaN        NaN   \n",
       "1750115  GGTAATTATCTTCATGCTTTG     GNYLHAL       NaN           NaN        NaN   \n",
       "1750116  CGTGAGCCTAGTCCTAGTAGG     REPSPSR       NaN           NaN        NaN   \n",
       "1750117  GTTGCTACTGCGCTGCGGCCG     VATALRP       NaN           NaN        NaN   \n",
       "\n",
       "         LY6C1_log2enr  \n",
       "0            -1.724217  \n",
       "1            -1.046010  \n",
       "2            -3.680601  \n",
       "3            -1.367466  \n",
       "4            -2.289964  \n",
       "...                ...  \n",
       "1750113            NaN  \n",
       "1750114            NaN  \n",
       "1750115            NaN  \n",
       "1750116            NaN  \n",
       "1750117            NaN  \n",
       "\n",
       "[1750118 rows x 6 columns]"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Compute enrichment and log2 enrichment, and store in a new dataframe\n",
    "df_enr = df.drop(columns=counts_cols)\n",
    "for sample in samples[1:]: # We only do this for non-starter-virus samples\n",
    "    enr = df_mean_cv_RPM[\"{}_mean_RPM\".format(sample)]/df_mean_cv_RPM[\"starter_virus_mean_RPM\"]\n",
    "    log2enr = np.log2(enr)\n",
    "    df_enr[\"{}_enr\".format(sample)] = enr\n",
    "    df_enr[\"{}_log2enr\".format(sample)] = log2enr\n",
    "df_enr\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Export processed data\n",
    "\n",
    "Now we'll write out the `log2enr` and `cv_RPM` values of the `LY6A`-Fc and `LY6C1`-Fc pulldown\n",
    "assays, along with the 7mer `AA_sequence`s and `mean_RPM` values for all samples\n",
    "(including the starter)\n",
    "to a .csv file, so we can use them to train the SVAE. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>AA_sequence</th>\n",
       "      <th>starter_virus_mean_RPM</th>\n",
       "      <th>LY6A_log2enr</th>\n",
       "      <th>LY6A_cv_RPM</th>\n",
       "      <th>LY6A_mean_RPM</th>\n",
       "      <th>LY6C1_log2enr</th>\n",
       "      <th>LY6C1_cv_RPM</th>\n",
       "      <th>LY6C1_mean_RPM</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>NAVSRAG</td>\n",
       "      <td>333.473218</td>\n",
       "      <td>-3.201301</td>\n",
       "      <td>0.405151</td>\n",
       "      <td>36.255449</td>\n",
       "      <td>-1.724217</td>\n",
       "      <td>0.103689</td>\n",
       "      <td>100.929909</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>VRPNGSG</td>\n",
       "      <td>328.441120</td>\n",
       "      <td>-3.052515</td>\n",
       "      <td>0.254412</td>\n",
       "      <td>39.587581</td>\n",
       "      <td>-1.046010</td>\n",
       "      <td>0.142009</td>\n",
       "      <td>159.065919</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>GTRPGLL</td>\n",
       "      <td>239.470574</td>\n",
       "      <td>-4.047358</td>\n",
       "      <td>0.655096</td>\n",
       "      <td>14.483582</td>\n",
       "      <td>-3.680601</td>\n",
       "      <td>0.484687</td>\n",
       "      <td>18.675866</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>FGGSRTP</td>\n",
       "      <td>215.893749</td>\n",
       "      <td>-2.866125</td>\n",
       "      <td>0.392306</td>\n",
       "      <td>29.610816</td>\n",
       "      <td>-1.367466</td>\n",
       "      <td>0.287778</td>\n",
       "      <td>83.674266</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>LDKRNLV</td>\n",
       "      <td>202.031793</td>\n",
       "      <td>-3.827988</td>\n",
       "      <td>0.322309</td>\n",
       "      <td>14.225926</td>\n",
       "      <td>-2.289964</td>\n",
       "      <td>0.198510</td>\n",
       "      <td>41.311582</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1750113</th>\n",
       "      <td>AARAATG</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1750114</th>\n",
       "      <td>GVPGYAW</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1750115</th>\n",
       "      <td>GNYLHAL</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1750116</th>\n",
       "      <td>REPSPSR</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1750117</th>\n",
       "      <td>VATALRP</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1750118 rows × 8 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        AA_sequence  starter_virus_mean_RPM  LY6A_log2enr  LY6A_cv_RPM  \\\n",
       "0           NAVSRAG              333.473218     -3.201301     0.405151   \n",
       "1           VRPNGSG              328.441120     -3.052515     0.254412   \n",
       "2           GTRPGLL              239.470574     -4.047358     0.655096   \n",
       "3           FGGSRTP              215.893749     -2.866125     0.392306   \n",
       "4           LDKRNLV              202.031793     -3.827988     0.322309   \n",
       "...             ...                     ...           ...          ...   \n",
       "1750113     AARAATG                0.000000           NaN          NaN   \n",
       "1750114     GVPGYAW                0.000000           NaN          NaN   \n",
       "1750115     GNYLHAL                0.000000           NaN          NaN   \n",
       "1750116     REPSPSR                0.000000           NaN          NaN   \n",
       "1750117     VATALRP                0.000000           NaN          NaN   \n",
       "\n",
       "         LY6A_mean_RPM  LY6C1_log2enr  LY6C1_cv_RPM  LY6C1_mean_RPM  \n",
       "0            36.255449      -1.724217      0.103689      100.929909  \n",
       "1            39.587581      -1.046010      0.142009      159.065919  \n",
       "2            14.483582      -3.680601      0.484687       18.675866  \n",
       "3            29.610816      -1.367466      0.287778       83.674266  \n",
       "4            14.225926      -2.289964      0.198510       41.311582  \n",
       "...                ...            ...           ...             ...  \n",
       "1750113       0.000000            NaN           NaN        0.000000  \n",
       "1750114       0.000000            NaN           NaN        0.000000  \n",
       "1750115       0.000000            NaN           NaN        0.000000  \n",
       "1750116       0.000000            NaN           NaN        0.000000  \n",
       "1750117       0.000000            NaN           NaN        0.000000  \n",
       "\n",
       "[1750118 rows x 8 columns]"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_for_SVAE = df_mean_cv_RPM[['AA_sequence', 'starter_virus_mean_RPM']].copy()\n",
    "df_for_SVAE['LY6A_log2enr'] = df_enr['LY6A_log2enr']\n",
    "df_for_SVAE['LY6A_cv_RPM'] = df_mean_cv_RPM['LY6A_cv_RPM']\n",
    "df_for_SVAE['LY6A_mean_RPM'] = df_mean_cv_RPM['LY6A_mean_RPM']\n",
    "df_for_SVAE['LY6C1_log2enr'] = df_enr['LY6C1_log2enr']\n",
    "df_for_SVAE['LY6C1_cv_RPM'] = df_mean_cv_RPM['LY6C1_cv_RPM']\n",
    "df_for_SVAE['LY6C1_mean_RPM'] = df_mean_cv_RPM['LY6C1_mean_RPM']\n",
    "df_for_SVAE"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_for_SVAE.to_csv(\"../data/pulldown_assay_SVAE_data.csv\", index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "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": "3.9.14"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {
    "height": "calc(100% - 180px)",
    "left": "10px",
    "top": "150px",
    "width": "288px"
   },
   "toc_section_display": true,
   "toc_window_display": true
  },
  "varInspector": {
   "cols": {
    "lenName": 16,
    "lenType": 16,
    "lenVar": 40
   },
   "kernels_config": {
    "python": {
     "delete_cmd_postfix": "",
     "delete_cmd_prefix": "del ",
     "library": "var_list.py",
     "varRefreshCmd": "print(var_dic_list())"
    },
    "r": {
     "delete_cmd_postfix": ") ",
     "delete_cmd_prefix": "rm(",
     "library": "var_list.r",
     "varRefreshCmd": "cat(var_dic_list()) "
    }
   },
   "types_to_exclude": [
    "module",
    "function",
    "builtin_function_or_method",
    "instance",
    "_Feature"
   ],
   "window_display": false
  },
  "vscode": {
   "interpreter": {
    "hash": "d3f6b1c0c8130950732fb13e0e1727bd86e6412f72db16aba2fe51cf9eae5320"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}