diagnostics_scripts/MiSeqRun_NNK.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "8b308f8b-5460-4ea6-9e91-151c5e92d354",
   "metadata": {
    "tags": [
     "parameters"
    ]
   },
   "outputs": [],
   "source": [
    "# Parameters\n",
    "# Refer to injected parameters for any run.ipynb (similar cell below)\n",
    "fname = 'MiSeqRun1-FKS1-HS2_variant_data_merge.tsv' # FKS2-HS1 in MiSeqRun2, others in MiSeqRun1\n",
    "strain = 'BY4741'\n",
    "locus = 'FKS1-HS2'\n",
    "pool_type = 'single'\n",
    "\n",
    "min_input_readcount_high = 10\n",
    "min_input_readcount_mid = 3"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "92739bd9-9f71-4f4b-800c-cde644e33f07",
   "metadata": {},
   "source": [
    "# Notebook to assess diversity from initial libraries sequenced in pre-runs"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1fa0cda4-2357-49a7-b5b3-3967fa25948c",
   "metadata": {},
   "source": [
    "## Import libraries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "13b73f43-866d-40d9-8944-b03d4a4c6d53",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "import os\n",
    "import glob\n",
    "import re\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import math\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "plt.rcParams['svg.fonttype'] = 'none'\n",
    "from upsetplot import from_indicators\n",
    "from upsetplot import UpSet"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "dc459c82-ea46-43e1-a619-9e2f79d92fea",
   "metadata": {
    "tags": []
   },
   "source": [
    "## Functions"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1a60ef3e-3bfa-4b4e-99e7-392dd959b240",
   "metadata": {},
   "source": [
    "## Specify paths"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "65cedf42-7b87-4b21-b755-85476c012c12",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "metadata_path = '../data/'\n",
    "variantdata_path = '../preruns/'\n",
    "DMSdata_path = '../processed_data/'\n",
    "codon_table_path = '../data/general_use/ScerevisiaeTAXID559292_Cocoputs_codon_table.csv'\n",
    "graph_outpath = '../diagnostics_graphs/'"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d32c96f1-3510-4ba3-8fc7-f92c9f0a5bad",
   "metadata": {},
   "source": [
    "## Analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "6ba18aa1-258c-4543-b3b7-d99396d7674a",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "f = variantdata_path+fname"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "33f43f13-aef3-4b06-bcbd-3afeef14a391",
   "metadata": {},
   "source": [
    "### Read dataframe"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "b2b79ead-cf09-4e1d-bc70-f1f4ac8cc103",
   "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>nt_seq</th>\n",
       "      <th>aa_seq</th>\n",
       "      <th>WT</th>\n",
       "      <th>STOP</th>\n",
       "      <th>STOP_readthrough</th>\n",
       "      <th>Nham_nt</th>\n",
       "      <th>Nham_aa</th>\n",
       "      <th>Nmut_codons</th>\n",
       "      <th>input1_e1_s0_bNA_count</th>\n",
       "      <th>output1_e1_s1_b1_count</th>\n",
       "      <th>barcode_valid</th>\n",
       "      <th>indel</th>\n",
       "      <th>constant_region</th>\n",
       "      <th>permitted</th>\n",
       "      <th>too_many_substitutions</th>\n",
       "      <th>mixed_substitutions</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>aagtgggtgagacgttatacacgt</td>\n",
       "      <td>KWVRRYTR</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>5.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>aagtgggtgagacgttatacactg</td>\n",
       "      <td>KWVRRYTL</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>aagtgggtgagacgttatacattg</td>\n",
       "      <td>KWVRRYTL</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>197</td>\n",
       "      <td>197</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>aagtgggtgagacgttatatgttg</td>\n",
       "      <td>KWVRRYML</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>aagtgggtgagacgttatcatttg</td>\n",
       "      <td>KWVRRYHL</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>5.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</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",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1208</th>\n",
       "      <td>ttttgggtgggacgttatacattg</td>\n",
       "      <td>FWVGRYTL</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1209</th>\n",
       "      <td>ttttgggttagacgttatacattg</td>\n",
       "      <td>FWVRRYTL</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1210</th>\n",
       "      <td>ttttggttgagacgttatacattg</td>\n",
       "      <td>FWLRRYTL</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1211</th>\n",
       "      <td>ttttgtgtgagacgttatacattg</td>\n",
       "      <td>FCVRRYTL</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1212</th>\n",
       "      <td>tttttgggtgagacgttatacattg</td>\n",
       "      <td>FLGETLYI</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1213 rows × 16 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                         nt_seq    aa_seq    WT   STOP  STOP_readthrough  \\\n",
       "0      aagtgggtgagacgttatacacgt  KWVRRYTR  <NA>  False             False   \n",
       "1      aagtgggtgagacgttatacactg  KWVRRYTL  <NA>  False             False   \n",
       "2      aagtgggtgagacgttatacattg  KWVRRYTL  <NA>  False             False   \n",
       "3      aagtgggtgagacgttatatgttg  KWVRRYML  <NA>  False             False   \n",
       "4      aagtgggtgagacgttatcatttg  KWVRRYHL  <NA>  False             False   \n",
       "...                         ...       ...   ...    ...               ...   \n",
       "1208   ttttgggtgggacgttatacattg  FWVGRYTL  <NA>  False             False   \n",
       "1209   ttttgggttagacgttatacattg  FWVRRYTL  <NA>  False             False   \n",
       "1210   ttttggttgagacgttatacattg  FWLRRYTL  <NA>  False             False   \n",
       "1211   ttttgtgtgagacgttatacattg  FCVRRYTL  <NA>  False             False   \n",
       "1212  tttttgggtgagacgttatacattg  FLGETLYI  <NA>  False             False   \n",
       "\n",
       "      Nham_nt  Nham_aa  Nmut_codons  input1_e1_s0_bNA_count  \\\n",
       "0         5.0      2.0          2.0                       1   \n",
       "1         3.0      1.0          2.0                       1   \n",
       "2         2.0      1.0          1.0                     197   \n",
       "3         4.0      2.0          2.0                       1   \n",
       "4         5.0      2.0          2.0                       1   \n",
       "...       ...      ...          ...                     ...   \n",
       "1208      3.0      2.0          2.0                       1   \n",
       "1209      3.0      1.0          2.0                       1   \n",
       "1210      3.0      2.0          2.0                       1   \n",
       "1211      3.0      2.0          2.0                       1   \n",
       "1212      NaN      NaN          NaN                       1   \n",
       "\n",
       "      output1_e1_s1_b1_count  barcode_valid  indel  constant_region permitted  \\\n",
       "0                          1           True  False             True      True   \n",
       "1                          1           True  False             True      True   \n",
       "2                        197           True  False             True      True   \n",
       "3                          1           True  False             True      True   \n",
       "4                          1           True  False             True      True   \n",
       "...                      ...            ...    ...              ...       ...   \n",
       "1208                       1           True  False             True      True   \n",
       "1209                       1           True  False             True      True   \n",
       "1210                       1           True  False             True      True   \n",
       "1211                       1           True  False             True      True   \n",
       "1212                       1           True   True             True       NaN   \n",
       "\n",
       "     too_many_substitutions mixed_substitutions  \n",
       "0                     False               False  \n",
       "1                     False               False  \n",
       "2                     False               False  \n",
       "3                     False               False  \n",
       "4                     False               False  \n",
       "...                     ...                 ...  \n",
       "1208                  False               False  \n",
       "1209                  False               False  \n",
       "1210                  False               False  \n",
       "1211                  False               False  \n",
       "1212                    NaN                 NaN  \n",
       "\n",
       "[1213 rows x 16 columns]"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.read_csv(f, sep='\\t', header=0, dtype={'WT': 'boolean'}) # Boolean type supports missing data, helps pandas interpret column content\n",
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "a6fd9d57-24c5-4869-8277-0562e95cb64c",
   "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>nt_seq</th>\n",
       "      <th>aa_seq</th>\n",
       "      <th>WT</th>\n",
       "      <th>STOP</th>\n",
       "      <th>STOP_readthrough</th>\n",
       "      <th>Nham_nt</th>\n",
       "      <th>Nham_aa</th>\n",
       "      <th>Nmut_codons</th>\n",
       "      <th>input1</th>\n",
       "      <th>output1_e1_s1_b1_count</th>\n",
       "      <th>barcode_valid</th>\n",
       "      <th>indel</th>\n",
       "      <th>constant_region</th>\n",
       "      <th>permitted</th>\n",
       "      <th>too_many_substitutions</th>\n",
       "      <th>mixed_substitutions</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>aagtgggtgagacgttatacacgt</td>\n",
       "      <td>KWVRRYTR</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>5.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                     nt_seq    aa_seq    WT   STOP  STOP_readthrough  Nham_nt  \\\n",
       "0  aagtgggtgagacgttatacacgt  KWVRRYTR  <NA>  False             False      5.0   \n",
       "\n",
       "   Nham_aa  Nmut_codons  input1  output1_e1_s1_b1_count  barcode_valid  indel  \\\n",
       "0      2.0          2.0       1                       1           True  False   \n",
       "\n",
       "   constant_region permitted too_many_substitutions mixed_substitutions  \n",
       "0             True      True                  False               False  "
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "raw_cols = [x for x in df.columns if 'input' in x]\n",
    "conditions = [x.split('_')[0] for x in raw_cols]\n",
    "df.rename(columns=dict(zip(raw_cols, conditions)), inplace=True)\n",
    "df.head(1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "3c14f9b6-c191-4ccf-9522-2677697cb280",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "# Special case - FKS2-HS1 sequenced during MiSeqRun2 instead of MiSeqRun1\n",
    "if locus == 'FKS2-HS1':\n",
    "    p = re.compile(r'input[DO][1-6]')\n",
    "    unwanted_cols = [x for x in df.columns if bool(p.fullmatch(x))]\n",
    "    conditions = [x for x in conditions if x not in unwanted_cols]\n",
    "    df.drop(unwanted_cols, axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "aa750e51-2689-460b-a69d-e2b4192cb389",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WT (nt): gattgggtgagacgttatacattg | (aa): DWVRRYTL\n"
     ]
    }
   ],
   "source": [
    "# Extract WT sequences (nucleotide and protein)\n",
    "wtseq = df.loc[df.WT==True, 'nt_seq'].values[0]\n",
    "wtaa = df.loc[df.WT==True, 'aa_seq'].values[0]\n",
    "print(f'WT (nt): {wtseq} | (aa): {wtaa}')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f8dcda7d-5ccd-4a7f-8d41-7a2426c2e765",
   "metadata": {},
   "source": [
    "### Mean readcount per variant (raw data)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7964fc9f-62fe-44a2-9041-f857bb44fb38",
   "metadata": {},
   "source": [
    "This matrix shows the average readcount per unique nucleotide sequence per number of mutated codons across samples."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "14c72e56-bd5b-446c-899e-f22533703c9a",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.lines.Line2D at 0x7f40998d0450>"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 600x300 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Purge graph space\n",
    "sns.set(rc = {'figure.figsize':(6,3)})\n",
    "sns.set_theme(style='whitegrid')\n",
    "f, ax = plt.subplots()\n",
    "\n",
    "count_per_mut_codon = df.groupby('Nmut_codons')[conditions].count()\n",
    "mean_rc_per_variant = df.groupby('Nmut_codons')[conditions].mean() / count_per_mut_codon\n",
    "mean_rc_per_variant.reset_index(inplace=True)\n",
    "mean_rc_long = mean_rc_per_variant.melt(id_vars=['Nmut_codons'], value_vars=conditions, var_name='sample')\n",
    "sns.barplot(mean_rc_long, x='Nmut_codons',y='value',hue='sample',palette='hls')\n",
    "plt.yscale('log')\n",
    "plt.ylabel('Average readcount per unique sequence')\n",
    "plt.xlabel('Number of mutated codons')\n",
    "plt.axhline(y=1, linestyle='--', c='k', zorder=-10)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "31139d1f-11d3-4d9a-9f48-04c2bc55cee2",
   "metadata": {},
   "source": [
    "### Overrepresented variants (raw data)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b4076115-05bd-41c9-9043-193498219182",
   "metadata": {},
   "source": [
    "The table below shows the 5 most represented sequences (mean readcount across samples) in % read depth per sample."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "0242b24c-0671-422b-a5d5-ab20a26fdf9e",
   "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>input1</th>\n",
       "      <th>nt_seq</th>\n",
       "      <th>aa_seq</th>\n",
       "      <th>Nmut_codons</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>595</th>\n",
       "      <td>2.78</td>\n",
       "      <td>gattgggtgagacgttatacattg</td>\n",
       "      <td>DWVRRYTL</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>843</th>\n",
       "      <td>0.87</td>\n",
       "      <td>gattgggtgttgcgttatacattg</td>\n",
       "      <td>DWVLRYTL</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>820</th>\n",
       "      <td>0.85</td>\n",
       "      <td>gattgggtgtagcgttatacattg</td>\n",
       "      <td>DWV*RYTL</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>665</th>\n",
       "      <td>0.82</td>\n",
       "      <td>gattgggtgagacgtttgacattg</td>\n",
       "      <td>DWVRRLTL</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>994</th>\n",
       "      <td>0.79</td>\n",
       "      <td>gatttggtgagacgttatacattg</td>\n",
       "      <td>DLVRRYTL</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     input1                    nt_seq    aa_seq  Nmut_codons\n",
       "595    2.78  gattgggtgagacgttatacattg  DWVRRYTL          0.0\n",
       "843    0.87  gattgggtgttgcgttatacattg  DWVLRYTL          1.0\n",
       "820    0.85  gattgggtgtagcgttatacattg  DWV*RYTL          1.0\n",
       "665    0.82  gattgggtgagacgtttgacattg  DWVRRLTL          1.0\n",
       "994    0.79  gatttggtgagacgttatacattg  DLVRRYTL          1.0"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "overrep_index = df[conditions].mean(axis=1).sort_values(ascending=False).index[:5]\n",
    "overrep = 100*round(df.iloc[overrep_index][conditions] / df[conditions].sum(axis=0), 4)\n",
    "pd.concat([overrep, df.iloc[overrep_index][['nt_seq', 'aa_seq', 'Nmut_codons']]], axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "890bdc78-9743-4290-baeb-2ee3c8873b1e",
   "metadata": {},
   "source": [
    "### Calculate % read depth per sample corresponding to variants with no indels + read count above threshold"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "fff39ee2-144e-41c8-bb3c-d2c83e9e589b",
   "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>Total_reads</th>\n",
       "      <th>Reads_no_indels</th>\n",
       "      <th>Reads_no_indels_%</th>\n",
       "      <th>No_indels_acceptable_RC</th>\n",
       "      <th>Acceptable_RC_over_no_indels_%</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>input1</th>\n",
       "      <td>36018</td>\n",
       "      <td>35813</td>\n",
       "      <td>99.4</td>\n",
       "      <td>35016</td>\n",
       "      <td>97.8</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        Total_reads  Reads_no_indels  Reads_no_indels_%  \\\n",
       "input1        36018            35813               99.4   \n",
       "\n",
       "        No_indels_acceptable_RC  Acceptable_RC_over_no_indels_%  \n",
       "input1                    35016                            97.8  "
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Sum all reads for each sample\n",
    "# Check proportion that corresponds to variants with no indels\n",
    "rcNoIndel = df[conditions].sum(axis=0).to_frame('Total_reads')\n",
    "rcNoIndel['Reads_no_indels'] = df[df.indel == False][conditions].sum(axis=0)\n",
    "rcNoIndel['Reads_no_indels_%'] = 100*round(rcNoIndel['Reads_no_indels']/rcNoIndel['Total_reads'], 3)\n",
    "\n",
    "# Check proportion that corresponds to variants with a read count >= 3 in any sample\n",
    "rcNoIndel['No_indels_acceptable_RC'] = df[((df[conditions] > 2).any(axis=1)) & (df.indel == False)][conditions].sum(axis=0)\n",
    "rcNoIndel['Acceptable_RC_over_no_indels_%'] = 100*round(rcNoIndel['No_indels_acceptable_RC']/rcNoIndel['Reads_no_indels'], 3)\n",
    "\n",
    "rcNoIndel"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e0e60bf3-92ce-470b-b3a9-84ede2e47eea",
   "metadata": {},
   "source": [
    "#### ⚠️ **Filtering step** ⚠️ - Discard variants with indels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "2e768508-89da-42ec-b3f4-24e1b01919b3",
   "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>nt_seq</th>\n",
       "      <th>aa_seq</th>\n",
       "      <th>WT</th>\n",
       "      <th>STOP</th>\n",
       "      <th>STOP_readthrough</th>\n",
       "      <th>Nham_nt</th>\n",
       "      <th>Nham_aa</th>\n",
       "      <th>Nmut_codons</th>\n",
       "      <th>input1</th>\n",
       "      <th>output1_e1_s1_b1_count</th>\n",
       "      <th>barcode_valid</th>\n",
       "      <th>indel</th>\n",
       "      <th>constant_region</th>\n",
       "      <th>permitted</th>\n",
       "      <th>too_many_substitutions</th>\n",
       "      <th>mixed_substitutions</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>aagtgggtgagacgttatacacgt</td>\n",
       "      <td>KWVRRYTR</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>5.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>aagtgggtgagacgttatacactg</td>\n",
       "      <td>KWVRRYTL</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>aagtgggtgagacgttatacattg</td>\n",
       "      <td>KWVRRYTL</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>197</td>\n",
       "      <td>197</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>aagtgggtgagacgttatatgttg</td>\n",
       "      <td>KWVRRYML</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>aagtgggtgagacgttatcatttg</td>\n",
       "      <td>KWVRRYHL</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>5.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</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",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1008</th>\n",
       "      <td>ttttgggtgatacgttatacattg</td>\n",
       "      <td>FWVIRYTL</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1009</th>\n",
       "      <td>ttttgggtgggacgttatacattg</td>\n",
       "      <td>FWVGRYTL</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1010</th>\n",
       "      <td>ttttgggttagacgttatacattg</td>\n",
       "      <td>FWVRRYTL</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1011</th>\n",
       "      <td>ttttggttgagacgttatacattg</td>\n",
       "      <td>FWLRRYTL</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1012</th>\n",
       "      <td>ttttgtgtgagacgttatacattg</td>\n",
       "      <td>FCVRRYTL</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1013 rows × 16 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                        nt_seq    aa_seq    WT   STOP  STOP_readthrough  \\\n",
       "0     aagtgggtgagacgttatacacgt  KWVRRYTR  <NA>  False             False   \n",
       "1     aagtgggtgagacgttatacactg  KWVRRYTL  <NA>  False             False   \n",
       "2     aagtgggtgagacgttatacattg  KWVRRYTL  <NA>  False             False   \n",
       "3     aagtgggtgagacgttatatgttg  KWVRRYML  <NA>  False             False   \n",
       "4     aagtgggtgagacgttatcatttg  KWVRRYHL  <NA>  False             False   \n",
       "...                        ...       ...   ...    ...               ...   \n",
       "1008  ttttgggtgatacgttatacattg  FWVIRYTL  <NA>  False             False   \n",
       "1009  ttttgggtgggacgttatacattg  FWVGRYTL  <NA>  False             False   \n",
       "1010  ttttgggttagacgttatacattg  FWVRRYTL  <NA>  False             False   \n",
       "1011  ttttggttgagacgttatacattg  FWLRRYTL  <NA>  False             False   \n",
       "1012  ttttgtgtgagacgttatacattg  FCVRRYTL  <NA>  False             False   \n",
       "\n",
       "      Nham_nt  Nham_aa  Nmut_codons  input1  output1_e1_s1_b1_count  \\\n",
       "0         5.0      2.0          2.0       1                       1   \n",
       "1         3.0      1.0          2.0       1                       1   \n",
       "2         2.0      1.0          1.0     197                     197   \n",
       "3         4.0      2.0          2.0       1                       1   \n",
       "4         5.0      2.0          2.0       1                       1   \n",
       "...       ...      ...          ...     ...                     ...   \n",
       "1008      3.0      2.0          2.0       1                       1   \n",
       "1009      3.0      2.0          2.0       1                       1   \n",
       "1010      3.0      1.0          2.0       1                       1   \n",
       "1011      3.0      2.0          2.0       1                       1   \n",
       "1012      3.0      2.0          2.0       1                       1   \n",
       "\n",
       "      barcode_valid  indel  constant_region permitted too_many_substitutions  \\\n",
       "0              True  False             True      True                  False   \n",
       "1              True  False             True      True                  False   \n",
       "2              True  False             True      True                  False   \n",
       "3              True  False             True      True                  False   \n",
       "4              True  False             True      True                  False   \n",
       "...             ...    ...              ...       ...                    ...   \n",
       "1008           True  False             True      True                  False   \n",
       "1009           True  False             True      True                  False   \n",
       "1010           True  False             True      True                  False   \n",
       "1011           True  False             True      True                  False   \n",
       "1012           True  False             True      True                  False   \n",
       "\n",
       "     mixed_substitutions  \n",
       "0                  False  \n",
       "1                  False  \n",
       "2                  False  \n",
       "3                  False  \n",
       "4                  False  \n",
       "...                  ...  \n",
       "1008               False  \n",
       "1009               False  \n",
       "1010               False  \n",
       "1011               False  \n",
       "1012               False  \n",
       "\n",
       "[1013 rows x 16 columns]"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Keeping only variants with no indels\n",
    "df.drop(df[df.indel].index, inplace=True)\n",
    "df.reset_index(drop=True, inplace=True)\n",
    "df"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3688f2ea-13a2-474b-b75a-caff98dffd12",
   "metadata": {},
   "source": [
    "#### ⚠️ **Filtering step** ⚠️ - Keep only variants with at least 3 reads in at least 1 input replicate"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "1cb73316-1b4b-4093-a37b-e85d7aacda6e",
   "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>nt_seq</th>\n",
       "      <th>aa_seq</th>\n",
       "      <th>WT</th>\n",
       "      <th>STOP</th>\n",
       "      <th>STOP_readthrough</th>\n",
       "      <th>Nham_nt</th>\n",
       "      <th>Nham_aa</th>\n",
       "      <th>Nmut_codons</th>\n",
       "      <th>input1</th>\n",
       "      <th>output1_e1_s1_b1_count</th>\n",
       "      <th>barcode_valid</th>\n",
       "      <th>indel</th>\n",
       "      <th>constant_region</th>\n",
       "      <th>permitted</th>\n",
       "      <th>too_many_substitutions</th>\n",
       "      <th>mixed_substitutions</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>aagtgggtgagacgttatacattg</td>\n",
       "      <td>KWVRRYTL</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>197</td>\n",
       "      <td>197</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>aagttggtgagacgttatacattg</td>\n",
       "      <td>KLVRRYTL</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>aattgggtgagacgttatacattg</td>\n",
       "      <td>NWVRRYTL</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>149</td>\n",
       "      <td>149</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>acgtgggtgagacgttatacattg</td>\n",
       "      <td>TWVRRYTL</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>147</td>\n",
       "      <td>147</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>acttgggtgagacgttatacattg</td>\n",
       "      <td>TWVRRYTL</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>148</td>\n",
       "      <td>148</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</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",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>249</th>\n",
       "      <td>tcttgggtgagacgttatacattg</td>\n",
       "      <td>SWVRRYTL</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>170</td>\n",
       "      <td>170</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>250</th>\n",
       "      <td>tggtgggtgagacgttatacattg</td>\n",
       "      <td>WWVRRYTL</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>126</td>\n",
       "      <td>126</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>251</th>\n",
       "      <td>tgttgggtgagacgttatacattg</td>\n",
       "      <td>CWVRRYTL</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>135</td>\n",
       "      <td>135</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>252</th>\n",
       "      <td>ttgtgggtgagacgttatacattg</td>\n",
       "      <td>LWVRRYTL</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>258</td>\n",
       "      <td>258</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>253</th>\n",
       "      <td>ttttgggtgagacgttatacattg</td>\n",
       "      <td>FWVRRYTL</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>256</td>\n",
       "      <td>256</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>254 rows × 16 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                       nt_seq    aa_seq    WT   STOP  STOP_readthrough  \\\n",
       "0    aagtgggtgagacgttatacattg  KWVRRYTL  <NA>  False             False   \n",
       "1    aagttggtgagacgttatacattg  KLVRRYTL  <NA>  False             False   \n",
       "2    aattgggtgagacgttatacattg  NWVRRYTL  <NA>  False             False   \n",
       "3    acgtgggtgagacgttatacattg  TWVRRYTL  <NA>  False             False   \n",
       "4    acttgggtgagacgttatacattg  TWVRRYTL  <NA>  False             False   \n",
       "..                        ...       ...   ...    ...               ...   \n",
       "249  tcttgggtgagacgttatacattg  SWVRRYTL  <NA>  False             False   \n",
       "250  tggtgggtgagacgttatacattg  WWVRRYTL  <NA>  False             False   \n",
       "251  tgttgggtgagacgttatacattg  CWVRRYTL  <NA>  False             False   \n",
       "252  ttgtgggtgagacgttatacattg  LWVRRYTL  <NA>  False             False   \n",
       "253  ttttgggtgagacgttatacattg  FWVRRYTL  <NA>  False             False   \n",
       "\n",
       "     Nham_nt  Nham_aa  Nmut_codons  input1  output1_e1_s1_b1_count  \\\n",
       "0        2.0      1.0          1.0     197                     197   \n",
       "1        3.0      2.0          2.0       3                       3   \n",
       "2        1.0      1.0          1.0     149                     149   \n",
       "3        3.0      1.0          1.0     147                     147   \n",
       "4        2.0      1.0          1.0     148                     148   \n",
       "..       ...      ...          ...     ...                     ...   \n",
       "249      2.0      1.0          1.0     170                     170   \n",
       "250      3.0      1.0          1.0     126                     126   \n",
       "251      2.0      1.0          1.0     135                     135   \n",
       "252      3.0      1.0          1.0     258                     258   \n",
       "253      2.0      1.0          1.0     256                     256   \n",
       "\n",
       "     barcode_valid  indel  constant_region permitted too_many_substitutions  \\\n",
       "0             True  False             True      True                  False   \n",
       "1             True  False             True      True                  False   \n",
       "2             True  False             True      True                  False   \n",
       "3             True  False             True      True                  False   \n",
       "4             True  False             True      True                  False   \n",
       "..             ...    ...              ...       ...                    ...   \n",
       "249           True  False             True      True                  False   \n",
       "250           True  False             True      True                  False   \n",
       "251           True  False             True      True                  False   \n",
       "252           True  False             True      True                  False   \n",
       "253           True  False             True      True                  False   \n",
       "\n",
       "    mixed_substitutions  \n",
       "0                 False  \n",
       "1                 False  \n",
       "2                 False  \n",
       "3                 False  \n",
       "4                 False  \n",
       "..                  ...  \n",
       "249               False  \n",
       "250               False  \n",
       "251               False  \n",
       "252               False  \n",
       "253               False  \n",
       "\n",
       "[254 rows x 16 columns]"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "inputRCdf = df[(df[conditions] > 2).any(axis=1)].reset_index(drop=True)\n",
    "inputRCdf"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2d2df4aa-1aaf-466e-9150-8f8937abf5fe",
   "metadata": {},
   "source": [
    "### Label variants with confidence score (related to readcount)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "b2cb9800-befa-483c-8417-52af90f56642",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "def get_confidence_score(g, thresh_high, thresh_mid):\n",
    "    # The read count in all input replicates is parsed to label each variant with a confidence score\n",
    "    # Score is an integer for easy aggregation of synonymous codons\n",
    "    if (g >= thresh_high).all():\n",
    "        return 1 # high confidence score\n",
    "    elif (g > thresh_mid).any():\n",
    "        return 2 # medium confidence score\n",
    "    else:\n",
    "        return 3 # low confidence score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "f93ed480-b24c-4ee8-b213-2076adc31902",
   "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>nt_seq</th>\n",
       "      <th>aa_seq</th>\n",
       "      <th>WT</th>\n",
       "      <th>STOP</th>\n",
       "      <th>STOP_readthrough</th>\n",
       "      <th>Nham_nt</th>\n",
       "      <th>Nham_aa</th>\n",
       "      <th>Nmut_codons</th>\n",
       "      <th>input1</th>\n",
       "      <th>output1_e1_s1_b1_count</th>\n",
       "      <th>barcode_valid</th>\n",
       "      <th>indel</th>\n",
       "      <th>constant_region</th>\n",
       "      <th>permitted</th>\n",
       "      <th>too_many_substitutions</th>\n",
       "      <th>mixed_substitutions</th>\n",
       "      <th>confidence_score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>aagtgggtgagacgttatacattg</td>\n",
       "      <td>KWVRRYTL</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>197</td>\n",
       "      <td>197</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>aagttggtgagacgttatacattg</td>\n",
       "      <td>KLVRRYTL</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>aattgggtgagacgttatacattg</td>\n",
       "      <td>NWVRRYTL</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>149</td>\n",
       "      <td>149</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>acgtgggtgagacgttatacattg</td>\n",
       "      <td>TWVRRYTL</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>147</td>\n",
       "      <td>147</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>acttgggtgagacgttatacattg</td>\n",
       "      <td>TWVRRYTL</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>148</td>\n",
       "      <td>148</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>1</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",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>249</th>\n",
       "      <td>tcttgggtgagacgttatacattg</td>\n",
       "      <td>SWVRRYTL</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>170</td>\n",
       "      <td>170</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>250</th>\n",
       "      <td>tggtgggtgagacgttatacattg</td>\n",
       "      <td>WWVRRYTL</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>126</td>\n",
       "      <td>126</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>251</th>\n",
       "      <td>tgttgggtgagacgttatacattg</td>\n",
       "      <td>CWVRRYTL</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>135</td>\n",
       "      <td>135</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>252</th>\n",
       "      <td>ttgtgggtgagacgttatacattg</td>\n",
       "      <td>LWVRRYTL</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>258</td>\n",
       "      <td>258</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>253</th>\n",
       "      <td>ttttgggtgagacgttatacattg</td>\n",
       "      <td>FWVRRYTL</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>256</td>\n",
       "      <td>256</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>254 rows × 17 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                       nt_seq    aa_seq    WT   STOP  STOP_readthrough  \\\n",
       "0    aagtgggtgagacgttatacattg  KWVRRYTL  <NA>  False             False   \n",
       "1    aagttggtgagacgttatacattg  KLVRRYTL  <NA>  False             False   \n",
       "2    aattgggtgagacgttatacattg  NWVRRYTL  <NA>  False             False   \n",
       "3    acgtgggtgagacgttatacattg  TWVRRYTL  <NA>  False             False   \n",
       "4    acttgggtgagacgttatacattg  TWVRRYTL  <NA>  False             False   \n",
       "..                        ...       ...   ...    ...               ...   \n",
       "249  tcttgggtgagacgttatacattg  SWVRRYTL  <NA>  False             False   \n",
       "250  tggtgggtgagacgttatacattg  WWVRRYTL  <NA>  False             False   \n",
       "251  tgttgggtgagacgttatacattg  CWVRRYTL  <NA>  False             False   \n",
       "252  ttgtgggtgagacgttatacattg  LWVRRYTL  <NA>  False             False   \n",
       "253  ttttgggtgagacgttatacattg  FWVRRYTL  <NA>  False             False   \n",
       "\n",
       "     Nham_nt  Nham_aa  Nmut_codons  input1  output1_e1_s1_b1_count  \\\n",
       "0        2.0      1.0          1.0     197                     197   \n",
       "1        3.0      2.0          2.0       3                       3   \n",
       "2        1.0      1.0          1.0     149                     149   \n",
       "3        3.0      1.0          1.0     147                     147   \n",
       "4        2.0      1.0          1.0     148                     148   \n",
       "..       ...      ...          ...     ...                     ...   \n",
       "249      2.0      1.0          1.0     170                     170   \n",
       "250      3.0      1.0          1.0     126                     126   \n",
       "251      2.0      1.0          1.0     135                     135   \n",
       "252      3.0      1.0          1.0     258                     258   \n",
       "253      2.0      1.0          1.0     256                     256   \n",
       "\n",
       "     barcode_valid  indel  constant_region permitted too_many_substitutions  \\\n",
       "0             True  False             True      True                  False   \n",
       "1             True  False             True      True                  False   \n",
       "2             True  False             True      True                  False   \n",
       "3             True  False             True      True                  False   \n",
       "4             True  False             True      True                  False   \n",
       "..             ...    ...              ...       ...                    ...   \n",
       "249           True  False             True      True                  False   \n",
       "250           True  False             True      True                  False   \n",
       "251           True  False             True      True                  False   \n",
       "252           True  False             True      True                  False   \n",
       "253           True  False             True      True                  False   \n",
       "\n",
       "    mixed_substitutions  confidence_score  \n",
       "0                 False                 1  \n",
       "1                 False                 3  \n",
       "2                 False                 1  \n",
       "3                 False                 1  \n",
       "4                 False                 1  \n",
       "..                  ...               ...  \n",
       "249               False                 1  \n",
       "250               False                 1  \n",
       "251               False                 1  \n",
       "252               False                 1  \n",
       "253               False                 1  \n",
       "\n",
       "[254 rows x 17 columns]"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.set_option('mode.chained_assignment', None) # Disable settingWithCopy warning\n",
    "inputRCdf['confidence_score'] = inputRCdf[conditions].apply(lambda row: get_confidence_score(row, min_input_readcount_high, min_input_readcount_mid), axis=1)\n",
    "inputRCdf"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b5f616c6-69ce-4088-8f44-85d1b160f5cc",
   "metadata": {},
   "source": [
    "### Merge with dataframe of expected variants"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "a02389a6-9342-47ed-b63c-1bd5a52208b8",
   "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>seq_type</th>\n",
       "      <th>nt_seq</th>\n",
       "      <th>aa_pos</th>\n",
       "      <th>alt_codons</th>\n",
       "      <th>mutated_codon</th>\n",
       "      <th>combination</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>WT</td>\n",
       "      <td>gattgggtgagacgttatacattg</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>single</td>\n",
       "      <td>ttttgggtgagacgttatacattg</td>\n",
       "      <td>0.0</td>\n",
       "      <td>ttt</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>single</td>\n",
       "      <td>ttgtgggtgagacgttatacattg</td>\n",
       "      <td>0.0</td>\n",
       "      <td>ttg</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>single</td>\n",
       "      <td>ctttgggtgagacgttatacattg</td>\n",
       "      <td>0.0</td>\n",
       "      <td>ctt</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>single</td>\n",
       "      <td>ctgtgggtgagacgttatacattg</td>\n",
       "      <td>0.0</td>\n",
       "      <td>ctg</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</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>55498</th>\n",
       "      <td>ortho</td>\n",
       "      <td>gattggatttatagatctattttg</td>\n",
       "      <td>2.0</td>\n",
       "      <td>att</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>55499</th>\n",
       "      <td>ortho</td>\n",
       "      <td>gattggatttatagatctattttg</td>\n",
       "      <td>3.0</td>\n",
       "      <td>tat</td>\n",
       "      <td>2.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>55500</th>\n",
       "      <td>ortho</td>\n",
       "      <td>gattggatttatagatctattttg</td>\n",
       "      <td>4.0</td>\n",
       "      <td>aga</td>\n",
       "      <td>3.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>55501</th>\n",
       "      <td>ortho</td>\n",
       "      <td>gattggatttatagatctattttg</td>\n",
       "      <td>5.0</td>\n",
       "      <td>tct</td>\n",
       "      <td>4.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>55502</th>\n",
       "      <td>ortho</td>\n",
       "      <td>gattggatttatagatctattttg</td>\n",
       "      <td>6.0</td>\n",
       "      <td>att</td>\n",
       "      <td>5.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>55503 rows × 6 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      seq_type                    nt_seq  aa_pos alt_codons  mutated_codon  \\\n",
       "0           WT  gattgggtgagacgttatacattg     NaN        NaN            NaN   \n",
       "1       single  ttttgggtgagacgttatacattg     0.0        ttt            1.0   \n",
       "2       single  ttgtgggtgagacgttatacattg     0.0        ttg            1.0   \n",
       "3       single  ctttgggtgagacgttatacattg     0.0        ctt            1.0   \n",
       "4       single  ctgtgggtgagacgttatacattg     0.0        ctg            1.0   \n",
       "...        ...                       ...     ...        ...            ...   \n",
       "55498    ortho  gattggatttatagatctattttg     2.0        att            1.0   \n",
       "55499    ortho  gattggatttatagatctattttg     3.0        tat            2.0   \n",
       "55500    ortho  gattggatttatagatctattttg     4.0        aga            3.0   \n",
       "55501    ortho  gattggatttatagatctattttg     5.0        tct            4.0   \n",
       "55502    ortho  gattggatttatagatctattttg     6.0        att            5.0   \n",
       "\n",
       "      combination  \n",
       "0             NaN  \n",
       "1             NaN  \n",
       "2             NaN  \n",
       "3             NaN  \n",
       "4             NaN  \n",
       "...           ...  \n",
       "55498         NaN  \n",
       "55499         NaN  \n",
       "55500         NaN  \n",
       "55501         NaN  \n",
       "55502         NaN  \n",
       "\n",
       "[55503 rows x 6 columns]"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Dataframe containing all expected mutants for the different libraries (single, double, ortholog) is imported\n",
    "# /!\\ Important to note that I did not actually separate by type of library /!\\\n",
    "# \"Double\" mutants do not include single mutants\n",
    "# \"Ortho\" are standalone (double mutants in orthologous sequences are duplicated in \"double\" and \"ortho\" seq types)\n",
    "source = metadata_path+locus+f'/{locus}_var_df.csv'\n",
    "vardf = pd.read_csv(source, header=0, index_col=0)\n",
    "vardf"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "39e0f228-6ef7-417a-af4a-edaa22108073",
   "metadata": {},
   "source": [
    "#### ⚠️ **Filtering step** ⚠️ - Keep only mutants corresponding to pool_type"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "2c75a014-5f0e-4537-ad6e-437bb061924e",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "251"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Depending on the type of pool (single or double), expected variants are filtered\n",
    "# The dictionary specifies which sequence types must be kept\n",
    "keep_seq_type = {'single' : ['WT', 'single'#, 'ortho'\n",
    "                            ]#,\n",
    "                 #'double': ['WT', 'single', 'double']\n",
    "                }\n",
    "\n",
    "var_filtered_df = vardf[vardf.seq_type.isin(keep_seq_type[pool_type])]\n",
    "len(var_filtered_df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "3be68045-d024-4ebc-958d-a16f32a59930",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "# Here we need to translate expected variants to aa,.. which will also us to count the expected number of unique aa variants\n",
    "# I did not do this in the first place because I didn't need it to merge with the dataframe of variant counts\n",
    "codon_table = pd.read_csv(codon_table_path, header=0)\n",
    "codon_table['codon'] = codon_table['codon'].str.lower()\n",
    "codon_dic = dict(zip(codon_table['codon'], codon_table['aminoacid']))\n",
    "def get_aa_seq(seq, cdic):\n",
    "    clist = [seq[i:i+3] for i in range(0, len(seq), 3)] # Converting nucleotide sequence to list of codons\n",
    "    return ''.join([cdic[x] for x in clist])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "0cdac710-9de7-43b9-b1d9-65e5e77d534c",
   "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>seq_type</th>\n",
       "      <th>nt_seq</th>\n",
       "      <th>aa_pos</th>\n",
       "      <th>alt_codons</th>\n",
       "      <th>mutated_codon</th>\n",
       "      <th>combination</th>\n",
       "      <th>aa_seq</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>WT</td>\n",
       "      <td>gattgggtgagacgttatacattg</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>DWVRRYTL</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>single</td>\n",
       "      <td>ttttgggtgagacgttatacattg</td>\n",
       "      <td>0.0</td>\n",
       "      <td>ttt</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>FWVRRYTL</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>single</td>\n",
       "      <td>ttgtgggtgagacgttatacattg</td>\n",
       "      <td>0.0</td>\n",
       "      <td>ttg</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>LWVRRYTL</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>single</td>\n",
       "      <td>ctttgggtgagacgttatacattg</td>\n",
       "      <td>0.0</td>\n",
       "      <td>ctt</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>LWVRRYTL</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>single</td>\n",
       "      <td>ctgtgggtgagacgttatacattg</td>\n",
       "      <td>0.0</td>\n",
       "      <td>ctg</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>LWVRRYTL</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>246</th>\n",
       "      <td>single</td>\n",
       "      <td>gattgggtgagacgttatacacgg</td>\n",
       "      <td>7.0</td>\n",
       "      <td>cgg</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>DWVRRYTR</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>247</th>\n",
       "      <td>single</td>\n",
       "      <td>gattgggtgagacgttatacaagt</td>\n",
       "      <td>7.0</td>\n",
       "      <td>agt</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>DWVRRYTS</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>248</th>\n",
       "      <td>single</td>\n",
       "      <td>gattgggtgagacgttatacaagg</td>\n",
       "      <td>7.0</td>\n",
       "      <td>agg</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>DWVRRYTR</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>249</th>\n",
       "      <td>single</td>\n",
       "      <td>gattgggtgagacgttatacaggt</td>\n",
       "      <td>7.0</td>\n",
       "      <td>ggt</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>DWVRRYTG</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>250</th>\n",
       "      <td>single</td>\n",
       "      <td>gattgggtgagacgttatacaggg</td>\n",
       "      <td>7.0</td>\n",
       "      <td>ggg</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>DWVRRYTG</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>251 rows × 7 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "    seq_type                    nt_seq  aa_pos alt_codons  mutated_codon  \\\n",
       "0         WT  gattgggtgagacgttatacattg     NaN        NaN            NaN   \n",
       "1     single  ttttgggtgagacgttatacattg     0.0        ttt            1.0   \n",
       "2     single  ttgtgggtgagacgttatacattg     0.0        ttg            1.0   \n",
       "3     single  ctttgggtgagacgttatacattg     0.0        ctt            1.0   \n",
       "4     single  ctgtgggtgagacgttatacattg     0.0        ctg            1.0   \n",
       "..       ...                       ...     ...        ...            ...   \n",
       "246   single  gattgggtgagacgttatacacgg     7.0        cgg            1.0   \n",
       "247   single  gattgggtgagacgttatacaagt     7.0        agt            1.0   \n",
       "248   single  gattgggtgagacgttatacaagg     7.0        agg            1.0   \n",
       "249   single  gattgggtgagacgttatacaggt     7.0        ggt            1.0   \n",
       "250   single  gattgggtgagacgttatacaggg     7.0        ggg            1.0   \n",
       "\n",
       "    combination    aa_seq  \n",
       "0           NaN  DWVRRYTL  \n",
       "1           NaN  FWVRRYTL  \n",
       "2           NaN  LWVRRYTL  \n",
       "3           NaN  LWVRRYTL  \n",
       "4           NaN  LWVRRYTL  \n",
       "..          ...       ...  \n",
       "246         NaN  DWVRRYTR  \n",
       "247         NaN  DWVRRYTS  \n",
       "248         NaN  DWVRRYTR  \n",
       "249         NaN  DWVRRYTG  \n",
       "250         NaN  DWVRRYTG  \n",
       "\n",
       "[251 rows x 7 columns]"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "var_filtered_df['aa_seq'] = var_filtered_df.nt_seq.apply(lambda x: get_aa_seq(x, codon_dic))\n",
    "var_filtered_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "8cf5a749-4bdd-40ff-8654-515882e59d97",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "251"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(var_filtered_df.nt_seq.unique())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "3b91d5a4-edcc-45d8-9aa5-88d20b59461d",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "161"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "unique_aa = len(var_filtered_df.aa_seq.unique())\n",
    "unique_aa"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "257e4367-c8e9-4145-8667-9d2c085b34d1",
   "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>seq_type</th>\n",
       "      <th>nt_seq</th>\n",
       "      <th>aa_pos</th>\n",
       "      <th>alt_codons</th>\n",
       "      <th>mutated_codon</th>\n",
       "      <th>combination</th>\n",
       "      <th>aa_seq</th>\n",
       "      <th>WT</th>\n",
       "      <th>STOP</th>\n",
       "      <th>STOP_readthrough</th>\n",
       "      <th>...</th>\n",
       "      <th>input1</th>\n",
       "      <th>output1_e1_s1_b1_count</th>\n",
       "      <th>barcode_valid</th>\n",
       "      <th>indel</th>\n",
       "      <th>constant_region</th>\n",
       "      <th>permitted</th>\n",
       "      <th>too_many_substitutions</th>\n",
       "      <th>mixed_substitutions</th>\n",
       "      <th>confidence_score</th>\n",
       "      <th>Location</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>WT</td>\n",
       "      <td>gattgggtgagacgttatacattg</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>DWVRRYTL</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>...</td>\n",
       "      <td>1003</td>\n",
       "      <td>1003</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>1</td>\n",
       "      <td>both</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>single</td>\n",
       "      <td>ttttgggtgagacgttatacattg</td>\n",
       "      <td>0.0</td>\n",
       "      <td>ttt</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>FWVRRYTL</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>...</td>\n",
       "      <td>256</td>\n",
       "      <td>256</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>1</td>\n",
       "      <td>both</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>single</td>\n",
       "      <td>ttgtgggtgagacgttatacattg</td>\n",
       "      <td>0.0</td>\n",
       "      <td>ttg</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>LWVRRYTL</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>...</td>\n",
       "      <td>258</td>\n",
       "      <td>258</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>1</td>\n",
       "      <td>both</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>single</td>\n",
       "      <td>ctttgggtgagacgttatacattg</td>\n",
       "      <td>0.0</td>\n",
       "      <td>ctt</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>LWVRRYTL</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>...</td>\n",
       "      <td>164</td>\n",
       "      <td>164</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>1</td>\n",
       "      <td>both</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>single</td>\n",
       "      <td>ctgtgggtgagacgttatacattg</td>\n",
       "      <td>0.0</td>\n",
       "      <td>ctg</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>LWVRRYTL</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>...</td>\n",
       "      <td>162</td>\n",
       "      <td>162</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>1</td>\n",
       "      <td>both</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",
       "      <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>249</th>\n",
       "      <td>single</td>\n",
       "      <td>gattgggtgagacgttatacaggt</td>\n",
       "      <td>7.0</td>\n",
       "      <td>ggt</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>DWVRRYTG</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>...</td>\n",
       "      <td>73</td>\n",
       "      <td>73</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>1</td>\n",
       "      <td>both</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>250</th>\n",
       "      <td>single</td>\n",
       "      <td>gattgggtgagacgttatacaggg</td>\n",
       "      <td>7.0</td>\n",
       "      <td>ggg</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>DWVRRYTG</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>...</td>\n",
       "      <td>86</td>\n",
       "      <td>86</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>1</td>\n",
       "      <td>both</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>251</th>\n",
       "      <td>NaN</td>\n",
       "      <td>aagttggtgagacgttatacattg</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>KLVRRYTL</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>...</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>3</td>\n",
       "      <td>right_only</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>252</th>\n",
       "      <td>NaN</td>\n",
       "      <td>gattggttgagacgttataggttg</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>DWLRRYRL</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>...</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>3</td>\n",
       "      <td>right_only</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>253</th>\n",
       "      <td>NaN</td>\n",
       "      <td>gattggttgagatgttatacattg</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>DWLRCYTL</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>...</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>3</td>\n",
       "      <td>right_only</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>254 rows × 23 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "    seq_type                    nt_seq  aa_pos alt_codons  mutated_codon  \\\n",
       "0         WT  gattgggtgagacgttatacattg     NaN        NaN            NaN   \n",
       "1     single  ttttgggtgagacgttatacattg     0.0        ttt            1.0   \n",
       "2     single  ttgtgggtgagacgttatacattg     0.0        ttg            1.0   \n",
       "3     single  ctttgggtgagacgttatacattg     0.0        ctt            1.0   \n",
       "4     single  ctgtgggtgagacgttatacattg     0.0        ctg            1.0   \n",
       "..       ...                       ...     ...        ...            ...   \n",
       "249   single  gattgggtgagacgttatacaggt     7.0        ggt            1.0   \n",
       "250   single  gattgggtgagacgttatacaggg     7.0        ggg            1.0   \n",
       "251      NaN  aagttggtgagacgttatacattg     NaN        NaN            NaN   \n",
       "252      NaN  gattggttgagacgttataggttg     NaN        NaN            NaN   \n",
       "253      NaN  gattggttgagatgttatacattg     NaN        NaN            NaN   \n",
       "\n",
       "    combination    aa_seq    WT   STOP  STOP_readthrough  ...  input1  \\\n",
       "0           NaN  DWVRRYTL  True  False             False  ...    1003   \n",
       "1           NaN  FWVRRYTL  <NA>  False             False  ...     256   \n",
       "2           NaN  LWVRRYTL  <NA>  False             False  ...     258   \n",
       "3           NaN  LWVRRYTL  <NA>  False             False  ...     164   \n",
       "4           NaN  LWVRRYTL  <NA>  False             False  ...     162   \n",
       "..          ...       ...   ...    ...               ...  ...     ...   \n",
       "249         NaN  DWVRRYTG  <NA>  False             False  ...      73   \n",
       "250         NaN  DWVRRYTG  <NA>  False             False  ...      86   \n",
       "251         NaN  KLVRRYTL  <NA>  False             False  ...       3   \n",
       "252         NaN  DWLRRYRL  <NA>  False             False  ...       3   \n",
       "253         NaN  DWLRCYTL  <NA>  False             False  ...       3   \n",
       "\n",
       "     output1_e1_s1_b1_count  barcode_valid  indel  constant_region  permitted  \\\n",
       "0                      1003           True  False             True       True   \n",
       "1                       256           True  False             True       True   \n",
       "2                       258           True  False             True       True   \n",
       "3                       164           True  False             True       True   \n",
       "4                       162           True  False             True       True   \n",
       "..                      ...            ...    ...              ...        ...   \n",
       "249                      73           True  False             True       True   \n",
       "250                      86           True  False             True       True   \n",
       "251                       3           True  False             True       True   \n",
       "252                       3           True  False             True       True   \n",
       "253                       3           True  False             True       True   \n",
       "\n",
       "     too_many_substitutions  mixed_substitutions confidence_score    Location  \n",
       "0                     False                False                1        both  \n",
       "1                     False                False                1        both  \n",
       "2                     False                False                1        both  \n",
       "3                     False                False                1        both  \n",
       "4                     False                False                1        both  \n",
       "..                      ...                  ...              ...         ...  \n",
       "249                   False                False                1        both  \n",
       "250                   False                False                1        both  \n",
       "251                   False                False                3  right_only  \n",
       "252                   False                False                3  right_only  \n",
       "253                   False                False                3  right_only  \n",
       "\n",
       "[254 rows x 23 columns]"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# The dataframe of sequenced variants ( with no indels and readcount > threshold in all input replicates) and the dataframe of expected mutants are compared\n",
    "# The column 'Location' indicates if a variant is found only in the set of expected variants (left), only in the sequenced dataset (right), or in both (covered variants)\n",
    "\n",
    "comparedf = pd.merge(left=var_filtered_df.drop(columns='aa_seq'), right=inputRCdf, how='outer', on='nt_seq', indicator='Location')\n",
    "comparedf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "dacdba6b-b384-4ff0-89d6-7e35f9f45aff",
   "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>Filtered_sequenced_variants</th>\n",
       "      <th>Filtered_sequenced_variants_rc</th>\n",
       "      <th>Usable_variants</th>\n",
       "      <th>AA</th>\n",
       "      <th>Covered_variants_%</th>\n",
       "      <th>AA_%</th>\n",
       "      <th>Usable_variants_rc</th>\n",
       "      <th>Usable_variants_%rc</th>\n",
       "      <th>Unsequenced_variants</th>\n",
       "      <th>Unexpected_variants</th>\n",
       "      <th>Unexpected_variants_rc</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>input1</th>\n",
       "      <td>254</td>\n",
       "      <td>35016</td>\n",
       "      <td>251</td>\n",
       "      <td>161</td>\n",
       "      <td>100.0</td>\n",
       "      <td>100.0</td>\n",
       "      <td>35007</td>\n",
       "      <td>100.0</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        Filtered_sequenced_variants  Filtered_sequenced_variants_rc  \\\n",
       "input1                          254                           35016   \n",
       "\n",
       "        Usable_variants   AA  Covered_variants_%   AA_%  Usable_variants_rc  \\\n",
       "input1              251  161               100.0  100.0               35007   \n",
       "\n",
       "        Usable_variants_%rc  Unsequenced_variants  Unexpected_variants  \\\n",
       "input1                100.0                     0                    3   \n",
       "\n",
       "        Unexpected_variants_rc  \n",
       "input1                       9  "
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# From that we build a summary with number of variants and read counts across samples\n",
    "# Careful, some variants and their associated reads are shared among different samples\n",
    "seqdf = inputRCdf[conditions].astype(bool).sum(axis=0).to_frame('Filtered_sequenced_variants') # No duplicates in this dataframe - corresponds to sequenced variants (expected AND unexpected)\n",
    "seqdf['Filtered_sequenced_variants_rc'] = inputRCdf[conditions].sum(axis=0)\n",
    "seqdf['Usable_variants'] = comparedf[comparedf.Location.str.contains('both')].groupby(['aa_seq','nt_seq'])[conditions].first().astype(bool).sum(axis=0)\n",
    "seqdf['AA'] = comparedf[comparedf.Location.str.contains('both')].groupby(['aa_seq','nt_seq'])[conditions].first().groupby('aa_seq')[conditions].sum().astype(bool).sum(axis=0)\n",
    "# Below we calculate how many variants (of the proper seq type) are found in our sequencing dataset after filtering steps\n",
    "# Only the highest percentage value is relevant, since variants are shared between input and output samples (some of them being depleted after selection)\n",
    "# Expected but unsequenced variants might have been absent from our initial libraries of mutants\n",
    "seqdf['Covered_variants_%'] = 100*round(seqdf['Usable_variants']/len(var_filtered_df.nt_seq.unique()),3)\n",
    "seqdf['AA_%'] = 100*round(seqdf['AA']/unique_aa,3)\n",
    "seqdf['Usable_variants_rc'] = comparedf[comparedf.Location.str.contains('both')].groupby('nt_seq')[conditions].first().sum(axis=0)\n",
    "seqdf['Usable_variants_%rc'] = 100*round(seqdf['Usable_variants_rc']/seqdf['Filtered_sequenced_variants_rc'],3)\n",
    "seqdf['Unsequenced_variants'] = comparedf[comparedf.Location.str.contains('left')].groupby('nt_seq')[conditions].first().astype(bool).sum(axis=0)\n",
    "# Readcount for unsequenced variants is by definition = 0\n",
    "seqdf['Unexpected_variants'] = comparedf[comparedf.Location.str.contains('right')][conditions].astype(bool).sum(axis=0)\n",
    "seqdf['Unexpected_variants_rc'] = comparedf[comparedf.Location.str.contains('right')][conditions].sum(axis=0)\n",
    "# %Readcount for unexpected variants is always 100 - %rc for covered variants\n",
    "\n",
    "seqdf"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e58468a4-c211-46ad-90dd-6ce083b1a4ac",
   "metadata": {},
   "source": [
    "### Read counts on **sequenced** variants"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5f1139b3-712e-4cdb-bf0e-df7c30248ee1",
   "metadata": {},
   "source": [
    "#### ⚠️ **Filtering step** ⚠️ - Keep only expected variants"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "58f60ed3-5aeb-4da0-8204-13e6d1fdc54b",
   "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>seq_type</th>\n",
       "      <th>nt_seq</th>\n",
       "      <th>aa_pos</th>\n",
       "      <th>alt_codons</th>\n",
       "      <th>mutated_codon</th>\n",
       "      <th>combination</th>\n",
       "      <th>aa_seq</th>\n",
       "      <th>WT</th>\n",
       "      <th>STOP</th>\n",
       "      <th>STOP_readthrough</th>\n",
       "      <th>...</th>\n",
       "      <th>input1</th>\n",
       "      <th>output1_e1_s1_b1_count</th>\n",
       "      <th>barcode_valid</th>\n",
       "      <th>indel</th>\n",
       "      <th>constant_region</th>\n",
       "      <th>permitted</th>\n",
       "      <th>too_many_substitutions</th>\n",
       "      <th>mixed_substitutions</th>\n",
       "      <th>confidence_score</th>\n",
       "      <th>Location</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>0 rows × 23 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "Empty DataFrame\n",
       "Columns: [seq_type, nt_seq, aa_pos, alt_codons, mutated_codon, combination, aa_seq, WT, STOP, STOP_readthrough, Nham_nt, Nham_aa, Nmut_codons, input1, output1_e1_s1_b1_count, barcode_valid, indel, constant_region, permitted, too_many_substitutions, mixed_substitutions, confidence_score, Location]\n",
       "Index: []\n",
       "\n",
       "[0 rows x 23 columns]"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.set_option('mode.chained_assignment', None) # Disable settingWithCopy warning\n",
    "covereddf = comparedf[comparedf.Location.str.contains('both')].reset_index(drop=True)\n",
    "covereddf[covereddf.duplicated(subset='nt_seq', keep=False)].head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "0072ab87-6c1e-4851-b422-9cc842a987f0",
   "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>total_readcount</th>\n",
       "      <th>WT</th>\n",
       "      <th>WT_%</th>\n",
       "      <th>NhamAA1_rc</th>\n",
       "      <th>NhamAA1_rc_%</th>\n",
       "      <th>NhamAA2_rc</th>\n",
       "      <th>NhamAA2_rc_%</th>\n",
       "      <th>NhamAA3+_rc</th>\n",
       "      <th>NhamAA3+_rc_%</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>input1</th>\n",
       "      <td>35007</td>\n",
       "      <td>2161</td>\n",
       "      <td>6.2</td>\n",
       "      <td>32846</td>\n",
       "      <td>93.8</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        total_readcount    WT  WT_%  NhamAA1_rc  NhamAA1_rc_%  NhamAA2_rc  \\\n",
       "input1            35007  2161   6.2       32846          93.8           0   \n",
       "\n",
       "        NhamAA2_rc_%  NhamAA3+_rc  NhamAA3+_rc_%  \n",
       "input1           0.0            0            0.0  "
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "totrc = covereddf.groupby('nt_seq')[conditions].first().sum(axis=0).to_frame('total_readcount')\n",
    "totrc['WT'] = covereddf[covereddf.Nham_aa == 0].groupby('nt_seq')[conditions].first().sum(axis=0)\n",
    "totrc['WT_%'] = 100*round(totrc['WT'] / totrc['total_readcount'], 3)\n",
    "totrc['NhamAA1_rc'] = covereddf[covereddf.Nham_aa == 1].groupby('nt_seq')[conditions].first().sum(axis=0)\n",
    "totrc['NhamAA1_rc_%'] = 100*round(totrc['NhamAA1_rc'] / totrc['total_readcount'], 3)\n",
    "totrc['NhamAA2_rc'] = covereddf[covereddf.Nham_aa == 2].groupby('nt_seq')[conditions].first().sum(axis=0)\n",
    "totrc['NhamAA2_rc_%'] = 100*round(totrc['NhamAA2_rc'] / totrc['total_readcount'], 3)\n",
    "totrc['NhamAA3+_rc'] = covereddf[covereddf.Nham_aa > 2].groupby('nt_seq')[conditions].first().sum(axis=0)\n",
    "totrc['NhamAA3+_rc_%'] = 100*round(totrc['NhamAA3+_rc'] / totrc['total_readcount'], 3)\n",
    "        \n",
    "totrc"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c560b2b7-5f80-4e2b-868c-017268ba2468",
   "metadata": {},
   "source": [
    "### Average read frequency"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "d2786b50-5b4d-471a-8ad6-38094b0ed4c5",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "cov_upset = covereddf.groupby(['aa_seq','nt_seq']).agg(dict(zip(conditions+['Nham_aa', 'confidence_score'], ['first']*(len(conditions)+2))))\n",
    "aggdic = dict(zip(conditions, ['median']*len(conditions)))\n",
    "aggdic['Nham_aa'] = 'first'\n",
    "aggdic['confidence_score'] = 'min'\n",
    "aa_upset = cov_upset.groupby('aa_seq').agg(aggdic)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "95946a42-652a-489d-8e5c-1c138a49beff",
   "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>aa_seq</th>\n",
       "      <th>Nham_aa</th>\n",
       "      <th>confidence_score</th>\n",
       "      <th>input1</th>\n",
       "      <th>average_freq_pool</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>*WVRRYTL</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.010655</td>\n",
       "      <td>0.010655</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>AWVRRYTL</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.002830</td>\n",
       "      <td>0.002830</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>CWVRRYTL</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.005969</td>\n",
       "      <td>0.005969</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>D*VRRYTL</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.011804</td>\n",
       "      <td>0.011804</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>DAVRRYTL</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.004421</td>\n",
       "      <td>0.004421</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>156</th>\n",
       "      <td>SWVRRYTL</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.007516</td>\n",
       "      <td>0.007516</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>157</th>\n",
       "      <td>TWVRRYTL</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.006521</td>\n",
       "      <td>0.006521</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>158</th>\n",
       "      <td>VWVRRYTL</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.004974</td>\n",
       "      <td>0.004974</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>159</th>\n",
       "      <td>WWVRRYTL</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.005571</td>\n",
       "      <td>0.005571</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>160</th>\n",
       "      <td>YWVRRYTL</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.008356</td>\n",
       "      <td>0.008356</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>161 rows × 5 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       aa_seq  Nham_aa  confidence_score    input1  average_freq_pool\n",
       "0    *WVRRYTL      1.0                 1  0.010655           0.010655\n",
       "1    AWVRRYTL      1.0                 1  0.002830           0.002830\n",
       "2    CWVRRYTL      1.0                 1  0.005969           0.005969\n",
       "3    D*VRRYTL      1.0                 1  0.011804           0.011804\n",
       "4    DAVRRYTL      1.0                 1  0.004421           0.004421\n",
       "..        ...      ...               ...       ...                ...\n",
       "156  SWVRRYTL      1.0                 1  0.007516           0.007516\n",
       "157  TWVRRYTL      1.0                 1  0.006521           0.006521\n",
       "158  VWVRRYTL      1.0                 1  0.004974           0.004974\n",
       "159  WWVRRYTL      1.0                 1  0.005571           0.005571\n",
       "160  YWVRRYTL      1.0                 1  0.008356           0.008356\n",
       "\n",
       "[161 rows x 5 columns]"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "samples = conditions\n",
    "pool_df = aa_upset.reset_index()[['aa_seq','Nham_aa','confidence_score']+samples]\n",
    "pool_df[samples] = pool_df[samples].div(pool_df[samples].sum(axis=0), axis=1)\n",
    "pool_df['average_freq_pool'] = pool_df[samples].mean(axis=1)\n",
    "pool_df"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "429907bd-989e-45ed-a4c6-3ee033438ffe",
   "metadata": {},
   "source": [
    "### Compare with input in DMS experiment"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "11a0fa03-8fd4-4e42-be2a-5ff7eef57e0a",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "DMSf = f'{DMSdata_path}{strain}_{locus}_single_ortho_none/selcoeff_all_libraries.csv'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "634a3acb-4a85-4727-8cd4-7825950f90d3",
   "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>seq_type</th>\n",
       "      <th>nt_seq</th>\n",
       "      <th>aa_pos</th>\n",
       "      <th>alt_codons</th>\n",
       "      <th>mutated_codon</th>\n",
       "      <th>combination</th>\n",
       "      <th>aa_seq</th>\n",
       "      <th>WT</th>\n",
       "      <th>STOP</th>\n",
       "      <th>STOP_readthrough</th>\n",
       "      <th>...</th>\n",
       "      <th>b3_log10</th>\n",
       "      <th>alt_aa</th>\n",
       "      <th>mutation_type</th>\n",
       "      <th>Log2FC_1</th>\n",
       "      <th>Log2FC_2</th>\n",
       "      <th>Log2FC_3</th>\n",
       "      <th>selcoeff_1</th>\n",
       "      <th>selcoeff_2</th>\n",
       "      <th>selcoeff_3</th>\n",
       "      <th>median_s</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>single</td>\n",
       "      <td>ttttgggtgagacgttatacattg</td>\n",
       "      <td>0.0</td>\n",
       "      <td>ttt</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>FWVRRYTL</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>...</td>\n",
       "      <td>-1.854191</td>\n",
       "      <td>F</td>\n",
       "      <td>non-synonymous</td>\n",
       "      <td>-0.071871</td>\n",
       "      <td>-0.095830</td>\n",
       "      <td>-0.091700</td>\n",
       "      <td>0.017002</td>\n",
       "      <td>0.152116</td>\n",
       "      <td>0.171600</td>\n",
       "      <td>0.152116</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>single</td>\n",
       "      <td>ttgtgggtgagacgttatacattg</td>\n",
       "      <td>0.0</td>\n",
       "      <td>ttg</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>LWVRRYTL</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>...</td>\n",
       "      <td>-2.708009</td>\n",
       "      <td>L</td>\n",
       "      <td>non-synonymous</td>\n",
       "      <td>-0.113504</td>\n",
       "      <td>-0.183024</td>\n",
       "      <td>-0.147477</td>\n",
       "      <td>-0.024631</td>\n",
       "      <td>0.064923</td>\n",
       "      <td>0.115823</td>\n",
       "      <td>0.064923</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>single</td>\n",
       "      <td>ctttgggtgagacgttatacattg</td>\n",
       "      <td>0.0</td>\n",
       "      <td>ctt</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>LWVRRYTL</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>...</td>\n",
       "      <td>-2.958482</td>\n",
       "      <td>L</td>\n",
       "      <td>non-synonymous</td>\n",
       "      <td>-0.057853</td>\n",
       "      <td>-0.136193</td>\n",
       "      <td>-0.068939</td>\n",
       "      <td>0.031020</td>\n",
       "      <td>0.111754</td>\n",
       "      <td>0.194362</td>\n",
       "      <td>0.111754</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>single</td>\n",
       "      <td>ctgtgggtgagacgttatacattg</td>\n",
       "      <td>0.0</td>\n",
       "      <td>ctg</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>LWVRRYTL</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>...</td>\n",
       "      <td>-2.592214</td>\n",
       "      <td>L</td>\n",
       "      <td>non-synonymous</td>\n",
       "      <td>-0.024083</td>\n",
       "      <td>-0.115885</td>\n",
       "      <td>-0.102083</td>\n",
       "      <td>0.064790</td>\n",
       "      <td>0.132062</td>\n",
       "      <td>0.161218</td>\n",
       "      <td>0.132062</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>single</td>\n",
       "      <td>atttgggtgagacgttatacattg</td>\n",
       "      <td>0.0</td>\n",
       "      <td>att</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>IWVRRYTL</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>...</td>\n",
       "      <td>-2.679478</td>\n",
       "      <td>I</td>\n",
       "      <td>non-synonymous</td>\n",
       "      <td>-0.059975</td>\n",
       "      <td>-0.127113</td>\n",
       "      <td>-0.143584</td>\n",
       "      <td>0.028898</td>\n",
       "      <td>0.120834</td>\n",
       "      <td>0.119717</td>\n",
       "      <td>0.119717</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",
       "      <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>550</th>\n",
       "      <td>WT</td>\n",
       "      <td>gattgggtgagacgttatacattg</td>\n",
       "      <td>3.0</td>\n",
       "      <td>aga</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>DWVRRYTL</td>\n",
       "      <td>True</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>R</td>\n",
       "      <td>synonymous</td>\n",
       "      <td>-0.111276</td>\n",
       "      <td>-0.027630</td>\n",
       "      <td>-0.157724</td>\n",
       "      <td>-0.022403</td>\n",
       "      <td>0.220317</td>\n",
       "      <td>0.105577</td>\n",
       "      <td>0.105577</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>551</th>\n",
       "      <td>WT</td>\n",
       "      <td>gattgggtgagacgttatacattg</td>\n",
       "      <td>4.0</td>\n",
       "      <td>cgt</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>DWVRRYTL</td>\n",
       "      <td>True</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>R</td>\n",
       "      <td>synonymous</td>\n",
       "      <td>-0.111276</td>\n",
       "      <td>-0.027630</td>\n",
       "      <td>-0.157724</td>\n",
       "      <td>-0.022403</td>\n",
       "      <td>0.220317</td>\n",
       "      <td>0.105577</td>\n",
       "      <td>0.105577</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>552</th>\n",
       "      <td>WT</td>\n",
       "      <td>gattgggtgagacgttatacattg</td>\n",
       "      <td>5.0</td>\n",
       "      <td>tat</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>DWVRRYTL</td>\n",
       "      <td>True</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Y</td>\n",
       "      <td>synonymous</td>\n",
       "      <td>-0.111276</td>\n",
       "      <td>-0.027630</td>\n",
       "      <td>-0.157724</td>\n",
       "      <td>-0.022403</td>\n",
       "      <td>0.220317</td>\n",
       "      <td>0.105577</td>\n",
       "      <td>0.105577</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>553</th>\n",
       "      <td>WT</td>\n",
       "      <td>gattgggtgagacgttatacattg</td>\n",
       "      <td>6.0</td>\n",
       "      <td>aca</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>DWVRRYTL</td>\n",
       "      <td>True</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>T</td>\n",
       "      <td>synonymous</td>\n",
       "      <td>-0.111276</td>\n",
       "      <td>-0.027630</td>\n",
       "      <td>-0.157724</td>\n",
       "      <td>-0.022403</td>\n",
       "      <td>0.220317</td>\n",
       "      <td>0.105577</td>\n",
       "      <td>0.105577</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>554</th>\n",
       "      <td>WT</td>\n",
       "      <td>gattgggtgagacgttatacattg</td>\n",
       "      <td>7.0</td>\n",
       "      <td>ttg</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>DWVRRYTL</td>\n",
       "      <td>True</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>L</td>\n",
       "      <td>synonymous</td>\n",
       "      <td>-0.111276</td>\n",
       "      <td>-0.027630</td>\n",
       "      <td>-0.157724</td>\n",
       "      <td>-0.022403</td>\n",
       "      <td>0.220317</td>\n",
       "      <td>0.105577</td>\n",
       "      <td>0.105577</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>555 rows × 48 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "    seq_type                    nt_seq  aa_pos alt_codons  mutated_codon  \\\n",
       "0     single  ttttgggtgagacgttatacattg     0.0        ttt            1.0   \n",
       "1     single  ttgtgggtgagacgttatacattg     0.0        ttg            1.0   \n",
       "2     single  ctttgggtgagacgttatacattg     0.0        ctt            1.0   \n",
       "3     single  ctgtgggtgagacgttatacattg     0.0        ctg            1.0   \n",
       "4     single  atttgggtgagacgttatacattg     0.0        att            1.0   \n",
       "..       ...                       ...     ...        ...            ...   \n",
       "550       WT  gattgggtgagacgttatacattg     3.0        aga            NaN   \n",
       "551       WT  gattgggtgagacgttatacattg     4.0        cgt            NaN   \n",
       "552       WT  gattgggtgagacgttatacattg     5.0        tat            NaN   \n",
       "553       WT  gattgggtgagacgttatacattg     6.0        aca            NaN   \n",
       "554       WT  gattgggtgagacgttatacattg     7.0        ttg            NaN   \n",
       "\n",
       "     combination    aa_seq    WT   STOP STOP_readthrough  ...  b3_log10  \\\n",
       "0            NaN  FWVRRYTL  <NA>  False            False  ... -1.854191   \n",
       "1            NaN  LWVRRYTL  <NA>  False            False  ... -2.708009   \n",
       "2            NaN  LWVRRYTL  <NA>  False            False  ... -2.958482   \n",
       "3            NaN  LWVRRYTL  <NA>  False            False  ... -2.592214   \n",
       "4            NaN  IWVRRYTL  <NA>  False            False  ... -2.679478   \n",
       "..           ...       ...   ...    ...              ...  ...       ...   \n",
       "550          NaN  DWVRRYTL  True    NaN              NaN  ...       NaN   \n",
       "551          NaN  DWVRRYTL  True    NaN              NaN  ...       NaN   \n",
       "552          NaN  DWVRRYTL  True    NaN              NaN  ...       NaN   \n",
       "553          NaN  DWVRRYTL  True    NaN              NaN  ...       NaN   \n",
       "554          NaN  DWVRRYTL  True    NaN              NaN  ...       NaN   \n",
       "\n",
       "     alt_aa   mutation_type  Log2FC_1  Log2FC_2  Log2FC_3  selcoeff_1  \\\n",
       "0         F  non-synonymous -0.071871 -0.095830 -0.091700    0.017002   \n",
       "1         L  non-synonymous -0.113504 -0.183024 -0.147477   -0.024631   \n",
       "2         L  non-synonymous -0.057853 -0.136193 -0.068939    0.031020   \n",
       "3         L  non-synonymous -0.024083 -0.115885 -0.102083    0.064790   \n",
       "4         I  non-synonymous -0.059975 -0.127113 -0.143584    0.028898   \n",
       "..      ...             ...       ...       ...       ...         ...   \n",
       "550       R      synonymous -0.111276 -0.027630 -0.157724   -0.022403   \n",
       "551       R      synonymous -0.111276 -0.027630 -0.157724   -0.022403   \n",
       "552       Y      synonymous -0.111276 -0.027630 -0.157724   -0.022403   \n",
       "553       T      synonymous -0.111276 -0.027630 -0.157724   -0.022403   \n",
       "554       L      synonymous -0.111276 -0.027630 -0.157724   -0.022403   \n",
       "\n",
       "     selcoeff_2  selcoeff_3  median_s  \n",
       "0      0.152116    0.171600  0.152116  \n",
       "1      0.064923    0.115823  0.064923  \n",
       "2      0.111754    0.194362  0.111754  \n",
       "3      0.132062    0.161218  0.132062  \n",
       "4      0.120834    0.119717  0.119717  \n",
       "..          ...         ...       ...  \n",
       "550    0.220317    0.105577  0.105577  \n",
       "551    0.220317    0.105577  0.105577  \n",
       "552    0.220317    0.105577  0.105577  \n",
       "553    0.220317    0.105577  0.105577  \n",
       "554    0.220317    0.105577  0.105577  \n",
       "\n",
       "[555 rows x 48 columns]"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "DMSdf = pd.read_csv(DMSf, header=0, index_col=0,\n",
    "                    dtype={'WT': 'boolean'}) # Boolean type supports missing data, helps pandas interpret column content\n",
    "DMSdf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "96d535e1-d176-4667-8c30-5f5a52439d41",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['input1_abundance', 'input2_abundance', 'input3_abundance']"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "p = re.compile(r'input[1-3]_abundance')\n",
    "input_samples = [x for x in DMSdf.columns if bool(p.fullmatch(x))]\n",
    "input_samples"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "61435ba1-897f-445c-8120-1f8d472dd981",
   "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>aa_seq</th>\n",
       "      <th>input1_abundance</th>\n",
       "      <th>input2_abundance</th>\n",
       "      <th>input3_abundance</th>\n",
       "      <th>average_freq_input</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>*WVRRYTL</td>\n",
       "      <td>0.005217</td>\n",
       "      <td>0.005489</td>\n",
       "      <td>0.005114</td>\n",
       "      <td>0.005274</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>AWVRRYTL</td>\n",
       "      <td>0.000113</td>\n",
       "      <td>0.000132</td>\n",
       "      <td>0.000097</td>\n",
       "      <td>0.000114</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>CWVRRYTL</td>\n",
       "      <td>0.000907</td>\n",
       "      <td>0.000847</td>\n",
       "      <td>0.000801</td>\n",
       "      <td>0.000852</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>DAVRRYTL</td>\n",
       "      <td>0.003412</td>\n",
       "      <td>0.003809</td>\n",
       "      <td>0.003891</td>\n",
       "      <td>0.003704</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>DFVRRYTL</td>\n",
       "      <td>0.001542</td>\n",
       "      <td>0.001601</td>\n",
       "      <td>0.001713</td>\n",
       "      <td>0.001619</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>120</th>\n",
       "      <td>RWVRRYTL</td>\n",
       "      <td>0.003163</td>\n",
       "      <td>0.003418</td>\n",
       "      <td>0.003202</td>\n",
       "      <td>0.003261</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>121</th>\n",
       "      <td>SWVRRYTL</td>\n",
       "      <td>0.000034</td>\n",
       "      <td>0.000068</td>\n",
       "      <td>0.000029</td>\n",
       "      <td>0.000044</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>122</th>\n",
       "      <td>TWVRRYTL</td>\n",
       "      <td>0.004100</td>\n",
       "      <td>0.004291</td>\n",
       "      <td>0.004270</td>\n",
       "      <td>0.004220</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>123</th>\n",
       "      <td>VWVRRYTL</td>\n",
       "      <td>0.000436</td>\n",
       "      <td>0.000377</td>\n",
       "      <td>0.000374</td>\n",
       "      <td>0.000396</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>124</th>\n",
       "      <td>YWVRRYTL</td>\n",
       "      <td>0.014196</td>\n",
       "      <td>0.015682</td>\n",
       "      <td>0.014973</td>\n",
       "      <td>0.014950</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>125 rows × 5 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       aa_seq  input1_abundance  input2_abundance  input3_abundance  \\\n",
       "0    *WVRRYTL          0.005217          0.005489          0.005114   \n",
       "1    AWVRRYTL          0.000113          0.000132          0.000097   \n",
       "2    CWVRRYTL          0.000907          0.000847          0.000801   \n",
       "3    DAVRRYTL          0.003412          0.003809          0.003891   \n",
       "4    DFVRRYTL          0.001542          0.001601          0.001713   \n",
       "..        ...               ...               ...               ...   \n",
       "120  RWVRRYTL          0.003163          0.003418          0.003202   \n",
       "121  SWVRRYTL          0.000034          0.000068          0.000029   \n",
       "122  TWVRRYTL          0.004100          0.004291          0.004270   \n",
       "123  VWVRRYTL          0.000436          0.000377          0.000374   \n",
       "124  YWVRRYTL          0.014196          0.015682          0.014973   \n",
       "\n",
       "     average_freq_input  \n",
       "0              0.005274  \n",
       "1              0.000114  \n",
       "2              0.000852  \n",
       "3              0.003704  \n",
       "4              0.001619  \n",
       "..                  ...  \n",
       "120            0.003261  \n",
       "121            0.000044  \n",
       "122            0.004220  \n",
       "123            0.000396  \n",
       "124            0.014950  \n",
       "\n",
       "[125 rows x 5 columns]"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "DMSfreq = DMSdf[DMSdf.seq_type.isin(keep_seq_type[pool_type])\n",
    "               ].groupby(['aa_seq','nt_seq'])[input_samples].first().groupby('aa_seq').median().reset_index()\n",
    "DMSfreq['average_freq_input'] = DMSfreq[input_samples].mean(axis=1)\n",
    "DMSfreq"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "36aafeef-64f6-4948-a200-35186ac20f69",
   "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>aa_seq</th>\n",
       "      <th>Nham_aa</th>\n",
       "      <th>confidence_score</th>\n",
       "      <th>input1</th>\n",
       "      <th>average_freq_pool</th>\n",
       "      <th>input1_abundance</th>\n",
       "      <th>input2_abundance</th>\n",
       "      <th>input3_abundance</th>\n",
       "      <th>average_freq_input</th>\n",
       "      <th>Location</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>*WVRRYTL</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.010655</td>\n",
       "      <td>0.010655</td>\n",
       "      <td>0.005217</td>\n",
       "      <td>0.005489</td>\n",
       "      <td>0.005114</td>\n",
       "      <td>0.005274</td>\n",
       "      <td>both</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>AWVRRYTL</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.002830</td>\n",
       "      <td>0.002830</td>\n",
       "      <td>0.000113</td>\n",
       "      <td>0.000132</td>\n",
       "      <td>0.000097</td>\n",
       "      <td>0.000114</td>\n",
       "      <td>both</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>CWVRRYTL</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.005969</td>\n",
       "      <td>0.005969</td>\n",
       "      <td>0.000907</td>\n",
       "      <td>0.000847</td>\n",
       "      <td>0.000801</td>\n",
       "      <td>0.000852</td>\n",
       "      <td>both</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>D*VRRYTL</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.011804</td>\n",
       "      <td>0.011804</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>left_only</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>DAVRRYTL</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.004421</td>\n",
       "      <td>0.004421</td>\n",
       "      <td>0.003412</td>\n",
       "      <td>0.003809</td>\n",
       "      <td>0.003891</td>\n",
       "      <td>0.003704</td>\n",
       "      <td>both</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>156</th>\n",
       "      <td>SWVRRYTL</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.007516</td>\n",
       "      <td>0.007516</td>\n",
       "      <td>0.000034</td>\n",
       "      <td>0.000068</td>\n",
       "      <td>0.000029</td>\n",
       "      <td>0.000044</td>\n",
       "      <td>both</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>157</th>\n",
       "      <td>TWVRRYTL</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.006521</td>\n",
       "      <td>0.006521</td>\n",
       "      <td>0.004100</td>\n",
       "      <td>0.004291</td>\n",
       "      <td>0.004270</td>\n",
       "      <td>0.004220</td>\n",
       "      <td>both</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>158</th>\n",
       "      <td>VWVRRYTL</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.004974</td>\n",
       "      <td>0.004974</td>\n",
       "      <td>0.000436</td>\n",
       "      <td>0.000377</td>\n",
       "      <td>0.000374</td>\n",
       "      <td>0.000396</td>\n",
       "      <td>both</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>159</th>\n",
       "      <td>WWVRRYTL</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.005571</td>\n",
       "      <td>0.005571</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>left_only</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>160</th>\n",
       "      <td>YWVRRYTL</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.008356</td>\n",
       "      <td>0.008356</td>\n",
       "      <td>0.014196</td>\n",
       "      <td>0.015682</td>\n",
       "      <td>0.014973</td>\n",
       "      <td>0.014950</td>\n",
       "      <td>both</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>161 rows × 10 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       aa_seq  Nham_aa  confidence_score    input1  average_freq_pool  \\\n",
       "0    *WVRRYTL      1.0                 1  0.010655           0.010655   \n",
       "1    AWVRRYTL      1.0                 1  0.002830           0.002830   \n",
       "2    CWVRRYTL      1.0                 1  0.005969           0.005969   \n",
       "3    D*VRRYTL      1.0                 1  0.011804           0.011804   \n",
       "4    DAVRRYTL      1.0                 1  0.004421           0.004421   \n",
       "..        ...      ...               ...       ...                ...   \n",
       "156  SWVRRYTL      1.0                 1  0.007516           0.007516   \n",
       "157  TWVRRYTL      1.0                 1  0.006521           0.006521   \n",
       "158  VWVRRYTL      1.0                 1  0.004974           0.004974   \n",
       "159  WWVRRYTL      1.0                 1  0.005571           0.005571   \n",
       "160  YWVRRYTL      1.0                 1  0.008356           0.008356   \n",
       "\n",
       "     input1_abundance  input2_abundance  input3_abundance  average_freq_input  \\\n",
       "0            0.005217          0.005489          0.005114            0.005274   \n",
       "1            0.000113          0.000132          0.000097            0.000114   \n",
       "2            0.000907          0.000847          0.000801            0.000852   \n",
       "3                 NaN               NaN               NaN                 NaN   \n",
       "4            0.003412          0.003809          0.003891            0.003704   \n",
       "..                ...               ...               ...                 ...   \n",
       "156          0.000034          0.000068          0.000029            0.000044   \n",
       "157          0.004100          0.004291          0.004270            0.004220   \n",
       "158          0.000436          0.000377          0.000374            0.000396   \n",
       "159               NaN               NaN               NaN                 NaN   \n",
       "160          0.014196          0.015682          0.014973            0.014950   \n",
       "\n",
       "      Location  \n",
       "0         both  \n",
       "1         both  \n",
       "2         both  \n",
       "3    left_only  \n",
       "4         both  \n",
       "..         ...  \n",
       "156       both  \n",
       "157       both  \n",
       "158       both  \n",
       "159  left_only  \n",
       "160       both  \n",
       "\n",
       "[161 rows x 10 columns]"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "corrdf = pool_df.merge(right=DMSfreq, on='aa_seq', how='outer', indicator='Location')\n",
    "corrdf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "020bbde6-26ea-45c6-8782-a6a61026b99b",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/tmp/ipykernel_26411/2413502137.py:1: FutureWarning: The default of observed=False is deprecated and will be changed to True in a future version of pandas. Pass observed=False to retain current behavior or observed=True to adopt the future default and silence this warning.\n",
      "  ratios = corrdf.groupby('Location')[['aa_seq']].nunique().reset_index()\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>Location</th>\n",
       "      <th>aa_seq</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>pool_only</td>\n",
       "      <td>36</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>input_only</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>both</td>\n",
       "      <td>125</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     Location  aa_seq\n",
       "0   pool_only      36\n",
       "1  input_only       0\n",
       "2        both     125"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ratios = corrdf.groupby('Location')[['aa_seq']].nunique().reset_index()\n",
    "ratios.Location = ratios.Location.replace({'left_only':'pool_only',\n",
    "                                           'right_only':'input_only',\n",
    "                                          })\n",
    "ratios"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "7f85e1e0-4b29-4004-ba91-b25c99e18f1e",
   "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>aa_seq</th>\n",
       "      <th>Nham_aa</th>\n",
       "      <th>confidence_score</th>\n",
       "      <th>input1</th>\n",
       "      <th>average_freq_pool</th>\n",
       "      <th>input1_abundance</th>\n",
       "      <th>input2_abundance</th>\n",
       "      <th>input3_abundance</th>\n",
       "      <th>average_freq_input</th>\n",
       "      <th>Location</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "Empty DataFrame\n",
       "Columns: [aa_seq, Nham_aa, confidence_score, input1, average_freq_pool, input1_abundance, input2_abundance, input3_abundance, average_freq_input, Location]\n",
       "Index: []"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Mutants sequenced in the input but not in the initial pool (not enough coverage)\n",
    "corrdf[corrdf.Location == 'right_only']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "eb50a523-d780-4f0b-a53e-1e5c07e5196f",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/tmp/ipykernel_26411/285900106.py:3: UserWarning: The palette list has more values (3) than needed (1), which may not be intended.\n",
      "  fig = sns.relplot(corrdf, x='average_freq_pool', y='average_freq_input',\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<seaborn.axisgrid.FacetGrid at 0x7f4099905250>"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 642.597x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "cscore_colors = ['green','orange','red']\n",
    "\n",
    "fig = sns.relplot(corrdf, x='average_freq_pool', y='average_freq_input',\n",
    "                  hue='confidence_score', palette=cscore_colors, alpha=.5\n",
    "                 )\n",
    "sns.kdeplot(corrdf,x='average_freq_pool', y='average_freq_input',\n",
    "            levels=5, log_scale=True)\n",
    "fig.set(xscale='log', yscale='log',\n",
    "        xlabel='Average allele frequency in initial pool',\n",
    "        ylabel='Average allele frequency in DMS input'\n",
    "       )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "414308d4-82b1-4c12-be4e-4a52b5511d54",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/tmp/ipykernel_26411/3711620569.py:20: UserWarning: The palette list has more values (3) than needed (2), which may not be intended.\n",
      "  sns.kdeplot(corrdf,x='average_freq_pool', y='average_freq_input',\n",
      "/tmp/ipykernel_26411/3711620569.py:20: UserWarning: KDE cannot be estimated (0 variance or perfect covariance). Pass `warn_singular=False` to disable this warning.\n",
      "  sns.kdeplot(corrdf,x='average_freq_pool', y='average_freq_input',\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAMQAAAC9CAYAAADsi9JiAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjguMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/SrBM8AAAACXBIWXMAAA9hAAAPYQGoP6dpAABEl0lEQVR4nO2deZgU1bm431NVvXfPPsOArIIKAhqIIiYaDIoKIoqIiTdiIMaoiUZzE41JvPGKGqPm+rtGExOTm3iTm4hREAiKipC4EAVUVFRE2ZeBGWafXms7vz+qu+lZmQZmBrXe5/GRrq6uOt1T3/nO+VYhpZS4uLgAoPT1AFxcjiZcgXBxycEVCBeXHFyBcHHJwRUIF5ccXIFwccnBFQgXlxy0vh7A0cj69euRUuLxePp6KC5HAMMwEEIwbty4g57raogOkFLi+iuPLFJKdF3vk981n7+nqyE6IKMZxo4d28cj+fQQj8fZuHEjI0aMIBgM9uq9N2zY0O1zXQ3h4pKDKxAuLjm4AuHikoO7h3DpFZ588kn+8pe/4PP5UBSFqqoqTj75ZPbv3589p7GxEYCXXnqp1Wd1Xee+++5j3bp1mKbJj370I8444wy2bt3K9773PWzb5o477mD8+PHYts3cuXP59a9/TTgcznucrkC49AqzZ89mzJgxjBo1iurqaq666iruvvtu+vXrB0AqlWL27NncfPPN7T77+9//noaGBhYvXszmzZuZN28eL7/8MgsWLOCaa65h4MCBPProo4wfP55FixZx7rnnHpIwgLtkcukD5s+fz3e/+92sMAD84Q9/YPTo0Zx55pntzl++fDlXX301QgiOO+44HnvssayfSNd1kskkHo+HZDLJ0qVL+cpXvnLIY3MFwqVXeeutt6itreWiiy7KHovFYvzf//0fN910U4ef2bFjB2+++SazZs3isssuo7a2FlVV+drXvsbChQt54IEHuO6663jssce44oorDsuh6i6ZXHqVJ598knnz5iGEyB5btmwZZ555ZiuNkYtlWezevZunnnqKTZs28c1vfpPly5czYMAA/vznPwNQX1/Pm2++yZVXXskPf/hDmpqauPzyy5k0aVJe43M1hEuvYZoma9as4dxzz211/MUXX2T69Omdfq6srIypU6cihGDkyJFUVlaybdu2Vuf8+te/5tvf/jZLly5lyJAhPPjgg9x77715j9EVCJdeY9euXQwaNKjVhldKybvvvsspp5zS6ee+/OUvs3z5cgB2797N3r17GTZsWPb9nTt30tDQwLhx49B1HU3TEEKQTCbzHqMrEC69RnV1NZWVla2ONTQ0oGkafr+/1fHHH3+cBx98EIAf/OAH1NXVccEFF3DNNdcwf/58IpFI9txf/vKXfPe73wVg+vTprFq1ipkzZ3LNNdfkPUbxWai6sWvXLm688UYWLVrUrfMzsS9uLNORIxPLNGrUqD6LZerO3/NTryGam5tZsGABoVCor4fi8gngU2VleuKJJ1i2bFn29QMPPEB5eTk333zzIanPzyKWLWlIWbQYNilLYgOqAJ8qCGkKEa+CX/30zqOfKoH4yle+clhOmc8iUkqq4iYfN+nsaDGoTpjYB1lE+1RBoVeh2KdS6lMp9auU+jXK/CqaIrr+8FHOp0ogjiRSSuLxeF8Po8doMWzebzL5sNmi2WgtAR4FCjSBTxUoAiwbkrYkbkpSNqQsSU3CoiZhtfqcApT6BBV+hf4BlQEBhUKPQAhBIpEAyP6/N5FStvJ7dMUnYlMdjUa5/PLLeeSRRxg4cCDguPMffvhhDMNgxowZXH/99Ufsfhs2bEDX9SN2vaMFCTTgY5cooIYApB8SVdqUkaBcJigkRQCTzh4fE0ESjQQacTRiwkMMD1E8mEJtd75XmpSQolCmKCJFGL1PNq5er7dbm+qjXkOsX7+en/70p60cMfv37+e+++5j4cKFRCIRrr76al555ZUO42AOFY/Hw4gRI47Y9foS05Z81GLxdoNBberA/HdMQGFMkcaxYRWPcmjBcBmklLSYkpqkTXXSZk/cpiZpowuNfWjsE45RQwGKvM6SK6wJQqrAr0JQExR4FEq84ogvuzZv3tztc/MSiK9+9avMmjWLadOm9ZrV5m9/+xu33347t9xyS/bY6tWrmThxIiUlJQBcfPHFPPvss0dUIIQQvW4ePNIkTZu3apO8tT9F1LQB0ASMLfXz+TI/ZYHWf37TliRNG1uCIsCvKXk9nCGgsvDAa8OWVMUMdkVNdjYnqYqZmEKhXpfU61aH11AFDAp7GFfm5/hCb7eXOl2RzzXyEojbbruNp59+ml//+tecdtppzJo1i1NPPTXvAebDPffc0+5YTU0NFRUV2dcVFRVUV1f36Dg+STTrFutqErxdl8Rw5ICIR2F8mZ9xZX78WvtFS9K0adJtctfPMdOi0Kt0eH538CiCIREvQyJe4oXwwcaNDBx+AnHhpTFtyYqbNnFTEjNsGnSLhCnZ3mKwvcVgYEhjxtAIBd72S7GeIi+BGDNmDGPGjOHWW2/l+eef5wc/+AFer5fZs2czZ84cAoFAT42zFbZttzt2JGaSTzo1CZM11Qk2NqTI/ELlfpXT+gUYVeRD7WS2N23ZThjA2XM06TaacmSWMQIo8ChUBr0dvp8wLHbFTDY36bzfkGJ3zOSvHzdx+YhCCn29IxR5CYSUktWrV7N48WLWrl3LlClTmDZtGq+++irXXHMNf/rTn3pqnK2orKxk7dq12dc1NTXtQgI+K0gp2dZisK4mwbYWI3t8cNjDxH4BhkU8B50skmZ7YcheP/1+uIdnadOWNBuSAq/K+PIAxxZ4eX5XlEbd5sU9MS4aGukVk25eAjFp0iRGjBjBrFmz+NnPfobX60j6uHHjmDVrVo8MsCNOP/10fvnLX1JbW0thYSFLly7l8ssv77X7Hw2kLJsN9Sne2p+kPuWsxwVwQpGX0yoC9A91Pycg1+8gpSRhSVp0mxbDJmbYGLbEkhLdlpi2IySKAK8iCGiCAq/jj6gIaBT7lEPS1m2FssinctaAEMt3Rfm4SWdvzGBQpGPNciTJSyAef/xxbNtm0KBBxONxPv74Y0aPHo2iKDz99NM9NcZ29OvXj5tvvpl58+ah6zqTJ09mypQpvXb/vmR/wuSt2iTv1R/YH/gUwdhSH6eUByjKc2mRNG12tOjsjJnUJi3qUxYp69At8UFNMCzi5fgiL8MLvN2e1TtyBvYLagwKaeyKmbzfkDr6BOKZZ55h5cqVPPHEEzQ1NfHjH/+YmTNnMnfu3B4a3gFWrVrV6vXUqVOZOnVqj9/3aEBKydZmg3X7E2zPWRaV+lXGl/kZU+LDl0c4xf6EyUdNOpubdPbGzXbvCyDkccyiXtXRAAHV+bcqBCFNoCoC3XKcdU26RW3SoiZhEjcl7zekeL8hhV8VnFzq5/Pl/oM+aJ3JzbEFXnbFTHZGjY5POMLkJRBLly7NRoz279+fJ598klmzZuUlEMlksl2or0vH2FKyqVHnX/vi7E8eWBYdV+hlfLmfIeGD7w8ypCyb9+tTvF2XbOdhLvGp9A9qFHoVSvwqRV4VVUBjyiLoUdoJm8ARxrazv2VLdscMtjQbbGxI0WLYrKlJsG5/gtGFGsVduOT8mkLMtNrtZSqDziNan3KWbp4e3kfkJRCZorEZDvbHuPTSS3nqqaeyr23b5tJLL20VgOfSHltKPmzQWb0vTl16f+BVBCeX+jilIkBhHhvcpGWzribBG/uT2aWQImBYxMPxRT6OjXiIpK+X8UMYtqQ2YdJk2FQnLOKWTcqU6b2Ec12/Kgh5FPyqoMirUhlUs3uJ4jKVU8p8VMUt3qxNsCtqsqHRRBXHIBoMJgbah1JoiuOsa2vtCmrOPZKWE3RYEehZX3JeVz/vvPO48sormTZtGkIIXnjhhXbpgABXXnll1go0cuTIAzfTNM4555zDHPKnFyklHzY6glCb1gh+VXBKeYBTyjv2H3SGLSXra5O8sjdOMv0UF/sUxpcFGFPiI5C+lpSS6rjJ7phBVcykOmFSl2w/U3eHAo9gcMTLyCIfYY8T/DdrWAHVCZOVu6PUJOGlGoNt8WbOOSaIX1VaOQAz/2/rHIx4FJKWRcywoYct+3nHMq1YsYI1a9agqioTJkzg7LPP7vTc+fPn89Of/vSwB9nb9HaCkJ0WhH/lCIJPFUyocAQhn/0BOP6IZ3dE2Zdw9gclPpUJ5X6GRTyoihO0tydm8kFjii1NOjGz/SPgUSDsUQh7FEKagk8VeBWBIgS2dLRFzLCJmZJm3SKacw0BDIk43uYirxMNm0rEWfHhbjarpZgSQppgysAwRT4VAV06AP/vo0Z2x0wuHhphZLEvr98C8vt75q1/hg4dSnFxcba8+Lp16zr1Vo8dO5bFixe3O37xxRfne9tPJaYteb8+xes1cRpSjsnIpwpOLQ9wSoW/w7yDrsIrpJSsrUnw0t44tnSu9YV+AQaHPVgSqhMmW5oNNje3FgKPAgNDHo4JeagMalQEnE10fapj/0Q8ZaJLmQ0BNC3J3rhBg26ncykcb/OOFoOxpT6+WBHAIwRDtSTDK728st+k2bB5fleUaYPDRLxqlw7AzL7BOFhc+hEgL4G48847WbFiBUOGDMkeE0J06pDLdZ7pus66deuYOHHiZ14gDFvydm2SNTUJomnbqV8VnFoR4POdhFZA1+EVEli2o4UtzY41ZkShlynHhEhakoRpsalJZ2ODnl0+eQQcW+BhZJGPYwu8+Dq4Z0iT7E+YWGnh86oKHkXg0wTxtNFHIGk2LIQQlPhUSnwqGrAzblKTsHi3LsWeqMG5/b1Ij5+IR2Ha4DDP747SkLJZuSfG9CGR7FKpIwdgRkbaxyccefISiJdffpnnnnuu20FvbeOQ6urqWgXpfdawpeSduiSv7o1nZ+iIR+HUigCfK/XjVTs3UnQVXrGjxeAfVTEadRtNwDkDw5xc6iNm2LQYFq/sjVOdOLAnGRzS6BdUCWkqEa9Ko25TCK0EMWk6yyG/ppCwbAxLIrEp8qjEpEjfGSx5wIdgS8fSFPIpnFjkpcRrsaVFpy5ls3hXkpGKh5ANBZrClIFhlm5voVG3eacuyefLAwdNTOoN8hKIAQMGZL3Th0JJSQm7d+8+5M9/kqlLmizbEc3a/Qu9Cqf3CzKmxNct51Vn4RX74iYr90QxbOealwwroF/aVLk/YfHszihxUyKAwWGNEp+CRBA3JUFNols2qlCoSZj4VEFtwmJv3GRP3KDFsEmYByxLkPZRaI6FqdinUuZXUIWjKZKWdAyrEpp1G78mmFjh5916nSbd5n0rTEHSIuiXBDWFL/QLsqoqxvsNKWcjrnVsPROdZmccefISiMGDB3P55Zdz+umntxKMzpJzfvSjH7V6vWnTJk488cRDGOYnm63NOou3taDbEp8qOLMyyLgyf6fBdh3R0ey5O2qwqiqGLaEyoHLZ8EK8qiCqW+yLmyzd0YJuQ0AVjCzyEvQIYsaBmT1m2IRUwcYWgx1Rg/0J66DLEglETUnUtLJaJ6wJSr0KYa+zpJKCbC52gUdlysAQL+yK0mwovN1oUhoy0RSFYq+g3K+yP2mxsTHFkEjHORnykGxeh0ZeAtGvX79Oyw12xIQJE7L/FkJw/vnn84UvfCGfW37i2Rk1WLi1GUvCoLDGjCGRrN0/H9rKTlXsgDAMCmtcMCiMIqAuadGkWzyzM4puQ5FXYUSBB7+mYOcYFDUBdSmLt2pTxHNUQEAVFPscy1KZX8WjCCxbgoCQpmLjxDglTcmumEl9ys4KSD9bcnyhBzt9Ha8KquJojS+Ve1i+J0mTofBBo86IQsdaNCTiYX/SYmuzThcrxl4jL4G4/vrrqa+vZ+vWrYwbN46mpqZskk5HzJw5k5qaGt5++20UReGkk076THX2TJo2S7Y5wjAs4uH8gaGs/b8tB0vOyfXkNqQsVu1xhGFw2MOXBwTxa45Ty7Alq/bESFmSUr/KGf387Ik5hQMyvjCBZHOzSaPu6ANVQJlfZVBQI6gJfB6FhCkp8CjEDQtLASnBsG0CmkKRV0HzCQaFPDTqNpuaUuxPZ8pJTPoHFVTheLibdYnH55hwh4goWylga7PBmBI/anpj/05dkrgp2R0zGRRu/3xk5Lg35CUvA/c///lPvvrVr/LDH/6Q+vp6pk6dysqVKzs9f/HixVx00UUsW7aMp59+mgsvvLBdTNKnmdX7nM1zgVfhi5VBErYzgyfN1guTpGlTl3Rs+XFLEjVlu/MynlzTlvxjTwxTQmVA46z+QYp9KqbtLCzerk3QqNv4VcEX+/nTyxYFTUgCqsCnwqYmIysMlQGVE4u89A9qSCFJ2ZKmlIVEkrJtWkxJzJA0604CT2PKxLBt4qaFISVF6XiqEQXO3FqTtIibkrBHYKQjZJt0Z99UrtkEVNBt2Nas41EUgppKZdr7vKO54zz2jP7qDYHIS0M89NBDPPHEE8ybN4/y8nKeeOIJbrjhhk6dc4888giLFy/OLrP27NnDtddey+TJkw9/5Ec5ejo8G+DzZf4DvgJaJ90cLDkHHO2R0Rzv1ydpNmxCmuCCwSEiXkcYWnSLupTF+w3OQzWqyEtDysarOOt5nyZQpGBzk0HclGgChhd4KPA62sC0JU2GTUATWNLRDroBhg2GdITNNiUhVWVvzCDiUUlJQQRImTZFXpUyv01t0mZri0lQE1hSOJpFE0gbFJ+fgYrKxy0WO6MmA0IWYY9Cqd+JaM04JduS+W16o8JN3o654uLi7L+HDh3aZTxTJBKhvLw8+3rAgAGfmSVTVcwgaUk8Cu2WAblJN10l5yQtm1TCxpt20EUNm7frHCG7IL0XyQhTzLB5py6FBEp9KmGP45tI2eBTQBOC3XGLupSNIuDkUh8VAZX6pIUtbaKGxLJJl52xsTWnBE1IE3gVBVNmli7OfqJZt/HbAsuUeNMlayr9CnHD0XJ7YhaDwxoeVckut4SqUeYRfAzUpywMy6ZZSkKa8ww1Gx1v6TMGBaUXsiLzNru++OKLgFMi8i9/+QvHHHNMp+cfd9xxXHXVVcycORNN03juueeoqKjIeq8/zQ66jOe5yKt2+IfMtd13hGlLooZNUDvw2Y+anAe+f1BjQFBrrVkE2YjYUp9jCs04o1M2FHokm5sc7TG6yMPQsEbckvhUiBugKc7eQhWChA2WdDTF/pRJwpJIBEjwqc79bQkJS5K0JWWaiqpAeUAjasL2qEnSknhVZx8kpSSm29RHUwwoDSBwrFx1KQufqpCynN8q2UkeRsYY0BtZwnntIe644w6WL1/Ovn37OOecc9i4cSN33HFHl5+prKzktdde45VXXiEUClFcXMyaNWtYs2bNYQ38aCfzp+3MtJo53NkyQLfsdGbagRP2RJ21+IgCLy26s8GWUpKybFp0O/tAhT1KjsA5PoiqhE0irbEKvQpxU1IVNdjcZNKs2yAdq1BABa8CqmLTrFs06o5FKWbamNK5T3XSJKhCuV+hOF1Optjr+DcCaVNRypLolsS2JXHdwpISFUjZznINHI0nHFEDnEmgIzKCrR1tGqKkpIT/+q//6vb5HVXM+KxQ6HXmmngHywDBAa9wZ3kAtgQVJ+w7Q7PhaIACr3BCPoSzfjcR2bIuPkVQ5FUQiiNMKVOiCNiT9hmENAULwZ6YgVco2NIiZoNiW5T4VKoSlvNw25AwnYdRE2BZgCIQiiSq27R4nQC/pC2xbEmxT6PUp9GSHqPEsV7ZSDRVwbBsFFVJL8sESIkihBMOImT2Mx1hpQVF6wUNkZdATJ48ucM9Q2eWptWrV/PAAw/Q1NREblBtV5apTwuZnObmdKmVYFoAMpGdmU22pjgZaG1jhlQBfm/r/OTMBBo3JYq0aTZshICIR0UVaWUvwJBQqAoMG1QhiBoWevrDHsVZFiiKQsqyKPMrpOwDyxUpIWmDZjvZcZm7C+F8TlPAxrm2lBLdtGkC/KqNRxVZE6lHcZZclgXNhsSyJBGPB1UBM32STxWO0CjO2DvzQ2SC+o66IgOZfl7gJAstX74cVe3cyTR//nxuvvlmRo4c+ZkrExPUFPoHNfbGTaqiBieV+jv0L+TGDOm2xJbOf2U+lbh9YNaU0jGbRtM1jIq9GrbhPMAthkU6WgPDkggp8akKdQkLG4khRTb8wpKg2zamZWPaENAUDGnjEY7AmbZzzczsHtKc5ZdAUuRVqU2aRE1J2ONUB1cEKIqgJmXiUQX70poo4lGw00EXipBIAQjHodEqyUhTaEhb0wKdqICMMHcV63WkyEsg2m6gr7vuOi655BK+9a1vdXh+UVHRZzoh6KRSH3vjJu/Wp5hYGWw3w+WaXIVwLDUZ4rZj4YmZjslTtx1HWzRqUp2wGBz2ENQU4qaNKSGiObNtRstYlqDMr5KwbaQu8aev3aLbNCctalI2HiEIWpKoaRNJ51BbPif0O6xByiNo1iWKIijQBHHLRpcQ9ggUaaMKAUIhZUvCmkJLymJv3BGIAUGVYNpShRDYQqDoBknL0ZxBTVDs1Qh5VHbFMvFd7SdXKWVWe/mONoGoqqrK/ltKyYcffkgsFuv0/C9+8Ys89thjTJo0CZ/vQGLHgAEDDmGonzzGlPhZvS9Bs2Hzxv4EE/u1jhI+WD0kcHKXk6aNtJ1c6h1RJ+F+VJEHj6oQFiqGbeMVCv0CKlVxi5qkTf+wgioEVtLEtE3CHgUFZzlVrzumVxvwqjBCpAglGojVNHNcURFKpJgmEaJ/UMGDRcyWeBVBynZMpMVeBVMKSK/9TdvZ73zUYiJxMvPCaf+IDZT6NGLYtLQkaRFOylu5X6XA52jLprRFrtTfXiAM+8BSMdALfSnyEogrrrgi+28hBCUlJfzHf/xHp+dnzKu5+RJCiM/EHgKcdfSk/kGe2Rnl1b1xTijyUZxTJuZg4c62dNbNmRyBgYrCoJDOrpjJW3UpJpQ7AYKqohD0KIwo9FEVj7OtxeDz6ff8qqMxZDquqV63qdcl/QMKmoDjZBPfv/4aVrzwQva+5557Lg/95lGagqVUhjVng21LCj0KSduiKeX4GgIeZ0kkpGR33AkFUQUMjzhaIKMR46bj8DM9HjYlnIf/pBIfEY+CZUvqU46GKO2ghE4spybtUbdkyjfs4rMUptEZY0p8vFefYkfUYOn2Fq44vtBZanBwz2vu+xlr1ISKAHu3t1CTcMo+Dg57UHCsUYPDHvyqIGFJtrWYjCj0ogiFioCHmoROmV+hUbexJNTrki9EdL5/TWthAHjhhRe44dpvcf/v/oTlC6PjaAjdlli2QFXAJ8G0QFMkDbpNSzqK9tRyH37VceTZUpI0HS3hUQV7khIDx6hwbIGXZsNZDtWnMnsIhaRpt8rLiKWtdCFP7xTRz0sg2oZztyVjZn3ooYe44YYbOj2/t8yxH374Ib/73e8IBAJMnTqVL37xi71y31yEEEwbEuYPHzayN27yzz0xzh7ohDl3ZnKF1qZZOBDLBDChIsC/qhO8W5eif0BjUFjDozgJ+6dWBHhlb5y365IMi3jSm14o9WsENCd4cEuLUz/ptQbBR9WNHY77hRde4N7mOsyyMF7NubYhwSMUgtLCtKFRt9kbt0g5FmDGlfkYUejHkja1cZMmnQNpodJwelIAo4p8TliKEOyNO8usAq9CIB2gmJtK2pIWiMjRKBChUIht27ZlPc/PPvsspmm2q5o3evRooHX4d18Qj8f54Q9/iKZp3H///X0iEOBsFi8YHGbRthbW7U8yIORhVLGv09IrbU2zGTIWqnGlPvYnLD5u1nm1OsE3iouyZt1TywOsr00SNWw+aEgxqthHwrJQhIJXkVT4VXTphFbEpMa1f3yGbetf593nFrF5zUvU7tyavV9zczNqqWNu9SgKpT6F/UmDppRkT9wJ4gPHWjS+1MeAkOY8zKgU+yUeTWbDPd6rszFxOgoNDKmk0rkhO9OF1walzdS5YS3g1IbK/Ia9QV4C8c4777BgwYKsqfXcc89l9uzZzJw5s9V5meC9tsd7ko4aLo4fP55t27bxox/9iCuvvLLXxtIRxxf5OK3CZE1Ngmd3tlDqd2qhdlZ6pTObu6YIIj6N6UPD/O+mJupTFs/saOHSYwsQwqmud9aAIMt2RHmnLsmoIi/9Aipx0wm/aExapKTEIyEZj7JT1xg2biLDxk0EIBltoX73dlpqq9kaGojVmEIIJwgxZspWZS4VYFhEY2yp37EyGTamLdEUgS0zFibY2WKyK+7Yc08t9QIHKndkKvINiRyI98rdW2Xq1hYfjdW/W1paaGxspLS0FHAacZtm+1KIfUFHDRc3bNjAsccey4IFC/jGN77BtGnT+mh0DpMGBKlOmGxvcZKG5p5QRCD98OdbXdunKlw8LML/bmpkS7PBO3UpPlfmVEQcXezsW7a3GKyqivO14wspkk7fCFtCyCswQyoi4eNvN32X5lA5J5wxhUFjxuMPRxgwciwwln02jpeuDU6ohkp5QKVfUCPsUdAUhYgAw7LRFDXbm64qpvNuvRNDNVjEKPMFsIXjrd7SrGNKJzy9PMfClDsXZOKzyjqwQPUEeQnEtddey4wZM/j85z+Pbdu8//773HnnnT01tsMmmUzyk5/8hJKSEiZNmpTXZ3uq6eK5/TSeSJo06TYLtzRy8UBfp1GcjhPNcWSpwtk45xpawsDpZR5e3W+wck+UAR7HvArw5XKVv8QMquImL2xv4owKLy0pG4HzR9cAEQjy3/fdy7ev/Ra/u/phVM1D6aBhnD39Yr7+7RuoEwF8qlM5T0ViSicTLmFJhJBoWDTELQxDoczn3DeoCTTLRLMl25tMNjY7E2Z/HxSmUkRTKUI+J5T8vTpHUIaGFFIpJ4pXACGfQtx0NuX707WlIsIgHu84PPxg9GjTxerqatavX4/f7+fkk09uFQ7eFQ0NDRQVFeXtse7thovQ800Xo3hYKyqxhMJg2cwJsqHdOYrPT4vhhGJnUBWFiAfsVDJ7TALrRCVNwkeljDFW1mbfq1XDrJeONj/Ok6DCirbT6KqiUBzyYyXjNDU3EykoAH+YZkshqEpSugFCUBAIUJM0MVGImxZYNradrjerCPp5BWYyToFPxTQt3pXF1OForGKZYICMIgCvplEZ1NiSEGyywijS5gRZj08RzsZa6tnv14yXNUp/NGlzltx1WAlCPdJ0MR6P89hjj7F9+3Z+/vOf88gjj3DjjTe26xxUX1/P7bffzhVXXMGpp57Kt7/9bV5//XVKSkr4zW9+w/HHH9+t+/VVw0Xo+aaLxS0mz1bp7BQFjD6mjOMiB/4UlnSK+3Y01QigJB3enaEsabNgR5J9IsRZQ0upSMcnlek2dpPJO40mHxsBjFCIU0o8He5PVAVCljMrK0JQ0uYUG7DS5tEW06btNFrkd+ox7U5YvNtgYqTHenKRxrGRIlJ6iMb6OvqVl2ArHnYnDUByQqGH4wqOQcHxRJflfLe36g3YbzAwrHHiwFF5/b659FjTxdtvv53hw4eze/dufD4fsViMW2+9lQcffLDVefPnz2fs2LGMGTOG5cuX89FHH/Hqq6+yY8cO7rnnHv74xz9263591XARer7p4klBqDNjrKlJsHKfweCiUHbjGNUtfKJzxS00QTBnzzE0CKObnTL0bzbaXDQ0REvSwueHcX5QtCTra5Nsj9nU6wbjyvwMDntahaYLKfF4O5+DU6aFN20LK/JI4oaNLp24qmbDZkvUpDZ1wIRc5lc5vV+AUr/ziCUVQW0qidB8fNzixG8FVMH4fuFWFb1zv9ueqiYAhhX6D+tv0WNNFzdv3sz999/Pc889h9/v58477+TCCy9sd96WLVv47//+bwBeeeUVzj//fMLhMKNHj6ampqbb9/u0N1ycNCDInpjB7pjJ39NOO6d2atef6+j9L1QGeL8hle2248kJczg5bQV6vSZOs2Hz0t44moCygEaBRyGoCYKqwEIgxIHcZSnBkjJdgMymyXCsTClLEjed/Iq29A9qnFjk5ZiQhqK09x006hYbGx3L0ikVgXbl7TPfLWnZ7IgeqELYW+QlEEIIdF3PSlxdXV2n52V4/fXXueuuu7KvD7eT/aep4aIiBBcOjfCHDxupijsNE0+vDOblwc5Q6tcYXuBhS7PBe/UpxpW3XsYOL/RSHlDYUJ9id9TJgtsXN9l3mN/BpwjK/Ar9gx6GpEvrO0GLFmGPxE6bWE0JiqrxZp1TAeSYkMaxkfbpxJnv9lGjji2dcI6MlukN8rrT3LlzmTt3Lvv37+fuu+/mxRdf7HBDO2DAAJ599llisRiJRCLroFu2bBnHHXfcYQ3409ZwsdCrcs4xIZ7ZGWX1vjijip1S8t31YOcyrizAlmaDTU06xxd50BQFryKyE0aBV+OUcsHoQpuo5UTaJtOZbUI4iTgpu/W9FOGEZfvTORpSOmEYAdUZe4HHCdPI1UiaIvAosDdmEkpn0qVSFh/JAhoNxyH3xcpgu4ks97u9W+dsrEeX5F/t+3DISyBOO+00Ro8ezeuvv45lWTzyyCOt+j9kuP322/npT39KbW0tv/jFL/B6vfz85z9n5cqV/O53vzusAX8aGy6OKfGxoT7FzqjBS1UxLhpWkJcHO8OAoJptLrIjalIR0FCBiNfxE0gpiRuSiE+jSBEMbHPtTLRpV07CtvWjLClp05AIMy1YPk3Fk/Y5bEva1OE83OceEyKkKZ1+t5qEye6YU1t87NEsEF//+td57rnnGD58eJfn9e/fv92Df91113HLLbd0uK7Mh09jw0UhBGcfE+KPmxrZ2KjzxYRJWUDLy4Nt2pIWQzI04uHDRp39cZMSr0oKiZ6UlPuddNCgp+Nr5IZMdOUkbOtEbE5ZtE3+zOSDq4qjTbY0G7zf5EjNuGKNIRGnkmBn321NtbOsPqHIe0hVDg+HvARi5MiRLFy4kM997nOtTK1t8xsefvjhLq+Tr9/gs9BwsV9Q4/hCLx816bxZm+S8QeG8PNiZ3IqhES8fNursjpscX+RDpp17McOmwKsgRecT0qFU3+5IPjNZf4Yleb/eyPbPHiCjnFBQ1i6sPZe6pMkHDY6Trm3+SG+Ql0C8++67vPvuu62OdZTf8Mgjj1BcXMyUKVOyYR4uB2d8uZ+PmnQ+aEhxzjGhQyqGXOJzuv2kLEl9yspWAtdUQcp20jk70zKHkrLcNmLXtG3qUxY7oia7oyap9MBOKtLw17cAZV3e5+W9cSROZZFMw8XepFt3vOeee/jRj37Ebbfd1q2qey+//DLPPvsszz//PLt27eKCCy5gypQphMMdV3d2cXDSQkW6zqnBkDz6MmceMtOWDAhqbGsx2BU1qAyqWBJSFqg4edQ+TcnWTMrQ1WYdnLyEnVGDfXGTJt0iaTnN3JV03rSVtiRFDTsbsg1OqMfJpT6KVJsWTevyPruiBpsanQiBLw3ofe0A3RSIFStW8KUvfYl7772XgoIC2kZ7tG2pVVpaypw5c5gzZw5VVVU8++yzfPOb36SiooILL7zwE7/m7ykUIRga8fJBQ4pdUTMvgcjM1E4BZEcg9iUsquOmE5ItnGoZYY/ATPsQwumWu7kb2rab5rqUxdqaBFubjbyK0hd4FAaHNQalHYB6SkdoTunMjjSULSUrdkeBTFXB3tcO0E2BuO6663j00Uepqalp55XuqqUWOPuLuXPnMmLECH7729/yk5/8xBWILugf1PigIUV1Ir8o4kxuRdywiHhVKgIqNQmLzc0GJxZ5sdPuNikdk2hAE+lKHhDyqNkNfKaCeMKweKsulV3/A1QEVI4JeSjxqQQ0p0yNna6ioVs2zSkbFCeZJ6A5JSwzhcs0TSCFha8TJfRWrdM/26cKJvUPHdqPdwTolkDMnj2b2bNn86tf/YrvfOc73bqwZVm8+uqrLF++nNdee41TTjmFb37zm0c8xOLTRiYMuq6Twr9d4dcU+oc81CRMTi7x8uKeBPUpmx1Rk8FhDU0IPIrABBAZ30LrostJy6ZFt1i7/0CD98FhjS9VBhnYicYybadaecxnteo1IYRTU6nIp5KyBTG74+8UNWxeqXIiiyf1DxLspey4jshLL3VXGH784x/z6quv8rnPfY5p06Zxxx13tKq64dI5Rel4pibdKQacsmS3zK4Z/JpCRUDDlpLR6XzuXTEn3HxgSCNpO6UjG5IgcQqLKYrjlBPCKZpWn7KzwjChwu9cD7LJP23JWLi8qkLCau1QtHCsXAJQLaPdZwFe3B0lZUv6B7VsTkdf0SMLtUWLFlFcXMx7773He++9x/3339/q/c9K1Y1DocDrlIuxJOyKmq1my0zH0YM1cPdrTv+3E4u8KEjea9BpNmw+aDx4SLvAKSOTYVfUxJOuLthZl9CMhUtTBGGPQtRo7VCUUlLoVajtIJlsW7POh406Ajh/ULhXKnx3RY8IhPvAHzqKcPYCDbpNk261EggJXfZzziXkUUlacFJpgH4BjZ0xk/qklS365VcFflUQ8ar4VOdh39bilPBXhKAyoLIv3YBxb9ykzK8ysSLA6BKlXchF7lB86WWYbh3YmBf7VDxW++WSLSUv7nHqeo0v92dNxH1JXiOwbZvHH3+cLVu28P3vf59Fixbxta99rZ33OVPh76OPPmLr1q34/X6GDx/OoEGDjtzIP8UU+1QadMeen6kRm6FtEn5n5BYwKEvnbstCsqEcHkVppW1MW/L01ma2tBgENcGQsIfKoE1VzGR/0qI2abFsZ5R36pNMPiZE/+CBcbX1RTgC64xP4Ain3sH24d26FHVJi4DmNKI8GshLIH72s5+h6zpvvPEGUkpefvlltm7dyu23397qvPr6em688UY+/PBDhg4dCsDWrVs59dRT+cUvfuH6Iw5CRUBja4vBvoTJ6A7e765H+UABA5EtCuYRAk0V7fYjmiKy+cvhdJxR2KNwYrEXTQi2pjuV7oqa/O+mJsaX+TlrQAivKrpVPaTtYs2WkteqnY30F/oFD7oM7C3yGsUbb7zB/Pnz8Xq9hMNhfvOb33TY5+Huu+9m7NixrF69mieffJInn3yS1157jWHDhn2mS+R3l8FhZ56qipnoHeQc5ONRzoRIlPg1yoMeigIaYa/absnVmLKcauLAkLBGxKNQ4FWIeBVCHoUvVAa4ZFgBx6dzE96qTfLYpsZszrNfUyj1q4TTuRVhTVDqVzt90Lc1GzTpNgFV9PlGOpe8BEIIgW3b2TVkKpXqMFjvgw8+4JZbbmnVy9rr9XLzzTezYcOGwxzyp59BYQ+FXgVLOl2DcjmYR9m0JVHdojllEdWtTpuQtKUqXXC43K9SEfRQ5FOdAsgelSKfiqY4gnHOMSG+OqKAiEehPmXx54+a2N7izP8Z4SvwqR0KXS6b0t9rVLGvXZJQX5KXQFx44YXcdNNNNDY28thjj3HFFVd0mDHXmYlVUZTDjnb9LOBRFSakE3zerUsRT9c3PVj4d3e6mXbG3rhjEu0X1NKVyBUCmup4snM20YpwAgjnjSxiUFhDtyVPbWlmV7Rjk2pnZATw2ILey4brDnk9nd/4xjf4yle+wvnnn8/evXu54YYbuOaaa9qd11UG2yc1u623GVfup19ARbclr+2LE1LpcgmSW1o/02YrYVokLZvG1ME1RcbvMCCoQc7nU5adDdXJaCfTdlplTR0YZkjYgylh4dbmbJW97tCS7glR3Jnruo/o1qZ63bp12X97vd5WAX7r1q1rF8u0ffv2DivlSSnZsWPHoY71M4UiBNOHOIXIMtW+Jw3oPKQh4xwzbacP3IFHUxLH6RxUGui8A2xd+mEu8qkY6YaP6Z5CqDj+kVK/1q6N8Jn9g8R2RalNWixpU8y5KzJX6Gu/Q1u6JRC//OUvO32vo1im3/72t4c3KhfA6eo5dXCYv++I8lp1goqAxqjijpejtnQmnNbC4GABTSmbQl/HnuaMAIBTEM2jKhQKQdK0HE85gpRhY3mdZViurtEUwZcHhFiyvYW9cZO16bzwgxFMFzaOGnavlansDt0SiNxWWuC00+qq33RfFzn+NDG6xE91wok4fXZnC2V+lfIOIkEVka7y19mFROf+i+a0k8CjkO2JbUknDVSmZ/C4hH0JJ3LW16ZxScijMKEiwKv74vyrOu7Uej1IPFKp3+mxXR032/Xx7kvyWsBt27aNWbNmMWXKFKqrq7nkkkvYunXrwT/oclicNSDI0IgHw4Yl21uyTQhz8WtKu7D8DALnQe9sGxFL93YI5Tjp2oZfwAFN0tF+ZHiBh34BFcOGNdUHLwGaqfa9vSW/zXhPk3ef6p/85CcUFRXRr18/5s6dy49//OOeGptLGkUILhwSIawp1CYt/rGnfRszTREUeNV25R4FpIsRi079F6l0aZ9MD7dMTnRH95Dp99sihMhaxt6pS3XoP8klY13aEdU7FPC+Ii+BaG5uZvz48dnXM2bMOOw6Sy7dI+RRuGCI4+F/qzbJjpb2gXoFXpVir0pQFfgVx0FW6FWzSUCdWagyz66WbsPb7vmUTjSsKmws2yau20R1i7hxwHolgOOLvBR5nW6qG+qSXfpCKgIqhV4Fw4YtzT1XRzdf8hKIQCBAVVVV1nT6xhtvdBnWvXjxYk477TRGjRrFqFGjGDlyJKNGHXqNzs86wwq8fK7U8eo+tyvabmbVFEGJXyXkcRxjwXTiT0f+i1wHnpGe8a30ebl1Yy1bEjcsNCGIGwLdktTrJrVJk7qkRXXcIGE4UbgeVeG4tCd7S7PeyheSaqNUhBCcUOQ8Ox93Iwq3t8i7pdZ1113Hzp07ueiii2hqasqWrOyIhx9+mMcee8wVgiPIl48JsrlZpyFl89q+OF9qY4rtTgOWTGZcRpxkeqEVM+xWSUaWtInqEr8msLGJGTaWFHhVJxtOVQWaAgnTJqqbJA2FYFqaapO5hl9o1m1UrfXjdlyhl7U1CbY069kiy31NXgIxZswYFi5cyLZt27Asi2OPPbZVeEZbysvLXWFoQ9uc5e4k/eTiUxXOGRhi8bYWXq9JMLrE167UY1fla9r6EeBA/7YWwyZu2AQ9TiTsjhaLZuOA2ER1C78iMBUFj+I0RzFsgWHbSCRhj4qaFoioYbfqyyABS21tTTompOFVnMJq+xPWJyf8u7vNFttyyimncNddd3HWWWe1Wlq1deT1FJs3b+bWW29lxIgRjBw5krlz5/bKfTuj7cwM3U/6yeWEQm+2juvzu2JcPqKgwwiAjoSvo97YAU3JNonf3qJzfJGPpOVssjUTkqbjuY4akiZpE/E43UWDHsXJE8Wp6KEpatYpl8mw8+Ssv2zZNo9CMCCksb3FYG/c/OQIxKH6Fd5++20ANm3alD12sKIER5I33niDsrIybNvm5JNP7pV7dkZHMzPkl/STQQjBlIFhdmxsYGfUKW48trR1xGhnwqchoZ0tCgaEPHzcpLO5SWdw2ONU3hNOmLYhHYeekyMNLaYkYUp0aRHQFDyKQKLQYpiY2Q26Y6b15CgqRTgFCaLp1l6KcIoZb28xsr3k+ppuCcRpp512SBdv69DrSTpqunjKKadw7rnnEg6Hufbaa/nDH/7Qa+NpS0czc4buJv3kUuRTOaN/kH9WxVm5J8axBd5sL+euhK85vU9oq1GOLXAE4uNmnYkVAWfml+mljgRFOeChtiRZa5QtwbYligBVCGrT4eBhT2u/hwA0VaU+ZXfY+yLRjQDE3qBbAnHFFVcghGjl+Mn8oDU1Ne1Cuv/jP/6DO++8kzlz5nSoyntCQ3TUdHH16tWcccYZeL3eLj3rvcGh9Hw4GKdWBPigIUVNwuK5XVEuGRZBCNGl8HlUBcOWeNXWf5fKgEaJT6U+ZfFGbZLx5Y5Q+FSFmGGh207ikGVb2bqtCKf3XNDjBCGGPSp70n3gygNq1u+R8YXsSJoUthlPZpxHiyuiWwLRtrZqLBZjyZIlLFiwAL+/fXJH5sG84YYbjsAQD53Bgwdz9913Ew6Hueyyy/p0LIfS8+FgqEJwweAI//tRIx836axPP8hdPVyZomWWbF2iWBGCSf2DPL29hXfrU/QPamkTrCDoUUiakqRtUxFwKgEmLWdZ5FOF00lUU6hOOOmmAjixyEeRT8WrONl5TTHZql9ehoxm8B4lQa957WI+/PBDHn/8cZYtW4au69x8883MmjWr3XljxowBjkxM0+E0XRw/fnwrR2I+HOkupFJCKtXxzC040HkzXyLAF9KdSF/cE6NQMSnyKiSNzqUi4hEEFdGuw2nEBycWqnzQZPHPqhhn9fM6TjnTBgkeKQkoCs26DbbAo0FAUdAtm2jK5u06JwzjhAKVSq+Nz9bBBt2EeLrLaDKnYSRATcL5TFixe6TrK+TXhbRbArF48WIWLFhAVVUVU6dO5Y9//CM33XRTjzdD78umi4ZhsHHjxiN6za46i9a2eVDywQ9UiDJqCLF4R5wJSq3TDquDGVlVFAoUE6uT/uKVCPaICppsPyv3JhnrjRPWBLbtRL/uqjcJ+r1IJLptkdB1kqqfHTKEgUKBMBjUsoetjW2u73PCOvZW7c0eSqFQJyqc/UrDXjZWtw9JOVJ05R7Ipdtm1/POO4+777472xuiNxJ9+rLpYk91IT1Y7+lD5ThbsmhXiuokvCn6MeUYH2FVtEv6L/AqnZaTzF7Lkvx9T4qqBKzXwwzzqJxYqjJMU4hb0qkAKCQ1KcnmFpNdcUfwynyCiwcWENTa7hQgnkjwwc59VFT2w+9zltnr6gyIWvTzK5w0ePAR+R064oh3IX3uuedYuHAh8+bNo7i4mAsuuACrgzo7R5q+bLrY011Ie4LLjwvwty3NVMVNlu1OcVpFgJNLfahC5OUEDAL/dnyQl/fGWVeTYFvMYlvMwqcICn2Ol7pZt7Ol7sEpUHz2MeF2m/VcIh4I+Pz4/H6qYgZboo5WPHNAiEio5yo75jN5d2srM2TIEP793/+df/7zn3zve9/jvffeo66ujrlz5/Liiy8e8kAPhU9T08UjjV9TuPy4QkYX+5DA6zUJ/vJxMxsbU6h5+DnA2XxPPibEVSOLOKnEh08RpGxJTcJif9IiZUs8Cowu9jH3hCKmDo50KQzgNJwv8TkFmf+ZruV6UomPEYVHT5nTvDbViqJw1llncdZZZ9HQ0MCSJUt46KGHOOecc3pqfO34tDVdPNJ4FKez6XFFXlbtidGs2/yjKs5Le+MML/ByYrGPEYXeble6KAtoTBsS4fzBktqklc2FDnsUyvxqXk1dALZFTVbsi2PYMDCkMWXQ0VWj65B95cXFxdmupL3Jp7HpYk8wssjHiAIv7zekeGt/guqE5TjemnS8iuDEYh/jyrpfPlIRgoqARkXg4Od2RNyUvCdK2VvlRLYOCXu45NjIUVWCBnqotmtP8mlsuthTaIrg5FI/J5f6qUmYbGxI8X5Dimbd5u26JG/XJRka8fCFyiCDeyiN07Alb+5P8K99CXThaIOJFQHOHBDsVjGC3kbIzvIOP8NkPO9jx47t45EceaSU7IwavF2b5MNGPWuFGhL2cGb/IAOPkGDolmR9bYK1NQli6QCniExx/pAChpdGjsg9uks+f89PnIZwOTyEEAyJeBkS8TIpZbGmJsE7dUl2RA12fNzE0IiHif0CDAl7DslY0ZiyWF/raJ9UOhWv0KswoURDVu2gf6D4SH+lI4orEJ9hinwq5w0KM7FfgH/ti7OhLsX2FoPtLQblfpWxpX6OL/Rmm7h0RrNusaVZZ2ODzs6cCn7FPoWJ/YKMKfaRSibYWNXT3+jwcQXChUKvytTBEU7vF2RtTYIN9Un2Jy1W7Ymxak+MIq9Cv6BGUbqXRCY6t0m3qU443YlyGRrx8PlyP8MLvEdFFlw+uALhkqXIp3LuoDBf6h/kg4YUGxtT7I6aNOo2jXrnec8Cp1nk8UVeRhX7KMwjjP1owxUIl3b4NYXx5QHGlwdIWjZ7001TWnTHISdwUlkjHoXygEr/oNaueNknFVcgXLrEryoMK/AyrKCvR9I7fDrE2sXlCOEKhItLDq5AuLjk4AqEi0sOrkC4uOTgCoSLSw6uQLi45OAKhItLDq5jrgMMw0BK6fbUPoJksgw2b97c6ym/uq4f2TI0nzXcHO0jjxCi26VgeuLe3f2buglCLi45uHsIF5ccXIFwccnBFQgXlxxcgXBxycEVCBeXHFyBcHHJwRUIF5ccXIFwccnBFQgXlxxcgXBxycGNZXLpcz7++GMeffRRIpEIZWVlfPvb3+6zsbgC4dLnNDQ08MMf/pCysjKuvvrqPh2LG9zn0qs88cQTLFu2LPv6gQceoLy8HIBHH32UiooKLr744j4anSsQLkcBqVSKe+65h7PPPvuIN8/MF1cgXPqc//qv/+L1119nyJAh+Hw+7r777j4biysQLkeEaDTK5ZdfziOPPMLAgQMBWL58OQ8//DCGYTBjxgyuv/76Ph7lwXHNri6Hzfr167n88svZtm1b9tj+/fu57777+POf/8wzzzzDG2+8wSuvvNKHo+werkC4HDZ/+9vfuP3221v1D1+9ejUTJ06kpKQEj8fDxRdfzLPPPtuHo+wertnV5bC555572h2rqalpJSAVFRVUV1f35rAOCVdDuPQItm23O/ZJKN7gCoRLj1BZWcn+/fuzr2tqaqisrOzDEXUPVyBceoTTTz+d119/ndraWgzDYOnSpZx11ll9PayD4u4hXHqEfv36cfPNNzNv3jx0XWfy5MlMmTKlr4d1UFw/hItLDu6SycUlB1cgXFxycAXCxSUHVyBcXHJwBcLFJQdXIFxccnAFwsUlB1cgepnt27czcuRIFi1alD02efJkdu/ezaJFi7j11lu7/PycOXNYs2ZNt+/35JNP8uUvf7lPk24+SbgC0cssXLiQc889lwULFvTK/f7+978zf/58fvKTn/TK/T7puALRi1iWxeLFi7nhhhvYu3cvGzdu7PTc9957j3/7t39j5syZXHnllWzZsqXdOf/zP//DxRdfzIwZM7jzzjsxTbPV+w8//DAbNmxg/vz5rFy5ksmTJ3PjjTdy3nnnsW/fPpYsWcIll1zCRRddxPe//32i0SgAixcv5rzzzmPWrFnceeedWa2V0WQAu3fvZvLkyQDU19dz/fXXc8kll3DJJZewatUqAB566CFuu+02rrzySs4++2zmz5+fHdsDDzzAeeedx7Rp0/jtb3/L7t27OfPMM7EsC4ANGzbwla985VB/6kPGFYhe5JVXXqGoqIjjjjuOqVOn8vjjj3d4nmEY/PjHP+a+++7j6aef5sYbb+Tmm29udc6rr77K22+/zVNPPcWSJUvQdb2d1rn++usZM2YMd911F2effTYAZ5xxBs8//zyxWIwFCxbw17/+lSVLljBixAh+9atfsW/fvmym24IFC1plwXXG3XffzYwZM1i0aBG///3vueeee6irqwNg48aN/P73v2fp0qWsWLGCTZs28cILL7B27VqWLl3K008/zYsvvohlWQwfPpzVq1cD8PTTTzNr1qy8f+PDxQ3u60UWLlzI9OnTAZg+fTpf//rXueWWW9qdt23bNnbu3Ml3vvOd7LH6+np0Xc++Xr16Ne+++272oUmlUqiqetAxjBs3DoDXXnuNHTt2ZGdh0zQZNGgQ69evZ/z48dnknpkzZ2Yf0s549dVX+fjjj/nVr36VvVZGo51++ul4vV68Xi9DhgyhqamJNWvWMHXqVHw+H+DscwAuvfRSlixZwsSJE1m1ahX//u//ftDvc6RxBaKXqK+v5x//+AfvvPMOTzzxBOBogiVLlrQ717ZtBg0alH1PSkl1dXWrLp6WZTF37lzmzZsHQEtLS7cScPx+f/bz06ZN47bbbgMgHo+j6zrr168nN95T01o/Ipn3DMNoNd4//elPFBUVAU7uQ0lJCWvWrMk+9OAkCEkpUVW11VirqqooKChgypQp3HvvvTz//PNMmDCBcDh80O9zpHGXTL3E0qVLGT9+PC+//DKrVq1i1apV/OAHP+hwc33sscfS1NTEunXrAGdjfO2117Y6Z+LEiSxZsoRYLIZlWXzve99j4cKF3R7PaaedxooVK6itrQWcNNBf//rXnHzyybz33nvs2bMH27Zb5UEXFxezadMmAJ577rlWY/nrX/8KOFa06dOn09TU1Om9J0yYwIoVK9B1HV3Xufbaa9m2bRs+n48pU6Zw//3398lyCVwN0WssXLiQG264odWxSy+9lIcffphEItHquNfr5cEHH+RnP/sZyWSSYDDIL37xi1bnTJ48mU2bNnHZZZdhWRYTJkzga1/7WrfHM3LkSK6//nrmzZuHbdsMHz6cW2+9lVAoxJ133sl1112Hpmmt8qJvuukm7rzzTn71q19x3nnnZY/fdttt3H777Vx44YVIKbn77rspLS3t9N7nnHMOH3zwAbNmzcK2bWbNmsXYsWMBmDZtGi+99BITJkzo9nc5krj5EC5dsmjRItauXcvPf/7zHr+XZVk89NBD+P3+dhqxt3A1hMtRw6xZs4hEIvzmN7/pszG4GsLFJQd3U+3ikoMrEEcBGzZsyCu0YuXKlTz44INdnvPggw+ycuVKwIl/ynDRRRd1+bnuxFMdDmvWrGk1nqMNdw9xFDB27NislaU7nH322VnPc2fceOON2X+vXbs2+++O/B4uB3A1xFFA7qw5Z84cfvGLX/DVr36VyZMn8/TTT7c7P3cWnzx5Mg8++CCXXXYZ5557brag8K233sqiRYu44447ALjkkksAOOGEEwCorq7mqquu4rLLLuOss87i3nvv7XKMc+bM4ac//SmXXHIJU6dO5aWXXgIgkUjw/e9/n+nTp3PhhReyePFiwHHW3XXXXUybNo3p06fz6KOPHuav1Du4GuIoJJFIsGDBAj744APmzZvHzJkzuzw/Eonwt7/9jRUrVvD//t//a9V05Pbbb+evf/1rq3BzgGXLlnH++ecze/ZsotEokyZNOmg7q1QqxcKFC/nwww/5xje+wUsvvcRDDz1EYWEhy5Yto76+ntmzZzNy5EjefPNN9uzZw9KlSzEMgzlz5nD88ccTCAQO/YfpBVwNcRQyadIkAEaNGkVjY2O3zx85cmS3zge46qqrGDRoEP/zP//DXXfdha7r7RyEbZk9ezZCCEaNGkVlZSWbNm3i9ddfZ/bs2QCUlJRw9tlns3btWtasWcOsWbPQNI1AIMCFF17Ia6+91q2x9SWuhjgKycT/dLc4cL7nA/z85z9nx44dzJgxg3POOYd//etfHMwCnxs8aNs2qqq2+4yUEtM02xU7zhw/2nE1xGcAVVXbPYyrV6/m6quvZurUqezdu5fq6uoOK3bn8swzzwCOVayxsZHjjz+eiRMnZqNV6+vrefHFFznllFOYOHEiCxcuxDRNEokEf//73zn11FN75gseQVwN8RlgypQpzJgxg6eeeip77JprruGWW26hoKCAkpISxo4dy65du7q8TnV1NTNnzsS2bR544AE0TeM73/kO//mf/8n06dOxLItvfetbnHTSSYwaNYrt27dz8cUXYxgG06dP5/zzz88r/bUvcD3VLt1izpw5XH/99Zx22ml9PZQexV0yubjk4GoIF5ccXA3h4pKDKxAuLjm4AuHikoMrEC4uObgC4eKSgysQLi45uALh4pKDKxAuLjn8f+EFNzLn6ZHUAAAAAElFTkSuQmCC",
      "text/plain": [
       "<Figure size 150x150 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "Nham_aa_colors = ['black', 'skyblue', 'pink']\n",
    "sns.set(rc = {'figure.figsize':(1.5,1.5),\n",
    "              'font.size':8,\n",
    "              'legend.title_fontsize':8, 'legend.fontsize':8,\n",
    "              'axes.labelsize':8,'axes.titlesize':8,\n",
    "              'xtick.labelsize':8, 'ytick.labelsize':8,\n",
    "              'xtick.major.pad':1, 'ytick.major.pad':1,\n",
    "             },\n",
    "       style='whitegrid')\n",
    "\n",
    "\n",
    "fig = sns.scatterplot(corrdf[corrdf.Nham_aa == 1],\n",
    "                      x='average_freq_pool', y='average_freq_input',\n",
    "                      color=Nham_aa_colors[1], alpha=.2, legend=False\n",
    "                     )\n",
    "sns.scatterplot(corrdf[corrdf.Nham_aa == 0],\n",
    "                x='average_freq_pool', y='average_freq_input',\n",
    "                color=Nham_aa_colors[0], legend=False\n",
    "               )\n",
    "sns.kdeplot(corrdf,x='average_freq_pool', y='average_freq_input',\n",
    "            hue='Nham_aa', palette=Nham_aa_colors, common_norm=False,\n",
    "            levels=5, log_scale=True,legend=False)\n",
    "fig.set(xscale='log', yscale='log',\n",
    "        xlabel='Allele frequency\\nin initial pool',\n",
    "        ylabel='Allele frequency\\nin DMS input'\n",
    "       )\n",
    "both_perc = ratios.loc[ratios.Location == 'both', 'aa_seq'].values[0] / unique_aa\n",
    "fig.annotate(f'{both_perc:.1%}',(.72,.9), xycoords='axes fraction')\n",
    "\n",
    "plt.savefig(f'{graph_outpath}/{\"_\".join([strain,locus,pool_type])}_bottleneck.svg', format='svg', dpi=300)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "470d3695-f182-4392-b42b-48670af7803d",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "dms-env",
   "language": "python",
   "name": "dms-env"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.11.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}