classification_NNK.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 97,
   "id": "97aab040-a997-4a24-85a3-2ec89b8bd0f4",
   "metadata": {
    "tags": [
     "parameters"
    ]
   },
   "outputs": [],
   "source": [
    "# Parameters\n",
    "# Refer to injected parameters for any run.ipynb (similar cell below)\n",
    "strain = 'BY4741'\n",
    "locus = 'FKS1-HS1'\n",
    "pool_type = 'single'"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "574d242e-a54d-487e-8965-20932b7b8626",
   "metadata": {},
   "source": [
    "# Notebook to classify variants using a Gaussian mixture model"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ee0e22d9-7097-4741-99bb-f6d09e5eacf3",
   "metadata": {},
   "source": [
    "This notebook currently supports:\n",
    "- BY FKS1-HS1 NNK (+orthos)\n",
    "- BY FKS1-HS2 NNK (+orthos)\n",
    "- R1158 FKS1-HS1 NNK  - skip manually section to include missing mutants\n",
    "- R1158 FKS2-HS1 NNK (+ orthos)\n",
    "- R1158 FKS2-HS2 NNK (+ orthos)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "90c9799e-4251-4972-b974-4931c98d3e63",
   "metadata": {},
   "source": [
    "## Import libraries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "id": "30da980b-701d-47ed-9025-5ea72e42cdf6",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "import os\n",
    "import pandas as pd\n",
    "from sklearn.mixture import GaussianMixture\n",
    "import numpy as np\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": "3093dbac-164c-4a3e-822f-1d13a6447da2",
   "metadata": {},
   "source": [
    "## Specify paths"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "id": "57972351-e6ef-4079-8bd7-cc4bf1f316ee",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "aggdata_outpath = 'aggregated_data/'\n",
    "missing_mut_path = 'aggregated_data/validation_DMS_missing_estimates.csv'\n",
    "classified_outpath = 'aggregated_data/classified_variants/'\n",
    "graph_outpath = 'manuscript_figures/'"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "70a36e16-2cde-4a12-ae77-d11790f79858",
   "metadata": {},
   "source": [
    "## GMM"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "id": "c677017d-e161-406c-ad6a-b908c661968c",
   "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>compound</th>\n",
       "      <th>seq_type</th>\n",
       "      <th>Nham_aa</th>\n",
       "      <th>aa_seq</th>\n",
       "      <th>nt_seq</th>\n",
       "      <th>s</th>\n",
       "      <th>cscore</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>anidulafungin</td>\n",
       "      <td>WT</td>\n",
       "      <td>0.0</td>\n",
       "      <td>FLVLSLRDP</td>\n",
       "      <td>tttttagttttatctttgagagatcca</td>\n",
       "      <td>0.059886</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>anidulafungin</td>\n",
       "      <td>ortho</td>\n",
       "      <td>0.0</td>\n",
       "      <td>FLVLSLRDP</td>\n",
       "      <td>tttttggttttgtctttgagagatcca</td>\n",
       "      <td>0.042567</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>anidulafungin</td>\n",
       "      <td>ortho</td>\n",
       "      <td>1.0</td>\n",
       "      <td>FLILSLRDP</td>\n",
       "      <td>tttttgattttgtctttgagagatcca</td>\n",
       "      <td>-0.223979</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>anidulafungin</td>\n",
       "      <td>ortho</td>\n",
       "      <td>1.0</td>\n",
       "      <td>FLTLSLRDP</td>\n",
       "      <td>tttttgactttgtctttgagagatcca</td>\n",
       "      <td>0.127139</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>anidulafungin</td>\n",
       "      <td>ortho</td>\n",
       "      <td>2.0</td>\n",
       "      <td>FLALSFRDP</td>\n",
       "      <td>tttttggctttgtcttttagagatcca</td>\n",
       "      <td>0.370152</td>\n",
       "      <td>1.0</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>1259</th>\n",
       "      <td>none</td>\n",
       "      <td>single</td>\n",
       "      <td>1.0</td>\n",
       "      <td>TLVLSLRDP</td>\n",
       "      <td>acgttagttttatctttgagagatcca</td>\n",
       "      <td>-0.449800</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1260</th>\n",
       "      <td>none</td>\n",
       "      <td>single</td>\n",
       "      <td>1.0</td>\n",
       "      <td>TLVLSLRDP</td>\n",
       "      <td>actttagttttatctttgagagatcca</td>\n",
       "      <td>0.033715</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1261</th>\n",
       "      <td>none</td>\n",
       "      <td>single</td>\n",
       "      <td>1.0</td>\n",
       "      <td>VLVLSLRDP</td>\n",
       "      <td>gtgttagttttatctttgagagatcca</td>\n",
       "      <td>0.037623</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1262</th>\n",
       "      <td>none</td>\n",
       "      <td>single</td>\n",
       "      <td>1.0</td>\n",
       "      <td>WLVLSLRDP</td>\n",
       "      <td>tggttagttttatctttgagagatcca</td>\n",
       "      <td>0.070207</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1263</th>\n",
       "      <td>none</td>\n",
       "      <td>single</td>\n",
       "      <td>1.0</td>\n",
       "      <td>YLVLSLRDP</td>\n",
       "      <td>tatttagttttatctttgagagatcca</td>\n",
       "      <td>0.080017</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1264 rows × 7 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "           compound seq_type  Nham_aa     aa_seq                       nt_seq  \\\n",
       "0     anidulafungin       WT      0.0  FLVLSLRDP  tttttagttttatctttgagagatcca   \n",
       "1     anidulafungin    ortho      0.0  FLVLSLRDP  tttttggttttgtctttgagagatcca   \n",
       "2     anidulafungin    ortho      1.0  FLILSLRDP  tttttgattttgtctttgagagatcca   \n",
       "3     anidulafungin    ortho      1.0  FLTLSLRDP  tttttgactttgtctttgagagatcca   \n",
       "4     anidulafungin    ortho      2.0  FLALSFRDP  tttttggctttgtcttttagagatcca   \n",
       "...             ...      ...      ...        ...                          ...   \n",
       "1259           none   single      1.0  TLVLSLRDP  acgttagttttatctttgagagatcca   \n",
       "1260           none   single      1.0  TLVLSLRDP  actttagttttatctttgagagatcca   \n",
       "1261           none   single      1.0  VLVLSLRDP  gtgttagttttatctttgagagatcca   \n",
       "1262           none   single      1.0  WLVLSLRDP  tggttagttttatctttgagagatcca   \n",
       "1263           none   single      1.0  YLVLSLRDP  tatttagttttatctttgagagatcca   \n",
       "\n",
       "             s  cscore  \n",
       "0     0.059886     1.0  \n",
       "1     0.042567     1.0  \n",
       "2    -0.223979     1.0  \n",
       "3     0.127139     1.0  \n",
       "4     0.370152     1.0  \n",
       "...        ...     ...  \n",
       "1259 -0.449800     1.0  \n",
       "1260  0.033715     1.0  \n",
       "1261  0.037623     1.0  \n",
       "1262  0.070207     1.0  \n",
       "1263  0.080017     1.0  \n",
       "\n",
       "[1264 rows x 7 columns]"
      ]
     },
     "execution_count": 100,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "master = pd.read_csv(f'{aggdata_outpath}/master.csv', index_col=0)\n",
    "slpNT = master[(master.strain == strain)\n",
    "                   & (master.locus == locus)\n",
    "                   & (master.pool_type == pool_type)\n",
    "                  ].groupby(['compound','seq_type','Nham_aa','aa_seq','nt_seq']\n",
    "                           )[['median_s','confidence_score']\n",
    "                            ].agg(s = ('median_s', 'first'),\n",
    "                                  cscore = ('confidence_score', 'first')\n",
    "                                 ).reset_index()\n",
    "slpNT"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "id": "3667a923-3900-4832-b8ef-1593deadeb32",
   "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>Nham_aa</th>\n",
       "      <th>s</th>\n",
       "      <th>cscore</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>1264.000000</td>\n",
       "      <td>1264.000000</td>\n",
       "      <td>1264.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>1.319620</td>\n",
       "      <td>0.743967</td>\n",
       "      <td>1.056962</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>1.092287</td>\n",
       "      <td>0.910137</td>\n",
       "      <td>0.257737</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>-1.144093</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.011592</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.140323</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.805569</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>7.000000</td>\n",
       "      <td>2.239924</td>\n",
       "      <td>3.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           Nham_aa            s       cscore\n",
       "count  1264.000000  1264.000000  1264.000000\n",
       "mean      1.319620     0.743967     1.056962\n",
       "std       1.092287     0.910137     0.257737\n",
       "min       0.000000    -1.144093     1.000000\n",
       "25%       1.000000     0.011592     1.000000\n",
       "50%       1.000000     0.140323     1.000000\n",
       "75%       1.000000     1.805569     1.000000\n",
       "max       7.000000     2.239924     3.000000"
      ]
     },
     "execution_count": 101,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "slpNT.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "de5fb6ea-6d8a-492f-aa7f-f728130b202f",
   "metadata": {},
   "source": [
    "### Optimize model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "id": "52831485-157c-4e41-a510-e7d0ccfa2fdd",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1264"
      ]
     },
     "execution_count": 102,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "GMMdf = slpNT\n",
    "len(GMMdf)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "id": "bcd6d7d5-bdf8-4863-b453-8c5d2c753082",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "# Set specific seed\n",
    "seed = 25 # Set value for repetability\n",
    "\n",
    "# Initialize dictionary to store parameters for each condition\n",
    "model_dict = {}\n",
    "\n",
    "# Specify order of conditions\n",
    "lcomp = ['none','dox','anidulafungin','caspofungin','micafungin']\n",
    "\n",
    "# Optimize model for every condition (here, compound) separately\n",
    "for j,c in enumerate(lcomp):\n",
    "    if c in GMMdf.compound.unique():\n",
    "        # The goal is to minimize AIC and BIC, but mostly BIC\n",
    "        X = GMMdf[GMMdf.compound == c].s.values.reshape(-1,1)\n",
    "        N = np.arange(1, 6)\n",
    "        models = [None for i in range(len(N))] # testing up to 1-5 components\n",
    "        # 5 is the max number of biologically relevant classes: deleterious, WT-like, intermediary, resistant, very resistant\n",
    "        for i in range(len(N)):\n",
    "            models[i] = GaussianMixture(n_components=N[i], random_state = seed).fit(X, y=None)\n",
    "\n",
    "        # Compute AIC and BIC\n",
    "        AIC = [m.aic(X) for m in models]\n",
    "        BIC = [m.bic(X) for m in models]\n",
    "    \n",
    "        # Save parameters to dictionary\n",
    "        model_dict[c.title()] = (N, AIC, BIC)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "id": "ca436c99-d081-4de7-95c6-08f9cceec5aa",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "model_df = pd.DataFrame.from_dict(model_dict, orient='index', columns=['components','AIC','BIC']).reset_index(names='compound')\n",
    "model_df = model_df.explode(['components','AIC','BIC']).melt(['compound','components'], var_name='IC')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "id": "9d7ec435-74e3-411e-a925-dda2d7d0b497",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 858.5x200 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Purge graph space\n",
    "sns.set(rc = {\n",
    "              'font.family':'Arial',\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':2, 'ytick.major.pad':2,\n",
    "              'xtick.bottom': True, 'ytick.left': True,\n",
    "              'xtick.major.size':2, 'ytick.major.size':2,\n",
    "             },\n",
    "       style='ticks')\n",
    "\n",
    "fig = sns.relplot(data=model_df, col='compound', x='components', y='value',\n",
    "                  hue='IC', style='IC',\n",
    "                  kind='line', height=2,\n",
    "                  facet_kws = {'sharey':False, 'despine':False})\n",
    "\n",
    "fig.set_titles(row_template='{row_name}', col_template='{col_name}')\n",
    "fig.set_axis_labels('Components', 'Information criterion')\n",
    "sns.move_legend(fig, 'upper left', bbox_to_anchor=(.175, .88), title=None)\n",
    "plt.savefig(f'{graph_outpath}/{\"_\".join([strain,locus,pool_type])}_GMM_BIC.svg', format='svg', dpi=300)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "id": "24dcce9c-43e2-452d-ac8a-143e3059d921",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "None 3\n",
      "Anidulafungin 4\n",
      "Caspofungin 5\n",
      "Micafungin 4\n"
     ]
    }
   ],
   "source": [
    "for c,y in model_dict.items(): # for each compound\n",
    "    print(c, np.argmin(y[1])+1) # display the optimal number of components for the GMM"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7ab9e077-dd8b-498d-9275-e1996b97f5e9",
   "metadata": {},
   "source": [
    "### Run model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "id": "f54d8447-8fc6-4881-87d8-d5066684b80d",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "# Define the number of components (number of mods)\n",
    "mods = 3\n",
    "\n",
    "# Define range\n",
    "mini = round(GMMdf.s.min()) -1\n",
    "maxi = round(GMMdf.s.max())+1\n",
    "\n",
    "# Initialize list of df\n",
    "proba_list = []\n",
    "\n",
    "for i,c in enumerate(model_dict.keys()): # For every compound\n",
    "    X = GMMdf[GMMdf.compound == c.lower()].s.values.reshape(-1,1)\n",
    "    M_best = GaussianMixture(n_components=mods, random_state = seed).fit(X, y=None)\n",
    "    GMMdf.loc[GMMdf.compound == c.lower(), 'label'] = M_best.predict(X)\n",
    "    \n",
    "    x = np.linspace(mini, maxi, len(GMMdf[GMMdf.compound == c.lower()]))\n",
    "    logprob = M_best.score_samples(x.reshape(-1, 1))\n",
    "    responsibilities = M_best.predict_proba(x.reshape(-1, 1))\n",
    "    pdf = np.exp(logprob)\n",
    "    pdf_individual = responsibilities * pdf[:, np.newaxis]\n",
    "    \n",
    "    # Convert to dataframe\n",
    "    indiv_df = pd.DataFrame(pdf_individual)\n",
    "    indiv_df['x'] = x\n",
    "    indiv_df = indiv_df.melt(id_vars = 'x', var_name = 'label', value_name='p(x)')\n",
    "    indiv_df['compound'] = c\n",
    "    proba_list.append(indiv_df)\n",
    "\n",
    "proba_df = pd.concat(proba_list, ignore_index=True)\n",
    "hist_df = GMMdf.copy()\n",
    "hist_df['compound'] = hist_df.compound.apply(lambda x: x.title())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "id": "60e568f8-b4e0-4704-9c4e-6ec563c757f2",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "stddf = pd.read_csv(f'{classified_outpath}/{\"_\".join([strain,locus,pool_type])}/thresholds.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "id": "aabe677f-fb18-4a57-aea1-c527f88dcccd",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAyUAAADECAYAAACBZI4lAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjguMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/SrBM8AAAACXBIWXMAAA9hAAAPYQGoP6dpAABcAklEQVR4nO3dd3xUZdbA8d+dkkZCQoCQIAgWlCIWFhB1VSQriqCAwiq42EFRV1ZBBJFFVMSKi+iirm1VEEH6u+C6Bgs2EEXpRQglhVTSJ5l23z8mM5kW0jN3Zs7383lfkrkzNyfrc3LvuU9TVFVVEUIIIYQQQogA0QU6ACGEEEIIIUR4k6JECCGEEEIIEVBSlAghhBBCCCECSooSIYQQQgghREBJUSKEEEIIIYQIKClKhBBCCCGEEAElRYkQQgghhBAioKQoEUIIIYQQQgSUFCVCCCGEEEKIgJKiJAxkZGRw7rnn8vXXX3u8PmTIEDIyMgIUlRB1O3LkCD179mTVqlV1vnfhwoWkpaX5vL5o0SIWLVp0ys9OmDCBLVu2nPI9r732GkOGDOGtt96qM5b62rlzJ7NmzWq284nQVVZWxtNPP82IESMYOXIkt956K9u3b2+Vn/34448zdOhQ1qxZ02znTEtLY+HChc12PhGenPc3M2fO9Hg9Pz+f3r17s2jRIiZOnEhOTk6Dz71nzx5SU1MZN24cqqo2V8iNjiccGAIdgGgdRqOROXPmsH79euLi4gIdjhD1snLlSoYOHcqyZcu48cYbT/neKVOmtGgsq1ev5u233+bMM89stnP27duXvn37Ntv5RGiy2+1MmjSJgQMHsmbNGgwGA9u3b2fy5MmsXr2alJSUFv35q1atYseOHURERDTbOVNTU0lNTW2284nw1a5dOzZv3ozFYsFoNAKwceNGEhISAPjXv/7VqPNu2rSJ6667jqlTpzZXqE2KJxxIURImkpKSuPTSS5k/fz7PPvusx7E33niDdevWodfrueyyy3j00UfJzs7m/vvvp0+fPuzevZuoqChefvllunbtyq5du3j22WcxmUzExcUxZ84czjrrrAD9ZiJU2Ww21qxZw7vvvstdd93F3r176dWrF0OGDGHkyJF89913FBUVMXv2bC6//HJmzJjBwIEDufHGG3n77bdZvnw57dq1o23btpx//vkAnHvuuezfvx+ALVu28Nprr/Hhhx+6fqbVauXJJ5/kwIEDFBQU0L17d1577TXmz59PTk4ODz74IM8//zxjxozxe54JEyZwwQUXsG3bNnJzc/nrX//K6NGjKSsr4/HHH+fgwYN07NgRRVG4//77Aer8rBBbtmwhJyeHhx56CJ3OMcDhoosu4sUXX8Rut/PEE0/4tFlFUZg+fTrHjh1DURRuvvlmbrnlFmbMmIGiKBw4cIDS0lImT57M6NGjsdvtPPvss3z//ffodDpuuOEGJk2axH333YeqqowdO5Y5c+Ywbdo0Nm3aBDiKla1bt/Lcc8/Vmpc5OTlMnz6dwsJCTjvtNHJycli0aBFbt26t87NC1EdMTAx9+vThu+++Y/DgwQB89tlnXH311YBjVMgHH3xAUlISTz/9NFu3bsVgMHD33Xdz4403snHjRt577z0qKyuprKzkqaeeoqqqio8//hgAg8FAdna26/rifs6tW7fyzTffUFZWxvHjx+nVqxcvvPACERERLF26lH//+9/ExMRw3nnnYbFYXO29rs+GKxm+FUZmzJjBDz/8wObNm12vff3113zxxResXLmS1atXc/ToUZYtWwbAgQMHGD9+POvWreP8889nyZIlWCwWHn/8cV544QVWr17NlClTePTRRwP1K4kQtnnzZhISEujRowfDhg1zXSAA4uLiWL58OY8++iivvPKKx+d27tzJihUrWLVqFe+//z4nTpyo98/cvn07Op2O5cuX88UXX2A2m/nmm2946qmnSEpK4q233qqzZ8NkMrFs2TJee+01nnvuOQBef/11OnXqxMaNG5k3bx47duyo92eF2LNnD3379nUVJE6XX345WVlZftvst99+i6qqrsL+559/dn0uKyuLZcuW8cEHH/D888+Tk5PDxx9/TGZmJuvWrWPFihV8/vnnfPXVV7zxxhsArF27lqSkpFPG6S8v582bx1VXXcX69eu577772LdvX70/K0R9DRs2jI0bNwKQmZlJVFQUHTp08HjPkiVLKC4uZsOGDXz00Ue8/fbblJaW8vHHH7N48WLWrVvHvffey/vvv8+VV17JLbfcwi233FJnL/z27dt55ZVX2LBhA8eOHePbb79l//79vPvuu3zyySesWLGC/Pz8en82nElPSRiJjY3l6aefZvbs2axfvx5wPIEbMWIE0dHRANx0002sWbOGK6+8kvbt27tuwHr16sW2bdtIT0/n2LFjPPDAA67zFhYWYjabw7q6F81v5cqVjBgxAoARI0Zw++23M336dACuvPJKAHr27ElRUZHH57Zu3crgwYOJjY0F4Nprr8Vut9frZw4YMICEhASWLFnC4cOHOXLkCBUVFQ2K2xlbr169XLF9++23vPjiiwB07dqVSy+9tN6fFUKn09Xahmtrs5deeinz5s3j7rvv5sorr+Sxxx5zfWb06NEYjUaSk5Pp378/27dvZ8uWLdx0000YDAYMBgPXX389P/zwg+vJc334y8tvv/3W1Tt/4YUXcu6559b7s0LU11VXXcVzzz2H2Wxmw4YNXHfddWRlZXm8Z8uWLYwZMwa9Xk+7du3YsGEDAP/85z/ZtGkT6enpbN261af4r0u/fv1cw+J79OhBcXExR48eJTU11TWE7KabbuKLL76o12fDmfSUhJk//vGPXHbZZa6nsP4udFarFYDIyEjXa4qioKoqdrudrl27snbtWtauXcuaNWtYsWKFFCSiWRUWFvLll1+yZMkShgwZwt/+9jcsFgtr164Fatqmoig+n3W2VSfnGGMn5zGLxeLz2bS0NB555BGioqK48cYbGTBgQK0THGs7j7/Y9Hp9vSZKnur3EuHrvPPOY/fu3T5taPHixbW2WWfP3Pjx4zl8+DCjR4+mpKQEcAxHcbLb7ej1ep9rgaqqrmuBk3du1bft1+ehgLR90RTR0dH079+f7777ji+++IKhQ4f6vEev13u0r2PHjlFeXs5NN93E8ePHGTBgABMmTPB7fu+2bzabXV/7u1c61YMEd/4+G86kKAlDM2bM4LvvviM3N5dBgwbxf//3f5hMJqxWKytXrmTAgAG1fvbMM8+kuLiYn376CcDVJS9Ec1q3bh39+vXjm2++YdOmTWzatIlp06a5hhaeyiWXXMKmTZsoKSnBbDZ7PJ1q166da/jIZ5995vPZH374geHDh3PTTTfRoUMHfvrpJ2w2m8/76jqPt8suu4zVq1cDcOLECbZs2SI3X6Le+vfvT8eOHfnHP/7hKhR++uknPvroI1asWOG3za5fv54nn3yS1NRUnnjiCWJiYsjOzgYcbdZut5OZmclvv/3GH/7wBwYNGsTKlSuxWq2YTCbWr1/vcy2Ij4+nqKiI3NxcbDYbn3/+eZ2xu7f9PXv2cODAAWn7okVcd911vPnmmyQlJfld0GfgwIFs3LgRVVUpLi5mwoQJ7N+/H0VRmDx5MoMGDeKbb77x+zc/ISHB9Td/27Zt5OXlnTKWSy+9lK+++oqSkhLsdjvr16+Xdl8PMnwrDDmHcd19990MHjyYkpISxowZg9Vq5dJLL+W2226rdbm6iIgIFi5cyLPPPktlZSUxMTG89NJLrfwbiFC3cuVK/vrXv3q8NmbMGF577TVMJtMpP9urVy/uvPNOxowZQ3x8vMfKRI899hgPPPAAiYmJXHHFFRw9etTjs2PHjmXatGl89tlnREREcNFFF/ldNruu83ibPHkys2fP5vrrr6djx4507tyZqKioOn8XIcDxBHXx4sU899xzXH/99RgMBtq2bcvrr79OdHS03zZ7//33k5aWxvDhwzEajVxzzTWuoVMWi4UxY8ZQVVXF3LlzSUxM5Oabb+bIkSOMGjUKi8XCiBEjuPbaaz3iiI2N5b777uOWW26hQ4cO9O/fn8LCwlPGPmvWLGbMmMGqVavo2rUrHTt2JCoqqsX+txLh64orrmDGjBnccccdfo+PGzeOw4cPc8MNN6CqKlOnTuWCCy6gd+/eDBs2jKioKAYMGEBmZqZPj8X48eN55JFHGDFiBH369KFPnz6njKVHjx5MnDiRcePGERkZSZcuXTx6RYR/ihrufUVCCNHC1q1bR3JyMgMHDqSsrIwbb7yRFStWEB8fH+jQRJhxX6WuNXz00Uf079+fnj17kpOTw6233srnn3/e4HH7QgSTo0eP8vnnnzNx4kQAnn32Wbp168att94a4Mi0TXpKhBCihZ155pnMmTPHNSxgypQpUpCIsNCtWzcee+wxFEVBr9fz9NNPS0EiQl5KSgoHDhxg+PDh6HQ6zj//fMaOHRvosDRPekqEEEIIIYQQASWPK4QQQgghhBABJUWJEEIIIYQQIqCkKBFCCCGEEEIElBQlQgghhBBCiIAKudW3+vfvj9lspmPHjoEORYSpvLw8IiIi2LZtW8BikDwQgSZ5IMKdFnIAJA9EYDUkD0KuKKmqqvK7G2c4UFWVoqIiwLH7qOweGhhWq9Vn46XWFs550BK0lFtaiuVUJA9CixbanRZiaAgt5ABIHjSHYGh7Wo2xIXkQckVJUlISAGlpaQGOpPWZzWbmz58PwMyZM4mIiAhwROEpNTU10CGEdR60BC3llpZiORXJg9CihXanhRgaQgs5AJIHzSEY2p5WY2xIHsicEiGEEEIIIURASVEihBBCCCGECCgpSoQQQgghhBABJUVJiLJbLZhysrCZqwIdihCiFdiqKjHlZKHKhFYhhBAE372gFCUhypSdiSkni5Lf9wU6FCFEC1NVleL9uzDlZFFVVBDocIRoVc6VfbSw0pUQWlJ+PB1TThZlRw8FOpR6CbnVt8KZXq/nkksuAcBeVYkCqFYLqqpqZmk4IYKRe27p9XrNxWJ3ewpmq6wMSFwitGkhB2qLQVEUsrOzSUlJCUhcIvRpof3XxV+MltISAGymioDF1RBSlIQQvV7P0KFDASg98juWEnOAIxIiNLjnVqD5i8VjyJY8LRYtQAs5cKoYqqqCY3iKCE5aaP91CYYY69Lqw7fKysq4/vrrycjIAGDr1q2MHDmSa665hjlz5mC1Wv1+ZvLkyVx33XXcdNNNHDlypJWjDj46Y8361KpdxpgLEcpUW83fTbvN92+oEEKI8KVotHfHW6sWJdu3b2fcuHGkp6cDjo1eHnvsMf7xj3/w2WefUVlZyaeffurzuYULF9KnTx82bNjAtGnTmDFjRmuGHTScu3kWFRWhorofCFxQQoQAj9wKcD75i8Xu1lOi+nmwI0RTaSEHtBCDCE/B0Pa8Y9RqnKfSqkXJ8uXLmTNnjmt30R07dtClSxfOOOMMFEVhzJgxbNiwwedzX375JaNHjwbgkksuIS8vj6ysrNYMPShYLBYWLlzIwoULsVjcbkyCr10KoSmeuWXRXCzuw7dU6SkRLUALOaCFGER4Coa25xOjW1ESLAVKq84pmT9/vsf3OTk5dOrUyfV9UlISOTk5Pp/z974TJ07QuXNn12vObexlspuT9JQIES48hm9JT4kQQoQ9VbW7fxO4QBogoBPd7Xa7z2v+VonyV+HpdLKa8Sm5V8jSVSJESPOc6O77d1WIUKXa7ZhLilD83E8IEdbsng+ng2El1oAWJcnJyeTl5bm+z8vLIzk52ed9nTp1Ii8vz9UD4u99aWlpQE2PSdhTpadEiHDhvpiFKjdnIoxU5p3AlJNFjC44JvIK0VpUrwdUqt2Gotf2orsB7W644IILOHLkCIcOOTZ1WblyJYMHD/Z53+DBg1m5ciUAW7ZsISYmxm/xIty41yFSlAgR0jwKEVX1uRgJEarMxScB0Msqk0J48n5AZdf+vWBAS6aIiAhefPFFpk6dSmVlJeeffz7jx48HHCtuJSUlMW7cOKZMmcKsWbMYMWIEERERvPDCC4EMO0gE3wQnIUQj+bv4yINjEQYUQ81tTDAMTxGitXj3mgfDw6qAFCWbNm1yfT1w4EDWrFnj854pU6a4vm7bti2LFi1qjdBCh/SUCBE2fC4+dnvQrEsvRFPo3IajqFYLitseXUKEM58iJAjuBbU9uEw0iE6no3///o5vPB4Wab8hCqFl7rkV6EU2/MXiO3ZY+0/ERHDRQg74bfsyn0q0Ai20/7p4x6h6DdcKhlEzUpSEEIPBwPDhwwEoTT+IcyXtIGiHQmiae24Fmt9YgrCbXgQXLeSAvxg8Ng6VokS0EC20/7p4x2j2vg4EQX5os9wTzUuqEiFCms/NmEz6FWFCdduXR4oSIWr4zinR/r2g9JSEEFVVqaioALz2gAmChiiElrnnVkxMTEAn0/qLxftiIzdnorlpIQf8tn33AlzavWghWmj/dfGO0ScfgqAHXXpKQojFYuGll17ipZdewur+9EjmlAjRJO65ZbFY6v5Aa8fiZ6K7EM1JCzngv+27rzQp7V60DC20/7p4x+gz1zAIHlBLURKyZPNEIcKFswhxrbglRYkIE+43WqFWjJeVlXH99deTkZHhc2zZsmX88Y9/ZOTIkYwcOZJXXnklABEKLfOe6B4M1wUZvhWqZEd3IcKG84mYojeg2mwhd3MmRK3cnwaHULvfvn07f//730lPT/d7fMeOHcyZM4err766lSMTwSP4Vt+SnpIQpco+JUKEBVVVXTdjSvWeDcFw8WmK9957jxEjRjBixAhmzpyJ2WwOdEgiAEJ5LtXy5cuZM2cOSUlJfo/v3LmTFStWcMMNNzB9+nRKSkpaOUKhed7XgSC4LkhRErJUP18JIUKO24Wm0nlzHsJj63fs2MGqVatYvnw569evx2q1snTp0kCHJQLBZ8x86LT7+fPn1+w75sVut5OSksKUKVNYu3YtSUlJPPPMMz7vS01NJTU1lezs7JYOV2hREBbtMnwrVMnwLSHCgvuNmNVux4ifscQhpG3btsyePduxugzQs2dPsrKyAhyVCASfHsEguOlqDjqdjrfeesv1/cSJE2UYl/ARjDu6S09JqJKiJCjJxEbRYKqfXtEQemLsrXv37gwcOBCAgoIClixZQmpqqsd75AlxmPDesTpMihJnu3ey2+3onYtcuElLSyMtLY2UlJTWDE9ohXdPSRBcF6SnJITodDouuOACn9dDfXx5qJCJjdrlnls6XWCf5XjH4rwRUwEVx9r54ZDzGRkZ3HvvvYwdO5aLL7440OGEPC3kQG1t3ylcipKYmBgWLVpEv3796NWrFx9++KFcF1qYFtp/XXzyIwjnlEhREkIMBgOjRo0CoGjfTuw2514l2m+IomZi4/Tp0/0e37lzJ/n5+SxatIiePXvyxBNP0LZt21aOMjy551agecdiq6oEnEVJtRC/Odu7dy+TJk1i0qRJTJgwwed4WloagE8Pimg8LeRAbW3fJQhuuppi4sSJPPTQQ/Tt25eXX36ZGTNmUFVVxZlnnsnzzz8f6PBCmhbaf128Y6ySOSVCM2T4VtCZP39+rcfcJzb27t2bl19+mWeeeYYXXnjB433Om7Ds7Gzpsg8T/npFgqGbvrEKCwu55557mDNnDkOHDg10OCKAfFbfCsF2v2nTJtfX//rXv1xfX3bZZaxduzYQIYlgIT0lIpBUVXXtNGp3++McDkM5Qp1MbAws99wyGo0oiqKZWHAN31JQFaW6yyR0c/7f//43ZWVlvP7667z++usADB48mIcffjjAkYU2LeSAT9sPwpsuEZy00P7r4h2jDN8SAWWxWFxP2yePHl7zHzcIGqI4tYKCAj777DNuvfVW4NQTG0GGrTQ399yaOXMmERERmolFqc5v9+FbwdBN31gPP/ywFCABoIUc8I5B5z2nRK51ooVoof3XxTtG50IQdhyrWgVDfmhzto5oOu23PdEAzomNe/fuBZCJjaKGvyErQXDxEaKpwnVJYCHqQ6XmgZXjC+1fF6QoCVUypyQkTJw4kZ07dxIdHe2a2Hjttdeyb98+Hn300UCHJzRA9egpqV59S27ORDjw2TxRrnVCuLgN7YXgyA8ZvhWiHBVy8DREUUMmNooGcZ9TUv2S5LwIB8E4Zl6I1uPdU6L9h1XSUxKqpKdEiLDgfmMWTBcfIZrMpyiRdi+EU00vuuJ8IYDR1E/Ae0o+/vhjli1b5vo+KyuLwYMH8+KLL7pe27x5M9OmTSM5ORmA3r17n3L5VIFX49N+QxRCNJLzwuO2GIxql5wXoc93SWBp90K42N2uDWpw5EfAi5Jx48Yxbtw4ANLT07n77ruZNm2ax3t27NjB5MmTueOOOwIQYfALgnYohGgkfzu6yxNjERZcc6dcd12BjEYITamZ6C49JY3y1FNP8dBDD9GpUyeP13fu3EllZSVr1qwhJSWFOXPmuHpNRA2dTkfv3r1RVdVzDe0gaIhCaJkzt5xfaykWuyu/FVdvSTA8ERPBRQs54B2DtbqdK3odqs0mCzyIFqOF9l8XnxjtnnNKguG6oJmiZNu2beTn5zNy5EifY/Hx8YwfP54rrriCpUuXMnXqVJYsWeLxHtnJGgwGA2PHjkW12zm56xe3I9pviEJomTO3tMA7Fota01PiIjdnoplpIQe8Y3AVJTo9qs0mD+BEi9FC+6+LT4yq5+pbwZAfmin3Pv74Y+68806/u2Q+//zzXHHFFQCMHz+e/fv3U1pa2tohBhFZkUSIsOG2JLDVZqt+SfX4V4hQpFbfdCnVT66lvQtRw325eMcX2n9YpYmixGKx8P333zN06FCfY1VVVbz55pser6mqisHg2cmTlpZGWlpa2PaSuPNdJjEwcQghWp7qNq7eYrECYLNaSE9P9/uQR4iQ4bzW6fXV32v/pkuIVqPK8K1GOXDgAN26dSM2NtbnWGRkJKtWraJHjx4MGTKElStXcuGFFxIdHR2ASLXNbDa7ViW77/prMFYXbqpUJUI0iXtuzZw5k4iICM3E4n7hqXkiBiaTKSDxidCkhRzwjsFZkCs6R1ESDDddIjhpof3XxSc/nBPdFaX6AqH9/NBET8mxY8d8ejhmzZpFWloaAAsWLGDx4sUMHz6ctWvX8swzzwQizOAVBA1RCNE4qp+iRHEcCFBEQrQS10R3vcf3QghkontjDRs2jGHDhnm8Nm/ePNfXffr0YcWKFa0dVujQfjsUQjSW25AVV1Eio7ZEOHBNdNe5vvdZfVKIcBWEmydqoqdEtCwZviVECLPXdNG7Z7rclolQp7qtvuX2YoCiEUJbnPlhr3khYLHUlxQl4SAIGqIQonFU1X3zRCHCiHP1Lb3brYxc74Rw8OopCYbhW1KUhIMgaIhCiEZSPftHnN8qUqKIEOevp0SVFbiEqM4NZy+668WAxVNfUpSEBe03RCFE46jeT8OqX5fhWyLk2WvmlNSsPCfXOyHc86Bmorv2C3ZNTHQXzUOn09GjRw/sNpvHRD/5Gy1E0zhzy/m1pmKxe+7oLkWJaAlayAHvGFw3WR7XO7ngieanhfZfF48Y3XMiiCa6S1ESQgwGA+PHj8daUU7J73vdjmi/IQqhZc7c0gLvWLxvwiTbRUvQQg74xOAcvqXocJThqqv3RIjmpIX2Xxf3GO1Wq+v1YOpF1Ga5J5rIe0d37TdEIUQjqbX1lEjei9DmKsgVxTVuPhiGqAjR4jyGbwVPT6IUJSHIp81puw0KIZqg1jklkvci1DlX31IUCKIhKkK0NPdNdb0OtHosDSHDt0KI2WzmpZdeAlXl7mGpGA2O/7yyT4kQTePKLWDatGlERERoJhbnnBInVUUmlIhmp4Uc8I7BvafESXpKREvQQvuvi3uMD//1r4CfpeKlKBGtyWKx+L6o8UYoRDDwm1sB4h6L64mY80Fx9esyfEs0Ny3kgEcMzvkjOh2qolTfgUm7Fy1DC+2/Lq4YaynOVdWOgt7vMS1ocFHyww8/8OWXX3Ls2DF0Oh3dunUjNTWV/v37t0R8ojnIH2khQpdXN72sviXChsfwreqXQuh6V1ZWxrhx41i8eDFdunTxOHbgwAGeeOIJSkpKOOecc3juueeIiYkJUKRCazyGb7nlh9bvB+s9p2Tfvn385S9/4eOPP6Zr166MHTuWm2++mS5duvDBBx/wl7/8hb1799Z9IiGEEM2mZriK7FMiwovnRPfqFh8iq29t376dcePGkZ6e7vf4o48+yrRp0/jss88466yzWLx4cStHKDRN9boSKMGxq3u9e0rWrFnDq6++SmJios+xW2+9lfz8fN544w2eeOKJZg1QNJ3WG6EQogns/ntKZIULEfI8lgR2vhYac0qWL1/OnDlzmD59us+x7OxsSktLGThwIABjx45lwoQJTJ06tbXDFBrlvO1zlSaKDlW1hU5PyYwZM0hMTPQ7pq6goIAOHTpIQaJVGm+EokZZWRnXX389GRkZPscOHDjAn//8Z6699loeeughKioqAhCh0Bq11iWBhQhtnpsnBseT4PqaP39+rcPic3Jy6NSpk+v7pKQkcnJyfN6XmppKamoq2dnZLRan0CqvPFCCY3W6Bi8JPGbMGH7//XfX9xs2bGDkyJHNGpRoJq5xhNpuhMJBuutFQ6mq6tNNL0WJCBtuPSWq4vlaKLPbfXuD3OfVCOE91zDkhm85/fWvf2XSpEmMHz+eXbt2kZ2dzTvvvNMSsYkGUhSFbt26YbdaUBTFrbsu0JGJ+pDueu1y5pbza83E4va6q6dE8l20AC3kgHcMqmv1LfeektAYvnUqycnJ5OXlub7Py8sjOTnZ531paWmAo8dENI0W2n9d3GN0v/GzWq1gdPaU2FFVVbO/Q4OLkj/96U9UVlYyffp0EhMTWbp0KaeffnpLxCYayGg0cscdd1B1soDy4+mOP9R27VfGwmH+/Pm1HmtIdz04ipiUlJTmDzJMOXNLC9xjUW021+s+w7ck70Uz0kIO+MTgZ/WtcKjKO3fuTHR0NFu3bmXgwIGsXLmSK6+8MtBhhTQttP+6uMdoLj5JFY5NdS0WC2ZUjEBWZhan9zgnkGGeUoOLkqlTp7Jv3z4++eQTDh8+zG233cbtt9/OnXfe2RLxicZw79J2vBDIaEQzkO564Y+/p8IyfEuEA4+HbW6rb6khsvqWPxMnTuShhx6ib9++vPzyy8yePZvS0lK6dOnCyy+/HOjwNMvZM6DlHoJm51WcOzfVNVdVBSaeempwUZKUlMT8+fOJiIigb9++DBo0iL///e9SlGiIzy63ofs3OmxId73wx3UDpig1Y4arj4XJpVeEK7ebrlBcfctp06ZNrq//9a9/ub4+55xz+OSTTwIRUtBRFCXsRg/Utqmu1tV7ovuOHTsAeOyxx4iIiHC93qlTJ958802P94jAMJvNvPjii7z69rtYrNagWW1B1M29ux6Q7vpW5sytF198EbPZrJ1YqiodL7pvHufnKyGaSgs54B5DlbPtQ0iuviWah6qq7P7Pj/z8xmfs+3xbo8+jhfZfl1PF6HpWHYC4GqLePSVpaWm8++67jBo1ioEDB7p2DjWZTGzZsoWVK1fSvXt3zj///AYH8cgjj7B7926ioqIAePDBB7n66qtdx3Nycpg2bRr5+fl07NiRV155hfbt2zf454QD92VizWYLBkCVm5OgJd312qGlJZidsah+9mmQnhLRUrSQA64YfIZv4fu6CHvHtu5n9/ofAdix6lvikhI47cKzG3UuLbT/unhfG1QUPB9QaTs/6l2UPPzwwxw8eJB33nmHxx57zPFhgwFVVbnyyit56KGH6NGjR6OC2LVrF8uXLychIcHv8blz53LjjTcyevRoPv30U+bNm8eCBQsa9bPCid21e462G6HwJN31ot6q5xopOsV1rZGiRIQD96GLihJeq2+J+lFVlb0bt3q8tnfjT40uSoKK95yS6n+1fl1o0D4lPXr04KmnnmLSpElceOGF9O3blwcffJB58+Y1uiA5efIkhYWFTJ8+neuvv57XXnvNo/vVYrGwZcsWRowYAcCoUaP46quv/G7iKDzVPDkKaBhCiBZSM3/MrafEle+S+CKEeWyc6P66tHvhUHQ8j5ITheiNes6/92oUnY7CozmUnCgMdGgtr5ZNdbWuwRPdH3/8caxWKzfffDN2u521a9dy6NAhZs+e3agACgoKuPTSS5k7dy6RkZHcd999JCcnM2bMGACKiopo06YNRqPREbDBQGxsLIWFhR5LpMpSqL5kbLkQoc45fMt3TonWn4gJ0SReQxedq28RwqtviYbJ2nkYgOQ+3Ylq14aknl3I2XOM7F1HaJucGODoWlZtc6u0fl1ocFGye/duNm7c6Pp+yJAhDB8+vNEBnH322bz66quu7ydMmMDatWtdRYm/pVABdLoGb0YfhtxuVMJpKTwhwoTq/Puo00H1liU1+5QEJCQhWoXPKpOu12X4lnC0j7wDmQAk9eyK1Wql07mnk7PnGHkHMjj3T/0CHGEL89rRPVgeUjf4zr5z586kp6e7vs/Ly2tSz8TOnTs9xtDb7Xb0er3r+8TERMrKyhw7UuLYmbK8vNxn/klaWhppaWnSS+LGo+lJl7YQoUeVnhIRprzaviqrTQo3dpud/ENZABRRgdlspuM5XQDI/z0zpPezAe+J7sFzXWhwT0llZSWjRo2iX79+6PV6tm/fTqdOnbjtttsA+OCDDxp0PovFwrPPPsvAgQOJiIhg2bJl3HTTTa7jRqORgQMHsn79ekaPHs369esZOHCgaziXqKEoCp07d8ZuMTs2Cgp0QEKECGduOb/WSiyuGzCdDmdXSbA8ERPBRQs54NH2qa2nRNq9gJNHc7BbbeiiDCixjm0s2p2ehCHSiLmiiqLMPNp1Tar3+bTQ/uviEaPXsWDJigYXJVOmTGnWAPr168f48eMZO3YsNpuNa665hhEjRjBr1iyGDBlCamoqc+bMYebMmbz99tvEx8fz0ksvNWsMocJoNDJx4kQqsjOozDuB2b1Zyh9qIRrNmVta4B5LVWE+ZqSnRLQ8LeSAewyW0hIqAcU1lNvZUyLDtwQUpJ8AIKpTvOvvo06vo/0ZKeTsO0bhkZwGFSVaaP91cY+xIus4NnyLEa1fFxpclAwcOLDZg7jrrru46667PF6bN2+e6+uUlBTef//9Zv+5IctrLKHja1XzjVEI0TCu8fNhuE9JWVkZ48aNY/HixXTp0iXQ4YhWpnqtvuXauVoewAkcK28BRHZo4/F6QteO5Ow75joeqrzzIFjSQmaLhyDnZokyp0SIEOdv88QwWBJ4+/btjBs3zmN+owgzPvOpZPUtUaMoo7ooaR/r8Xq70zs6jh/PbfWYWpXz4bTiNadE4/eCDe4pEdplsVh4/fXXUW1Wxl91ORgNqFT/qdZ2OxRC05y5BfDAAw8EdE6beyx3ja2ef6cLr+Fby5cvZ86cOUyfPj3QoYQNLeSAewz33DrO8aLi+WxVVt8SNouVkuwCwLcoSageslWUmY/dbq/3Sq5aaP91cY/x9pGeq+IGyy2gFCUhRFVViouLnd85KuQweGoqREtzz61ADw/xjKV6R/cwG741f/78Ux6XfauanxZywH/bl9W3hKfirAJUu4o+yoi+TYTHsdikBPQRBmxmK2U5RbRNqd9+JVpo/3XxuAe0e26e6KT164IM3woTWk0iIUTj1exT4r+nRPJehCrXkq46754SafPhrjgzH4CoDrE+K2XpdDriO7cHcPWmhCTV/7daL0qkpyTE1Qzfkj/UInxUnCxj/+fbsNtsnPOnPxDbsWYFlpDaSNTfnBLv46HyuzZAWloaUNNjIkKQz+aJ0lMiHEpOFAIQmVgzdMtqtbqGa7VNTqTwSA7F2YWE6hIZqteCR8GyVLwUJSFM201PiJZRVWZi04ufUFFYCsDxnw9y9ePjyS1xPBU744wzAhles/JegQi8Vt1T7SjSIS5CUG3Dt2ROiSg9cRKAyMSalbcsFgu7du3CYDBgjVKq31cYkPhaheq/b0Trj6jkahUu5OmRCAOqqrJnwxYqCktp074tbVMSMZdXsmPVZkwmEyaTKdAhNq/qISyKzxAWz+NChByfnpJq0ubDnmuSezvPSe4WiwWTyYSuejPFUB6+VXtPibZJURLCVNwapBQlIgyYisr5/esdAPT/y58YdM91ABz/5SCVBaWBDK1l+OkpAcUt70P7qfGmTZtkj5Jw5TN0UYZvhTtVVbFZrJTnlwBgiI/y+z7nsK7SnJPY7aH6N9Lz95I5JaLVKYpCx44dsZvNOJqe8//kj7QIXe5zRNJ/2I1qsxOdEk+Hc05Dr9dz2kVnk7n9d/J3HKfLVb0b9TOcueX8OpDcY3FdaLyXRXV+IU+NRTPRQg64x+DdU1KzeWKo3mSKuiiKwr5fdqGqKrpIA3aD/5vwiLbR6Ax6bBYbFQUlxHZMqNe5A93+6+IvP2oeUAUmpoaSoiSEGI1G7r//fkrTD2IpLfY8GCwtUogGUhSF9PR0KioqOPjNbwDEntOJ3bt3YzAYMHSNg+1QcjAH9YqejfoZztzSAvdYStMPYgEUnedFMlx6SkTr0UIOuMdQkXUcG+5tv6anJKQWsxANUpLtmCcSkRBTaxtQdApxndpRnJlPSXZhvYoSLbT/urjHWHxwDzbAd06Jtu8FZfhWSKqpkGu2KdF2QxSiKUwmE8WZBZiLKlD0OmK7t3eNH9Z3bIMuyoi90kLZ8dCa2Fgz0b2WnhIpSkSIcg1Jrm77qtQgAqgsLAPAmBB9yvc59ycpCdXJ7jKnRGiFv/kjatA0SSEax5RVBEBUSjy6iJpOYEWnENvdsS592dH8QITWclzj6mvpKZHhWyJUea2+5fFEWIrxsFVVWA44ekpOxVWUZIdmUeKa6O7sQKx+XdH4JUGGb4UQi8XCv/71L2zmKv58xaXgvpGp9JSIEFd5wjFkMTol3udYdJd2lOw7QemxxhUlztwCmDhxIkajsfGBNpF7LH++6o+OJ0veq2+5vpCbs3CXdzCD3z7djN1m47wbLqXz+Wc26jxayAH3GMYNu9rxop8hOqpd9e48FGGi6qSjKDHWVZR0chQl9V0WWAvtvy7uMY69fBB6t2PBcgcoRUkIUVWVvLw853dew7cCE5MQrUFVVUzZpyhKOieAAuaTFZQXlNCmfdsGn9+ZW4Feyc4jz2vrKaneNTXQsYrAKs8v5ptFa7CZrQB898b/kTr9zyR2T27wubSQAx5t3+7sKfFTfUi7D0t2ux1zUQUAEXUM34pz9pTknKzXHCQttP+6+ItRlTklQotk+JYIZVWFZdirrCgGHZEdY32O6yMNRCU5CpETe462dngtRnUuZ1lbT0nILncp6mP78q+xma20PyOZlPO6o9rt/LZys2ZvqhrCNTRRp2C326msqqpZiUvafViqKCxFtdlR9AqGWP/LAYNjd/c2HdqCApaKKipLylsxylZS3Usuc0qEhiiyNKgIC+UZjh18ozq19dlI0Cm6SwIAOSFUlLgmM6qOC63vYcn7cFWSXUjWjsMoOoUBtw3lD+NT0Rn05B3MJO9ARqDDa7rqtp2dfYJdu3ZRXl6OXQ22W7Dabdy4keHDhzN06FBee+01n+ObN2/m4osvZuTIkYwcOZKZM2cGIEptcQ7FMsZH+6xI6M5isbB3/z6McdHVnzvZKvG1Ju/5hMGyT4kUJSFMBexSjAQNuQg1XnlWdVGS7Dt0yym6cwIAeb9nhszNutVqAeDQ4UNYLBbX687nxPLEODypqsqhzTsBSOl7Bm1TEolJjKP7Jb0ASP9hTyDDaxbOHK4ym11t35nWwZ7feXl5vPDCC3z44Yf85z//Ydu2bWzevNnjPTt27GDy5MmsXbuWtWvXMn/+/ABFqx0l1cWFMf7U80nAUZgY4x1FSUlO6BUlwdpTInNKwkawNMnw5LwIrVy5kri4OCZOnMjmzZu5/PLLXe9xXoTuuOOOwAWqQaqqUpHpuKj4m0/iFNUxDkWvo6rURFluEXGd2rVWiC2n+ubLbLHi9xlYkN+cicZR7SrpP+wGwNgtnj17HEVIZLd2sBkyt/+O5ZarMEZFnOo0GudbcIfKsMXvvvuOQYMGkZjomPcwatQoNmzY4HE92LlzJ5WVlaxZs4aUlBTmzJlDcnLD5wqFktIc5x4lp55P4mRMiIbj9Z/sHoxcc0qcK2gHLpR6kZ6SEKYqsk9JsHC/CBmNRtdFyN3OnTv56quvGDVqFJMnT+bEiRMBilZbSnNOYq0wo+gVIjvG1fo+Ra8jupNjXkn+oazWCq9FOS8w3tktmyeGt4LDWVhNZnSRBvQd22AymTCZTOjaRRHbMR5rlSXohzHWTOR1e63mYGuH06xyc3NJSkpyfZ+UlEROTo7He+Lj47nrrrtYs2YNl19+OVOnTvU5T2pqKqmpqWRnZ7d4zFrgHIZV18pbTs5lg0N2rxI3NcO3tJ0bUpSEEEVRiI+PJy4mBu96ONi7s0OdXIQaL+9gJgCRSW3RGU79Jy0mxdE7kv97w4oSZ27Fx8cHfKdoj1iq87y2oiTYnxiLxsn89RAAMacneoytVxTFtSRw9q4jDTqnFnLAPQZn2/ZXlAT79c7uJ2+9/zd//vnnueKKKwAYP348+/fvp7S0tFXi06rS6mFYEfEN6CmhfnNKtND+6+KRH17XhmDJCE0M33rvvfdYuXIlAH379mXu3LlERNR0K2/evJlp06a5uiZ79+4t4yf9MBqN/O1vf6No3w7sZjMQPA0x3NX3IuQ0fvx4FixYQGlpKXFxtfcOhANnURKdXPcyv206J5D/c8N7Spy5pQXOWOw2K0W7fwV8B7KEyth60TDOpU2zdqYD0KZbe5/3pJx3BgfStpO9K71eS6E6aSEH3GMo2rcTO/4GcRH0+/MkJyezdetW1/e5ubkeQ7Oqqqp4//33uffee12vqaqKweB5S5eWlgY4HlaFOnN5JZUljuWAG9pTUlFYirXKgiGy9r1HtND+6+K6NljMFO3d4XFMJrrX044dO1i1ahXLly9n/fr1WK1Wli5d6vMemdDVAKrHP9XfyM2JliUnJ7vtMeP/IvTmm296fKa2i1BaWhopKSktG7BGqKpK3kHHSkJRp5hP4hSTkgA4nqg5L2BBy/mk2E9qy+aJ4UlRFPb/toey3CJQIPq0BJ/3dDi7M4ZII5UlFRQdz/M5HjT8Dd8KkWL8kksu4ccffyQ/Px+LxcK6desYPHiw63hkZCSrVq1i06ZNAKxcuZILL7yQ6Oj69RCEImcviaFNJDqjvo53O+ijjOijjB6fDwWu9q8oNctkVx9T0HZ+BLwoadu2LbNnzyYmJgZFUejZsydZWZ5PMWUsfcPUNDglZLqzQ51chBqnPL8E08kyFJ1CVKe6e0r0UUbiOzueHgf7vBLVY/iK13BN13sk78ONc2hiZIc49BG+gyH0RgNJPbsCDR/CpQWuuSR+ivJQmVPSqVMnHn30Ue68805GjBjBueeey9VXX82sWbNcvR8LFixg8eLFDB8+nLVr1/LMM88EOOrAcs4LiUxs06DPOd8fUvNK7M5NdWtu8T0yQsPDegM+fKt79+50794dgIKCApYsWeLTExIfH8/48eO54oorWLp0KVOnTmXJkiUe73F2T2ZnZ4fNU2JvFouF999/H6upgpsuH4Rav4cFQgPcL0Jms5khQ4a4LkJDhgwhNTWVBQsW8OSTT/Lyyy/Tvn17j+Fc4crZSxLdKR6doX4NvsPZnSnOKiD/UBZdLjq7Xp9x5hbAHXfcgdFYezd/S3PGotrtjBp4Ieh8f2/pKQlf5RmOm6vozrX3HCb36U7Wb4fJ2XuU3tcNrNd5tZADzhiqqqoYc9kAIvR6/3NKNHzTVV/Dhg1j2LBhHq/NmzfP9XWfPn1YsWJFa4elWc6ejsh2DStKjAkxkFXkmldS25BGLbT/unhfG4xuIyk8exTtKGjzBjHgRYlTRkYG9957L2PHjuXiiy/2OCZj6etHVVVXL5PPyiRB/uQoHMhFqOGc80liOtd/ed8OZ3Xm0Dc7G9RT4i+3AqU+sUgPaXhSVdWtKEnwOW61WrHb7XTqeToABYez6xxL737uQOeAewzYVdCH5upbouGcRUVDixJDXCQAGQeO0CY9mTPOOMPv+7TQ/uviHaOi6NySQql+Tds96AEfvgWwd+9exo0bxy233MLkyZM9jslY+sbzv5+nEKHD2VPS5rSGFSUAJ4/mYjVb6ni39vnLbFl9KzyV55dgKa2EWoYzWiwWdu3aRV5ZITGJcdhtdvJ/zwxApE2neG6/4PG1Vm8aRcspqd6jpME9Je0ck90rC0oxmUzNHldDubfdJrdjxf+wXi33oAe8KCksLOSee+5h9uzZTJgwwee4jKVvPNVtnLn8kRahpqKwlPL8EhSdQoyfp8K1iWnfluiEWFS7ncL04J+f5i+1nQ/CQmEYi6i/3APHAcdGobVN9rVYLFRWVrp6S3L2HW+1+FqC3yubhm+6RPOzWW2U5RYDENHAoiSievd3c7FJEz0IiqKQnp5Oenp6k5ceVnSet/jOrNDydSHgw7f+/e9/U1ZWxuuvv87rr78OwODBg8nPz5ex9M1AurNFqHIO3WrXNckxoddUv14PRVHocHZnjm87QN7vmSSd27Ulw2xx/i4voTS2XtRfbnWBcar5JE5JPbuS/v1ucvYda+mwWpT7lS0YbrpE8yvNOYlqt2OMisAYG4mtsrLenzXERqLodag2O+aSwPeUAM3XY6PocM+QYOgpCXhR8vDDD/Pwww+f8j0ylr5xVNyWSJQ/0iLEOIdudehxWoM/2/Hs0xxFycHgHLrizt/jhpqbM1trhiICSFVVcvc7i5KEOt/vLMaLMvKoKjMRGRt8ow8cD7bdRwRU/yvXu7BSnJEPQHyXDg3uXVB0Csb4aMyF5VSdLG+J8ALG0VNScw0IhlUZAz58S7Qs6SkRocpZUCSd06XBn3UWMgWHs7HbgvvGXXpKBNTsvaPodUQm1b08dnR8G8fy2CquYibYeF/VXK1d2n1YKcp07LcTf1qHRn0+onpeSVVBWbPFpAXew7dq5lxpNz+kKAkxMdHRREVEOL5RlKBohEI0lKmozLEEpAKJZyZjtVrr9Tnn6kPxKe0xxkRiM1s5eax+G8jFxMQQE1O/nYJbWkxMDNFRUUAtPSWuJ8bBXXCJ+nMO3YpJSUBnqN+l3blfSX3nlWghB2JiYoisvsZ5t3253oWn4kxHT0lCY4uS9o55KKa8klO+Twvtvy7u1wafosSZIBou2gM+fEs0n4iICB752xSK9+2sHkvovgqP9JSI0JFXvWJQVPs4Dhz+HYulfvNJnKsPxcXF0eGszmTvTCf/90zan5F8ys9FRETw6KOPNjnu5uCMpSI7g8q8E1T4qTtclxybdi8+onnlHqheia5LYr0/06nn6Rzc9Cu59ZhXooUccMZwcPdOjLYqLF6XNekhDE/uw7dyK4sa/PnI9rEAVOaV1voeLbT/ujhjNOVmYzqRCbX2lGj3flCKklCjOnfydIyrlCdHIhQ5h5tEJretd0HiZLFYMJlMdDz7NLJ3ppP3eybnXv2Hlgiz2bhv6OX82tkLcurhW7ZaNwMToUO118wnie1ad1Hi7DHs2OM0FJ1CWV4x5QUltGlf97CvQHBvw3a73TXksrbhW1KUhI+qMhOmYsdckPjOHcg9XNTgc0RW95SYiyqwVJoxRkU0Z4h1ysspYOv3v7B/7+8UFhRRUVFOu8R4iq8tp/+gizAaDX7/jvu7LoBb+1cUj1EEwbBUvBQlIcZVAes8ixKZUyJCSe6+6p3c67HKUG069HDsV5L/exaqXUXRaffG3blMJODa3Eut7gXxd3nxeE1VfdarF6GlKDMPc3klhkgj0UltqTRXnfL9zh5Dg8FATHIC5Vknydl3nDMv69NKETeMs/2bTCasVitGqxUMvgMAaoanyLDFcOEcutWmQ3yjiwl9dAT6NhHYys0UZeTR8eyGL57SGL/8tIN3Xl/Cd19vxe6nUFjx0f8R1zaWoSOuZMr0e0lo53m983ddAFxFR0FhIRa37sRgKNqlKAkhFouFj5YuxWaqYNQVl3nsdqvl7johGqK8sISyvCJQFKJTGl+UJHbrhCEqAnN5JSeP55LYrVOt77VYLCxZsgSAW2+9FaOx7h2w66uqsgqdTocxwlhrr4bdbqe01DG0oKqqio8//hhbpYnrL+6Hiu9+FB4bytltPmOLRWjJ2esYftXxnC4o+vr9t7ZYLFgsFmK6tKM86yS5+46dsihpqRyoqqziv//5ki3f/UJeTj6xcbFc0K8PI0ZfTfuONb0+JpOJsrIytmzZgl6BMZcNxK54tn0ZvhV+Th53zAls7HwSp8j2sVSUF1J03H9R0pzt/0RWLvPnLOTLz791vda777lc+Ic+dEhqT2ZmFhnHsti74yAlxWWsXPof/rv+K+6bcju33jUGvb6m3TsLdbvdjs1mY8mSJdiqKrl+4EU4+kjcVqdz/qvhol2KkhCiqirHjjueIKuuLr3qg/JHWoSIvP3VvSSd2qKLaPyfMJ1eT9K5Xcj67TA5e4+esihRVZWjR4+6vm6KkuJS1q/6L5s3/cjOX/dSWuJY8aVdYjzn9+vD+X/oxR8uPo+o6CjXZ6xWq2uY2u7duz1i8Z/ZCnZUdIBqs4Gh+YoooT3OoqRTz9Op35IPNWK7JJK39TA5+46fcqhfc+aA8xzrV/2Xf8x/k/y8Qo9jX2z8mtdeeptb7xrD5L/dQWRUpOszhYWFrq9Vr1CD4UmwaF6F6dkAtOte+9/v+ohs34aKY46ixJ/mav8b16XxzKwFlJaUodfrGfXnYdxx7zi6nVGziuSePXswmUzYbDZ+27aHtcs/51h6Ji8980/SPtvMvFcep8vpnV3vd/Z8Ah4x+uRHECyAIkVJqPKZUyI9JSI0OFcKim3AhN7aJPfqRtZvhzmx5xi9rh3Y5POdSnlZBe8uXsqHby+nstJ3eM3JwmK+/uJ7vv7ie9rExnD92KsZcu1lGI2ef6a959DUdvulogCqoygRIctmsZJfvfBDp16nk1mU06DPRycnoI8wUFVaQXFWQZOfONdHZWUVs6fO57//9yUAnVI6MnLMMGLiIikqLGHbjzvY9ete3l28lM1f/sjCf83zex7vti89JeGnIP0EQJ2LldQlonqye+HRhuVPfVnMFubNfoVVy/4DwHkX9uLpF2dw1jnda/2MXq+n38V9ueSK/uzbmc7Lz/yT7dt2csuISTy/6O9cdmXNNctisfisQllrfmh4ARQpSkKU90R3Le/gKUR92W12snc5xtC2Ob19k8+X3KcbAAWHslp0guPmTT8yZ/rzrifCPXqeyaixwxhwyUV07pLseAqXnsGP3/7MiiVrycnOY9l7a/nqv99zz0PjOeucbrWeu7bHDXZAD1KUhLi8g5nYLDai49sQ2ykBa37DNgTVGXR0PPs0Tuw5Su6+Yy1elBQXlfDQ3Y+zfdtODAY9kx++kzsm3YLeoHc97Z0y/V6+SfuBJ2e8yMF9h7ntxgd49MnJJHb0HK7pPadE9ikJL5Ul5VQUljqWhj9FT3d9RHVyLPJQnJWP2VRFRHRkc4QIQNHJYh657+9s+/FXdDodkx66jYkPTvB44OTspbTb7T7FhU6nY8y467n08gFMf3AuO7bv4f7bp/PQ9In0v+y8Wn+u97PoYNhUVwYahyrvnhJZEliEgILD2ZjLK4mIiaRNPXatro1zDG5sxwTadIjHbrO3yAZyFrOFZ2f/gwfufIz8vEJO734ar7z5NJ9+9i4T7vkzPfv0oG18HPEJbTn/ot5M+usEFvxrDnfe/2cS2rXlRFYe8x5/lXXLP/c7ERLAVktqq9Vjie22hg7oEcHkxO4jAER2jmf37t0NXo0OGr5fSWOVFJdyz7iH2b5tJ3FtY3lr6StMfHACGZkZ7Nq1yzXPZffu3XTvcRorNr7DOb3OIj+vkLnTF3AsPcPjfD77lDiHp0ghHhacvSRtkxMxNrGIMMREYGwbDSoUHs5ujvAASD90jL+Mmsy2H3+lTWwMi96dz/0P3+nTA+6ctO7MA386d0nm3U8Wcs31g1FVlYXPv8W7ry/DXkvPRzCuTidFSYhSFK//tNJTIkJA5m+HAEjpe0aTJm87x+Du2bOHNt0cPS4Z239vlhidCguKmHjrIyz7YDUAE+4ey8r/vkfqtVeccolevV7PFX8axDMLpzPoin6odpXVyz7j9Rfex2Sq9Hl/bZltdz6YkKIkZKmq6mq3kafFN6ogAcdcFIC8Axk1y+1W3+E319Bfk6mSv941k/17fqd9x0Te/3QR/S++oPqYySN257LdSZ068O4nC7loQF8qyk28+tx7Huf0bvvOUsS5FLYIbbnV8ws79mie1bKcD7ryDzVPUfLLTzuYMPp+jh3JpHOXZD5Y9TqXXzWo1vd754E/EZER3Hn/zdx692gUReHL/37Pohfeo6rK7PNen+FbzpTQcNEuRUmokjklIsSoqkrWb4cB6Hz+mU0+n/PGp80ZjqIk67dD2CxWj1xpSN64v/fAvkOMv+Feftm6g9i4Nrz+3vM8+vcHXRN2a/uce9d9m9gY7v3bX7jrgVswGPT8snUXz8xYSO6JfI/P13Z5cT0V0/AFSDRN0fE8KgpLUQw6Yrq2a/R5Erp0JKJNFNYqi+uGTFEUsrOzm2WPG4vZwtT7/u7qIXnzw5focW79crhtfBz/fP8FzuzRjbLSco9jtT0JBmn34cDZu510btdmOV9MZ0cOOTcibYr/bfiaSbdOpaS4lL4X9WbJ2jfq3ebr40/DL+f+abdhMBr49afdvDhnsU9+eGeA9JSIVmc0GDDo9a6eEplTIkJFcWY+ZXlF6Ax6knvXPseioWJSEoiOb4PFZCZn7zFXN3p6errHDZnRaDzlMpDOzy1fsprbbnyArIwTdO12Gh+t/ieXD6n96Zjzc3v27PHbdX956kBmPP0ACe3aknU8h+dnv46iKBj0+uox9bWsllT9umqVnpJQlfGro5ckrlsHdAbfpaHrw2q1oqK6Cv1jP+13Hauq8lyQoa4cAN9C3m6388TU+Xz71RaioiJ57d3nOKfXWQ2KsU1sDDOefpCU05KwWmzYrHaKist85pQ4Vp2rjkOKkpBWWVLu2qOk4zld6nh3/Tg3Hi04nIW5oqZX2tmm69v+l7z3KdPun4O5ysxVQ//I2x+/QvsOvg8Nansg5c053Nj7ff0vuYBHn7yPNrHRHDpwlBfmLEan02GoXjLYZx8f578anlMiE91DSEREBA9PuhtTdgYRUZFQVlkzxlbmlIggd+THvYBj6FZTxw+7s9lsnHbR2fz+1W8c//kAyed1p7S0FIOh5s9jREQEjz/++CnPo6oqyz9cy9J316CqKgMGXcjLbzxFfILvLtneO1SXlpaestv+rHO7M+fFR1j0/LscPniMVe99yZR7R6MzGKhtDVi7ooAqc0pClaqqZPxyEIC2ZyU1+jzOoYwRp8fDD5Dx8wEuunkweq8ipz45AJ4bHaqqykf/WsnGdWkYDHpefuMpLhrQt1Fxto2P5ZEnJjHv8UUU5J1k97cHmfnUZJSYGI/31aw6ZwWa7++E0BbnMtjxp3UgKi6mjnfXT0R8DHGd2lGac5LcfcextHNcA84444x6tX+73c4r89/k328tA+DPfxnJzKemeOwr4s57U9DargHum516v++cXmfy+LMPseDpt8jOyGXjJ9/z7Kw7MRoM2L1O5xreqOGCXXpKQoyrsel0jura+bqGK2Mh6mKz2ji2dR8A3Qf1atZzWywWLImOi8+xbQf4bdv2Bo/Nr6qsYtYjz7LkndWoqsqQay9jyqx7OFlU6Hf4S129I/4kJLblsafuZ9AV/RwXv8UreXXxSqxW/7ltd050b+Q8A6FtBenZlJ44id5oIO6Mjk06l8Viwdgxlqj4Npgrqjix60iTzmcymTCZTCz791r+szoNRVF4ZsHjpxxP783f0+F27ROYNuc+EuJjOXg4kxeefRez11h66SkJDxm/OHoJT7uwYb1udUnu0x2ArB2HXe24PirKK5h2/5OuguSWO0Yy65mHay1InJw/o65rgHO4sb/3de7SiSfmT6H7WV0oK61g6hNv8PmXP/vMKXE+m7ZbLZod0i9FSYhxFh/FJY4nr+5LJGq1EQpRl4yfD1JZUkFU2xjXRaM56ROjiUhsg2qzU7jH/5Kq/vJHVVXycgq465a/8X+rPken03HrPaP5y8QbsVjMlJaW+l01y9k7Up+LkbuIyAgmTbmV8X+5DoD1G79nwdNv+YwlBrA5h29ZfCdAiuCX/u1uALr+oQf6yKZvjqnoFE4fcK7j3N/vbvL5vtiwmbWf/BeAGU8+xHUj/9Sgz7svRuFeuCd37sDzT06kTUwU+/YcZvHLH2JzK0BcwxalhzBkWassnNhzBIDOF9TM0zjVEKj66nKRo8jJ/PUQ9loe+IDn9eDYkQxuHXU/X2z8GoPRwP1Tb2fETX+q856rOeJ1Skhsy4ynH6D/oL5YrDaeXbCUFR/+xyM3XF+pqmaXzZbhW81s4fNv8dn6TSxd9wbtEhNa9WdbrVZWbvwcu8XM5Zc4nkh5TvyzosjOziLIqKrKgbRfADh78IU+w0qag6IotO2dQv63v1O8M5P43ikex61WK8uXL6eiooKLL74YvV5PdHQ0edknmTnlGfJyC2gbH8dfH7uLs3vWzHdx73b3Pl9jV0qy2+106BLHnZOH88m/v2DvzoM8PWMhD06/g67danb5tTmHh1mlpyTUmIrLOVrdc3jmH/uSay5qlvOe+cfzOPDFL2TtPExpzkmPY84cAPjzn//s06bdbVidxpK3HavO3Th+GDffNgqoGbboPXyxthsz5xLBTjabjZ9/3ka0As/MuoPpc9/m1227efOVJUyaMh6D0VAzbFHafcg6tu0ANouNiPgYjp88wYmyAqBpf1edOpx1GjGJcVQUllJ6OI/4c5Jd5/Zu/+np6Xz/9U8sXvABJcWldExqz4I3nsIYrcNkMvn92++uOeJ1ZzAaGHhlb7r36MCqpV+zYfUmDu5N596H/0L7ju1QUbCjogPsVmudvTiBID0lzeydfy4h83g2n3y4ptV/tt1uJz0jk6M5edhdFbpM/BPBLeOX3zl5LBdDpJGzrmjcePT6iOvRCX1MBNbyKkr2n/A4ZrfbOXjwIJmZmVRUVFBaUsr7byxj4vhHyMst4Kwe3Vm67g36XtTT57zObnf3/2vKhUhVVTJz8yipMjH3mcm079iO3Ox8nnr0Ff677quaIS+uJ8Y2Gb4ZYg588Qt2q432Z6bQ/qyUuj9QT22TE0npewaosPeznzyOOXPg4MGDte6Zo6oqi//xPh+89SkAw0ZdxYib/sSuXbs8Fo5ozPBF5/lzc/M4mpNH715ncP/U29Ab9Pz0/a+8+ty7VFVWuXoI7dJDGJJUVeXQ178B0O6807Barc3yd9VJ0Sl0G+j4O16w47hrGKF3+y8qKuHFua/z/JzXKSku5fyLevPx/71F34t6u4psf3/7m+s64I+qquTk5aEaFaY/PI7omCgO7ktnztSX+OHrn1FV1W0BFG0W7VKUtBB/a0Y3laqqqKpKeUEJmb8eIvO3Q5iKyvy+1+62Io90Z4tgZak0s2PVZgDO+VM/ImOjW+xn6Qw62l3oWFqycNtRrBVVft93YO9h5kx9mdXLPkNVVW68ZThL1r3B6d2bZwWYhjitawp/f/5vXPCH3litNpa9v47nZr/O0cMZqIBS/STMVuX/dxHBpzS3iINf/gpAr2sHoKpqswwBcd589b5uIABHftxDSWaB63hdS2WXl1fw6ANPsvgVx14iI2++hrETRqAoChaLxWMoY2OHL7qzARcOOI8pM+4iIjKCndv38czMVzmelef4GTKXKiTl7DnKyWO56Ax62vVunv1JvJ115QXo9Doqsk5ScjyfXbt2sW/fPtfxNxe9y4grxvHtlz+h6BRG/vkaps+dTH5BboOK7JZ08SXn8+RLUznj7K6Ul5l4a+ESXpr7BunHcwHt9iTK8K1mVNvTo+ZSdDyP31ZtJtdr192U87rT5/pLiE3xv069HdCDa1MsIYKBqqps/+QrygtKiEmM49yr/9DiP7NtrxRK9p/AXFBOxv9207ffhSg6z4nqC55+C5vVTnxCHLOfncrQ4Ve1eFyeam4IzSq0TYhjyuN38/X/fmTZ+2s5uDeduY++wuChl3DvX4aRnBCDrdKEIbp5VqgRgWO32dn24f+wW2106n06lbE0202Q+1DD+HOSKT5wgiMbt9Nn4AXojQaPBRu8F2/YtuU35jz6PMePZqI36Jkw8UauvPqSWs/fHMNWLCqgQN9+vZj+5H28+vx7ZBzN5uEpLzH5zhGMGp3apPML7bFZbfy26lsAzh58AYboCCz1nIjeEDHtYuk2qBfp3+2m4IfDRCW1xezW8/buPz/BZrWTfFoSdz1wMz16noHFasGioRt9qwpJye15fN5f2bj2S9Z/+j/27DjIpAef509X9mPiA7dy7oCWv6Y2lCZ6SjZu3Mjw4cMZOnQor732ms/xnJwcJkyYwLBhw7jtttsoKCjwc5bA8zfZtDlUllawbUka/5u/lNx9x1EUhYSuHUno0gEUyN51hC/mf8xPH37h+ozd7aLhWgZO9ivQtFDJg+agqio7137PkR/2gAIDbx+KMSrC9YS2OScIulN0CklXnoOi11F2NJ8f3v8vG9elMWXirJr3KApXXXMpL781JwAFieeTJOclUFEUBg+9hHmvPsbFl1+Eqqp8+d/vGX/735nz3L/ZvOkHLJbQyf+6ciUUqXaVX5ZtIu9gJoaoCPrdfBWVlZXN+lTWOdwkYUA3dFFGKvPL2PLuZ7U+0Dp04AiPPvAkd/35IY4fzSSxQzumPznZpyDxPn+zFFFunTVnnduduS9N5ZzeZ2IyVbLgn59y7wPz+WbTD0G7wItcDzypqspvn35DcWY+EW2i6DVsYIv+vPNuuBRdhIGso9l8/MoyZj/8kutYdHQUt9xxA08vmEaPnme0aByN5cxYg9HA9WOu5pl/PMaFA/pgt6t8/uXPjB3zCBPHP8KGtV9QUV4R0FjdBbynJC8vjxdeeIGVK1cSFxfHxIkT2bx5M5dffrnrPXPnzuXGG29k9OjRfPrpp8ybN48FCxYEMGr/SoprhlJVmpo+XMJiquLQ5p3s2bAVa6WjSj99wLn0HXUZ0e1iOXToEGV5ReRuOUzx/myOb9sPjjlZVFRaoHpCsMWuEqUDW1VlbT9KBFgo5UFTFWcVsGP1t2TvTAfgopuvcu3Y6xyLXte+Ho1lsVjJKSnmQEQ5v323g/T/rafKakVv0DHqzMEAzH7+b5ze/TSio1tuKNmpxHg8SvJ8Yt2+Qzvue3gCqcP+yMbVX7L9p118/d0Ovv5uBwlPv8kfrxrEgEEXcmH/vnTt1vmUkzC1qj65EmpKc07y64qvyd51xFGk33Y1cZ3aQUF2i/w8Q0wEnYacy4nPdpOx/Xe+emUVfW+6zHV89Sf/4bP1X/Lzlt9cr914y3BGjElFaaVHnRavWiMhsS3T507mm/9t4dMP17P3wDEevHMGZ5x1OsNGpnJl6qX06HlmULR5uR54slSa+XXFN6R/twtwPKSKbBPV7D/HbrNz5PBxDuz9nV+27mDTD99wItexQaPeUNOwn1k43e/+U4FXkxRVXvmRlNyeKTPvJv9YNquX/ocfftrLlu9+Zst3P2OMMNL/4gsYcMlFnHdBT3r3PZe28XGtHLtDwLPzu+++Y9CgQSQmOnbSHDVqFBs2bHAln8ViYcuWLSxcuNB1/Nlnn3Wsq17HzpreLGYz+7bu8hxmpapuu567j5kF9//AHg9b3I+pNec4eqxmKdH0vYfY9vn3qO7vQ605j9v5XD/LrmKtsmAqduxUmn84B3v1082o9rF06t+d9v3OIOtEFhnbMmt22z0rjsgEHQW/1Qzr2vLm/4hJiiciMYbMuEjio/QYY7KITspwdL0rSvW/oNSyI3Tzaenzt76krp1I7t5841lbMw8qTSb2/bQLVLd2rdrddnt1fOV8wuj+vas9O19zvk91tG0VUN3yy/Ga6vzC9XlVdZ7X0earyqswFZZSlFVAaU4RoKLodZw9+HxKjBZ+/m4bKo45VRnHj2O12bxysiYexfWS6n4IUDFXWaisrKLSVEVlZRXl5SZOFpZQWFBMYX4R2Vm52GyewzDjIqO44PSaFbXiS21Upudhj4kkR4lyrGCiKKioVOUUolptUJ1bPpqYCnrAoNY8tTYXm7D7WUGlW3In/vbwHWDXsfy9T/jup10UnSzh/1Z9zv+t+hwAo9FA166d6dw5icTEBBITE4iPjyM6OpLIyEiioyOJiorEYDCg0+nQ6ZTqf2u+Pq3H6XTqkty0X6qB6sqV+qooLePA9r2eeVDdxhxfVrdp599nt/bkaltUt2tn4wfUU+USKqqrebn/LPf2qqLa7FSVmTCVVFCUWUBxZj6q3THv6dw/XUROZQk53/9MZmYmZnOV6xzOn1GTAzW/b21LWuPnvRaLhYpyE8WxVWTtOMrJXdspWvYJ/a53LDTx3JOvYrPaURSFyy7tx+2338jZZ51ORmYWZnOV4+c7e+vd23xT279bj01xkQlF7/tgInXwIK4ZcB6frvofn325jfRDx/jngvf454L3iIqKpEeP7qSkJJGc3JH27ROIiYl2/Z/RaHC1cb1e59HevYesuX9/es/utEtq37Rfzk1rXg9KThZxeOdBt3aJ63qguu5v3K4FHm3VXv0vrr/v7m3c/biqeuWQs/273U/Z7TXXEpvZQmVpBaU5RRQcPoGlyvHf+uwr+5JRnM/xr3PJyMxw7dOkqtXbZrrlsvN1Z+yVlWYqK6uoqjRjMlViMlVRVFhMYWExhQXF5Oee9JkPrFMUzuiQRN8uXWtePFZEVbEFXYQenUGHotN5tG3F+f8U7xeb7lSniXBb0KSsyITBT3506pjErAfHkZN3kq+2/84Xm34kKyuXHzZv44fN21zvS0hoS0pKR5KTHdeHuLg2xMbGEBcXS3R0FEajAaNRj8FgwGg0YDAYOLdfTxKTOjTp9wt4UZKbm0tSUs1utElJSeTk5Li+Lyoqok2bNq5EMxgMxMbGUlhYSKdOnVzvS011jB/NyMhAr9e7vvf4Wdl5mFt6CEOE458vf/6cL3/+vHnPveHUhxUFYmIdTw8q3HZzF81PUaBDh0Qi/ewsnp2d3eCl9lozD3Iyc7EEy/yin1a07s/Tg86oYDTqiTAaiYgwgK2SX/IKMK91/MHeaFvfujH5YdY7LvYbV9UzFh0Y29qx2mxY7TZsdjtVKuzNyGZvRuPjUBRI6tQBY4TvjVBj8qA+6soVqF8eZGfkYNPoWv2ntHV5QH6sosDxtYcBsBjNGCL1RBgM/LL7G36Z/k2rxdGgth8JBoMVc3W7r7LDtv0ZsL95Y1IUSO6c5NPeG5sDrXU9UFWV7MwcRzEQLLavatHTK5FgNDhuuCMjjUREGCmwZfC/o+lUHXNcN7VwDahNg68NQGS86nFtsNlVciryyDl0CA7V/2frFIWULp18CviG5EHAixJ/k8Pdf6HaJo/rdP77iBVFqbV71oYVvRFSUppvCUWtqqx0dOuH+u+ana2939NgMBAREdGgz7RmHth1NvQ6bf1v1lKas320afIZmo93LKGSB/VRV654v17rcB29Hb1eW/+btYQWaRv+11RpFafKQ63lQWNzoLWuB4qioOhVyYN6MMZEEwxLhbTB8btWop3/pg3Jg4AXJcnJyWzdutX1fW5uLsnJNcMBEhMTKSsrw2q1ulbtKC8vJyEhweM8aWlpdf4s51OC+rw32IXL7xoqv6fkQcsIl981XH5PqDtXQPLAXbj8nhA6v6tcD5pfuPyeENy/a8BX37rkkkv48ccfyc/Px2KxsG7dOgYPHuw6bjQaGThwIOvXO7qi1q9fz8CBAxs8blIILZM8EKJ+6soVIYKdXA9EuFJUDayXt3HjRv75z39iNpsZMmQIjz32GLNmzWLIkCGkpqaSnZ3NzJkzycvLIz4+npdeeonOnTsHOmwhmpXkgRD14y9XhAglcj0Q4UgTRYkQQgghhBAifAV8+FZrO378OLfffjs33HADo0eP5ocffgh0SM0uXDYWe++99xgxYgQjRoxg5syZmM3muj8kAMmDUCJ50HiSB6FD8qDxJA9CR9DngRpmJk6cqK5evVpVVVX9/fff1UsuuUS1Wq2BDaoZ5ebmqoMHD1YLCgpUs9ms3n777eo333wT6LCa3W+//aaOGDFCLS8vV+12uzpt2jT1vffeC3RYQUPyIDRIHjSN5EFokDxoGsmD0BAKeRB2PSU33HAD11xzDQDdu3fHbDZTUVER4Kiaj/umS0aj0bXpUqhp27Yts2fPJiYmBkVR6NmzJ1lZWYEOK2hIHoQGyYOmkTwIDZIHTSN5EBpCIQ/CrigZMWIE0dHRALzzzjv06tWLuLi4AEfVfOqzsVgo6N69OwMHDgSgoKCAJUuW+N0gTfgneRAaJA+aRvIgNEgeNI3kQWgIhTwI+D4lLWXjxo3Mnz/f47WUlBQ++eQTAN566y2WL1/ORx99FIjwWkxDNhYLBRkZGdx7772MHTuWiy++ONDhaI7kQQ3Jg/AleVBD8iB8SR7UkDzQppAtSoYNG8awYcN8XldVlblz57J9+3aWLl3qUT2HgvpsLBYq9u7dy6RJk5g0aRITJkwIdDiaJHngIHkQ3iQPHCQPwpvkgYPkgXaF3fCtRYsWceDAAZYsWRJyiQfhs7FYYWEh99xzD7Nnzw7KxAs0yYPQIHnQNJIHoUHyoGkkD0JDKORBWO1TUl5ezqBBg0hKSiI2Ntb1+htvvEFKSkoAI2te4bCx2CuvvML7779P9+7dXa8NHjyYhx9+OHBBBQnJg9AhedB4kgehQ/Kg8SQPQkco5EFYFSVCCCGEEEII7Qm74VtCCCGEEEIIbZGiRAghhBBCCBFQUpQIIYQQQgghAkqKEiGEEEIIIURASVEihBBCCCGECCgpSoQQQgghhBABJUWJEEIIIYQQIqCkKBFCCCGEEEIElBQlQgghhBBCiICSokQA8OGHH3LLLbegqiq7du3iqquu4uTJk4EOS4hWJXkghOSBECB5EAiKqqpqoIMQgaeqKrfffjtXX301n3zyCY899hiXX355oMMSolVJHggheSAESB4EghQlwiUzM5MRI0Zwww03MHfu3ECHI0RASB4IIXkgBEgetDYZviVcsrKyaNOmDXv37sVmswU6HCECQvJACMkDIUDyoLVJUSIAqKioYNasWbz66qskJyfzzjvvBDokIVqd5IEQkgdCgORBIBgCHYDQhhdeeIHLLruMfv360aVLF0aNGsXgwYM555xzAh2aEK1G8kAIyQMhQPIgEGROiRBCCCGEECKgZPiWEEIIIYQQIqCkKBFCCCGEEEIElBQlQgghhBBCiICSokQIIYQQQggRUFKUCCGEEEIIIQJKihIhhBBCCCFEQElRIoQQQgghhAgoKUqEEEIIIYQQASVFiRBCCCGEECKgpCgRQgghhBBCBJQUJUIIIYQQQoiA+n/1liU4aQjC2wAAAABJRU5ErkJggg==",
      "text/plain": [
       "<Figure size 800x200 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Purge graph space\n",
    "sns.set(rc = {\n",
    "              'font.family':'Arial',\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':2, 'ytick.major.pad':2,\n",
    "              'xtick.bottom': True, 'ytick.left': True,\n",
    "              'xtick.major.size':2, 'ytick.major.size':2,\n",
    "             },\n",
    "       style='ticks')\n",
    "\n",
    "fig = sns.relplot(proba_df, col='compound',\n",
    "                  x='x', y='p(x)', hue='label',\n",
    "                  kind='line', height=2, legend=False,\n",
    "                  facet_kws = {'sharey':False, 'despine':False}\n",
    "                 )\n",
    "\n",
    "fig.set_titles(row_template='{row_name}', col_template='{col_name}')\n",
    "\n",
    "for i,c in enumerate(proba_df.compound.unique()):\n",
    "    fig.axes[0][i].hist(hist_df[hist_df.compound == c].s,\n",
    "                        30, density=True, histtype='stepfilled',\n",
    "                        color='.8')\n",
    "    fig.axes[0][i].axvline(x=stddf[stddf.compound==c.lower()].val.tolist()[0], ls='--', color='grey')\n",
    "    fig.axes[0][i].axvline(x=stddf[stddf.compound==c.lower()].val.tolist()[1], ls='--', color='grey')\n",
    "    \n",
    "plt.savefig(f'{graph_outpath}/{\"_\".join([strain,locus,pool_type])}_GMM_densities.svg', format='svg', dpi=300)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "id": "96789d0d-282c-4a8b-a0ac-8b74626c3079",
   "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 tr th {\n",
       "        text-align: left;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th>compound</th>\n",
       "      <th>label</th>\n",
       "      <th colspan=\"2\" halign=\"left\">s</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th>size</th>\n",
       "      <th>median</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>anidulafungin</td>\n",
       "      <td>1.0</td>\n",
       "      <td>132</td>\n",
       "      <td>-0.012576</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>anidulafungin</td>\n",
       "      <td>2.0</td>\n",
       "      <td>79</td>\n",
       "      <td>1.492155</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>anidulafungin</td>\n",
       "      <td>0.0</td>\n",
       "      <td>105</td>\n",
       "      <td>2.057751</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>caspofungin</td>\n",
       "      <td>1.0</td>\n",
       "      <td>129</td>\n",
       "      <td>0.005718</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>caspofungin</td>\n",
       "      <td>2.0</td>\n",
       "      <td>84</td>\n",
       "      <td>1.448548</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>caspofungin</td>\n",
       "      <td>0.0</td>\n",
       "      <td>103</td>\n",
       "      <td>1.930501</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>micafungin</td>\n",
       "      <td>1.0</td>\n",
       "      <td>141</td>\n",
       "      <td>-0.037772</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>micafungin</td>\n",
       "      <td>2.0</td>\n",
       "      <td>96</td>\n",
       "      <td>1.450873</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>micafungin</td>\n",
       "      <td>0.0</td>\n",
       "      <td>79</td>\n",
       "      <td>2.121986</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>none</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1</td>\n",
       "      <td>-1.144093</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>none</td>\n",
       "      <td>1.0</td>\n",
       "      <td>38</td>\n",
       "      <td>-0.262749</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>none</td>\n",
       "      <td>0.0</td>\n",
       "      <td>277</td>\n",
       "      <td>0.038049</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         compound label    s          \n",
       "                        size    median\n",
       "1   anidulafungin   1.0  132 -0.012576\n",
       "2   anidulafungin   2.0   79  1.492155\n",
       "0   anidulafungin   0.0  105  2.057751\n",
       "4     caspofungin   1.0  129  0.005718\n",
       "5     caspofungin   2.0   84  1.448548\n",
       "3     caspofungin   0.0  103  1.930501\n",
       "7      micafungin   1.0  141 -0.037772\n",
       "8      micafungin   2.0   96  1.450873\n",
       "6      micafungin   0.0   79  2.121986\n",
       "11           none   2.0    1 -1.144093\n",
       "10           none   1.0   38 -0.262749\n",
       "9            none   0.0  277  0.038049"
      ]
     },
     "execution_count": 110,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Looking at the predicted labels for each condition and the median value to properly convert the label into a class\n",
    "label_median = GMMdf.groupby(['compound', 'label'])[['s']].agg(['size','median']).reset_index().sort_values(by=[('compound',''),('s','median')])\n",
    "label_median"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ce76f502-5418-485d-8622-bb6181e3f250",
   "metadata": {},
   "source": [
    "### Convert classification labels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "id": "ea4ade44-7654-48d7-8b93-23cb8f7325a0",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'anidulafungin': {1.0: 'WT-like', 2.0: 'intermediary', 0.0: 'resistant'},\n",
       " 'caspofungin': {1.0: 'WT-like', 2.0: 'intermediary', 0.0: 'resistant'},\n",
       " 'micafungin': {1.0: 'WT-like', 2.0: 'intermediary', 0.0: 'resistant'},\n",
       " 'none': {2.0: 'deleterious', 1.0: 'slightly deleterious', 0.0: 'WT-like'}}"
      ]
     },
     "execution_count": 111,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "class_dict = {}\n",
    "class_list = ['WT-like', 'intermediary', 'resistant']\n",
    "none_list = {'FKS1-HS1': ['deleterious', 'slightly deleterious', 'WT-like'],\n",
    "             'FKS1-HS2': ['WT-like', 'WT-like', 'WT-like'],\n",
    "             'FKS2-HS1': ['slightly deleterious', 'WT-like', 'WT-like'],\n",
    "             'FKS2-HS2': ['slightly deleterious', 'WT-like', 'WT-like']\n",
    "            }\n",
    "dox_list = {'FKS1-HS1': ['deleterious', 'slightly deleterious', 'WT-like'],\n",
    "            'FKS2-HS1': ['deleterious', 'slightly deleterious', 'WT-like'],\n",
    "            'FKS2-HS2': ['deleterious', 'slightly deleterious', 'WT-like']\n",
    "           }\n",
    "\n",
    "for c in label_median.compound.unique():\n",
    "    class_dict[c] = {}\n",
    "\n",
    "i=0\n",
    "for c,l in zip(label_median.compound, label_median.label):\n",
    "    i = len(class_dict[c])\n",
    "    if c in ['anidulafungin','caspofungin','micafungin']:\n",
    "        class_dict[c][l] = class_list[i]\n",
    "    elif c=='none':\n",
    "        if locus in none_list.keys():\n",
    "            class_dict[c][l] = none_list[locus][i]\n",
    "        else:\n",
    "            class_dict[c][l] = 'error'\n",
    "    elif c=='dox':\n",
    "        if locus in dox_list.keys():\n",
    "            class_dict[c][l] = dox_list[locus][i]\n",
    "        else:\n",
    "            class_dict[c][l] = 'error'\n",
    "    else:\n",
    "        class_dict[c][l] = 'error'\n",
    "\n",
    "class_dict"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 112,
   "id": "e15797ba-79a7-4ed9-bca1-24d76da09c47",
   "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>compound</th>\n",
       "      <th>seq_type</th>\n",
       "      <th>Nham_aa</th>\n",
       "      <th>aa_seq</th>\n",
       "      <th>nt_seq</th>\n",
       "      <th>s</th>\n",
       "      <th>cscore</th>\n",
       "      <th>label</th>\n",
       "      <th>rclass</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>anidulafungin</td>\n",
       "      <td>WT</td>\n",
       "      <td>0.0</td>\n",
       "      <td>FLVLSLRDP</td>\n",
       "      <td>tttttagttttatctttgagagatcca</td>\n",
       "      <td>0.059886</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>WT-like</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>anidulafungin</td>\n",
       "      <td>ortho</td>\n",
       "      <td>0.0</td>\n",
       "      <td>FLVLSLRDP</td>\n",
       "      <td>tttttggttttgtctttgagagatcca</td>\n",
       "      <td>0.042567</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>WT-like</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>anidulafungin</td>\n",
       "      <td>ortho</td>\n",
       "      <td>1.0</td>\n",
       "      <td>FLILSLRDP</td>\n",
       "      <td>tttttgattttgtctttgagagatcca</td>\n",
       "      <td>-0.223979</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>WT-like</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>anidulafungin</td>\n",
       "      <td>ortho</td>\n",
       "      <td>1.0</td>\n",
       "      <td>FLTLSLRDP</td>\n",
       "      <td>tttttgactttgtctttgagagatcca</td>\n",
       "      <td>0.127139</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>WT-like</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>anidulafungin</td>\n",
       "      <td>ortho</td>\n",
       "      <td>2.0</td>\n",
       "      <td>FLALSFRDP</td>\n",
       "      <td>tttttggctttgtcttttagagatcca</td>\n",
       "      <td>0.370152</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>WT-like</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1259</th>\n",
       "      <td>none</td>\n",
       "      <td>single</td>\n",
       "      <td>1.0</td>\n",
       "      <td>TLVLSLRDP</td>\n",
       "      <td>acgttagttttatctttgagagatcca</td>\n",
       "      <td>-0.449800</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>slightly deleterious</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1260</th>\n",
       "      <td>none</td>\n",
       "      <td>single</td>\n",
       "      <td>1.0</td>\n",
       "      <td>TLVLSLRDP</td>\n",
       "      <td>actttagttttatctttgagagatcca</td>\n",
       "      <td>0.033715</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>WT-like</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1261</th>\n",
       "      <td>none</td>\n",
       "      <td>single</td>\n",
       "      <td>1.0</td>\n",
       "      <td>VLVLSLRDP</td>\n",
       "      <td>gtgttagttttatctttgagagatcca</td>\n",
       "      <td>0.037623</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>WT-like</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1262</th>\n",
       "      <td>none</td>\n",
       "      <td>single</td>\n",
       "      <td>1.0</td>\n",
       "      <td>WLVLSLRDP</td>\n",
       "      <td>tggttagttttatctttgagagatcca</td>\n",
       "      <td>0.070207</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>WT-like</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1263</th>\n",
       "      <td>none</td>\n",
       "      <td>single</td>\n",
       "      <td>1.0</td>\n",
       "      <td>YLVLSLRDP</td>\n",
       "      <td>tatttagttttatctttgagagatcca</td>\n",
       "      <td>0.080017</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>WT-like</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1264 rows × 9 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "           compound seq_type  Nham_aa     aa_seq                       nt_seq  \\\n",
       "0     anidulafungin       WT      0.0  FLVLSLRDP  tttttagttttatctttgagagatcca   \n",
       "1     anidulafungin    ortho      0.0  FLVLSLRDP  tttttggttttgtctttgagagatcca   \n",
       "2     anidulafungin    ortho      1.0  FLILSLRDP  tttttgattttgtctttgagagatcca   \n",
       "3     anidulafungin    ortho      1.0  FLTLSLRDP  tttttgactttgtctttgagagatcca   \n",
       "4     anidulafungin    ortho      2.0  FLALSFRDP  tttttggctttgtcttttagagatcca   \n",
       "...             ...      ...      ...        ...                          ...   \n",
       "1259           none   single      1.0  TLVLSLRDP  acgttagttttatctttgagagatcca   \n",
       "1260           none   single      1.0  TLVLSLRDP  actttagttttatctttgagagatcca   \n",
       "1261           none   single      1.0  VLVLSLRDP  gtgttagttttatctttgagagatcca   \n",
       "1262           none   single      1.0  WLVLSLRDP  tggttagttttatctttgagagatcca   \n",
       "1263           none   single      1.0  YLVLSLRDP  tatttagttttatctttgagagatcca   \n",
       "\n",
       "             s  cscore  label                rclass  \n",
       "0     0.059886     1.0    1.0               WT-like  \n",
       "1     0.042567     1.0    1.0               WT-like  \n",
       "2    -0.223979     1.0    1.0               WT-like  \n",
       "3     0.127139     1.0    1.0               WT-like  \n",
       "4     0.370152     1.0    1.0               WT-like  \n",
       "...        ...     ...    ...                   ...  \n",
       "1259 -0.449800     1.0    1.0  slightly deleterious  \n",
       "1260  0.033715     1.0    0.0               WT-like  \n",
       "1261  0.037623     1.0    0.0               WT-like  \n",
       "1262  0.070207     1.0    0.0               WT-like  \n",
       "1263  0.080017     1.0    0.0               WT-like  \n",
       "\n",
       "[1264 rows x 9 columns]"
      ]
     },
     "execution_count": 112,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "GMMdf['rclass'] = GMMdf.apply(lambda row: class_dict[row.compound][row.label], axis=1)\n",
    "GMMdf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "id": "b4a9b59e-6034-49ba-ae72-cbcec10236c2",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "# Create output folder if it does not already exists\n",
    "if not os.path.exists(f'{classified_outpath}/{\"_\".join([strain,locus,pool_type])}'):\n",
    "    os.makedirs(f'{classified_outpath}/{\"_\".join([strain,locus,pool_type])}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "id": "6642911b-dc71-45a9-91f7-e0e0a39d9707",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "GMMdf.to_csv(f'{classified_outpath}/{\"_\".join([strain,locus,pool_type])}/nt_raw_classification.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "055fb5df-ef8c-4c38-a5de-5fe38d3f9849",
   "metadata": {},
   "source": [
    "## Breakdown: number of variants per class"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 115,
   "id": "28d42cd0-db32-43cb-b03b-dcef65660b4a",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "GMMdf = pd.read_csv(f'{classified_outpath}/{\"_\".join([strain,locus,pool_type])}/nt_raw_classification.csv', index_col=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 116,
   "id": "44777ce7-0102-4a19-a920-f7e11940e984",
   "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 tr th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe thead tr:last-of-type th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th colspan=\"2\" halign=\"left\">s</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th>size</th>\n",
       "      <th>median</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>compound</th>\n",
       "      <th>rclass</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"3\" valign=\"top\">anidulafungin</th>\n",
       "      <th>WT-like</th>\n",
       "      <td>132</td>\n",
       "      <td>-0.012576</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>intermediary</th>\n",
       "      <td>79</td>\n",
       "      <td>1.492155</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>resistant</th>\n",
       "      <td>105</td>\n",
       "      <td>2.057751</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"3\" valign=\"top\">caspofungin</th>\n",
       "      <th>WT-like</th>\n",
       "      <td>129</td>\n",
       "      <td>0.005718</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>intermediary</th>\n",
       "      <td>84</td>\n",
       "      <td>1.448548</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>resistant</th>\n",
       "      <td>103</td>\n",
       "      <td>1.930501</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"3\" valign=\"top\">micafungin</th>\n",
       "      <th>WT-like</th>\n",
       "      <td>141</td>\n",
       "      <td>-0.037772</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>intermediary</th>\n",
       "      <td>96</td>\n",
       "      <td>1.450873</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>resistant</th>\n",
       "      <td>79</td>\n",
       "      <td>2.121986</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"3\" valign=\"top\">none</th>\n",
       "      <th>WT-like</th>\n",
       "      <td>277</td>\n",
       "      <td>0.038049</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>deleterious</th>\n",
       "      <td>1</td>\n",
       "      <td>-1.144093</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>slightly deleterious</th>\n",
       "      <td>38</td>\n",
       "      <td>-0.262749</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                      s          \n",
       "                                   size    median\n",
       "compound      rclass                             \n",
       "anidulafungin WT-like               132 -0.012576\n",
       "              intermediary           79  1.492155\n",
       "              resistant             105  2.057751\n",
       "caspofungin   WT-like               129  0.005718\n",
       "              intermediary           84  1.448548\n",
       "              resistant             103  1.930501\n",
       "micafungin    WT-like               141 -0.037772\n",
       "              intermediary           96  1.450873\n",
       "              resistant              79  2.121986\n",
       "none          WT-like               277  0.038049\n",
       "              deleterious             1 -1.144093\n",
       "              slightly deleterious   38 -0.262749"
      ]
     },
     "execution_count": 116,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "GMMdf.groupby(['compound', 'rclass'])[['s']].agg(['size','median'])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "58ad4f5a-43d8-4482-96f1-ca17fbe43add",
   "metadata": {
    "tags": []
   },
   "source": [
    "## Refine classification (resolve overlaps)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "69c5f59c-ff21-4999-a707-7f8803ab6c33",
   "metadata": {},
   "source": [
    "To resolve overlaps at the nucleotide level, I use the distributions to define thresholds. However, because only the \"resistant\" and \"WT-like\" distributions (as called by the GMM) are narrow, we define the thresholds only for these groups. \"Intermediary\", \"deleterious\", etc end up being the label of variants which **did not** cross the thresholds.\n",
    "\n",
    "In order to have a single class per aminoacid variant, we reapply the thresholds on the aggregated dataframe"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d7c3fd59-317a-44df-9627-e834130bd97e",
   "metadata": {
    "tags": []
   },
   "source": [
    "### Step 1 - Set thresholds from narrow distributions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "id": "38dceda2-ad1d-4aa7-916c-24cb3f0a1843",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "zthresh = 3 # set threshold in z-score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 118,
   "id": "f5cd137a-5aad-46de-b49d-68d1ed323cd2",
   "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>compound</th>\n",
       "      <th>rclass</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>lower_bound</th>\n",
       "      <th>upper_bound</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>anidulafungin</td>\n",
       "      <td>WT-like</td>\n",
       "      <td>-0.013664</td>\n",
       "      <td>0.226409</td>\n",
       "      <td>-0.692890</td>\n",
       "      <td>0.665562</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>anidulafungin</td>\n",
       "      <td>intermediary</td>\n",
       "      <td>1.398613</td>\n",
       "      <td>0.365794</td>\n",
       "      <td>0.301232</td>\n",
       "      <td>2.495995</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>anidulafungin</td>\n",
       "      <td>resistant</td>\n",
       "      <td>2.044907</td>\n",
       "      <td>0.063072</td>\n",
       "      <td>1.855692</td>\n",
       "      <td>2.234123</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>caspofungin</td>\n",
       "      <td>WT-like</td>\n",
       "      <td>0.022527</td>\n",
       "      <td>0.207676</td>\n",
       "      <td>-0.600500</td>\n",
       "      <td>0.645554</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>caspofungin</td>\n",
       "      <td>intermediary</td>\n",
       "      <td>1.413736</td>\n",
       "      <td>0.296348</td>\n",
       "      <td>0.524693</td>\n",
       "      <td>2.302779</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>caspofungin</td>\n",
       "      <td>resistant</td>\n",
       "      <td>1.926463</td>\n",
       "      <td>0.051259</td>\n",
       "      <td>1.772685</td>\n",
       "      <td>2.080240</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>micafungin</td>\n",
       "      <td>WT-like</td>\n",
       "      <td>-0.029174</td>\n",
       "      <td>0.195351</td>\n",
       "      <td>-0.615226</td>\n",
       "      <td>0.556878</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>micafungin</td>\n",
       "      <td>intermediary</td>\n",
       "      <td>1.396550</td>\n",
       "      <td>0.413320</td>\n",
       "      <td>0.156589</td>\n",
       "      <td>2.636511</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>micafungin</td>\n",
       "      <td>resistant</td>\n",
       "      <td>2.125015</td>\n",
       "      <td>0.051190</td>\n",
       "      <td>1.971445</td>\n",
       "      <td>2.278586</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>none</td>\n",
       "      <td>WT-like</td>\n",
       "      <td>0.038281</td>\n",
       "      <td>0.032359</td>\n",
       "      <td>-0.058796</td>\n",
       "      <td>0.135357</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>none</td>\n",
       "      <td>deleterious</td>\n",
       "      <td>-1.144093</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>none</td>\n",
       "      <td>slightly deleterious</td>\n",
       "      <td>-0.273800</td>\n",
       "      <td>0.127841</td>\n",
       "      <td>-0.657323</td>\n",
       "      <td>0.109722</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         compound                rclass      mean       std  lower_bound  \\\n",
       "0   anidulafungin               WT-like -0.013664  0.226409    -0.692890   \n",
       "1   anidulafungin          intermediary  1.398613  0.365794     0.301232   \n",
       "2   anidulafungin             resistant  2.044907  0.063072     1.855692   \n",
       "3     caspofungin               WT-like  0.022527  0.207676    -0.600500   \n",
       "4     caspofungin          intermediary  1.413736  0.296348     0.524693   \n",
       "5     caspofungin             resistant  1.926463  0.051259     1.772685   \n",
       "6      micafungin               WT-like -0.029174  0.195351    -0.615226   \n",
       "7      micafungin          intermediary  1.396550  0.413320     0.156589   \n",
       "8      micafungin             resistant  2.125015  0.051190     1.971445   \n",
       "9            none               WT-like  0.038281  0.032359    -0.058796   \n",
       "10           none           deleterious -1.144093       NaN          NaN   \n",
       "11           none  slightly deleterious -0.273800  0.127841    -0.657323   \n",
       "\n",
       "    upper_bound  \n",
       "0      0.665562  \n",
       "1      2.495995  \n",
       "2      2.234123  \n",
       "3      0.645554  \n",
       "4      2.302779  \n",
       "5      2.080240  \n",
       "6      0.556878  \n",
       "7      2.636511  \n",
       "8      2.278586  \n",
       "9      0.135357  \n",
       "10          NaN  \n",
       "11     0.109722  "
      ]
     },
     "execution_count": 118,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "distri_params = GMMdf.groupby(['compound','rclass'])[['s']].agg(['mean','std']).reset_index()\n",
    "distri_params.columns = [x[0] if x[1]=='' else x[1] for x in distri_params.columns]\n",
    "distri_params['lower_bound'] = distri_params['mean'] - zthresh*distri_params['std']\n",
    "distri_params['upper_bound'] = distri_params['mean'] + zthresh*distri_params['std']\n",
    "distri_params"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 119,
   "id": "379d3fd7-d46d-41f8-a526-a79e3abb1a8e",
   "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>compound</th>\n",
       "      <th>thresh</th>\n",
       "      <th>val</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>anidulafungin</td>\n",
       "      <td>low</td>\n",
       "      <td>0.665562</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>anidulafungin</td>\n",
       "      <td>up</td>\n",
       "      <td>1.855692</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>caspofungin</td>\n",
       "      <td>low</td>\n",
       "      <td>0.645554</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>caspofungin</td>\n",
       "      <td>up</td>\n",
       "      <td>1.772685</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>micafungin</td>\n",
       "      <td>low</td>\n",
       "      <td>0.556878</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>micafungin</td>\n",
       "      <td>up</td>\n",
       "      <td>1.971445</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>none</td>\n",
       "      <td>low</td>\n",
       "      <td>-0.058796</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>none</td>\n",
       "      <td>up</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        compound thresh       val\n",
       "0  anidulafungin    low  0.665562\n",
       "1  anidulafungin     up  1.855692\n",
       "2    caspofungin    low  0.645554\n",
       "3    caspofungin     up  1.772685\n",
       "4     micafungin    low  0.556878\n",
       "5     micafungin     up  1.971445\n",
       "6           none    low -0.058796\n",
       "7           none     up       NaN"
      ]
     },
     "execution_count": 119,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "stdlist = []\n",
    "\n",
    "for c in distri_params.compound.unique():\n",
    "    if c in ['caspofungin','micafungin','anidulafungin']:\n",
    "        stdlist.append([c, 'low',\n",
    "                        distri_params.loc[(distri_params.compound == c)\n",
    "                                          & (distri_params.rclass == 'WT-like'),\n",
    "                                          'upper_bound'\n",
    "                                         ].values[0]\n",
    "                       ])\n",
    "        if locus in ['FKS1-HS1','FKS2-HS1']:\n",
    "            stdlist.append([c, 'up',\n",
    "                            distri_params.loc[(distri_params.compound == c)\n",
    "                                              & (distri_params.rclass == 'resistant'),\n",
    "                                              'lower_bound'\n",
    "                                             ].values[0]                    \n",
    "                           ])\n",
    "        elif locus in ['FKS1-HS2','FKS2-HS2']:\n",
    "            stdlist.append([c, 'up',\n",
    "                            distri_params.loc[(distri_params.compound == c)\n",
    "                                              & (distri_params.rclass == 'intermediary'),\n",
    "                                              'upper_bound'\n",
    "                                             ].values[0]\n",
    "                           ])\n",
    "        else:\n",
    "            print('Please define locus-specific bound')\n",
    "    else:\n",
    "        stdlist.append([c, 'low',\n",
    "                        distri_params.loc[(distri_params.compound == c)\n",
    "                                          & (distri_params.rclass == 'WT-like'),\n",
    "                                          'lower_bound'\n",
    "                                         ].values[0]\n",
    "                       ])\n",
    "        stdlist.append([c, 'up', None])\n",
    "    \n",
    "stddf = pd.DataFrame(stdlist, columns=['compound', 'thresh', 'val'])\n",
    "stddf.to_csv(f'{classified_outpath}/{\"_\".join([strain,locus,pool_type])}/thresholds.csv')\n",
    "stddf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 120,
   "id": "ca8dec7f-8bc6-44c1-a6f3-73cf0bc32c81",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Axes: xlabel='compound', ylabel='s'>"
      ]
     },
     "execution_count": 120,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 631.625x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = sns.catplot(GMMdf, x='compound', y='s', hue='rclass',\n",
    "            hue_order=['resistant','intermediary','WT-like','slightly deleterious','deleterious'],\n",
    "            palette='icefire_r',\n",
    "            #alpha=.3\n",
    "                 )\n",
    "\n",
    "sns.boxplot(stddf, x='compound', y='val',\n",
    "            hue='thresh', hue_order=['low','up'], palette=['black','orange'], fill=False,\n",
    "            legend=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c2a1b99a-4255-45cb-b4a6-1d5c008ba867",
   "metadata": {
    "tags": []
   },
   "source": [
    "### Step 2 - Refine classification for all nucleotide variants"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 121,
   "id": "504e0dbe-e524-4c34-9a3f-5503853f8172",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "def refine_class(v, comp, threshdf):\n",
    "    wtbound = threshdf.loc[(threshdf.compound == comp)\n",
    "                           & (threshdf.thresh == 'low'),\n",
    "                           'val'].values[0]\n",
    "    resbound = threshdf.loc[(threshdf.compound == comp)\n",
    "                           & (threshdf.thresh == 'up'),\n",
    "                           'val'].values[0]\n",
    "    \n",
    "    if (comp in ['none','dox']) & (v > -.8):\n",
    "        if (v >= wtbound):\n",
    "            return 'WT-like'\n",
    "        else:\n",
    "            return 'slightly deleterious'\n",
    "    elif (comp in ['none','dox']) & (v < -.8):\n",
    "        return 'deleterious'\n",
    "    else:        \n",
    "        if v <= wtbound:\n",
    "            return 'WT-like'\n",
    "        elif v >= resbound:\n",
    "            return 'resistant'\n",
    "        else:\n",
    "            return 'intermediary'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 122,
   "id": "d20da0db-bbcd-468a-8238-b20d35f35314",
   "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>compound</th>\n",
       "      <th>seq_type</th>\n",
       "      <th>Nham_aa</th>\n",
       "      <th>aa_seq</th>\n",
       "      <th>nt_seq</th>\n",
       "      <th>s</th>\n",
       "      <th>cscore</th>\n",
       "      <th>label</th>\n",
       "      <th>rclass</th>\n",
       "      <th>refined_class</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>anidulafungin</td>\n",
       "      <td>WT</td>\n",
       "      <td>0.0</td>\n",
       "      <td>FLVLSLRDP</td>\n",
       "      <td>tttttagttttatctttgagagatcca</td>\n",
       "      <td>0.059886</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>WT-like</td>\n",
       "      <td>WT-like</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>anidulafungin</td>\n",
       "      <td>ortho</td>\n",
       "      <td>0.0</td>\n",
       "      <td>FLVLSLRDP</td>\n",
       "      <td>tttttggttttgtctttgagagatcca</td>\n",
       "      <td>0.042567</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>WT-like</td>\n",
       "      <td>WT-like</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>anidulafungin</td>\n",
       "      <td>ortho</td>\n",
       "      <td>1.0</td>\n",
       "      <td>FLILSLRDP</td>\n",
       "      <td>tttttgattttgtctttgagagatcca</td>\n",
       "      <td>-0.223979</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>WT-like</td>\n",
       "      <td>WT-like</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>anidulafungin</td>\n",
       "      <td>ortho</td>\n",
       "      <td>1.0</td>\n",
       "      <td>FLTLSLRDP</td>\n",
       "      <td>tttttgactttgtctttgagagatcca</td>\n",
       "      <td>0.127139</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>WT-like</td>\n",
       "      <td>WT-like</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>anidulafungin</td>\n",
       "      <td>ortho</td>\n",
       "      <td>2.0</td>\n",
       "      <td>FLALSFRDP</td>\n",
       "      <td>tttttggctttgtcttttagagatcca</td>\n",
       "      <td>0.370152</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>WT-like</td>\n",
       "      <td>WT-like</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>1259</th>\n",
       "      <td>none</td>\n",
       "      <td>single</td>\n",
       "      <td>1.0</td>\n",
       "      <td>TLVLSLRDP</td>\n",
       "      <td>acgttagttttatctttgagagatcca</td>\n",
       "      <td>-0.449800</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>slightly deleterious</td>\n",
       "      <td>slightly deleterious</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1260</th>\n",
       "      <td>none</td>\n",
       "      <td>single</td>\n",
       "      <td>1.0</td>\n",
       "      <td>TLVLSLRDP</td>\n",
       "      <td>actttagttttatctttgagagatcca</td>\n",
       "      <td>0.033715</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>WT-like</td>\n",
       "      <td>WT-like</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1261</th>\n",
       "      <td>none</td>\n",
       "      <td>single</td>\n",
       "      <td>1.0</td>\n",
       "      <td>VLVLSLRDP</td>\n",
       "      <td>gtgttagttttatctttgagagatcca</td>\n",
       "      <td>0.037623</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>WT-like</td>\n",
       "      <td>WT-like</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1262</th>\n",
       "      <td>none</td>\n",
       "      <td>single</td>\n",
       "      <td>1.0</td>\n",
       "      <td>WLVLSLRDP</td>\n",
       "      <td>tggttagttttatctttgagagatcca</td>\n",
       "      <td>0.070207</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>WT-like</td>\n",
       "      <td>WT-like</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1263</th>\n",
       "      <td>none</td>\n",
       "      <td>single</td>\n",
       "      <td>1.0</td>\n",
       "      <td>YLVLSLRDP</td>\n",
       "      <td>tatttagttttatctttgagagatcca</td>\n",
       "      <td>0.080017</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>WT-like</td>\n",
       "      <td>WT-like</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1264 rows × 10 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "           compound seq_type  Nham_aa     aa_seq                       nt_seq  \\\n",
       "0     anidulafungin       WT      0.0  FLVLSLRDP  tttttagttttatctttgagagatcca   \n",
       "1     anidulafungin    ortho      0.0  FLVLSLRDP  tttttggttttgtctttgagagatcca   \n",
       "2     anidulafungin    ortho      1.0  FLILSLRDP  tttttgattttgtctttgagagatcca   \n",
       "3     anidulafungin    ortho      1.0  FLTLSLRDP  tttttgactttgtctttgagagatcca   \n",
       "4     anidulafungin    ortho      2.0  FLALSFRDP  tttttggctttgtcttttagagatcca   \n",
       "...             ...      ...      ...        ...                          ...   \n",
       "1259           none   single      1.0  TLVLSLRDP  acgttagttttatctttgagagatcca   \n",
       "1260           none   single      1.0  TLVLSLRDP  actttagttttatctttgagagatcca   \n",
       "1261           none   single      1.0  VLVLSLRDP  gtgttagttttatctttgagagatcca   \n",
       "1262           none   single      1.0  WLVLSLRDP  tggttagttttatctttgagagatcca   \n",
       "1263           none   single      1.0  YLVLSLRDP  tatttagttttatctttgagagatcca   \n",
       "\n",
       "             s  cscore  label                rclass         refined_class  \n",
       "0     0.059886     1.0    1.0               WT-like               WT-like  \n",
       "1     0.042567     1.0    1.0               WT-like               WT-like  \n",
       "2    -0.223979     1.0    1.0               WT-like               WT-like  \n",
       "3     0.127139     1.0    1.0               WT-like               WT-like  \n",
       "4     0.370152     1.0    1.0               WT-like               WT-like  \n",
       "...        ...     ...    ...                   ...                   ...  \n",
       "1259 -0.449800     1.0    1.0  slightly deleterious  slightly deleterious  \n",
       "1260  0.033715     1.0    0.0               WT-like               WT-like  \n",
       "1261  0.037623     1.0    0.0               WT-like               WT-like  \n",
       "1262  0.070207     1.0    0.0               WT-like               WT-like  \n",
       "1263  0.080017     1.0    0.0               WT-like               WT-like  \n",
       "\n",
       "[1264 rows x 10 columns]"
      ]
     },
     "execution_count": 122,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "GMMdf['refined_class'] = GMMdf.apply(lambda row: refine_class(row.s, row.compound, stddf), axis=1)\n",
    "GMMdf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 123,
   "id": "b2737bb6-1d15-4157-bce9-a46a6ca53283",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Axes: xlabel='compound', ylabel='s'>"
      ]
     },
     "execution_count": 123,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 631.625x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = sns.catplot(GMMdf, x='compound', y='s', hue='refined_class',\n",
    "            hue_order=['resistant','intermediary','WT-like','slightly deleterious','deleterious'],\n",
    "            palette='icefire_r',\n",
    "            #alpha=.3\n",
    "                 )\n",
    "\n",
    "sns.boxplot(stddf, x='compound', y='val',\n",
    "            hue='thresh', hue_order=['low','up'], palette=['black','orange'], fill=False,\n",
    "            legend=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 124,
   "id": "c5c7716b-3e03-4169-883a-78b0f50867b1",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "GMMdf.to_csv(f'{classified_outpath}/{\"_\".join([strain,locus,pool_type])}/nt_refined_classification.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "36efa543-0e94-4a04-bfcc-4d6421b672a3",
   "metadata": {
    "tags": []
   },
   "source": [
    "### Step 3 - Classify aminoacid variants"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 125,
   "id": "da669932-0e82-4e64-9267-b31b91db0b8b",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "slpAA = GMMdf.groupby(['compound','seq_type','Nham_aa','aa_seq']\n",
    "                         )[['s','cscore']\n",
    "                          ].agg(s = ('s', 'median'),\n",
    "                                cscore = ('cscore', 'min')).reset_index()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9fd10d16-879d-4575-ba97-65e0093f1b4c",
   "metadata": {},
   "source": [
    "Here I'm adding a little snippet to include FKS1-HS1 mutants originally missing from the DMS dataset but that were individually reconstructed and for which we have an inferred DMS score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 126,
   "id": "c25f10f0-7657-4e66-ad5d-0d2ca98921aa",
   "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>compound</th>\n",
       "      <th>aa_seq</th>\n",
       "      <th>s</th>\n",
       "      <th>seq_type</th>\n",
       "      <th>Nham_aa</th>\n",
       "      <th>cscore</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>anidulafungin</td>\n",
       "      <td>CLVLSLRDP</td>\n",
       "      <td>2.009071</td>\n",
       "      <td>single</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>caspofungin</td>\n",
       "      <td>CLVLSLRDP</td>\n",
       "      <td>1.918283</td>\n",
       "      <td>single</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>micafungin</td>\n",
       "      <td>CLVLSLRDP</td>\n",
       "      <td>1.941420</td>\n",
       "      <td>single</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>none</td>\n",
       "      <td>CLVLSLRDP</td>\n",
       "      <td>0.122445</td>\n",
       "      <td>single</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         compound     aa_seq         s seq_type  Nham_aa  cscore\n",
       "8   anidulafungin  CLVLSLRDP  2.009071   single        1       1\n",
       "9     caspofungin  CLVLSLRDP  1.918283   single        1       1\n",
       "10     micafungin  CLVLSLRDP  1.941420   single        1       1\n",
       "11           none  CLVLSLRDP  0.122445   single        1       1"
      ]
     },
     "execution_count": 126,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "missmut = pd.read_csv(missing_mut_path, index_col=0)[['compound','aa_seq','estimated_s']]\n",
    "missmut.rename(columns={'estimated_s':'s'}, inplace=True)\n",
    "missmut['seq_type'] = 'single'\n",
    "missmut['Nham_aa'] = 1\n",
    "missmut['cscore'] = 1\n",
    "missmut.head(4)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 127,
   "id": "e33455e2-d7ec-495f-8122-266f196d348f",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "# Replacing the two DMS datapoints which were overestimated to replace by validation data\n",
    "# Note: even though values obtained with caspofungin and anidulafungin were very well correlated, we still replace with the inferred score + that way we don't have duplicate values in the df\n",
    "slpAA.drop(slpAA[slpAA.aa_seq.isin(missmut.aa_seq.unique())].index, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 128,
   "id": "913891a2-d95a-49bc-93d8-22a0b7fa495d",
   "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>compound</th>\n",
       "      <th>seq_type</th>\n",
       "      <th>Nham_aa</th>\n",
       "      <th>aa_seq</th>\n",
       "      <th>s</th>\n",
       "      <th>cscore</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>anidulafungin</td>\n",
       "      <td>WT</td>\n",
       "      <td>0.0</td>\n",
       "      <td>FLVLSLRDP</td>\n",
       "      <td>0.059886</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>anidulafungin</td>\n",
       "      <td>ortho</td>\n",
       "      <td>0.0</td>\n",
       "      <td>FLVLSLRDP</td>\n",
       "      <td>0.042567</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>anidulafungin</td>\n",
       "      <td>ortho</td>\n",
       "      <td>1.0</td>\n",
       "      <td>FLILSLRDP</td>\n",
       "      <td>-0.223979</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>anidulafungin</td>\n",
       "      <td>ortho</td>\n",
       "      <td>1.0</td>\n",
       "      <td>FLTLSLRDP</td>\n",
       "      <td>0.127139</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>anidulafungin</td>\n",
       "      <td>ortho</td>\n",
       "      <td>2.0</td>\n",
       "      <td>FLALSFRDP</td>\n",
       "      <td>0.370152</td>\n",
       "      <td>1.0</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>919</th>\n",
       "      <td>none</td>\n",
       "      <td>single</td>\n",
       "      <td>1.0</td>\n",
       "      <td>FLVLSLRDQ</td>\n",
       "      <td>0.007008</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>920</th>\n",
       "      <td>anidulafungin</td>\n",
       "      <td>single</td>\n",
       "      <td>1.0</td>\n",
       "      <td>FLWLSLRDP</td>\n",
       "      <td>1.445682</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>921</th>\n",
       "      <td>caspofungin</td>\n",
       "      <td>single</td>\n",
       "      <td>1.0</td>\n",
       "      <td>FLWLSLRDP</td>\n",
       "      <td>1.040897</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>922</th>\n",
       "      <td>micafungin</td>\n",
       "      <td>single</td>\n",
       "      <td>1.0</td>\n",
       "      <td>FLWLSLRDP</td>\n",
       "      <td>0.903476</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>923</th>\n",
       "      <td>none</td>\n",
       "      <td>single</td>\n",
       "      <td>1.0</td>\n",
       "      <td>FLWLSLRDP</td>\n",
       "      <td>0.108415</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>924 rows × 6 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          compound seq_type  Nham_aa     aa_seq         s  cscore\n",
       "0    anidulafungin       WT      0.0  FLVLSLRDP  0.059886     1.0\n",
       "1    anidulafungin    ortho      0.0  FLVLSLRDP  0.042567     1.0\n",
       "2    anidulafungin    ortho      1.0  FLILSLRDP -0.223979     1.0\n",
       "3    anidulafungin    ortho      1.0  FLTLSLRDP  0.127139     1.0\n",
       "4    anidulafungin    ortho      2.0  FLALSFRDP  0.370152     1.0\n",
       "..             ...      ...      ...        ...       ...     ...\n",
       "919           none   single      1.0  FLVLSLRDQ  0.007008     1.0\n",
       "920  anidulafungin   single      1.0  FLWLSLRDP  1.445682     1.0\n",
       "921    caspofungin   single      1.0  FLWLSLRDP  1.040897     1.0\n",
       "922     micafungin   single      1.0  FLWLSLRDP  0.903476     1.0\n",
       "923           none   single      1.0  FLWLSLRDP  0.108415     1.0\n",
       "\n",
       "[924 rows x 6 columns]"
      ]
     },
     "execution_count": 128,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "if locus == 'FKS1-HS1':\n",
    "    slpAA = pd.concat([slpAA, missmut], ignore_index=True)\n",
    "elif locus in ['FKS1-HS2','FKS2-HS2']:\n",
    "    print('No mutants added')\n",
    "else:\n",
    "    print('Locus not supported')\n",
    "slpAA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 129,
   "id": "21b32d37-e26a-413c-a7e8-cfa821f25076",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "slpAA['refined_class'] = slpAA.apply(lambda row: refine_class(row.s, row.compound, stddf), axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 130,
   "id": "653c941a-6043-424e-867c-a3d830c901ef",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 400x200 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Purge graph space\n",
    "sns.set(rc = {\n",
    "              'font.family':'Arial',\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':2, 'ytick.major.pad':2,\n",
    "              'xtick.bottom': True, 'ytick.left': True,\n",
    "              'xtick.major.size':2, 'ytick.major.size':2,\n",
    "             },\n",
    "       style='ticks')\n",
    "\n",
    "lclasses = ['resistant','intermediary','WT-like','slightly deleterious','deleterious']\n",
    "custom_palette = ['#C75DAB','#D691C1','#F1F1F1','#42B7B9','#009B9E']\n",
    "\n",
    "fig = sns.catplot(slpAA, x='s', y='compound', order=['none','anidulafungin','caspofungin','micafungin'],\n",
    "                  hue='refined_class', hue_order=lclasses, palette=custom_palette, height=2, aspect=2,\n",
    "                  alpha=.4, legend=False\n",
    "                 )\n",
    "\n",
    "sns.boxplot(stddf, x='val', y='compound',\n",
    "            hue='thresh', hue_order=['low','up'], palette=['#D691C1','#C75DAB'], fill=False,\n",
    "            legend=False)\n",
    "\n",
    "fig.set(xlabel='Selection coefficient', ylabel='')\n",
    "plt.savefig(f'{graph_outpath}/{\"_\".join([strain,locus,pool_type])}_classification_stripplot.svg', format='svg', dpi=300)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 131,
   "id": "626ec667-377c-4bb5-9192-a97cef882cb7",
   "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>compound</th>\n",
       "      <th>seq_type</th>\n",
       "      <th>Nham_aa</th>\n",
       "      <th>aa_seq</th>\n",
       "      <th>s</th>\n",
       "      <th>cscore</th>\n",
       "      <th>refined_class</th>\n",
       "      <th>sensres</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>anidulafungin</td>\n",
       "      <td>WT</td>\n",
       "      <td>0.0</td>\n",
       "      <td>FLVLSLRDP</td>\n",
       "      <td>0.059886</td>\n",
       "      <td>1.0</td>\n",
       "      <td>WT-like</td>\n",
       "      <td>sensitive</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>anidulafungin</td>\n",
       "      <td>ortho</td>\n",
       "      <td>0.0</td>\n",
       "      <td>FLVLSLRDP</td>\n",
       "      <td>0.042567</td>\n",
       "      <td>1.0</td>\n",
       "      <td>WT-like</td>\n",
       "      <td>sensitive</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>anidulafungin</td>\n",
       "      <td>ortho</td>\n",
       "      <td>1.0</td>\n",
       "      <td>FLILSLRDP</td>\n",
       "      <td>-0.223979</td>\n",
       "      <td>1.0</td>\n",
       "      <td>WT-like</td>\n",
       "      <td>sensitive</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>anidulafungin</td>\n",
       "      <td>ortho</td>\n",
       "      <td>1.0</td>\n",
       "      <td>FLTLSLRDP</td>\n",
       "      <td>0.127139</td>\n",
       "      <td>1.0</td>\n",
       "      <td>WT-like</td>\n",
       "      <td>sensitive</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>anidulafungin</td>\n",
       "      <td>ortho</td>\n",
       "      <td>2.0</td>\n",
       "      <td>FLALSFRDP</td>\n",
       "      <td>0.370152</td>\n",
       "      <td>1.0</td>\n",
       "      <td>WT-like</td>\n",
       "      <td>sensitive</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>919</th>\n",
       "      <td>none</td>\n",
       "      <td>single</td>\n",
       "      <td>1.0</td>\n",
       "      <td>FLVLSLRDQ</td>\n",
       "      <td>0.007008</td>\n",
       "      <td>1.0</td>\n",
       "      <td>WT-like</td>\n",
       "      <td>sensitive</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>920</th>\n",
       "      <td>anidulafungin</td>\n",
       "      <td>single</td>\n",
       "      <td>1.0</td>\n",
       "      <td>FLWLSLRDP</td>\n",
       "      <td>1.445682</td>\n",
       "      <td>1.0</td>\n",
       "      <td>intermediary</td>\n",
       "      <td>resistant</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>921</th>\n",
       "      <td>caspofungin</td>\n",
       "      <td>single</td>\n",
       "      <td>1.0</td>\n",
       "      <td>FLWLSLRDP</td>\n",
       "      <td>1.040897</td>\n",
       "      <td>1.0</td>\n",
       "      <td>intermediary</td>\n",
       "      <td>resistant</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>922</th>\n",
       "      <td>micafungin</td>\n",
       "      <td>single</td>\n",
       "      <td>1.0</td>\n",
       "      <td>FLWLSLRDP</td>\n",
       "      <td>0.903476</td>\n",
       "      <td>1.0</td>\n",
       "      <td>intermediary</td>\n",
       "      <td>resistant</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>923</th>\n",
       "      <td>none</td>\n",
       "      <td>single</td>\n",
       "      <td>1.0</td>\n",
       "      <td>FLWLSLRDP</td>\n",
       "      <td>0.108415</td>\n",
       "      <td>1.0</td>\n",
       "      <td>WT-like</td>\n",
       "      <td>sensitive</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>924 rows × 8 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          compound seq_type  Nham_aa     aa_seq         s  cscore  \\\n",
       "0    anidulafungin       WT      0.0  FLVLSLRDP  0.059886     1.0   \n",
       "1    anidulafungin    ortho      0.0  FLVLSLRDP  0.042567     1.0   \n",
       "2    anidulafungin    ortho      1.0  FLILSLRDP -0.223979     1.0   \n",
       "3    anidulafungin    ortho      1.0  FLTLSLRDP  0.127139     1.0   \n",
       "4    anidulafungin    ortho      2.0  FLALSFRDP  0.370152     1.0   \n",
       "..             ...      ...      ...        ...       ...     ...   \n",
       "919           none   single      1.0  FLVLSLRDQ  0.007008     1.0   \n",
       "920  anidulafungin   single      1.0  FLWLSLRDP  1.445682     1.0   \n",
       "921    caspofungin   single      1.0  FLWLSLRDP  1.040897     1.0   \n",
       "922     micafungin   single      1.0  FLWLSLRDP  0.903476     1.0   \n",
       "923           none   single      1.0  FLWLSLRDP  0.108415     1.0   \n",
       "\n",
       "    refined_class    sensres  \n",
       "0         WT-like  sensitive  \n",
       "1         WT-like  sensitive  \n",
       "2         WT-like  sensitive  \n",
       "3         WT-like  sensitive  \n",
       "4         WT-like  sensitive  \n",
       "..            ...        ...  \n",
       "919       WT-like  sensitive  \n",
       "920  intermediary  resistant  \n",
       "921  intermediary  resistant  \n",
       "922  intermediary  resistant  \n",
       "923       WT-like  sensitive  \n",
       "\n",
       "[924 rows x 8 columns]"
      ]
     },
     "execution_count": 131,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Convert to 2 levels only = sensitive or resistant\n",
    "slpAA['sensres'] = slpAA.refined_class.replace({'intermediary':'resistant',\n",
    "                                                'WT-like':'sensitive', 'slightly deleterious':'sensitive', 'deleterious':'sensitive'})\n",
    "slpAA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 132,
   "id": "f3f55df0-c096-4d37-9520-f4a8f7db86d9",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "slpAA.drop(slpAA[(slpAA.Nham_aa == 0) & (slpAA.seq_type != 'single')].index, inplace=True)\n",
    "slpAA.reset_index(drop=True, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 133,
   "id": "cb3ea03d-0210-403e-9af7-c3f82789512f",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "slpAA.to_csv(f'{classified_outpath}/{\"_\".join([strain,locus,pool_type])}/aa_refined_classification.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4a22a74b-b091-4b48-bdab-6717156705b3",
   "metadata": {
    "tags": []
   },
   "source": [
    "### Step 4 - Keep info from all nucleotide variants but refined class is the one from aminoacid variants"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9996bd30-412d-43c9-8067-808bfa783c5f",
   "metadata": {},
   "source": [
    "This is useful to display error bars on our plot for validations. Each dot is an aminoacid variant colored by the refined class, but the error bars correspond to the 1st and 3rd quartiles obtained across nucleotide variants (encoding the given aminoacid sequence)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 134,
   "id": "743d619e-a3ff-4f15-a7cc-f00ee8a3e0b1",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "# The following merge operation automatically excludes all WT instances except the ones of seq_type == 'single', which correspond to median of synonymous mutants (only 1 mutated codon), excluding the WT nucleotide sequence\n",
    "lcols = ['compound','seq_type','Nham_aa','aa_seq']\n",
    "GMMdf = GMMdf.merge(right=slpAA[lcols+['refined_class']].rename(columns={'refined_class':'aa_class'}), on=lcols)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 135,
   "id": "2774e221-73e9-4f86-aab2-7f0128f9a7da",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "GMMdf.to_csv(f'{classified_outpath}/{\"_\".join([strain,locus,pool_type])}/nt_refined_classification.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3779ae64-e508-4057-bc45-407e983aaae3",
   "metadata": {
    "tags": []
   },
   "source": [
    "### Counts / %"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "id": "ce8da4a4-4dc6-4646-937d-84ff2aa2b614",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>seq_type</th>\n",
       "      <th>compound</th>\n",
       "      <th>sensres</th>\n",
       "      <th>aa_seq</th>\n",
       "      <th>totals</th>\n",
       "      <th>%</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>ortho</td>\n",
       "      <td>anidulafungin</td>\n",
       "      <td>resistant</td>\n",
       "      <td>3</td>\n",
       "      <td>62</td>\n",
       "      <td>4.8%</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>ortho</td>\n",
       "      <td>anidulafungin</td>\n",
       "      <td>sensitive</td>\n",
       "      <td>59</td>\n",
       "      <td>62</td>\n",
       "      <td>95.2%</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>ortho</td>\n",
       "      <td>caspofungin</td>\n",
       "      <td>resistant</td>\n",
       "      <td>6</td>\n",
       "      <td>62</td>\n",
       "      <td>9.7%</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>ortho</td>\n",
       "      <td>caspofungin</td>\n",
       "      <td>sensitive</td>\n",
       "      <td>56</td>\n",
       "      <td>62</td>\n",
       "      <td>90.3%</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>ortho</td>\n",
       "      <td>dox</td>\n",
       "      <td>sensitive</td>\n",
       "      <td>62</td>\n",
       "      <td>62</td>\n",
       "      <td>100.0%</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>ortho</td>\n",
       "      <td>micafungin</td>\n",
       "      <td>resistant</td>\n",
       "      <td>2</td>\n",
       "      <td>62</td>\n",
       "      <td>3.2%</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>ortho</td>\n",
       "      <td>micafungin</td>\n",
       "      <td>sensitive</td>\n",
       "      <td>60</td>\n",
       "      <td>62</td>\n",
       "      <td>96.8%</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>ortho</td>\n",
       "      <td>none</td>\n",
       "      <td>sensitive</td>\n",
       "      <td>62</td>\n",
       "      <td>62</td>\n",
       "      <td>100.0%</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>single</td>\n",
       "      <td>anidulafungin</td>\n",
       "      <td>resistant</td>\n",
       "      <td>22</td>\n",
       "      <td>125</td>\n",
       "      <td>17.6%</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>single</td>\n",
       "      <td>anidulafungin</td>\n",
       "      <td>sensitive</td>\n",
       "      <td>103</td>\n",
       "      <td>125</td>\n",
       "      <td>82.4%</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>single</td>\n",
       "      <td>caspofungin</td>\n",
       "      <td>resistant</td>\n",
       "      <td>37</td>\n",
       "      <td>125</td>\n",
       "      <td>29.6%</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>single</td>\n",
       "      <td>caspofungin</td>\n",
       "      <td>sensitive</td>\n",
       "      <td>88</td>\n",
       "      <td>125</td>\n",
       "      <td>70.4%</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>single</td>\n",
       "      <td>dox</td>\n",
       "      <td>sensitive</td>\n",
       "      <td>125</td>\n",
       "      <td>125</td>\n",
       "      <td>100.0%</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>single</td>\n",
       "      <td>micafungin</td>\n",
       "      <td>resistant</td>\n",
       "      <td>43</td>\n",
       "      <td>125</td>\n",
       "      <td>34.4%</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>single</td>\n",
       "      <td>micafungin</td>\n",
       "      <td>sensitive</td>\n",
       "      <td>82</td>\n",
       "      <td>125</td>\n",
       "      <td>65.6%</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>single</td>\n",
       "      <td>none</td>\n",
       "      <td>sensitive</td>\n",
       "      <td>125</td>\n",
       "      <td>125</td>\n",
       "      <td>100.0%</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   seq_type       compound    sensres  aa_seq  totals       %\n",
       "0     ortho  anidulafungin  resistant       3      62    4.8%\n",
       "1     ortho  anidulafungin  sensitive      59      62   95.2%\n",
       "2     ortho    caspofungin  resistant       6      62    9.7%\n",
       "3     ortho    caspofungin  sensitive      56      62   90.3%\n",
       "4     ortho            dox  sensitive      62      62  100.0%\n",
       "5     ortho     micafungin  resistant       2      62    3.2%\n",
       "6     ortho     micafungin  sensitive      60      62   96.8%\n",
       "7     ortho           none  sensitive      62      62  100.0%\n",
       "8    single  anidulafungin  resistant      22     125   17.6%\n",
       "9    single  anidulafungin  sensitive     103     125   82.4%\n",
       "10   single    caspofungin  resistant      37     125   29.6%\n",
       "11   single    caspofungin  sensitive      88     125   70.4%\n",
       "12   single            dox  sensitive     125     125  100.0%\n",
       "13   single     micafungin  resistant      43     125   34.4%\n",
       "14   single     micafungin  sensitive      82     125   65.6%\n",
       "15   single           none  sensitive     125     125  100.0%"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "counts = slpAA[slpAA.cscore <3].groupby(['seq_type','compound','sensres']).aa_seq.nunique().to_frame().reset_index()\n",
    "tots = slpAA[slpAA.cscore <3].groupby('seq_type').aa_seq.nunique().to_frame(name='totals').reset_index()\n",
    "perc = pd.merge(left=counts, right=tots, on='seq_type')\n",
    "perc['%'] = perc.apply(lambda row: f'{row.aa_seq/row.totals:.1%}', axis=1)\n",
    "perc"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a76cfd63-1e07-4f2a-b9bb-fb739b29b2d4",
   "metadata": {
    "tags": []
   },
   "source": [
    "### Show distributions only for high confidence single mutants"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "id": "7f35dfaf-6689-4d3a-ad09-ad556f81a346",
   "metadata": {},
   "outputs": [],
   "source": [
    "slpAA = pd.read_csv(f'{classified_outpath}/{\"_\".join([strain,locus,pool_type])}/aa_refined_classification.csv', index_col=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "id": "36d7307d-a4a4-4706-82ee-9823ca2bcaf4",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "graphdf = slpAA[(slpAA.Nham_aa <=1) & (slpAA.cscore < 3)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "id": "c263e650-bf5a-4646-8995-64a9912c49b5",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 350x190 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "list_drugs = ['none','dox','caspofungin','micafungin','anidulafungin']\n",
    "abrev_drugs = {'none':'ctrl',\n",
    "               'dox':'rep. ctrl',\n",
    "               'caspofungin':'CF',\n",
    "               'micafungin':'MF',\n",
    "               'anidulafungin':'AF'\n",
    "              }\n",
    "custom_palette = ['#C75DAB','#D691C1','#F1F1F1','#42B7B9','#009B9E']\n",
    "\n",
    "sns.set(rc = {'figure.figsize':(3.5,1.9),\n",
    "              'font.family':'Arial',\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':2, 'ytick.major.pad':2,\n",
    "              'xtick.bottom': True, 'ytick.left': True,\n",
    "              'xtick.major.size':2, 'ytick.major.size':2\n",
    "             },\n",
    "       style='whitegrid')\n",
    "\n",
    "fig = sns.boxplot(graphdf, x='s', y='compound', hue='refined_class',\n",
    "                  order=[x for x in list_drugs if x in graphdf.compound.unique()],\n",
    "                  hue_order=['resistant','intermediary','WT-like','slightly deleterious','deleterious'],\n",
    "                  palette=custom_palette, fliersize=0,\n",
    "                  width=1.5, # Width will draw wider boxes but strips have to be realigned manually\n",
    "                 )\n",
    "sns.stripplot(graphdf, x='s', y='compound', hue='refined_class',\n",
    "              order=[x for x in list_drugs if x in graphdf.compound.unique()],\n",
    "              hue_order=['resistant','intermediary','WT-like','slightly deleterious','deleterious'],\n",
    "              palette=custom_palette,\n",
    "              dodge=True, alpha=.5, s=3, linewidth=.5,\n",
    "              legend=False\n",
    "             )\n",
    "\n",
    "fig.set_yticks(np.arange(0,len(graphdf.compound.unique())),\n",
    "               [abrev_drugs[x] for x in list_drugs if x in graphdf.compound.unique()],\n",
    "              )\n",
    "\n",
    "fig.set(xlabel='Selection coefficient of NNK variants', ylabel='')\n",
    "fig.legend(bbox_to_anchor=(1,1), title=None)\n",
    "plt.savefig(f'{graph_outpath}/{\"_\".join([strain,locus,pool_type])}_distribution_effects.svg', format='svg', dpi=300)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8c862cbb-5b37-4cc1-ab91-2585c765a2c8",
   "metadata": {
    "tags": []
   },
   "source": [
    "### Rug plot with separate panels"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6b25838b-e821-41e7-85c8-beb596649bf7",
   "metadata": {},
   "source": [
    "This version means to highlight overall distributions, as well as separate the impact on protein function (none) from the impact on resistance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "id": "a20a4bbd-2d0a-4f0e-86ae-4c987bfdcdf0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAU4AAAC1CAYAAAAwcRYHAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjguMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/SrBM8AAAACXBIWXMAAA9hAAAPYQGoP6dpAAAcfklEQVR4nO3dfXAT190v8K8kS7bkF5CJwTbhPoHYUFqYXjwEyLTFBCcUG9OkJJlMpyVJ7xAaZmJ48uabtEnapkyh4CRNbIbJ0KeUhiaEhJQXB+fNEC5lwtvYDTY41IOdBPwGWCayLNkrrfb+YaRItmy09tq7Wn8/M5lIu2ePfkcyP53V7jnHIEmSBCIiippR7QCIiGINEycRkUxMnEREMjFxEhHJxMRJRCQTEycRkUxMnEREMsWpHYDSqqurIUkSzGaz2qEQUQzxer0wGAyYM2fODcvqrscpSRKUvKdfkiQIgqBonWoTRRGCIKCrqws+n0/tcBSjx88K0Ge7tNgmOblDdz3OQE9z9uzZitTndrtRV1eHrKws2Gw2RepUU1NTE3Jzc+HxeAAAVqsVR44cweTJk1WObPj09lkF6LFdWmxTTU1N1GV11+OkwTkcDng8HpSUlOCJJ56Ax+OBw+FQOyyimKK7HidFJysrCyaTSe0wiGISe5xERDIxcRIRycTESUQkk2YT59tvv43y8vKI+xYvXoxLly6NckRERL00mzirq6shCILaYRAR9aOZq+ovv/wyPvzwQ5hMJhQUFODQoUM4fvw4JkyYgIqKCnR0dODrr7/GunXr1A6ViMY4TfQ4P/roI5w8eRL79+/HP//5Txw5cgQLFy7E2rVrkZubCwBITk5GRUUFli5dqnK0RDTWaSJxnjhxAvn5+YiPj0d8fDzeeecdxMWFd4ajGT9KRDQaNJE4TSYTDAZD8HlzczO6urrCylit1tEOi4goIk0kznnz5uHjjz+GIAgQBAGPPvooOjs7IYqi2qEREfWjiYtDd955J86dO4d7770Xfr8f9957LzIyMrB582YkJiaqHd6o6OjpwX2HjwAA3r0jF/b4+KiP9fh8KPzkEACg/M7FABB8npmQgC05/ztY1uVyAQB8Ph+mTJkS3OZ0OiPWnZSUBKNRE9+vRJqhicQJAGvXrsXatWvDtuXn5wMACgoKwrYfOnRo1OIiIuqLXQkiIpmYOImIZNLMqTqNnObmZnR0dAAAGhoaVI6GKPYxcepc99WrKFz1f4IzvgO9t3bZ7faIydRutyMzM3PU4ySKJUycOuft7ITH48GmTZswbdo0AOHJ0Wq1ori4OFjearXitddeg91uh91ux/Tp01WJm0jLDJKWVktSQFVVFSRJgsViUaQ+SZLg9XphNpvDbtJXml+S0NbdDfj9sJvNsn58lgC09/QAACZcv40p8NwgivA6nZg0aVLElT9FUYTf7w8+bm9vDy5YZTAYkJGR0W8Ul1aN1mc12vTYLi22SRAEGAwG5OTk3LBsbPyLkEHpD6G1tRUAkJGRoWi9fRkNBmQMY3RUZp8virDn48f3K9/S0gKgf7u0snDWUIzWZzXa9NguLbbJYDBEnT901+NUWl5eHgCgsrJS5UiUpcd26bFNgD7bFett4u1IREQyscdJRCQTe5xERDIxcRIRycTESUQkExMnEZFMuruPs7q6GpIkRbzZm4hoIF6vFwaDIaplenSXOCVJgpI3CmhxhIMcoigOOJO+KIpISEiIyXZFEuuf1UD02C4ttklO3tBd4gz0NGfPnq1IfW63G3V1dcjKyoq5UTVNTU3Izc0Nm+AjVHx8PD766CNkZWWNcmQjI5Y/q8HosV1abFNNTU3UZXWXOOlbDocDHo8HpaWlyM7ODttXW1uLp556KjhDEhFFj4lzDMjOzu7XA+/u7lYpGqLYx6vqREQyMXESEcnExElEJJNmE+fFixfx61//OuK+0tJSlJaWjnJERES9NJs4m5ubcfHiRbXDICLqJ+qr6idOnMCmTZsgSRJuueUWrF+/Hn/4wx/wxRdfQBRFrFy5Evfffz/ee+89fPjhh3C5XLh8+TJyc3Pxm9/8ZtCbXI8fP44NGzbA7/dj4sSJKCkpwYsvvojm5ma88MILWLZsWdhrT506VZHGExENhazbkRobG3H48GGMGzcOL7/8MqZPn44NGzbA7Xbj5z//efCWlzNnzmD//v0YN24cHnzwQXzwwQfIz8+PWKcgCHjyySfx+uuvY9asWdi6dSt27dqFF154AWVlZXjxxRdx4sSJsNfmaToRqUlW4pw2bRrGjRsHAPjXv/4Fj8eDvXv3AgBcLhfOnz8PALjjjjuQlpYGACgoKMCpU6cGTJz/+c9/kJaWhlmzZgEA1qxZA6C3hzvQaxMRqUlW4rSGLCbm9/tRUlKC733vewCA9vZ2JCcno7y8PGxVRL/fD6Nx4J9S4+Liwk7ju7q6cO3atUFfm4hITUO+OLRgwQK89dZbkCQJDocDP/3pT3HhwgUAwNGjR+F0OtHT04P3338fP/zhDwesZ+rUqejo6EB9fT0AYOfOnfjrX/8Kk8kEn8831PCIiEbMkIdcPvbYY/j973+P5cuXw+fzYc2aNZg5cybq6uqQlpaGX/3qV3A4HCgsLMSiRYsAAI888gjWrl0bNvwvPj4eJSUlePbZZyEIAjIyMoIXglwuF5544gk88MADw25oJF1eL5J27Iy4z/XQL9Dl9WHSm2/3bqg+M2hdrod+gcTrE4x4fD4UfnIIAPDuHbm47/CRYLnyOxcDQHB/qL5lA+WvCQJ+8f/+NeBrxxuN2LVgXv+YXK7g/51OZ9i+wJeSz+cL7ktKSgoeAwDJycmambmGSEuiTpzz58/H/Pnzg8+TkpKwefPmiGWnTZuGjRs39tu+bdu2iOXnzp2Ld999t9/2AwcOhL1+QFFRUbRhExEpjpN8jHENDQ3BxzabDW63GwBgt9sxY8YMtcIi0jTFE+eKFSuwYsUKpaslhdntdlitVhQXF0fcb7Va8emnn+Lmm28e5ciItI89Th0Q2ttx9uzZfttDe5N9ZWZmory8PGw+zkCPs6GhAcXFxXA4HEycRBEwcd6A6do1GK+fvvblt9kgjh8/KnF429shujr7bfd1dqJ5SxnuE4SIx1mtVtjt9oj7MjMzkZmZGXze9+IQEUXGxInIyfFsbS2aL19GWtmrMHq9EY/zm8248ti6iMkzkOjO1U5A91dfBrfXXp+eP3RbQEtT5CUsWpub0fDcryENkByNFgu2bdsWMUHa7faw5EhEw2eQlFzZTAOqqqogSRIsFssNy0oAvvzmG5icTmCgt8FggJiYBBjDb8sxiCKMXV3wJyZCMpmQabMhUEIURVy5enXgOgdgMBhgSkkBQgYMTIiPhyAI6HA4YEpOhiGu/3edwWDEJNvwBwgYjUb4/X54vV60tbUhPT09qvdRK7S4AJgS9NguLbZJEAQYDAbk5OTcsKzuEqfSywO3tLQAADIyMhSpTyv02C49tgnQZ7u02CY5ywPrLnEqLS8vDwBQWVmpciTK0mO79NgmQJ/tivU2aXY+TiIirWKPk4hIJvY4iYhkYuIkIpKJiZOISCYmTiIimXQ3ckjp+ziJaGyQcx+n7hKnJElQ8kYBLY5wkEMURYiiGHFfoF3RMplMMJlMSoWmuFj/rAaix3ZpsU1y8obuEmcgEYTOMj8cbrcbdXV1yMrKgs1mU6TO0dLU1ITc3Fx4PB5F6rNarThy5AgmT56sSH1Ki+XPajB6bJcW21RzfR6JaOgucdK3HA4HPB4PSktLkZ2dHbavu7sbjY2NmDp1KhISEm5YV319PYqKiuBwODSbOIlGCxPnGJCdnd2vB+52u2E0GjFz5kzNfOMTxQpeVScikomJk4hIJp6qa1RTUxMcDsew6gisVU9EytJM4nS5XHjllVdw4sQJmEwmJCUl4amnnkJaWhqWLl2KW2+9Naz8Sy+9hKysyDOmxzolr4ZbrVakpqYqEBURBWgicfr9fqxevRrz5s3D3r17ERcXh+rqaqxZswZlZWWYOHEi9u3bp3aYo2awq+Fypaam8io4kcI0kThPnDiBtrY2rF27Fsbry0bMmTMHmzdv1szNsWqIdDWciNSnicR57tw5zJ49O5g0A370ox/h0qVLuHz5Mu6+++7g9sWLF2PdunWjHSYREQCNJM7AImEDGWun6kSkbZpInLNmzcLOnTshSVLYqfnWrVuRnJysYmRERP1p4j7OuXPnIi0tDX/+85/h8/kAAKdOncLOnTuRm5urcnREROE00eM0GAzYunUrNm7ciOXLlyMuLg4pKSnYsmXLmLs4JEkSXC5X8LEkSejs7ByR18rOzobP54PT6bxh2UBMLpcrWD4xMTE4W1JonImJiejq6gIANL3xOSDKn63q1odug2Vc/7Xi/V4R57ceAwDMWPMDAAg+BwAbAH+W2K9s9qoFqP/L8eBxRvPAszyJPT6cemQvAOC2bfcAQNhzU3xcxLKhvv/SUnz+5Af9Hgf0rQcABGc3qh4rBwDklBXCkjL4HAJ9XzuaY+SI1LZIcQ+Fv8cH76uNOINGxeocTZqJ1m63409/+lPEfYcOHRrlaIiIBqaZxEmxoaGhIfjYarWG9TgDvUyr1Rq8ef9Ka2O/HmeKLRmTUm4apYiJlMfESVGx2+2wWq0oLi4edl0J5nj89p7HMd6WMmAZzzkrzEn9Tzv9PhFftTYCAITaFKTaOSqKRh8TJ0UlMzMTlZWVaG5uDm6Lqsd5sD6sx3nN7cTv976CZ9/ZOPgL7ogiqB29r/WXhzexB0ujiomTojZ58mTY7fbg86guDlX5+p2q/8+qEjjdg1/wunn5dwfucb7zOQDAOycRa/97HZzuTkUSZ2BiFVHwocF5EQCQUNs7K3joc5Ml5OJQSNlQprqzwe2hjwP61gMAXldPsJzlbC3MSfFITU0Ne89JG5g4adRNSrnphonu1u/OGvCquuVo71V9Iav3VP/r9qawMs5zFtiSEsNO67vPJuFiyCm+MS78qnp7eztWrVrVf2KVn5QM/nwgPxvgcbT13Ne732q1YsuWLfjmm2/g9/vDZuvvm7QDyVYpkb4UEmprkJY+cczPf2CQlFzZTAOqqqogSRIsFosi9amxqFRPTw/a2tqQnp4Oi8Uy6Kiq0dR3hFffIbKBfaHlRJcwpNeKS4qHwRj5/fY6uwEABlscWlpaFFucz2AwIC0tDUajEUJHbwK12HuTd9/noQL7QpnHJcD7TXe/xwGR6oEkQbjWW84yPgF+ScKVK1cUXXxQCaHv01BJfgneb7phMhhhTUsGNHDXoSAIMBgMyMnJuWFZ3SVOpZcHbmlpAQBkZGQoUp9W6LFdemwToM92abFNcpYH1l3iVFpeXh4AoLKyUuVIlKXHdumxTYA+2xXrbdLEkEsioljCHicRkUzscRIRycTESUQkExMnEZFMTJxERDLpbuSQ0vdxEtHYIOc+Tt0lzsDkv0rWN9ojh+QQRRGiKA7p2EC7QplMpuD481ij9c9qqPTYLi22SU7e0F3iDCQCpZbVdbvdqKurQ1ZWFmw2myJ1KqWpqQm5ubn9x1cPg9VqxZEjR2JyLLKWP6vh0GO7tNimmpqaqMvqLnGOJQ6HAx6PB6WlpcjOzpZ1bHd3NxobGzF16tTgxBH19fUoKiqCw+GIycRJNFqYOHUgOztbdg/b7XbDaDRi5syZmvnGJ4oVvKpORCQTEycRkUxMnEREMqn+G+elS5eQl5eHFStWYMOGDcHtV69excKFC7FmzRqcPHkSra2tYb/FLVu2DKtXr1YjZCIa41RPnEDvCopHjx4Nu6+woqIC48ePD5ZZv3495s+fr1KERETf0sSpus1mw5w5c3Ds2LHgtg8++AB33XWXilEREUWmiR4nAOTn56OiogKLFi1CU1MTEhIScNNN3y7o9dxzz4Wdqu/bt0+NMImItJM477jjDmzcuBGCIODgwYMoKCgIW8Obp+pEpBWaOFUHeof6zZ07F8eOHcMnn3yCJUuWqB0SEVFEmulxAkBBQQFef/11TJw4EcnJyWqHMyKamprgcDgUqau+vl6Rekay3tTUVA7fJN3RVOJcuHAhnnnmGTz88MNqh6I4SZJw/vx5FBYWKj4ph9lshtPplH1sdnY2fD5f8Fiz2Qyr1YqioiJF4xto0pDuqy40vlk1cHyrFqD+L8cjPr/1odtwYcep4PY4mwV+rwjbsS58dew0Zqz5AYzmoc/y5PeKOL+192JloK7QbaGxRPNaYo8Ppx7ZCwC4bds9MMXHhW37/ktL8fmTH/R7fNu2ewAAktePM48eHPD4vnLKClH1WHnYttB6c8oKYUlJiOq9iBR7NPsGq2fWq71nlP4eH44/+m5Ux2uJ6lHefPPNOHToEADAYrHg9OnTwX1K/gPWgo6ODng8Hrz22muYPn06gN6E2tXVFSyTmJgIAGHbBmO325GZmalIfJmZmSgvL0dHR0fUx1w5WA+Ikafj+rq9CRvKt3DSENId1RPnWBQ6KYckSejs7AzuC/xEEbptNGVmZspKxM3VIiSffwQjItIeJk4acQP9ZtpzzY3m1sYBj+s+m4SLIfttzVMUj41oKJg4acSk2JKH95vpjvCn1t1W/OXhTZiUclPk8jpzpduBmtoaeP0+NDRcgOhsAQAk1NbAZImDKPjQ4LwY8VjL2dqwfcmWxFGJeaxg4qQRMynlJhz+5BCudX4TcX/PNTeaK74Y8PgpP52Ni//snZU78Hup0905JhLnlW4H1n22AT0/Efrv/EnJjSu4L7xMvNGCj1sWKRMcMXHSyJo8eTKmmP9XxH3dV11IrO4Z8Njs781CwmeusG1ftzcBADznrLh0/TS++2wS4hLMcLu60HZ9m1CbAmPcMK6q+0R81aeu0G2hPyNE81qhvcNIPUZT3dmwx3XXGtDjF/Dqy3/Gf027BQ31FyDu7u1xfvf5RcHjz/3h04iv953/+yN88aejAIBL7ja8evYNnKo6Db+zDUBvj9ScFB/VexEp9mj2DVaP4dxZfNV0Eb5uAQ3OizHXIzZISq5spgFVVVWQJAkWi0WR+pRcVKqnpwdtbW1IT08Pi8/v//biitFo7LdNy0RXhB5RCPMgt7xIfgk+18CJ05wcD29n737RL+JKp0PRhfi0zgADMjIzYDKZIAgC4PQBAOJvSgQMACSg52rkuy8sE2wQ2t0AAFHyo737GiRo970LtDUuTr2+nCAIMBgMyMnJuWFZ3SVOpZcHbmnp/ZbPyMhQpD6t0GO79NgmQJ/t0mKb5CwPrLvEqbS8vDwAQGVlpcqRKEuP7dJjmwB9tivW26SZsepERLGCPU4iIpnY4yQikomJk4hIJiZOIiKZmDiJiGTS3cghpe/jJKKxQc59nLpLnJIkKTq6RMmRQyNBFEWIojikY0OXY+7LZDLBZBr6kEU1aP2zGio9tkuLbZKTN3SXOAOJIDDf5XC53W7U1dUhKysrbJVNLWhqakJubq6iM8oHDDZzu1Zp+bMaDj22S4ttqqmpibqs7hLnWOJwOODxeFBaWors7GxZx3Z3d6OxsRFTp05FQkL4ePL6+noUFRVx5naiATBx6kDojPLRcrvdMBqNmDlzpma+8YliBa+qExHJxMRJRCQTEycRkUya+43zyy+/xNKlS/HHP/4RK1asAACsXLkSra2tYb/FLVu2DKtXr1YrTCIawzSXOPfs2YMlS5Zg165dwcQJAOvXr8f8+fNVjIyIqJemTtVFUcTevXtRVFSElpYW1NXVqR0SEVE/mupxHj16FOPHj0d2djby8/Px1ltv4cUXXwQAPPfcc2Gn6vv27VMrTCIa4zSVOPfs2YPCwkIAQGFhIR566CEUFxcD4Km6Gurr60ek3tTUVN5YTzFNM4nT4XDg8OHD+Pzzz/H2228D6B1LzZ7l6EtNTYXVakVRUdGI1B+LwzmJQmkmce7fvx85OTn4+9//Htz2t7/9Dbt27cL48ePVC2wMmjx5Mo4cOQKHw6F43RzOSXqgmcS5Z8+efj2c++67D2VlZWhoaFApquGTJAmdnZ0AgOTk5H4zwYTuD4hULtIxLpcLAOByueB0OmXHlp2dDZ/PF/HY5ORkJCcny64zwGazwe1299s+UMzJycmQfH6c33ps0HpnrPkBjObeWZv8XjFYfsaaH0Qs37dM4NgblY3mtfrWG+m1+tY5UBxy4hxOGwFA7PHh1CN7AQC3bbsHpvi4iPsi7Y8Vg7VRCZp5Rw4cONBvW1JSEk6fPq1CNEREA9PU7UhERLGAiZOISCbNnKrT2NL3d+vExERIoh9ftTYOepxQmwJj3PXfHX1isLxQm4LE5KSRCZaoDyZOGlV2ux1WqzV4f65sOwbebrVaUVZWhpkzZw45PqJoMHGOkubmZgiCEPGqeldXV9i2xMTEG15V7+rqism7DTIzM1FeXo6Ojo6w7cEe5zufD3r8f93//fAe5/Xy3jmJWPvf63D27FlMmDAhOKt9315p4NhIQsv2Ld+3HgD96o30Wn3rDN03UgMBmpqaBr2VTBR8aHBeBAAk1NbAZImLuK/vfg5c+JZBUnJlMw2oqqqCJEmwWCyK1KfEolJerxetra2KLiIXYDAYkJ6errmF1YxGI/x+v6zyAOB1dg9azpwSvsxHoLzBFoeWlpYReY9HisFgQFpaWrDtvi4BABCXGP63K0kSfD4f4uLiBvwbDBxrTIjDlatXRuxvLTTe4YimTcMldHhgMhhhTUsGoniJQMcmJyfnhmV11+NU+kNobW0FAGRkZAy5DrPZjClTpigVkiJaWloADK9dNzKUf2B9E6Oc8nFxvX/OI9mmkTTQl300n1XosVr7W4tkNP7+LJPkdZ4MBkPU+UN3PU6l5eXlAQAqKytVjkRZemyXHtsE6LNdsd4m3o5ERCQTe5xERDKxx0lEJBMTJxGRTEycREQyMXHKsGfPHjzzzDNqhzFsFRUVWLZsGZYsWYKysjK1w1GMy+XC8uXLcenSJbVDUcz27dtRWFiIwsJCPPvssxAEQe2QFFFSUoKCggIsW7YM27dvVzsc2Zg4o9DT04PNmzdj/fr1aocybFeuXMGmTZvwxhtv4P3338fp06dx9OhRtcMaturqavzsZz9DY+PgY91jyZkzZ/Dee+9h9+7dOHDgAHw+H9588021wxq2Tz/9FP/+979x4MAB7NmzB2+88UbMjYJj4ozC8ePHIUkSnn76abVDGbZjx45hwYIFSE1Nhdlsxj333IODBw+qHdaw7d69G7/97W8xceJEtUNRTEpKCp5//nnYbDYYDAZ85zvfQXNzs9phDduiRYuwfft2mEwmtLe3QxTFsIUYYwETZxRyc3NRXFwcHP8cyy5fvhyWXCZOnIi2tjYVI1LGhg0bMHfuXLXDUNQtt9yCefPmAQDa29vxj3/8I3jjeKwzm8145ZVXUFBQgNtvvx2TJk1SOyRZmDhDVFRUYOHChWH/PfDAA2qHpahI48dHaqwwKePSpUt48MEHcf/99+tqpdfHH38cn332GZqbm7F79261w5FFd2PVhyM/Px/5+flqhzGi0tPTcfLkyeDzy5cvIz09XcWIaDB1dXVYvXo1Vq9ejZUrV6odjiLq6+vh9/sxY8YM2Gw2/PjHP8b58+fVDksW9jjHmNtvvx3Hjx/H1atX4fV6sX//fixatEjtsCgCh8OBVatW4fnnn9dN0gSACxcu4He/+x28Xi8EQcDHH38ccz+zsMc5xkyaNAlPP/00fvnLX0IQBCxevBh33XWX2mFRBDt27IDL5cKWLVuwZcsWAL0XVh5//HGVIxuepUuX4ty5c7j77rthMpmwdOlSFBQUqB2WLByrTkQkE0/ViYhkYuIkIpKJiZOISCYmTiIimZg4iYhkYuIkIpKJiZOISCYmTiIimZg4iYhkYuIkIpLp/wP9iFx6CL7NJgAAAABJRU5ErkJggg==",
      "text/plain": [
       "<Figure size 350x190 with 5 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "list_drugs = ['none','dox','caspofungin','micafungin','anidulafungin']\n",
    "abrev_drugs = {'none':'ctrl',\n",
    "               'dox':'rep. ctrl',\n",
    "               'caspofungin':'CF',\n",
    "               'micafungin':'MF',\n",
    "               'anidulafungin':'AF'\n",
    "              }\n",
    "lclasses = ['resistant','intermediary','WT-like','slightly deleterious','deleterious']\n",
    "custom_palette = ['#C75DAB','#D691C1','#F1F1F1','#42B7B9','#009B9E']\n",
    "\n",
    "rc_dict = {'figure.figsize':(3.5,1.9),\n",
    "      'font.family':'Arial',\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':2, 'ytick.major.pad':2,\n",
    "      'xtick.bottom': True, 'ytick.left': False,\n",
    "      'xtick.major.size':2, 'ytick.major.size':2\n",
    "     }\n",
    "plt.rcParams.update(rc_dict)\n",
    "fig, ax = plt.subplots(len(graphdf.compound.unique()),1, sharex=True,\n",
    "                       gridspec_kw = {'hspace':.4})\n",
    "\n",
    "for i, c in enumerate([x for x in list_drugs if x in graphdf.compound.unique()]):\n",
    "    axdf = graphdf[graphdf.compound == c]\n",
    "    ax[i].hist(axdf.s, histtype='step', color='k')\n",
    "    ax[i].scatter(axdf.s, np.zeros(axdf.s.shape), marker=2, c=axdf.refined_class.map(dict(zip(lclasses, custom_palette))))\n",
    "    ax[i].set_yticks([])\n",
    "    ax[i].set_ylabel(abrev_drugs[c], rotation=0, ha='right', va='center')\n",
    "\n",
    "plt.savefig(f'{graph_outpath}/{\"_\".join([strain,locus,pool_type])}_distribution_effects_rug.svg', format='svg', dpi=300)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0698b2aa-0d1a-4332-a26f-df41d0bb0a70",
   "metadata": {
    "tags": []
   },
   "source": [
    "### Show distributions only for high confidence ortho mutants"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "id": "7b94408a-a741-46eb-add4-c7b364385bda",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "graphdf = slpAA[(slpAA.Nham_aa >=2) & (slpAA.cscore < 3)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "id": "9486b450-1a36-4743-aacc-d1a1f808ea98",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 350x190 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "list_drugs = ['none','dox','caspofungin','micafungin','anidulafungin']\n",
    "abrev_drugs = {'none':'ctrl',\n",
    "               'dox':'rep. ctrl',\n",
    "               'caspofungin':'CF',\n",
    "               'micafungin':'MF',\n",
    "               'anidulafungin':'AF'\n",
    "              }\n",
    "custom_palette = ['#C75DAB','#D691C1','#F1F1F1','#42B7B9','#009B9E']\n",
    "\n",
    "sns.set(rc = {'figure.figsize':(3.5,1.9),\n",
    "              'font.family':'Arial',\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':2, 'ytick.major.pad':2,\n",
    "              'xtick.bottom': True, 'ytick.left': True,\n",
    "              'xtick.major.size':2, 'ytick.major.size':2\n",
    "             },\n",
    "       style='whitegrid')\n",
    "\n",
    "fig = sns.boxplot(graphdf, x='s', y='compound', hue='refined_class',\n",
    "                  order=[x for x in list_drugs if x in graphdf.compound.unique()],\n",
    "                  hue_order=['resistant','intermediary','WT-like','slightly deleterious','deleterious'],\n",
    "                  palette=custom_palette, fliersize=0,\n",
    "                  width=1.5, # Width will draw wider boxes but strips have to be realigned manually\n",
    "                 )\n",
    "sns.stripplot(graphdf, x='s', y='compound', hue='refined_class',\n",
    "              order=[x for x in list_drugs if x in graphdf.compound.unique()],\n",
    "              hue_order=['resistant','intermediary','WT-like','slightly deleterious','deleterious'],\n",
    "              palette=custom_palette,\n",
    "              dodge=True, alpha=.5, s=3, linewidth=.5,\n",
    "              legend=False\n",
    "             )\n",
    "\n",
    "fig.set_yticks(np.arange(0,len(graphdf.compound.unique())),\n",
    "               [abrev_drugs[x] for x in list_drugs if x in graphdf.compound.unique()],\n",
    "              )\n",
    "\n",
    "fig.set(xlabel='Selection coefficient of ortholog variants', ylabel='')\n",
    "fig.legend(bbox_to_anchor=(1,1), title=None)\n",
    "plt.savefig(f'{graph_outpath}/{\"_\".join([strain,locus,pool_type])}_ortho_distribution_effects.svg', format='svg', dpi=300)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "id": "90164eed-bc9d-436d-9b44-1876a3225bc9",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 350x190 with 5 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "list_drugs = ['none','dox','caspofungin','micafungin','anidulafungin']\n",
    "abrev_drugs = {'none':'ctrl',\n",
    "               'dox':'rep. ctrl',\n",
    "               'caspofungin':'CF',\n",
    "               'micafungin':'MF',\n",
    "               'anidulafungin':'AF'\n",
    "              }\n",
    "lclasses = ['resistant','intermediary','WT-like','slightly deleterious','deleterious']\n",
    "custom_palette = ['#C75DAB','#D691C1','#F1F1F1','#42B7B9','#009B9E']\n",
    "\n",
    "rc_dict = {'figure.figsize':(3.5,1.9),\n",
    "      'font.family':'Arial',\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':2, 'ytick.major.pad':2,\n",
    "      'xtick.bottom': True, 'ytick.left': False,\n",
    "      'xtick.major.size':2, 'ytick.major.size':2\n",
    "     }\n",
    "plt.rcParams.update(rc_dict)\n",
    "fig, ax = plt.subplots(len(graphdf.compound.unique()),1, sharex=True,\n",
    "                       gridspec_kw = {'hspace':.4})\n",
    "\n",
    "for i, c in enumerate([x for x in list_drugs if x in graphdf.compound.unique()]):\n",
    "    axdf = graphdf[graphdf.compound == c]\n",
    "    ax[i].hist(axdf.s, histtype='step', color='k')\n",
    "    ax[i].scatter(axdf.s, np.zeros(axdf.s.shape), marker=2, c=axdf.refined_class.map(dict(zip(lclasses, custom_palette))))\n",
    "    ax[i].set_yticks([])\n",
    "    ax[i].set_ylabel(abrev_drugs[c], rotation=0, ha='right', va='center')\n",
    "\n",
    "plt.savefig(f'{graph_outpath}/{\"_\".join([strain,locus,pool_type])}_ortho_distribution_effects_rug.svg', format='svg', dpi=300)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "36c51377-6909-4e8f-9725-a534de723246",
   "metadata": {
    "tags": []
   },
   "source": [
    "## UpSet plot on classified variants"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "id": "b8fd93d5-3b5a-4930-b503-b208c8d11d20",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "upsetdf = slpAA[(slpAA.Nham_aa <=1) & (slpAA.cscore < 3)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "id": "429b32ac-db7f-4fc8-9529-4ac5fa9cb687",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "parse_dict = {'resistant':True, 'sensitive':False}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "id": "dffec941-f90a-45a3-a3c1-6131ddbbcfb1",
   "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>AF</th>\n",
       "      <th>CF</th>\n",
       "      <th>dox</th>\n",
       "      <th>MF</th>\n",
       "      <th>ctrl</th>\n",
       "      <th>max_s</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>0.448349</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>1.305855</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>1.214658</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>0.630822</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>0.703303</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>130</th>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>0.136393</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>131</th>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>1.457989</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>132</th>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>-0.005603</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>133</th>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>0.144368</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>134</th>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>1.245412</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>135 rows × 6 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        AF     CF    dox     MF   ctrl     max_s\n",
       "0    False  False  False   True  False  0.448349\n",
       "1     True   True  False   True  False  1.305855\n",
       "2     True   True  False   True  False  1.214658\n",
       "3    False  False  False   True  False  0.630822\n",
       "4    False  False  False   True  False  0.703303\n",
       "..     ...    ...    ...    ...    ...       ...\n",
       "130  False  False  False  False  False  0.136393\n",
       "131   True   True  False   True  False  1.457989\n",
       "132  False  False  False  False  False -0.005603\n",
       "133  False  False  False  False  False  0.144368\n",
       "134  False   True  False   True  False  1.245412\n",
       "\n",
       "[135 rows x 6 columns]"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "BYnnkCompWide = upsetdf.pivot_table(index='aa_seq',\n",
    "                                    columns='compound',\n",
    "                                    values=['s','sensres'],\n",
    "                                    aggfunc='first'\n",
    "                                   ).reset_index()\n",
    "BYnnkCompWide['max_s'] = BYnnkCompWide['s'].max(axis=1)\n",
    "bincat = BYnnkCompWide['sensres'].replace(parse_dict)\n",
    "bincat.rename(columns={'anidulafungin':'AF',\n",
    "                       'caspofungin':'CF',\n",
    "                       'micafungin':'MF',\n",
    "                       'none':'ctrl'\n",
    "                      }, inplace=True)\n",
    "compounds = bincat.columns.tolist()\n",
    "bincat = pd.concat([bincat, BYnnkCompWide[['max_s']].droplevel(1, axis=1)], axis=1)\n",
    "bincat"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "id": "1df0c819-1d8f-43b3-9902-5158dba0a254",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([False])"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bincat.ctrl.unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "id": "b7b83814-44bc-492c-bc52-22b89c7d5e49",
   "metadata": {},
   "outputs": [],
   "source": [
    "bincat.to_csv(f'{classified_outpath}/{\"_\".join([strain,locus,pool_type])}/upsetplot.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "id": "ee71e415-be6f-4878-9213-dc0eace85017",
   "metadata": {},
   "outputs": [],
   "source": [
    "bincat = pd.read_csv(f'{classified_outpath}/{\"_\".join([strain,locus,pool_type])}/upsetplot.csv', index_col=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "id": "2e517894-90be-4292-a36c-e3f2a89cb8cd",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "No support for this hotspot\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 700x380 with 5 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Manuscript version\n",
    "# For this plot, all sizes are doubled and the graph is ultimately reduced by 50% for the final panel\n",
    "\n",
    "rc_dict = {'font.family':'Arial',\n",
    "           'font.size':16,\n",
    "           'axes.linewidth': 2, 'grid.linewidth': 2,\n",
    "           'axes.labelsize':16,'axes.titlesize':16,\n",
    "           'xtick.labelsize':16, 'ytick.labelsize':16,\n",
    "           'xtick.major.pad':2, 'ytick.major.pad':2,\n",
    "           'xtick.major.size':4, 'ytick.major.size':4\n",
    "          }\n",
    "\n",
    "upset_obj = UpSet(from_indicators(compounds, data=bincat),\n",
    "                  orientation='vertical',\n",
    "                  facecolor='#cd9abe', other_dots_color='#F1F1F1',\n",
    "                  #show_percentages=True,\n",
    "                  show_counts=True,\n",
    "                  #min_subset_size=10,\n",
    "                  sort_by='cardinality',\n",
    "                  element_size=None,\n",
    "                  intersection_plot_elements=3, # height/width of intersection barplot in matrix elements # needs to be 0 to be able to draw stacked_barplot\n",
    "                  totals_plot_elements = 2, # width/height of totals barplot in matrix elements\n",
    "                 )\n",
    "\n",
    "upset_obj.add_catplot(value='max_s', kind='box',\n",
    "                      color='grey', fliersize=3, linewidth=2, linecolor='k',\n",
    "                      elements=5 # height in number of matrix elements\n",
    "                      )\n",
    "\n",
    "fig = plt.figure(figsize=(7, 3.8))\n",
    "\n",
    "with plt.rc_context(rc_dict):\n",
    "    d = upset_obj.plot(fig=fig) # Assigns all plots to a dictionary containing axes subplots - same keys as gridspec returned by upset_obj.make_grid()\n",
    "\n",
    "axt = d['totals']\n",
    "axm = d['matrix']\n",
    "ax0 = d['intersections'] # Key corresponding to stacked barplot ('intersections' = intersection barplot)\n",
    "ax1 = d['extra1'] # Key corresponding to additional plot, in this case catplot\n",
    "\n",
    "for i in [0,1,2]:\n",
    "    axt.get_children()[i].set_color('#cd9abe')\n",
    "    axt.get_children()[i].set_edgecolor('#ba6aa5')\n",
    "    axt.get_children()[i].set_linewidth(2)\n",
    "\n",
    "axm.set_xticklabels(axm.get_xticklabels(), rotation=0, ha='center')\n",
    "axm.get_children()[0].set_edgecolor('#ba6aa5')\n",
    "axm.get_children()[0].set_linewidth(2)\n",
    "\n",
    "ax0.set_xlabel('# Variants') # (unique aminoacid sequences)\n",
    "ax1.set_xlabel('Max(s)')\n",
    "\n",
    "if locus=='FKS1-HS1':\n",
    "    for i in [0,1,2,3,4,5,6,7]:\n",
    "        ax0.get_children()[i].set_linewidth(2)\n",
    "        if i <=5:\n",
    "            ax0.get_children()[i].set_color('grey')\n",
    "            ax0.get_children()[i].set_edgecolor('k')\n",
    "        elif i==6:\n",
    "            ax0.get_children()[i].set_color('#F1F1F1')\n",
    "            ax0.get_children()[i].set_edgecolor('#b0b0b0')\n",
    "        else:\n",
    "            ax0.get_children()[i].set_edgecolor('#ba6aa5')\n",
    "    \n",
    "    ax1.get_children()[42].set_facecolor('#F1F1F1')\n",
    "    ax1.get_children()[49].set_facecolor('#cd9abe')\n",
    "\n",
    "elif locus=='FKS1-HS2':\n",
    "    for i in [0,1,2,3,4,5,6]:\n",
    "        ax0.get_children()[i].set_linewidth(2)\n",
    "        if i <=4:\n",
    "            ax0.get_children()[i].set_color('grey')\n",
    "            ax0.get_children()[i].set_edgecolor('k')\n",
    "        elif i==5:\n",
    "            ax0.get_children()[i].set_edgecolor('#ba6aa5')\n",
    "        else:\n",
    "            ax0.get_children()[i].set_color('#F1F1F1')\n",
    "            ax0.get_children()[i].set_edgecolor('#b0b0b0')\n",
    "    ax1.get_children()[35].set_facecolor('#cd9abe')\n",
    "    ax1.get_children()[42].set_facecolor('#F1F1F1')\n",
    "\n",
    "else:\n",
    "    print('No support for this hotspot')\n",
    "\n",
    "plt.savefig(f'{graph_outpath}/upsetplot_resistance_overlap_{strain}_{locus}.svg', format='svg', dpi=300)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "146ad688-a83c-4562-8705-0b82e7043e7d",
   "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
}