call_strong_escape_sites.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Call sites of \"strong escape\"\n",
    "This Python Jupyter notebook call sites of srong escape from each antibody / sera"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Import Python modules:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "\n",
    "from dms_variants.constants import CBPALETTE\n",
    "\n",
    "from IPython.display import display, HTML\n",
    "\n",
    "import pandas as pd\n",
    "\n",
    "from plotnine import *\n",
    "\n",
    "import yaml"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Versions of key software:"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Read the configuration file:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "with open('config.yaml') as f:\n",
    "    config = yaml.safe_load(f)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Create output directory:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "os.makedirs(config['escape_profiles_dir'], exist_ok=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Extract from configuration what we will use as the site- and mutation-level metrics:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "At site level, quantifying selection by site_total_escape_frac_epistasis_model\n",
      "At mutation level, quantify selection by mut_escape_frac_epistasis_model\n"
     ]
    }
   ],
   "source": [
    "site_metric = config['site_metric']\n",
    "mut_metric = config['mut_metric']\n",
    "\n",
    "print(f\"At site level, quantifying selection by {site_metric}\")\n",
    "print(f\"At mutation level, quantify selection by {mut_metric}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Read the escape fractions.\n",
    "We only retain the **average** of the libraries for plotting here, not the individual libraries.\n",
    "Also, we work in the full-Spike rather than RBD numbering, which means we use `label_site` as `site` (and so rename as such below):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Reading escape fractions from results/escape_scores/escape_fracs.csv\n",
      "First few lines of escape-fraction data frame with sample-information added:\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th>condition</th>\n",
       "      <th>site</th>\n",
       "      <th>wildtype</th>\n",
       "      <th>mutation</th>\n",
       "      <th>protein_chain</th>\n",
       "      <th>protein_site</th>\n",
       "      <th>mut_escape_frac_epistasis_model</th>\n",
       "      <th>mut_escape_frac_single_mut</th>\n",
       "      <th>site_total_escape_frac_epistasis_model</th>\n",
       "      <th>site_total_escape_frac_single_mut</th>\n",
       "      <th>site_avg_escape_frac_epistasis_model</th>\n",
       "      <th>site_avg_escape_frac_single_mut</th>\n",
       "      <th>nlibs</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>CAB-A17_139</td>\n",
       "      <td>331</td>\n",
       "      <td>N</td>\n",
       "      <td>A</td>\n",
       "      <td>E</td>\n",
       "      <td>331</td>\n",
       "      <td>0.001006</td>\n",
       "      <td>0.000179</td>\n",
       "      <td>0.01628</td>\n",
       "      <td>0.003478</td>\n",
       "      <td>0.001017</td>\n",
       "      <td>0.000217</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>CAB-A17_139</td>\n",
       "      <td>331</td>\n",
       "      <td>N</td>\n",
       "      <td>D</td>\n",
       "      <td>E</td>\n",
       "      <td>331</td>\n",
       "      <td>0.001013</td>\n",
       "      <td>0.000167</td>\n",
       "      <td>0.01628</td>\n",
       "      <td>0.003478</td>\n",
       "      <td>0.001017</td>\n",
       "      <td>0.000217</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>CAB-A17_139</td>\n",
       "      <td>331</td>\n",
       "      <td>N</td>\n",
       "      <td>E</td>\n",
       "      <td>E</td>\n",
       "      <td>331</td>\n",
       "      <td>0.000979</td>\n",
       "      <td>0.000237</td>\n",
       "      <td>0.01628</td>\n",
       "      <td>0.003478</td>\n",
       "      <td>0.001017</td>\n",
       "      <td>0.000217</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>CAB-A17_139</td>\n",
       "      <td>331</td>\n",
       "      <td>N</td>\n",
       "      <td>F</td>\n",
       "      <td>E</td>\n",
       "      <td>331</td>\n",
       "      <td>0.001090</td>\n",
       "      <td>0.000095</td>\n",
       "      <td>0.01628</td>\n",
       "      <td>0.003478</td>\n",
       "      <td>0.001017</td>\n",
       "      <td>0.000217</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>CAB-A17_139</td>\n",
       "      <td>331</td>\n",
       "      <td>N</td>\n",
       "      <td>G</td>\n",
       "      <td>E</td>\n",
       "      <td>331</td>\n",
       "      <td>0.001025</td>\n",
       "      <td>0.000109</td>\n",
       "      <td>0.01628</td>\n",
       "      <td>0.003478</td>\n",
       "      <td>0.001017</td>\n",
       "      <td>0.000217</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "print(f\"Reading escape fractions from {config['escape_fracs']}\")\n",
    "escape_fracs = (pd.read_csv(config['escape_fracs'])\n",
    "                .query('library == \"average\"')\n",
    "                .drop(columns=['site', 'selection', 'library'])\n",
    "                .rename(columns={'label_site': 'site'})\n",
    "                )\n",
    "print('First few lines of escape-fraction data frame with sample-information added:')\n",
    "display(HTML(escape_fracs.head().to_html(index=False)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now identify sites of \"strong\" selection using three criteria for each threshold.\n",
    "\n",
    " 1. The selection at a site has to exceed the median selection across all sites for that condition (e.g., antibody or sera) by some threshold.\n",
    " \n",
    " 2. The selection at a site has to be greater than some fraction of the maximum selection observed at any site for that condition (e.g., antibody or sera).\n",
    " \n",
    " 3. The selection at a site has to exceed some absolute minimum."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Calling sites of strong escape for the default calling parameters:\n",
      "Here are criteria used to call sites of strong escape for the default params:\n",
      "  1. Selection at site exceeds median by >10 fold\n",
      "  2. Selection at site is >0.1 of the max for any site\n",
      "  3. Selection at site is >0\n",
      "\n",
      "Calling sites of strong escape for the sensitive calling parameters:\n",
      "Here are criteria used to call sites of strong escape for the sensitive params:\n",
      "  1. Selection at site exceeds median by >5 fold\n",
      "  2. Selection at site is >0.05 of the max for any site\n",
      "  3. Selection at site is >0\n",
      "\n",
      "Calling sites of strong escape for the sensitive_max_mut calling parameters:\n",
      "Here are criteria used to call sites of strong escape for the sensitive_max_mut params:\n",
      "  1. Selection at site exceeds median by >5 fold\n",
      "  2. Selection at site is >0.05 of the max for any site\n",
      "  3. Selection at site is >0\n",
      "Sites are also called as significant if they have **any** mutation with escape fraction > 0.5\n"
     ]
    }
   ],
   "source": [
    "site_threshold_df = pd.DataFrame()\n",
    "\n",
    "for threshold, params in config['strong_escape_sites_calling_params'].items():\n",
    "    print(f\"\\nCalling sites of strong escape for the {threshold} calling parameters:\")\n",
    "    \n",
    "    exceed_median = params['exceed_median']\n",
    "    frac_max = params['frac_max']\n",
    "    min_value = params['min_value']\n",
    "\n",
    "    print(f\"Here are criteria used to call sites of strong escape for the {threshold} params:\\n\"\n",
    "          f\"  1. Selection at site exceeds median by >{exceed_median} fold\\n\"\n",
    "          f\"  2. Selection at site is >{frac_max} of the max for any site\\n\"\n",
    "          f\"  3. Selection at site is >{min_value}\")\n",
    "    \n",
    "    if 'min_value_mut' in params:\n",
    "        min_value_mut = params['min_value_mut']\n",
    "        print(f\"Sites are also called as significant if they have **any** mutation with escape fraction > {min_value_mut}\")\n",
    "    else:\n",
    "        min_value_mut = None\n",
    "    \n",
    "    site_threshold_df = site_threshold_df.append(\n",
    "        escape_fracs\n",
    "        .assign(max_mut=lambda x: x.groupby(['condition', 'site'])[mut_metric].transform('max'))\n",
    "        [['condition', 'site', site_metric, 'max_mut']]\n",
    "        .drop_duplicates()\n",
    "        .assign(\n",
    "            threshold=threshold,\n",
    "            median=lambda x: x.groupby('condition')[site_metric].transform('median'),\n",
    "            max=lambda x: x.groupby('condition')[site_metric].transform('max'),\n",
    "            exceed_median=lambda x: x['median'] * exceed_median,\n",
    "            meet_exceed_median=lambda x: x[site_metric] > x['exceed_median'],\n",
    "            frac_max=lambda x: x['max'] * frac_max,\n",
    "            meet_frac_max=lambda x: x[site_metric] > x['frac_max'],\n",
    "            min_value=lambda x: min_value,\n",
    "            meet_min_value=lambda x: x[site_metric] > x['min_value'],\n",
    "            meet_mut_min_value=lambda x: False if min_value_mut is None else x['max_mut'] > min_value_mut,\n",
    "            strong_site=lambda x: (x['meet_exceed_median'] & x['meet_frac_max'] & x['meet_min_value']) | x['meet_mut_min_value'],\n",
    "            n_strong_sites=lambda x: x.groupby('condition')['strong_site'].transform('sum'),\n",
    "            )\n",
    "        )\n",
    "                    \n",
    "# orders for plotting\n",
    "site_threshold_df = (\n",
    "        site_threshold_df\n",
    "        .assign(condition=lambda x: pd.Categorical(x['condition'],\n",
    "                                                   reversed(x['condition'].unique()),\n",
    "                                                   ordered=True),\n",
    "                )\n",
    "        )"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now plot the selection for all sites for all conditions, indicating which sites are strong escape and using lines to draw the three thresholds:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1200x75 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "max_sel = site_threshold_df[site_metric].max()  # max y-value plotted\n",
    "\n",
    "p = (ggplot(site_threshold_df.assign(alpha=lambda x: x['strong_site'].astype(int))) +\n",
    "     aes(site_metric, 'condition', alpha='alpha', fill='strong_site') +\n",
    "     geom_jitter(size=2.5, height=0.1, width=0, stroke=0, random_state=1) +\n",
    "     scale_fill_manual(values=CBPALETTE) +\n",
    "     scale_alpha(range=(0.15, 0.5), guide=None) +\n",
    "     theme(figure_size=(4 * site_threshold_df['threshold'].nunique(),\n",
    "                        0.25 * site_threshold_df['condition'].nunique()),\n",
    "           legend_position='top',\n",
    "           ) +\n",
    "     xlab('site-level selection') +\n",
    "     expand_limits(x=1.2 * max_sel) +\n",
    "     geom_text(data=site_threshold_df[['condition', 'threshold', 'n_strong_sites']].drop_duplicates(),\n",
    "               mapping=aes(y='condition', label='n_strong_sites'),\n",
    "               inherit_aes=False,\n",
    "               x=1.1 * max_sel,\n",
    "               color=CBPALETTE[1],\n",
    "               alpha=1,\n",
    "               size=8,\n",
    "               ha='left',\n",
    "               va='center',\n",
    "               ) +\n",
    "     geom_linerange(data=site_threshold_df\n",
    "                         .melt(id_vars=['condition', 'threshold'],\n",
    "                               value_vars=['exceed_median', 'frac_max', 'min_value'],\n",
    "                               var_name='threshold_type',\n",
    "                               value_name='threshold_line')\n",
    "                         .reset_index()\n",
    "                         .drop_duplicates()\n",
    "                         .assign(ymin=lambda x: x['condition'].cat.codes + 1 - 0.3,\n",
    "                                 ymax=lambda x: x['condition'].cat.codes + 1 + 0.3),\n",
    "                   mapping=aes(x='threshold_line', ymin='ymin', ymax='ymax', color='threshold_type'),\n",
    "                   inherit_aes=False,\n",
    "                   size=0.1) +\n",
    "     scale_color_manual(values=CBPALETTE[2: ]) +\n",
    "     facet_wrap('~ threshold', nrow=1)\n",
    "     )\n",
    "\n",
    "_ = p.draw()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now get the strong sites for each condition, and write to a CSV file:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of sites of strong escape for each threshold:\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>n_sites</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>threshold</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>default</th>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sensitive</th>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sensitive_max_mut</th>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Writing to results/escape_profiles/strong_escape_sites.csv, and printing the first few lines below:\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>condition</th>\n",
       "      <th>site</th>\n",
       "      <th>site_total_escape_frac_epistasis_model</th>\n",
       "      <th>max_mut</th>\n",
       "      <th>threshold</th>\n",
       "      <th>median</th>\n",
       "      <th>max</th>\n",
       "      <th>exceed_median</th>\n",
       "      <th>meet_exceed_median</th>\n",
       "      <th>frac_max</th>\n",
       "      <th>meet_frac_max</th>\n",
       "      <th>min_value</th>\n",
       "      <th>meet_min_value</th>\n",
       "      <th>meet_mut_min_value</th>\n",
       "      <th>strong_site</th>\n",
       "      <th>n_strong_sites</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>CAB-A17_139</td>\n",
       "      <td>456</td>\n",
       "      <td>0.5772</td>\n",
       "      <td>0.5368</td>\n",
       "      <td>default</td>\n",
       "      <td>0.01313</td>\n",
       "      <td>0.5772</td>\n",
       "      <td>0.1313</td>\n",
       "      <td>True</td>\n",
       "      <td>0.05772</td>\n",
       "      <td>True</td>\n",
       "      <td>0</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>CAB-A17_139</td>\n",
       "      <td>475</td>\n",
       "      <td>0.4891</td>\n",
       "      <td>0.4616</td>\n",
       "      <td>default</td>\n",
       "      <td>0.01313</td>\n",
       "      <td>0.5772</td>\n",
       "      <td>0.1313</td>\n",
       "      <td>True</td>\n",
       "      <td>0.05772</td>\n",
       "      <td>True</td>\n",
       "      <td>0</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>CAB-A49_222</td>\n",
       "      <td>417</td>\n",
       "      <td>2.3510</td>\n",
       "      <td>0.9329</td>\n",
       "      <td>default</td>\n",
       "      <td>0.01927</td>\n",
       "      <td>2.8220</td>\n",
       "      <td>0.1927</td>\n",
       "      <td>True</td>\n",
       "      <td>0.28220</td>\n",
       "      <td>True</td>\n",
       "      <td>0</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>CAB-A49_222</td>\n",
       "      <td>420</td>\n",
       "      <td>1.8470</td>\n",
       "      <td>0.9687</td>\n",
       "      <td>default</td>\n",
       "      <td>0.01927</td>\n",
       "      <td>2.8220</td>\n",
       "      <td>0.1927</td>\n",
       "      <td>True</td>\n",
       "      <td>0.28220</td>\n",
       "      <td>True</td>\n",
       "      <td>0</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>CAB-A49_222</td>\n",
       "      <td>456</td>\n",
       "      <td>1.3500</td>\n",
       "      <td>0.9806</td>\n",
       "      <td>default</td>\n",
       "      <td>0.01927</td>\n",
       "      <td>2.8220</td>\n",
       "      <td>0.1927</td>\n",
       "      <td>True</td>\n",
       "      <td>0.28220</td>\n",
       "      <td>True</td>\n",
       "      <td>0</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "strong_sites = site_threshold_df.query('strong_site').reset_index(drop=True)\n",
    "\n",
    "print('Number of sites of strong escape for each threshold:')\n",
    "display(HTML(\n",
    "    strong_sites\n",
    "    .groupby('threshold')\n",
    "    .aggregate(n_sites=pd.NamedAgg('site', 'nunique'))\n",
    "    .to_html()\n",
    "    ))\n",
    "\n",
    "print(f\"Writing to {config['strong_escape_sites']}, and printing the first few lines below:\")\n",
    "display(HTML(strong_sites.head().to_html()))\n",
    "\n",
    "strong_sites.to_csv(config['strong_escape_sites'], index=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Plot which sites are auto-identified for each condition (e.g., antibody or sera):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 90x180 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "p = (ggplot(strong_sites\n",
    "            .assign(site=lambda x: pd.Categorical(x['site'],  # categorical so only plot observed sites\n",
    "                                                  sorted(x['site'].unique()),\n",
    "                                                  ordered=True)\n",
    "                    )\n",
    "            ) +\n",
    "     aes('site', 'condition') +\n",
    "     geom_tile() +\n",
    "     theme(axis_text_x=element_text(angle=90),\n",
    "           figure_size=(0.18 * strong_sites['site'].nunique(),\n",
    "                        0.2 * strong_sites['condition'].nunique() * strong_sites['threshold'].nunique()),\n",
    "           ) +\n",
    "     facet_wrap('~ threshold', ncol=1)\n",
    "     )\n",
    "\n",
    "_ = p.draw()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "language_info": {
   "name": "python"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}