lineplots_by_group.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Summary plots showing antibody escape as line plots and mapped to RBD surface, averaged across groups.\n",
    "This Python Jupyter notebook plots escape map line plots averaged across groups.\n",
    "\n",
    "Also maps the group-average site-total escape to PDB 6M0J surface. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Imports and configuration\n",
    "Import Python modules:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import re\n",
    "\n",
    "import collections\n",
    "\n",
    "import Bio.PDB\n",
    "\n",
    "import dms_variants.pdb_utils\n",
    "\n",
    "from IPython.display import display, HTML\n",
    "\n",
    "import math\n",
    "import numpy as np\n",
    "\n",
    "import pandas as pd\n",
    "from plotnine import *\n",
    "\n",
    "import yaml"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Get configuration:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "with open('config.yaml') as f:\n",
    "    config = yaml.safe_load(f)\n",
    "    \n",
    "with open(config['lineplots_by_group_config']) as f:\n",
    "    line_plot_config = yaml.safe_load(f)\n",
    "    \n",
    "results_dir = config['lineplots_by_group_dir']\n",
    "os.makedirs(results_dir, exist_ok=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Read escape fractions\n",
    "Read the escape fractions for both the B.1.351 plasmas and the early 2020 plasmas, only keeping average of libraries:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "First few lines of escape-fraction data frame:\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>protein_chain</th>\n",
       "      <th>site_escape</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>12C_d152_80</td>\n",
       "      <td>331</td>\n",
       "      <td>E</td>\n",
       "      <td>0.04926</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>12C_d152_80</td>\n",
       "      <td>331</td>\n",
       "      <td>E</td>\n",
       "      <td>0.04926</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>12C_d152_80</td>\n",
       "      <td>331</td>\n",
       "      <td>E</td>\n",
       "      <td>0.04926</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>12C_d152_80</td>\n",
       "      <td>331</td>\n",
       "      <td>E</td>\n",
       "      <td>0.04926</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>12C_d152_80</td>\n",
       "      <td>331</td>\n",
       "      <td>E</td>\n",
       "      <td>0.04926</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "escape_frac_files = [config['early2020_escape_fracs'], config['escape_fracs']]\n",
    "\n",
    "site_metrics = {config['early2020_escape_fracs']: config['early2020_site_metric'], \n",
    "                config['escape_fracs']: config['site_metric'],\n",
    "               }\n",
    "\n",
    "escape_fracs_dfs = []\n",
    "\n",
    "for f in escape_frac_files:\n",
    "    \n",
    "    site_metric = site_metrics[f]\n",
    "\n",
    "    df = (pd.read_csv(f)\n",
    "          .query('library == \"average\"')\n",
    "          .drop(columns=['site', 'selection', 'library'])\n",
    "          .rename(columns={'label_site': 'site',\n",
    "                           site_metric: 'site_escape'\n",
    "                          }\n",
    "                 )\n",
    "          [['condition', 'site', 'protein_chain', 'site_escape']]\n",
    "         )\n",
    "    escape_fracs_dfs.append(df)\n",
    "\n",
    "escape_fracs=pd.concat(escape_fracs_dfs)\n",
    "\n",
    "# make \"padding\" data frame covering all conditions, sites, and mutations\n",
    "first_site = escape_fracs['site'].min()\n",
    "last_site = escape_fracs['site'].max()\n",
    "\n",
    "pad_df = pd.concat([pd.DataFrame({'condition': condition,\n",
    "                                  'site': list(range(first_site, last_site + 1)),\n",
    "                                 })\n",
    "                    for condition in escape_fracs['condition'].unique()\n",
    "                   ])\n",
    "\n",
    "# pad escape fracs data frame\n",
    "escape_fracs = (\n",
    "    escape_fracs\n",
    "    .merge(pad_df, how='outer')\n",
    "    .fillna(0)\n",
    "    .assign(protein_chain='E')\n",
    "    )\n",
    "\n",
    "escape_fracs.query('site_escape==361').head()\n",
    "\n",
    "print('First few lines of escape-fraction data frame:')\n",
    "display(HTML(escape_fracs.head().to_html(index=False)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Make group-averaged line plots"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Saving to results/lineplots_by_group/b1351_early2020_day30_grouped_lineplots.pdf\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/fh/fast/bloom_j/computational_notebooks/agreaney/2021/SARS-CoV-2-RBD_B.1.351/env/lib/python3.8/site-packages/plotnine/scales/scale_alpha.py:68: PlotnineWarning: Using alpha for a discrete variable is not advised.\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 500x400 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "for supergroup, subgroup in line_plot_config.items():\n",
    "\n",
    "    lineplot_df = pd.DataFrame()\n",
    "    annotation_df = pd.DataFrame()\n",
    "\n",
    "    for name, conditions in subgroup.items():\n",
    "        \n",
    "        assert set(conditions).issubset(set(escape_fracs['condition']))\n",
    "\n",
    "        df = (escape_fracs\n",
    "              .query('condition in @conditions')\n",
    "              .assign(group_name=name,))\n",
    "        \n",
    "        lineplot_df = (pd.concat([lineplot_df,df], axis=0, ignore_index=True))\n",
    "\n",
    "        label_df = (pd.DataFrame({'xmin': [382,482,454,442,416],\n",
    "                                 'xmax': [387,487,457,451,418],\n",
    "                                 'antibody_class': ['class 4', 'class 2', 'class 1', 'class 3', 'class 1'],\n",
    "                                 'label': ['383-386','483-486','455-456','443-450','417'],\n",
    "                                }\n",
    "                               )\n",
    "                    .assign(group_name=name,)\n",
    "                   )\n",
    "        annotation_df = pd.concat([annotation_df, label_df], axis=0, ignore_index=True)\n",
    "\n",
    "    lineplot_df = (lineplot_df\n",
    "        [['condition', 'site', 'site_escape', 'group_name']]\n",
    "        .drop_duplicates()\n",
    "        )\n",
    "\n",
    "    # add mean in each group\n",
    "    df_with_mean = (\n",
    "        lineplot_df\n",
    "        .append(lineplot_df.groupby(['group_name', 'site'], as_index=False)\n",
    "                      .aggregate('mean')\n",
    "                      .assign(condition='mean')\n",
    "                )\n",
    "        .assign(is_mean=lambda x: x['condition'] == 'mean')\n",
    "        )\n",
    "\n",
    "    annotation_df = (annotation_df\n",
    "                     .assign(group_name=lambda x: pd.Categorical(x['group_name'], ordered=True,\n",
    "                                                                 categories=subgroup.keys()\n",
    "                                                                )\n",
    "                            ) \n",
    "                    )\n",
    "\n",
    "    # space breaks every ? sites\n",
    "    site_break_freq = 10\n",
    "    def get_site_breaks(tup):\n",
    "        start = int(math.ceil(tup[0] / site_break_freq)) * site_break_freq\n",
    "        end = int(math.floor(tup[1] / site_break_freq)) * site_break_freq\n",
    "        return list(range(start, end + 1, site_break_freq))\n",
    "\n",
    "    # make plot\n",
    "    p = (ggplot(df_with_mean\n",
    "                .assign(group_name=lambda x: pd.Categorical(x['group_name'], ordered=True,\n",
    "                                                            categories=subgroup.keys()\n",
    "                                                           )\n",
    "                       )\n",
    "               ) +\n",
    "         aes('site', 'site_escape',\n",
    "             group='condition',\n",
    "             alpha='is_mean', size='is_mean') +\n",
    "\n",
    "         geom_rect(data=annotation_df.assign(is_mean=False),\n",
    "                   mapping=aes(xmin='xmin',\n",
    "                               xmax='xmax',\n",
    "                               ymin=-np.inf,\n",
    "                               ymax=np.inf,\n",
    "                               fill='antibody_class',\n",
    "                              ),\n",
    "                   alpha=0.2,\n",
    "                   inherit_aes=False,\n",
    "                  ) +\n",
    "\n",
    "         geom_line() +\n",
    "         facet_wrap('~ group_name', ncol=1) +\n",
    "         theme_classic() +\n",
    "         theme(figure_size=(5, 2 * df_with_mean['group_name'].nunique()),\n",
    "               axis_text_x=element_text(rotation=90, hjust=0.5),\n",
    "               strip_background=element_blank(),\n",
    "               ) +\n",
    "         scale_x_continuous(expand=(0, 0), breaks=get_site_breaks) +\n",
    "         scale_y_continuous(limits=(None, (df_with_mean.query('is_mean')['site_escape'].max()*1.5))) +\n",
    "         scale_alpha_discrete(range=(0.25, 1)) +  # transparency of individual and mean lines\n",
    "         scale_size_manual(values=(0.25, 0.5)) +  # size of individual and mean lines\n",
    "         scale_color_manual(values=['#e52794', '#6a0dad', '#66ccee', '#E69F00']) +\n",
    "         scale_fill_manual(values=['#e52794', '#6a0dad', '#66ccee', '#E69F00']) +\n",
    "         guides(alpha=False, size=False, )+ #color=False, fill=False\n",
    "         labs(x=\"RBD site\",\n",
    "              y=\"site-total antibody escape\",\n",
    "              color='antibody epitope',\n",
    "              fill='antibody epitope',\n",
    "             )\n",
    "         ) \n",
    "        \n",
    "\n",
    "    plotfile = f'{results_dir}/{supergroup}_grouped_lineplots.pdf'\n",
    "    print(f\"Saving to {plotfile}\")\n",
    "    p.save(plotfile, verbose=False)\n",
    "\n",
    "    _ = p.draw()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Map the average site-total escape for a given antibody class or plasma/serum group to the RBD surface.\n",
    "\n",
    "Outputs modified 6M0J PDBs with the escape scores as B factors.\n",
    "\n",
    "Though we will want more elaborate series of commands to codify our visualization of these RBD structures colored by escape, the series of commands below, when executed in a PyMol session with one of these PDBs open, will color the RBD surface according to escape scores."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Making PDB mappings for the average of 11 conditions for early2020 to data/pdbs/6M0J.pdb\n",
      "Mapping to the following chain: E\n",
      "  Writing B-factor re-assigned PDBs for early2020 to:\n",
      "    results/lineplots_by_group/early2020_6m0j_mean_total_escape.pdb\n",
      "\n",
      "Making PDB mappings for the average of 9 conditions for B1351 to data/pdbs/6M0J.pdb\n",
      "Mapping to the following chain: E\n",
      "  Writing B-factor re-assigned PDBs for B1351 to:\n",
      "    results/lineplots_by_group/B1351_6m0j_mean_total_escape.pdb\n"
     ]
    }
   ],
   "source": [
    "pdbfile = 'data/pdbs/6M0J.pdb'\n",
    "assert os.path.isfile(pdbfile)\n",
    "rbd_chain = config['escape_frac_protein_chain']\n",
    "assert isinstance(rbd_chain, str)\n",
    "\n",
    "for supergroup, subgroup in line_plot_config.items():\n",
    "\n",
    "    for name, conditions in subgroup.items():\n",
    "        \n",
    "        name=name.replace('.', '')\n",
    "        name=name.replace(' ', '')\n",
    "        name=re.search(r\"^[^\\(]*\", name).group(0)\n",
    "        \n",
    "        print(f\"\\nMaking PDB mappings for the average of {len(conditions)} conditions for {name} to {pdbfile}\")\n",
    "\n",
    "        # get escape fracs just for conditions of interest\n",
    "        df = escape_fracs.query('condition in @conditions')\n",
    "\n",
    "        # assign average total_escape at each site across all the conditions in ab_class\n",
    "        df = (df\n",
    "              .groupby(['site'])\n",
    "              .aggregate(mean_total_escape=pd.NamedAgg('site_escape', 'mean'),\n",
    "                          )\n",
    "              .reset_index()\n",
    "              .drop_duplicates()\n",
    "             )\n",
    "\n",
    "        # get chains\n",
    "        print(f'Mapping to the following chain: {rbd_chain}')\n",
    "        df = df.assign(chain=rbd_chain)\n",
    "\n",
    "\n",
    "        # make mappings for each condition and metric\n",
    "        print(f\"  Writing B-factor re-assigned PDBs for {name} to:\")\n",
    "\n",
    "        for metric in ['mean_total_escape']: # keeping this as list because we might need to normalize\n",
    "\n",
    "            # what do we assign to missing sites?\n",
    "            missing_metric = collections.defaultdict(lambda: 0)  # non-RBD chains always fill to zero\n",
    "            \n",
    "            # note that the next line DOESN'T ACTUALLY WORK AS INTENDED because I'm using padded escape fracs\n",
    "            missing_metric[rbd_chain] = -1  # missing sites in RBD are -1 for non-normalized metric PDBs\n",
    "\n",
    "            fname = os.path.join(results_dir, f\"{name}_6m0j_{metric}.pdb\")\n",
    "            print(f\"    {fname}\")\n",
    "\n",
    "            dms_variants.pdb_utils.reassign_b_factor(input_pdbfile=pdbfile,\n",
    "                                                     output_pdbfile=fname,\n",
    "                                                     df=df,\n",
    "                                                     metric_col=metric,\n",
    "                                                     missing_metric=missing_metric)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "language_info": {
   "name": "python"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}