Fig2_3_4.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 61,
   "id": "25c80549-c2fc-4a39-8344-bbe7f5b9622e",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import os\n",
    "import seaborn as sns\n",
    "from matplotlib import pyplot as plt\n",
    "\n",
    "\n",
    "color_scheme = {'Mouse':'#FF9E01',\n",
    "               'Kymouse':'#000000',\n",
    "                'Human':'#5695C1'}\n",
    "row_colors = {'Human':'#1D00FF','Kymouse':'#000000'}\n",
    "data_dir = 'tables'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "id": "7af1bac5-39cd-48df-b1b4-a36cc61c1d0a",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "def figure2(matrix, transpose = False, row_colors = row_colors, figsize = (20,10), order = True, col_cluster = True, row_cluster = False):\n",
    "    subject_ids = matrix.index\n",
    "    genes = matrix.columns\n",
    "    row_cols = pd.Series(subject_ids.map(lambda x:row_colors[x.split()[0]]))\n",
    "    row_cols.index = subject_ids\n",
    "    row_cols.name = 'Species'\n",
    "    if order:\n",
    "        gene_order = sorted(genes, key = lambda x:int(x[4]))\n",
    "    else:\n",
    "        gene_order = genes\n",
    "    matrix = matrix.T if transpose else matrix\n",
    "    if transpose:\n",
    "        print(matrix)\n",
    "        ax = sns.clustermap(matrix.loc[gene_order],figsize=figsize,cmap='RdBu_r',z_score=0,dendrogram_ratio=(.1,0.1),col_colors=row_cols,\n",
    "                       edgecolor='white',lw=.8,vmin=-4,vmax=5,cbar_kws={'label':'Z-normalized frequency'},row_cluster=False)\n",
    "        \n",
    "        ax.ax_row_dendrogram.set_visible(False)\n",
    "    else:\n",
    "        ax = sns.clustermap(matrix[gene_order],figsize=figsize,cmap='RdBu_r',z_score=1,dendrogram_ratio=(.1,0.1), row_colors=row_cols,\n",
    "                       edgecolor='white',lw=.8,vmin=-4,vmax=5,cbar_kws={'label':'Z-normalized frequency'},col_cluster=False)\n",
    "\n",
    "        ax.ax_col_dendrogram.set_visible(False)        \n",
    "        ax.cax.yaxis.set_label_position('left')\n",
    "        ax.cax.yaxis.set_ticks_position('left')\n",
    "    for tick in ax.ax_heatmap.get_xticklabels():\n",
    "        tick.set_fontsize(18)\n",
    "    for tick in ax.ax_heatmap.get_yticklabels():\n",
    "        tick.set_fontsize(18)\n",
    "    ax.cax.set_visible(False)\n",
    "    ax.ax_heatmap.figure.axes[-3].tick_params(labelsize=18)  \n",
    "    \n",
    "def plot_confidence_interval(mean, ci_lower, ci_upper, x, ax, color='black', horizontal_line_width=0.25):\n",
    "    left = x - horizontal_line_width / 2\n",
    "    top = max([ci_lower,ci_upper])\n",
    "    right = x + horizontal_line_width / 2\n",
    "    bottom = min([ci_lower,ci_upper])\n",
    "    ax.plot([x, x], [top, bottom], color=color)\n",
    "    ax.plot([left, right], [top, top], color=color)\n",
    "    ax.plot([left, right], [bottom, bottom], color=color)\n",
    "    ax.plot(x, mean, 'o',markersize=4,markeredgecolor='black',markerfacecolor=\"white\")\n",
    "\n",
    "            \n",
    "def figure3_estimationplot(factors):\n",
    "    fig, ax = plt.subplots(1,1)\n",
    "    xticks = ['n']\n",
    "    mapping = {'NP1':'VD insertion','NP2':'DJ insertion','IGHD':'IGHD gene','IGHJ':'IGHJ gene','IGHV':'IGHV gene'}\n",
    "    for x,(k,p) in enumerate(factors.sort_values('mean').iterrows()):\n",
    "        plot_confidence_interval(p['mean'],p['lower'],p['upper'],x=x, ax=ax)\n",
    "        xticks.append(k)\n",
    "    ax.set_xticklabels(xticks)\n",
    "    ax.set_ylabel('Difference in CDRH3 length\\nbetween Kymouse and humans\\n (nt)')\n",
    "    ax.set_xlabel('Factor')\n",
    "    ax.axhline(0,color='grey')\n",
    "    \n",
    "\n",
    "def figure4EF(diversities):\n",
    "    fig, axs = plt.subplots(2,1,figsize=(5,10))\n",
    "    sns.lineplot(data=diversities[diversities['key']=='clone_id'], ax=axs[1], x='subsample',y='shannon',hue='species',palette=[color_scheme['Human'],color_scheme['Kymouse'],color_scheme['Mouse']])\n",
    "    axs[1].set_ylabel('Shannon diversity (clonotypes)')\n",
    "    axs[1].set_xlabel('subsample proportion')\n",
    "    sns.despine()\n",
    "    sns.lineplot(data=diversities[diversities['key']=='cdr3_aa'], ax=axs[0], x='subsample',y='shannon',hue='species',palette=[color_scheme['Human'],color_scheme['Kymouse'],color_scheme['Mouse']])\n",
    "    axs[0].set_ylabel('Shannon diversity (CDRH3 A.A.)')\n",
    "    axs[0].set_xlabel('subsample proportion')\n",
    "    \n",
    "def figure4GH(sharing):\n",
    "    fig, axs = plt.subplots(2,1,figsize=(5,10))\n",
    "    sns.boxplot(data=sharing[sharing['key']=='cdr3_aa'].sort_values('order'), x='type', y='shared',palette=[color_scheme['Human'],'white',color_scheme['Kymouse'],'white','white',color_scheme['Mouse']],boxprops=dict(edgecolor='black',alpha=.4),ax=axs[0])\n",
    "    axs[0].set_xlabel('Comparison')\n",
    "    axs[0].set_ylabel('% CDRH3s shared')\n",
    "    for tick in axs[0].get_xticklabels():\n",
    "        tick.set_rotation(15)\n",
    "    sns.despine()\n",
    "\n",
    "    plt.setp(axs[0].artists, edgecolor = 'k')\n",
    "    sns.boxplot(data=sharing[sharing['key']=='clone_id'], x='type',y='shared',palette=[color_scheme['Human'],'white',color_scheme['Kymouse']],boxprops=dict(alpha=.3),ax=axs[1])\n",
    "    sns.swarmplot(data=sharing[sharing['key']=='clone_id'], x='type',y='shared',color='black',alpha=0.5,ax=axs[1])\n",
    "    axs[1].set_xlabel('Comparison')\n",
    "    axs[1].set_ylabel('% clonotypes shared')\n",
    "    for tick in axs[1].get_xticklabels():\n",
    "        tick.set_rotation(15)\n",
    "    sns.despine()\n",
    "    \n",
    "def figure4CD(diversities):\n",
    "    fig, axs = plt.subplots(2,1,figsize=(5,10))\n",
    "    ax = axs[1]\n",
    "    sns.lineplot(data = diversity[(diversity['cdr3_length']<=37)&(diversity['type']=='clonotype')], x= 'cdr3_length', y = 'diversity',hue='species',palette = [color_scheme['Human'],color_scheme['Kymouse'],color_scheme['Mouse']],ax=ax)\n",
    "    ax.set_ylabel('Shannon diversity (clonotypes)')\n",
    "    ax.set_xlabel('CDRH3 length (A.A.)')\n",
    "    sns.despine()\n",
    "    ax = axs[0]\n",
    "    ax = sns.lineplot(data = diversity[(diversity['cdr3_length']<=37)&(diversity['type']=='cdrh3')], x= 'cdr3_length', y = 'diversity',hue='species',palette = [color_scheme['Human'],color_scheme['Kymouse'],color_scheme['Mouse']],ax=ax)\n",
    "    ax.set_ylabel('Shannon diversity (CDRH3s)')\n",
    "    ax.set_xlabel('CDRH3 length (A.A.)')\n",
    "    sns.despine()\n",
    "    \n",
    "def figure4AB(ratios):\n",
    "    fig, axs = plt.subplots(2,1,figsize=(2,10))\n",
    "    sns.stripplot(data=table, ax=axs[1], x='species',y='ratio clonotypes:sequences',hue='species',palette=[color_scheme['Human'],color_scheme['Kymouse'],color_scheme['Mouse']])\n",
    "    axs[1].legend(bbox_to_anchor=(1.05, 1), loc=2, borderaxespad=0.)\n",
    "\n",
    "    axs[1].set_ylabel('Ratio unique clonotypes:sequences',size=14)\n",
    "\n",
    "    sns.despine()\n",
    "    axs[1].set_xticks([])\n",
    "\n",
    "    xlims = axs[1].get_xlim()\n",
    "\n",
    "    for k,p in enumerate(['Human','Kymouse','Mouse']):\n",
    "        y = table[table['species']==p]\n",
    "        mean = y['ratio clonotypes:sequences'].mean()\n",
    "        sem = y['ratio clonotypes:sequences'].sem()*2\n",
    "        axs[1].axhspan(mean+sem, mean-sem,xlims[0],xlims[1],color=color_scheme[p.replace('Humanised mouse','Kymouse')],alpha=0.2)\n",
    "        axs[1].axhline(y=mean,color=color_scheme[p.replace('Humanised mouse','Kymouse')],alpha=0.9)\n",
    "\n",
    "\n",
    "    sns.stripplot(data=table, ax=axs[0], x='species',y='ratio H3:sequences',hue='species',palette=[color_scheme['Human'],color_scheme['Kymouse'],color_scheme['Mouse']])\n",
    "    xlims = axs[0].get_xlim()\n",
    "    for k,p in enumerate(['Human','Kymouse','Mouse']):\n",
    "        y = table[table['species']==p]\n",
    "        mean = y['ratio H3:sequences'].mean()\n",
    "        sem = y['ratio H3:sequences'].sem()*2\n",
    "        axs[0].axhspan(mean+sem, mean-sem,xlims[0],xlims[1],color=color_scheme[p.replace('Humanised mouse','Kymouse')],alpha=0.2)\n",
    "        axs[0].axhline(y=mean,color=color_scheme[p.replace('Humanised mouse','Kymouse')],alpha=0.9)\n",
    "\n",
    "    axs[0].legend(bbox_to_anchor=(1.05, 1), loc=2, borderaxespad=0.)\n",
    "    axs[0].set_ylabel('Ratio unique CDRH3s:sequences',size=14)\n",
    "    axs[0].set_xticks([])\n",
    "    sns.despine()\n"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}