Metaomic_correlations_to_IBD_severity.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Meta -omic correlations to IBD severity"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "11/03/2020<br>\n",
    "Author: Robert Mills<br>\n",
    "Environment: Qiime2-2019.7<br>\n",
    "<br>\n",
    "<i>This notebook contains code for reproducing several figures related to meta -omic associations to IBD from the manuscript, <b> \"Meta–omics Reveals Microbiome Driven Proteolysis as a Contributing Factor to Severity of Ulcerative Colitis Disease Activity\" </b>by Mills et al. <br><br>\n",
    "Data sheets used here will be made available on Massive.ucsd.edu"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Import dependencies\n",
    "import math\n",
    "import re\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import scipy.stats as stats\n",
    "import seaborn as sns\n",
    "import sys\n",
    "import scipy\n",
    "from scipy.stats import linregress\n",
    "from matplotlib_venn import venn2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Makes figures with white background\n",
    "sns.set_style(\"whitegrid\")\n",
    "#Allows visualization within the ipython notebook format\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Optional for disabling warnings\n",
    "if not sys.warnoptions:\n",
    "    import warnings\n",
    "    warnings.simplefilter(\"ignore\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Option for re-enabling warnings\n",
    "if not sys.warnoptions:\n",
    "    import warnings\n",
    "    warnings.simplefilter(\"default\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### User defined functions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "def LinearRegressions(df,metadata,variable):\n",
    "    \"\"\"\n",
    "    This function will perform linear regressions on all features in a dataframe. Each row is a feature and each column is a sample. The function appends appends new columns for the stats associated with the linear regression.\n",
    "    A mask is in place to detect missing values, and remove them from the regression. The fraction of samples missing the feature is listed as an additional column.\n",
    "\n",
    "\n",
    "    \n",
    "    Parameters\n",
    "    ----------\n",
    "    df : pd.DataFrame\n",
    "    metadata : pd.DataFrame\n",
    "    variable: String\n",
    "\n",
    "    Returns\n",
    "    -------\n",
    "    Dataframe with stats on linear regressions\n",
    "    \"\"\"\n",
    "    \n",
    "    samples = list(df.columns)\n",
    "    meta2 = metadata[metadata.index.isin(samples)]\n",
    "\n",
    "    #Remove non-numeric samples\n",
    "    meta3 =\tmeta2[pd.to_numeric(meta2[variable], errors='coerce').notnull()]\n",
    "    samples2 = list(meta3.index)\n",
    "    df2 = df[samples2]\n",
    "\n",
    "    df2.loc['%s'%(variable)] = meta3[variable].astype('float')\n",
    "\n",
    "    #Order patients by the variable & add linregress stats\n",
    "    Order = list(meta3[variable].sort_values().index)\n",
    "    linregstats =['slope','intercept','rvalue','pvalue','stderr']\n",
    "    for i in linregstats:\n",
    "        Order.append(i)\n",
    "\n",
    "    #Reindex to sort the data columns\n",
    "    df2 = df2.reindex(columns = Order)\n",
    "\n",
    "    #Remove the features that have no values (or are a missing value)\n",
    "    df2 = df2.dropna(axis=0,how='all')\n",
    "    df2 = df2.loc[df2.index.dropna()]\n",
    "\n",
    "    #Apply linear regressions to the variable\n",
    "    slope = []\n",
    "    intercept = []\n",
    "    rvalue = []\n",
    "    pvalue = []\n",
    "    stderr = []\n",
    "    sparsity = []\n",
    "\n",
    "    for i in df2.index:\n",
    "        a = df2.loc[i]\n",
    "        b = df2.loc[variable]\n",
    "        mask = ~np.isnan(df2.loc[i]) & ~np.isnan(df2.loc[variable])\n",
    "        linres = linregress(a[mask],b[mask])\n",
    "        #linres = linregress(a[samples],b[samples])\n",
    "        slope.append(linres.slope)\n",
    "        intercept.append(linres.intercept)\n",
    "        rvalue.append(linres.rvalue)\n",
    "        pvalue.append(linres.pvalue)\n",
    "        stderr.append(linres.stderr)\n",
    "        sparsity.append((np.isnan(df2[samples2].loc[i])).astype(int).sum()/len(a[samples2]))\n",
    "    df2['slope'] = slope\n",
    "    df2['intercept'] = intercept\n",
    "    df2['rvalue'] = rvalue\n",
    "    df2['pvalue'] = pvalue\n",
    "    df2['stderr'] = stderr\n",
    "    df2['sparsity_prcnt'] = sparsity\n",
    "\n",
    "    global df_out\n",
    "    df_out = df2  \n",
    "    return 'Linear regressions performed on %s. New dataframe stored as \"df_out\"'%(variable)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 158,
   "metadata": {},
   "outputs": [],
   "source": [
    "def PlotByCategory_MultiSampleGroups2(df,lists,Taxa_col,category,metadata,display_number,save_descrip,human=\"True\",col_order=False):\n",
    "    \"\"\"\n",
    "    This function allows you to average the abundance of a category by the levels within that category\n",
    "    and plot stacked barplots of the relative abundance of the x most abundant groups within \n",
    "    each level. Can handle multiple groups of samples input into lists\n",
    "    \n",
    "    Parameters\n",
    "    ----------\n",
    "    df : pd.DataFrame\n",
    "    lists: list\n",
    "    Taxa_col : String\n",
    "    Sparcity: np.float\n",
    "    variable : Name of variable you're comparing ratios of\n",
    "    display_number: np.integer\n",
    "    human: String\n",
    "    save_descrip : String\n",
    "    col_order : list\n",
    "\n",
    "    Returns\n",
    "    -------\n",
    "    pdf visualization\n",
    "    global of Others and top_list for input into other functions\n",
    "    \"\"\"\n",
    "    global Others\n",
    "    global top_list\n",
    "    \n",
    "    num_lists = len(lists)\n",
    "    all_samples = []\n",
    "    for i in lists:\n",
    "        for j in i:\n",
    "            all_samples.append(j)\n",
    "\n",
    "    df_all = df[all_samples].groupby(df['%s' % Taxa_col]).agg(['sum'])\n",
    "    \n",
    "    #Find the sum per variable\n",
    "    df_all['sum'] = df_all.sum(axis=1)\n",
    "        \n",
    "    #Find the most abundantly entry within the variable, group others\n",
    "    top_list = list(df_all.sort_values('sum', ascending=False)[0:display_number].index)\n",
    "    Others = set(df_all.index) - set(top_list)\n",
    "    top_list.append('Others')\n",
    "\n",
    "    #A dictionary of dataframes will be needed to store all summed dataframes:\n",
    "    #df_dict = {}\n",
    "\n",
    "    #Initialize the figure\n",
    "    figlength = num_lists*1.75\n",
    "    fig, axes = plt.subplots(nrows=num_lists,ncols=1,figsize=(1.25,figlength))\n",
    "    \n",
    "    for num, i in enumerate(lists, start=1):       \n",
    "        #Sum your dataframe by a category of interest \n",
    "        dfcategories = df[i].groupby(df['%s' % Taxa_col]).agg(['sum'])\n",
    "        #Restructure the columns which become multilevel with the sum function                  \n",
    "        dfcategories.columns = dfcategories.columns.droplevel(1)\n",
    "\n",
    "        if human != \"True\":\n",
    "            dfcategories.drop('Homo sapiens (Human)', inplace=True)   \n",
    "        \n",
    "        #Save the postive and negative correlations per genera or eggNOG category\n",
    "        counts = dfcategories\n",
    "        \n",
    "        #Find averages according to the levels in a category of interest\n",
    "        \n",
    "        #subset metadata to only list with patients of interest\n",
    "        metadata.index = metadata.index.astype('str')\n",
    "        meta2 = metadata.loc[i]\n",
    "        #Identify the levels within the metadata category\n",
    "        levels_list = list(meta2[category].drop_duplicates().sort_values())\n",
    "        #Add new column for patients regarding the level of the category of interest.\n",
    "        counts.loc['%s' % category] = metadata['%s' % category]\n",
    "\n",
    "        #Create a new column averaging by the levels within the category\n",
    "        col_list = []\n",
    "        \n",
    "        ##NOTE - THIS STEP MAY BE REORDERING LIST SHOULD CHECK\n",
    "        for j in levels_list:\n",
    "            a = counts[i].T\n",
    "            patients_in = []\n",
    "            patients_in = list(a[a[category] == j].index.astype('str'))\n",
    "            b = counts[patients_in].drop(category)\n",
    "            b = b.astype('float')\n",
    "            counts['%s' % j] = b[patients_in].mean(axis=1)\n",
    "            col_list.append('%s' % (j))\n",
    "        \n",
    "\n",
    "        counts.loc['Others'] = counts.loc[Others].sum()\n",
    "        plot_df = counts[col_list]\n",
    "        plot_df = plot_df.loc[top_list]\n",
    "\n",
    "        #Convert to ratio for display\n",
    "        for i in col_list:\n",
    "            plot_df[i] = plot_df[i]/plot_df[i].sum()\n",
    "        \n",
    "        plot_df = plot_df[col_list]\n",
    "        \n",
    "        #Remove index name\n",
    "        del plot_df.index.name\n",
    "        \n",
    "        #Transpose\n",
    "        plot_df = plot_df.transpose()\n",
    "        #Get the dates in the correct order\n",
    "        if col_order != False:\n",
    "            plot_df = plot_df.reindex(col_order)\n",
    "        #Get variable into an order which depends on the abundance of each variable for easier interpretation\n",
    "        #plot_df = plot_df[plot_df.sum().sort_values(ascending = False).index]\n",
    "        #df_dict['Group_list%s' % (num)] = plot_df\n",
    "        \n",
    "        axenum = num-1\n",
    "        plot1 = plot_df.plot(kind='bar', ax=axes[axenum], width = 0.8, stacked=True, colors=['black','red','white'], title='')\n",
    "        #plot1.get_xaxis().set_ticklabels([])\n",
    "        plot1.get_legend().remove()\n",
    "    fig.autofmt_xdate(rotation = 90, ha = 'center')\n",
    "    plot1.legend(loc='center left', bbox_to_anchor=(1, 0.5))\n",
    "    fig.tight_layout()\n",
    "    fig.savefig('Barplot_of_%s_by_%s_%s.pdf' %(Taxa_col,category,save_descrip),bbox_inches=\"tight\")\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "def TaxaRatioCompareSigLevelGroups(df_dict,sig_level,Taxa_col,variable,Sparsity,display_number,save_descrip,human=\"True\",relative=\"True\"):\n",
    "    \"\"\"\n",
    "    This function combines several lines of code to speed up comparative analyses between two data sets.\n",
    "    Input the data frames you want to compare and the name of the taxonomic level under Taxa.\n",
    "    \n",
    "    Parameters\n",
    "    ----------\n",
    "    df_dict : pd.dict\n",
    "    Taxa_col : String\n",
    "    sig_level : np.float\n",
    "    Sparcity: np.float\n",
    "    variable : Name of variable you're comparing ratios of\n",
    "    display_number: np.integer\n",
    "    human: String\n",
    "    relative: String\n",
    "    save_descrip : String\n",
    "\n",
    "    Returns\n",
    "    -------\n",
    "    Plot of the composition of significant proteins.\n",
    "    \"\"\"\n",
    "    num_df = len(df_dict)\n",
    "    \n",
    "    #Combine all of the proteins in the dictionary and find most common categories\n",
    "    df_concat = pd.concat(df_dict)\n",
    "\n",
    "    #Remove spurious correlations with high sparsity\n",
    "    df_sig = df_concat[df_concat['sparsity_prcnt'] < Sparsity]\n",
    "\n",
    "    #Subset out the significant positive and negative correlations\n",
    "    Cal_neg = df_sig[df_sig['rvalue'] < -sig_level]\n",
    "    Cal_pos = df_sig[df_sig['rvalue'] > sig_level]    \n",
    "\n",
    "    Cal_neg['%s_Pos' % variable] = 0\n",
    "    Cal_pos['%s_Pos' % variable] = 1\n",
    "    Cal_neg['%s_Neg' % variable] = 1\n",
    "    Cal_pos['%s_Neg' % variable] = 0\n",
    "    \n",
    "    df_sig = pd.concat([Cal_neg, Cal_pos])\n",
    "    \n",
    "    #Sum by category\n",
    "\n",
    "    #Use this for summing by taxonomy\n",
    "    dfcategories = df_sig[['%s_Neg' % variable,'%s_Pos' % variable]].groupby(df_sig['%s' % Taxa_col]).agg(['sum'])\n",
    "    dfcategories.columns = dfcategories.columns.droplevel(1)\n",
    "\n",
    "    if human != \"True\":\n",
    "        dfcategories.drop('Homo', inplace=True)\n",
    "\n",
    "    #Save the postive and negative correlations per genera or eggNOG category\n",
    "    counts = dfcategories\n",
    "\n",
    "    #Find the sum per variable\n",
    "    counts['sum'] = counts.sum(axis=1)\n",
    "\n",
    "    #Find the most abundantly correlated entry within the variable, group others\n",
    "    top_list = list(counts.sort_values('sum', ascending=False)[0:display_number].index)\n",
    "    Others = set(counts.index) - set(top_list)\n",
    "    counts.loc['Others'] = counts.loc[Others].sum()\n",
    "    top_list.append('Others')\n",
    "\n",
    "    figwidth = num_df*2\n",
    "    fig, axes = plt.subplots(nrows=1,ncols=num_df,figsize=(figwidth,3))\n",
    "    \n",
    "    for num, i in enumerate(df_dict, start=1):      \n",
    "        #Remove spurious correlations with high sparsity\n",
    "        df_sig = df_dict[i][df_dict[i]['sparsity_prcnt'] < Sparsity]\n",
    "\n",
    "        #Subset out the significant positive and negative correlations\n",
    "        Cal_neg = df_sig[df_sig['rvalue'] < -sig_level]\n",
    "        Cal_pos = df_sig[df_sig['rvalue'] > sig_level]\n",
    "\n",
    "        #Subset out the significant positive and negative correlations MG\n",
    "        #Cal_neg2 = df2_sig[df2_sig['rvalue'] < 0]\n",
    "        #Cal_pos2 = df2_sig[df2_sig['rvalue'] > 0]\n",
    "\n",
    "\n",
    "        #pDB Use this code if you want to simply look at quantity of significant proteins\n",
    "        Cal_neg['%s_Pos' % variable] = 0\n",
    "        Cal_pos['%s_Pos' % variable] = 1\n",
    "        Cal_neg['%s_Neg' % variable] = 1\n",
    "        Cal_pos['%s_Neg' % variable] = 0\n",
    "\n",
    "        #MG Use this code if you want to simply look at quantity of significant proteins\n",
    "        #Cal_neg2['%s_Pos2' % variable] = 0\n",
    "        #Cal_pos2['%s_Pos2' % variable] = 1\n",
    "        #Cal_neg2['%s_Neg2' % variable] = 1\n",
    "        #Cal_pos2['%s_Neg2' % variable] = 0\n",
    "\n",
    "        df_sig = pd.concat([Cal_neg, Cal_pos])\n",
    "\n",
    "        #df_sig2 = pd.concat([Cal_neg2, Cal_pos2])\n",
    "\n",
    "        #Sum by category\n",
    "\n",
    "        #Use this for summing by taxonomy\n",
    "        dfcategories = df_sig[['%s_Neg' % variable,'%s_Pos' % variable]].groupby(df_sig['%s' % Taxa_col]).agg(['sum'])\n",
    "        dfcategories.columns = dfcategories.columns.droplevel(1)\n",
    "\n",
    "        #Use this for summing by taxonomy\n",
    "        #dfcategories2 = df_sig2[['%s_Neg2' % variable,'%s_Pos2' % variable]].groupby(df_sig2['%s' % Taxa_col]).agg(['sum'])\n",
    "        #dfcategories2.columns = dfcategories2.columns.droplevel(1)\n",
    "\n",
    "        if human != \"True\":\n",
    "            dfcategories.drop('Homo', inplace=True)\n",
    "\n",
    "\n",
    "        #Save the postive and negative correlations per genera or eggNOG category\n",
    "        counts = dfcategories\n",
    "\n",
    "        plot_df = counts.loc[top_list]\n",
    "\n",
    "        #Convert to ratio for display - switch comment activity for relative versus total correlations\n",
    "        plot_vars = ['%s_Neg' % variable,'%s_Pos' % variable]\n",
    "\n",
    "        if relative != \"True\":\n",
    "            pass\n",
    "        else:\n",
    "            for j in plot_vars:\n",
    "                plot_df[j] = plot_df[j]/plot_df[j].sum()\n",
    "\n",
    "        plot_df = plot_df[plot_vars]\n",
    "\n",
    "        #Transpose the plot to make it fit the code structure I'm following for the stacked bar chart\n",
    "\n",
    "        #Remove index name\n",
    "        del plot_df.index.name\n",
    "        #Transpose\n",
    "        plot_df = plot_df.transpose()\n",
    "        #Get the dates in the correct order\n",
    "        plot_df = plot_df.reindex(plot_vars)\n",
    "        #Get variable into an order which depends on the abundance of each variable for easier interpretation\n",
    "        plot_df = plot_df[top_list]\n",
    "        axenum = num-1   \n",
    "        plot1 = plot_df.plot(kind='bar', ax=axes[axenum], width = 0.8, stacked=True, colormap='tab20c', title='%s' %i)\n",
    "        #plot1.get_xaxis().set_ticklabels([])\n",
    "        plot1.get_legend().remove()\n",
    "    plot1.legend(loc='center left', bbox_to_anchor=(1, 0.5))\n",
    "    fig.tight_layout()\n",
    "    fig.savefig('Barplot_of_%s_correlations_at_%s_sparsity_%s.pdf' %(variable,Sparsity,save_descrip),bbox_inches=\"tight\")\n",
    "\n",
    "\n",
    "    \n",
    "    return 'New plot saved.'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def TaxaLogRatioCompareSigLevelGroupsCombine(df_dict,sig_level,Taxa_col,variable,Sparsity,\n",
    "                                             MinCount,human=\"True\",compareall=\"False\"):\n",
    "    \"\"\"\n",
    "    This function provides a dictionary of comparative analyses between \n",
    "    the groupings of significant features. Input a dictionary of different datasets to compare\n",
    "    under the same settings for significance level, sparsity and filtering human features.\n",
    "    Compare all switches between comparing only features above or below the absolute value\n",
    "    of the significance threshold or comparing all within a significance threshold to\n",
    "    the values below that threshold. Input the column of the category you want to compare in\n",
    "    \"Taxa_col\"\n",
    "    \n",
    "    Parameters\n",
    "    ----------\n",
    "    df_dict : pd.dict\n",
    "    Taxa_col : String\n",
    "    sig_level : np.float\n",
    "    Sparcity: np.float\n",
    "    variable : Name of variable you're comparing ratios of\n",
    "    display_number: np.integer\n",
    "    human: String\n",
    "    relative: String\n",
    "    save_descrip : String\n",
    "    MinCount : np.integer\n",
    "    compareall: String\n",
    "\n",
    "    Returns\n",
    "    -------\n",
    "    Plot of the composition of significant proteins.\n",
    "    \"\"\"\n",
    "    global ratio_dict\n",
    "    num_df = len(df_dict)\n",
    "\n",
    "    #figwidth = num_df*5\n",
    "    #fig, axes = plt.subplots(nrows=1,ncols=num_df,figsize=(figwidth,6))\n",
    "    \n",
    "    ratio_dict = {}\n",
    "    for num, i in enumerate(df_dict, start=1):      \n",
    "        #Remove spurious correlations with high sparsity\n",
    "        df_sig = df_dict[i][df_dict[i]['sparsity_prcnt'] < Sparsity]\n",
    "        \n",
    "        if compareall==\"False\":\n",
    "            #Subset out the significant positive and negative correlations\n",
    "            Cal_neg = df_sig[df_sig['rvalue'] < -abs(sig_level)]\n",
    "            Cal_pos = df_sig[df_sig['rvalue'] > abs(sig_level)]\n",
    "        else:\n",
    "            if sig_level > 0:\n",
    "                Cal_neg = df_sig[df_sig['rvalue'] < sig_level]\n",
    "                Cal_pos = df_sig[df_sig['rvalue'] >= sig_level]\n",
    "            else:\n",
    "                Cal_neg = df_sig[df_sig['rvalue'] <= sig_level]\n",
    "                Cal_pos = df_sig[df_sig['rvalue'] > sig_level]                  \n",
    "\n",
    "\n",
    "        #Subset out the significant positive and negative correlations MG\n",
    "        #Cal_neg2 = df2_sig[df2_sig['rvalue'] < 0]\n",
    "        #Cal_pos2 = df2_sig[df2_sig['rvalue'] > 0]\n",
    "\n",
    "\n",
    "        #pDB Use this code if you want to simply look at quantity of significant proteins\n",
    "        Cal_neg['%s_Pos' % variable] = 0\n",
    "        Cal_pos['%s_Pos' % variable] = 1\n",
    "        Cal_neg['%s_Neg' % variable] = 1\n",
    "        Cal_pos['%s_Neg' % variable] = 0\n",
    "\n",
    "        df_sig = pd.concat([Cal_neg, Cal_pos])\n",
    "\n",
    "        #Sum by category\n",
    "\n",
    "        #Use this for summing by taxonomy\n",
    "        dfcategories = df_sig[['%s_Neg' % variable,'%s_Pos' % variable]].groupby(df_sig['%s' % Taxa_col]).agg(['sum'])\n",
    "        dfcategories.columns = dfcategories.columns.droplevel(1)\n",
    "\n",
    "        dfcategories['sum'] = dfcategories['%s_Neg' % variable] + dfcategories['%s_Pos' % variable]\n",
    "\n",
    "\n",
    "        if human != \"True\":\n",
    "            dfcategories.drop('Homo sapiens (Human)', inplace=True)\n",
    "\n",
    "\n",
    "        #Save the postive and negative correlations per genera or eggNOG category\n",
    "        counts = dfcategories\n",
    "\n",
    "        #Add a Pseudocount of 1 so that the categories without anything are still shown\n",
    "        dfcategories = counts.replace(0,1)\n",
    "        \n",
    "        #Find the ratios necessary for calculating the ratio of ratios\n",
    "        dfcategories['%s_Percent_Neg' % variable] = \\\n",
    "        dfcategories['%s_Neg' % variable]/dfcategories['%s_Neg' % variable].sum()\n",
    "        dfcategories['%s_Percent_Pos' % variable] = \\\n",
    "        dfcategories['%s_Pos' % variable]/dfcategories['%s_Pos' % variable].sum()\n",
    "\n",
    "        dfcategories['%s_Ratio of Percentages (Pos/Neg)' % variable] = dfcategories['%s_Percent_Pos' % variable]/dfcategories['%s_Percent_Neg' % variable]\n",
    "        \n",
    "        dfcategories['Log(Pos/Neg)'] = np.log10(dfcategories['%s_Ratio of Percentages (Pos/Neg)' % variable])\n",
    "\n",
    "        dfcategories['Category'] = (\"%s\" % i)\n",
    "        dfcategories['#Proteins'] = dfcategories['sum'].astype('str')\n",
    "        #dfcategories = dfcategories.loc[top_list]\n",
    "        #dfcategories.index = dfcategories['Category and #Proteins']\n",
    "        df = dfcategories[dfcategories['sum'] > MinCount]\n",
    "        df = df.sort_values('Log(Pos/Neg)')\n",
    "        df.reset_index(inplace=True)\n",
    "        \n",
    "        ratio_dict[\"%s\"%i] = df\n",
    "\n",
    "\n",
    "    \n",
    "    return 'All ratio dataframes combined into \"ratio_dict\".'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [],
   "source": [
    "def PlotByCategory_MultiSampleGroups(df,lists,Taxa_col,category,metadata,display_number,save_descrip,human=\"True\",size=9):\n",
    "    \"\"\"\n",
    "    This function allows you to average the abundance of a category by the levels within that category\n",
    "    and plot stacked barplots of the relative abundance of the x most abundant groups within \n",
    "    each level. Can handle multiple groups of samples input into lists\n",
    "    \n",
    "    Parameters\n",
    "    ----------\n",
    "    df : pd.DataFrame\n",
    "    lists: list\n",
    "    Taxa_col : String\n",
    "    variable : Name of variable you're comparing ratios of\n",
    "    display_number: np.integer\n",
    "    human: String\n",
    "    save_descrip : String\n",
    "    size : np.integer\n",
    "\n",
    "    Returns\n",
    "    -------\n",
    "    pdf visualization\n",
    "    global of Others and top_list for input into other functions\n",
    "    \"\"\"\n",
    "    global Others\n",
    "    global top_list\n",
    "    \n",
    "    num_lists = len(lists)\n",
    "    all_samples = []\n",
    "    for i in lists:\n",
    "        for j in i:\n",
    "            all_samples.append(j)\n",
    "\n",
    "    df_all = df[all_samples].groupby(df['%s' % Taxa_col]).agg(['sum'])\n",
    "    \n",
    "    #Find the sum per variable\n",
    "    df_all['sum'] = df_all.sum(axis=1)\n",
    "        \n",
    "    #Find the most abundantly entry within the variable, group others\n",
    "    top_list = list(df_all.sort_values('sum', ascending=False)[0:display_number].index)\n",
    "    Others = set(df_all.index) - set(top_list)\n",
    "    top_list.append('Others')\n",
    "\n",
    "    #A dictionary of dataframes will be needed to store all summed dataframes:\n",
    "    #df_dict = {}\n",
    "\n",
    "    #Initialize the figure\n",
    "    figlength = num_lists*2\n",
    "    figwidth = 2*(size/9)\n",
    "    fig, axes = plt.subplots(nrows=num_lists,ncols=1,figsize=(figwidth,figlength))\n",
    "    \n",
    "    for num, i in enumerate(lists, start=1):       \n",
    "        #Sum your dataframe by a category of interest \n",
    "        dfcategories = df[i].groupby(df['%s' % Taxa_col]).agg(['sum'])\n",
    "        #Restructure the columns which become multilevel with the sum function                  \n",
    "        dfcategories.columns = dfcategories.columns.droplevel(1)\n",
    "\n",
    "        if human != \"True\":\n",
    "            dfcategories.drop('Homo', inplace=True)   \n",
    "        \n",
    "        #Save the postive and negative correlations per genera or eggNOG category\n",
    "        counts = dfcategories\n",
    "        \n",
    "        #Find averages according to the levels in a category of interest\n",
    "        \n",
    "        #subset metadata to only list with patients of interest\n",
    "        metadata.index = metadata.index.astype('str')\n",
    "        meta2 = metadata.loc[i]\n",
    "        #Identify the levels within the metadata category\n",
    "        levels_list = list(meta2[category].drop_duplicates().sort_values())\n",
    "        #Add new column for patients regarding the level of the category of interest.\n",
    "        counts.loc['%s' % category] = metadata['%s' % category]\n",
    "\n",
    "        #Create a new column averaging by the levels within the category\n",
    "        col_list = []\n",
    "        \n",
    "        ##NOTE - THIS STEP MAY BE REORDERING LIST SHOULD CHECK\n",
    "        for j in levels_list:\n",
    "            a = counts[i].T\n",
    "            patients_in = []\n",
    "            patients_in = list(a[a[category] == j].index.astype('str'))\n",
    "            b = counts[patients_in].drop(category)\n",
    "            b = b.astype('float')\n",
    "            counts['Patients_in_%s_%s_Average' % (category, j)] = b[patients_in].mean(axis=1)\n",
    "            col_list.append('Patients_in_%s_%s_Average' % (category, j))\n",
    "        \n",
    "\n",
    "        counts.loc['Others'] = counts.loc[Others].sum()\n",
    "        plot_df = counts[col_list]\n",
    "        plot_df = plot_df.loc[top_list]\n",
    "\n",
    "        #Convert to ratio for display\n",
    "        for i in col_list:\n",
    "            plot_df[i] = plot_df[i]/plot_df[i].sum()\n",
    "        \n",
    "        plot_df = plot_df[col_list]\n",
    "        \n",
    "        #Remove index name\n",
    "        del plot_df.index.name\n",
    "        \n",
    "        #Transpose\n",
    "        plot_df = plot_df.transpose()\n",
    "        #Get the dates in the correct order\n",
    "        #plot_df = plot_df.reindex(col_list)\n",
    "        #Get variable into an order which depends on the abundance of each variable for easier interpretation\n",
    "        #plot_df = plot_df[plot_df.sum().sort_values(ascending = False).index]\n",
    "        #df_dict['Group_list%s' % (num)] = plot_df\n",
    "        \n",
    "        axenum = num-1\n",
    "        plot1 = plot_df.plot(kind='bar', ax=axes[axenum], width = 0.8, stacked=True, colormap='tab20', title='')\n",
    "        plot1.get_xaxis().set_ticklabels([])\n",
    "        plot1.get_legend().remove()\n",
    "    plot1.legend(loc='center left', bbox_to_anchor=(1, 0.5))\n",
    "\n",
    "    fig.savefig('Barplot_of_%s_by_%s_%s.pdf' %(Taxa_col,category,save_descrip),bbox_inches=\"tight\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "metadata": {},
   "outputs": [],
   "source": [
    "def DotPlotCategory(df,samples,Taxa_col,category,metadata,save_descrip,width=2):\n",
    "    \"\"\"\n",
    "    This function creates a dotplot based on the averages of a given category in a metadata file.\n",
    "\n",
    "    Parameters\n",
    "    ----------\n",
    "    df : pd.DataFrame\n",
    "    samples: list\n",
    "    Taxa_col : String\n",
    "    variable : String\n",
    "    display_number: np.integer\n",
    "    save_descrip : String\n",
    "    width : np.integer\n",
    "\n",
    "    Returns\n",
    "    -------\n",
    "    pdf visualization\n",
    "    \"\"\"\n",
    "    #A dictionary of dataframes will be needed to store all summed dataframes:\n",
    "    #df_dict = {}\n",
    "\n",
    "    #Initialize the figure\n",
    "    df2 = df[samples]\n",
    "    #convert to averages per category\n",
    "    #subset metadata to only list with patients of interest\n",
    "    metadata.index = metadata.index.astype('str')\n",
    "    meta2 = metadata.loc[samples]\n",
    "    #Identify the levels within the metadata category\n",
    "    \n",
    "    levels_list = list(meta2[category].drop_duplicates())\n",
    "    #Add new column for patients regarding the level of the category of interest.\n",
    "    df2.loc['%s' % category] = metadata['%s' % category]\n",
    "\n",
    "    #Create a new column averaging by the levels within the category\n",
    "    col_list = []\n",
    "    \n",
    "    ##NOTE - THIS STEP MAY BE REORDERING LIST SHOULD CHECK\n",
    "    for j in levels_list:\n",
    "        a = df2[samples].T\n",
    "        patients_in = []\n",
    "        patients_in = list(a[a[category] == j].index.astype('str'))\n",
    "        b = df2[patients_in].drop(category)\n",
    "        b = b.astype('float')\n",
    "        df2['Patients_in_%s_%s_Average' % (category, j)] = b[patients_in].mean(axis=1)\n",
    "        col_list.append('Patients_in_%s_%s_Average' % (category, j))\n",
    "\n",
    "    df2 = df2[col_list]\n",
    "\n",
    "    #Convert to ratio for display\n",
    "    for j in col_list:\n",
    "        df2[j] = df2[j]/df2[j].sum()\n",
    "    \n",
    "    df2 = df2[col_list]\n",
    "    \n",
    "    df2 = df2.T.reset_index()\n",
    "\n",
    "    df_melt = df2.melt(id_vars='index',value_vars=Taxa_col)\n",
    "    df_melt.rename(columns={'index':'Patient ID', 'value':'% MG Composition'}, inplace = True)\n",
    "    \n",
    "    fig_width = 2*(width/9)\n",
    "    fig, ax = plt.subplots(figsize = (fig_width,0.5))\n",
    "    sns.stripplot(x='Patient ID',y='% MG Composition',order=col_list, palette=['black'], data=df_melt).set(ylim=(0, 1))\n",
    "    fig.autofmt_xdate(rotation = 90, ha = 'center')\n",
    "    plt.savefig('Stripplot_of_%s_by_%s_%s.pdf' %(Taxa_col,category,save_descrip),bbox_inches=\"tight\")\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Analysis start"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Metadata Correlations"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Import metadata from UC patients in Cohort 1\n",
    "meta = pd.read_csv('../UC_Severity/UC_MP_Emperor_Map.txt',sep ='\\t', index_col = 'id')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Select numerical data\n",
    "numerics = ['int16', 'int32', 'int64', 'float16', 'float32', 'float64']\n",
    "newdf = meta.select_dtypes(include=numerics)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Remove some columns not wanted for the visualization\n",
    "Remove_cols = ['Relative_Activity_Index','anonymized_name',\n",
    "               'faith_pd','pielou_e','Num_85ALC_denovo_peptides',\n",
    "               'SerpinA1_SerumVFecal','TAXON_ID','ELEVATION',\n",
    "               'LATITUDE','LONGITUDE','host_taxid','shannon',\n",
    "              'observed_otus']\n",
    "newdf = newdf.drop(columns = Remove_cols)\n",
    "\n",
    "#Rename some of the columns for the visualization\n",
    "newdf.rename(columns={'histologic_remission':'Histologic Remission','Intact_crypts_without_erosions_or_ulcers':'Crypts with no Erosions or Ulcers',\n",
    "                     'height':'Height','smoker':'Smoker','current_steroids':'Current Steroids','disease_duration':'Disease Duration',\n",
    "                     'age':'Age','age_diagnosis':'Age of Diagnosis','current_IM':'Current Intramuscular','IM_exposure':'Intramuscular Exposure',\n",
    "                     'biologic_exposure':'Biologics Exposure','current_biologic':'Current Biologic','current_5ASA':'Current 5ASA',\n",
    "                     'SerpinA1_SerumVFecal':'Ratio of Serum to Fecal SerpainA1','ASA_exposure':'5ASA Exposure','stool_frequency':'Stool Frequency',\n",
    "                     'rectal_bleeding':'Rectal Bleeding','partial_Mayo':'Partial Mayo','UCEIS_endoscopic_score':'UCEIS Endoscopic Score',\n",
    "                     'PGA':'Physician General Assessment','mayo_endoscopic_score':'Mayo Endoscopic Score','steroid_exposure':'Steroid Exposure'\n",
    "                     },inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 216x432 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Perform spearman correlations on each metadata variable\n",
    "corr = newdf.corr(method='spearman')\n",
    "ax=sns.clustermap(corr,figsize=(3,6))\n",
    "#plt.savefig('UC40_Spearman_correlation_metadata_heatmap_final.pdf',bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Compare compositions of metaproteome from human proteins by IBD subtypes and activity"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 148,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Load in necessary data\n",
    "df = pd.read_csv('./MGDB_Proteomics/All_proteins_normalized_MGDBFinal.csv',index_col='Master Protein', low_memory=False)\n",
    "df3 = pd.read_csv('../UC_Severity/pDB_Proteomics/2Search/CSVs/NormalizedDataAll.txt', sep = '\\t', index_col=\"datarest$ProteinID\")\n",
    "\n",
    "protein_ann = pd.read_csv('./Metagenomics/MGDB/MGDB_Full_Protein_Annotations_AllProteomic_hits2.csv')\n",
    "protein_meta = pd.read_csv('./Metagenomics/MGDB/UC40_Full_Protein_Annotations_All_hitsV2.csv')\n",
    "\n",
    "meta = pd.read_csv('./Combined_Metadata_Allomics_MP_V1.txt', sep = '\\t', index_col = 'id')\n",
    "meta3 = pd.read_csv('../UC_Severity/UC_MP_Emperor_Map.txt',sep ='\\t', index_col = 'id')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 149,
   "metadata": {},
   "outputs": [],
   "source": [
    "df2=df.merge(protein_ann,left_index=True,right_on='queryID',how='left')\n",
    "df4=df3.merge(protein_meta,left_index=True,right_on='queryID',how='left')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 150,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Remove less certain proteins:\n",
    "df2=df2[df2['Phylum']!='unknown']\n",
    "df2=df2[~df2['Phylum'].isna()]\n",
    "\n",
    "df4=df4[df4['Phylum']!='unknown']\n",
    "df4=df4[~df4['Phylum'].isna()]\n",
    "\n",
    "\n",
    "#Create a column for Human vs microbial origin\n",
    "Human_Prot=[]\n",
    "for i in df2['Phylum']:\n",
    "    if i == 'Chordata':\n",
    "        Human_Prot.append('Human')\n",
    "    else:\n",
    "        Human_Prot.append('Microbial')\n",
    "        \n",
    "Human_Prot2=[]\n",
    "for i in df4['Phylum']:\n",
    "    if i == 'Chordata':\n",
    "        Human_Prot2.append('Human')\n",
    "    else:\n",
    "        Human_Prot2.append('Microbial')\n",
    "                           \n",
    "df2['Origin'] = Human_Prot\n",
    "df4['Origin'] = Human_Prot2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 151,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Make separate metadata for healthy patients because they won't have activity\n",
    "meta2 = meta[meta[\"Relative_Activity\"] != 'Missing']\n",
    "meta2.index = meta2.index.astype('str')\n",
    "\n",
    "#Make lists of patients within categories of interest\n",
    "meta2.index = meta2.index.astype('str')\n",
    "\n",
    "#IBD200 only\n",
    "#Make lists of patient groups\n",
    "UC_patients = list(meta2[meta2['Diagnosis'] == 'UC'].index.astype('str'))\n",
    "Healthy_controls = list(meta[meta['Diagnosis'] == 'Healthy_control'].index.astype('str'))\n",
    "CD_patients =  list(meta[meta['Diagnosis'] == 'CD'].index.astype('str'))\n",
    "\n",
    "Colonic_CD = list(meta2[(meta2['Diagnosis'] == 'CD') & (meta2['Location'] == 'Colonic')].index.astype('str'))\n",
    "Ileal_CD = list(meta2[(meta2['Diagnosis'] == 'CD') & (meta2['Location'] == 'Ileal')].index.astype('str'))\n",
    "Ileocolonic_CD = list(meta2[(meta2['Diagnosis'] == 'CD') & (meta2['Location'] == 'Ileocolonic')].index.astype('str'))\n",
    "\n",
    "#IBD 200 only\n",
    "#Remove samples missing from the original dataframe\n",
    "UC_patients = list(set(UC_patients) & set(df.columns))\n",
    "Healthy_controls = list(set(Healthy_controls) & set(df.columns))\n",
    "CD_patients = list(set(CD_patients) & set(df.columns))\n",
    "Colonic_CD = list(set(Colonic_CD) & set(df.columns))\n",
    "Ileal_CD = list(set(Ileal_CD) & set(df.columns))\n",
    "Ileocolonic_CD = list(set(Ileocolonic_CD) & set(df.columns))\n",
    "\n",
    "#IBD 200 only\n",
    "#Resort the samples by the activity index\n",
    "UC_patients=list(meta2.loc[UC_patients].sort_values('Relative_Activity').index.astype('str'))\n",
    "Colonic_CD=list(meta2.loc[Colonic_CD].sort_values('Relative_Activity').index.astype('str'))\n",
    "Ileal_CD=list(meta2.loc[Ileal_CD].sort_values('Relative_Activity').index.astype('str'))\n",
    "Ileocolonic_CD=list(meta2.loc[Ileocolonic_CD].sort_values('Relative_Activity').index.astype('str'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 152,
   "metadata": {},
   "outputs": [],
   "source": [
    "UC_Cohort1 = list(meta3.sort_values('Relative_Activity_Index').index)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 153,
   "metadata": {},
   "outputs": [],
   "source": [
    "Activity_cat=[]\n",
    "for i in meta3['partial_Mayo']:\n",
    "    try:\n",
    "        a = float(i)\n",
    "        if a < 3:\n",
    "            Activity_cat.append('Low')\n",
    "        elif a <6:\n",
    "            Activity_cat.append('Moderate')\n",
    "        elif a <11:\n",
    "            Activity_cat.append('High')\n",
    "        else:\n",
    "            Activity_cat.append('not applicable')\n",
    "    except ValueError:\n",
    "        Activity_cat.append('not applicable')\n",
    "    \n",
    "meta3['Activity_Level'] = Activity_cat"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 155,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Combine the metadata\n",
    "combined_meta = pd.concat([meta3,meta],join='inner')\n",
    "\n",
    "#Change the indexes to contain dataset ID\n",
    "df4['index_wdataset'] = df4.index.astype(str) + 'UC40'\n",
    "df2['index_wdataset'] = df2.index.astype(str) + 'IBD200'\n",
    "df4.index = df4['index_wdataset']\n",
    "df2.index = df2['index_wdataset']\n",
    "\n",
    "#For origin plots:\n",
    "cols = ['Origin','Phylum']\n",
    "df2_key = df2[cols]\n",
    "df4_key = df4[cols]\n",
    "ind_key = pd.concat([df4_key,df2_key])\n",
    "\n",
    "#Marge data together\n",
    "df_merged = df2.merge(df4, left_index=True, right_index =True,how='outer')\n",
    "#df_merged.replace(np.nan,0,inplace=True)\n",
    "\n",
    "#Combine origin columns when one is null\n",
    "df_merged['Origin'] = np.where(df_merged['Origin_y'].isnull(), df_merged['Origin_x'],df_merged['Origin_y'])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 160,
   "metadata": {},
   "outputs": [],
   "source": [
    "col_order=['Low','Moderate','High']\n",
    "lists = [UC_Cohort1,UC_patients,CD_patients]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 162,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 90x378 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "PlotByCategory_MultiSampleGroups2(df=df_merged,lists=lists,Taxa_col='Origin',\n",
    "                                 category='Activity_Level',\n",
    "                                 metadata=combined_meta,display_number=2,\n",
    "                                 save_descrip='Bothdatasets_MP_Origin_Diagnosis_ActivityLevel_Final',\n",
    "                                 human=\"True\",col_order=col_order)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Correlating individual meta -omic features to severity states"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 332,
   "metadata": {},
   "outputs": [],
   "source": [
    "###LOAD DATA\n",
    "\n",
    "#Metaproteomic abundances from first cohort of patients:\n",
    "#df = pd.read_csv('../UC_Severity/pDB_Proteomics/2Search/CSVs/NormalizedDataAll.txt', sep = '\\t', index_col=\"datarest$ProteinID\")\n",
    "#meta = pd.read_csv('../UC_Severity/UC_MP_Emperor_Map.txt',sep ='\\t', index_col = 'id')\n",
    "\n",
    "##WARNING - METAGENOMIC DATA IS LARGE AND MAY TAKE SIGNIFICANT COMPUTING TIME\n",
    "#Metagenomic gene counts from first cohort of patients:\n",
    "#df=pd.read_csv('./UC40_Salmon_CPMs.csv)\n",
    "#meta = pd.read_csv('../UC_Severity/UC_MP_Emperor_Map.txt',sep ='\\t', index_col = 'id')\n",
    "\n",
    "#Metabolomic abundaces from first cohort of patients:\n",
    "#df = pd.read_csv('../UC_Severity/Metabolomics/StandardizedMzMIne2020Rerun/UC40_MB_Table_normalized.txt',\n",
    "                #sep='\\t',index_col='#OTU ID')\n",
    "#meta = pd.read_csv('../UC_Severity/UC_MP_Emperor_Map.txt',sep ='\\t', index_col = 'id')\n",
    "\n",
    "##Metaproteomic data from second cohort of patients:\n",
    "#df = pd.read_csv('./MGDB_Proteomics/All_proteins_normalized_MGDBFinal.csv',index_col='Master Protein')\n",
    "#meta = pd.read_csv('./Combined_Metadata_Allomics_MP_V1.txt', sep = '\\t', index_col = 'id')\n",
    "\n",
    "##Metabolomics data from second cohort of patients:\n",
    "df = pd.read_csv('./Metabolomics/MB_Standardized_Table_Normalized_Final.txt', sep = '\\t', index_col = '#OTU ID')\n",
    "meta = pd.read_csv('./Combined_Metadata_Allomics_MB_V2.txt',sep='\\t',index_col='#SampleID')\n",
    "\n",
    "##Metagenomic data from second cohort of patients:\n",
    "#df = pd.read_csv('./Metagenomics/Salmon_CPMs.csv',index_col='Name')\n",
    "#meta = pd.read_csv('./Combined_Metadata_Allomics_MG_V1.txt', sep = '\\t',index_col = '#SampleID')\n",
    "\n",
    "meta.index = meta.index.astype('str')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 333,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Make lists of patients within categories of interest\n",
    "\n",
    "###SECOND COHORT ONLY\n",
    "meta2=meta\n",
    "\n",
    "#Make lists of patient groups\n",
    "UC_patients = list(meta2[meta2['Diagnosis'] == 'UC'].index.astype('str'))\n",
    "Healthy_controls = list(meta[meta['Diagnosis'] == 'Healthy_control'].index.astype('str'))\n",
    "Colonic_CD = list(meta2[(meta2['Diagnosis'] == 'CD') & (meta2['Location'] == 'Colonic')].index.astype('str'))\n",
    "Ileal_CD = list(meta2[(meta2['Diagnosis'] == 'CD') & (meta2['Location'] == 'Ileal')].index.astype('str'))\n",
    "Ileocolonic_CD = list(meta2[(meta2['Diagnosis'] == 'CD') & (meta2['Location'] == 'Ileocolonic')].index.astype('str'))\n",
    "\n",
    "###SECOND COHORT ONLY\n",
    "#Remove samples missing from the original dataframe\n",
    "UC_patients = list(set(UC_patients) & set(df.columns))\n",
    "Healthy_controls = list(set(Healthy_controls) & set(df.columns))\n",
    "Colonic_CD = list(set(Colonic_CD) & set(df.columns))\n",
    "Ileal_CD = list(set(Ileal_CD) & set(df.columns))\n",
    "Ileocolonic_CD = list(set(Ileocolonic_CD) & set(df.columns))\n",
    "\n",
    "CD = Colonic_CD + Ileocolonic_CD + Ileal_CD\n",
    "\n",
    "###SECOND COHORT ONLY\n",
    "#Resort the samples by the activity index\n",
    "UC_patients=list(meta2.loc[UC_patients].sort_values('Relative_Activity').index.astype('str'))\n",
    "Colonic_CD=list(meta2.loc[Colonic_CD].sort_values('Relative_Activity').index.astype('str'))\n",
    "Ileal_CD=list(meta2.loc[Ileal_CD].sort_values('Relative_Activity').index.astype('str'))\n",
    "Ileocolonic_CD=list(meta2.loc[Ileocolonic_CD].sort_values('Relative_Activity').index.astype('str'))\n",
    "\n",
    "CD=list(meta2.loc[CD].sort_values('Relative_Activity').index.astype('str'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 334,
   "metadata": {},
   "outputs": [],
   "source": [
    "###ONLY For metabolite averages:\n",
    "MB_ann1 = pd.read_csv('./Metabolomics/Qemistree_IDs.txt',sep='\\t')\n",
    "MB_ann2 = pd.read_csv('../UC_Severity/Metabolomics/StandardizedMzMIne2020Rerun/UC40_MB_Annotations.txt',sep='\\t')\n",
    "\n",
    "##FIRST COHORT:\n",
    "patients=list(df.columns)\n",
    "\n",
    "#df['#OTU ID']=df.index\n",
    "\n",
    "#Change metabolite ID columns to strings\n",
    "MB_ann1['rowid']=MB_ann1['rowid'].astype('str')\n",
    "MB_ann2['rowID']=MB_ann2['rowID'].astype('str')\n",
    "\n",
    "df.index=df.index.astype('str')\n",
    "\n",
    "#Add annotation information for metabolites\n",
    "#SECOND COHORT:\n",
    "df2=df.merge(MB_ann1,left_index=True,right_on='rowid')\n",
    "#FIRST COHORT:\n",
    "#df2=df.merge(MB_ann2,left_index=True,right_on='rowID')\n",
    "\n",
    "#Find averages of r-values within checmical classes\n",
    "##SECOND COHORT BY TYPE:\n",
    "df2 = df2[CD].groupby([df2['direct_parent']]).mean()\n",
    "##FIRST COHORT:\n",
    "#df2 = df2[patients].groupby([df2['direct_parent']]).mean()\n",
    "\n",
    "#Remove 0s from metabolomics data when calculting spearman correlations\n",
    "df2.replace(np.nan,0,inplace=True)\n",
    "df = df2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 335,
   "metadata": {},
   "outputs": [],
   "source": [
    "###SECOND COHORT ONLY\n",
    "#Subset dataframe to only patients of interest in lists above:\n",
    "\n",
    "#df = df[UC_patients]\n",
    "df = df[CD]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 336,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/rhmills/miniconda3/envs/qiime2-2019.7/lib/python3.6/site-packages/ipykernel_launcher.py:27: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/indexing.html#indexing-view-versus-copy\n",
      "/Users/rhmills/miniconda3/envs/qiime2-2019.7/lib/python3.6/site-packages/scipy/stats/_stats_mstats_common.py:130: RuntimeWarning: invalid value encountered in double_scalars\n",
      "  slope = r_num / ssxm\n",
      "/Users/rhmills/miniconda3/envs/qiime2-2019.7/lib/python3.6/site-packages/scipy/stats/_stats_mstats_common.py:142: RuntimeWarning: divide by zero encountered in double_scalars\n",
      "  sterrest = np.sqrt((1 - r**2) * ssym / ssxm / df)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "'Linear regressions performed on CDAI. New dataframe stored as \"df_out\"'"
      ]
     },
     "execution_count": 336,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "##USED CDAI as variable for CD patient correlations, partial_Mayo for UC\n",
    "\n",
    "LinearRegressions(df=df,metadata=meta,variable='CDAI')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 338,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_out.to_csv('./IBD200_MB_parent_CDAICorrelations.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Top metabolite class correlations to UC disease activity\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 339,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_n = pd.read_csv('./IBD200_MB_parent_pMayoCorrelations.csv')\n",
    "df2 = pd.read_csv('./UC40_MB_parent_pMayoCorrelations.csv')\n",
    "df3 = pd.read_csv('./IBD200_MB_parent_CDAICorrelations.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 340,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Add a column for which dataset is being represented\n",
    "df_n['Experiment'] = 'IBD200'\n",
    "df2['Experiment'] = 'UC40'\n",
    "df3['Experiment'] = 'IBD200'\n",
    "\n",
    "df_n['Diagnosis'] = 'UC (Cohort 2)'\n",
    "df2['Diagnosis'] = 'UC (Cohort 1)'\n",
    "df3['Diagnosis'] = 'CD (Cohort 2)'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 341,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/rhmills/miniconda3/envs/qiime2-2019.7/lib/python3.6/site-packages/ipykernel_launcher.py:3: FutureWarning: Sorting because non-concatenation axis is not aligned. A future version\n",
      "of pandas will change to not sort by default.\n",
      "\n",
      "To accept the future behavior, pass 'sort=False'.\n",
      "\n",
      "To retain the current behavior and silence the warning, pass 'sort=True'.\n",
      "\n",
      "  This is separate from the ipykernel package so we can avoid doing imports until\n"
     ]
    }
   ],
   "source": [
    "#Combine UC datasets\n",
    "interest_columns = ['Experiment', 'direct_parent', 'rvalue', 'Diagnosis']\n",
    "combine_df = pd.concat([df_n,df2],names = interest_columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 342,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/rhmills/miniconda3/envs/qiime2-2019.7/lib/python3.6/site-packages/ipykernel_launcher.py:16: FutureWarning: Sorting because non-concatenation axis is not aligned. A future version\n",
      "of pandas will change to not sort by default.\n",
      "\n",
      "To accept the future behavior, pass 'sort=False'.\n",
      "\n",
      "To retain the current behavior and silence the warning, pass 'sort=True'.\n",
      "\n",
      "  app.launch_new_instance()\n"
     ]
    }
   ],
   "source": [
    "combine_df = combine_df[interest_columns]\n",
    "combine_df['chemical_class'] = 'direct_parent'\n",
    "\n",
    "#Combine chemicals by rvalues to IBD severity\n",
    "combine_r = combine_df['rvalue'].groupby([combine_df['direct_parent']]).sum()\n",
    "\n",
    "#Drop non-compounds\n",
    "combine_r.drop('partial_Mayo',inplace=True)\n",
    "combine_r.drop('SMILE parse error',inplace=True)\n",
    "\n",
    "#Identify top correlations to UC\n",
    "TopCorrelations=list(combine_r.nlargest(10).index)\n",
    "BottomCorrelations=list(combine_r.nsmallest(10).index)\n",
    "\n",
    "#Make final combined df to display UC cohorts and CD patients\n",
    "combine_df2 = pd.concat([df_n,df2,df3],names = interest_columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 373,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_melt2.index=df_melt2['direct_parent']\n",
    "df_melt2['combinedr']=combine_r"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 365,
   "metadata": {
    "collapsed": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/rhmills/miniconda3/envs/qiime2-2019.7/lib/python3.6/site-packages/ipykernel_launcher.py:10: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/indexing.html#indexing-view-versus-copy\n",
      "  # Remove the CWD from sys.path while we load stuff.\n",
      "/Users/rhmills/miniconda3/envs/qiime2-2019.7/lib/python3.6/site-packages/ipykernel_launcher.py:13: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/indexing.html#indexing-view-versus-copy\n",
      "  del sys.path[0]\n",
      "/Users/rhmills/miniconda3/envs/qiime2-2019.7/lib/python3.6/site-packages/ipykernel_launcher.py:15: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/indexing.html#indexing-view-versus-copy\n",
      "  from ipykernel import kernelapp as app\n",
      "/Users/rhmills/miniconda3/envs/qiime2-2019.7/lib/python3.6/site-packages/ipykernel_launcher.py:16: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/indexing.html#indexing-view-versus-copy\n",
      "  app.launch_new_instance()\n"
     ]
    }
   ],
   "source": [
    "#Melt into longform dataframe\n",
    "interest_cols = ['direct_parent', 'Experiment','Diagnosis']\n",
    "df_melt = pd.melt(combine_df2, value_vars = 'rvalue', id_vars = interest_cols)\n",
    "\n",
    "#Isolate top correlations\n",
    "df_melt2 = df_melt[df_melt['direct_parent'].isin(TopCorrelations)]\n",
    "df_melt3 = df_melt[df_melt['direct_parent'].isin(BottomCorrelations)]\n",
    "\n",
    "df_melt2.index=df_melt2['direct_parent']\n",
    "df_melt2['combinedr']=combine_r\n",
    "\n",
    "df_melt3.index=df_melt3['direct_parent']\n",
    "df_melt3['combinedr']=combine_r\n",
    "\n",
    "df_melt3.sort_values('combinedr',inplace=True)\n",
    "df_melt2.sort_values('combinedr',ascending=False,inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 371,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Manual reordered the cohorts to fit color scheme\n",
    "#df_melt2.to_csv('./Metabolomics_UC_class_avgs_top.csv')\n",
    "#df_melt3.to_csv('./Metabolomics_UC_class_avgs_bot.csv')\n",
    "\n",
    "df_melt2=pd.read_csv('./Metabolomics_UC_class_avgs_top.csv')\n",
    "df_melt3=pd.read_csv('./Metabolomics_UC_class_avgs_bot.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 372,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 108x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axes = plt.subplots(figsize=(1.5,6),nrows=2)\n",
    "ax1 = sns.barplot(x = 'value', y = 'direct_parent',ax=axes[1], hue = 'Diagnosis',palette=['blue','gray','red'], data = df_melt3).get_legend().remove()\n",
    "ax2 = sns.barplot(x = 'value', y = 'direct_parent',ax=axes[0], hue = 'Diagnosis',palette=['blue','gray','red'], data = df_melt2)#.yaxis.tick_right()\n",
    "\n",
    "axes[0].legend().remove()\n",
    "plt.savefig('./MetabolomeClass_normalized_top_correlationsbymean_to_Activity_vert_redo.pdf',bbox_inches=\"tight\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Annotate dataframes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "##Load correlations to IBD severity:\n",
    "\n",
    "#df_n = pd.read_csv('./pMayo_correlations_MGDBMP_naMask.csv', index_col = 'Master Protein')\n",
    "#df_n = pd.read_csv('./pMayo_correlations_MGDBMP_naMask_UC40.csv', index_col = 'datarest$ProteinID')\n",
    "#df_n = pd.read_csv('./CDAI_Colonic_correlations_MGDB_naMask.csv', index_col = 'Master Protein')\n",
    "#df_n = pd.read_csv('./CDAI_Ileocolonic_correlations_MGDB_naMask.csv', index_col = 'Master Protein')\n",
    "#df_n = pd.read_csv('./CDAI_Ileal_correlations_MGDB_naMask.csv', index_col = 'Master Protein')\n",
    "#df_n = pd.read_csv('./CDAI_correlations_MGDB_naMask.csv', index_col = 'Master Protein')\n",
    "#df_n = pd.read_csv('./CDAI_correlations_MGDB_naMask.csv', index_col = 'Master Protein')\n",
    "\n",
    "\n",
    "#df_n = pd.read_csv('./Metagenomics/pMayo_UC40_correlations_MG_naMask_combined.csv')\n",
    "#df_n = pd.read_csv('./Metagenomics/pMayo_correlations_MG_naMask.csv', index_col='Name')\n",
    "#df_n = pd.read_csv('./Metagenomics/CDAI_Colonic_correlations_MG_naMask.csv', index_col='Name')\n",
    "#df_n = pd.read_csv('./Metagenomics/CDAI_Ileal_correlations_MG_naMask.csv', index_col='Name')\n",
    "#df_n = pd.read_csv('./Metagenomics/CDAI_Ileocolonic_correlations_MG_naMask.csv', index_col='Name')\n",
    "df_n = pd.read_csv('./Metagenomics/CDAI_correlations_MG_naMask_combined.csv', index_col='Name')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "##Load protein/gene annotation information\n",
    "\n",
    "#IBD200 Protein Annotation Information\n",
    "#MP\n",
    "#protein_ann = pd.read_csv('./Metagenomics/MGDB/MGDB_Full_Protein_Annotations_AllProteomic_hits2.csv')\n",
    "#MG\n",
    "protein_ann = pd.read_csv('./Metagenomics/MGDB/MGDB_Full_Protein_Annotations2.txt', sep='\\t')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "##Original UC40 Metaproteomics annotations\n",
    "\n",
    "#protein_meta = pd.read_csv('./Metagenomics/MGDB/UC40_Full_Protein_Annotations_All_hitsV2.csv')\n",
    "#protein_meta = pd.read_csv('../UC_Severity/Annotations/UniKB_Taxa_ghostKO_Merge.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "##For UC40 only\n",
    "#protein_ann = protein_meta2\n",
    "\n",
    "#Others\n",
    "protein_ann.index = protein_ann['queryID']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_n = df_n.merge(protein_ann, left_index=True, right_index=True, how='left')\n",
    "len(df_n)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "##Annotate and save dataframe\n",
    "#df_n.to_csv('./CDAI_correlations_MG_naMask_annotated.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Identify the biases in gene and protein level data for identifying taxa and functions related to severity within the different subtypes of IBD"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "#UC40\n",
    "df_m40 = pd.read_csv('./pMayo_correlations_MG_MGDB_annotated_UC40.csv', index_col=0)\n",
    "df_n40 = pd.read_csv('./pMayo_correlations_MP_MGDB_annotated_UC40.csv', index_col=0)\n",
    "\n",
    "#Second cohort data:\n",
    "df_n3 = pd.read_csv('./pMayo_correlations_MP_MGDB_annotated.csv', index_col = 0)\n",
    "df_m3 = pd.read_csv('./pMayo_correlations_MG_MGDB_annotated.csv', index_col = 0)\n",
    "df_n5 = pd.read_csv('./CDAI_correlations_MP_naMask_annotated.csv', index_col = 0)\n",
    "df_m5 = pd.read_csv('./CDAI_correlations_MG_naMask_annotated.csv', index_col = 0)\n",
    "\n",
    "\n",
    "#Supplemental figures generated with these dataframes subsetting data by CD subtypes\n",
    "#df_n = pd.read_csv('./CDAI_Colonic_correlations_MP_MGDB_annotated.csv', index_col = 0)\n",
    "#df_m = pd.read_csv('./CDAI_Colonic_correlations_MG_naMask_annotated.csv', index_col = 0)\n",
    "#df_n2 = pd.read_csv('./CDAI_Ileal_correlations_MP_MGDB_annotated.csv', index_col = 0)\n",
    "#df_m2 = pd.read_csv('./CDAI_Ileal_correlations_MG_naMask_annotated.csv', index_col = 0)\n",
    "#df_n4 = pd.read_csv('./Ileolcolonic_CDAI_correlations_MP_annotated.csv', index_col = 0)\n",
    "#df_m4 = pd.read_csv('./CDAI_Ileocolonic_correlations_MG_naMask_annotated.csv', index_col = 0)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_dict = {'UC Cohort 1 (MP)':df_n40,'UC Cohort 2 (MP)':df_n3,\n",
    "           'CD (MP)':df_n5}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'New plot saved.'"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x216 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "TaxaRatioCompareSigLevelGroups(df_dict,sig_level=0.3,Taxa_col='Genus', save_descrip=\"UCswCD_MDB_MP_r03_sprc60_Genera_NoRelative_smaller3\",human=\"False\",relative=\"False\", \n",
    "                               variable='Genus',Sparsity=0.60,display_number=10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Compare the ratios of significant proteins to significant genes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Create a new dictionary including the metagenomic correlations\n",
    "df_dict = {'UC Cohort 1 (MP)':df_n40,'UC Cohort 2 (MP)':df_n3,\n",
    "           'CD (MP)':df_n5,'UC Cohort 1 (MG)':df_m40,'UC Cohort 2 (MG)':df_m3,'CD (MG)':df_m5}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "#To remove human proteins from all dataframes\n",
    "for i in df_dict:\n",
    "    df_dict[i] = df_dict[i][df_dict[i]['Genus'] != 'Homo']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'All ratio dataframes combined into \"ratio_dict\".'"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Comparing only significant proteins\n",
    "TaxaLogRatioCompareSigLevelGroupsCombine(df_dict,sig_level=0.3,Taxa_col='Genus',human=\"True\",\n",
    "                                         variable='Genus',MinCount=10,Sparsity=0.6,compareall=\"False\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Concatenate ratio data into one dataframe\n",
    "ratio_df2=pd.concat(ratio_dict)\n",
    "\n",
    "#Create separate lists for most commonly found genera within significant proteins or genes\n",
    "top_mp_genera = ['Bacteroides','Blautia','Faecalibacterium','Clostridium','Roseburia','Eubacterium',\n",
    "                'Ruminococcus','Escherichia','Subdoligranulum','Collinsella']\n",
    "\n",
    "top_mg_genera = ['Bacteroides','Roseburia','Blautia','Anaerostipes','Escherichia','Flavonifractor',\n",
    "                 'Fusicatenibacter','Clostridium','Lachnoclostridium']\n",
    "\n",
    "#Subset out ratio dataframes to only include these top genera for plot\n",
    "top_mp_ratios = ratio_df2[ratio_df2['Genus'].isin(top_mp_genera)]\n",
    "top_mg_ratios = ratio_df2[ratio_df2['Genus'].isin(top_mg_genera)]\n",
    "\n",
    "#Optional - manually sort the dataframes by top genera and cohorts to align with manuscript figures:\n",
    "#top_mp_ratios.to_csv('./TopMPGenus_ratios.csv')\n",
    "#top_mg_ratios.to_csv('./TopMGGenus_ratios.csv')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Reordered for being in order of genera with most MP or MG correlations\n",
    "top_mp_ratios = pd.read_csv('./TopMPGenus_ratios.csv')\n",
    "#top_mg_ratios = pd.read_csv('./TopMGGenus_ratios.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 420.389x280.8 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Make the plot\n",
    "plot1 = sns.catplot(x=\"Log(Pos/Neg)\", y = \"Genus\",hue=\"Category\",kind=\"bar\",palette=['blue','#BDD0FF','gray','#E1E1E1','red','#ffcccc'],\n",
    "                    data=top_mp_ratios,\n",
    "                    height=3.9,aspect=1.08)\n",
    "\n",
    "#Option to save the plot\n",
    "#plot1.savefig('./MGDB_wMG_Top10MPGenus_Log10_Plot_r03_sprc06_10min.pdf')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Identify the types of proteins overrepresented in <i> Bacteroides </i> positive associations to UC activity"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_dict = {'UC Cohort 1 (MP)':df_n40,'UC Cohort 2 (MP)':df_n3,\n",
    "           'CD (MP)':df_n5}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Filter to leave only the Bacteroides proteins\n",
    "for i in df_dict:\n",
    "    df_dict[i] = df_dict[i][df_dict[i]['Genus'] == 'Bacteroides']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'All ratio dataframes combined into \"ratio_dict\".'"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Identify the functional differences (via kegg functional categories) in significant proteins and all other proteins within each genera:\n",
    "\n",
    "TaxaLogRatioCompareSigLevelGroupsCombine(df_dict,sig_level=0.3,Taxa_col='KEGG_Functional_Category',\n",
    "                                         human=\"True\", compareall=\"True\",\n",
    "                                         variable='Genus',MinCount=10,Sparsity=0.95)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "ratio_df = pd.concat(ratio_dict)\n",
    "ratio_df2 = ratio_df.sort_values(by=[\"Category\",\"Log(Pos/Neg)\"], ascending=[False,False])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Save and reorder by the top UC Cohort 1 findings:\n",
    "#ratio_df2.to_csv('./Bacteroides_r03vs_sprct95_KEGGRatios.csv')\n",
    "ratio_df2 = pd.read_csv('./Bacteroides_r03vs_sprct95_KEGGRatios.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYgAAADfCAYAAADsv0TdAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy8QZhcZAAAgAElEQVR4nOydd5hV1dWHX5Bm7wY7GHWJxhYjNoLYS0ws0ViwYFQssZegWAATo+ZLLCSxYUEjTY2xayQiKiJqYuzwiygIUbFhRRkYh++PtQ/3zJ3bZrh37pT9Ps88c+fcc/bZ59w7Z+299lq/1WHRokVEIpFIJJJNx2p3IBKJRCItk2ggIpFIJJKTaCAikUgkkpNoICKRSCSSk2ggIpFIJJKTaCAikUgkkpNoICKRLN56661FQKN/Zs6c2aTjWmmbkXZANBCRSBa1tbVNOu7bb78tc09g9dXXLmt7NTV1FelnpG3SqdodiJQHM+sH3AW8mdr8saRDqtOjDHn6NlrSTU1sb29gPeBxYKyk7c1sLHC0pAVL2t+WxLLLdqZnz5lla2/GjB5layvS9okGom0xQdJh1e5EHsrWN0mPAZhZj9S2lnrdkUirJRqIdoCZTQReBn4ArAAcAqwFXB52WQ1YNvy9kaTzzGypcMzBwB3AbKAHMDa0szXwsKTBZrY5MBzoAHwK/FLSFyX0awXgZmCl0IcRkq4P/X0lnOdr4Blgr7DfnsD+wCbADam2ZoZtqwM3Ad2A+cBA4GN8BrMisDTwa0kTS7l31WbBgoVFR/21tbV06lTav3JNTV0ZehVpL0QD0bbYNTxcEx6W9H/h9QuSzjSzy4DDJV0B9DOzVYCHgKOBt4CXzOx8YG/gSaAG2AB/MC8NzADWBr4B3gUGAyNwo/CmmR0H/Bq4sEjfdgM2xF1E95rZWsBTwPWp/p5hZo8B30jaw8xuB3Yucg/+AAyX9KiZ7QZcAfwO6A7sDqwBbFykjRZDly6dGTZsWMF9hgwZUnJ7XbvGZcdI6UQD0bYo5Mb5T/g9G39YYmbLAfcBl0h6KWx7Ch+tHwtcGo55R9IXZlYDfChpbtg3iWbpBVxnZgCdgf+W0jczmwOcaWYHAV+GYxNeCr8/J7N28Rk+MyjE5sBgMxuEz2gWSHrDzP4CjAnnGF6kjUgkQjQQ7Yl6oYlm1gX4G/AXSf9MvTUCGASsJunV4OcvFtYofIF4lpntBKxZYp/OBZ4LbqVdgJ/k628jmAb8QdJkM9sE2Dm4wJaX9BMzWxOYjM+aWjx18+c3aoaQs42aGjp27VqmHkXaE9FAtC2y3TgA++TZ9wzgh0AnMzs5bDtI0vNmtiHwl0ac92TgjrBuAXBcicc9CFxvZv3xtYtaM1vSJ9m5oc1uuEvsDNx1NsTMjgYWAJcs4TmajY7dujGzZ88laqPHjBll6k2kvdEh1oOIpDGzjsCzwF6Svqx2f6rB1KlTF/Xq1aspx9GU44pRbgNRpn52WNIGIi2fOIOILMbMegJ/B25sr8ahpbFw3rwlngFEF1OkqUQDEVmMpBnAVtXuRyTD9Fmzlni0H41DpKnEmLdIpAWz3nobNvqYmOsQKRdxBhGJtGCaIrUR5TQi5SLOICqMmfUzs8/NbN3UtivMbEAj2xlqZic14fwDzayzmW1lZg2id8xsbNBKKqWtRWZ2fda24SGLudBxpxZ5f2IISS2lDyPNbO/wM7CUYyKRSNOIM4jmYQFwm5ntIam5w8YGA3dIehmXzlgSPsXzCjpJqg1hrT8q4biLgD8v4bnrkegxtXUWLFjIW2+tU7KUBriLKWZMR8pBNBDNwwR8tvYrUg/KkIQ2VtL24e8pwGG4jMVIXHuoAy6DQeq4y4G+oc2rJN1tZlsDfwK+wzWITgD2wLOmx5rZNcBJkg4zs18BxwMf4NITmFlnXNtoo9DuRTn0imqBiaHdR3H5jX8m/culyQScCqxiZtcB55NDeym0famZrYZLexwt6WMz+yPQJ7w/WtK1qXswANdeGkoOnSUzm44nxG0U7v+KQG9Ako6ildClS+fiO2URjUOkXMRvUvNxMnCWmW1Uwr4XAg9I2jG87p28YWb7AD0l7QTsAlxoZivhGdCnStoZuA43HLcAc3Cjkxy/Ip48tj0uetclvHU88ImkvmF7vkS50an2jgBGpd4bAfxKUj/gEfxhfRkwV9IpZLSX9gT2A85OHXuvpF3x5LkLzGw/oGfoZx/giGCAsvk+bgR/GvqzTNjeA5+59AVOD/dkO6BPuF+RSKQI0UA0E5I+Bc7EZwb57nuSfGTAc+G4CZLSD+HNgW1CxvRjuLbQ+sBawY0E8DSwWZ5zbAK8IalG0kLghVS7+4Z2/4ZnWK+a4/hnga3De6vign0JiSbTRHz2sFbWsXOAA8zsTvzhnR4ePx1+Tw7X3wt4RtKi0M8pwKbZnZH0Bm7MxuBGILm3n0qaFY6dJ+nN4N77guJ6TlWnrqbGf8+f3+QCRpHIkhINRDMi6UFct2hA2DQfWMPMlgqj2iRldiqwLYCZ9TWzK1PNTAOeDKP0XXH3yjvA+2a2RdhnZzKCeXXU/5zfATY1s6XDGsLWqXbHhHb3Ae7GxfGyr2ERPju4Hhf6q/c27h7qhyu6Phy2J4Yv0V46MrSfzsZNZkk/Bl4P96BPuAedgR1xyYx6pHWWgGNwNxu08rKYHbt2ZWbPnnTs1q1R6w+RSDmJBqL5ORP4FkDSHGA88CJew2B62Od3wP5hJD4MuDF1/IPA12b2DPBvYJGkr/A1hz+H7WcAZ4X9n8Ef6B3COT/GtYgm4+sI88J+NwKbBDXXycC7kvIF1I/C3VB3Z21PNJmewWW2Xw3b3wyzhgeBM8xsUrgPae2lA8L17gFcIekhYIaZPYfPHu5JFGezeAuXLX8h9KfV6CxFIi2dqMUUiWTRErSYEnmMhfPm0XnZZcvSZkLUYoqUSpxBRCItkEQeY/qsWVXuSaQ9Ew1EJBKJRHISDUQk0oJprBZT1GGKlJMYHtGGCJIZJ+Uo7TkWjy5akOOY7njJ0VOytl8BTJM0MrVtAHAbsL2k58O2znjC3Z8lDc3Tr/WALUMUV673e5BKGCzhOudI6h6S/66S1Gb9MI3VYoo6TJFyEg1EO6BAneokkuqUfO/nYBpwOPB8+HtvPLegELvi+Rc5DURTkXRmOduLRCL1iQaiHRDE9DbBpTQ6AOsCy+ESGfMJo3cz+zmewPYxnmE9LUdzjwJ7mVnHEAZ7OJ6klpzrNDyjeREwFk9iOx9Yxswm48YkKbK8TOjDAmB1M3sAl/54WNJvwsziFjyhbhFwuqRXUueaCJyEJ+z9EViI5270B36OZ1cvjdfIvhYPzf0BcK6k+xt3F6vDggULyz4raEoEU21tbczHaIeU9Imb2fckfVjpzkSahbclHWNm+wK/x2UoEn6PJ6zNJZPkls0CPMt7ZzP7F7AC8D+gu5ltChyKJ7gtwnWa/oHnRGwi6QEzOwU4UtL7ZjYYOATPq1gOOAr4GngmGIuLgeGS7jezrXBjkUsc8ADgXuAPwM+AlcP25SXtaWaH4Xkh2wP98DyRVmEgunTpzLBhw6rdDYYMGVJ8p0ibo9RF6r+Z2d/NbL9QszjSepkQfieSFoAPAoAvJX0asqUnF2hjND5zOAh/MCf8AJf9eCKcZ1VcfynNe8BwMxuJa0klchuvSPpC0ne4/MfGuNzG0wBBRmRdcvM7fObxBHAwPpMA+E/4/TkwNVzXZ7QCqY1IpCVQ0gxCUh8z64Xr61xkZk8At0h6p6K9i1SCbYBJwE7AG6ntnwIrmtnqIdt6W3xmkIuJwDW41tIR4QdcauMNYB9Ji8zsLOA1fNSeDCxuBjaQ9JWZ3U4m4aqXmS2Hu7y2wzPLp+LSGw+EGcScPP3pD4yUdK6ZXQAMxDWiWmQWaGNqRNfNn98iRu/RxdQ+acwn/j6u47MNPlK81sz+IylKG7Qs9gyun4Qjst7fx8z2B5YiowlFqO9wLPAPM5tLZhTeAEl1ZjYeWFfSl2aWbH8lDB4mBQmNF/AZw2u46uxLwF+B583sM+BDMoJ+c4FxwOrAOElvmtm5wIjwuzNwXJ4uvQjcbmZf4y6wgbgeVYsk0VkqhR4zZpT9/E3JpI7GoX1SktSGmd2FG4U78ZHa+2H7vySVUjAm0gIIbp2x7aXYTlNpDqmN1mYgchClNtoBpQ4Lbpb0eI7tfXJsi0QiBairqSn5wV8JLaZIpFRKXXC+KNdGSfPL2JdIhZE0IM4eqk+p6w8QtZgi1aXUGcQiM/s7vghZByBpcMV6FYlEgOJSG7H+dKSSlGogbq1oLyKthlAnehu8zOcyeODCx5IOKeM5hgJzJN2Q5/3F0h1tXW6jmNRGlNaIVJJSDcQo4ES85ON/8WpikXaIpHNgsS7TJpLOr0I3Fkt3RLmNSKRylGogbsSTjcbj4YM34xIJkUgiEnglHmJ6E14x71dkIl0OxqPgBoV9euKhrJeZ2UFh+0JgJqnvVSiJeiOeILcqLvMxlPrSHWfjchtz8Ci7FfDv9UWSJpjZq8BTwBZ4XsT+koppRy2mMfH/5SoWlKaY1EZ0MUUqSakGYiNJfcPr+8I/ZiSSppuk7QCChMZPJH1jZjcCe+H5EOvjD+queF7NZXhG9tWSxprZ0fgDPmFdYIqk482sG/A/SRcHpdlEuuPssO9FwHhJ15rZ2nguxvdDe2MknWZmo/B622NLvahOnTpVVeqiWJJcNA6RSlLqt6ubmS0DYGZL40lWkUgapV5/hCeu3YYbhERO4zVJtZLmEepy4zOAvqEW9o6EIIjAXGDb8GC/Gjcs+UjLcrwHfIkn3UFGcmM2UWYjEimZUmcQ1wKvmNnr+DrE0Ir1KNJaqQMwsxWBYcB6Yft4Mq6mXFmZA4Ghkj4Ks40DU+8NAD6XdKKZbQgMNLMO4VzZg5tEluM/YQaxMi4fku+8JVFbW1uy1EVjJDRKpW7+fDp2izYtUh1K1WIaZWaPAhsAMyR9WuyYSLvlS+BZ4CVgHi6OtxaQLzPsBWC8mX0KfAU8BJwW3nsCGGtmPw5tvRXaSkt3JPwOuNXMDsYlvgcG+ZAlupjGSEw0RkKjVCqRSR2JlEqpUhtPUn8UthCfrv9W0szKdC0SqQ5NldqA0iU0SiVKbUSqSanDo5n4qPAZYAe8EMtzuD7/bhXpWSTSylg4b17ZH+hRaiNSTUpdpF5P0s1yRgIrSLqFWJEuEllMJWQxotRGpJqU+oDvYmZ74bOGHYHOZrYBnkkbiUQqRCGpjZgDEak0pRqIAcD/4aGGr+OFg7bHQxQjBTCzQcCZQM9SxA3NbG98xnbTEp53O9wluJOkFwvsNwCYK+mBrO1zJHXP2jYR+J6kXqltBwF/w69vZp5zHAg8n8jE53h/KAWkNXL0dxO8YNElkk4pdkxrppDURpTZiFSaUqOY3jaz8/Hyka8C78VqciXTH0/MOgwYWWznMqqtHg/8Ec9oHlDgfEX7lI2ZbRVKgIJf17tFDjkDz3bOaSCagqQ5QJs2DpFItSnJQJjZqXh8+ir4Q24j4NTKdattECQo3gZuIBRbCqPwV3Dpia/xUf5ewErAnsD++Aj5BmAMHi32feAFSSeb2UrkkJTIOu9yuF7RZsBrZraapE/MbA3881sJj0I5Gjdgc4ARuEzGZqHP+QL6x+DZzy+HvnQLxyc5ELfgshgAp+P5EFsBd5hZHzxH4kfA8nid6GPDvgea2S9wt+Xpkl4ws/747KsGD3EdmLrGHnjxo+3N7LJwvR3xrOlrit1nSZ/lub6qkE/So5DURnQxRSpNqd+uw4Dd8aSla/GawZHiHI8XWxJQE9w+4A/73fCH8DeS9gDepGGZzI3xMpu9gX3NrDsZSYm+wCHALWaW/TkeBtwbXFrjyJTqvBB4QNKO4XXv1DH74HIZ2wMXkH996cHQlw64xtI9qfcGA09I2gV/mF8v6WHgZdwYdQM+C9e7I7B9SGoDz6/ZNfT1BjNbFTcmu0rqg2uBnZinT0fjpVX7ksnQhtLvc9VJJD2yf7p06Zz3mGgcIpWm1G9Ysl+SC1FTgb60KcxsZWBf4AwzewxYkcysK0nw+hx/YIEnlGWnzE6X9JWk74APwvuFJCUSjgd2COftC5wYjIjhgQZImiBpVOqYzfCkNYJ09uw8l/YtLl2xAz6rvDf13ubAL8PofQSezZx97BpmNgYX4VuOjAxHck1v4FLiGwBvSPoq9f5mefp0GHA58A98hpBQ6n2ORCI5KHWRegz+D7q+mT0C3Fe5LrUZjgRukXQeQNCymgF8QunSD7n2KyQpgZltDiwVZgLJtvHAfuHYbXHZlL7AT8iMuKfhrqNrzWwtIBnZ52I0HqAwV9LXqWzlacCdkkYHd9bxYXsijbEPsK6kQ81sddzAJAlXvYHRof+z8Hu1qZktG7Sbdsal5uthZl3xmdThoa03zCwR42uyxEZzkkh05JL0iFIbkWpSqoG4Hvgn7s8V/g8cKczxwFHJH0HZ9G9kHppNJaekROr9E4C/Zh0zAp+99A/HHok/PI8jyGtLut/M+pjZ8/ii8ycF+jAeuB04Nmv7ZbjLayC+RjI0bJ8M3AH8DLjYzKbgs9B3cOkMgJ5mNgF3B50Y1kyGAE+aWR0wHZf5Pix9Qkk1ZjYXd2N9BjxOK/t+FpLoiFIbkWpSUGoj+LxXwP+5j8JHaEsBt0vqnffASKQV01SpjSWRsGhOAxGlNiKlUmwGsT0eomh4hAu4u+AflexUJNKeqKupyWsIotRGpJoUNBCS7sMLBO0r6ZFm6lMk0q4oJBE+fdasilSqi0RKodQ1iLlBq78zPrVcS9JeletWJBKB/FIbMQci0hyUaiCG4zIbB+Na/F0q1qNIJLKYfFIbUWYj0hyUaiA+lzTGzPaUNDSUh2w0QeDv98A6wDd4iOWvJb1hZiPxzNhySU3k60M/4C48Lr4DPiv6paRpZnYNcFXIA6jEuQcCtwE7ASdJOqzIIfna6U6FdYjM7CSgu6ShqW39gCeBwySNS21/FXhJ0oA8ba0C7C1pdIHzNdB+KrDvTDzb/ExggqQXSjkuEok0jlINxCIz2wxYxjzovaR/5DQhD+AB4ARJz4VtvYG/AP0a294SMiF5OJvZnsAfgP0knVnh8w7GI8KWiCrrECX5EuNgcd5FsVXULfAQ17wGoilIuqKc7bVECkltNJWmrmnkkwOJtF1K/bTPxrNYh+P/5Nc34Vw/xR/MzyUbgt7OLumdzGw0MErSw2bWC394H4yPvNfHR/ynAW8AN+OZs6sBIyRdH7J4X8ZzNlYADpFUSExuZbwgUqJWehKuLdRA7yhb8yf06SVgY0nfmdmVwL9wgbx6fcClSrrjwn3XABuFMq5rAA+Gmdnm+D3ugCe//RJ3540L5+wc+vcVBXSIsu7nqcBB4dgvwusj8CzvZXCdpysljQxaSdcCc4HvgCk57tcrwMZmtpKkz/GEwFGEGtRmdgj+ffkOmCTpfFzWY8swg5oMXBX6uxKuuzQZ6BoS3NbFBSFPwbPP8+pOJbNOPJ9iJF7psBbP7dgIlwypCW3eEO7TlsC1kpryHW52unTpzLBhw6rdDYCSa3NH2g5FV7nMbBVJb0i6CxdM2yv7IVQiPfFkp6Td+8MDeZqZrZPabwRwTHj9S1z87SRgpqQdcGXS7XBl2bGS9sSzhNPS4y9I2h1P6Do8R192NbOJZvYccCv19YQgv95RPc0fSV8Ak4C9zGwpPFP4/lx9CAWW5pBJ9OoGHIBnRScSHCOAX0nqBzwC/BrPMP4itH06/rBMk0+HiNDnVYHdJf0YNxLbhrdXlLQfPrI/P2y7OvR1D/LXkAaX1zgw6DH1xh/6iStpGLBb0E9a28z2wBPoJgQJ882Ac8K9uYpMst3SwCBJO4U+/5TSdKcA9gD+jRvhy8hIfKwD/Bw4ObR1VLiP+TSdIpFIioIzCDPbGVfh3CqoX24J3Gxm/SVNauS5ZuMqngBI2j+cY0pWPyYCw4NUw564W+ZPwKPhuNeB14McxJmhHsGXZDR9wLWCknPmcoelXUwGPJcSjQPXOxoVzveemSV6R4nmT/ekP/hD/XTc2P5T0oIgPVGsD69Lqgl9SDKhewHXheM749ISj+Kj4fvxEfJvs9rJ1SdC3+vMbAEwxsy+xh+YyX1K5Lpnk9EmWltSImfxLG6Ec5HMIt/BVVITNsTv0yPhGpbHNZWU2uc9PJv62/D+l2H7rNRMbzKee5Pvc8jmFmAQ8BhuTAeH7a9LWmhmnwNvh8+mVWkx1c2fX/aReyLt0Viii6n9UWwG8Vtg50QaWdJ4fLR2eRPOdT+wu5mlNYI2xB9ai9O5JS3C3QrXAo9LWkhGQwgz2yC4oc4FnpN0JHA39TM7G6PB82GObYneESm9o8/JaP7sCgwws/WDofw+LltxS5E+JJpE+d4XcHSYQfwaeBhfn/kgzJR+i0ttEPqW1iFa3KfU+1sAB0g6FHfLdSRzn3Kdf05w60FmptGwk14LZFncMN6ZemsGbnD2CNfwJ+D5rOseDgyRdAweEZf0Zx0zWzO87oMXpsr1OSzWnUqxP/BMUG69GzcW+a6xVdGxWzdm9uxZ1p+mGAcgGod2SLFPvFZZVcIk/Tdo4zSKIOr2U+CK8CDohPuLB0p6NyX4Bu5Pno0vboIrf94aoqeWwqNXlgeuN68Z8ClQGx6YpbBrcG99F9o5W9K3qT7k0jsqpPkzCl/reKPIeZ/BXUf5nMon4zO2pcLfx4VrG2dmZ4b+XprsXIIO0XRgnpn9C/fFf0BG+ygXRwK3m9lX+DpHoZoJ44Cjwvdhg9Cfj83sKuCpcA0z8YixlYHNwzXcCdxvZh8C/8PXjwjXOTy4GydLejToQjXQncr6roCv+9wZZmJ1wFk0dMVFIpFGUkyL6SlgF0l1qW1L4YuPO1SqU2G0eEcYEbZ4zOzXwCeSbq12XyJLTjW0mPJRCamNMrUZtZjaAcVcTHfi/ustzWx5M9s0bBtX5LgmY2Y/x33JF1TqHOUkRNL0pb6rJRIpC9NnlT8lpxJtRtomxbSYRoSFwatx18RM4LZ0klS5kfQ34G+Var/c5EsOi0QikdZO0VWnYAxyGgQzGyKpZQRpRyJtkFxaTFGHKdJcLGlYQouq7VtIyqNM7R9IJiqnJKmLRBZCXh861/sH4NFJN0kaXo5+5jlPo/teoK2JeJLdvLDpOzz66v0l7WcT+zMAr273QDXOX0lyaTFFHaZIc7GkBqLFLFQ1k5THGbiG0jTKJ3WxH3CBpAfL1F4+yt33o0NbmNnJeNjx2YUPqQySRlbjvJFIW2dJDURLijMvKOVhZuviRY+6AfOBgXjI7Bg8pPb7ePbzyWa2Ip7TsGpo6nRcSmIrPAz1SDzKansz2w9IMpn+gz+EG4QBm9lQPJt8DVwy5Cy8vOZ+QG8z+wRPKjsTD0l9K/SxP55R3jGc50Y8kWwjYAIuR9HbL1dHmdkPyJKywMNMc/V9D3z2Mp+MtMdWeB7BgtDfcZIuK3LvVwG+NhfzuzIcexOeOZ7d/pd4LkRvXEZkiLzc6eX4Yn9HXDDxbjM7Bc+qr8Mj584LiZGD8KTBmXgm+SXhXNNy9T3k24wMx7wL9Ah5GmWn3MlkubSYoosp0ly0pcyXBlIe+MNzTTPbDdd0Gh7i63cDrsA1gjbGM7a/Ad4xV0o9C3giaDtthC/M9zGzl3HZjwXhHJ2APwO9JX1kZpfg7q18YSI1kvYJD+ZzJO0dHnhj8azpvwJbS/rKzK7GJSG+Bj5LZZ73wJPiPsA1k7bDk+DeMbOVyEhZvGZmRwDHSjohR9874A/xPiFL+QxcjuIh3IBtgRuw93H5imzuMLNv8Ie38MS+bYBukrYL7b+To/1ngdUk9Q73+tSQ7d1T0k5m1g2YYmbjcRmO0yRNMbOTw/0+HLha0lgzO5qG+Q65+v5/wO8kPWJmJwA98nw+S0ynTp3Kqp2UK4s6GodIc9FmXEwUl/LYHBhsZoPwfi8Iu06X9FXY9wN8hrE5nkx3aNgn0fbJZjX84f1ROOelefZLSMtvZMs9bAC8kfQFeBo3XM9TX6riUwU5cjObJ+nN8PqL0GY+KYtcff9S0nup8/0ONxCvSarFkw+/zXP8YhdTQkhgS/qar/1PgOdgsSrtRSGPZJuwtgEuB7I+biDONRdBfA7/3M4GLghuranAfVn9ytX3XgS9KDxZsX+ea4pEIimKaTHtme89SY/j0/uWwv3A+Wa2vaQp0EDKYxrwB0mTzWwTMgvsudxk04A7JY0214Q6PmxPS0YAfASsZC5oONfMhofj8tUnKOSSmwFsambLSpoX+pfoIqVdVsXcesOB/pKmmtkwMqPl7L5/AqxgZmtK+iDrfEviOkz6mq/9qbg8CMGVdxe+TvSkpIHmYnwX47OP3+Iuu/lm9g9gR1yQb2iYsd0IHJh1/lx9fx3YAdeq2j7H+2Wjtra2rNpJdfPn07FbeaWjSk3ma6pmU6TtUGwGkUsJFfyf8HFJs8vcnyZTgpTHubg0RzdctuGMAs1dhiuHDsRdGEPD9sl4PYeB4Zx1wU/+sJl9h88QXmxi/z8xsyHAk0HKZDqustrYokL5pCyy+74ouFvuDef7DFfK/UFT+p9NgfY/xTW5JuGf0TA8MbKfmT0DLAf8PbjZXgNeNLOP8ZnR8/jnMd7MPsXlQB7CXWyFGIRLdpyLi/ktLMc15qLcekWJFlM16DGjkKBvpD1QUGojH6lRYSTS4jHX63pe0nQzOx7YUdIv8+3fkqQ2gJZqIFqSezlSIUoa7gRXxSl41MkyuKtgswr2KxIpJ7OBsWFR/TtcBLFVsHDevKqN5KOLKVLqfHgf3Jd/NR5CeV3FehSJlBlJT5MKYGhNTJ81q+yzklJnOq3dOJiXSf49PqhdDldSHhpKCmTvux6wZTPkI7UqSo2X+1Re3GZ5SdPxGx6JRCpATU0mJiGX1EakOCHkeyxwpqRd8OCEzclfTXBXYKdm6l6rodQZxP/M7Jd4bYHLiVr7zUZIPrsLeDO1+WNJh1ShLxOB70nqldp2EC6u2DO7dkhqnwPxNYCcUhwhiXCOpBtK6MMAYBO8rvcSSYa0VLp27bhYXiPKajSZ/fHE2bcA5DXjjwYWmpU0hn8AACAASURBVNnNeJ3yVfHItqF4QMgyZjYZjyjMrg3/JR5t9yM8KbMnnpwLnlTbGQ/eOV3SK2b2Lh4NKbz+e+8Q6XgysJyk/6vw9ZeFUmcQJwJPAOfhyUeNjayJLBkTJPVL/TS7cUhjZlul/jwMz04uxBmUeVAhaU5bNA6RsrEWHiq9GElfA2sCUyTthVcuPFnSd3ji7Oig55WrNvzPgFUl9cbXsNYNzSYJuH3x73lSVXJd4AhJp+MFxZJn5lF4NGGroNQZxLJ4xm43PEzwR9Qf0UaqQBjRv4yHpq6AV7V718xOA47ARzRjJQ03r1uxavj5CR7Kmx4N7Y9XpCs20hmDhz+/HKbx3UIbSV5DIYmSPnhY64/wJL6pko4N+x5oZr/A3ZenB5mU/jSUHkmuvUe4tu3N7DLcRdARGCPpmnBvXgn35ms8QW4vXH5kT4Uyui2Rmpo63nprHTp16sSCBQvp0qVz8YMi2bwL/DC9wcx64g/ubYMEz5d4xn02uWrD9yKT4PmxmU1L7ft02P6yuaQPeAGxpDzuLXhVyKfxmXKuMsctklJnEPfjFrRX+NmkYj2K5GJXM5uY+jkv9d4LknYHxgOHmxd1OhQfHfUBDjBbXKNzgqQdcc2j7NFQHaWNdB4E9g1SGgcD96TeG4xLlOyCP8yvl/QwbsSOxo3JZ5L2wJPetjevHggwQ9KuoT83mNmquDHZVVIfvCZ4Pv/x0bhB7Isr+KbvzW74Q+CbcN43aWEqxNl07dpxsWRHNA5N5iFgbzP7PoCZdcYDbLYCPpfUH/gj7lbqQP1E0ly14ZNkS8xsZVyiB+rXTd+KMFgildwalA8+x6V90nXrWzylziA6Sjqyoj2JFGKCpHxuvbR8R3d8xLw+7hIElwlJVjoTGYx8o6FSRjrfhnPugGcxH0pGHbaYRMm3wBpmNgYf1S+Hj9AgMwp7I2g0FZIeyeYw4PJw/Y+mtr8Ufn9OZsb7GQ1lTiJtDElfmtkxwIiQnb88Prh5Ag95/jEuV/8W7o56DbjQzF4id234t4B9whrFHFy7bSGuYjwiJGF2Jn8I9Qh8XaNVPUdLNRCvmtl2+EhwEYCkBYUPiTQT2SF7At4A9gnZzGfhX/5DyIxqXsdnCNekR0OSZplZKSOd0bgm0tyQwZ5sLyZRsg+wrqRDzWx13MAkCVe9gdFmtjkudlhIemQxZtY1XNvhoa03zGxsnntTNZqSUzBkyJCKSG20FyT9G3c9ZrN5jm3vAZb6u1/6zSDP84ykX4XZ7Ru4G6kG2CPHubtnbeoE3BrWO1oNpRqIncms2IP/421Q/u5E8rBrSsguYZ9cO4YIiieASeHh+QL+5U/zMLlHQ1DaSGc8cDsuppemmETJz3AhwSn4usI7+OgNoKeZTcDdQSeWKj0iqcbM5uKDl8/wdZQWV3S5Y9euTcqIjnIXLYbZwJVmdiZeJmBQMA5FMbPf4W6o/SvYv4pQstRG8NOtjudEtCorGKlPGA1tFSSzk9HQ+uFh+wvgB5IuqW4vq0elpDZaioEokyRIlNpoB5QqtdEPuBWPYFrZzE6QNL6SHYtUlJyjodY80mnp1NXUNOlhv3DePDovu2wFehSJFKdUF9Nv8cIv74eok3txN0OkFRJ8+g2MgKTBVehOu6CpshWVkNqIREql1DDX75Is2FAAZn7luhSJtF/SMhsQpTYi1aXUGcSXIfnqaTzWfG7lutSyaYwAWI5j++EFcIpmoqcTwRrZv72BwyQNKHH/gjIYWftOJEptVJS0zAZEqY1IdSnVQByJ1xO+DI8nz6ul35ZJCYAdJOmtECd9N57AVfTh1kI5A69VXdRAJJjZVpJeDn+WKrXRqHMUQ16utE0Zh7bA/Pl1c7p16/i9Mrb3YbduHbNDRheTa9BlZlcA0ySNDJnNfwTWwAuF/RsX8FuQ1U6jBn5BmWCspMcae00hyu82SQ0KV4XB1CGSjihw3lOBPwMH4AO2mvDeD8P17RJ2TzTcFuHXPkrSn8zsJOAtSU9QhGIlR9eR9D/ge3j4Y8LqeEhheyOfANiCYCxuJCUCJuniLImL/wM2Mi+fuSqeaXyLmW0N/AmvVTAfOCGcb3Uzuw/Xj3kVN0T/JUsOA88avRVP/JlH+GxSgmFTgZvxTNKOuNzE6XgiW1oG40SyJDpy3IMotVFB0jIbQKuT2ujWreP30jOgJWXGjB5NNjbhf/J+XG/p+bDtWuBSPGQ62a+5B36D8bDvegYi9G0vPGS7ASEB9d+p3KMP8HD3pC57f+rrTy1OsA0h7zKzv+LPgvFmNrFYRGqxNYizw+8b8Rt1Y+p1eySnAFgYjaxLlghYardE4uIzPNvyp3i00KCQMDYCOFXSznitjavCcSvguQY7ALvh5UNzyWH8Bne37I7nHCQkgmFn4gWezgn7XAUcmyWDsSH5JTrSRKmNCpKW2YhSG0tMH2B2YhwCg3ADkabBwA//Tt0KYGZ/NLPnw0+6VPGJZjbBzP5tZr3DvueY2Ytm9pyZXRm2DTWzx81sspn9Cs/4H0tDJlP/uZHNaVnHJYM1Qrb4D8lf8nh5fABaK6kWVxn4SYFzAUVmEJISA3GVUoU0wkivPVJIAOxl8ouAKfV6SjK9NbM3gR7AWimXzdO4siTAO8lI18w+wkfXDeQwwvT4hXDMs7iUBtQXDHsPT1L7Fv+yfJl1bfkkOpS1X5TaiLQGFpF7QJcrwCaf8itmth8uZrk9/rycFBI6wUfzvw3rYgPC/9Yv8MFPLfC3cDz4bPmM0OZ55FDEljQuuMwaYGZLA+tJ+ji1+QXgIDNbFv9/fBLYNPV+kmBbh89WTkuuC/dI9AMeyHW+hGIupv3wIhqHm9kOYXNH3OLeVejYNspDwGAzu17S25YRABtPRgTsRDPbEBgYRtmQEu4CtjazTrgB6QW8DbxvZltIepX6khIN/J955DCm4V+Qx4BtU7unzzsc6C9pqnkJ2R6pfTqSX6IjF1Fqo8xkS3EMGTLEt0epjWJ8S0NF1uXC9v8BP0+/EWamO0h6KLW50MCvFy6xsQivJTGFzEP43+H3HHzwtgk+AFwY2niGTGnm7IFWY1kZ+CTH9gfw5/Hu+BrxZan3Cmm4fUBuGZJ6FHMxvYL/039LpvjFG4RpTXtD0pdAIgA2EZiC36Pr8ZH3vubyFdeTEQHLZj4+yp2IL4LNxdcc/hy+UGcAZxXpygjcRZUskJ2CG64ncFn2XNwJ3B/OsXGqb4kMxuxwDZPM7F/ARjSU6EgYH84/Jmv7ZcAvwr15DNd8Sp/jX8AG4Z/sHnJLbdxAkNoAEqmNKbh77frsjoQFukRqYwItVGqjGIkUR/ZPNA5FmYoPutYEMLNuuKvxJfz/s2fK/dMBl3/pm9VGPuXXH4T2+6S274j/b0PDAcg0YDsz6xTO1ZfMoCY9WEsrx5bKp/jMP5tRuDtsTUlvN6K9lYGPiu1UktRGGA1uIemfwYc2StLnjehMpIxYlMOoKJWS2ihGLimOVii1sajMi9SFzgUsDrW+CNcU6wKMkDQivLcBHvGzbPiZApydI4ppGzyIJK38OizMpv+Ae1K6AHdJujIdxWSp0HIzOxt3u3YEJuEz7SGkwrjN7HbcnbtLdpSUFQiFD4O7n0v6KOv8/wZukXRdmD0na8R5Q+rN7Crg8WJRWKUaiPHAjZLuMbMj8IXP/YodFyk/lpLDCLOPSJmphoHIp/ZaCamNShqI5g5zbU+Y2eFAd0lXL2E7nXAvwO5LGsWUsKykewAkjcb9bZEqIGmwpB9H49C2yCfFMX1W6/KWhYd5h3L9RONQj7HAD81suSVsZyBweTHjAKUnyi0wsz3w6Vlv6vvTIpFIGaipqaNr1/pjtii1EUkI7qijytDOdaXuW6qBOJ5QnBsPF8wXj14xQmLUq2RCF8FX6bNjmpP9R5KV6RgWsKZJ6mFm1+Dhu0s8RAsROTfg0RMd8KiI0yV9W/DAzPEDgdvwiIef5bumSmKeXdld0tDUtn5ksjE74CGp10iqeARbWJg+TPnlO/riUWOvmtm9kg6qdJ8qTbbMBkSpjUh1KclASJqOp3VXmzfldWKXmJA8Vi7OA8anFqGuwaUlSvUVDgbuCLkQObMoq0g6G3M54Ckz+28qb6Na/BKfcr/aFoxDJNISKbUexGC8ePc3+EhykaRcIZxVwcz+SAhFA0ZLujb13nJ4KNjKeFWyZPtE/CF+GJ4IswYeWXCWpH+EHJBL8RoYn+EPoqF5uvAucLCZTccT1c4lhMCZ2TnhHLXA05IGmYvT7YjPOEYRMisTwyLpMDN7K7RlwId4PHcXPFx0LTwstW/252BmpwIH4aP9L8LrI4B98bWj7wNXynVq+gDX4mGi3+EuxLyEnIcb8Qzql83scjyUryM+G7s7hBT+BfgKD6Obj4cWPoiH6j2CJ7sNCc0ugxeI/6+5ZMbe4dpWC9ezDh7e2g2X8Lg0vL837o99E8+Y7m65JUs64uG4s8O1vyCpULZqxaitrV0soZGP7BlDa5PaiLQtSnUx/QLP9v2mkp0pgU2tfunN/sDW5M90BBgAvC7pQvO62rmSQ2ok7RPWWc4xs3/i7rQdQqbyqCL9uh7PFTkP13CZBJxirvHSmMzKtHLrBrjMxGwzexZPgNsOl6Q4xLwq3BvpTpin26+KRyfUmWs+JYlzK0ray8w2wh/WI/EZzuHh4dwgxyAPH+IP5n1wBdedgutuSoh2uwE4KmREXwYkUhrdgW0kLTCzU4Aj5fVFBgOHmNn9uLHZFjecSaz5JsAfJU00sx3x0MM9zOwx3IU4K5WoNwI4XtLLZrY/Hst+Lp73sSc+wHnHzLrLxf6alURCozEkCXORSDUo1UDMpL7GTbVo4GIKYbf5Mh3B/fqPAUh63swaKCji0hHgo8xuuBjhl5I+DNufwR9w+dgFdxHdGjJ7f43LUY+l6ZmVn0iandWvXqlrmWZm6bR7glFYAIwxs6+BdchIWbyc1RbA2pKSRJ5ncWmNYqyPZ6huDmyTMtidw3trSUoM1zNkJAVmpGLP3wOGhz6uHc69GfAvSXW4vHySxf0BcJGZHYfPygoNp/NJlkxPJDvM7AOi1EbFWLhw4ZzOnTuXLcx14cKFH3bu3LnNq7maC13eieuvdcFzNZ7Lc96WoeaaogvwWuqfdpHyyNFWgam4oN3VqUzH23FZB8jIUNwfXBC5HjDZX4KPgOXNbHW59sn2uJHMxxn4iH+EvHTnG/jDfBo+I+mEuz364i6iLSmeWZnri/l6uJb7Qtbnauk3zWwL4ABJ25nZMviXJYlXz9XeHDPrJWkqPnIvqHBqZsvjbpuDcdfXk5IGhpnLxXhm9Gwz21TSm9SfEaWv92ZgA0lfhaShDrjBPD20tTQZI/8b/L4+ambH4jPCpL3se1ayZEk1qK2trTcjyJf7kKa1SW107tz5e42dJRViyJAh7UXN9Wxc6PIa8ynxGBrKfzS7mmupBuLKEvdrdiQ9ZGb9zOw5MpmOL6XcDn8BbjOzSfgDu6aENuuCL/8RM/sCfxC9Bb52kWOh/CTguuA6+Rb4GP9Svm9md+Ej5CSz8j7cQKR5Bh+xFPvPugUYaS7U9y4NK/tNB+aZS2XU4F+gQmtFRwK3m9lX+JpBLgORCH59h39fhkiSmf0X6BdmRcsBfw8P/FOAW8PsYAG55Tr+CjxvZp/hLqu1glvoblyN8n0yMgB347ONOaTWJvB1jCvMLJ1qnEiWdMBdescVuPZmJ3v9IZHXKEQlMqnbEfnUXLMHFnll/KHgGueJZvZrYEX8//2Fxqw5Uj/w52oyz6ZO5K7aeRquX5aQqLneZ41UczWzRM21oFhfqZnUR2dvk3RH0QNbMWZ2Ab7wWmNmd+Jp6XeY2bXJ2kEV+rQjsJykx8NawmOSvl+NvuTDXIrlLkkfm9lvgQXVCNtdEpozk7oaBqLSUhtlnkEUOlchF9NU/AHfW1JBbbPwv/5NOrgl9d5+eMTczwlrnHjY/zm46zJRc+2Nr0XejBuTWrzS4q14/ZOVU2uOM4FNlENZ1lzJ+FHcDfZUavvSgCStF/4eiRuZK3AZkB1wI7QpmVlP4mJK1FyvlfRIOP4YYEtlFLtzUuoMIvk2dcBVS+fiU6S2zFf4wus3uHtpXNj+h6r1yKePY8xsCO4q+1UV+5KPD4HHwwziC1zcMJKDupqaogagElIbbYw2o+ZqrmQ8Fjg3bRwCVVFzLTUP4oLkdZi+P1Rg9zaBpD/jC0HZ22fn2L1ZCJE3uxTdsYrIJVnuKbpjpOj6A7jURhlG+22ZxWqukj6wjJrrNfiAqqeZ9Q7un0TN9VvqP8MKyfgXWuPMpebapDVHM9sUd6ceKumVHNdZSM31Wnxd+G3LWeMrJyWpuZakxWRmXZIfvIRk4XlxJBIpmZqa/Mo1UWqjMHIJ/rOBh8M649PAnyRNDxFxhwBDzewp3D/fAVd+zW4jp4x/mGnMCGucU4B7JKXVHNLtvIa7dZ7Fi/nMJLOAnOYZfH0z7Tq7HI+uu9bMJpqHfafbrsGDStbI2i486vJBGsd2ZIqD5aXUNYgZuLXsgFvf/5N0WyM7FIm0Cqqh5ppPIrsSUhuVXINo7jDX9oRVQc21WEW5TpJqJbWIGUMTYpUnksmWXqzHnmO/kTQhpjlEJxyBR90swn2hgyVNbEw7rRVL6eBnbT8GH5F9hz9Ifh8W1s/H/aIvNGis8eceivt+p1Al/apIQ8r9MO/cOWaRpxgL3GFmyylTOrQplE3N9XHCQoaZHVvNWUMVYpVL5SplNJh64T7BHxY+pO1invBzMbBpyJpeC3jBzNaTdEWRwxuNWqZ+VaOoqanLO1Moh9RGKRIfkZaPWqCaa3oaeRSuOFotisUqN9AFym4gOyTOzOZI6p56fwU8TG0lPN5+hKTrw0zkY3xhZ68ClncVICl2/i6+aDUVXzC7BY88WoQrvb5inh18MrAUcL+koWZ2CO5T/Q6YJOl8M9sJzwZdiOcq9AfWxOUyFuIhdUdLei/XfQi5Ccfgi2OTJJ2XdV8OxiOiks/7YLzc4qBwf3sC4yRdFozgrcC88JOdO/E1/r062cweCgtn3w+5JSNxI/8UOTSlwn1+OZx7BeAQSe+Ga/oRvkg3VdKxqb73Ywn1q6pNtsR3mi5dOjdaniObKNcRaSrFFqlbRAZqYC3qZwki6eswSl2sC4RH+VwYZhyNZUPc1bQnsB/+oE4YLSmXz+7ssKj0RNj/hLB9Xbzy3pkEqXRJffGs61vCYtP5eHW4bYAVzWw9PFluN0l9gLXN9aEOAO7Fs4NvxQ3VHniYXRLetnKB+3AscIakHXAtouyBwcbAT0ICoIC9wvb18YfsDrh8CHhm8yWSdsdrTdcj3J/d8ZrWjwVD+cus3Qbi0hs74VElaZ/1C6Ht8cDhwWh/JmkPPIJkezNbm9xsAFwcrnN1PDu80LkikUgBis0glg0JWR2BZcLrDgDKaPg0F4VilfPpAhUje6FtDnCmeY3bL6kvy5EvjnmxiymLTyR9Gl73wqMrCBnD6+IPs9eVqRlxlrkS6up4hAP4iHkD4HfAhXjUwXt4FvEt+Aj/MTzfYDD578OxwLlmdiXwXI7r/gjPqP4aj+VONGBek1QL1JpZ0s/N8AgN8NF6vdXO4FJaWtKp4e+NcUMxKbVbIU2ptC5WdzwoYg0zG4PPTpYjvx5To/WrWjp18+c3aQaQlvGILqZIUyk2g/gGuAn38X8bXt9IdXz+DwF7m2sQkYpV/gHuynkyjIB3xUPN3snRxnzcNYOZrY+7hNKcCzwn6Uh8fSP9IG1sFb30/lPxmQJmthVuiN4GNjHXSMHM7sHdIrOBPcK1/Ak3Bv2BkZJ2wRVcB+Iut2ck7Rb6OqjAfTgBd8PsjKvf7ph0LKwZDMMX8o/HP+dC+k2JthVklGLTdAdGmdnK4e938QSftDhaoimFNdSUyj7nPsC6kg7HjeDSFMjizbGt0LlaPB27dWNmz56N/knnWETjEGkqBb854YGUFzPbX9L9hfYpF5K+DNExI8x1R5bHY38TmepcukDZzfwL+NzMnscf2tlprA8C15tZfzwxpTZ5gC8h54Z+n4uPfo8LUhRX4gV4FgEPBn/7VWHbUngc9V14dFQywl+AG4iOwJ1mVosbo7Pw0Xeu+/Aa8GIYPSczkIQv8ZnAS2TWFNbKcW8STgHGmUuUf0yWZoxcB2s4MME8C70TcLMkpT6PYppSaV4ALjbPYK3BDV5j1hAac67IElI3f/6cjt26lc2NVzd//ocdu3UrGBmVL7oRnz0nVSg74P9Hd4Yk2Ow2VgV+J+lEcymMqZL2Sb1/Ni4736FQ9KKZ/QW4VBkl6FZNSXkQ+TCzCZKKpmtHImmsGTWlmnKuauRB5KOpUhuFlGIrrcVUTF+qMQQpkkJaTCvhiWfZ0Y2P467FsZK2D/t2xhPXbpD0YFY71+OJca8GA/EVsIukT8L7/wR+KGmVJMQ6O3pR0g/NR0GDJGWvu7VKSsqkLkDeDy4SKcA7wAXmhZBGUVlNqeY8V9mZPqtpJdNLkfFoIzSIbgSOxoM56iHXSLoWODS9PQRCbCuXiU+4G8/Cxrw419vUd5OmWRy9GDKbe4UZSatnSZ2TLSnKKdJKUDNqSjXnuSJVIWd0I0AOFzP4Ol/2OtT2NAxCGYOvuV6Py+KPwo1Rwtlmdhgejv45mehF8HW6nSgipd0aiKtXkUiVqampy5sLEbWYilIoujHX9CupiJhmNdxwpJkNdAgRhzvhyZ9p8kUvgiulxhkE0cVUEiGZK13+bwV81NNfWaUPS2gnSQq7V9JBjezHQDzZcTOqJE9hXu6wu6ShqW39KMP9KXDOskl8VIKuXTs2qxZTG6OQEms9AxECTs7AhfHSfIQnx2YzFk9QfU7SojwzklyUpJTaGijZQIT49k64rGwSa35VRXrVNqmnzW5mo4Gf0URp7MYah8BgvHZ2S5SnKOv9SVMJiY9Iy6BIdOP6wKYhL6gOjyAcJemfWc1MIXfVzLuB4XgNnMawNamSpq2ZYmJ9mwJ/DpFKE/AQyLXN7ARJ/8iOBIiUhrls+poEmYo88hgTcV/mJvhMLXthbY6k7ma2Hb7w1gEPYe2PV7dKsquWwRftfkwodWhm15CZifQHzsRDSN/CQ2j7A/uGY78PXClpZNb5TwUOwv/pvgivj8h1nJn1CX2ci/tsp1To/nyM5+msi0/xH5V0sWUkPrrjWd0d8dKpReWOm4NyaTG1hIS4uvnzP+wxY0a5w1wL7iPp3+QufjMTn40WO/5rM3vRzLaW9B9JPcJb83GplmS/7uH30HxthWfm65K+KHbe1kCxb9OVZCQWPpC0i5ltCIwA/lHRnrU9ktrOa+CjmZskPZGWxzAvdjLFzMaHYyZLOslcS2kwLreRzU24ourUsF8v3IV0pLwm9mBc0+gyM7sYT4hLwv5WxZPktg75Elfj4odfAytK2iuEhj6I6z4RjuuIP4B3DxpL/yCTNJfruKuBwyX9N4QTVur+XIVX9Do+7Ps/GvqOP5O0Py2IcmkxtQTNpWI5C01or5zNFeISXLLmhGI7FuE0Gn7nWi3FwlyXkfSv8PoLAEnTyS91EMnPhJDh/GM8XC5JREvLYzxGfZmQCeH3ZFyALhffkzQVXKVRXszkPWB4GDnvQv7PawPgDUlfhb+fJlMiMXFBJZIVi5EXYlmAlz+9BVgndY5cx62dkmZ5Nk9fynF/5gLbmtko3CjlivUsWvox0v6Q9JGkJTUOSDpZ0gfl6FNLoNgMYunkhaQDUtsXVqY7bR9Jn5rZkcCT5rIbiTzGwDAyv5hM2N42+Ch4J1xiIxfvm9lGIUloEPBffFaxQZgV3E4mmCC71OEM3Ee7rKR5uBhg8iDPG8JsZlsAB0jazsyWwUUDC8lzzDGzXsGQbUtDBdjFLOH9GQB8HrJhNwQGWv2qXck9qAgbrrde2dtsjBZTS3AxRdoWxb5N71mo55psMBeUm1PZbrVtJL0Z5CiGA78gv0zIgJDiPw+XW988R3MnAreaWR0eXncN8FfgeTP7DA/fS6QpnsFlCIaFfnxiZkPwh3EdMB1fXMtX6DxhOjDPzP6Fr118QGH5iyNxqZCv8AzVvAYi9Kup9ydZY/lx2PZWkX6Vlc7LLks5s4hhcSZxSUTjECk3BaU2QjzxA7iK6HTcJbEb8FNJTUvxjJREcKmcJGlatfvSEqnk/Wmq1AZQVQNRKhWW2oi0IYqJ9c0IM4af4kVjXsT19uc1R+cikdbEwnnzyv5Ab6oWUyRSDkrRYuoq6S58BtEZOCgIYkUqiKR+cfaQn5Z4f5qqm9TcbUYipVLQQJjZAUCSVHIxHkLZHzipwv2KRNoFNTWF18yj1EakmhRb1TqNTPnJzyRdYF5gZjzwl4r2rJ0QZCaexHMZxqW2vwq8JGlAE9vtQUrquNKY2ep45il45ul/8YJTf5V0S3P0oRghN+JISTdXuy8JhWQ2IEptRKpLMQPRUZmymU8BSPrCvBBMpHxMAw4HxgGY2eZAq3I8S/oY6ActeoG9O141r8UYiEikJdOYPIihqe1xDaK8vAJsbGYrSfqcjLzwegBmdghwNi5TMUnS+aFoyYa4EuUqwHXAz4GNgWPwUOTVzewBPDv5YUm/Cclzq4afn+LZ8rmkKWqAHrjkxQB5pbgG/Sjl4pbwnCNx2Y5uwB8kjTOzN/GQ3c3w5LjD8eS6W8O+S+GSHOOCsfoYF1BL8j4uqYZQYS4KyWxA46Q2ihHzJCKNpdi35TkzO1WpEn1BjfO5AsdEmsa9wIHhgdgbf4iuZ2ar4HkLP5L0jZn91cz2CMd8K2nvoFa6r6SfmtmxeB7DNXjewFG4dMYzwViAZy1fHdxQ+aQp3g0JZyfgCWeDc/VDUiJ7UYymnPM8PBP8R3gC3p5hv2Vw0bWnzez3eC7IAuATSUeZ2fLAS2aWaC2NlvT3cO7NW4pxgMIyG9A4qY1itAQpUl1zYAAAFJpJREFUjkjropiBuBBPwvolnr3aM/w+utIda4eMxhUo38FHxwkbAqsDj4TksOXxfBTwWrvgBUveDK8/IyNx8UoiGmZmL+CzC8jITSTSFLvgtanT0hT/Cb9n45nKhfpRCo0+Z0iIOxXPDF8BuDO8v1DS0+H1ZGAfoJYQUBGOexOfTaTPHYlEGkGxPIhvgMPM7Hv41P9/kt5rjo61NyS9Y2bLAqcDF5B5+M7AH5h7SFpoZgNwvaMDKF7Rr5eZLYerUm6HP2h/QkZuYgD5pSmy287Xj1Jp9DnNbE1gG0kHhtnGbDP7K9DZzLaU9AoZmY06XMfp72EGsTkZPae61O8lLbPbrDRGaqNUkiS5QnWrIxEoLvd9rKTbJH1oZqslxsHMhkgqz7w3kmYccFRQPd0AfPHXzK4Cngr5JzPx4jqlMDe0uTowLkhYpN9/ghKlKZawH006J76O0t3M/oO7yf4gqTZcwyAzWw8vCnMRblxGmNkkfO1smKSPsq73I6CLmV0paVAT+t7sdOzWrezZ2QmVyNKOtC2KSW1MkNeCyPs6EmluzGwmsImk+ZVov6lSG2WSsGhACzUQUWqjHVBsDaJDCa8jkUiFqIR8R0J0MUWKUcwfu6iE15FIsyKpR6VmDy2NSkhtTJ06FSAah0hRis0gVjWzPfEZwyrp1xXvWSTShqmpqSsa4gpRaiNSXYoZiE/wJCTwEMTk9YoV61E7I0ht3IWHqS7CF1hHSfrTErTZgxJlNsxsZeAPwEZ4gtls4MSQMT9HoQ5vpTCzvnhU06sl7NsduETSKU08VzdgmjI1h6tGMYmNhCi1EakmxYYwawC/lnSspGPxgu9vU9ywRBrHhKBOugte1e0cM1upmc49BnhIUl9JOwHPAzc207nBv1MlFfWRNKepxiESiTSeYg/6YXhi1G641PcoXA5h60p3rB2zPC5lUWtmWwN/Cn/PB06QNMvMzsGzpWuBpyUNCtIbO+LZ08cBmNnGwJ2Seoe/x+Ghoi+Gv9cHukv6e+r8w0MbAF3NbDQu+fEpcDDwPTyhrxsulXGppPvM7HVcoK8GOC/PPvsBSVD/f3BDtDfww5DYth25JUXS13Ubnl39aGinE7A9XpN6Hbzw/Hf4QOZEPBFvFC61Mb30j6GyFJPYSIhSG5FqUixR7h4z64Srt64MXCspqriWn12DZlAdXu/7NElfm9kI4HhJL5vZ/sBVZjYML8O5I24g/hYevABTJZ0RXEyEfIpvzWxTPKegZ2IcAmuRSSYjHPMd8EX4czlgsKSZoX9b4xnNf5Q00cx2xAcR94V9fyPpP2a2e/Y+ZvYQ8Gegd8hPuATXSHoMGIvnOeSTFMm+rm/xMqQd8PKqd+AZ6I8CfUL7v8GT8roCr0u60My2A1pEeHYp6w8QpTYi1aXocELS2GAkTgBGVL5L7ZIJknLVgV5LUpKt/DRwBbAJrmW0EMC8VvNmYZ9ckhIj8AflLDJSFQmz8FH3YsysM3CIpNHAXEkzw1tzcA2kD4CLzOw4fM0kPbxNzp9rn9VwyfiPABI9pFQiWyEpj3xSGX/ypnSTma2Bi/zdFY5fGng8nPexcM7nzWxhnrYikUgWxTKpx+D/4B1wXZtJZjYdQNIRle9eu+d9M9siLODujLtwpuFrFJ1wV0pffAS9JRlJiTT3AOfiLqJD0m9Ies/MPjGz/SXdHzafgYsFjiZ3OPNvgBGSHg3CgANS79UV2OcjYCUzW0XSXDMbjhusRP6ikKRIg+sKM4QOkn4TNn2CC//tHxbYf4bPSjYHdgDuDy678vhrmolCUhuNzWOILqZIYyn2bbkhz+tI83AC8OfgSqkFjguaTXcBz+IP1km4i2fLXA1Imm9mTwOrS5qbY5ejgL+Y2blAF9x3f0KBPt0NDDezOfgDfbVS9pFUZ2anAA+b2Xf4GsSLuNvqCuBQoCQpDzPbFtermhhcXwCX4sbtYTPriAsBHo3PvG4LEhzT8DWSVkMhqY3GJtBF4xBpLAWlNiJtAzO7DrhH0oRq96U10FqkNpqaYV2mfkY1hXZAHFK0cczsceD9aBxaJ4WkNqJURqTSRAPRxpG0Z/G9Ii2V6bNm5R3tR+MQqTStShs/EmlN1NTkihloHFFqI1JN4gwiEqkQpcppFCJKbUSqSdkNRNAWehI4TNK41PZXgZckDchz3CrA3iH+Pl/bJWsDJTUDgDPxPIMXSryERhEkMR4Fvkq7c0Ikzfl4Oczv8JDR0yW9VqCty4C9gHMlTSzx/OsBW0p6sMkXUQbMbG/8Mx+Q2nYHMFHSraltZ+HZ1V8DE4BXgSMl3bwE5x6A52w8UGzfSCRSOpWaQUzDhf3GAZjZ5sCyRY7ZAvgZHn9fNiRdUc72cvADfBH451nbf42HgO4cQjy3xWPxLUlyy8GhwNaSvmrE+XfFDWFVDUQebgJ+C9ya2nYMcECSgBeyo48HmmwgJI1scg8rSKlyGoUop9RGQrEIppgvEUmo1LfgFWBjM1tJ0ufAkbgeznoAZnYIWZo7wIXAlmY2EC9EfxW+RrISPvKejGsDjQXWxUeep+DKsnfiEhCdgIvSETtmNhKXcngHGIlLWdTiMfIb4fH0NaHNG/AH7pa4rMj16YvK1kACLsblI9Yys2GS0hlNA/F6ynUAkl40s21DElgDjSU8mWwdPI5/H+Da0KdVgUclXWxmG+EP0i7AN8AR+CxlGTObjOccZLfbETcenwKP4CP3Y/Dks0mSzsu6xoOBX5EJYzwYN4KDgAVAT7x86WVm1gt/+M8LP5+l25I0ycxWN7P1Jb0bjOScIN2RfC4/BzYN0hu3klvD6bVwvzfHs6o/xBMEa4B98e/OHHxgkqufB4XtC/H8iqOTz6WSlCqnUYhySm2USpTkiCRUcpH6XuDAkOTVG3/oJ66kYcBukvoAawfNnctwV9BNuHTEOZJ2xw3FsaHNpYFBQXV0VeCneD3i8ZL64pnCtwT3TjZ7AP8Gdg/nWjlsXwd/SJ0c2joKdwudmD44zIISDaQdceOyJxkXVvZ/1TKSsh+Yn4aXI4BTJe0MXAdcFaQn5oQ2V8flNPYC+oS+gctyXy5pB1zobks8yWx0cK80aDcc1x3YU9Lvw708I7TxTsjITrMx8BNJ/fCH8V5h+/rhPu2Az47AM6YvCZ/TZHJzCz5AIJw7Wyn2MuDNcP2b4BpOewCn4oYKXHZjdPiMfwxMDq+7kJEZScjVz8P5//buPdiqsg7j+FfEQlPLNBO8pDL2pJZaWOYVKHO0vIQ1k5fMWxGaWTo6pmnqlGmlOTpjlqYjVCY2k2PjbXQQ72gqWljy84LhJUUgFRQvIPbH712c5Wbtwz4H8HB5PjOO56y113vW3oez3/V7136fF84v/95uJi8mzGwRlmYdeSV5NTgFuLO2vV3mTj1v5zngNEmvl/2zyvanI2Jq+foeMsFzS7I6qaIjZpX2W11GXkXeRIbRnVK2P1Ku6l8GnoyItyS9RF7F1rXLQLqvzfN/SdLaEVGdO5JGAONozliq+x/wWUnDy3OvPs8oYEJ5rleXNg+rHdeu3aci4q3y9eHACZJ+UdpqnfD0IjBa0qvlOU8o2ydFxDwyZfb1sm1roLq3czf5u2g1Bhgn6TxgGHBsw2Mq3eU8TSz/f5lcOwOyYmn9PTWd5/HAyZKOAh4lZ54vF7qL2ui4DUdyWC8ttQoiIqaQ9x2O5d0hcfXMnWHkkMh9dGXyQEZOnx4RhwKT6HoT20jSwPL1LsAj5B/8rgCSNiQrg+pKvW4/4M6I+CIZBXFS2d7pVPLJwA6S+peqaDcyG6md0cDp5bGUVNNfk0M//5W0TXlclbFUdxi5iM7BwHnkENIq5bl+trR3sKTv8+7XrV279eGU7wCjSpXxabIaorT5QbK6O4C8L/A6Xa990+s0mbxSpzqvVhExo5z3acA15c27rn7+PwXGRMQh5Acd6p1Xp7+npseNBM4oz3kVYESHbfW5Kmpjcf5r7RyqJUfbcedglaX9L2EscEiJnd4cICKmS2rK3FkH+JSkH5IdyrWSppEBbFXez0wy42cjcpjhRkn3AZeXsfPVgZERMa+WElp5APijpHnkm9Jx9GCoISImtclAGtrmkF+Rb3gTSoLoXGDfUqEslLHUcuw44CpJu5Jj+4+T0dwnAr+TdCp5D+Kb5JDKjyVNpCG7qeG8JgH3S5pOVmr1CmhWeX4T6bqnsFAkeM3RwFhJJ5LR3e3Wib6UvP+x0C+FrFjeVyqaTnKeeuPvwC2SZgKzgeuWULtmKzRnMZm1WJaymOa+9hqrfWBRHwDsXusQk7OYrFOeSW22DHvi6acXuw1HclhvuYMwM7NG7iDMeqEpZ2lpRH07i8n6kj+uYB3rTYzK8hqhsihLImepE85isr7kCsJ6qopRATqKUakiVJaoiDinrzoHs5WFKwjrqbYxKstrhEpvLImcpUp3E9OWRhaTWafcQVhvVDEqV5AxKr8AtiMn2W0fEXMk/aEWoTIqIi6R9A0yQmWSpIPIWd330BWhMrXMNdmHnIh4S0RcUCZA3iVpcMO5VBEqx5MTJusRKtsBQ8j5FYOBDYFryBn+i6UpZ6m3Hx/t379/27wl5yJZX/IQk/XGleRs693oilHpT1eEym3AVmSESl0VoTKaDAGsLo3bRajcARmhQk7iaxehMoOMUDmGrCKgRKiQ0RxPlqiRpmgOM2vDFYT1WERMkVTFqJxMdgTv0BWhMrdkRD1MDhHVI1QOjohHJZ0JbFq2byRpYEQ8T0aoXAZ8lKwIHuowQuVMSQeSESqj6Tyao8/NmzevbaUw/4036Dcg+zSvQW3vNVcQ1ltjgY0josp7mk7eX7i9xJ/sRWZBPcnCESp3kqmxg8qxVYTKBGBqRNwI/Bz4gqQ7yEiTkQ05TpARKmeVNkeR2V7Lle6yj+pZTO4c7L3mqA2zFstS1AbAfzbbDIBNn2oXidUzjtqwTnmIyWwZNve11xZ0DB5isveaKwizFg8++OB0YOoiH7hymzFkyJA9+/okbOlyB2FmZo18k9rMzBq5gzAzs0buIMzMrJE7CDMza+QOwszMGnkehFkPSFqdnBG+PjAbODQiprc85jgyqwrghhIDsgrwLPB42T6BjA/Zlkyd/XZEPFFrYx/gJ2S21OURcamkfsBvujlmUfsPJNfReJuSmhsR8yU9BLxSHvZURBy+OK+RrThcQZj1zFHApIjYFRgDnFrfKWlz4GBgJ2BHYA9J25BpshMjYlhEDAPuBwZExI7Aj4Dzam2sBpwP7AEMBUZK2gD4artjirb7S8f2M2B4ROxExqnvLWkAQHVe7hyszh2EWc/sQibHAtwI7N6y/xlyBb23I2I+mVj7Bhk7vqGk8ZJuICPNbwKIiHuB7WttbAk8EREvlRTau8jgwl26OYZF7H8T2Cki5pTv+5fz2hZYQ9LNkm6V9PmeviC24vIQk1kbko4EjmvZPI2u4ZjZ5JX4AiVifEYZUvoV8FBEPFYqgLMj4i+SdiHXpbi6dujbkvqXQMK1az+j/nNat9ePobv9pbOaVp7X94E1gVuATwLnAr8nF1q6UZLaBCPaSsYdhFkbEXEZGT2+gKS/AmuVb9ci15ug5TEDgMvJN/ajy+YHKGtVRMRdZchnrdph/WpvyrNa9lU/p3V7v5Y38m73l3sUvySTdL8WEe9IeoysVt4BHpM0ExhIVkK2kvMQk1nP3A18uXy9F10LJgFQKodrgX9ExHcj4u2y63TyBjGStgX+W7VThnUm1Zp5FNhC0oclvY9cmGlC/Wc3HEMH+39HLpj01dpQ0xGUexWSBpFVyPMdvha2gnMWk1kPSFqDXJBoIPAWcFBEvCDpeOAJYFXgz8C9tcNOBiaTn35ak6wkjgF+AGxDRmcfDnwGWLMsz1p9iqkf+Smmi2qfUlpwTERMrp3bQvurNskK5gGyQ6v+6C8ArifX9N6kbD+prBNu5g7CzMyaeYjJzMwauYMwM7NG7iDMzKyROwgzM2vkDsLMzBp5opxZhyQNA0ZFxAGLemw3bexKfvT0WjIwbyL58dIBwPiIOKUXbV5IZjdNIec4XFu27wkcEBGH9bC9UcDjETGup+diKxZXEGbvkTKJ7gzg4rLp3yUgbziwMzC8BPv11GYR8RQwB/i1pPUW81R/D5wqadXFbMeWc64gzBaDpC+RKalvADPJmcmvABeRYXkvAJuR4XwfJzuFtyS1NrU68H5gjqRNyYiP1cjq4tiI+IekK8hU2AHAuRExVtLWwL9LG7PJWdG/Bb7ecp5DgbPIqO8nge+Sf/9jgEFktMZuETEoIuZJmgh8BfjbYr5EthxzBWHWS6UiuATYPyKGAreT8d/7AutGxOeAI4GNyyHDyGGlylaSbpM0nhxyuqCs33AucGFE7EbOtr5M0lrAcGB/MuKjurrfG7iu1ubFwNqSDmo5z0tr5/kccBgwklz/YWeysvlorZ1/lvO1lZg7CLPeWw+YFRHPle/vALYm47onAJTFhCbXHj+tdvyCIaaI2D0iRpftW5a2iIiHgY0jYjYZz3EJMJasNiDXnFgQjVFC944gq5pBZfNHyGiQqyXdRq4zsUn5OfeU4yYD9YWPngfW7flLYisSdxBmvTeDvFofWL4fCjwGPEK+cSNpHXJoCeBF4EMdtPsouf4DkrYDXig/Y0hEjCCHfn4paX3glVogIAAR8SxZEZxTO89ngf3KYkVnAeNbznMw2YFV1innaysx34Mw65k9JD1Q+/5s4K+S5gMvkUM3M4G9JN1D3oOYA8wFbgNGkOP+3TkBuFTSCeR9iCNLOxuU5UFfJYehdqdr8aJ3iYgxkkaUr+dL+gFwfQn0mwV8i6xyrpB0BzCVvI9S2QG4eZGvhq3QHNZntoRJ+gSwXURcJWld4F/Ax8hO4lZgj7JSXJ+StBOZHnuzpC2AmyJisKT+5GJCu7dWJ7Zy8RCT2ZL3DHCgpHvJK/yTIuLNsqrbmXQtItTXpgAnS7ob+BPwvbJ9JLn6nTuHlZwrCDMza+QKwszMGrmDMDOzRu4gzMyskTsIMzNr5A7CzMwauYMwM7NG/wdKkDvszDr24AAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 461.6x230.4 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot1 = sns.catplot(x=\"Log(Pos/Neg)\", y = \"KEGG_Functional_Category\",hue=\"Category\",kind=\"bar\",palette=['blue','gray','red'],\n",
    "                    data=ratio_df2,\n",
    "                    height=3.2,aspect=1.5)#,order=combine_list)\n",
    "#plot1.savefig('./MGDB_MP_Bacteroides_Log10_KEGGPlot_r03_sprc95_10min_small.pdf')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Specifically plot out the ratios of peptidases and proteases correlated with disease severity"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Add a new column for when the protein name contains peptidase or protease\n",
    "for i in df_dict:\n",
    "    df_dict[i]['Peptidase'] = np.where(df_dict[i]['Protein names'].str.contains('peptidase|protease',case=False) == True, 'Peptidase', 'Not Peptidase')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'All ratio dataframes combined into \"ratio_dict\".'"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Identify the differences in significant vs not significant peptidases\n",
    "TaxaLogRatioCompareSigLevelGroupsCombine(df_dict,sig_level=0.1,Taxa_col='Peptidase',\n",
    "                                         human=\"True\", compareall=\"True\",\n",
    "                                         variable='Peptidase',MinCount=10,Sparsity=0.95)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [],
   "source": [
    "pept_ratios = pd.concat(ratio_dict)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [],
   "source": [
    "##Optional: manually reordered to have data presented in order of manuscript:\n",
    "#pept_ratios.to_csv('./Bacteroides_PeptidaseProteaser03_sprct95_ratios.csv')\n",
    "pept_ratios=pd.read_csv('./Bacteroides_PeptidaseProteaser03_sprct95_ratios.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 386x90 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot1 = sns.catplot(x=\"Log(Pos/Neg)\", y = \"Peptidase\",hue=\"Category\",kind=\"bar\",palette=['blue','gray','red'],\n",
    "                    data=pept_ratios,\n",
    "                    height=1.25,aspect=3)\n",
    "#plot1.savefig('./MGDB_MP_BacteroidesPeptidasesProteases_Log10_r03_sprc95_small.pdf')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Identify Bacteroides MG abundance & composition of each Bacteroides species"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Data\n",
    "df = pd.read_csv('./Metagenomics/Salmon_CPMs.csv',index_col='Name')\n",
    "df2 = pd.read_csv('../UC_Severity/Genomics/Shotgun/Salmon_CPMs.csv', index_col=\"Name\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Patient metadata\n",
    "meta = pd.read_csv('./Combined_Metadata_Allomics_MG_V1.txt', sep = '\\t',index_col = 'id')\n",
    "meta2 = pd.read_csv('../UC_Severity/UC_MP_Emperor_Map.txt',sep ='\\t', index_col = 'id')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Protein metadata\n",
    "MG_ann = pd.read_csv('./Metagenomics/MGDB/MGDB_Full_Protein_Annotations.txt', sep = '\\t')\n",
    "MG_ann2 = pd.read_csv('../UC_Severity/Annotations/UniKB_Taxa_ghostKO_Merge.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Isolate list of UC patients from cohort 1 and sort by activity\n",
    "UC_patients2 = list(meta2.index)\n",
    "UC_patients2 = list(set(UC_patients2) & set(df2.columns))\n",
    "UC_patients2=list(meta2.loc[UC_patients2].sort_values('partial_Mayo').index)\n",
    "\n",
    "#Isolate list of UC patients from cohort 2 and sort by activity\n",
    "UC_patients = list(meta[meta['Diagnosis'] == 'UC'].index)\n",
    "UC_patients = list(set(UC_patients) & set(df.columns))\n",
    "UC_patients=list(meta.loc[UC_patients].sort_values('partial_Mayo').index)\n",
    "\n",
    "#Remove duplicates\n",
    "MG_ann = MG_ann.loc[list(MG_ann['queryID'].drop_duplicates().index)]\n",
    "MG_ann2 = MG_ann2.loc[list(MG_ann2['queryID'].drop_duplicates().index)]\n",
    "\n",
    "#Merge data with annotation information\n",
    "df_merge = df.merge(MG_ann, left_index = True, right_on = 'queryID', how = 'left')\n",
    "df_merge2 = df2.merge(MG_ann2, left_index = True, right_on = 'queryID', how = 'left')\n",
    "\n",
    "#List of sample names\n",
    "samples = list(df.columns)\n",
    "samples2 = list(df2.columns)\n",
    "\n",
    "#Create dataframes for genus level sums of genes\n",
    "df_g = df_merge[samples].groupby(df_merge['Genus']).agg('sum')\n",
    "df_g2 = df_merge2[samples2].groupby(df_merge2['Genus']).agg('sum')\n",
    "\n",
    "\n",
    "#Pull out the reads for each Bacteroides species\n",
    "df_s = df_merge[df_merge['Genus'] == ' g__Bacteroides']\n",
    "df_s2 = df_merge2[df_merge2['Genus'] == ' g__Bacteroides']\n",
    "#Aggregate the species and overwrite the dataframe\n",
    "df_s = df_s[samples].groupby(df_s['Species']).agg('sum')\n",
    "df_s2 = df_s2[samples2].groupby(df_s2['Species']).agg('sum')\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "#Combine both datasets\n",
    "dfs_merge = df_s.merge(df_s2, left_index=True,right_index=True, how='outer')\n",
    "dfs_merge.replace(np.nan,0,inplace=True)\n",
    "#Change index\n",
    "dfs_merge['Species'] = dfs_merge.index"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Combine both dataset metadata files\n",
    "combined_meta = pd.concat([meta,meta2],join='inner')\n",
    "lists = [UC_patients2,UC_patients]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Create a 3 tiered activity level indication\n",
    "Activity_cat=[]\n",
    "for i in combined_meta['partial_Mayo']:\n",
    "    try:\n",
    "        a = float(i)\n",
    "        if a < 3:\n",
    "            Activity_cat.append('Low')\n",
    "        elif a <6:\n",
    "            Activity_cat.append('Moderate')\n",
    "        elif a <11:\n",
    "            Activity_cat.append('High')\n",
    "        else:\n",
    "            Activity_cat.append('not applicable')\n",
    "    except ValueError:\n",
    "        Activity_cat.append('not applicable')\n",
    "    \n",
    "combined_meta['Activity_Level'] = Activity_cat"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 64x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Plot stacked barplots\n",
    "PlotByCategory_MultiSampleGroups(df=dfs_merge,lists=lists,Taxa_col='Species',category='TAXON_ID',\n",
    "                                 metadata=combined_meta,display_number=10, size=4,\n",
    "                                 save_descrip='Bothdatasets_MGBacteroidesSpecies_comp_UC_OverallAvg_top10_final',\n",
    "                                 human=\"True\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 144x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "PlotByCategory_MultiSampleGroups(df=dfs_merge,lists=lists,Taxa_col='Species',category='Activity_Level',\n",
    "                                 metadata=combined_meta,display_number=10,size=9,\n",
    "                                 save_descrip='Bothdatasets_MGBacteroidesSpecies_comp_UC_ActivityLevel_Averaged_top10_final',\n",
    "                                 human=\"True\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Plot the composition of Bacteroides in overall MG reads"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Identify compositions\n",
    "df_g4 = df_g2\n",
    "for j in df_g4:\n",
    "    df_g4[j] = df_g4[j]/df_g4[j].sum()\n",
    "\n",
    "df_g3 = df_g\n",
    "for j in df_g3:\n",
    "    df_g3[j] = df_g3[j]/df_g3[j].sum()\n",
    "\n",
    "#Reset indexes\n",
    "df_g4=df_g4.T.reset_index()\n",
    "df_g3=df_g3.T.reset_index()\n",
    "\n",
    "#Change to longform dataframes\n",
    "df_melt = df_g4.melt(id_vars='index',value_vars=' g__Bacteroides')\n",
    "df_melt2 = df_g3.melt(id_vars='index',value_vars=' g__Bacteroides')\n",
    "\n",
    "#Change columns names\n",
    "df_melt.rename(columns={'index':'Patient ID', 'value':'% MG Composition'}, inplace = True)\n",
    "df_melt2.rename(columns={'index':'Patient ID', 'value':'% MG Composition'}, inplace = True)\n",
    "\n",
    "#Subset to only the UC patients in the IBD200 data\n",
    "df_melt2=df_melt2[df_melt2['Patient ID'].isin(UC_patients)]\n",
    "\n",
    "df_melt['Diagnosis'] = 'UC'\n",
    "df_melt2['Diagnosis'] = 'UC'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 64x36 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Make swarm plots to summarize the composition data\n",
    "fig_width = 2*(4/9)\n",
    "fig, ax = plt.subplots(figsize = (fig_width,0.5))\n",
    "sns.violinplot(x='Diagnosis',y='% MG Composition',data=df_melt,palette=['blue']).set(ylim=(0, 1))\n",
    "plt.savefig('./UC40_BacteroidesMGPrcntages_Swarmplot_final.pdf', bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAFkAAABkCAYAAADt75zcAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy8QZhcZAAAMoklEQVR4nO2de3BU1R3HP7vLwmZDQiThkSBoef14xHaUQRQHamkobY2W2lbwMe04CjpDax2gWhmrjmOtdbRVS50qImokPAZbSRysCVSbEHxUfKBifggRgpRHEtQSY0Ky2f5xbuIKZPdukntZlv3M3Nm9d+/d+93fnv2dc37nd856wuEwKZzFe7IFnA6kjOwCKSO7QMrILpAysgukjOwCKSO7QMrILtAn1gkiMhNYCPTrOKaqM5wUlWzENDLwZ+BmYK/DWpIWO0auVdWNjitJYjyxYhci8hTQDLwNhAFU9XHHlSURdkryx9bjUCeFJDMxSzKAiFwCTARUVdc7rirJsOMu/gCMATYD04EaVV3sgrakwY67mK6qFwGIyMPAa85KSj7sdEb8ItJxnger8kthHzsleQ1QJSKvAVOs/RRxYLfiywfGAdWq+r7jqpKMLo0sIter6hNWxfe1k1R1iRvikoVo7qKjG119zPGUT46XcDgcdRs7duzSY/afiXVNavv61mVJFpEFwO3AQBG53DrsAba78eUnE3Y6I0tU9V6X9CQl0Sq+QlV9QURu4PiKLxUgioNoFV+29ZgKDPUQO+7CA2QC7cBs4AVV/dQFbUmDnR7f00AZMBXTDb8c+LGTopINO7GLs1X1WWC8qt6IKdUp4sCOkfuKyBXAdhHJ4StfncImdtzF/cAcYBFwE6btnCIO4gkQjQc+UtV3HFeVZMR0FyJyE7AMuAh4XERsjYqIyBQReaVn8pIDOz75SmCaqt6MMfScWBeIyC3AE0CgZ/KSAztG9qhqG4CqtgKtNq7ZhWnqpcBexbdZRNYBlcA0oCrWBar6nIicbVfEW2+9FfZ4PHZPT0ja29vrJ02aNOhEr8U0sqoutlICxgFPquqG3haYlpbG+PHje/ttXWXr1q17unrNTsU3GPgeMBOYISJn9KK20wI7PnkN8CFwK1ADFDmqKAmx45NR1b9ZT9+1en92rtkNXNBNXb1KcXEx1dXV3H333Sfl/naMXC0iVwMvA5OABhEZC6CqO5wU11usXr0agCNHjpCRkeH6/e0YeZy1XR9x7DFMID/hk8EbGxs7n1dXVzN58mTXNdhpXXzHDSFO8frrr3c+r6qqSkwji8g9wHVEDEGpap6TonqT8vJyAoEA6enpVFVVMX/+fILBoKsa7LiLQkxMucVpMb3Nvn372L59O3l5eWRkZNDQ0EBlZSWzZs1yVYedJtzbnKIxiNLSUjweDzk5OaSnpxMMBikpKaG9vd1VHXaM/D6wX0RqRORjEalxWlRvcPDgQcrKysjOzsbv9+PxeBg8eDB79+5l8+bNrmqxY+Q5wDcw8eRx1mPCs2rVKtrb28nNzaWhoYGGhgYGDhxIMBikqKiIUCjkmhY7PnkP8MWp5JObmpqoqKggOzubvn37Ul9fD0B2djZDhw6lpqaGbdu2ce6557qix46RhwO7ItxEWFWnOqipx+zevZu2tjYGDBhw3GuZmWYceMeOHQll5JhB+kTj0KFDAPTt2/e413w+H36/n7q6Otf02PHJIeABYAPwECbpMGFpaWlhw4YN+P1+AoETN4rS09PZsmULBw4ccEWTHSMvw0TeLsIkuix3VFEP2LZtG4sWLaK6upozzzyTrgYC8vLyaG5uZuHChZSVlTleCdpxFwFVLbGePy8iC50UFC+tra28+uqrlJaWoqr069eP0aNHn9Afd5CWloaIsGfPHpYuXcqaNWsoLCykoKDAkQCSHSP3EZFzVPU9ETmHBMm0r6urY8OGDZSVlXHkyBECgQDDhw8nJycHr/erH2g4HKa1tZVQKERdXR05OTl4PB4CgQBjx47l888/5+DBg6xYsYKioiKmTZtGYWEhY8aM6TWtdoz8K+BJEckF/gvM77W7d4MDBw5QXFxMRUUF4XCYzMxMRo8eTWZm5gndQ319PS0tpvVZW1sLwKBBZijO4/GQlZVFVlYWTU1N1NfXU1lZycsvv8yECRO45ppryM/P77HmqEYWEZ+VzDJZRDKALztGrt2mra2NdevWsXbtWtrb28nJyWHIkCEnbEFE8tlnnx2332HkSILBICNGjGDYsGHU19ezc+dOlixZwvTp05k/f35n0687dFnxWVlDGjGm911rf0K379ZNdu3axeLFiykuLiYjI4OJEycyfPjwmAYGjotTxIpb+Hw+hgwZwsSJE8nNzaWyspIFCxawefNmursaZLSS/DAwtyMXWVWfF5FDwCNAQbfuFicNDQ2sXLmSTZs20adPH0aOHMkZZ8Q3jhsKhQgEAsycOZPy8nLbLQmv10teXh5ZWVnU1tZy//33k5+fz7XXXhu3v45mZK+qvhl5QFW3iEjs4tNDwuEwJSUlFBUV0dbWxqBBg8jNzaVPH1tDkl8jFAoxa9Ys5s2bB8BLL70U1/XBYBARoa6uDlVl0aJFFBQUcMMNN9CvX7/Yb0B0I/u6OO6P9abWXOxHgW8BLcD1qrrTliLMaMby5cvJzMxkxIgRtj/MifD5fJSXlwMmgO/zdfWxuqYjgpednc3+/fvZuHEjaWlpnV9cLKJ1Rl4UkQdEZACAiPQXkQeAf9l439mY9vWFwG+BB22psegwxNGjRzl8+DBNTU3d9od+v5/m5mZKS0tpbm7G749ZRk5IS0tLpxYgrl9VtNlPHuAWTJMtDfgU0+N7QFWj1h4i8ifgDVVdbe3vU9VhXZ2/devWOky0r5OmpqasxsbGoW1tbekdx3w+H16vt3NzMrUrFArR3t7euXXYyev1Hg0Gg4cyMjIOeTyeSOOd1VWalq385HgRkSeA51T1RWu/Fhh5spp/JxunFt/7HxDZP/WergYG54xcBfwQQEQuAN5z6D6nBLa8t4jkWeeGVdXOInz/AGaKyBZMaPTa7ks89YlW8U0AlqrqDBGpxlR8w4B5qhpfY/M0J5q7+COmdQGw32qOzcA0yVLEQTQjByN6fJ8DWB2K7jU0T2Oi+eS0jieqOjviuJ05IycFEbkYuFFV50Ycuw+z+swmTKdoMOazbQVuVtWjTuuKVpL3icj5kQesfXcGxnoXH7AeeFBVL1bVKZjC4krCcrSSfCuwXkQ2ATuBkZhw56VuCOtlzgT2qurrEcduxaVFuru8iarWAOcDW4B04E1gqqrWuiGsl9mNmYrRiao2q2qTGzePFrSfDkzGuIdXgU+ASdbxROVLIlYst+gPnIVJ0ulERLJFpNANUdHcxSuYSY//sfY7ojFhoMJBTT3hQ+BcEclV1f0iEsAs4vpT4Ecicr6qvmEFv+7CfCkvOC0qWmfkPOAq4DxMeHOlqn58wpMTCGvlr9uBJqAvsExVl4nISGApxvWlYxZ2XehG68Lu8jgzMAYfCpSo6mNOC0smYtauqhrGVH4brfOvj35FimOJ5i78wA8wJXgsUAIUnyrTyhKJaEb+FJPMsgpT+UVOzClzRV2SEK11sR5j2FHW1kEYs7pWCps4MvyU4uvEn8jgMFaQZy1m4VUPJur3ELADuExVXZ0gLSLfB0b0ZOnMhCvJx0bSRKQ/8G/gulN1kSnbJVlERgPpqvqug3qOQ1UbReQxYKmIfKKqc0Xkl5jld/yYWPflmEjbM0AeZoHt6aqaZy029Q6Qj1k48GequkdEFgFzgTagQlVvFZGLMOHQVsxI0NXATzCzvu7C/MIGYEKlt6jqK3Y+g60olLWC1q+B60RkrZ1repmDQI6lxYtZALBAVadhDD0Zkx/ysfV3HXcBQyKuf0NVC4By4Eorz/oKzNKYU4ExVhxjNvB34NvAk0Bk4t0oTGfsUkyz1vbc4WgBot9E5L2NBu4E7sAEW9zmLOBZACux5iiwSkSWY8KYfsz8wi3WOdVA5Mybt63HvZjZteOA11S11epsVWL+EeheTFB/Eybe0TlAoaofAH/FNGkfJY4wabQTtwDPishcTCbnUmAF8Du7b94bWHnR87CMJiLfBGar6hxMgroXU0G+D1xonTMKq+RbHFvxVANTRKSPFTaYjqlYrwaeslZG+ICIhHer9Geo6iXAL4C/2P0MXfpkVa3C/L/IVZhS/Ih1zA1mWL40hNF4J3AY+A5mAOELEXkTk8y4H+OHlwNPiUgFJuWruas3t6ZmrMXkh3gxf7n0PCZ+/rSINGJ+LfMxrgPgI+BOEfm59doddj9MtB5fPiZO0YiZAXUl5j+gfm8F9BMKEZkK9FfVMhEZA/xTVUfFus4NormLxzDu4UXgHlW9D7gNUwEmIjXAbSJSBawEFpxkPZ1EK8mbMF3r/sAQVU1U4yY80drJl2HWg2vEhDlTdJOE6/ElI6n/rXaBlJFdIGVkF0gZ2QX+D1L5ZhgxDaBBAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 64x36 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Make swarm plots to summarize the composition data\n",
    "fig_width = 2*(4/9)\n",
    "fig, ax = plt.subplots(figsize = (fig_width,0.5))\n",
    "sns.violinplot(x='Diagnosis',y='% MG Composition',data=df_melt2,palette=['gray']).set(ylim=(0, 1))\n",
    "plt.savefig('./IBD200UCpatientsBacteroidesMGPrcntages_final.pdf', bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "#Reperform the compositions\n",
    "df_g3 = df_g\n",
    "for j in df_g3:\n",
    "    df_g3[j] = df_g3[j]/df_g3[j].sum()\n",
    "\n",
    "df_g4 = df_g2\n",
    "for j in df_g4:\n",
    "    df_g4[j] = df_g4[j]/df_g4[j].sum()\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 144x36 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Use the dotplotcategory function to create plots\n",
    "DotPlotCategory(df=df_g3,samples=UC_patients,Taxa_col=' g__Bacteroides',\n",
    "                category='Activity_Level',metadata=combined_meta,\n",
    "                save_descrip='IBD200_DotPlot_Activity_final',width=9)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 144x36 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Use the dotplotcategory function to create plots\n",
    "DotPlotCategory(df=df_g4,samples=UC_patients2,Taxa_col=' g__Bacteroides',\n",
    "                category='Activity_Level',metadata=combined_meta,\n",
    "                save_descrip='UC40_DotPlot_Activity_final',width=9)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Compare correlations to activity of Bacteroides proteases by species"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Load in the correlation data\n",
    "df_n40 = pd.read_csv('./pMayo_correlations_MP_MGDB_annotated_UC40.csv', index_col=0)\n",
    "df_n3 = pd.read_csv('./pMayo_correlations_MP_MGDB_annotated.csv', index_col = 0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Subset for Bacteroides\n",
    "df_n40=df_n40[df_n40['Genus']=='Bacteroides']\n",
    "\n",
    "#Pull out the enzymes and proteases, then concatenate them to one dataframe\n",
    "df_n40_enz = df_n40[df_n40['KEGG_Functional_Category'] == 'Enzyme Families'] \n",
    "df_n40_pept = df_n40[df_n40['Protein names'].str.contains('peptidase|protease',case=False)==True]\n",
    "df_n40_all = pd.concat([df_n40_pept,df_n40_enz])\n",
    "\n",
    "df_n3=df_n3[df_n3['Genus']=='Bacteroides']\n",
    "df_n3_enz = df_n3[df_n3['KEGG_Functional_Category'] == 'Enzyme Families'] \n",
    "df_n3_pept = df_n3[df_n3['Protein names'].str.contains('peptidase|protease',case=False)==True]\n",
    "df_n3_all = pd.concat([df_n3_pept,df_n3_enz])\n",
    "\n",
    "#Compile the correlations by both the species and the cohort, start by making new column\n",
    "df_n40_all['Species_cohort']=df_n40_all['Species']+\" (UC Cohort 1)\"\n",
    "df_n3_all['Species_cohort']=df_n3_all['Species']+\" (UC Cohort 2)\"\n",
    "\n",
    "#Identify the columns of interest\n",
    "Columns = ['Species_cohort','rvalue','Protein names']\n",
    "#Subset the columns of interest\n",
    "UC1=df_n40_all[Columns]\n",
    "UC2=df_n3_all[Columns]\n",
    "#Concatenate the two datasets\n",
    "UC_cat=pd.concat([UC1,UC2])\n",
    "\n",
    "#Aggregate by the sum of the correlations within each protein of each species\n",
    "UC_cat['sumbby']=UC_cat['Species_cohort']+UC_cat['Protein names']\n",
    "groupcols=['Protein names','Species_cohort']\n",
    "UC_cat2=UC_cat.groupby(groupcols)['rvalue'].sum().reset_index()\n",
    "\n",
    "#Pivot the table for visualization\n",
    "Bprot_rtable=UC_cat2.pivot(index='Protein names',columns='Species_cohort',values='rvalue')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Determine bacteria order by species abundance\n",
    "Abund_bac=[\"Bacteroides vulgatus (UC Cohort 1)\",\"Bacteroides vulgatus (UC Cohort 2)\",\"Bacteroides dorei (UC Cohort 1)\",\"Bacteroides dorei (UC Cohort 2)\",\n",
    "\"Bacteroides uniformis (UC Cohort 1)\",\"Bacteroides uniformis (UC Cohort 2)\",\n",
    "\"Bacteroides ovatus (UC Cohort 1)\",\"Bacteroides ovatus (UC Cohort 2)\",\n",
    "\"Bacteroides fragilis (UC Cohort 1)\",\"Bacteroides fragilis (UC Cohort 2)\",\n",
    "\"Bacteroides thetaiotaomicron (UC Cohort 1)\",\"Bacteroides thetaiotaomicron (UC Cohort 2)\"]\n",
    "\n",
    "abund_df=Bprot_rtable[Abund_bac]\n",
    "abund_df=abund_df.dropna(axis=0,how='all')\n",
    "abund_df['Number_missing']=abund_df.isnull().sum(axis=1)\n",
    "abund_df=abund_df[abund_df['Number_missing']<11].drop(columns='Number_missing')\n",
    "abund_df=abund_df.drop(index='Deleted.')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Determine protein order by molecular function\n",
    "proteins=abund_df.index\n",
    "dfall=pd.concat([df_n3_all,df_n40_all])\n",
    "cols=['Gene ontology (molecular function)','Protein names']\n",
    "prot_funct_df=dfall[cols].drop_duplicates()\n",
    "prot_funct_df=prot_funct_df[prot_funct_df['Protein names'].isin(proteins)]\n",
    "prot_order=list(prot_funct_df.sort_values(by='Gene ontology (molecular function)')['Protein names'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Manually sorted by the class type activities of each protease\n",
    "#prot_funct_df.sort_values(by='Gene ontology (molecular function)').to_csv('./abundbacteroidesproteases_molfunct.csv')\n",
    "\n",
    "prot_order=[\"Dipeptidase (EC 3.4.-.-)\",\n",
    "\"Dipeptidyl aminopeptidase IV\",\n",
    "\"Dipeptidyl peptidase IV\",\n",
    "\"Dipeptidyl peptidase IV N-terminal domain protein (Fragment)\",\n",
    "\"Dipeptidyl-peptidase 7 (DPP7) (EC 3.4.14.-)\",\n",
    "\"Dipeptidyl-peptidase III (EC 3.4.14.4)\",\n",
    "\"Peptidase S9A/B/C family catalytic domain protein\",\n",
    "\"Peptidase, S9A/B/C family, catalytic domain protein (EC 3.4.-.-)\",\n",
    "\"Prolyl oligopeptidase family protein\",\n",
    "\"Prolyl tripeptidyl peptidase (EC 3.4.14.12)\",\n",
    "\"Protease Do\",\n",
    "\"Putative Tricorn-like protease\",\n",
    "\"Serine protease\",\n",
    "\"Signal peptidase I (EC 3.4.21.89)\",\n",
    "\"Signal peptide peptidase SppA, 67K type (EC 3.4.-.-)\",\n",
    "\"Tricorn protease\",\n",
    "\"Tricorn protease homolog (EC 3.4.21.-)\",\n",
    "\"Uncharacterized protein\",\n",
    "\"Aminopeptidase P domain protein (EC 3.4.-.-)\",\n",
    "\"Endothelin-converting enzyme 1 (EC 3.4.24.71)\",\n",
    "\"Leucine aminopeptidase\",\n",
    "\"Peptidase M16 inactive domain protein\",\n",
    "\"Peptidase M16 inactive domain protein (EC 3.4.24.-)\",\n",
    "\"Peptidase, M24 family (EC 3.4.-.-)\",\n",
    "\"Peptidase, M28 family (EC 3.4.-.-)\",\n",
    "\"Putative endothelin-converting enzyme\",\n",
    "\"Putative peptidyl-dipeptidase dcp\",\n",
    "\"TldD/PmbA family protein\",\n",
    "\"Xaa-His dipeptidase (EC 3.4.13.3)\",\n",
    "\"Xaa-Pro aminopeptidase (EC 3.4.11.9)\",\n",
    "\"Aminopeptidase C\",\n",
    "\"Aminopeptidase C (Bleomycin hydrolase)\",\n",
    "\"Peptidase C1-like family protein (Fragment)\",\n",
    "\"Peptidase C1A papain\",\n",
    "\"Aminoacyl-histidine dipeptidase (Cytosol non-specific dipeptidase) (EC 3.4.13.18)\",\n",
    "\"Carboxypeptidase regulatory-like domain protein\",\n",
    "\"Carboxypeptidase regulatory-like domain protein (TonB-dependent Receptor Plug Domain protein)\",\n",
    "\"Cytosol non-specific dipeptidase (EC 3.4.13.18)\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 230.4x576 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Reorder the data\n",
    "abund_df=abund_df.loc[prot_order]\n",
    "\n",
    "#Create a heatmap of the data\n",
    "ax=plt.subplots(figsize=(3.2,8))\n",
    "ax=sns.heatmap(abund_df,cmap='coolwarm',vmax=1,vmin=-1,cbar_kws = dict(use_gridspec=False,location=\"left\"))\n",
    "ax.tick_params(left=False,bottom=False,right=True, top=True, \n",
    "               labelright=True, labeltop=True,labelbottom=False,\n",
    "               labelleft=False,rotation=0)\n",
    "plt.xticks(rotation='vertical')\n",
    "bottom, top = ax.get_ylim()\n",
    "ax.set_ylim(bottom + 0.5, top - 0.5)\n",
    "plt.savefig('Combined_UC_Bacteroides_ProteasesRvaluesSum_HeatmapReds_final.pdf',bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Identify patient samples overproducing Bacteroides vulgatus proteases"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 163,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Proteases from Bv or Bd overlap\n",
    "df1 = pd.read_csv('./pMayo_correlations_MP_MGDB_annotated.csv', index_col = 'Master Protein')\n",
    "df2 = pd.read_csv('./pMayo_correlations_MP_MGDB_annotated_UC40.csv', index_col=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 167,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Find the positive correlations\n",
    "df2 = df2[df2['rvalue'] > 0.3]\n",
    "df1 = df1[df1['rvalue'] > 0.3]\n",
    "\n",
    "#Pull out \n",
    "species_interest = ['Bacteroides dorei', 'Bacteroides vulgatus']\n",
    "df2 = df2[df2['Species'].isin(species_interest)]\n",
    "df1 = df1[df1['Species'].isin(species_interest)]\n",
    "\n",
    "df1_enz = df1[df1['KEGG_Functional_Category'] == 'Enzyme Families'] \n",
    "df1_pept = df1[df1['Protein names'].str.contains('peptidase|protease',case=False)==True]\n",
    "df1_all = pd.concat([df1_pept,df1_enz])\n",
    "#IBD200 = set(df1_all['Protein names'])\n",
    "\n",
    "df2_enz = df2[df2['KEGG_Functional_Category'] == 'Enzyme Families'] \n",
    "df2_pept = df2[df2['Protein names'].str.contains('peptidase|protease',case=False)==True]\n",
    "df2_all = pd.concat([df2_pept,df2_enz])\n",
    "#UC40 = set(df2_all['Protein names'])\n",
    "\n",
    "df1_all.index = df1_all.index + 'IBD200'\n",
    "df2_all.index = df2_all.index + 'UC40'\n",
    "\n",
    "BvBdProteasesMPdf=pd.concat([df1_all,df2_all])\n",
    "#Instead only going to match by protease names\n",
    "#UC40 = set(df2_pept['Protein names'])\n",
    "#IBD200 = set(df1_pept['Protein names'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 187,
   "metadata": {},
   "outputs": [],
   "source": [
    "## Bring in metagenomics data\n",
    "\n",
    "#Data\n",
    "#df = pd.read_csv('./Metagenomics/Salmon_CPMs.csv',index_col='Name')\n",
    "#df2 = pd.read_csv('../UC_Severity/Genomics/Shotgun/Salmon_CPMs.csv', index_col=\"Name\")\n",
    "\n",
    "#Patient metadata\n",
    "meta = pd.read_csv('./Combined_Metadata_Allomics_MG_V1.txt', sep = '\\t',index_col = 'id')\n",
    "meta2 = pd.read_csv('../UC_Severity/UC_MP_Emperor_Map.txt',sep ='\\t', index_col = 'id')\n",
    "meta2['Proteome_ID'] = meta2.index\n",
    "\n",
    "#Protein metadata\n",
    "#MG_ann = pd.read_csv('./Metagenomics/MGDB/MGDB_Full_Protein_Annotations.txt', sep = '\\t')\n",
    "#MG_ann2 = pd.read_csv('../UC_Severity/Annotations/UniKB_Taxa_ghostKO_Merge.csv')\n",
    "\n",
    "#Isolate list of UC patients from cohort 1 and sort by activity\n",
    "UC_patients2 = list(meta2.index)\n",
    "UC_patients2 = list(set(UC_patients2) & set(df2.columns))\n",
    "UC_patients2=list(meta2.loc[UC_patients2].sort_values('partial_Mayo').index)\n",
    "\n",
    "#Isolate list of UC patients from cohort 2 and sort by activity\n",
    "UC_patients = list(meta[meta['Diagnosis'] == 'UC'].index)\n",
    "UC_patients = list(set(UC_patients) & set(df.columns))\n",
    "UC_patients=list(meta.loc[UC_patients].sort_values('partial_Mayo').index)\n",
    "\n",
    "#Remove duplicates\n",
    "MG_ann = MG_ann.loc[list(MG_ann['queryID'].drop_duplicates().index)]\n",
    "MG_ann2 = MG_ann2.loc[list(MG_ann2['queryID'].drop_duplicates().index)]\n",
    "\n",
    "#Merge data with annotation information\n",
    "df_merge = df.merge(MG_ann, left_index = True, right_on = 'queryID', how = 'left')\n",
    "df_merge2 = df2.merge(MG_ann2, left_index = True, right_on = 'queryID', how = 'left')\n",
    "\n",
    "#List of sample names\n",
    "samples = list(df.columns)\n",
    "samples2 = list(df2.columns)\n",
    "\n",
    "#Pull out the reads for each Bacteroides species\n",
    "df_s = df_merge[df_merge['Genus'] == ' g__Bacteroides']\n",
    "df_s2 = df_merge2[df_merge2['Genus'] == ' g__Bacteroides']\n",
    "#Aggregate the species and overwrite the dataframe\n",
    "df_s = df_s[samples].groupby(df_s['Species']).agg('sum')\n",
    "df_s2 = df_s2[samples2].groupby(df_s2['Species']).agg('sum')\n",
    "\n",
    "\n",
    "#Combine both datasets\n",
    "dfs_merge = df_s.merge(df_s2, left_index=True,right_index=True, how='outer')\n",
    "dfs_merge.replace(np.nan,0,inplace=True)\n",
    "#Change index\n",
    "dfs_merge['Species'] = dfs_merge.index\n",
    "\n",
    "#Combine both dataset metadata files\n",
    "combined_meta = pd.concat([meta,meta2],join='inner')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 198,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Subset out species of interest from above\n",
    "BvBd = [' s__Bacteroides vulgatus',' s__Bacteroides dorei']\n",
    "BvMGdf = dfs_merge[dfs_merge.index.isin(BvBd)]\n",
    "\n",
    "#Sum the MG abundances\n",
    "BvMGdf.loc['sum'] = BvMGdf.sum(axis=0)\n",
    "\n",
    "#Identify sum of Bacteroides species\n",
    "samples=set(BvMGdf.columns)&set(combined_meta.index)\n",
    "\n",
    "#Subset out to only listed samples\n",
    "BvMGdf2=BvMGdf[samples]\n",
    "combined_meta2 = combined_meta.loc[samples]\n",
    "#Change sample names to be compatible with proteome data\n",
    "BvMGdf2.columns = combined_meta2['Proteome_ID']\n",
    "#Drop unused samples\n",
    "BvMGdf2.drop(columns='not_analyzed',inplace=True)\n",
    "#Rename\n",
    "BvMGdf2.rename({'sum':'BvandBdMGSum'},axis='index',inplace=True)\n",
    "#Save and reload for column names error\n",
    "BvMGdf2.to_csv('./BvBdMGSums2.csv')\n",
    "#Couldn't figure how to get axis labels proper but this worked to save and load\n",
    "BvMGdf2 = pd.read_csv('./BvBdMGSums2.csv',index_col='Species')\n",
    "BvBdProteasesMPdf.loc['BvBdMGSum'] = BvMGdf2.loc['BvandBdMGSum']\n",
    "\n",
    "#Remove unnecessary data in metadata\n",
    "combined_meta2.index = combined_meta2['Proteome_ID']\n",
    "combined_meta2=combined_meta2[combined_meta2.index!='not_analyzed']\n",
    "combined_meta3=combined_meta2\n",
    "combined_meta3=combined_meta3[combined_meta3['Proteome_ID']!='not_analyzed']\n",
    "combined_meta3.index = combined_meta3['Proteome_ID']\n",
    "combined_meta4 = combined_meta3[combined_meta3['partial_Mayo']!='Missing']\n",
    "combined_meta4 = combined_meta4[combined_meta4['partial_Mayo']!='Healthy_control']\n",
    "UC_patients = list(combined_meta4.index)\n",
    "\n",
    "#Subset out dataframes to overlapping patients\n",
    "df = BvBdProteasesMPdf[set(UC_patients)&set(BvBdProteasesMPdf.columns)]\n",
    "combined_meta4 = combined_meta4.loc[combined_meta4.index.isin(set(UC_patients)&set(BvBdProteasesMPdf.columns))]\n",
    "combined_meta4=combined_meta4.drop_duplicates()\n",
    "df.loc['Relative_Disease_Activity'] = combined_meta4['Relative_Activity_Index']\n",
    "\n",
    "dfT=df.T\n",
    "bvproteases=list(dfT.columns)\n",
    "bvproteases.remove('BvBdMGSum')\n",
    "bvproteases.remove('Relative_Disease_Activity')\n",
    "\n",
    "dfT['MPsum']=dfT[bvproteases].sum(axis=1)\n",
    "dfT['MPaverage']=dfT[bvproteases].mean(axis=1)\n",
    "dfT['Relative_Disease_Activity']=dfT['Relative_Disease_Activity'].astype('float')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 213,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "dfT['BvBdMGSum']=dfT['BvBdMGSum'].astype('float')\n",
    "dfT['MPsum']=dfT['MPsum'].astype('float')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 214,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<seaborn.axisgrid.FacetGrid at 0x4a806a5f8>"
      ]
     },
     "execution_count": 214,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.lmplot(x='BvBdMGSum', y='MPsum', data=dfT)\n",
    "#plot1.legend(loc='center left', bbox_to_anchor=(1, 0.5))\n",
    "#plt.tight_layout()\n",
    "#plt.savefig('BVMGabund_BVProteaseSum_lmplot.pdf',bbox_inches=\"tight\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#dfT.to_csv('./MGvsProteaseAbundances_.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Identify strength of associations between disease activity, endoscopic, and histology and Bacteroides protease production\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": [
    "#List of patient samples by BV protease abundance category:\n",
    "High_overpros = ['H19','L18','H9','59','H5','77', 'H2','H7']\n",
    "Med_overpros = ['57','H4','193','25']\n",
    "Low_overpros = ['H10','137','12','H11']\n",
    "Low_underpros = ['129','8','154']\n",
    "Med_underpros = ['102','60','138','L7','179','60','L10','132']\n",
    "High_underpros = ['200','L11','L19','L1','H15','L4','H20']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 234,
   "metadata": {},
   "outputs": [],
   "source": [
    "meta = pd.read_csv('./Combined_metadata_metaproteome_metabolome_v1.txt', sep = '\\t', index_col='Proteome_ID')\n",
    "meta2 = pd.read_csv('../UC_Severity/UC_MP_Emperor_Map.txt',sep ='\\t', index_col = 'id')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 239,
   "metadata": {},
   "outputs": [],
   "source": [
    "meta2.rename(columns={'UCEIS_endoscopic_score':'UCEIS'},inplace=True)\n",
    "\n",
    "meta['Experiment'] = 'IBD200'\n",
    "meta2['Experiment'] = 'UC40'\n",
    "\n",
    "cols = ['Experiment','partial_Mayo','UCEIS']\n",
    "\n",
    "meta = meta[cols]\n",
    "meta2 = meta2[cols]\n",
    "\n",
    "meta3 = pd.concat([meta,meta2])\n",
    "meta3 = meta3[meta3['partial_Mayo'] != 'Missing'] \n",
    "meta3 = meta3[meta3['partial_Mayo'] != 'Follow_up_with_MD']\n",
    "meta3 = meta3[meta3['partial_Mayo'] != 'Healthy_control']\n",
    "\n",
    "meta3['partial_Mayo'] = meta3['partial_Mayo'].astype('float')\n",
    "\n",
    "palette = {'IBD200':'gray','UC40':'blue'}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 242,
   "metadata": {},
   "outputs": [],
   "source": [
    "Protease_cat = []\n",
    "for i in meta3.index:\n",
    "    if i in High_overpros:\n",
    "        Protease_cat.append('Overproducer')\n",
    "    elif i in Med_overpros:\n",
    "        Protease_cat.append('Overproducer')\n",
    "    elif i in Low_overpros:\n",
    "        Protease_cat.append('Overproducer')    \n",
    "    elif i in Med_underpros:\n",
    "        Protease_cat.append('Underproducer')\n",
    "    elif i in Low_underpros:\n",
    "        Protease_cat.append('Others')\n",
    "    elif i in High_underpros:\n",
    "        Protease_cat.append('Underproducer')    \n",
    "    else:\n",
    "        Protease_cat.append('Others')    \n",
    "meta3['Bv Protease Producer Category'] = Protease_cat"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 243,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[Text(0, 0, 'Others'), Text(0, 0, 'Underproducer'), Text(0, 0, 'Overproducer')]"
      ]
     },
     "execution_count": 243,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 144x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.set(rc={'figure.figsize':(2,4)})\n",
    "sns.set_style(\"whitegrid\")\n",
    "plot1 = sns.swarmplot(x='Bv Protease Producer Category', y='partial_Mayo',hue='Experiment', palette=palette, data=meta3, size=7)\n",
    "ax = sns.boxplot(x='Bv Protease Producer Category', y='partial_Mayo',color='white', linewidth='2',data=meta3)\n",
    "# iterate over boxes\n",
    "for i,box in enumerate(ax.artists):\n",
    "    box.set_edgecolor('black')\n",
    "    box.set_facecolor('white')\n",
    "\n",
    "    # iterate over whiskers and median lines\n",
    "    for j in range(6*i,6*(i+1)):\n",
    "         ax.lines[j].set_color('black')\n",
    "\n",
    "plot1.legend(loc='center left', bbox_to_anchor=(1, 0.5))\n",
    "\n",
    "plot1.set_xticklabels(plot1.get_xticklabels(),rotation=90)\n",
    "#plt.savefig('Bv_protease_outliers_analysis_boxplots.pdf', bbox_inches=\"tight\")\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Stats on association to disease activity"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 244,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Ttest_indResult(statistic=3.782479069637194, pvalue=0.0007505816369892496)"
      ]
     },
     "execution_count": 244,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "a = list(meta3[meta3['Bv Protease Producer Category'] == 'Overproducer']['partial_Mayo'])\n",
    "b = list(meta3[meta3['Bv Protease Producer Category'] == 'Underproducer']['partial_Mayo'])\n",
    "\n",
    "scipy.stats.ttest_ind(a=a,b=b)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 245,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Ttest_indResult(statistic=5.036173141634479, pvalue=2.390131497985491e-06)"
      ]
     },
     "execution_count": 245,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "a = list(meta3[meta3['Bv Protease Producer Category'] == 'Overproducer']['partial_Mayo'])\n",
    "b = list(meta3[meta3['Bv Protease Producer Category'] == 'Others']['partial_Mayo'])\n",
    "\n",
    "scipy.stats.ttest_ind(a=a,b=b)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 246,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Ttest_indResult(statistic=-0.25970869519310863, pvalue=0.7956883134787964)"
      ]
     },
     "execution_count": 246,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "a = list(meta3[meta3['Bv Protease Producer Category'] == 'Underproducer']['partial_Mayo'])\n",
    "b = list(meta3[meta3['Bv Protease Producer Category'] == 'Others']['partial_Mayo'])\n",
    "\n",
    "scipy.stats.ttest_ind(a=a,b=b)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Relationship to protease producing categories and endoscopy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 247,
   "metadata": {},
   "outputs": [],
   "source": [
    "meta4 = meta3[meta3['UCEIS']!='Missing']\n",
    "meta4['UCEIS']=meta4['UCEIS'].astype('float')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 248,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[Text(0, 0, 'Others'), Text(0, 0, 'Underproducer'), Text(0, 0, 'Overproducer')]"
      ]
     },
     "execution_count": 248,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 144x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.set(rc={'figure.figsize':(2,4)})\n",
    "sns.set_style(\"whitegrid\")\n",
    "plot1 = sns.swarmplot(x='Bv Protease Producer Category', y='UCEIS',hue='Experiment', palette=palette, data=meta4, size=7)\n",
    "ax = sns.boxplot(x='Bv Protease Producer Category', y='UCEIS',color='white', linewidth='2',data=meta4)\n",
    "# iterate over boxes\n",
    "for i,box in enumerate(ax.artists):\n",
    "    box.set_edgecolor('black')\n",
    "    box.set_facecolor('white')\n",
    "\n",
    "    # iterate over whiskers and median lines\n",
    "    for j in range(6*i,6*(i+1)):\n",
    "         ax.lines[j].set_color('black')\n",
    "\n",
    "plot1.legend(loc='center left', bbox_to_anchor=(1, 0.5))\n",
    "\n",
    "plot1.set_xticklabels(plot1.get_xticklabels(),rotation=90)\n",
    "#plt.savefig('Bv_protease_outliers_analysis_boxplots_UCEIS.pdf', bbox_inches=\"tight\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 249,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Ttest_indResult(statistic=5.291233533028236, pvalue=1.2514208976853215e-05)"
      ]
     },
     "execution_count": 249,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "a = list(meta4[meta4['Bv Protease Producer Category'] == 'Overproducer']['UCEIS'])\n",
    "b = list(meta4[meta4['Bv Protease Producer Category'] == 'Underproducer']['UCEIS'])\n",
    "\n",
    "scipy.stats.ttest_ind(a=a,b=b)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 250,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Ttest_indResult(statistic=4.803777892838683, pvalue=6.625328700792065e-06)"
      ]
     },
     "execution_count": 250,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "a = list(meta4[meta4['Bv Protease Producer Category'] == 'Overproducer']['UCEIS'])\n",
    "b = list(meta4[meta4['Bv Protease Producer Category'] == 'Others']['UCEIS'])\n",
    "\n",
    "scipy.stats.ttest_ind(a=a,b=b)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 251,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Ttest_indResult(statistic=-0.9403747677770924, pvalue=0.34975416602861975)"
      ]
     },
     "execution_count": 251,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "a = list(meta4[meta4['Bv Protease Producer Category'] == 'Underproducer']['UCEIS'])\n",
    "b = list(meta4[meta4['Bv Protease Producer Category'] == 'Others']['UCEIS'])\n",
    "\n",
    "scipy.stats.ttest_ind(a=a,b=b)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Proportions of histological remission in producer categories"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [],
   "source": [
    "meta = pd.read_csv('./Combined_metadata_allomics_hist.txt', sep = '\\t', index_col='Proteome_ID')\n",
    "meta2 = pd.read_csv('../UC_Severity/UC_MP_Emperor_Map_v2_2020.txt',sep ='\\t', index_col = 'id')\n",
    "\n",
    "meta['Experiment'] = 'IBD200'\n",
    "meta2['Experiment'] = 'UC40'\n",
    "\n",
    "cols = ['Experiment','partial_Mayo','Geboes_Grade_3_Numeric']\n",
    "\n",
    "meta = meta[cols]\n",
    "meta2 = meta2[cols]\n",
    "\n",
    "meta3 = pd.concat([meta,meta2])\n",
    "\n",
    "Protease_cat = []\n",
    "for i in meta3.index:\n",
    "    if i in High_overpros:\n",
    "        Protease_cat.append('Overproducer')\n",
    "    elif i in Med_overpros:\n",
    "        Protease_cat.append('Overproducer')\n",
    "    elif i in Low_overpros:\n",
    "        Protease_cat.append('Overproducer')    \n",
    "    elif i in Med_underpros:\n",
    "        Protease_cat.append('Underproducer')\n",
    "    elif i in Low_underpros:\n",
    "        Protease_cat.append('Others')\n",
    "    elif i in High_underpros:\n",
    "        Protease_cat.append('Underproducer')    \n",
    "    else:\n",
    "        Protease_cat.append('Others')\n",
    "\n",
    "meta3['Bv Protease Producer Category'] = Protease_cat\n",
    "\n",
    "meta3 = meta3[meta3['Geboes_Grade_3_Numeric']!='not applicable']\n",
    "meta3['Geboes_Grade_3_Numeric']=meta3['Geboes_Grade_3_Numeric'].astype('float')\n",
    "meta3 = meta3[meta3['Geboes_Grade_3_Numeric'].isna()==False]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Categorize patients as having histological remission\n",
    "\n",
    "hist_rem=[]\n",
    "for i in meta3['Geboes_Grade_3_Numeric']:\n",
    "    if i >0:\n",
    "        hist_rem.append('no')\n",
    "    else:\n",
    "        hist_rem.append('yes')\n",
    "meta3['Histological_Remission'] = hist_rem"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "81.25"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Percentage of patients with overproduction in not in remission:\n",
    "Total=len(meta3[meta3['Bv Protease Producer Category']=='Overproducer'])\n",
    "meta4 = meta3[meta3['Bv Protease Producer Category']=='Overproducer']\n",
    "NotHistRemission = len(meta4[meta4['Geboes_Grade_3_Numeric']>0])\n",
    "\n",
    "100*(NotHistRemission/Total)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "18.75"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "100-81.25"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "61.53846153846154"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Percentage of patients with underproduction in not in remission:\n",
    "Total=len(meta3[meta3['Bv Protease Producer Category']=='Underproducer'])\n",
    "meta4 = meta3[meta3['Bv Protease Producer Category']=='Underproducer']\n",
    "NotHistRemission = len(meta4[meta4['Geboes_Grade_3_Numeric']>0])\n",
    "\n",
    "100*(NotHistRemission/Total)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "38.462"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "100-61.538"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "55.00000000000001"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Percentage of other patients not in remission:\n",
    "Total=len(meta3[meta3['Bv Protease Producer Category']=='Others'])\n",
    "meta4 = meta3[meta3['Bv Protease Producer Category']=='Others']\n",
    "NotHistRemission = len(meta4[meta4['Geboes_Grade_3_Numeric']>0])\n",
    "\n",
    "100*(NotHistRemission/Total)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "45"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "100-55"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "60"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Total"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.6.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}