6.[GO set]Plot_GO_analysis_figs.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Notebook for the figures generated in Workflow1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "import seaborn as sns\n",
    "from glob import glob\n",
    "import re\n",
    "from scipy.stats import wilcoxon, mannwhitneyu\n",
    "\n",
    "# Set default configuration for all plots\n",
    "plt.rcParams['svg.fonttype'] = 'none'\n",
    "plt.rcParams['pdf.fonttype'] = 42\n",
    "plt.rcParams.update({'font.size': 7, 'font.family': 'sans-serif'})\n",
    "plt.rcParams['xtick.labelsize'] = 6\n",
    "plt.rcParams['ytick.labelsize'] = 6\n",
    "plt.rcParams['axes.labelsize'] = 7\n",
    "plt.rcParams['axes.titlesize'] = 7\n",
    "plt.rcParams['axes.linewidth'] = 0.5\n",
    "plt.rcParams['hatch.linewidth'] = 0.5\n",
    "plt.rcParams['xtick.major.width'] = 0.4\n",
    "plt.rcParams['ytick.major.width'] = 0.4\n",
    "plt.rcParams['xtick.minor.width'] = 0.3\n",
    "plt.rcParams['ytick.minor.width'] = 0.3\n",
    "plt.rcParams['legend.frameon'] = False\n",
    "plt.rcParams['xtick.major.size'] = 3\n",
    "plt.rcParams['ytick.major.size'] = 3\n",
    "plt.rcParams['xtick.minor.size'] = 2\n",
    "plt.rcParams['ytick.minor.size'] = 2\n",
    "plt.rcParams['xtick.major.pad'] = 1\n",
    "plt.rcParams['ytick.major.pad'] = 1\n",
    "plt.rcParams['axes.labelpad'] = 1\n",
    "plt.rcParams['patch.linewidth'] = 0.25\n",
    "plt.rcParams[\"axes.unicode_minus\"] = True\n",
    "plt.figsize = (2,2)    \n",
    "\n",
    "%load_ext autoreload\n",
    "%autoreload 2\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "## save the final result data to the supplemntary table\n",
    "import pandas as pd\n",
    "df = pd.read_csv('./data/GO_term_analysis/simrank_pval_LLM_processed_selected_1000_go_terms.tsv', sep='\\t')\n",
    "# check if there is any NA \n",
    "df.isna().sum()\n",
    "\n",
    "df.to_csv('./supplementary_information/Supplementary Table 2', sep='\\t', index=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Extendend data fig 1\n",
    "\n",
    "GO term size distribution "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x300 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "## distribution of original go term sizes\n",
    "import pandas as pd \n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "import numpy as np\n",
    "\n",
    "go_terms = pd.read_csv('data/go_terms.csv', sep = ',',index_col=0)\n",
    "# only show go terms from 3-100\n",
    "filtered_go_terms = go_terms[(go_terms['Gene_Count']>=3) & (go_terms['Gene_Count']<=100)]\n",
    "\n",
    "input_file = 'data/GO_term_analysis/1000_selected_go_terms.csv'\n",
    "thousand_GO = pd.read_csv(input_file, sep = ',')\n",
    "\n",
    "\n",
    "fig, ax = plt.subplots(nrows=1, ncols=2, figsize=(8,3))\n",
    "# Apply log transformation to the 'Gene_Count' column\n",
    "# go_terms['log_Gene_Count'] = np.log2(go_terms['Gene_Count'])\n",
    "ax[0].set_yscale(\"log\")\n",
    "ax[1].set_yscale(\"log\")\n",
    "# Plotting the distribution of the log-transformed data\n",
    "sns.histplot(filtered_go_terms['Gene_Count'], bins=100, edgecolor='black',color='white', linewidth=0.5, rasterized=True, ax=ax[0])\n",
    "sns.histplot(thousand_GO['Gene_Count'], bins=100, edgecolor='black',color='white', linewidth=0.5, rasterized=True, ax=ax[1])\n",
    "\n",
    "plt.xlabel('Gene Count')\n",
    "ax[0].set_ylabel('Log10(Number of GO terms)')\n",
    "ax[1].set_ylabel('Log10(Number of selected GO terms)')\n",
    "sns.despine()\n",
    "plt.savefig('figures/GO_BP_size_distribution.svg', bbox_inches='tight')\n",
    "plt.show()\n"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Fig. 2 \n",
    "### Panel b: CDF of the similarity quantile with random curve\n",
    "random similarity from picking a random GO from the thousand GO terms"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Count cutoff at similarity percentile 95.00%: 603\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 300x300 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df = pd.read_csv('data/GO_term_analysis/simrank_LLM_processed_selected_1000_go_terms.tsv', sep='\\t')\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(3,3))\n",
    "\n",
    "\n",
    "# Calculate the cumulative counts\n",
    "df_sorted = df[\"true_GO_term_sim_percentile\"].sort_values(ascending=False)\n",
    "cumulative_counts = np.arange(1, len(df_sorted) + 1)\n",
    "\n",
    "# Draw the ECDF with raw counts on the y-axis\n",
    "# x axis be the percentage not the raw fraction\n",
    "#66C5CC,#F6CF71,#F89C74,#DCB0F2,#87C55F,#9EB9F3,#FE88B1,#C9DB74,#8BE0A4,#B497E7,#D3B484,#B3B3B3\n",
    "ax.plot([perc*100 for perc in df_sorted], cumulative_counts, color='#4f75ad', linewidth=1, label='Assigned GO')\n",
    "\n",
    "rand_percentiles = df['random_sim_percentile'].sort_values(ascending=False)\n",
    "cumulative_counts_random = np.arange(1, len(rand_percentiles) + 1)\n",
    "ax.plot([perc*100 for perc in rand_percentiles], cumulative_counts_random, color='#818181', linewidth=0.5, linestyle='--', label='Random GO')\n",
    "\n",
    "\n",
    "# find the sim percentile at count_cutoff\n",
    "for sim_percentile_cutoff in [0.95]:\n",
    "    \n",
    "    count_cutoff = df_sorted[df_sorted >= sim_percentile_cutoff].count()\n",
    "    print(f'Count cutoff at similarity percentile {sim_percentile_cutoff*100:.2f}%: {count_cutoff}')\n",
    "\n",
    "    # print(f'Similarity percentile at count cutoff {count_cutoff}: {sim_percentile_at_count_cutoff*100:.2f}%')\n",
    "\n",
    "    ax.plot([103, sim_percentile_cutoff*100, sim_percentile_cutoff*100], [count_cutoff, count_cutoff, 0], color='red', linestyle=':', alpha= 0.8,linewidth=1)\n",
    "\n",
    "    \n",
    "\n",
    "    # Set ticks on the x-axis and y-axis\n",
    "    ax.set_xticks(list(ax.get_xticks()) + [sim_percentile_cutoff*100])\n",
    "    ax.set_yticks(list(ax.get_yticks()) + [count_cutoff] )\n",
    "\n",
    "    # Get the tick labels\n",
    "    xticklabels = ax.get_xticklabels()\n",
    "    yticklabels = ax.get_yticklabels()\n",
    "\n",
    "\n",
    "    # Modify specific tick label color on x-axis\n",
    "    xticklabels[-1].set_color('red')\n",
    "    xticklabels[-1].set_alpha(0.8)\n",
    "\n",
    "    # Modify specific tick label color on y-axis\n",
    "    yticklabels[-1].set_color('red')\n",
    "    yticklabels[-1].set_alpha(0.8)\n",
    "\n",
    "# invert x axis\n",
    "plt.gca().invert_xaxis()\n",
    "\n",
    "        \n",
    "ax.set_xlim(103, -1)\n",
    "ax.set_ylim(0, 1010)\n",
    "ax.set_xlabel('Similarity threshold\\n(Percent of GO terms less similar)')\n",
    "plt.ylabel('Cumulative counts of GPT-4 names')\n",
    "# ax.text(0.12, percentage_at_0_01, 'True GO in the top 1%', horizontalalignment='left', verticalalignment='bottom', \n",
    "#          transform=ax.transAxes, fontsize=14, color='red', alpha=0.8)\n",
    "\n",
    "sns.despine()\n",
    "plt.savefig(f\"figures/LLM_name_GO_term_percentile_CDF_w_random_95_perc_cutoff.svg\", bbox_inches='tight')\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Extended Data Figure 3"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CC: Counts at similarity percentile cutoff 95.00%: 642\n",
      "MF: Counts at similarity percentile cutoff 95.00%: 757\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 600x300 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "fig, axes = plt.subplots(nrows=1, ncols=2, figsize=(6, 3))\n",
    "\n",
    "branches = ['CC', 'MF']\n",
    "\n",
    "for idx, branch in enumerate(branches):\n",
    "    ax = axes[idx]\n",
    "    \n",
    "    # Load the dataset\n",
    "    df = pd.read_csv(f'data/GO_term_analysis/CC_MF_branch/sim_rank_LLM_processed_selected_1000_go_{branch}terms.tsv', sep='\\t')\n",
    "    \n",
    "    # Calculate the cumulative counts\n",
    "    df_sorted = df[\"true_GO_term_sim_percentile\"].sort_values(ascending=False)\n",
    "    cumulative_counts = np.arange(1, len(df_sorted) + 1)\n",
    "    \n",
    "    # Plot for Assigned GO\n",
    "    ax.plot([perc*100 for perc in df_sorted], cumulative_counts, color='#4f75ad', linewidth=1, label='Assigned GO')\n",
    "    \n",
    "    # Plot for Random GO\n",
    "    rand_percentiles = df['random_sim_percentile'].sort_values(ascending=False)\n",
    "    cumulative_counts_random = np.arange(1, len(rand_percentiles) + 1)\n",
    "    ax.plot([perc*100 for perc in rand_percentiles], cumulative_counts_random, color='#818181', linewidth=0.5, linestyle='--', label='Random GO')\n",
    "\n",
    "    # Set cutoff based on true_GO_term_sim_percentile == 0.98\n",
    "    cutoff_percentile = 0.95\n",
    "    # Find the index closest to the cutoff percentile\n",
    "    cutoff_index = (np.abs(df_sorted - cutoff_percentile)).argmin()\n",
    "    cutoff_count = cumulative_counts[cutoff_index]\n",
    "    print(f'{branch}: Counts at similarity percentile cutoff {cutoff_percentile*100:.2f}%: {cutoff_count}')\n",
    "    # Plot the cutoff line\n",
    "    ax.plot([103, cutoff_percentile*100, cutoff_percentile*100], [cutoff_count, cutoff_count, 0], color='red', linestyle=':', alpha=0.8, linewidth=0.5)\n",
    "    ax.set_xticks(list(ax.get_xticks()) + [cutoff_percentile*100])\n",
    "    ax.set_yticks(list(ax.get_yticks()) + [cutoff_count])\n",
    "    ax.get_xticklabels()[-1].set_color('red')\n",
    "    ax.get_xticklabels()[-1].set_alpha(0.8)\n",
    "    ax.get_yticklabels()[-1].set_color('red')\n",
    "    ax.get_yticklabels()[-1].set_alpha(0.8)\n",
    "    \n",
    "    # Invert x axis and set limits and labels\n",
    "    ax.invert_xaxis()\n",
    "    ax.set_xlim(103, -1)\n",
    "    ax.set_ylim(0, 1010)\n",
    "    ax.set_xlabel(f'Similarity threshold\\n(Percent of {branch} terms less similar)')\n",
    "    ax.set_ylabel(f'Cumulative {branch} term counts of GPT-4 names')\n",
    "    ax.legend()\n",
    "\n",
    "    sns.despine()\n",
    "\n",
    "# Adjust layout\n",
    "plt.tight_layout()\n",
    "\n",
    "# Save the figure\n",
    "plt.savefig(\"figures/LLM_name_GO_CC_MF_term_percentile_CDF_w_random_combined_95_cutoff.svg\", bbox_inches='tight')\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "median of the similarity scores: 0.49230277499999997\n",
      "mean of the similarity scores: 0.50620477945\n",
      "std of the similarity scores: 0.1434311227733863\n",
      "median of the quantile scores: 0.990438364224772\n",
      "mean of the quantile scores: 0.9315825242718447\n",
      "std of the quantile scores: 0.14723561855745396\n"
     ]
    }
   ],
   "source": [
    "print(f\"median of the similarity scores: {df['LLM_name_GO_term_sim'].median()}\")\n",
    "print(f\"mean of the similarity scores: {df['LLM_name_GO_term_sim'].mean()}\")\n",
    "print(f\"std of the similarity scores: {df['LLM_name_GO_term_sim'].std()}\")\n",
    "\n",
    "\n",
    "print(f\"median of the quantile scores: {df['true_GO_term_sim_percentile'].median()}\")\n",
    "print(f\"mean of the quantile scores: {df['true_GO_term_sim_percentile'].mean()}\")\n",
    "print(f\"std of the quantile scores: {df['true_GO_term_sim_percentile'].std()}\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "median +sd 1.2118443912473749\n",
      "number of terms with quantile within median + std 1000\n"
     ]
    }
   ],
   "source": [
    "print('median +sd', df['true_GO_term_sim_percentile'].median()+df['true_GO_term_sim_percentile'].std())\n",
    "print('number of terms with quantile within median + std', sum((df['true_GO_term_sim_percentile'] <= df['true_GO_term_sim_percentile'].median()+df['true_GO_term_sim_percentile'].std())))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Fig. 2c: example GO term for conversion between similarity to percentile ranking\n",
    "\n",
    "randomly select a term and convers semantic similarity to percentile score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "response to X-ray 0.5444585\n",
      "gpt4name:  DNA Damage Response and Repair\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 300x300 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import numpy as np\n",
    "import random \n",
    "import pandas as pd\n",
    "\n",
    "df = pd.read_csv('data/GO_term_analysis/simrank_LLM_processed_selected_1000_go_terms.tsv', sep='\\t')\n",
    "# load the list of all go term similarities\n",
    "with open (\"data/GO_term_analysis/all_go_sim_scores_gpt4.txt\", \"r\") as f:\n",
    "    all_go_sim_list = [float(line.strip()) for line in f]\n",
    "binsize = len(all_go_sim_list)//1000\n",
    "start = 0 \n",
    "end = binsize\n",
    "\n",
    "\n",
    "# get a random index\n",
    "random.seed(42)\n",
    "random_index = random.randint(0, len(df)-1)\n",
    "\n",
    "go_term = df.loc[random_index, 'Term_Description']\n",
    "gpt4_name = df.loc[random_index, 'gpt_4_default Name']\n",
    "sim = df.loc[df['Term_Description'] == go_term, 'LLM_name_GO_term_sim'].values[0]\n",
    "\n",
    "\n",
    "print(go_term, sim)\n",
    "print('gpt4name: ', gpt4_name)\n",
    "\n",
    "# Select a distribution related to the random GO term\n",
    "random_distribution = all_go_sim_list[random_index * binsize: (random_index+1) * binsize]\n",
    "\n",
    "# Plot histogram for counts\n",
    "plt.figure(figsize=(3,3))\n",
    "sns.histplot(random_distribution, bins=30, color='grey', kde=True, stat='count', element='poly')\n",
    "\n",
    "\n",
    "# Highlight the specific value of `sim` on the plot\n",
    "plt.axvline(x=sim, color='#457135', linestyle='-', linewidth=1)\n",
    "\n",
    "# Plot the 95th percentile in the distribution\n",
    "percentile_95 = np.percentile(random_distribution, 95)\n",
    "plt.axvline(x=percentile_95, color='red', linestyle=':', linewidth=1, label='95th Percentile')\n",
    "\n",
    "# Annotate the similarity score\n",
    "plt.text(sim+0.02, plt.ylim()[1]*0.15, f'semantic similarity between\\nassigned GO term and\\nGPT-4 name: {sim:.2f}', color='#457135', ha='left')\n",
    "# set x lim \n",
    "plt.xlim(0, max(random_distribution))\n",
    "plt.xlabel('Semantic similarity with GPT-4 name')\n",
    "plt.ylabel('Count of GO names')\n",
    "plt.legend(loc='upper left', bbox_to_anchor=(0.65, 0.95))\n",
    "sns.despine()\n",
    "\n",
    "plt.savefig(f\"figures/LLM_name_GO_term_sim_against_random_count_kde_{go_term}.svg\", bbox_inches='tight')\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Fig.3b Confidence score distribution between real contaminated and random gene sets\n",
    "\n",
    "Among 5 models + functional enrichment analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Gene set type</th>\n",
       "      <th>Unassigned GO terms</th>\n",
       "      <th>Assigned GO terms</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Genes</td>\n",
       "      <td>0</td>\n",
       "      <td>100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>50perc_contaminated_Genes</td>\n",
       "      <td>3</td>\n",
       "      <td>97</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>100perc_contaminated_Genes</td>\n",
       "      <td>85</td>\n",
       "      <td>15</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                Gene set type Unassigned GO terms Assigned GO terms\n",
       "0                       Genes                   0               100\n",
       "1   50perc_contaminated_Genes                   3                97\n",
       "2  100perc_contaminated_Genes                  85                15"
      ]
     },
     "execution_count": 1,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## check the go term enrichment analysis result \n",
    "import pandas as pd\n",
    "enrichr_res = pd.read_csv('./data/GO_term_analysis/100_GO_terms_enricher_res.tsv', sep='\\t')\n",
    "genes_col = [col for col in enrichr_res.columns if col.endswith('Genes')]\n",
    "# create the count table\n",
    "\n",
    "p_cutoff = 0.05\n",
    "count_table = pd.DataFrame(columns=['Gene set type','Unassigned GO terms', 'Assigned GO terms'])\n",
    "\n",
    "for col in genes_col: \n",
    "    num_enriched = sum(enrichr_res[f'{col} adj p-value'] <= p_cutoff)\n",
    "    num_unassigned = sum(enrichr_res[f'{col} adj p-value'] > p_cutoff)\n",
    "    count_row = pd.DataFrame({'Gene set type': col, 'Unassigned GO terms': num_unassigned, 'Assigned GO terms': num_enriched}, index=[0])\n",
    "    count_table = pd.concat([count_table, count_row], ignore_index=True)\n",
    "\n",
    "count_table\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x300 with 6 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Defining the custom order for Score Types and Models\n",
    "score_type_order = ['Real', 'Contaminated', 'Random']\n",
    "model_order = ['GPT-4', 'Gemini Pro', 'GPT-3.5', 'Mixtral Instruct', 'Llama2 70b']\n",
    "score_bin_order =[ 'Name not assigned', 'Low Confidence', 'Medium Confidence', 'High Confidence']\n",
    "color_palette = ['#A5AA99','#e39cc5','#af549e', '#6c2167']\n",
    "\n",
    "combine_df = pd.read_csv('./data/GO_term_analysis/model_compare/combined_scores_df.tsv', sep='\\t')\n",
    "# Creating a categorical type for ordering\n",
    "combine_df['Score Type'] = pd.Categorical(combine_df['Score Type'], categories=score_type_order, ordered=True)\n",
    "combine_df['Model'] = pd.Categorical(combine_df['Model'], categories=model_order, ordered=True)\n",
    "combine_df['Score Bin'] = pd.Categorical(combine_df['Score Bin'], categories=score_bin_order, ordered=True)\n",
    "\n",
    "# Grouping and stacking data \n",
    "stacked_bins_data_option2 = combine_df.groupby(['Model', 'Score Type', 'Score Bin']).size().unstack().fillna(0)\n",
    "\n",
    "# Modify the layout to include an additional subplot\n",
    "fig, axes = plt.subplots(1, len(model_order) + 1, figsize=(8,3), sharey=True)\n",
    "\n",
    "# Plotting for each model (existing code)\n",
    "for i, model in enumerate(model_order):\n",
    "    model_data = stacked_bins_data_option2.loc[model]\n",
    "    model_data.plot(kind='bar', stacked=True, ax=axes[i], width=0.8, legend=False, color=color_palette)\n",
    "    axes[i].set_xlabel(model)\n",
    "    axes[i].set_xticklabels(['Real', '50/50 Mix', 'Random'], rotation=45)\n",
    "    if i == 0:\n",
    "        axes[i].set_ylabel('Counts')\n",
    "    else:\n",
    "        axes[i].yaxis.set_visible(False)\n",
    "\n",
    "# Plotting the new stacked bar plot\n",
    "count_table.plot(kind='bar', stacked=True, ax=axes[-1], x='Gene set type',y=['Unassigned GO terms', 'Assigned GO terms'], color=['#A5AA99','#6c2167'], width=0.8, legend=False)\n",
    "axes[-1].set_xticklabels(['Real', '50/50 Mix', 'Random'], rotation=45)\n",
    "axes[-1].set_xlabel('Functional Enrichment')\n",
    "axes[-1].yaxis.set_visible(False)\n",
    "\n",
    "# Adjusting the layout for the new plot\n",
    "for i, ax in enumerate(axes):\n",
    "    if i != 0:\n",
    "        sns.despine(ax=ax, left=True)\n",
    "    else:\n",
    "        sns.despine(ax=ax)\n",
    "        \n",
    "plt.tight_layout()\n",
    "\n",
    "handles, labels = axes[0].get_legend_handles_labels()\n",
    "\n",
    "fig.legend(handles, labels, title='Score Bin', loc='upper left', bbox_to_anchor=(1, 1), fontsize=6)\n",
    "\n",
    "# Saving the figure\n",
    "plt.savefig('./figures/model_compare/combined_scores_bin_counts_by_models_with_enrichr.svg', bbox_inches='tight')\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "llm",
   "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": "3.11.5"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}