data_origins.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "import numpy as np\n",
    "import math\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Plots describing origins of data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "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>PublicationTitle</th>\n",
       "      <th>DatabaseIDs</th>\n",
       "      <th>PublicationDate</th>\n",
       "      <th>PublicationDOI</th>\n",
       "      <th>StudyType</th>\n",
       "      <th>SegmentationUse</th>\n",
       "      <th>SourceData</th>\n",
       "      <th>TrainingData</th>\n",
       "      <th>LabelData</th>\n",
       "      <th>Code</th>\n",
       "      <th>Modality</th>\n",
       "      <th>FileFormatsSource</th>\n",
       "      <th>FileFormatsTraining</th>\n",
       "      <th>FileFormatsLabel</th>\n",
       "      <th>DataOrigin</th>\n",
       "      <th>SegmentationType</th>\n",
       "      <th>SegmentationMethod</th>\n",
       "      <th>BiologicalScale</th>\n",
       "      <th>Notes</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>A novel framework for segmentation of secretor...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2014</td>\n",
       "      <td>10.1016/j.media.2013.12.008</td>\n",
       "      <td>Software</td>\n",
       "      <td>Proof-of-concept</td>\n",
       "      <td>No</td>\n",
       "      <td>NaN</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>TEM</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>this study</td>\n",
       "      <td>Semantic</td>\n",
       "      <td>Automated</td>\n",
       "      <td>Subcellular</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Cryo-FIB-SEM serial milling and block face ima...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2016</td>\n",
       "      <td>10.1016/j.jsb.2016.09.016</td>\n",
       "      <td>Method</td>\n",
       "      <td>Qualitative visualisation</td>\n",
       "      <td>No</td>\n",
       "      <td>NaN</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>Cryo-FIB-SEM</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>this study</td>\n",
       "      <td>Semantic</td>\n",
       "      <td>Automated</td>\n",
       "      <td>Subcellular</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Convolutional neural networks for automated an...</td>\n",
       "      <td>EMD-8594, EMD-8589, EMD-8590, EMD-8591, EMD-85...</td>\n",
       "      <td>2017</td>\n",
       "      <td>10.1038/nmeth.4405</td>\n",
       "      <td>Software</td>\n",
       "      <td>STA</td>\n",
       "      <td>No</td>\n",
       "      <td>NaN</td>\n",
       "      <td>EMDB</td>\n",
       "      <td>Github</td>\n",
       "      <td>Cryo-ET</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>map</td>\n",
       "      <td>this study, public data</td>\n",
       "      <td>Instance</td>\n",
       "      <td>Automated</td>\n",
       "      <td>Subcellular</td>\n",
       "      <td>Code integrated into existing software EMAN2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Extracellular space preservation aids the conn...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2015</td>\n",
       "      <td>10.7554/eLife.08206</td>\n",
       "      <td>Method</td>\n",
       "      <td>Proof-of-concept</td>\n",
       "      <td>Dryad</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>SBF SEM</td>\n",
       "      <td>zipped</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>this study</td>\n",
       "      <td>Instance</td>\n",
       "      <td>Automated</td>\n",
       "      <td>Tissue (connectomics)</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Domain adaptation for microscopy imaging</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2015</td>\n",
       "      <td>10.1109/TMI.2014.2376872</td>\n",
       "      <td>Software</td>\n",
       "      <td>Proof-of-concept</td>\n",
       "      <td>Some (DIADEM)</td>\n",
       "      <td>Some (DIADEM)</td>\n",
       "      <td>No</td>\n",
       "      <td>Institutional - working</td>\n",
       "      <td>3D EM</td>\n",
       "      <td>tiff</td>\n",
       "      <td>swc</td>\n",
       "      <td>NaN</td>\n",
       "      <td>this study</td>\n",
       "      <td>Semantic</td>\n",
       "      <td>Automated</td>\n",
       "      <td>Subcellular</td>\n",
       "      <td>Need to find code from reference to another pa...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                    PublicationTitle  \\\n",
       "0  A novel framework for segmentation of secretor...   \n",
       "1  Cryo-FIB-SEM serial milling and block face ima...   \n",
       "2  Convolutional neural networks for automated an...   \n",
       "3  Extracellular space preservation aids the conn...   \n",
       "4           Domain adaptation for microscopy imaging   \n",
       "\n",
       "                                         DatabaseIDs  PublicationDate  \\\n",
       "0                                                NaN             2014   \n",
       "1                                                NaN             2016   \n",
       "2  EMD-8594, EMD-8589, EMD-8590, EMD-8591, EMD-85...             2017   \n",
       "3                                                NaN             2015   \n",
       "4                                                NaN             2015   \n",
       "\n",
       "                PublicationDOI StudyType            SegmentationUse  \\\n",
       "0  10.1016/j.media.2013.12.008  Software           Proof-of-concept   \n",
       "1    10.1016/j.jsb.2016.09.016    Method  Qualitative visualisation   \n",
       "2           10.1038/nmeth.4405  Software                        STA   \n",
       "3          10.7554/eLife.08206    Method           Proof-of-concept   \n",
       "4     10.1109/TMI.2014.2376872  Software           Proof-of-concept   \n",
       "\n",
       "      SourceData   TrainingData LabelData                     Code  \\\n",
       "0             No            NaN        No                       No   \n",
       "1             No            NaN        No                       No   \n",
       "2             No            NaN      EMDB                   Github   \n",
       "3          Dryad             No        No                       No   \n",
       "4  Some (DIADEM)  Some (DIADEM)        No  Institutional - working   \n",
       "\n",
       "       Modality FileFormatsSource FileFormatsTraining FileFormatsLabel  \\\n",
       "0           TEM               NaN                 NaN              NaN   \n",
       "1  Cryo-FIB-SEM               NaN                 NaN              NaN   \n",
       "2       Cryo-ET               NaN                 NaN              map   \n",
       "3       SBF SEM            zipped                 NaN              NaN   \n",
       "4         3D EM              tiff                 swc              NaN   \n",
       "\n",
       "                DataOrigin SegmentationType SegmentationMethod  \\\n",
       "0               this study         Semantic          Automated   \n",
       "1               this study         Semantic          Automated   \n",
       "2  this study, public data         Instance          Automated   \n",
       "3               this study         Instance          Automated   \n",
       "4               this study         Semantic          Automated   \n",
       "\n",
       "         BiologicalScale                                              Notes  \n",
       "0            Subcellular                                                NaN  \n",
       "1            Subcellular                                                NaN  \n",
       "2            Subcellular       Code integrated into existing software EMAN2  \n",
       "3  Tissue (connectomics)                                                NaN  \n",
       "4            Subcellular  Need to find code from reference to another pa...  "
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Import csv of review data\n",
    "df = pd.read_csv(\"FAIRly_Depositing_Segmentations.csv\")\n",
    "df.head()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Categories of data origins\n",
    "\n",
    "origin_lookup = {\n",
    "    'other': 'Other studies',\n",
    "    'public challenge': 'Public challenge',\n",
    "    'public data': 'Public data',\n",
    "    'this study': 'Collected own data',\n",
    "    'unclear': 'Unclear'\n",
    "}\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "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>Category</th>\n",
       "      <th>StudyType</th>\n",
       "      <th>Count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Public challenge</td>\n",
       "      <td>All</td>\n",
       "      <td>13</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Public data</td>\n",
       "      <td>All</td>\n",
       "      <td>17</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Unclear</td>\n",
       "      <td>All</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Other studies</td>\n",
       "      <td>All</td>\n",
       "      <td>23</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Collected own data</td>\n",
       "      <td>All</td>\n",
       "      <td>181</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             Category StudyType  Count\n",
       "0    Public challenge       All     13\n",
       "1         Public data       All     17\n",
       "2             Unclear       All      4\n",
       "3       Other studies       All     23\n",
       "4  Collected own data       All    181"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Sort records into groups - all data, software papers, connectomics\n",
    "\n",
    "origins = {}\n",
    "\n",
    "for category in set(origin_lookup.values()):\n",
    "    origins[category] = {\"All\": 0, \"Software\": 0, \"Connectomics\": 0}\n",
    "for i in range(len(df)):\n",
    "    for value, category in origin_lookup.items():\n",
    "        if value.lower() in df.DataOrigin.iloc[i].lower():\n",
    "            # Note that each record can belong to >1 group, e.g., a software paper in connectomics is also included in all papers\n",
    "            origins[category][\"All\"] += 1\n",
    "            if df.StudyType.iloc[i] == \"Software\":\n",
    "                origins[category][\"Software\"] += 1\n",
    "            if df.BiologicalScale.iloc[i] == \"Tissue (connectomics)\":\n",
    "                origins[category][\"Connectomics\"] += 1\n",
    "origins_df = pd.DataFrame.from_dict(origins, orient=\"index\")\n",
    "origins_df[\"Category\"] = origins_df.index\n",
    "origins_df = origins_df.melt(\n",
    "    id_vars=\"Category\",\n",
    "    var_name=\"StudyType\",\n",
    "    value_name=\"Count\"\n",
    ")\n",
    "origins_df.head()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "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>Category</th>\n",
       "      <th>StudyType</th>\n",
       "      <th>Count</th>\n",
       "      <th>Percentage</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Public challenge</td>\n",
       "      <td>All</td>\n",
       "      <td>13</td>\n",
       "      <td>5.726872</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Public data</td>\n",
       "      <td>All</td>\n",
       "      <td>17</td>\n",
       "      <td>7.488987</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Unclear</td>\n",
       "      <td>All</td>\n",
       "      <td>4</td>\n",
       "      <td>1.762115</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Other studies</td>\n",
       "      <td>All</td>\n",
       "      <td>23</td>\n",
       "      <td>10.132159</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Collected own data</td>\n",
       "      <td>All</td>\n",
       "      <td>181</td>\n",
       "      <td>79.735683</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             Category StudyType  Count  Percentage\n",
       "0    Public challenge       All     13    5.726872\n",
       "1         Public data       All     17    7.488987\n",
       "2             Unclear       All      4    1.762115\n",
       "3       Other studies       All     23   10.132159\n",
       "4  Collected own data       All    181   79.735683"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Add percentages\n",
    "\n",
    "origins_df[\"Percentage\"] = np.nan  # initialise values for percentage\n",
    "\n",
    "# For all studies\n",
    "all_pct = 100 * origins_df.loc[origins_df.StudyType == \"All\", \"Count\"] / df.DataOrigin.count()\n",
    "origins_df.loc[origins_df.StudyType == \"All\", \"Percentage\"] = all_pct\n",
    "\n",
    "# For software studies\n",
    "software_pct = 100 * origins_df.loc[origins_df.StudyType == \"Software\",  \"Count\"] / df.loc[df.StudyType == \"Software\", \"DataOrigin\"].count()\n",
    "origins_df.loc[origins_df.StudyType == \"Software\", \"Percentage\"] = software_pct\n",
    "\n",
    "# For connectomics studies\n",
    "conn_pct = 100 * origins_df.loc[origins_df.StudyType == \"Connectomics\",  \"Count\"] / df.loc[df.BiologicalScale == \"Tissue (connectomics)\", \"DataOrigin\"].count()\n",
    "origins_df.loc[origins_df.StudyType == \"Connectomics\", \"Percentage\"] = conn_pct\n",
    "\n",
    "origins_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x241b7daab10>"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x300 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axs = plt.subplots(1, 3, figsize=(8,3), layout=\"constrained\")\n",
    "sns.set_style(\"white\")\n",
    "subfig_labels = [\"a\", \"b\", \"c\"]\n",
    "\n",
    "for subplot, studytype in enumerate([\"All\", \"Software\", \"Connectomics\"]):\n",
    "    study_df = origins_df[origins_df.StudyType == studytype]\n",
    "    axs[subplot].bar(\n",
    "        x=0,\n",
    "        height=study_df.loc[study_df.Category == \"Collected own data\", \"Percentage\"],\n",
    "    )\n",
    "    axs[subplot].text(\n",
    "        0,\n",
    "        study_df.loc[study_df.Category == \"Collected own data\", \"Percentage\"].iloc[0] / 2,\n",
    "        study_df[study_df.Category == 'Collected own data'].Count.iloc[0],\n",
    "        horizontalalignment=\"center\"\n",
    "    )\n",
    "\n",
    "    bottom = 0\n",
    "    for i in [\"Other studies\", \"Public data\", \"Public challenge\"]:\n",
    "        axs[subplot].bar(\n",
    "            x=1,\n",
    "            height=study_df.loc[study_df.Category == i, \"Percentage\"],\n",
    "            bottom=bottom,\n",
    "            label=i,\n",
    "        )\n",
    "        axs[subplot].text(\n",
    "            1,\n",
    "            bottom+1,\n",
    "            study_df.loc[study_df.Category == i, 'Count'].iloc[0],\n",
    "            horizontalalignment=\"center\"\n",
    "        )\n",
    "        bottom += study_df.loc[study_df.Category == i, 'Percentage'].iloc[0]\n",
    "\n",
    "    axs[subplot].set_ylim([0, 100])\n",
    "    axs[subplot].set_xticks([0, 1], [\"New data\", \"Reused data\"])\n",
    "    axs[subplot].set_xlabel(f\"{studytype} studies\", fontsize=12)\n",
    "    axs[subplot].set_ylabel(r\"% of studies surveyed\")\n",
    "    axs[subplot].text(-1, 97, subfig_labels[subplot], fontsize=16, fontweight=\"bold\")\n",
    "\n",
    "sns.despine()\n",
    "plt.savefig(\"./figs/data_origins.svg\")\n",
    "fig.legend(\n",
    "    [\"New data\", \"Other studies\", \"Public data\", \"Public challenge\"],\n",
    "    loc=\"upper center\",\n",
    "    bbox_to_anchor=(0.5, -0.05),\n",
    "    ncol=4\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Breakdown by imaging modality"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Categorise into X-ray, room-temp EM or cryoEM\n",
    "modality_lookup = {\n",
    "    '3D EM': [\"Room-temp EM\"],\n",
    "    'ATUM SEM, FIB SEM': [\"Room-temp EM\"],\n",
    "    'Cryo-EM': [\"Cryo EM\"],\n",
    "    'Cryo-ET': [\"Cryo EM\"],\n",
    "    'Cryo-FIB-SEM': [\"Cryo EM\"],\n",
    "    'EM': [\"Room-temp EM\"],\n",
    "    'Electron Tomography': [\"Room-temp EM\"],\n",
    "    'Electron tomography': [\"Room-temp EM\"],\n",
    "    'FIB SEM': [\"Room-temp EM\"],\n",
    "    'FIB SEM, SBF SEM': [\"Room-temp EM\"],\n",
    "    'FIB SEM, ssTEM': [\"Room-temp EM\"],\n",
    "    'MicroCT': [\"X-ray\"],\n",
    "    'MicroCT, CryoET': [\"X-ray\", \"Cryo EM\"],\n",
    "    'MicroCT, FIBSEM, Electron tomography': [\"X-ray\", \"Room-temp EM\"],\n",
    "    'MicroCT, Neutron imaging': [\"X-ray\"],\n",
    "    'SBF SEM': [\"Room-temp EM\"],\n",
    "    'SBF SEM ': [\"Room-temp EM\"],\n",
    "    'SBF SEM, FIB SEM': [\"Room-temp EM\"],\n",
    "    'SEM': [\"Room-temp EM\"],\n",
    "    'Soft XCT': [\"X-ray\"],\n",
    "    'Synchrotron X-ray microCT': [\"X-ray\"],\n",
    "    'Synchrotron X-ray microCT, SEM': [\"X-ray\", \"Room-temp EM\"],\n",
    "    'TEM': [\"Room-temp EM\"],\n",
    "    'TEM, SEM': [\"Room-temp EM\"],\n",
    "    'TEM, electron tomography, STEM': [\"Room-temp EM\"],\n",
    "    'X-ray microCT': [\"X-ray\"],\n",
    "    'X-ray microCT, FIB SEM': [\"X-ray\", \"Room-temp EM\"],\n",
    "    'X-ray microCT, SEM': [\"X-ray\", \"Room-temp EM\"],\n",
    "    'Cryo-ET & fluor': [\"Cryo EM\"],\n",
    "    'ssSEM': [\"Room-temp EM\"],\n",
    "    'ssTEM': [\"Room-temp EM\"],\n",
    "    'ssTEM, SBF SEM': [\"Room-temp EM\"],\n",
    "    'ssTEM, ssSEM, SBF SEM': [\"Room-temp EM\"],\n",
    "    }"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "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>Category</th>\n",
       "      <th>Modality</th>\n",
       "      <th>Count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Public challenge</td>\n",
       "      <td>Room-temp EM</td>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Public data</td>\n",
       "      <td>Room-temp EM</td>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Unclear</td>\n",
       "      <td>Room-temp EM</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Other studies</td>\n",
       "      <td>Room-temp EM</td>\n",
       "      <td>14</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Collected own data</td>\n",
       "      <td>Room-temp EM</td>\n",
       "      <td>66</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             Category      Modality  Count\n",
       "0    Public challenge  Room-temp EM     12\n",
       "1         Public data  Room-temp EM     12\n",
       "2             Unclear  Room-temp EM      2\n",
       "3       Other studies  Room-temp EM     14\n",
       "4  Collected own data  Room-temp EM     66"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "origins_mod = {}\n",
    "\n",
    "for category in set(origin_lookup.values()):\n",
    "    origins_mod[category] = {}\n",
    "    for modality_category in [\"Room-temp EM\", \"X-ray\", \"Cryo EM\"]:\n",
    "        origins_mod[category][modality_category] = 0\n",
    "\n",
    "for i in range(len(df)):\n",
    "    for value, category in origin_lookup.items():\n",
    "        if value.lower() in df.DataOrigin.iloc[i].lower():\n",
    "            try:\n",
    "                modality_groups = modality_lookup[df.Modality.iloc[i]]\n",
    "                for g in modality_groups:\n",
    "                    origins_mod[category][g] += 1\n",
    "            except:\n",
    "                pass  # ignore papers which are not modality-specific\n",
    "\n",
    "origins_mod_df = pd.DataFrame.from_dict(origins_mod, orient=\"index\")\n",
    "origins_mod_df[\"Category\"] = origins_mod_df.index\n",
    "origins_mod_df = origins_mod_df.melt(\n",
    "    id_vars=\"Category\",\n",
    "    var_name=\"Modality\",\n",
    "    value_name=\"Count\"\n",
    ")\n",
    "origins_mod_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "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>Category</th>\n",
       "      <th>Modality</th>\n",
       "      <th>Count</th>\n",
       "      <th>Percentage</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Public challenge</td>\n",
       "      <td>Room-temp EM</td>\n",
       "      <td>12</td>\n",
       "      <td>11.320755</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Public data</td>\n",
       "      <td>Room-temp EM</td>\n",
       "      <td>12</td>\n",
       "      <td>11.320755</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Unclear</td>\n",
       "      <td>Room-temp EM</td>\n",
       "      <td>2</td>\n",
       "      <td>1.886792</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Other studies</td>\n",
       "      <td>Room-temp EM</td>\n",
       "      <td>14</td>\n",
       "      <td>13.207547</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Collected own data</td>\n",
       "      <td>Room-temp EM</td>\n",
       "      <td>66</td>\n",
       "      <td>62.264151</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             Category      Modality  Count  Percentage\n",
       "0    Public challenge  Room-temp EM     12   11.320755\n",
       "1         Public data  Room-temp EM     12   11.320755\n",
       "2             Unclear  Room-temp EM      2    1.886792\n",
       "3       Other studies  Room-temp EM     14   13.207547\n",
       "4  Collected own data  Room-temp EM     66   62.264151"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Add percentages\n",
    "\n",
    "origins_mod_df[\"Percentage\"] = np.nan  # initialise values for percentage\n",
    "\n",
    "for modality_category in [\"Room-temp EM\", \"X-ray\", \"Cryo EM\"]:\n",
    "    each_df = origins_mod_df.loc[origins_mod_df.Modality == modality_category, \"Count\"]\n",
    "    pct = 100 * each_df / each_df.sum()\n",
    "    origins_mod_df.loc[origins_mod_df.Modality == modality_category, \"Percentage\"] = pct\n",
    "\n",
    "origins_mod_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x241b9a97fb0>"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x300 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axs = plt.subplots(1, 3, figsize=(8,3), layout=\"constrained\")\n",
    "sns.set_style(\"white\")\n",
    "subfig_labels = [\"a\", \"b\", \"c\"]\n",
    "\n",
    "for subplot, mod in enumerate([\"X-ray\", \"Room-temp EM\", \"Cryo EM\"]):\n",
    "    mod_df = origins_mod_df[origins_mod_df.Modality == mod]\n",
    "    axs[subplot].bar(\n",
    "        x=0,\n",
    "        height=mod_df.loc[mod_df.Category == \"Collected own data\", \"Percentage\"],\n",
    "    )\n",
    "    axs[subplot].text(\n",
    "        0,\n",
    "        mod_df.loc[mod_df.Category == \"Collected own data\", \"Percentage\"].iloc[0] / 2,\n",
    "        mod_df[mod_df.Category == 'Collected own data'].Count.iloc[0],\n",
    "        horizontalalignment=\"center\"\n",
    "    )\n",
    "\n",
    "    bottom = 0\n",
    "    for i in [\"Other studies\", \"Public data\", \"Public challenge\"]:\n",
    "        axs[subplot].bar(\n",
    "            x=1,\n",
    "            height=mod_df.loc[mod_df.Category == i, \"Percentage\"],\n",
    "            bottom=bottom,\n",
    "            label=i,\n",
    "        )\n",
    "        if mod_df.loc[mod_df.Category == i, 'Count'].iloc[0] > 0:\n",
    "            axs[subplot].text(\n",
    "                1,\n",
    "                bottom+1,\n",
    "                mod_df.loc[mod_df.Category == i, 'Count'].iloc[0],\n",
    "                horizontalalignment=\"center\"\n",
    "            )\n",
    "        bottom += mod_df.loc[mod_df.Category == i, 'Percentage'].iloc[0]\n",
    "\n",
    "    axs[subplot].set_ylim([0, 100])\n",
    "    axs[subplot].set_xticks([0, 1], [\"New data\", \"Reused data\"])\n",
    "    axs[subplot].set_xlabel(mod, fontsize=12)\n",
    "    axs[subplot].set_ylabel(r\"% of studies surveyed\")\n",
    "    axs[subplot].text(-1, 97, subfig_labels[subplot], fontsize=16, fontweight=\"bold\")\n",
    "\n",
    "sns.despine()\n",
    "plt.savefig(\"./figs/data_origins_by_modality.svg\")\n",
    "fig.legend(\n",
    "    [\"New data\", \"Other studies\", \"Public data\", \"Public challenge\"],\n",
    "    loc=\"upper center\",\n",
    "    bbox_to_anchor=(0.5, -0.05),\n",
    "    ncol=4\n",
    ")"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "analysis",
   "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": "3.12.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}