notebooks/06_global_analysis_of_protein_folding_stability.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Summary\n",
    "\n",
    "----"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Imports"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import concurrent.futures\n",
    "import itertools\n",
    "import os\n",
    "from pathlib import Path\n",
    "\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "\n",
    "import matplotlib.image as mpimg\n",
    "import matplotlib.pyplot as plt\n",
    "import proteinsolver\n",
    "import psutil\n",
    "import pyarrow as pa\n",
    "import pyarrow.parquet as pq\n",
    "import torch\n",
    "from kmbio import PDB\n",
    "from scipy import stats\n",
    "from torch.utils.data import DataLoader, Dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "DEBUG = \"CI\" not in os.environ"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "if DEBUG:\n",
    "    %load_ext autoreload\n",
    "    %autoreload 2"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Parameters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "UNIQUE_ID = \"191f05de\"  # No attention\n",
    "# UNIQUE_ID = \"0007604c\"  # 5-layer graph-conv with attention, batch_size=1\n",
    "# UNIQUE_ID = \"91fc9ab9\"  # 4-layer graph-conv with attention, batch_size=4"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "BEST_STATE_FILES = {\n",
    "    #\n",
    "    \"191f05de\": \"protein_train/191f05de/e53-s1952148-d93703104.state\"\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "PosixPath('06_global_analysis_of_protein_folding')"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "NOTEBOOK_NAME = \"06_global_analysis_of_protein_folding\"\n",
    "NOTEBOOK_PATH = Path(NOTEBOOK_NAME)\n",
    "NOTEBOOK_PATH.mkdir(exist_ok=True)\n",
    "NOTEBOOK_PATH"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "PosixPath('/home/kimlab1/database_data/datapkg_input_dir')"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "INPUT_PATH = Path(os.getenv(\"DATAPKG_INPUT_DIR\"))\n",
    "INPUT_PATH"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "PosixPath('/home/kimlab1/database_data/datapkg_output_dir')"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "DATAPKG_DATA_DIR = Path(f\"~/datapkg_data_dir\").expanduser().resolve()\n",
    "DATAPKG_DATA_DIR"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "device(type='cuda', index=0)"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "device = torch.device(\"cuda:0\" if torch.cuda.is_available() else \"cpu\")\n",
    "device"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'/home/kimlab1/database_data/datapkg_output_dir'"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "proteinsolver.settings.data_url = DATAPKG_DATA_DIR.as_posix()\n",
    "proteinsolver.settings.data_url"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "toc-hr-collapsed": false
   },
   "source": [
    "## Load data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "aan0693_designed-PDB-files\t   hhpred\n",
      "aan0693_designed-PDB-files.gz.zip  hhpred2\n",
      "aan0693_SI_datasets\t\t   mutation_structures_for_rosetta\n",
      "aan0693_SI_datasets.tar.gz.zip\t   swissmodel\n"
     ]
    }
   ],
   "source": [
    "!ls {INPUT_PATH}/global_analysis_of_protein_folding"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "aan0693_designed-PDB-files.gz  nmr  other  other2  rd1\trd2  rd3  rd4\n"
     ]
    }
   ],
   "source": [
    "!ls {INPUT_PATH}/global_analysis_of_protein_folding/aan0693_designed-PDB-files"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "counts_and_ec50s.tar.gz\t\t  protein_and_dna_sequences.tar.gz\n",
      "design_scripts.tgz\t\t  stability_scores\n",
      "design_structural_metrics\t  stability_scores.tar.gz\n",
      "design_structural_metrics.tar.gz  unfolded_state_model_params\n",
      "fig1_thermodynamic_data.csv\n"
     ]
    }
   ],
   "source": [
    "!ls {INPUT_PATH}/global_analysis_of_protein_folding/aan0693_SI_datasets"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### aan0693_SI_datasets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "rd1_stability_scores  rd3_stability_scores  ssm2_stability_scores\n",
      "rd2_stability_scores  rd4_stability_scores\n"
     ]
    }
   ],
   "source": [
    "!ls {INPUT_PATH}/global_analysis_of_protein_folding/aan0693_SI_datasets/stability_scores"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "def remove_controls(df):\n",
    "    df = df[\n",
    "        (~df[\"name\"].str.endswith(\"_hp\")) & (~df[\"name\"].str.endswith(\"_random\")) & (~df[\"name\"].str.endswith(\"_buryD\"))\n",
    "    ]\n",
    "    return df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "def load_stability_scores(key):\n",
    "    stability_scores = pd.read_csv(\n",
    "        INPUT_PATH\n",
    "        / \"global_analysis_of_protein_folding\"\n",
    "        / \"aan0693_SI_datasets\"\n",
    "        / \"stability_scores\"\n",
    "        / f\"{key}_stability_scores\",\n",
    "        sep=\"\\t\",\n",
    "    )\n",
    "    stability_scores = remove_controls(stability_scores)\n",
    "\n",
    "    for energy_function in [\"talaris2013\", \"betanov15\"]:\n",
    "        rosetta_energies_file = (\n",
    "            INPUT_PATH\n",
    "            / \"global_analysis_of_protein_folding\"\n",
    "            / \"aan0693_SI_datasets\"\n",
    "            / \"design_structural_metrics\"\n",
    "            / f\"{key}_relax_scored_{'filtered_' if energy_function == 'betanov15' else ''}{energy_function}.sc\"\n",
    "        )\n",
    "        if not rosetta_energies_file.is_file():\n",
    "            print(f\"Not loading Rosetta energies for {energy_function}!\")\n",
    "            continue\n",
    "\n",
    "        before_ = len(stability_scores)\n",
    "        relax_scored_filtered = pd.read_csv(\n",
    "            rosetta_energies_file, sep=\"\\t\" if energy_function == \"betanov15\" else \" +\", engine=\"python\"\n",
    "        ).rename(columns={\"description\": \"name\", \"total_score\": f\"{energy_function}_score\"})\n",
    "        stability_scores = stability_scores.merge(\n",
    "            relax_scored_filtered[[\"name\", f\"{energy_function}_score\"]], on=\"name\", how=\"outer\"\n",
    "        )\n",
    "#         assert len(stability_scores) == before_, (len(stability_scores), before_)\n",
    "\n",
    "    stability_scores[\"library_name\"] = key\n",
    "    return stability_scores"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### stability_scores"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "# stability_scores = {}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "# for key in [\"rd1\", \"rd2\", \"rd3\", \"rd4\", \"ssm2\"]:\n",
    "#     print(key)\n",
    "#     stability_scores[key] = load_stability_scores(key)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "# stability_scores[\"fig1\"] = pd.read_csv(\n",
    "#     INPUT_PATH / \"global_analysis_of_protein_folding\" / \"aan0693_SI_datasets\" / \"fig1_thermodynamic_data.csv\"\n",
    "# ).assign(library_name=\"fig1\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "# with open(NOTEBOOK_PATH.joinpath(\"stability_scores.raw.incremental.torch\"), \"wb\") as fout:\n",
    "#     torch.save(stability_scores, fout)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "stability_scores = torch.load(NOTEBOOK_PATH.joinpath(\"stability_scores.torch\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "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>name</th>\n",
       "      <th>sequence</th>\n",
       "      <th>assay_library</th>\n",
       "      <th>ec50_t</th>\n",
       "      <th>ec50_95ci_lbound_t</th>\n",
       "      <th>ec50_95ci_ubound_t</th>\n",
       "      <th>ec50_95ci_t</th>\n",
       "      <th>ec50_pred_t</th>\n",
       "      <th>ec50_rise_t</th>\n",
       "      <th>stabilityscore_t</th>\n",
       "      <th>...</th>\n",
       "      <th>ec50_95ci_lbound_c</th>\n",
       "      <th>ec50_95ci_ubound_c</th>\n",
       "      <th>ec50_95ci_c</th>\n",
       "      <th>ec50_pred_c</th>\n",
       "      <th>ec50_rise_c</th>\n",
       "      <th>stabilityscore_c</th>\n",
       "      <th>stabilityscore</th>\n",
       "      <th>talaris2013_score</th>\n",
       "      <th>betanov15_score</th>\n",
       "      <th>library_name</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1A11</td>\n",
       "      <td>GSEKMSTAISVLLAQAVFLLLTSQRGSGNGSGGNGSGGNGSGNGSG...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.69</td>\n",
       "      <td>2.5</td>\n",
       "      <td>2.9</td>\n",
       "      <td>0.4</td>\n",
       "      <td>3.01</td>\n",
       "      <td>-0.32</td>\n",
       "      <td>-0.15</td>\n",
       "      <td>...</td>\n",
       "      <td>2.4</td>\n",
       "      <td>2.7</td>\n",
       "      <td>0.3</td>\n",
       "      <td>1.41</td>\n",
       "      <td>1.12</td>\n",
       "      <td>0.53</td>\n",
       "      <td>-0.15</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>rd4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1AIE</td>\n",
       "      <td>EYFTLQIRGRERFEMFRELNEALELKDAQAGSGNGSGGNGSGGNGS...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.07</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.1</td>\n",
       "      <td>0.1</td>\n",
       "      <td>0.49</td>\n",
       "      <td>3.58</td>\n",
       "      <td>1.71</td>\n",
       "      <td>...</td>\n",
       "      <td>3.9</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0.1</td>\n",
       "      <td>0.61</td>\n",
       "      <td>3.34</td>\n",
       "      <td>1.60</td>\n",
       "      <td>1.60</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>rd4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1AMB</td>\n",
       "      <td>DAEFRHDSGYEVHHQKLVFFAEDVGSNKGNGSGGNGSGGNGSGNGS...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>1.24</td>\n",
       "      <td>1.1</td>\n",
       "      <td>1.4</td>\n",
       "      <td>0.3</td>\n",
       "      <td>2.40</td>\n",
       "      <td>-1.16</td>\n",
       "      <td>-0.55</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.5</td>\n",
       "      <td>1.2</td>\n",
       "      <td>1.7</td>\n",
       "      <td>1.47</td>\n",
       "      <td>-1.11</td>\n",
       "      <td>-0.53</td>\n",
       "      <td>-0.55</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>rd4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1AMC</td>\n",
       "      <td>DAEFRHDSGYEVHHQKLVFFAEDVGSNKNGSGGNGSGGNGSGNGSG...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.42</td>\n",
       "      <td>2.3</td>\n",
       "      <td>2.6</td>\n",
       "      <td>0.3</td>\n",
       "      <td>2.39</td>\n",
       "      <td>0.03</td>\n",
       "      <td>0.01</td>\n",
       "      <td>...</td>\n",
       "      <td>0.6</td>\n",
       "      <td>1.1</td>\n",
       "      <td>0.5</td>\n",
       "      <td>1.47</td>\n",
       "      <td>-0.53</td>\n",
       "      <td>-0.25</td>\n",
       "      <td>-0.25</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>rd4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1AML</td>\n",
       "      <td>DAEFRHDSGYEVHHQKLVFFAEDVGSNKGAIIGLMVGGVVGSGGNG...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.20</td>\n",
       "      <td>2.1</td>\n",
       "      <td>2.3</td>\n",
       "      <td>0.2</td>\n",
       "      <td>2.09</td>\n",
       "      <td>0.10</td>\n",
       "      <td>0.05</td>\n",
       "      <td>...</td>\n",
       "      <td>0.5</td>\n",
       "      <td>1.1</td>\n",
       "      <td>0.6</td>\n",
       "      <td>1.35</td>\n",
       "      <td>-0.49</td>\n",
       "      <td>-0.23</td>\n",
       "      <td>-0.23</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>rd4</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 21 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   name                                           sequence  assay_library  \\\n",
       "0  1A11  GSEKMSTAISVLLAQAVFLLLTSQRGSGNGSGGNGSGGNGSGNGSG...            4.0   \n",
       "1  1AIE  EYFTLQIRGRERFEMFRELNEALELKDAQAGSGNGSGGNGSGGNGS...            4.0   \n",
       "2  1AMB  DAEFRHDSGYEVHHQKLVFFAEDVGSNKGNGSGGNGSGGNGSGNGS...            4.0   \n",
       "3  1AMC  DAEFRHDSGYEVHHQKLVFFAEDVGSNKNGSGGNGSGGNGSGNGSG...            4.0   \n",
       "4  1AML  DAEFRHDSGYEVHHQKLVFFAEDVGSNKGAIIGLMVGGVVGSGGNG...            4.0   \n",
       "\n",
       "   ec50_t  ec50_95ci_lbound_t  ec50_95ci_ubound_t  ec50_95ci_t  ec50_pred_t  \\\n",
       "0    2.69                 2.5                 2.9          0.4         3.01   \n",
       "1    4.07                 4.0                 4.1          0.1         0.49   \n",
       "2    1.24                 1.1                 1.4          0.3         2.40   \n",
       "3    2.42                 2.3                 2.6          0.3         2.39   \n",
       "4    2.20                 2.1                 2.3          0.2         2.09   \n",
       "\n",
       "   ec50_rise_t  stabilityscore_t  ...  ec50_95ci_lbound_c  ec50_95ci_ubound_c  \\\n",
       "0        -0.32             -0.15  ...                 2.4                 2.7   \n",
       "1         3.58              1.71  ...                 3.9                 4.0   \n",
       "2        -1.16             -0.55  ...                -0.5                 1.2   \n",
       "3         0.03              0.01  ...                 0.6                 1.1   \n",
       "4         0.10              0.05  ...                 0.5                 1.1   \n",
       "\n",
       "   ec50_95ci_c  ec50_pred_c  ec50_rise_c  stabilityscore_c  stabilityscore  \\\n",
       "0          0.3         1.41         1.12              0.53           -0.15   \n",
       "1          0.1         0.61         3.34              1.60            1.60   \n",
       "2          1.7         1.47        -1.11             -0.53           -0.55   \n",
       "3          0.5         1.47        -0.53             -0.25           -0.25   \n",
       "4          0.6         1.35        -0.49             -0.23           -0.23   \n",
       "\n",
       "   talaris2013_score  betanov15_score  library_name  \n",
       "0                NaN              NaN           rd4  \n",
       "1                NaN              NaN           rd4  \n",
       "2                NaN              NaN           rd4  \n",
       "3                NaN              NaN           rd4  \n",
       "4                NaN              NaN           rd4  \n",
       "\n",
       "[5 rows x 21 columns]"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "stability_scores[\"rd4\"].head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Load model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "%run protein_train/{UNIQUE_ID}/model.py"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "batch_size = 1\n",
    "num_features = 20\n",
    "adj_input_size = 2\n",
    "hidden_size = 128\n",
    "frac_present = 0.5\n",
    "frac_present_valid = frac_present\n",
    "info_size= 1024"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'protein_train/191f05de/e53-s1952148-d93703104.state'"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "state_file = BEST_STATE_FILES[UNIQUE_ID]\n",
    "state_file"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "net = Net(\n",
    "    x_input_size=num_features + 1, adj_input_size=adj_input_size, hidden_size=hidden_size, output_size=num_features\n",
    ")\n",
    "net.load_state_dict(torch.load(state_file, map_location=device))\n",
    "net.eval()\n",
    "net = net.to(device)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "toc-hr-collapsed": false
   },
   "source": [
    "## Structure probabilities"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "structure_file = \"/home/kimlab1/database_data/datapkg_input_dir/global_analysis_of_protein_folding/aan0693_designed-PDB-files/rd1/EEHEE_rd1_0004.pdb\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "structure = PDB.load(structure_file)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "pdata = proteinsolver.utils.extract_seq_and_adj(structure, 'A')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tensor([-2.1866, -2.8192, -0.8576, -3.1876, -6.9420, -1.0521, -0.1976, -1.1227,\n",
      "        -1.4856, -2.1950, -2.0287, -0.3771, -2.8255, -1.1229, -2.2655, -3.1545,\n",
      "        -0.8361, -0.3876, -1.8040, -1.5459, -2.0852, -0.1505, -1.7250, -2.2637,\n",
      "        -3.0158, -0.1281, -0.4505, -3.4455, -3.2797, -0.2760, -1.1259, -2.8295,\n",
      "        -2.0394, -0.4193, -4.0072, -1.1092, -2.4877, -1.2500, -3.2661, -2.8514,\n",
      "        -3.5015])\n"
     ]
    }
   ],
   "source": [
    "data = proteinsolver.datasets.protein.row_to_data(pdata)\n",
    "data = proteinsolver.datasets.protein.transform_edge_attr(data)\n",
    "data.to(device)\n",
    "\n",
    "print(\n",
    "    proteinsolver.utils.get_node_outputs(\n",
    "        net, data.x, data.edge_index, data.edge_attr, 20, output_transform=\"logproba\"\n",
    "    )\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### structure_proba"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_structure_proba(net, filename):\n",
    "    try:\n",
    "        library_name = filename.split(\"_\")[1]\n",
    "    except IndexError:\n",
    "        return np.nan, np.nan\n",
    "    filepath = (\n",
    "        INPUT_PATH / \"global_analysis_of_protein_folding\" / \"aan0693_designed-PDB-files\" / library_name / filename\n",
    "    )\n",
    "    if not filepath.is_file():\n",
    "        return np.nan, np.nan\n",
    "    structure = PDB.load(filepath)\n",
    "    pdata = proteinsolver.utils.extract_seq_and_adj(structure, \"A\")\n",
    "    data = proteinsolver.datasets.protein.row_to_data(pdata)\n",
    "    data = proteinsolver.datasets.protein.transform_edge_attr(data)\n",
    "    data = data.to(device)\n",
    "    #     log_prob = proteinsolver.utils.get_node_proba(net, data.x, data.edge_index, data.edge_attr, 20)\n",
    "    log_prob = proteinsolver.utils.get_node_outputs(\n",
    "        net, data.x, data.edge_index, data.edge_attr, 20, output_transform=\"logproba\"\n",
    "    )\n",
    "    return log_prob.sum().item(), len(pdata.sequence)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_filepath(filename):\n",
    "    try:\n",
    "        library = filename.split(\"_\")[1]\n",
    "    except IndexError:\n",
    "        #         print(f\"Cound not guess library for file: {filename}!\")\n",
    "        return None\n",
    "    filepath = INPUT_PATH / \"global_analysis_of_protein_folding\" / \"aan0693_designed-PDB-files\" / library / filename\n",
    "    if not filepath.is_file():\n",
    "        #         print(f\"File {filename} does not exist!\")\n",
    "        return None\n",
    "    return filepath\n",
    "\n",
    "\n",
    "class DesignsDataset(Dataset):\n",
    "    \"\"\"Face Landmarks dataset.\"\"\"\n",
    "\n",
    "    @staticmethod\n",
    "    def transform(pdata):\n",
    "        data = proteinsolver.datasets.protein.row_to_data(pdata)\n",
    "        data = proteinsolver.datasets.protein.transform_edge_attr(data)\n",
    "        return data\n",
    "\n",
    "    @staticmethod\n",
    "    def add_domain(filepath, data):\n",
    "        domain = filepath.name.split(\"_\")[0]\n",
    "        data.domain = domain\n",
    "        return data\n",
    "\n",
    "    def __init__(self, filenames, sequences, use_ref_graph=True):\n",
    "        self.use_ref_graph = use_ref_graph\n",
    "\n",
    "        ref_files = {\n",
    "            \"EEHEE\": (\"EEHEE_nmr_5up1.cif\", \"TTVKLGDIKVTFDNPEKAKKYAQKLAKIYQLTVHVHGDTIHVK\"),\n",
    "            \"EHEE\": (\"EHEE_nmr_5up5.cif\", \"TQTQEFDNEEEARKAEKELRKENRRVTVTQENGRWRVTWD\"),\n",
    "            \"HEEH\": (\"HEEH_nmr_5uyo.cif\", \"DVEEQIRRLEEVLKKNQPVTWNGTTYTDPNEIKKVIEELRKSM\"),\n",
    "            \"HHH\": (\"HHH_nmr_5uoi.cif\", \"RKWEEIAERLREEFNINPEEAREAVEKAGGNEEEARRIVKKRL\"),\n",
    "        }\n",
    "        domains = list({f.split(\"_\")[0] for f in filenames})\n",
    "        if set(domains) - set(ref_files):\n",
    "            print(f\"Do not have reference for the following domains: {set(domains) - set(ref_files)}!\")\n",
    "        pdata_ref = {}\n",
    "        for domain, (filename, sequence) in ref_files.items():\n",
    "            filepath = get_filepath(filename)\n",
    "            structure = PDB.load(filepath)\n",
    "            pdata = proteinsolver.utils.extract_seq_and_adj(structure, \"A\")\n",
    "            assert pdata.sequence == sequence, (domain, filename, pdata.sequence, sequence)\n",
    "            pdata_ref[domain] = pdata\n",
    "        #         pdata_ref[\"EEHEE\"] = pdata_ref[\"EEHEE\"]._replace(sequence=pdata_ref[\"EEHEE\"].sequence + \"LESSGGS\")\n",
    "        self.data_ref = {domain: self.transform(pdata) for domain, pdata in pdata_ref.items()}\n",
    "\n",
    "        self.filepaths, self.sequences = list(\n",
    "            zip(\n",
    "                *[\n",
    "                    (f, s)\n",
    "                    for f, s in itertools.zip_longest((get_filepath(f) for f in filenames), sequences)\n",
    "                    if f is not None\n",
    "                ]\n",
    "            )\n",
    "        )\n",
    "\n",
    "    def __len__(self):\n",
    "        return len(self.filepaths)\n",
    "\n",
    "    def __getitem__(self, idx):\n",
    "        filepath = self.filepaths[idx]\n",
    "        sequence = self.sequences[idx]\n",
    "        structure = PDB.load(filepath)\n",
    "        pdata = proteinsolver.utils.extract_seq_and_adj(structure, \"A\")\n",
    "\n",
    "        name = filepath.name\n",
    "        domain = filepath.name.split(\"_\")[0]\n",
    "\n",
    "        start = None\n",
    "        end = None\n",
    "        if pd.isnull(sequence):\n",
    "            pass\n",
    "        elif pdata.sequence != sequence:\n",
    "            mismatch_size = len(sequence) - len(pdata.sequence)\n",
    "            if mismatch_size <= 0:\n",
    "                print(f\"Sequence mismatch of size '{mismatch_size}' for protein {filepath.name}!\")\n",
    "                return name, domain, None\n",
    "            else:\n",
    "                for start in range(mismatch_size + 1):\n",
    "                    end = start + len(pdata.sequence)\n",
    "                    sequence_slice = sequence[start:end]\n",
    "                    frac_identical = sum(\n",
    "                        a1 == a2 for (a1, a2) in itertools.zip_longest(pdata.sequence, sequence_slice)\n",
    "                    ) / len(pdata.sequence)\n",
    "                    if frac_identical >= 1.0:\n",
    "                        break\n",
    "                if frac_identical < 1.0:\n",
    "                    print(f\"Sequence mismatch for protein {filepath.name}: ({frac_identical:.2%} identical).\")\n",
    "                    return name, domain, None\n",
    "\n",
    "        data = self.transform(pdata)\n",
    "\n",
    "        if self.use_ref_graph and data.domain in self.data_ref:\n",
    "            data_ref = self.data_ref[data.domain].clone()\n",
    "            if start is not None:\n",
    "                data_ref.x = data_ref.x[start:end]\n",
    "                mask = ((data_ref.edge_index >= start) & (data_ref.edge_index < end)).all(dim=0)\n",
    "                data_ref.edge_index = data_ref.edge_index[:, mask] - start\n",
    "                data_ref.edge_attr = data_ref.edge_attr[mask, :]\n",
    "                assert (data_ref.edge_index >= 0).all().item()\n",
    "                assert (data_ref.edge_index < data_ref.x.size(0)).all().item()\n",
    "\n",
    "            if data.x.size(0) != data_ref.x.size(0):\n",
    "                print(f\"Reference structure length mismatch for protein {filepath.name}.\")\n",
    "                data.edge_index_ref = None\n",
    "                data.edge_attr_ref = None\n",
    "                data.identity_to_ref = None\n",
    "            else:\n",
    "                assert data.x.size(0) == data_ref.x.size(0)\n",
    "                assert data_ref.edge_index.min() >= 0\n",
    "                assert data_ref.edge_index.max() < data.x.size(0)\n",
    "                data.edge_index_ref = data_ref.edge_index\n",
    "                data.edge_attr_ref = data_ref.edge_attr\n",
    "                data.identity_to_ref = float((data.x == data_ref.x).sum().item()) / data.x.size(0)\n",
    "        else:\n",
    "            data.edge_index_ref = data.edge_index\n",
    "            data.edge_attr_ref = data.edge_attr\n",
    "            data.identity_to_ref = 1\n",
    "\n",
    "        return name, domain, data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "design_dataset = DesignsDataset(\n",
    "    stability_scores[\"rd2\"][\"name\"], stability_scores[\"rd2\"][\"sequence\"], use_ref_graph=False\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "('EHEE_rd2_1032.pdb',\n",
       " 'EHEE',\n",
       " Data(edge_attr=[1186, 2], edge_attr_ref=[1186, 2], edge_index=[2, 1186], edge_index_ref=[2, 1186], identity_to_ref=[1], x=[40]))"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "design_dataset[4000]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "# for library in [\"rd1\", \"rd2\", \"rd3\", \"rd4\"]:\n",
    "#     stability_scores[library][\"network_score\"], stability_scores[library][\"sequence_length\"] = np.array(\n",
    "#         [get_structure_proba(net, name) for name in stability_scores[library][\"name\"]]\n",
    "#     ).T"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "def apply_scores(df):\n",
    "    results = []\n",
    "    design_dataset = DesignsDataset(df[\"name\"], df[\"sequence\"], use_ref_graph=False)\n",
    "    for i, (name, domain, data) in enumerate(\n",
    "        DataLoader(design_dataset, batch_size=1, shuffle=False, num_workers=12, collate_fn=lambda lst: lst[0])\n",
    "    ):\n",
    "        if (i + 1) % 200 == 0:\n",
    "            print(i + 1)\n",
    "\n",
    "        if data is None:\n",
    "            results.append((name, None, None, None, domain, None))\n",
    "            continue\n",
    "\n",
    "        data = data.to(device)\n",
    "        network_score = proteinsolver.utils.get_node_outputs(\n",
    "            net, data.x, data.edge_index, data.edge_attr, 20, output_transform=\"logproba\"\n",
    "        )\n",
    "        if data.edge_index_ref is not None:\n",
    "            network_score_wref = proteinsolver.utils.get_node_outputs(\n",
    "                net, data.x, data.edge_index_ref, data.edge_attr_ref, 20, output_transform=\"logproba\"\n",
    "            )\n",
    "        else:\n",
    "            network_score_wref = None\n",
    "        results.append(\n",
    "            (\n",
    "                name,\n",
    "                network_score.mean().item(),\n",
    "                network_score_wref.mean().item() if network_score_wref is not None else None,\n",
    "                data.x.size(0),\n",
    "                domain,\n",
    "                data.identity_to_ref,\n",
    "            )\n",
    "        )\n",
    "    columns = [\"name\", \"network_score\", \"network_score_wref\", \"sequence_length\", \"domain\", \"identity_to_ref\"]\n",
    "    results_df = pd.DataFrame(results, columns=columns)\n",
    "    return results_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "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>name</th>\n",
       "      <th>sequence</th>\n",
       "      <th>assay_library</th>\n",
       "      <th>ec50_t</th>\n",
       "      <th>ec50_95ci_lbound_t</th>\n",
       "      <th>ec50_95ci_ubound_t</th>\n",
       "      <th>ec50_95ci_t</th>\n",
       "      <th>ec50_pred_t</th>\n",
       "      <th>ec50_rise_t</th>\n",
       "      <th>stabilityscore_t</th>\n",
       "      <th>...</th>\n",
       "      <th>ec50_95ci_lbound_c</th>\n",
       "      <th>ec50_95ci_ubound_c</th>\n",
       "      <th>ec50_95ci_c</th>\n",
       "      <th>ec50_pred_c</th>\n",
       "      <th>ec50_rise_c</th>\n",
       "      <th>stabilityscore_c</th>\n",
       "      <th>stabilityscore</th>\n",
       "      <th>talaris2013_score</th>\n",
       "      <th>betanov15_score</th>\n",
       "      <th>library_name</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>EEHEE_rd1_0005.pdb</td>\n",
       "      <td>GSVELEDNGRRVEATSTQEARDRAKKEGATTWTESGTRIEVRG</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.04</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.1</td>\n",
       "      <td>0.1</td>\n",
       "      <td>0.95</td>\n",
       "      <td>-0.91</td>\n",
       "      <td>-0.44</td>\n",
       "      <td>...</td>\n",
       "      <td>0.1</td>\n",
       "      <td>0.1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.56</td>\n",
       "      <td>-2.46</td>\n",
       "      <td>-1.18</td>\n",
       "      <td>-1.18</td>\n",
       "      <td>-71.705</td>\n",
       "      <td>-115.069</td>\n",
       "      <td>rd1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1 rows × 21 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                 name                                     sequence  \\\n",
       "4  EEHEE_rd1_0005.pdb  GSVELEDNGRRVEATSTQEARDRAKKEGATTWTESGTRIEVRG   \n",
       "\n",
       "   assay_library  ec50_t  ec50_95ci_lbound_t  ec50_95ci_ubound_t  ec50_95ci_t  \\\n",
       "4            1.0    0.04                 0.0                 0.1          0.1   \n",
       "\n",
       "   ec50_pred_t  ec50_rise_t  stabilityscore_t  ...  ec50_95ci_lbound_c  \\\n",
       "4         0.95        -0.91             -0.44  ...                 0.1   \n",
       "\n",
       "   ec50_95ci_ubound_c  ec50_95ci_c  ec50_pred_c  ec50_rise_c  \\\n",
       "4                 0.1          0.0         2.56        -2.46   \n",
       "\n",
       "   stabilityscore_c  stabilityscore  talaris2013_score  betanov15_score  \\\n",
       "4             -1.18           -1.18            -71.705         -115.069   \n",
       "\n",
       "   library_name  \n",
       "4           rd1  \n",
       "\n",
       "[1 rows x 21 columns]"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "stability_scores[\"rd1\"][stability_scores[\"rd1\"][\"name\"] == \"EEHEE_rd1_0005.pdb\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "# with open(NOTEBOOK_PATH.joinpath(\"stability_scores_wpreds.torch\"), \"wb\") as fout:\n",
    "#     torch.save(stability_scores_wpreds, fout)\n",
    "stability_scores_wpreds = {}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "rd1\n",
      "200\n",
      "400\n",
      "600\n",
      "800\n",
      "1000\n",
      "1200\n",
      "1400\n",
      "1600\n",
      "1800\n",
      "2000\n",
      "2200\n",
      "2400\n",
      "2600\n",
      "2800\n",
      "3000\n",
      "3200\n",
      "3400\n",
      "3600\n",
      "3800\n",
      "4000\n"
     ]
    },
    {
     "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>name</th>\n",
       "      <th>sequence</th>\n",
       "      <th>assay_library</th>\n",
       "      <th>ec50_t</th>\n",
       "      <th>ec50_95ci_lbound_t</th>\n",
       "      <th>ec50_95ci_ubound_t</th>\n",
       "      <th>ec50_95ci_t</th>\n",
       "      <th>ec50_pred_t</th>\n",
       "      <th>ec50_rise_t</th>\n",
       "      <th>stabilityscore_t</th>\n",
       "      <th>...</th>\n",
       "      <th>stabilityscore_c</th>\n",
       "      <th>stabilityscore</th>\n",
       "      <th>talaris2013_score</th>\n",
       "      <th>betanov15_score</th>\n",
       "      <th>library_name</th>\n",
       "      <th>network_score</th>\n",
       "      <th>network_score_wref</th>\n",
       "      <th>sequence_length</th>\n",
       "      <th>domain</th>\n",
       "      <th>identity_to_ref</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>EEHEE_rd1_0001.pdb</td>\n",
       "      <td>GSQEVNSGTQTYKNASPEEAERIARKAGATTWTEKGNKWEIRI</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.90</td>\n",
       "      <td>1.8</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.2</td>\n",
       "      <td>1.06</td>\n",
       "      <td>0.84</td>\n",
       "      <td>0.40</td>\n",
       "      <td>...</td>\n",
       "      <td>0.17</td>\n",
       "      <td>0.17</td>\n",
       "      <td>-78.360</td>\n",
       "      <td>-132.050</td>\n",
       "      <td>rd1</td>\n",
       "      <td>-1.987571</td>\n",
       "      <td>-1.987572</td>\n",
       "      <td>41</td>\n",
       "      <td>EEHEE</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>EEHEE_rd1_0002.pdb</td>\n",
       "      <td>GSYELEVGNYRYRADDPEQLKEEAKKVGARDVQTDGNNFKVRG</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.18</td>\n",
       "      <td>-0.5</td>\n",
       "      <td>0.7</td>\n",
       "      <td>1.2</td>\n",
       "      <td>0.48</td>\n",
       "      <td>-0.30</td>\n",
       "      <td>-0.14</td>\n",
       "      <td>...</td>\n",
       "      <td>0.65</td>\n",
       "      <td>-0.14</td>\n",
       "      <td>-78.022</td>\n",
       "      <td>-126.266</td>\n",
       "      <td>rd1</td>\n",
       "      <td>-1.901309</td>\n",
       "      <td>-1.901310</td>\n",
       "      <td>41</td>\n",
       "      <td>EEHEE</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>EEHEE_rd1_0003.pdb</td>\n",
       "      <td>GSWRVHFRGETYTADTEDDAKQLAKDAGARRIESSNGEVRVEL</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.65</td>\n",
       "      <td>1.6</td>\n",
       "      <td>1.7</td>\n",
       "      <td>0.1</td>\n",
       "      <td>1.59</td>\n",
       "      <td>0.06</td>\n",
       "      <td>0.03</td>\n",
       "      <td>...</td>\n",
       "      <td>0.06</td>\n",
       "      <td>0.03</td>\n",
       "      <td>-73.631</td>\n",
       "      <td>-129.425</td>\n",
       "      <td>rd1</td>\n",
       "      <td>-1.465720</td>\n",
       "      <td>-1.465720</td>\n",
       "      <td>41</td>\n",
       "      <td>EEHEE</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>EEHEE_rd1_0004.pdb</td>\n",
       "      <td>GSYTIEENGEKYTFRTRDEAEEWARRQGAQTVETRGTELRSRA</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.26</td>\n",
       "      <td>2.2</td>\n",
       "      <td>2.4</td>\n",
       "      <td>0.2</td>\n",
       "      <td>1.51</td>\n",
       "      <td>0.74</td>\n",
       "      <td>0.35</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.32</td>\n",
       "      <td>-0.32</td>\n",
       "      <td>-79.304</td>\n",
       "      <td>-130.277</td>\n",
       "      <td>rd1</td>\n",
       "      <td>-1.953686</td>\n",
       "      <td>-1.953686</td>\n",
       "      <td>41</td>\n",
       "      <td>EEHEE</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>EEHEE_rd1_0005.pdb</td>\n",
       "      <td>GSVELEDNGRRVEATSTQEARDRAKKEGATTWTESGTRIEVRG</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.04</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.1</td>\n",
       "      <td>0.1</td>\n",
       "      <td>0.95</td>\n",
       "      <td>-0.91</td>\n",
       "      <td>-0.44</td>\n",
       "      <td>...</td>\n",
       "      <td>-1.18</td>\n",
       "      <td>-1.18</td>\n",
       "      <td>-71.705</td>\n",
       "      <td>-115.069</td>\n",
       "      <td>rd1</td>\n",
       "      <td>-1.957716</td>\n",
       "      <td>-1.957716</td>\n",
       "      <td>41</td>\n",
       "      <td>EEHEE</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 26 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                 name                                     sequence  \\\n",
       "0  EEHEE_rd1_0001.pdb  GSQEVNSGTQTYKNASPEEAERIARKAGATTWTEKGNKWEIRI   \n",
       "1  EEHEE_rd1_0002.pdb  GSYELEVGNYRYRADDPEQLKEEAKKVGARDVQTDGNNFKVRG   \n",
       "2  EEHEE_rd1_0003.pdb  GSWRVHFRGETYTADTEDDAKQLAKDAGARRIESSNGEVRVEL   \n",
       "3  EEHEE_rd1_0004.pdb  GSYTIEENGEKYTFRTRDEAEEWARRQGAQTVETRGTELRSRA   \n",
       "4  EEHEE_rd1_0005.pdb  GSVELEDNGRRVEATSTQEARDRAKKEGATTWTESGTRIEVRG   \n",
       "\n",
       "   assay_library  ec50_t  ec50_95ci_lbound_t  ec50_95ci_ubound_t  ec50_95ci_t  \\\n",
       "0            1.0    1.90                 1.8                 2.0          0.2   \n",
       "1            1.0    0.18                -0.5                 0.7          1.2   \n",
       "2            1.0    1.65                 1.6                 1.7          0.1   \n",
       "3            1.0    2.26                 2.2                 2.4          0.2   \n",
       "4            1.0    0.04                 0.0                 0.1          0.1   \n",
       "\n",
       "   ec50_pred_t  ec50_rise_t  stabilityscore_t  ...  stabilityscore_c  \\\n",
       "0         1.06         0.84              0.40  ...              0.17   \n",
       "1         0.48        -0.30             -0.14  ...              0.65   \n",
       "2         1.59         0.06              0.03  ...              0.06   \n",
       "3         1.51         0.74              0.35  ...             -0.32   \n",
       "4         0.95        -0.91             -0.44  ...             -1.18   \n",
       "\n",
       "   stabilityscore  talaris2013_score  betanov15_score  library_name  \\\n",
       "0            0.17            -78.360         -132.050           rd1   \n",
       "1           -0.14            -78.022         -126.266           rd1   \n",
       "2            0.03            -73.631         -129.425           rd1   \n",
       "3           -0.32            -79.304         -130.277           rd1   \n",
       "4           -1.18            -71.705         -115.069           rd1   \n",
       "\n",
       "   network_score  network_score_wref  sequence_length  domain  identity_to_ref  \n",
       "0      -1.987571           -1.987572               41   EEHEE                1  \n",
       "1      -1.901309           -1.901310               41   EEHEE                1  \n",
       "2      -1.465720           -1.465720               41   EEHEE                1  \n",
       "3      -1.953686           -1.953686               41   EEHEE                1  \n",
       "4      -1.957716           -1.957716               41   EEHEE                1  \n",
       "\n",
       "[5 rows x 26 columns]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "rd2\n",
      "200\n",
      "400\n",
      "600\n",
      "800\n",
      "1000\n",
      "1200\n",
      "1400\n",
      "1600\n",
      "1800\n",
      "2000\n",
      "2200\n",
      "2400\n",
      "2600\n",
      "2800\n",
      "3000\n",
      "3200\n",
      "3400\n",
      "3600\n",
      "3800\n",
      "4000\n"
     ]
    },
    {
     "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>name</th>\n",
       "      <th>sequence</th>\n",
       "      <th>assay_library</th>\n",
       "      <th>ec50_t</th>\n",
       "      <th>ec50_95ci_lbound_t</th>\n",
       "      <th>ec50_95ci_ubound_t</th>\n",
       "      <th>ec50_95ci_t</th>\n",
       "      <th>ec50_pred_t</th>\n",
       "      <th>ec50_rise_t</th>\n",
       "      <th>stabilityscore_t</th>\n",
       "      <th>...</th>\n",
       "      <th>stabilityscore_c</th>\n",
       "      <th>stabilityscore</th>\n",
       "      <th>talaris2013_score</th>\n",
       "      <th>betanov15_score</th>\n",
       "      <th>library_name</th>\n",
       "      <th>network_score</th>\n",
       "      <th>network_score_wref</th>\n",
       "      <th>sequence_length</th>\n",
       "      <th>domain</th>\n",
       "      <th>identity_to_ref</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>EEHEE_rd2_0001.pdb</td>\n",
       "      <td>GSTTVDLDGETYKFPSREEAKKFIKKVNPTTYEIRGEEIRVQT</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.39</td>\n",
       "      <td>-0.5</td>\n",
       "      <td>7.0</td>\n",
       "      <td>7.5</td>\n",
       "      <td>1.59</td>\n",
       "      <td>-1.20</td>\n",
       "      <td>-0.57</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.16</td>\n",
       "      <td>-0.57</td>\n",
       "      <td>-88.467</td>\n",
       "      <td>-150.197</td>\n",
       "      <td>rd2</td>\n",
       "      <td>-2.013789</td>\n",
       "      <td>-2.013789</td>\n",
       "      <td>41</td>\n",
       "      <td>EEHEE</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>EEHEE_rd2_0002.pdb</td>\n",
       "      <td>GSTEVDKRGQTYRFTTKEEAREWARRQGAKEVREDGTRIEVRQ</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.84</td>\n",
       "      <td>0.6</td>\n",
       "      <td>1.1</td>\n",
       "      <td>0.5</td>\n",
       "      <td>0.54</td>\n",
       "      <td>0.31</td>\n",
       "      <td>0.15</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.13</td>\n",
       "      <td>-0.13</td>\n",
       "      <td>-85.578</td>\n",
       "      <td>-137.308</td>\n",
       "      <td>rd2</td>\n",
       "      <td>-1.888334</td>\n",
       "      <td>-1.888334</td>\n",
       "      <td>41</td>\n",
       "      <td>EEHEE</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>EEHEE_rd2_0003.pdb</td>\n",
       "      <td>GSKTLTVNGVTITSTDVTWLEKWAKKFGATRIEVRGDTIDVQS</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.43</td>\n",
       "      <td>2.4</td>\n",
       "      <td>2.4</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.49</td>\n",
       "      <td>0.94</td>\n",
       "      <td>0.45</td>\n",
       "      <td>...</td>\n",
       "      <td>0.66</td>\n",
       "      <td>0.45</td>\n",
       "      <td>-75.917</td>\n",
       "      <td>-127.802</td>\n",
       "      <td>rd2</td>\n",
       "      <td>-1.751409</td>\n",
       "      <td>-1.751409</td>\n",
       "      <td>41</td>\n",
       "      <td>EEHEE</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>EEHEE_rd2_0004.pdb</td>\n",
       "      <td>GSTTLESNGTTLTFPSEEAAREWARQYGASSIERRGTKFHVHE</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.38</td>\n",
       "      <td>0.3</td>\n",
       "      <td>0.5</td>\n",
       "      <td>0.2</td>\n",
       "      <td>1.14</td>\n",
       "      <td>-0.76</td>\n",
       "      <td>-0.36</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.26</td>\n",
       "      <td>-0.36</td>\n",
       "      <td>-80.530</td>\n",
       "      <td>-134.274</td>\n",
       "      <td>rd2</td>\n",
       "      <td>-1.887325</td>\n",
       "      <td>-1.887325</td>\n",
       "      <td>41</td>\n",
       "      <td>EEHEE</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>EEHEE_rd2_0005.pdb</td>\n",
       "      <td>GSTTLTRNGQTYTFRSRDEAKKWADKSGATTVRFDGDEIEVRE</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.94</td>\n",
       "      <td>2.9</td>\n",
       "      <td>2.9</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.19</td>\n",
       "      <td>0.75</td>\n",
       "      <td>0.36</td>\n",
       "      <td>...</td>\n",
       "      <td>0.32</td>\n",
       "      <td>0.32</td>\n",
       "      <td>-80.184</td>\n",
       "      <td>-135.376</td>\n",
       "      <td>rd2</td>\n",
       "      <td>-1.764627</td>\n",
       "      <td>-1.764627</td>\n",
       "      <td>41</td>\n",
       "      <td>EEHEE</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 26 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                 name                                     sequence  \\\n",
       "0  EEHEE_rd2_0001.pdb  GSTTVDLDGETYKFPSREEAKKFIKKVNPTTYEIRGEEIRVQT   \n",
       "1  EEHEE_rd2_0002.pdb  GSTEVDKRGQTYRFTTKEEAREWARRQGAKEVREDGTRIEVRQ   \n",
       "2  EEHEE_rd2_0003.pdb  GSKTLTVNGVTITSTDVTWLEKWAKKFGATRIEVRGDTIDVQS   \n",
       "3  EEHEE_rd2_0004.pdb  GSTTLESNGTTLTFPSEEAAREWARQYGASSIERRGTKFHVHE   \n",
       "4  EEHEE_rd2_0005.pdb  GSTTLTRNGQTYTFRSRDEAKKWADKSGATTVRFDGDEIEVRE   \n",
       "\n",
       "   assay_library  ec50_t  ec50_95ci_lbound_t  ec50_95ci_ubound_t  ec50_95ci_t  \\\n",
       "0            2.0    0.39                -0.5                 7.0          7.5   \n",
       "1            2.0    0.84                 0.6                 1.1          0.5   \n",
       "2            2.0    2.43                 2.4                 2.4          0.0   \n",
       "3            2.0    0.38                 0.3                 0.5          0.2   \n",
       "4            2.0    2.94                 2.9                 2.9          0.0   \n",
       "\n",
       "   ec50_pred_t  ec50_rise_t  stabilityscore_t  ...  stabilityscore_c  \\\n",
       "0         1.59        -1.20             -0.57  ...             -0.16   \n",
       "1         0.54         0.31              0.15  ...             -0.13   \n",
       "2         1.49         0.94              0.45  ...              0.66   \n",
       "3         1.14        -0.76             -0.36  ...             -0.26   \n",
       "4         2.19         0.75              0.36  ...              0.32   \n",
       "\n",
       "   stabilityscore  talaris2013_score  betanov15_score  library_name  \\\n",
       "0           -0.57            -88.467         -150.197           rd2   \n",
       "1           -0.13            -85.578         -137.308           rd2   \n",
       "2            0.45            -75.917         -127.802           rd2   \n",
       "3           -0.36            -80.530         -134.274           rd2   \n",
       "4            0.32            -80.184         -135.376           rd2   \n",
       "\n",
       "   network_score  network_score_wref  sequence_length  domain  identity_to_ref  \n",
       "0      -2.013789           -2.013789               41   EEHEE                1  \n",
       "1      -1.888334           -1.888334               41   EEHEE                1  \n",
       "2      -1.751409           -1.751409               41   EEHEE                1  \n",
       "3      -1.887325           -1.887325               41   EEHEE                1  \n",
       "4      -1.764627           -1.764627               41   EEHEE                1  \n",
       "\n",
       "[5 rows x 26 columns]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "rd3\n",
      "200\n",
      "400\n",
      "600\n",
      "800\n",
      "1000\n",
      "1200\n",
      "1400\n",
      "1600\n",
      "1800\n",
      "2000\n",
      "2200\n",
      "Sequence mismatch of size '0' for protein HEEH_rd3_0164.pdb!\n",
      "Sequence mismatch of size '0' for protein HEEH_rd3_0218.pdb!\n",
      "2400\n",
      "Sequence mismatch of size '0' for protein HEEH_rd3_0359.pdb!\n",
      "Sequence mismatch of size '0' for protein HEEH_rd3_0480.pdb!\n",
      "2600\n",
      "Sequence mismatch of size '0' for protein HEEH_rd3_0522.pdb!\n",
      "Sequence mismatch of size '0' for protein HEEH_rd3_0523.pdb!\n",
      "2800\n",
      "3000\n",
      "Sequence mismatch of size '0' for protein HEEH_rd3_1103.pdb!\n",
      "Sequence mismatch of size '0' for protein HEEH_rd3_1108.pdb!\n",
      "3200\n",
      "3400\n",
      "3600\n",
      "Sequence mismatch of size '0' for protein HEEH_rd3_1575.pdb!\n",
      "Sequence mismatch of size '0' for protein HEEH_rd3_1614.pdb!\n",
      "Sequence mismatch of size '0' for protein HEEH_rd3_1640.pdb!\n",
      "3800\n",
      "Sequence mismatch of size '0' for protein HEEH_rd3_1760.pdb!\n",
      "Sequence mismatch of size '0' for protein HEEH_rd3_1811.pdb!\n",
      "4000\n",
      "4200\n"
     ]
    },
    {
     "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>name</th>\n",
       "      <th>sequence</th>\n",
       "      <th>assay_library</th>\n",
       "      <th>ec50_t</th>\n",
       "      <th>ec50_95ci_lbound_t</th>\n",
       "      <th>ec50_95ci_ubound_t</th>\n",
       "      <th>ec50_95ci_t</th>\n",
       "      <th>ec50_pred_t</th>\n",
       "      <th>ec50_rise_t</th>\n",
       "      <th>stabilityscore_t</th>\n",
       "      <th>...</th>\n",
       "      <th>stabilityscore_c</th>\n",
       "      <th>stabilityscore</th>\n",
       "      <th>talaris2013_score</th>\n",
       "      <th>betanov15_score</th>\n",
       "      <th>library_name</th>\n",
       "      <th>network_score</th>\n",
       "      <th>network_score_wref</th>\n",
       "      <th>sequence_length</th>\n",
       "      <th>domain</th>\n",
       "      <th>identity_to_ref</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>EEHEE_rd2_0005.pdb</td>\n",
       "      <td>GSTTLTRNGQTYTFRSRDEAKKWADKSGATTVRFDGDEIEVRE</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.85</td>\n",
       "      <td>3.6</td>\n",
       "      <td>4.1</td>\n",
       "      <td>0.5</td>\n",
       "      <td>2.19</td>\n",
       "      <td>1.66</td>\n",
       "      <td>0.79</td>\n",
       "      <td>...</td>\n",
       "      <td>0.20</td>\n",
       "      <td>0.20</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>rd3</td>\n",
       "      <td>-1.764627</td>\n",
       "      <td>-1.764628</td>\n",
       "      <td>41.0</td>\n",
       "      <td>EEHEE</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>EEHEE_rd2_0030.pdb</td>\n",
       "      <td>GSTTVTFKGQTFTFRSDEEAERWAKKQGVTSIQKDGTTITLHG</td>\n",
       "      <td>3.0</td>\n",
       "      <td>4.14</td>\n",
       "      <td>4.1</td>\n",
       "      <td>4.1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.26</td>\n",
       "      <td>1.89</td>\n",
       "      <td>0.90</td>\n",
       "      <td>...</td>\n",
       "      <td>0.88</td>\n",
       "      <td>0.88</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>rd3</td>\n",
       "      <td>-1.937184</td>\n",
       "      <td>-1.937185</td>\n",
       "      <td>41.0</td>\n",
       "      <td>EEHEE</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>EEHEE_rd2_0320.pdb</td>\n",
       "      <td>GSKKLTWNGQTVTSTDPDELKRLAKKLGATRIEEHGDTIRVQG</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.32</td>\n",
       "      <td>2.2</td>\n",
       "      <td>2.5</td>\n",
       "      <td>0.3</td>\n",
       "      <td>1.14</td>\n",
       "      <td>1.18</td>\n",
       "      <td>0.56</td>\n",
       "      <td>...</td>\n",
       "      <td>1.01</td>\n",
       "      <td>0.56</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>rd3</td>\n",
       "      <td>-1.699440</td>\n",
       "      <td>-1.699440</td>\n",
       "      <td>41.0</td>\n",
       "      <td>EEHEE</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>EEHEE_rd2_0499.pdb</td>\n",
       "      <td>GSKTVEVNGQTYTATTSTEAKEYATRVGATTWTQDGDKFTFTG</td>\n",
       "      <td>3.0</td>\n",
       "      <td>4.90</td>\n",
       "      <td>4.9</td>\n",
       "      <td>4.9</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4.33</td>\n",
       "      <td>0.57</td>\n",
       "      <td>0.27</td>\n",
       "      <td>...</td>\n",
       "      <td>0.22</td>\n",
       "      <td>0.22</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>rd3</td>\n",
       "      <td>-1.592018</td>\n",
       "      <td>-1.592018</td>\n",
       "      <td>41.0</td>\n",
       "      <td>EEHEE</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>EEHEE_rd2_0753.pdb</td>\n",
       "      <td>GSTTLEYNGRTWTFNTTEEAKEWAKRWGATTIHVSNGTLRVQE</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.63</td>\n",
       "      <td>3.6</td>\n",
       "      <td>3.7</td>\n",
       "      <td>0.1</td>\n",
       "      <td>1.88</td>\n",
       "      <td>1.76</td>\n",
       "      <td>0.84</td>\n",
       "      <td>...</td>\n",
       "      <td>0.77</td>\n",
       "      <td>0.77</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>rd3</td>\n",
       "      <td>-2.050423</td>\n",
       "      <td>-2.050424</td>\n",
       "      <td>41.0</td>\n",
       "      <td>EEHEE</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 26 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                 name                                     sequence  \\\n",
       "0  EEHEE_rd2_0005.pdb  GSTTLTRNGQTYTFRSRDEAKKWADKSGATTVRFDGDEIEVRE   \n",
       "1  EEHEE_rd2_0030.pdb  GSTTVTFKGQTFTFRSDEEAERWAKKQGVTSIQKDGTTITLHG   \n",
       "2  EEHEE_rd2_0320.pdb  GSKKLTWNGQTVTSTDPDELKRLAKKLGATRIEEHGDTIRVQG   \n",
       "3  EEHEE_rd2_0499.pdb  GSKTVEVNGQTYTATTSTEAKEYATRVGATTWTQDGDKFTFTG   \n",
       "4  EEHEE_rd2_0753.pdb  GSTTLEYNGRTWTFNTTEEAKEWAKRWGATTIHVSNGTLRVQE   \n",
       "\n",
       "   assay_library  ec50_t  ec50_95ci_lbound_t  ec50_95ci_ubound_t  ec50_95ci_t  \\\n",
       "0            3.0    3.85                 3.6                 4.1          0.5   \n",
       "1            3.0    4.14                 4.1                 4.1          0.0   \n",
       "2            3.0    2.32                 2.2                 2.5          0.3   \n",
       "3            3.0    4.90                 4.9                 4.9          0.0   \n",
       "4            3.0    3.63                 3.6                 3.7          0.1   \n",
       "\n",
       "   ec50_pred_t  ec50_rise_t  stabilityscore_t  ...  stabilityscore_c  \\\n",
       "0         2.19         1.66              0.79  ...              0.20   \n",
       "1         2.26         1.89              0.90  ...              0.88   \n",
       "2         1.14         1.18              0.56  ...              1.01   \n",
       "3         4.33         0.57              0.27  ...              0.22   \n",
       "4         1.88         1.76              0.84  ...              0.77   \n",
       "\n",
       "   stabilityscore  talaris2013_score  betanov15_score  library_name  \\\n",
       "0            0.20                NaN              NaN           rd3   \n",
       "1            0.88                NaN              NaN           rd3   \n",
       "2            0.56                NaN              NaN           rd3   \n",
       "3            0.22                NaN              NaN           rd3   \n",
       "4            0.77                NaN              NaN           rd3   \n",
       "\n",
       "   network_score  network_score_wref  sequence_length  domain  identity_to_ref  \n",
       "0      -1.764627           -1.764628             41.0   EEHEE              1.0  \n",
       "1      -1.937184           -1.937185             41.0   EEHEE              1.0  \n",
       "2      -1.699440           -1.699440             41.0   EEHEE              1.0  \n",
       "3      -1.592018           -1.592018             41.0   EEHEE              1.0  \n",
       "4      -2.050423           -2.050424             41.0   EEHEE              1.0  \n",
       "\n",
       "[5 rows x 26 columns]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "rd4\n",
      "Do not have reference for the following domains: {'PF05433.12', '3M48', 'PF16049.2', 'PF17106.2', 'PF13396.3', '3P7K', '1ZWD', '1Z2T', '1HPH', 'PF08799.8', 'PF13137.3', 'PF11665.5', 'PF09514.7', '2YSI', 'PF02370.13', 'PF12465.5', 'PF12901.4', 'PF11002.5', '2L9A', 'PF08983.7', '2N4V', 'PF07557.8', 'PF08067.8', 'PF07589.8', '2ZAJ', 'PF14766.3', 'PF11776.5', 'PF16777.2', 'PF15696.2', 'PF05186.10', '1S4T', 'PF02184.13', '1JO6', 'PF07012.9', 'PF00376.20', '2BN6', '2LZQ', 'PF05509.8', '2MCE', 'PF07661.10', 'PF15447.3', 'PF00681.17', 'PF12426.5', '2ME2', '2KWT', '1ZDB', 'PF01585.20', 'PF16517.2', '2MDU', '2M9R', '1PV0', 'PF08104.8', '2HI5', 'PF07595.9', '2B19', 'PF16563.2', 'PF12113.5', 'PF10281.6', '2JY0', 'PF12521.5', '2L91', 'PF10525.6', 'PF13438.3', '1AIE', 'PF02987.13', 'PF12390.5', 'PF11591.5', '2L64', 'PF02513.14', '1ZWC', '2MHW', 'PF00746.18', 'PF02818.12', '2JOF', 'PF09457.7', '1FRY', '1E0G', 'PF15954.2', '1HZ3', '2LJT', '2MWA', 'PF05658.11', '1AMC', '1BH0', '2KPA', 'PF07054.8', 'PF13284.3', 'PF13056.3', '2XL1', 'PF07533.13', '2M0W', '2JR8', '1NMJ', '3V1A', '2K44', 'PF00036.29', 'PF10391.6', 'PF04650.14', 'PF06200.11', 'PF12434.5', 'PF13174.3', 'PF08485.7', 'PF12573.5', 'PF08024.8', '2JXF', '2BBL', '1IYT', 'PF08970.7', '1E0M', 'PF05725.9', 'PF13713.3', '2LCV', 'PF00741.15', '2HN8', 'PF02839.11', '2ODC', 'PF09042.8', 'PF11052.5', '1PP5', 'PF04627.10', 'PF12593.5', 'PF00401.17', '1BZG', 'PF14207.3', 'PF16751.2', 'PF10338.6', 'PF11752.5', '1IFN', 'PF12338.5', 'PF14852.3', 'PF13533.3', 'PF12075.5', 'PF13488.3', '1L1K', '1VD8', 'PF06696.8', 'PF10197.6', '2KEH', 'PF01281.16', 'PF12111.5', 'PF04680.10', 'PF02183.15', 'PF06596.8', '1DU1', 'PF13986.3', '1EDX', 'PF10945.5', '1GK7', 'PF08990.8', 'PF16741.2', 'PF08945.7', 'PF10766.6', 'PF08465.7', '2L36', '4MGP', '1T0C', '1USD', '1ODP', '2M9S', 'PF00832.17', 'PF08557.7', 'PF06565.9', 'PF09649.7', '2RMH', '1KX6', 'PF04135.9', 'PF14960.3', 'PF14141.3', '4CVD', '2GOH', 'PF06435.8', '2RQ2', '1RET', 'PF06145.8', 'PF12728.4', '2MTW', 'PF00123.17', '2LL2', 'PF15285.3', 'PF16121.2', 'PF08025.8', 'PF09604.7', 'PF12123.5', 'PF01839.20', 'PF01816.14', 'arc', 'PF00473.14', '2N4W', 'PF01290.17', 'PF08035.8', 'PF00159.15', 'PF02468.12', 'PF01469.15', '2ME3', 'PF15864.2', 'PF09853.6', 'PF00460.17', '2MG1', '1YIU', 'PF11522.5', 'PF17063.2', 'PF14990.3', 'PF10632.6', '2MWB', '1FME', 'PF02533.12', 'PF03406.10', '2K1A', 'PF03752.10', '1RIJ', 'PF12579.5', 'PF06582.9', 'PF00397.23', 'PF11953.5', 'PF10304.6', 'PF07599.8', '2N4U', 'PF09006.8', 'PF16488.2', 'PF12841.4', 'PF07835.9', 'PF06633.8', '1PGY', '2MDC', '2KDR', '1IIJ', 'PF12178.5', 'PF07208.8', '1BJC', '2L96', '1CKX', 'PF11900.5', '1XOP', '1HO0', '2LVG', 'PF11688.5', 'PF13128.3', '2MBY', 'PF09610.7', 'PF11423.5', 'PF00249.28', '2LMF', 'PF07504.10', 'PF12277.5', '1L2Y', 'PF12743.4', '1D0R', 'PF11384.5', 'PF08227.8', '2L63', 'PF05115.11', 'PF06298.8', 'PF01402.18', '1HGV', 'PF06568.8', 'PF07849.8', 'PF16300.2', 'PF03858.10', 'PF08725.8', 'PF09811.6', 'PF14146.3', '1Q2F', '2LS2', 'PF02755.12', 'PF15172.3', 'PF11784.5', '2WQ0', '1YYB', 'GCN4', 'PF00880.15', 'PF03640.12', '1WR7', '3IFM', '2MVJ', 'PF07806.8', 'PF04806.9', 'PF01296.15', 'PF02986.11', 'PF12415.5', '2MW3', 'PF00796.15', '2MWE', '2OP7', 'PF02817.14', 'PF16551.2', '1ZDA', 'PF00440.20', '1HO2', '1USE', 'PF12323.5', 'PF14038.3', 'PF14824.3', '2JNH', '1ZWB', 'PF03271.14', 'PF08513.8', '1YMZ', 'PF15805.2', 'PF08198.8', '4UZW', 'PF11464.5', '2GDL', '2KJM', 'PF05969.8', '1WR3', 'PF14379.3', '2LBG', 'PF11413.5', '2RT4', '1XC0', '2RMF', 'PF00402.15', '2MW9', '1Q2I', 'PF14195.3', 'PF09145.7', 'PF13404.3', 'PF05829.9', 'PF06594.8', 'PF10642.6', '2MU8', 'PF02796.12', 'PF02037.24', '1VDA', 'PF11542.5', '1FW5', 'PF16982.2', 'PF07706.9', 'PF08071.9', 'PF05968.8', 'PF07122.8', 'PF08476.7', 'PF14475.3', 'PF09957.6', 'PF10419.6', '2K98', 'PF12557.5', 'PF15214.3', 'PF12427.5', 'PF12949.4', '1V92', 'PF06739.8', 'PF01851.19', 'PF08038.9', 'PF12418.5', '2KE3', 'PF14244.3', '1ZR7', '2MG3', 'PF13428.3', '2KPB', 'PF09121.7', '1WN8', 'PF00614.19', '2M73', 'PF07676.9', 'PF13940.3', '2KWY', 'PF08113.8', '1MEQ', '2MMT', '5AHT', 'PF12399.5', 'PF01788.14', 'PF06451.8', '2MR9', '2M5L', 'PF17070.2', '1P9C', '1FSD', '1W4E', '2GOF', 'PF05079.9', '2JPX', '2M1A', 'PF12510.5', 'PF01465.17', 'PF07870.8', 'PF08776.8', 'PF12423.5', 'PF13070.3', '2RND', 'PF04995.11', '2RN0', 'PF02526.11', 'PF15314.3', 'PF13018.3', 'PF04419.11', '1OEI', 'PF11798.5', 'PF10516.6', '2KCF', '2DCI', 'PF04714.10', '2JPJ', 'PF10381.6', '2PV6', '1CFG', 'PF12459.5', 'PF12581.5', 'PF07492.8', '2WQ2', 'PF16320.2', 'PF00624.15', '1W4G', '1GCN', 'PF08183.8', 'PF07498.9', 'PF11661.5', '1ZRV', 'PF12551.5', 'PF07918.8', 'PF03875.10', '1HGZ', 'PF02419.14', 'PF04644.9', '2K84', 'PF12202.5', 'PF13431.3', '1VTP', 'PF03494.10', '1PI8', 'PF11616.5', '2LJQ', '1FVY', 'PF10689.6', '1DPK', 'PF02208.13', 'PF08049.8', 'PF13516.3', '2MSU', '2LE3', 'PF06308.8', '2MU7', '1SMZ', '1I6C', 'PF09047.7', 'PF03778.10', 'hYAP65', 'PF11590.5', '2BTH', 'PF01436.18', '2YSB', 'PF16423.2', '1YRF', '1HH0', 'PF08736.8', '4L5E', 'PF01744.17', 'PF12391.5', '1FSV', 'tryp', 'PF12556.5', 'PF05570.8', 'PF11601.5', '4REX', 'PF13134.3', 'PF06994.8', 'PF06392.8', '1CKZ', 'PF09060.7', 'PF14903.3', 'PF08179.9', 'PF09425.7', 'PF14406.3', 'PF00879.15', 'PF07710.8', 'PF10587.6', 'PF12765.4', 'PF12799.4', '1BH7', '2KGN', '1VD7', 'PF06671.8', '4GWT', 'PF10617.6', 'PF03911.13', 'PF01372.14', '2L79', '2LZP', 'PF00560.30', 'PF08173.8', 'PF11438.5', '2KI0', 'PF08200.8', 'PF08225.8', '1YTR', '2MLV', '1PSV', '2KBL', 'PF07142.9', '3V86', 'PF07442.8', 'PF00023.27', 'PF11632.5', 'PF11585.5', '2LQ2', 'PF13040.3', 'PF16484.2', '1ZWA', 'PF11848.5', 'PF05553.8', 'PF10948.5', 'PF02218.12', 'PF12611.5', 'PF10562.6', 'PF01011.18', 'PF07411.9', '2GP8', 'PF11401.5', '2DK1', '1E0N', 'PF13414.3', 'PF10417.6', '1SOL', 'PF13400.3', '1JJS', '2DMV', 'PF11629.5', 'PF12393.5', '1BWX', '2RRI', '1IBO', 'PF00556.17', '2LNY', '2KDT', 'PF08671.7', 'PF11426.5', 'PF03832.10', '2MU6', '1IWF', 'PF03475.11', 'PF01352.24', '1ODQ', 'pin1', '1HNS', '1HNR', '1WNK', '2LLM', '1P9I', 'PF07196.10', 'PF13153.3', 'PF16808.2', 'PF08109.8', '3HRO', '1CEK', 'PF08587.8', '1ZRJ', 'PF13576.3', '2IFO', '2RMG', 'PF11270.5', '2L0N', 'PF11883.5', '2MRW', 'PF08136.8', '1GW4', 'PF06344.8', 'PF03533.11', 'PF16237.2', '2L9V', '4APD', 'PF00805.19', '2MWD', 'PF02260.17', 'PF08383.8', 'PF09107.8', 'PF11502.5', '2K8J', 'PF03540.10', '1KDL', '1Y03', '1MOT', '2LE7', '1WR4', '2L0G', '1PSM', 'PF07538.8', '2OJN', 'PF12488.5', 'PF13080.3', 'PF07280.8', 'PF07872.8', 'PF07046.8', 'PF02285.12', 'PF12412.5', 'PF08483.8', 'PF02609.13', 'PF17046.2', '2KYA', 'PF06975.8', '2LMA', 'PF00634.15', '1WZ4', '1WAZ', '1WNN', '2N6U', 'PF08186.8', 'PF08977.7', '1CKW', '1IBN', 'PF06715.9', '2PDE', '2KHG', '2M8F', '2WQ3', 'PF03482.10', 'PF11636.5', 'PF09528.7', 'PF12121.5', 'PF09689.7', 'PF12153.5', 'PF16852.2', 'PF16833.2', 'PF02944.17', 'PF12149.5', 'PF04485.9', 'PF12360.5', 'PF05233.10', 'PF08140.8', 'PF01479.22', 'PF10872.5', 'PF07811.9', '2HFR', 'PF11532.5', '2B4N', 'PF07797.11', '1ZTA', 'PF13436.3', '4LH9', 'PF07048.8', '2B9K', 'PF06528.9', '2MKB', 'PF07004.9', 'PF08815.7', 'PF17195.1', '2RRH', '2N58', 'PF17199.1', 'PF10097.6', 'PF00467.26', 'PF07721.11', '2KYK', 'PF05552.9', 'PF04624.11', 'PF14152.3', 'PF12513.5', 'PF08802.7', '2M0S', 'PF13213.3', '2KHF', 'PF07853.8', '1ODR', '1Z0Q', '2MWT', 'PF03527.11', 'PF07440.9', 'PF17101.2', 'PF00284.17', 'PF08947.7', '2PDD', '2WQ1', '4NCU', 'PF08182.8', '2N6V', 'PF04325.10', '1E0L', 'PF08705.8', '2MMW', '1JRJ', '2N1C', 'PF04382.10', 'PF05933.10', '2ME4', 'PF01476.17', 'PF08061.8', 'PF00356.18', '2MUE', 'PF08338.8', 'PF10524.6', 'PF01505.15', 'PF06131.8', '2MMM', '2IMU', 'PF16713.2', 'PF03990.11', '2N68', '2LKE', 'PF05671.8', '1Y04', 'PF12035.5', '2K7Y', 'PF10855.5', '2LL5', 'PF09268.7', 'PF08105.8', 'PF14312.3', 'PF11627.5', 'GCN4-VNVV', 'PF12951.4', '2YSC', '2AJO', 'PF07043.10', 'PF10979.5', 'PF08213.8', '4GWV', 'PF07644.8', '2LS3', '2FCG', 'PF02197.14', 'PF08188.8', 'PF14804.3', '1PPT', '1IFY', 'PF12108.5', 'PF17162.1', 'PF08085.8', 'PF12424.5', '2MVH', 'PF13405.3', 'PF17164.1', 'PF12373.5', 'PF14013.3', 'PF13955.3', 'PF14304.3', '2CPS', 'PF12552.5', 'PF08880.8', 'PF10055.6', 'PF05906.8', '2RMY', 'PF11178.5', '1W4H', '2N4R', 'PF15363.3', 'PF14695.3', '2LUV', '2KNU', '1V1D', '2M9J', '2MV6', 'PF02189.12', 'PF07460.8', 'PF08238.9', 'PF17102.2', 'PF03701.11', 'PF14485.3', 'PF00018.25', '2LX0', 'PF08081.8', 'PF08518.8', 'PF02529.12', '2O00', '2LS9', '1P82', '2BTG', 'PF12365.5', '2MUA', '2LAT', '1CKY', 'PF02493.17', 'PF04021.9', 'PF07465.10', '1EMZ', 'PF05480.8', 'PF02209.16', 'PF03491.10', 'PF02205.17', '1AML', 'PF02180.14', 'PF14383.3', '2M8J', 'PF13441.3', '1BY6', 'PF05678.11', 'PF08102.8', '1FAC', '1HN3', 'PF02532.11', 'chymo', 'PF13217.3', 'PF14006.3', 'PF14202.3', 'PF00272.16', '2LUG', '2KZQ', '1S4W', 'PF08026.8', '1X32', '2ARI', 'PF10931.5', 'PF13653.3', 'PF09693.7', '3ZY1', '1VD9', 'PF08009.8', 'PF12453.5', '2M8M', '1MP6', 'PF12177.5', '2NR1', 'PF12344.5', 'PF01239.19', '1WNM', 'PF13939.3', '2JPK', '1OEG', 'PF06631.8', '2OBU', '1BY0', '1YWJ', '2L70', '1OEF', 'PF12198.5', '1AMB', '2JTW', 'PF07897.8', 'PF02188.14', 'PF14710.3', 'PF10509.6', 'PF12451.5', 'PF12221.5', '2RLW', 'PF09851.6', 'PF05920.8', 'PF13072.3', '2LFM', 'PF13823.3', 'PF09163.8', 'PF13215.3', '2OVC', '1PJE', 'PF07129.8', '2YSG', 'PF05545.8', 'PF05345.9', 'PF09200.7', 'PF08078.9', '2N2C', '3I1G', 'PF16086.2', '2LRX', 'PF12431.5', '1W4F', 'PF11880.5', '1SUT', '1X22', '2M8I', 'PF03742.11', 'PF06740.9', 'PF02845.13', '2KXA', '2M8O', '2M9I', '2E5U', '2MWF', 'PF14414.3', '2I9N', 'PF08133.8', 'PF13076.3', '1A11', 'PF03523.10', 'PF08509.8', 'PF12651.4', 'PF16621.2', 'PF08858.7', '1EDW', 'PF03597.12', '2KLV', '1OPP', '2MLU', 'PF09125.7', '2JPM', 'PF12407.5', '2J8P', '2AJM', 'PF11416.5', 'PF08127.10', 'PF11346.5', 'PF08041.8', 'PF11411.5', 'PF04726.10', 'PF08159.9', 'PF12515.5', 'PF11121.5', 'PF06116.9', 'PF08789.7', 'PF14115.3', '2L0O', '2M3E', 'PF12778.4', 'PF05372.8', 'PF06612.8', '2BBP', '2KJN', '1DPQ', '1PI7', 'PF09040.8', '2YSF', '1IWC', 'PF17112.2', 'PF13633.3', 'PF13368.3', '2AP7', '1WN4', '2XKM', 'PF11044.5', '2C0X', 'PF13724.3', '1QLO', '2I9O', 'PF06301.8', 'PF17041.2', '2KSG', 'PF13110.3', 'PF10708.6', 'PF01846.16', '1BBA', 'PF08621.7', '2L0L', 'PF07719.14', '1JDM', 'PF07581.9', '2AMN', 'PF05151.9', 'PF08632.7', 'PF16832.2', '1T5Q', 'PF10685.6', 'PF09282.7', '2ORU', 'PF03869.11', 'PF10346.6', 'PF04671.9', 'PF08254.8', 'PF12135.5', '1FDF', '2JUI', 'PF07558.8', '1LBJ', 'PF14084.3', 'PF14615.3', '1YRI', 'PF04394.11', 'PF05593.11', 'PF07553.8', '2L99', 'PF13544.3', 'PF11610.5', 'PF08954.8', 'PF12154.5', '2MG2', 'PF01693.13', 'PF00468.14', '1GW3', 'PF02151.16', '1ZRX', '2KEG', 'PF12197.5', 'PF15789.2', 'PF08528.8', 'PF04839.10', '2MVL', '2PPZ', '1ZWE', 'PF14657.3', 'PF02173.14', 'PF01023.16', '2E5T', 'PF07625.8', 'PF13081.3', 'PF07878.8', 'PF15448.3', 'PF16211.2', '1C26', '2AJN', '4GIF', '2L5I', '1UJL', 'PF16102.2', '2LSA', 'PF06598.8', '2M9F', 'PF08181.8', 'PF16973.2', '1LYP', '1LV4', 'PF11295.5', 'PF10518.6', '1Q71', '2LKS', 'PF13570.3', 'PF08230.8', 'PF00395.17', 'PF12779.4', 'PF03789.10', '2LTI', '2NZZ', 'PF07629.8', '2L4U', '2IFN', 'PF07966.9', 'PF12180.5', '2ME1', 'PF15018.3', '1K1V', 'PF01304.14', '2OJO', '2N16', '1BJB', 'PF17105.2', '2LQY', '2M76', 'PF13021.3', '2LSW', '1VII', '2MVM', 'PF00807.14', '2PJV', 'PF12162.5', '2K76', 'PF12478.5', '2LUF', 'PF14373.3', '1XOO', '2RMZ', '1VDB', '4HLR', 'PF10413.6', '1U0P', '2PCO', 'PF09501.7', '2LCM', '2M3F', 'PF03047.11', '2KJO', '2EKK', 'PF00515.25', '1IFP', 'PF07425.8', 'PF10968.5', '2MUF', 'PF08312.9', 'PF00325.17', 'PF07453.10', '2MUJ', 'PF03989.10', 'PF05931.8', 'PF09198.7', '1O06', 'PF06031.10', 'PF10584.6', 'PF06004.9', 'PF01701.15', 'PF07720.9', 'PF10581.6', '1CQ0', '2I2J', '1ZRW', 'PF07963.9', '1ND9', 'PF09453.7', '1VPU', 'PF08176.9', 'PF07877.8', 'PF14760.3', '2MU9', '3DWU', 'PF07896.9', 'PF08525.8', 'PF14209.3', 'PF02234.16', 'PF09303.7', 'PF15088.3', 'PF10399.6', 'PF04964.11', 'PF12359.5', '2N4T', 'PF08563.8', 'PF11433.5', '2P81', '1EZE', 'PF01779.14', '2KNS', 'PF16078.2', 'PF01679.14', '2RPW', 'PF12622.4', '1HJ0', 'PF14553.3', 'PF03889.10', '1GP8', '5E4X', 'PF08272.8', '1K91', '2MOF', 'PF04014.15', 'PF12522.5', 'PF12398.5', 'PF13253.3', '2L77', 'PF14555.3', 'PF03736.14', '2AJJ', 'PF13605.3', 'PF06376.9', 'PF09575.7', 'PF14140.3', 'PF01405.14', '4HTM', '2N4S', '2OJM', 'PF00608.14', 'PF13152.3', '1HO7', '2MVN', 'PF10389.6', 'PF05465.10', '2L0M', '2KFE', 'PF11421.5', '2JPL', 'PF00283.16', '1KV4', '1RES', 'PF02237.14', 'PF09373.7', 'villin', '2K6O', 'PF12136.5', '1CWX', '2L4J', 'PF03984.10', 'PF16772.2', 'PF07623.8', 'PF13318.3', '2L5R', '1P83', '2YSH', 'PF08946.7', 'PF11654.5', '1QG9', 'PF08062.8', '3U91', 'BBL', 'PF07549.11', 'PF14268.3', 'PF10590.6', 'PF08515.9', '2LK9', '1DX7', '2MTY', 'PF14174.3', 'PF08197.8', 'PF12073.5', '2L1X', 'p53tetS-gabeshortened', 'PF04022.9', 'PF06481.11', '4IFM', '2LQ1', 'PF08054.8', 'PF11540.5', '2B22', 'PF07499.10', '2K9D', 'PF05713.8', 'PF08177.8', '1JO5', 'PF05627.8', 'PF06543.9', 'PF10636.6', 'PF08180.8', '3MRA', '2M9E', 'PF02070.12', 'PF05792.10', 'PF08550.7', 'PF16857.2', 'PF00140.17', '4DZK', '1RKL', 'PF03455.16', 'PFRD-XC4', 'PF15116.3', '2WXC', 'PF04886.9', 'PF00612.24', 'PF16225.2', 'PF14237.3', '2L92'}!\n",
      "200\n",
      "400\n",
      "600\n",
      "800\n",
      "1000\n",
      "1200\n",
      "1400\n",
      "1600\n",
      "1800\n",
      "2000\n",
      "2200\n",
      "2400\n",
      "2600\n",
      "Sequence mismatch for protein HEEH_rd3_0218.pdb: (11.63% identical).\n",
      "2800\n",
      "Sequence mismatch for protein HEEH_rd4_0248.pdb: (13.95% identical).\n",
      "3000\n",
      "Sequence mismatch for protein HEEH_rd4_0458.pdb: (6.98% identical).\n",
      "Sequence mismatch for protein HEEH_rd4_0462.pdb: (16.28% identical).\n",
      "3200\n",
      "Sequence mismatch for protein HEEH_rd4_0543.pdb: (27.91% identical).\n",
      "3400\n",
      "Sequence mismatch for protein HEEH_rd4_0697.pdb: (20.93% identical).\n",
      "3600\n",
      "Sequence mismatch for protein HEEH_rd4_0890.pdb: (18.60% identical).\n",
      "3800\n",
      "4000\n",
      "4200\n",
      "4400\n",
      "4600\n",
      "Sequence mismatch for protein HHH_rd4_0076.pdb: (6.98% identical).\n",
      "4800\n",
      "Sequence mismatch for protein HHH_rd4_0307.pdb: (6.98% identical).\n",
      "5000\n",
      "Sequence mismatch for protein HHH_rd4_0542.pdb: (6.98% identical).\n",
      "5200\n",
      "Sequence mismatch for protein HHH_rd4_0759.pdb: (4.65% identical).\n",
      "Sequence mismatch for protein HHH_rd4_0819.pdb: (9.30% identical).\n",
      "5400\n"
     ]
    },
    {
     "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>name</th>\n",
       "      <th>sequence</th>\n",
       "      <th>assay_library</th>\n",
       "      <th>ec50_t</th>\n",
       "      <th>ec50_95ci_lbound_t</th>\n",
       "      <th>ec50_95ci_ubound_t</th>\n",
       "      <th>ec50_95ci_t</th>\n",
       "      <th>ec50_pred_t</th>\n",
       "      <th>ec50_rise_t</th>\n",
       "      <th>stabilityscore_t</th>\n",
       "      <th>...</th>\n",
       "      <th>stabilityscore_c</th>\n",
       "      <th>stabilityscore</th>\n",
       "      <th>talaris2013_score</th>\n",
       "      <th>betanov15_score</th>\n",
       "      <th>library_name</th>\n",
       "      <th>network_score</th>\n",
       "      <th>network_score_wref</th>\n",
       "      <th>sequence_length</th>\n",
       "      <th>domain</th>\n",
       "      <th>identity_to_ref</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1A11</td>\n",
       "      <td>GSEKMSTAISVLLAQAVFLLLTSQRGSGNGSGGNGSGGNGSGNGSG...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.69</td>\n",
       "      <td>2.5</td>\n",
       "      <td>2.9</td>\n",
       "      <td>0.4</td>\n",
       "      <td>3.01</td>\n",
       "      <td>-0.32</td>\n",
       "      <td>-0.15</td>\n",
       "      <td>...</td>\n",
       "      <td>0.53</td>\n",
       "      <td>-0.15</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>rd4</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1AIE</td>\n",
       "      <td>EYFTLQIRGRERFEMFRELNEALELKDAQAGSGNGSGGNGSGGNGS...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.07</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.1</td>\n",
       "      <td>0.1</td>\n",
       "      <td>0.49</td>\n",
       "      <td>3.58</td>\n",
       "      <td>1.71</td>\n",
       "      <td>...</td>\n",
       "      <td>1.60</td>\n",
       "      <td>1.60</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>rd4</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1AMB</td>\n",
       "      <td>DAEFRHDSGYEVHHQKLVFFAEDVGSNKGNGSGGNGSGGNGSGNGS...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>1.24</td>\n",
       "      <td>1.1</td>\n",
       "      <td>1.4</td>\n",
       "      <td>0.3</td>\n",
       "      <td>2.40</td>\n",
       "      <td>-1.16</td>\n",
       "      <td>-0.55</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.53</td>\n",
       "      <td>-0.55</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>rd4</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1AMC</td>\n",
       "      <td>DAEFRHDSGYEVHHQKLVFFAEDVGSNKNGSGGNGSGGNGSGNGSG...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.42</td>\n",
       "      <td>2.3</td>\n",
       "      <td>2.6</td>\n",
       "      <td>0.3</td>\n",
       "      <td>2.39</td>\n",
       "      <td>0.03</td>\n",
       "      <td>0.01</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.25</td>\n",
       "      <td>-0.25</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>rd4</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1AML</td>\n",
       "      <td>DAEFRHDSGYEVHHQKLVFFAEDVGSNKGAIIGLMVGGVVGSGGNG...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.20</td>\n",
       "      <td>2.1</td>\n",
       "      <td>2.3</td>\n",
       "      <td>0.2</td>\n",
       "      <td>2.09</td>\n",
       "      <td>0.10</td>\n",
       "      <td>0.05</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.23</td>\n",
       "      <td>-0.23</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>rd4</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 26 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   name                                           sequence  assay_library  \\\n",
       "0  1A11  GSEKMSTAISVLLAQAVFLLLTSQRGSGNGSGGNGSGGNGSGNGSG...            4.0   \n",
       "1  1AIE  EYFTLQIRGRERFEMFRELNEALELKDAQAGSGNGSGGNGSGGNGS...            4.0   \n",
       "2  1AMB  DAEFRHDSGYEVHHQKLVFFAEDVGSNKGNGSGGNGSGGNGSGNGS...            4.0   \n",
       "3  1AMC  DAEFRHDSGYEVHHQKLVFFAEDVGSNKNGSGGNGSGGNGSGNGSG...            4.0   \n",
       "4  1AML  DAEFRHDSGYEVHHQKLVFFAEDVGSNKGAIIGLMVGGVVGSGGNG...            4.0   \n",
       "\n",
       "   ec50_t  ec50_95ci_lbound_t  ec50_95ci_ubound_t  ec50_95ci_t  ec50_pred_t  \\\n",
       "0    2.69                 2.5                 2.9          0.4         3.01   \n",
       "1    4.07                 4.0                 4.1          0.1         0.49   \n",
       "2    1.24                 1.1                 1.4          0.3         2.40   \n",
       "3    2.42                 2.3                 2.6          0.3         2.39   \n",
       "4    2.20                 2.1                 2.3          0.2         2.09   \n",
       "\n",
       "   ec50_rise_t  stabilityscore_t  ...  stabilityscore_c  stabilityscore  \\\n",
       "0        -0.32             -0.15  ...              0.53           -0.15   \n",
       "1         3.58              1.71  ...              1.60            1.60   \n",
       "2        -1.16             -0.55  ...             -0.53           -0.55   \n",
       "3         0.03              0.01  ...             -0.25           -0.25   \n",
       "4         0.10              0.05  ...             -0.23           -0.23   \n",
       "\n",
       "   talaris2013_score  betanov15_score  library_name  network_score  \\\n",
       "0                NaN              NaN           rd4            NaN   \n",
       "1                NaN              NaN           rd4            NaN   \n",
       "2                NaN              NaN           rd4            NaN   \n",
       "3                NaN              NaN           rd4            NaN   \n",
       "4                NaN              NaN           rd4            NaN   \n",
       "\n",
       "   network_score_wref  sequence_length  domain  identity_to_ref  \n",
       "0                 NaN              NaN     NaN              NaN  \n",
       "1                 NaN              NaN     NaN              NaN  \n",
       "2                 NaN              NaN     NaN              NaN  \n",
       "3                 NaN              NaN     NaN              NaN  \n",
       "4                 NaN              NaN     NaN              NaN  \n",
       "\n",
       "[5 rows x 26 columns]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "for library in [\"rd1\", \"rd2\", \"rd3\", \"rd4\"]:\n",
    "    print(library)\n",
    "    if library in stability_scores_wpreds:\n",
    "        continue\n",
    "    df = apply_scores(stability_scores[library])\n",
    "    stability_scores_wpreds[library] = stability_scores[library].merge(df, on=\"name\", how=\"left\", validate=\"1:1\")\n",
    "    display(stability_scores_wpreds[library].head())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "torch.save(stability_scores_wpreds, NOTEBOOK_PATH.joinpath(\"stability_scores_for_selections.logproba.incremental.torch\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "# for library in [\"rd1\", \"rd2\", \"rd3\", \"rd4\"]:\n",
    "#     stability_scores[library][\"sum_log_prob\"], stability_scores[library][\"sequence_length\"] = np.array(\n",
    "#         [get_structure_proba(net, name) for name in stability_scores[library][\"name\"]]\n",
    "#     ).T"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Correlations"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_corr_95conf(r, num):\n",
    "    import math\n",
    "\n",
    "    stderr = 1.0 / math.sqrt(num - 3)\n",
    "    delta = 1.96 * stderr\n",
    "    lower = math.tanh(math.atanh(r) - delta)\n",
    "    upper = math.tanh(math.atanh(r) + delta)\n",
    "    return r - lower, upper - r "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "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>name</th>\n",
       "      <th>sequence</th>\n",
       "      <th>assay_library</th>\n",
       "      <th>ec50_t</th>\n",
       "      <th>ec50_95ci_lbound_t</th>\n",
       "      <th>ec50_95ci_ubound_t</th>\n",
       "      <th>ec50_95ci_t</th>\n",
       "      <th>ec50_pred_t</th>\n",
       "      <th>ec50_rise_t</th>\n",
       "      <th>stabilityscore_t</th>\n",
       "      <th>...</th>\n",
       "      <th>ec50_95ci_lbound_c</th>\n",
       "      <th>ec50_95ci_ubound_c</th>\n",
       "      <th>ec50_95ci_c</th>\n",
       "      <th>ec50_pred_c</th>\n",
       "      <th>ec50_rise_c</th>\n",
       "      <th>stabilityscore_c</th>\n",
       "      <th>stabilityscore</th>\n",
       "      <th>talaris2013_score</th>\n",
       "      <th>betanov15_score</th>\n",
       "      <th>library_name</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>4152</th>\n",
       "      <td>HHH_rd1_0995.pdb</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-90.153</td>\n",
       "      <td>-152.471</td>\n",
       "      <td>rd1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1 rows × 21 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                  name sequence  assay_library  ec50_t  ec50_95ci_lbound_t  \\\n",
       "4152  HHH_rd1_0995.pdb      NaN            NaN     NaN                 NaN   \n",
       "\n",
       "      ec50_95ci_ubound_t  ec50_95ci_t  ec50_pred_t  ec50_rise_t  \\\n",
       "4152                 NaN          NaN          NaN          NaN   \n",
       "\n",
       "      stabilityscore_t  ...  ec50_95ci_lbound_c  ec50_95ci_ubound_c  \\\n",
       "4152               NaN  ...                 NaN                 NaN   \n",
       "\n",
       "      ec50_95ci_c  ec50_pred_c  ec50_rise_c  stabilityscore_c  stabilityscore  \\\n",
       "4152          NaN          NaN          NaN               NaN             NaN   \n",
       "\n",
       "      talaris2013_score  betanov15_score  library_name  \n",
       "4152            -90.153         -152.471           rd1  \n",
       "\n",
       "[1 rows x 21 columns]"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "stability_scores[\"rd1\"][stability_scores[\"rd1\"][\"name\"] == \"HHH_rd1_0995.pdb\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "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>library</th>\n",
       "      <th>domain</th>\n",
       "      <th>feature</th>\n",
       "      <th>corr_t</th>\n",
       "      <th>pvalue_t</th>\n",
       "      <th>corr_c</th>\n",
       "      <th>pvalue_c</th>\n",
       "      <th>corr</th>\n",
       "      <th>pvalue</th>\n",
       "      <th>corr_conf_lower</th>\n",
       "      <th>corr_conf_upper</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>rd1</td>\n",
       "      <td>EEHEE</td>\n",
       "      <td>network_score</td>\n",
       "      <td>0.034046</td>\n",
       "      <td>0.254934</td>\n",
       "      <td>0.003777</td>\n",
       "      <td>0.899520</td>\n",
       "      <td>0.018060</td>\n",
       "      <td>0.546002</td>\n",
       "      <td>0.058621</td>\n",
       "      <td>0.058497</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>rd1</td>\n",
       "      <td>EEHEE</td>\n",
       "      <td>talaris2013_score</td>\n",
       "      <td>0.028906</td>\n",
       "      <td>0.333802</td>\n",
       "      <td>0.060923</td>\n",
       "      <td>0.041502</td>\n",
       "      <td>0.060169</td>\n",
       "      <td>0.044092</td>\n",
       "      <td>0.058572</td>\n",
       "      <td>0.058161</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>rd1</td>\n",
       "      <td>EEHEE</td>\n",
       "      <td>betanov15_score</td>\n",
       "      <td>-0.001259</td>\n",
       "      <td>0.966427</td>\n",
       "      <td>-0.013897</td>\n",
       "      <td>0.642224</td>\n",
       "      <td>0.000672</td>\n",
       "      <td>0.982077</td>\n",
       "      <td>0.058580</td>\n",
       "      <td>0.058575</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>rd1</td>\n",
       "      <td>EHEE</td>\n",
       "      <td>network_score</td>\n",
       "      <td>0.068902</td>\n",
       "      <td>0.031543</td>\n",
       "      <td>0.051711</td>\n",
       "      <td>0.106775</td>\n",
       "      <td>0.074328</td>\n",
       "      <td>0.020344</td>\n",
       "      <td>0.062763</td>\n",
       "      <td>0.062179</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>rd1</td>\n",
       "      <td>EHEE</td>\n",
       "      <td>talaris2013_score</td>\n",
       "      <td>-0.037468</td>\n",
       "      <td>0.242707</td>\n",
       "      <td>-0.025343</td>\n",
       "      <td>0.429496</td>\n",
       "      <td>-0.057332</td>\n",
       "      <td>0.073705</td>\n",
       "      <td>0.062385</td>\n",
       "      <td>0.062836</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  library domain            feature    corr_t  pvalue_t    corr_c  pvalue_c  \\\n",
       "0     rd1  EEHEE      network_score  0.034046  0.254934  0.003777  0.899520   \n",
       "1     rd1  EEHEE  talaris2013_score  0.028906  0.333802  0.060923  0.041502   \n",
       "2     rd1  EEHEE    betanov15_score -0.001259  0.966427 -0.013897  0.642224   \n",
       "3     rd1   EHEE      network_score  0.068902  0.031543  0.051711  0.106775   \n",
       "4     rd1   EHEE  talaris2013_score -0.037468  0.242707 -0.025343  0.429496   \n",
       "\n",
       "       corr    pvalue  corr_conf_lower  corr_conf_upper  \n",
       "0  0.018060  0.546002         0.058621         0.058497  \n",
       "1  0.060169  0.044092         0.058572         0.058161  \n",
       "2  0.000672  0.982077         0.058580         0.058575  \n",
       "3  0.074328  0.020344         0.062763         0.062179  \n",
       "4 -0.057332  0.073705         0.062385         0.062836  "
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results = []\n",
    "\n",
    "for library in [\"rd1\", \"rd2\", \"rd3\", \"rd4\"]:\n",
    "    df = (\n",
    "        stability_scores_wpreds[library]\n",
    "        .dropna(\n",
    "            subset=[\n",
    "                \"network_score\",\n",
    "                \"talaris2013_score\",\n",
    "                \"betanov15_score\",\n",
    "                \"domain\",\n",
    "                \"stabilityscore_t\",\n",
    "                \"stabilityscore_c\",\n",
    "            ]\n",
    "        )\n",
    "        .copy()\n",
    "    )\n",
    "    #     df[\"domain\"] = df[\"name\"].str.split(\"_\").str[0]\n",
    "    for domain, gp in df.groupby(\"domain\"):\n",
    "        for score in [\"network_score\", \"talaris2013_score\", \"betanov15_score\"]:\n",
    "            corr_t = stats.spearmanr(gp[score] / gp[\"sequence_length\"], gp[\"stabilityscore_t\"])\n",
    "            corr_c = stats.spearmanr(gp[score] / gp[\"sequence_length\"], gp[\"stabilityscore_c\"])\n",
    "            corr = stats.spearmanr(\n",
    "                gp[score] / gp[\"sequence_length\"], gp[[\"stabilityscore_t\", \"stabilityscore_c\"]].mean(axis=1)\n",
    "            )\n",
    "            corr_conf = get_corr_95conf(corr[0], len(gp[score]))\n",
    "            results.append(\n",
    "                (\n",
    "                    library,\n",
    "                    domain,\n",
    "                    score,\n",
    "                    corr_t[0],\n",
    "                    corr_t[1],\n",
    "                    corr_c[0],\n",
    "                    corr_c[1],\n",
    "                    corr[0],\n",
    "                    corr[1],\n",
    "                    corr_conf[0],\n",
    "                    corr_conf[1],\n",
    "                )\n",
    "            )\n",
    "\n",
    "results_df = pd.DataFrame(\n",
    "    results,\n",
    "    columns=[\n",
    "        \"library\",\n",
    "        \"domain\",\n",
    "        \"feature\",\n",
    "        \"corr_t\",\n",
    "        \"pvalue_t\",\n",
    "        \"corr_c\",\n",
    "        \"pvalue_c\",\n",
    "        \"corr\",\n",
    "        \"pvalue\",\n",
    "        \"corr_conf_lower\",\n",
    "        \"corr_conf_upper\",\n",
    "    ],\n",
    ")\n",
    "results_df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Final plot"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x720 with 20 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fg, axs = plt.subplots(4, 5, sharex=False, figsize=(12, 10))\n",
    "domains = [\"EEHEE\", \"EHEE\", \"HEEH\", \"HHH\"]\n",
    "for i, domain in enumerate(domains):\n",
    "    for j, library in enumerate([None, \"rd1\", \"rd2\", \"rd3\", \"rd4\"]):\n",
    "        ax = axs[i, j]\n",
    "        if i == 0 and j > 0:\n",
    "            ax.set_title(f\"Round {j}\")\n",
    "        if j == 0:\n",
    "            img = mpimg.imread(\n",
    "                (\n",
    "                    INPUT_PATH\n",
    "                    / \"global_analysis_of_protein_folding\"\n",
    "                    / \"aan0693_designed-PDB-files\"\n",
    "                    / \"nmr\"\n",
    "                    / f\"{domain}.png\"\n",
    "                ).as_posix()\n",
    "            )\n",
    "            ax.imshow(img)\n",
    "            ax.axis(\"off\")\n",
    "            ax.set_xlim(500, 3500)\n",
    "            ax.set_ylim(4250, -400)\n",
    "            ax.text(\n",
    "                -0.25,\n",
    "                0.5,\n",
    "                domain,\n",
    "                rotation=90,\n",
    "                fontsize=\"x-large\",\n",
    "                horizontalalignment=\"center\",\n",
    "                verticalalignment=\"center\",\n",
    "                transform=ax.transAxes,\n",
    "            )\n",
    "        else:\n",
    "            df = results_df[(results_df[\"domain\"] == domain) & (results_df[\"library\"] == library)]\n",
    "            data = df[\"corr\"].values\n",
    "            data = [data[0]] + (-data[1:]).tolist()\n",
    "            ax.bar([1, 2, 3], data, yerr=df[[\"corr_conf_upper\", \"corr_conf_lower\"]].values.T)\n",
    "            ax.set_ylim(-0.1, 0.5)\n",
    "            ax.set_xticks([1, 2, 3])\n",
    "            if j == 1:\n",
    "                ax.set_ylabel(\"Spearman correlation\")\n",
    "            ax.set_xticklabels([\"GCNN\", \"Rosetta\\n(talaris)\", \"Rosetta\\n(betanov)\"], fontsize=9)\n",
    "fg.tight_layout()\n",
    "fg.savefig(NOTEBOOK_PATH.joinpath(\"correlations_with_designs.svg\"), dpi=2000)\n",
    "fg.savefig(NOTEBOOK_PATH.joinpath(\"correlations_with_designs.pdf\"), dpi=2000)\n",
    "fg.savefig(NOTEBOOK_PATH.joinpath(\"correlations_with_designs.png\"), dpi=2000)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "ename": "KeyError",
     "evalue": "'sum_log_prob'",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mKeyError\u001b[0m                                  Traceback (most recent call last)",
      "\u001b[0;32m/home/strokach/env/lib/python3.7/site-packages/pandas/core/indexes/base.py\u001b[0m in \u001b[0;36mget_loc\u001b[0;34m(self, key, method, tolerance)\u001b[0m\n\u001b[1;32m   2645\u001b[0m             \u001b[0;32mtry\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 2646\u001b[0;31m                 \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_engine\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget_loc\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mkey\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   2647\u001b[0m             \u001b[0;32mexcept\u001b[0m \u001b[0mKeyError\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32mpandas/_libs/index.pyx\u001b[0m in \u001b[0;36mpandas._libs.index.IndexEngine.get_loc\u001b[0;34m()\u001b[0m\n",
      "\u001b[0;32mpandas/_libs/index.pyx\u001b[0m in \u001b[0;36mpandas._libs.index.IndexEngine.get_loc\u001b[0;34m()\u001b[0m\n",
      "\u001b[0;32mpandas/_libs/hashtable_class_helper.pxi\u001b[0m in \u001b[0;36mpandas._libs.hashtable.PyObjectHashTable.get_item\u001b[0;34m()\u001b[0m\n",
      "\u001b[0;32mpandas/_libs/hashtable_class_helper.pxi\u001b[0m in \u001b[0;36mpandas._libs.hashtable.PyObjectHashTable.get_item\u001b[0;34m()\u001b[0m\n",
      "\u001b[0;31mKeyError\u001b[0m: 'sum_log_prob'",
      "\nDuring handling of the above exception, another exception occurred:\n",
      "\u001b[0;31mKeyError\u001b[0m                                  Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-46-5caf3a0ccd63>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m      1\u001b[0m \u001b[0;32mfor\u001b[0m \u001b[0mlibrary_name\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mlist\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mstability_scores\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;36m4\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      2\u001b[0m     df = stability_scores[library_name][\n\u001b[0;32m----> 3\u001b[0;31m         \u001b[0;34m(\u001b[0m\u001b[0mstability_scores\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mlibrary_name\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m\"sum_log_prob\"\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0misnull\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m      4\u001b[0m         \u001b[0;34m&\u001b[0m \u001b[0;34m(\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mstability_scores\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mlibrary_name\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m\"talaris2013_score\"\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mnotnull\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      5\u001b[0m     ]\n",
      "\u001b[0;32m/home/strokach/env/lib/python3.7/site-packages/pandas/core/frame.py\u001b[0m in \u001b[0;36m__getitem__\u001b[0;34m(self, key)\u001b[0m\n\u001b[1;32m   2798\u001b[0m             \u001b[0;32mif\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcolumns\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mnlevels\u001b[0m \u001b[0;34m>\u001b[0m \u001b[0;36m1\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2799\u001b[0m                 \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_getitem_multilevel\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mkey\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 2800\u001b[0;31m             \u001b[0mindexer\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcolumns\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget_loc\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mkey\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   2801\u001b[0m             \u001b[0;32mif\u001b[0m \u001b[0mis_integer\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mindexer\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2802\u001b[0m                 \u001b[0mindexer\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0mindexer\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/home/strokach/env/lib/python3.7/site-packages/pandas/core/indexes/base.py\u001b[0m in \u001b[0;36mget_loc\u001b[0;34m(self, key, method, tolerance)\u001b[0m\n\u001b[1;32m   2646\u001b[0m                 \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_engine\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget_loc\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mkey\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2647\u001b[0m             \u001b[0;32mexcept\u001b[0m \u001b[0mKeyError\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 2648\u001b[0;31m                 \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_engine\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget_loc\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_maybe_cast_indexer\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mkey\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   2649\u001b[0m         \u001b[0mindexer\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget_indexer\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mkey\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mmethod\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mmethod\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtolerance\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mtolerance\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2650\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mindexer\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mndim\u001b[0m \u001b[0;34m>\u001b[0m \u001b[0;36m1\u001b[0m \u001b[0;32mor\u001b[0m \u001b[0mindexer\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msize\u001b[0m \u001b[0;34m>\u001b[0m \u001b[0;36m1\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32mpandas/_libs/index.pyx\u001b[0m in \u001b[0;36mpandas._libs.index.IndexEngine.get_loc\u001b[0;34m()\u001b[0m\n",
      "\u001b[0;32mpandas/_libs/index.pyx\u001b[0m in \u001b[0;36mpandas._libs.index.IndexEngine.get_loc\u001b[0;34m()\u001b[0m\n",
      "\u001b[0;32mpandas/_libs/hashtable_class_helper.pxi\u001b[0m in \u001b[0;36mpandas._libs.hashtable.PyObjectHashTable.get_item\u001b[0;34m()\u001b[0m\n",
      "\u001b[0;32mpandas/_libs/hashtable_class_helper.pxi\u001b[0m in \u001b[0;36mpandas._libs.hashtable.PyObjectHashTable.get_item\u001b[0;34m()\u001b[0m\n",
      "\u001b[0;31mKeyError\u001b[0m: 'sum_log_prob'"
     ]
    }
   ],
   "source": [
    "for library_name in list(stability_scores)[:4]:\n",
    "    df = stability_scores[library_name][\n",
    "        (stability_scores[library_name][\"sum_log_prob\"].isnull())\n",
    "        & ((stability_scores[library_name][\"talaris2013_score\"].notnull()))\n",
    "    ]\n",
    "    print(library_name, len(df))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "for library_name in list(stability_scores)[:4]:\n",
    "    df = stability_scores[library_name][\n",
    "        (stability_scores[library_name][\"sum_log_prob\"].notnull())\n",
    "        & ((stability_scores[library_name][\"talaris2013_score\"].isnull()))\n",
    "    ]\n",
    "    print(library_name, len(df))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "for library_name in list(stability_scores)[:4]:\n",
    "    df = stability_scores[library_name][\n",
    "        ((stability_scores[library_name][\"talaris2013_score\"].isnull()))\n",
    "    ]\n",
    "    print(library_name, len(df))"
   ]
  },
  {
   "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": "3.7.6"
  },
  "toc-showcode": false,
  "toc-showmarkdowntxt": false
 },
 "nbformat": 4,
 "nbformat_minor": 4
}