ensemble_analysis.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c4a7d6ba-4b34-4b2e-b46c-141e6945919e",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "from utils.pdb import tmscore, pdb_to_npy\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import tqdm\n",
    "from Bio import pairwise2\n",
    "from scipy.spatial.transform import Rotation as R\n",
    "from scipy.stats import pearsonr, spearmanr, kendalltau, linregress"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "813cc8cd-52ec-4c79-aa77-14d9fa479c73",
   "metadata": {},
   "source": [
    "## Paths"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "71b484b0-afa6-4241-845b-067723d1249c",
   "metadata": {},
   "outputs": [],
   "source": [
    "PDBS_DIR = './structures'\n",
    "CODNAS_SAMPLES_DIR = './pretrained_model/codnas.csv.ep7.num5.step0.5.alpha1.0.beta3.0.elbo0.2'\n",
    "APO_SAMPLES_DIR = './pretrained_model/apo.csv.ep7.num5.step0.5.alpha1.0.beta3.0.elbo0.2'"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "79918d08-fe0b-4900-af60-677b25fcec22",
   "metadata": {},
   "source": [
    "## Load structures"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "c0a90afe-bd77-4a34-bf38-a2dac76605b9",
   "metadata": {},
   "outputs": [],
   "source": [
    "codnas_df = pd.read_csv('splits/codnas.csv', index_col='name').sort_index()\n",
    "apo_df = pd.read_csv('splits/apo.csv', index_col='name').sort_index()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "770725d9-bf78-4516-b300-0455777fadf5",
   "metadata": {},
   "outputs": [],
   "source": [
    "codnas_csv = pd.read_csv(CODNAS_SAMPLES_DIR + '.csv');\n",
    "codnas_csv['name'] = [path.split('/')[-1] for path in codnas_csv.path]; codnas_csv = codnas_csv.set_index('name')\n",
    "apo_csv = pd.read_csv(APO_SAMPLES_DIR + '.csv');\n",
    "apo_csv['name'] = [path.split('/')[-1] for path in apo_csv.path]; apo_csv = apo_csv.set_index('name')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "86fbddf1-c45d-46e4-a61b-2e7abe426ced",
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_structures(path1, path2, seq):\n",
    "    struct1, seq1 = pdb_to_npy(path1)\n",
    "    align1 = pairwise2.align.globalxx(seq, seq1)[0]\n",
    "    struct1_fixed = np.ones((len(seq), 3)) * np.nan\n",
    "    i, j = 0, 0\n",
    "    for c, d in zip(align1.seqA, align1.seqB):\n",
    "        if c != '-' and d != '-': struct1_fixed[i] = struct1[j, 0]\n",
    "        if c != '-': i+=1\n",
    "        if d != '-': j+=1\n",
    "\n",
    "    struct2, seq2 = pdb_to_npy(path2)\n",
    "    align2 = pairwise2.align.globalxx(seq, seq2)[0]\n",
    "    struct2_fixed = np.ones((len(seq), 3)) * np.nan\n",
    "    i, j = 0, 0\n",
    "    for c, d in zip(align2.seqA, align2.seqB):\n",
    "        if c != '-' and d != '-': struct2_fixed[i] = struct2[j, 0]\n",
    "        if c != '-': i+=1\n",
    "        if d != '-': j+=1\n",
    "    \n",
    "    struct1 = struct1_fixed - np.nanmean(struct1_fixed, 0)\n",
    "    struct2 = struct2_fixed - np.nanmean(struct2_fixed, 0)\n",
    "    mask = ~np.isnan(struct1[:,0]) & ~np.isnan(struct2[:,0])\n",
    "    \n",
    "    rot = R.align_vectors(struct1[mask], struct2[mask])[0].as_matrix()\n",
    "    struct2 = struct2 @ rot.T\n",
    "    \n",
    "    return struct1, struct2"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4d743092-9ca4-40e4-aca7-53ae62d91b2b",
   "metadata": {},
   "source": [
    "## Analyze CODNAS (fold-switching) pairs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "d595406e-ddcc-4e93-9d98-a2278f1b9f76",
   "metadata": {},
   "outputs": [],
   "source": [
    "def analyze(pairs_df, samples_csv, samples_dir):\n",
    "    tmpair = []\n",
    "    tm1max_list = []\n",
    "    tm2max_list = []\n",
    "    ensvar_list = []\n",
    "\n",
    "    rmsd_dict = {}\n",
    "    rmsf_dict = {}\n",
    "\n",
    "    for i, (name1, row) in tqdm.tqdm(enumerate(pairs_df.iterrows()), total=len(pairs_df)):\n",
    "        tm1, tm2 = [], []\n",
    "        seq = row.seqres\n",
    "        try: name2 = row.holo\n",
    "        except: name2 = row.other\n",
    "        path1 = f'{PDBS_DIR}/{name1[:2]}/{name1}'\n",
    "        path2 = f'{PDBS_DIR}/{name2[:2]}/{name2}'\n",
    "\n",
    "        subdf = samples_csv[samples_csv.index == name1]\n",
    "        ensvar = []\n",
    "\n",
    "        struct1, struct2 = get_structures(path1, path2, seq)\n",
    "        rmsd_dict[name1] = np.square(struct1-struct2).sum(-1)**0.5\n",
    "        rmsf_dict[name1] = []\n",
    "\n",
    "        for j in range(len(subdf)):\n",
    "            path3 = f\"{samples_dir}/{name1}.{j}.pdb\"\n",
    "            res = tmscore(path1, path3, lddt=False)\n",
    "            tm1.append(res['tm'])\n",
    "            res = tmscore(path2, path3, lddt=False)\n",
    "            tm2.append(res['tm'])\n",
    "\n",
    "            for k in range(j+1, len(subdf)):\n",
    "                path4 = f\"{samples_dir}/{name1}.{k}.pdb\"\n",
    "                res = tmscore(path4, path3, lddt=False)\n",
    "                ensvar.append(res['tm'])\n",
    "                struct1, struct2 = get_structures(path3, path4, seq)\n",
    "                rmsf_dict[name1].append(np.square(struct1-struct2).sum(-1)**0.5)\n",
    "\n",
    "        rmsf_dict[name1] = np.mean(np.stack(rmsf_dict[name1])**2, 0)**0.5\n",
    "\n",
    "        ensvar_list.append(np.mean(ensvar))\n",
    "\n",
    "        tm1 = np.array(tm1); tm2 = np.array(tm2)\n",
    "        tm1max_list.append(tm1.max()); tm2max_list.append(tm2.max())\n",
    "        res = tmscore(path1, path2, lddt=False)\n",
    "        vline = res['tm']\n",
    "\n",
    "        res = tmscore(path2, path1, lddt=False)\n",
    "        hline = res['tm']\n",
    "\n",
    "        tmpair.append((hline+vline)/2)\n",
    "    pairs_df['tm1max'] = tm1max_list\n",
    "    pairs_df['tm2max'] = tm2max_list\n",
    "    pairs_df['ensvar'] = ensvar_list\n",
    "    pairs_df['tmpair'] = tmpair\n",
    "\n",
    "    return rmsd_dict, rmsf_dict"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "a8b1679d-be7c-454e-9a63-d3f7f6f315ea",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|████████████████████████████████████████████████████████| 77/77 [02:31<00:00,  1.97s/it]\n"
     ]
    }
   ],
   "source": [
    "rmsd_dict, rmsf_dict = analyze(codnas_df, codnas_csv, CODNAS_SAMPLES_DIR)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "88bbca38-1aa8-49fc-938e-ec9ebec8f38e",
   "metadata": {},
   "outputs": [],
   "source": [
    "def getmask(name):\n",
    "    return ~np.isnan(rmsd_dict[name]) & ~np.isnan(rmsf_dict[name])\n",
    "\n",
    "correls = {name: {\n",
    "        'pearson': pearsonr(rmsd_dict[name][getmask(name)], rmsf_dict[name][getmask(name)])[0],\n",
    "        'spearman': spearmanr(rmsd_dict[name][getmask(name)], rmsf_dict[name][getmask(name)]).correlation,\n",
    "        'kendall': kendalltau(rmsd_dict[name][getmask(name)], rmsf_dict[name][getmask(name)]).correlation\n",
    "    } for name in codnas_df.index}\n",
    "codnas_df = codnas_df.join(pd.DataFrame(correls).T)\n",
    "\n",
    "global_rmsd = np.concatenate(list(rmsd_dict.values()))\n",
    "global_rmsf = np.concatenate(list(rmsf_dict.values()))\n",
    "mask = ~np.isnan(global_rmsd) & ~np.isnan(global_rmsd)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "31f28afd-2f26-4077-b080-da4ccf57b9d8",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "TM correlation 0.36\n",
      "RMSD global correlation 0.23\n",
      "RMSD per-target mean 0.28 median 0.26\n"
     ]
    }
   ],
   "source": [
    "print('TM correlation', round(pearsonr(codnas_df.ensvar, codnas_df.tmpair)[0], 2))\n",
    "print('RMSD global correlation', round(pearsonr(global_rmsd[mask], global_rmsf[mask])[0], 2))\n",
    "print('RMSD per-target mean', round(codnas_df.pearson.mean(), 2), 'median', round(codnas_df.pearson.median(), 2))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "063477a7-f3d0-411e-9bc7-763eed3a0b0b",
   "metadata": {},
   "source": [
    "## Analyze apo/holo pairs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "f01d0a63-076b-428d-ad7e-993296290be4",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|████████████████████████████████████████████████████████| 90/90 [02:27<00:00,  1.64s/it]\n"
     ]
    }
   ],
   "source": [
    "rmsd_dict, rmsf_dict = analyze(apo_df, apo_csv, APO_SAMPLES_DIR)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "8c571b75-f812-4f89-9194-f4e9a63c6a2d",
   "metadata": {},
   "outputs": [],
   "source": [
    "def getmask(name):\n",
    "    return ~np.isnan(rmsd_dict[name]) & ~np.isnan(rmsf_dict[name])\n",
    "\n",
    "correls = {name: {\n",
    "        'pearson': pearsonr(rmsd_dict[name][getmask(name)], rmsf_dict[name][getmask(name)])[0],\n",
    "        'spearman': spearmanr(rmsd_dict[name][getmask(name)], rmsf_dict[name][getmask(name)]).correlation,\n",
    "        'kendall': kendalltau(rmsd_dict[name][getmask(name)], rmsf_dict[name][getmask(name)]).correlation\n",
    "    } for name in apo_df.index}\n",
    "apo_df = apo_df.join(pd.DataFrame(correls).T)\n",
    "\n",
    "global_rmsd = np.concatenate(list(rmsd_dict.values()))\n",
    "global_rmsf = np.concatenate(list(rmsf_dict.values()))\n",
    "mask = ~np.isnan(global_rmsd) & ~np.isnan(global_rmsd)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "ca520be9-9878-41cc-ab84-87bc4d2a544d",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "TM correlation 0.12\n",
      "RMSD global correlation 0.13\n",
      "RMSD per-target mean 0.41 median 0.4\n"
     ]
    }
   ],
   "source": [
    "print('TM correlation', round(pearsonr(apo_df.ensvar, apo_df.tmpair)[0], 2))\n",
    "print('RMSD global correlation', round(pearsonr(global_rmsd[mask], global_rmsf[mask])[0], 2))\n",
    "print('RMSD per-target mean', round(apo_df.pearson.mean(), 2), 'median', round(apo_df.pearson.median(), 2))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "69b1f959-432f-4872-9c80-746d26ba2d09",
   "metadata": {},
   "source": [
    "## Plots"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "b9f3b58b-2331-4568-83dd-6acbd338a3db",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 600x450 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(4,3), dpi=150)\n",
    "plt.scatter(codnas_df.tmpair, (codnas_df.tm1max+codnas_df.tm2max)/2, s=3, label='Fold-switch')\n",
    "plt.scatter(apo_df.tmpair, (apo_df.tm1max+apo_df.tm2max)/2, s=3, label='Apo/holo')\n",
    "plt.xlim(0, 1); plt.ylim(0, 1); plt.legend()\n",
    "plt.xlabel('TM$_\\mathregular{conf1/conf 2}$')\n",
    "plt.ylabel('TM$_{ens}$')\n",
    "plt.plot((0,1),(0.5,1), c='gray')\n",
    "plt.tight_layout()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "16d2f31a-dddb-4f1c-9fd9-3bc7716571e5",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 600x450 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(4,3), dpi=150)\n",
    "plt.scatter(codnas_df.tmpair, codnas_df.ensvar, s=3, label='Fold-switch')\n",
    "plt.scatter(apo_df.tmpair, apo_df.ensvar, s=3, label='Apo/holo')\n",
    "plt.xlim(0, 1); plt.ylim(0, 1);\n",
    "plt.xlabel('TM$_\\mathregular{conf1/conf 2}$')\n",
    "plt.ylabel('Sample diversity (TM$_{var}$)')\n",
    "reg = linregress(codnas_df.tmpair, codnas_df.ensvar)\n",
    "plt.plot((0, 1), (reg.intercept, reg.slope+reg.intercept))\n",
    "reg = linregress(apo_df.tmpair, apo_df.ensvar)\n",
    "plt.plot((0, 1), (reg.intercept, reg.slope+reg.intercept))\n",
    "plt.legend(loc='lower right')\n",
    "plt.tight_layout()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "08838446-dc78-420c-9fc3-5676e7f41002",
   "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": "3.10.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}