proteolytic_degradation.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "09deb9c4-506b-43bc-b378-be4f9b087bd2",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "import os\n",
    "import re\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "\n",
    "from scipy.stats import norm\n",
    "from scipy.stats import normaltest\n",
    "from scipy.optimize import curve_fit\n",
    "from sklearn.mixture import GaussianMixture as GMM\n",
    "from tqdm import tqdm\n",
    "from typing import Dict, List, Tuple\n",
    "from upsetplot import from_contents, UpSet, plot"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c4d7f824-9c26-4dda-a3ec-0ba1d1dd8c71",
   "metadata": {},
   "source": [
    "# Identification of HLA-independent proteolytic fragments in immunopeptidomics data\n",
    "\n",
    "This notebook is a implemention of the the method proposed by Jens Fritsche and Daniel J. Kowalewski to detect \"fake\" ligand that were generated as the result of proteolytic degradation (https://doi.org/10.1016/j.mcpro.2021.100110).\n",
    "\n",
    "In order to identify proteolytic peptides three metrics are computed:\n",
    "- protein coverage ratio\n",
    "- HLA ligand propensity\n",
    "- peptide coverage ratio\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d26adb27-3e41-4d4f-9ea5-a46635fb8d90",
   "metadata": {},
   "source": [
    "## Protein coverage ratio\n",
    "\n",
    "Protein coverage ratio for protein $P$\n",
    "\n",
    "$P = \\frac{1}{L(P)} \\sum_{p \\in P}L(p)$\n",
    "\n",
    "with peptide $p$\n",
    "\n",
    "where $L(x)$ = Number of amino acids per protein or peptide x "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "e51c6dc5-dad7-40ca-9ed3-d7623931aad0",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "def proteome_parser(file: str) -> Dict[str, str]:\n",
    "    \"\"\"\n",
    "    Read fasta-format file, returns dict of form scaffold:sequence.\n",
    "    Note: Uses only the unique identifier of each sequence, rather than the \n",
    "    entire header, for dict keys. \n",
    "    \"\"\"\n",
    "    proteome_dict = {}\n",
    "    uniprot_id = ''\n",
    "    sequence = []\n",
    "    for line in open(file):\n",
    "        if line.startswith(\">\") and uniprot_id == '':\n",
    "            uniprot_id = line.split(\"|\")[1]\n",
    "        elif line.startswith(\">\") and uniprot_id != '':\n",
    "            proteome_dict[uniprot_id] = ''.join(sequence)\n",
    "            uniprot_id = line.split('|')[1]\n",
    "            sequence = []\n",
    "        else:\n",
    "            sequence.append(line.rstrip())\n",
    "    proteome_dict[uniprot_id] = ''.join(sequence)\n",
    "    \n",
    "    return proteome_dict"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "ba9092e9-8820-46c1-923c-e210f6259ce3",
   "metadata": {},
   "outputs": [],
   "source": [
    "def compute_protein_coverage_ratio(df: pd.DataFrame, uniprot_ref_dict: Dict[str, str]) -> pd.DataFrame:\n",
    "    protein_IDs = []\n",
    "    protein_coverages = []\n",
    "    for protein, protein_group in df.groupby(\"Uniprot\"):\n",
    "        L_peps = []\n",
    "        for peptide, peptide_group in protein_group.groupby([\"sequence\"]):\n",
    "            peps = list(peptide_group[\"sequence\"])\n",
    "            aa_per_peptide = len(''.join(peps))\n",
    "            L = aa_per_peptide / len(peps)\n",
    "            L_peps.append(L)\n",
    "    \n",
    "        protein_coverage = 1 / len(uniprot_ref_dict[protein]) * sum(L_peps)\n",
    "        protein_IDs.append(protein)\n",
    "        protein_coverages.append(protein_coverage)\n",
    "    \n",
    "    protein_coverage_df = pd.DataFrame([protein_IDs,protein_coverages], index=[\"Uniprot\",\"protein_coverage_ratio\"]).T\n",
    "    return protein_coverage_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "15fbf285-6736-44e0-881a-90c4ac8b27ce",
   "metadata": {},
   "outputs": [],
   "source": [
    "# read input file\n",
    "df_peptides = pd.read_csv(\"../data/liegand_input.tsv\", sep='\\t')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "5659fcdd-86ec-4cf1-8986-73e5553a55f4",
   "metadata": {},
   "outputs": [],
   "source": [
    "# read reference proteome\n",
    "uniprot_ref_dict = proteome_parser('../data/spHUMANwoi_201014.fasta')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "65995edf-0b17-4090-9495-75201deeb9aa",
   "metadata": {},
   "outputs": [],
   "source": [
    "# split the input dataframe into timsTOF and orbitrap data\n",
    "df_orbitrap = df_peptides[df_peptides[\"Sample\"].str.contains(\"orbitrap\")]\n",
    "df_timstof = df_peptides[df_peptides[\"Sample\"].str.contains(\"timsTOF\")]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "c9a9a5df-4981-4d21-9b73-5217640b555e",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "# compute the protein coverage ratio\n",
    "#timsTOF\n",
    "timstof_protein_coverage_ratio = compute_protein_coverage_ratio(df_timstof, uniprot_ref_dict)\n",
    "\n",
    "#orbitrap\n",
    "orbitrap_protein_coverage_ratio = compute_protein_coverage_ratio(df_orbitrap, uniprot_ref_dict)\n",
    "\n",
    "# both\n",
    "all_protein_coverage_ratio = compute_protein_coverage_ratio(df_peptides, uniprot_ref_dict)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "c4bb5779-85ec-4c24-ace7-4f81e971e8a2",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.hist(timstof_protein_coverage_ratio[\"protein_coverage_ratio\"], bins=np.linspace(0.1,5), facecolor='gray', density=True);\n",
    "plt.xlabel('Protein Coverage Ratio')\n",
    "plt.ylabel('Density')\n",
    "plt.title('TimsTOF');"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "7dff6f8b-0173-46d0-801c-648a094b9cc9",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.hist(orbitrap_protein_coverage_ratio[\"protein_coverage_ratio\"], bins=np.linspace(0.1,5), facecolor='gray', density=True);\n",
    "plt.xlabel('Protein Coverage Ratio')\n",
    "plt.ylabel('Density')\n",
    "plt.title('orbitrap');"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "d217a43d-04b3-4f9f-a673-35ee317b4b3c",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_peptides_protcov = pd.merge(df_peptides, all_protein_coverage_ratio, on=\"Uniprot\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1dc4d62c-7d82-42d7-989d-21d106b8e640",
   "metadata": {},
   "source": [
    "## HLA ligand propensity\n",
    "\n",
    "HLA ligand propensity of peptide $p = \\frac{1}{card(S)}\\sum_{s \\in S}min_{h \\in H(s)}Rank(p,h)$ \n",
    "\n",
    "with:\n",
    "\n",
    "sample $s$ in dataset $S$\n",
    "\n",
    "allele $h$ in all six $H$\n",
    "\n",
    "peptide $p$\n",
    "\n",
    "where card is the cardinality (size) of the dataset S"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "d10df1fb-c75e-440d-bb7c-5842e6a39ad3",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_orbitrap = df_peptides_protcov[df_peptides_protcov[\"Sample\"].str.contains(\"orbitrap\")]\n",
    "df_timstof = df_peptides_protcov[df_peptides_protcov[\"Sample\"].str.contains(\"timsTOF\")]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "93352ea3-05dc-4cdc-9a88-2bac230d1d9b",
   "metadata": {},
   "outputs": [],
   "source": [
    "def compute_ligand_propensity(df: pd.DataFrame) -> pd.DataFrame:\n",
    "    min_ranks = {}\n",
    "    for sample, sample_group in df.groupby(\"Sample\"):\n",
    "        for peptide, peptide_group in sample_group.groupby(\"sequence\"):\n",
    "            min_rank = min(peptide_group[\"best score_netmhcpan-4.1\"])\n",
    "            if peptide not in min_ranks:\n",
    "                min_ranks.update({peptide: [min_rank]})\n",
    "            else:\n",
    "                min_ranks[peptide].append(min_rank)\n",
    "                \n",
    "    peptide_propensities = {peptide: sum(min_rank)/len(min_rank) for peptide, min_rank in min_ranks.items()}\n",
    "    peptide_propensities_df = pd.DataFrame.from_dict(peptide_propensities.items())  \\\n",
    "                                      .rename(columns={0:\"sequence\",1:\"propensity\"})\n",
    "    return peptide_propensities_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "f11522f0-ce07-48b1-b1dd-54e91e143f53",
   "metadata": {},
   "outputs": [],
   "source": [
    "timstof_ligand_propensity = compute_ligand_propensity(df_timstof)\n",
    "orbitrap_ligand_propensity = compute_ligand_propensity(df_orbitrap)\n",
    "all_ligand_propensity = compute_ligand_propensity(df_peptides_protcov)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "649a9365-a738-4ee6-86a1-51d90ce7eb09",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x16014d120>]"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "tp = plt.hist(timstof_ligand_propensity[\"propensity\"], bins=np.power(np.linspace(0.4, 1, 100), 10), facecolor='gray', density=True)\n",
    "plt.xscale('log')\n",
    "plt.xlabel('HLA ligand propensity')\n",
    "plt.ylabel('Density')\n",
    "plt.title('TimsTOF');\n",
    "\n",
    "x = tp[1][:-1]\n",
    "y = tp[0]\n",
    "\n",
    "n = len(x)\n",
    "mean = sum(x * y) / n\n",
    "sigma = sum(y * (x - mean) ** 2) / n\n",
    "\n",
    "\n",
    "def gaus(x: float, a: float, x0: float, sigma: float) -> float:\n",
    "    return a * np.exp(-(x - x0) ** 2 / (2 * sigma ** 2))\n",
    "\n",
    "popt, pcov = curve_fit(gaus, x, y, p0=[1, mean, sigma])\n",
    "\n",
    "plt.plot(x, gaus(x, *popt), 'ro:', label='fit')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "cd46e100-c4c1-436a-833a-8db0ac8d55d0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "lp = plt.hist(orbitrap_ligand_propensity[\"propensity\"], bins=np.power(np.linspace(0.4,1, 100), 10), facecolor='gray',density=True)\n",
    "plt.xscale('log')\n",
    "plt.xlabel('HLA ligand propensity')\n",
    "plt.ylabel('Density')\n",
    "plt.title('orbitrap');"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "3a13fbee-ac06-4056-b6f2-ac82d2999493",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "ip = plt.hist(all_ligand_propensity[\"propensity\"], bins=np.power(np.linspace(0.4,1, 100), 10), facecolor='gray',density=True)\n",
    "plt.xscale('log')\n",
    "plt.xlabel('HLA ligand propensity')\n",
    "plt.ylabel('Density')\n",
    "plt.title('TimsTOF + orbitrap');"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "dabf9759-5716-43bf-9aeb-ede3652474b2",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_peptides_protcov_propensity = pd.merge(df_peptides_protcov, all_ligand_propensity, on=\"sequence\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bde5d1bf-3aaa-4c9f-9a16-9804286d34e5",
   "metadata": {},
   "source": [
    "## Peptide coverage ratio\n",
    "\n",
    "peptide coverage ratio for peptide $p = \\frac{1}{card(S)} \\sum_{s \\in S}  \\left(  \\frac{1}{L(P)} \\sum_{\\textit{q overlaps with p in s}} L(q) \\right)$ \n",
    "\n",
    "with:\n",
    "\n",
    "$q$ overlaps\n",
    "\n",
    "where card is the cardinality (size) of the dataset S and sample $s$ in the dataset\n",
    "\n",
    "where $L(x)$ = Number of amino acids per protein or peptide x "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "1bd705b6-c3d5-4356-bf4b-3ff0d6d15204",
   "metadata": {},
   "outputs": [],
   "source": [
    "def generate_uniprot_pept_dict(df: pd.DataFrame, sample_keyword: str) -> Dict[str, Dict[str, List]]:\n",
    "    \"\"\"\n",
    "    In order to fascilate the computation of the peptide coverage ratio, the information about all the peptides in one samples originating from the sample protein is required.\n",
    "    This function generates a hierarchical dictionary with the uniprot ID on the first level and the sample IDs on the second level:\n",
    "    \n",
    "    Example:\n",
    "    {'uniprot_id_1': {'sample_id_1': ['peptide_seq_1', 'peptide_seq_2', ... , peptide_seq_x],\n",
    "                      'sample_id_2': ['peptide_seq_3', 'peptide_seq_4', ... , peptide_seq_y],\n",
    "                      ...\n",
    "                      'sample_id_i': ['peptide_seq_5', 'peptide_seq_6', ... , peptide_seq_z]},              \n",
    "     'uniprot_id_2': {'sample_id_1': ['peptide_seq_1', 'peptide_seq_2', ... , peptide_seq_x],\n",
    "                      'sample_id_2': ['peptide_seq_3', 'peptide_seq_4', ... , peptide_seq_y],\n",
    "                      ...\n",
    "                      'sample_id_i': ['peptide_seq_5', 'peptide_seq_6', ... , peptide_seq_z]},\n",
    "    ...\n",
    "    'uniprot_id_xxx': {...}}\n",
    "    \n",
    "    \"\"\"\n",
    "    \n",
    "    uniprot_sample_peptide_dict = {}\n",
    "    for sequence, uniprot_id, sample  in zip(df['sequence'], df['Uniprot'], df['Sample']):\n",
    "        if sample_keyword in sample:\n",
    "            if uniprot_id not in uniprot_sample_peptide_dict.keys():\n",
    "                uniprot_sample_peptide_dict[uniprot_id] = {}\n",
    "            \n",
    "            if sample not in uniprot_sample_peptide_dict[uniprot_id].keys():\n",
    "                uniprot_sample_peptide_dict[uniprot_id][sample] = []\n",
    "            uniprot_sample_peptide_dict[uniprot_id][sample].append(sequence)\n",
    "     \n",
    "    return uniprot_sample_peptide_dict"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "8de5a9d8-fc25-4046-a281-1a12be9ef182",
   "metadata": {},
   "outputs": [],
   "source": [
    "def compute_peptide_coverage(peptide_list: List[str], protein_seq: str, protein_name: str) ->  Dict[str, float]:\n",
    "    \"\"\"\n",
    "    To compute the overlaps of peptides at each position in the protein, first the sum of the coverage of each postion is calculated.\n",
    "    Then for each peptide the sum of the coverage values of its mapping positions in the protein is taken and devided by peptide length.\n",
    "    \n",
    "    If there are no overlaps of a peptide with another one, this will result in a score of 1.\n",
    "    \n",
    "    \"\"\" \n",
    "    peptide_coverage_dict = {}\n",
    "    protein_freq_array = [0] * len(protein_seq) # a zero array is generated representing a the protein. Each position in the array holds information about the mapping coverage the respective position.\n",
    "    \n",
    "    all_peptides_matches = []\n",
    "    for peptide in peptide_list:\n",
    "        petides_matches = [m.start() for m in re.finditer('(?=' + peptide + ')', protein_seq)] # find match position of peptide in protein sequence\n",
    "        all_peptides_matches.append(petides_matches)\n",
    "        if len(petides_matches) >= 2: # if a peptide matches multiple time in a protein, it is excluded. In the case repetitive sequences, the score can be heavily skewed (e.g. EEEEEEEE)  \n",
    "            pass\n",
    "        for match_idx in petides_matches:\n",
    "            for i in range(match_idx, len(peptide) + match_idx):\n",
    "                protein_freq_array[i] += 1\n",
    "    \n",
    "    for peptide, petides_matches in zip(peptide_list, all_peptides_matches):\n",
    "        all_overlaps = []\n",
    "        for match_idx in [petides_matches[0]]: # for each start index (multimappers in the same protein) -currently only the first hit is taken\n",
    "            overlaps = 0\n",
    "            for i in range(match_idx, len(peptide) + match_idx):\n",
    "                overlaps += protein_freq_array[i]\n",
    "            all_overlaps.append(overlaps)\n",
    "        \n",
    "        max_overlaps = max(all_overlaps) # if there is a peptide mapping to multiple positions in the same protein, the highest number of overlaps are taken\n",
    "            \n",
    "        sample_peptide_coverage = max_overlaps / len(peptide)\n",
    "        peptide_coverage_dict[peptide] = sample_peptide_coverage\n",
    "    \n",
    "    return peptide_coverage_dict"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "d38f9f9d-d348-4c52-a0a8-64f64217bd68",
   "metadata": {},
   "outputs": [],
   "source": [
    "#peptide_coverage_all_samples\n",
    "def average_coverage_ratio(peptide_coverage_all_samples: List[Dict[str, float]], peptide_sample_count: Dict[str, int], uniprot_id: str) -> Dict[Tuple[str, str], float]:\n",
    "    \"\"\"\n",
    "    computes the average peptide coverage ratio\n",
    "    \"\"\"\n",
    "    overlap_coverage_sums = {}\n",
    "    for sample_peptide_coverage in peptide_coverage_all_samples:\n",
    "        for sequence, coverage in sample_peptide_coverage.items():\n",
    "            if sequence not in overlap_coverage_sums.keys():\n",
    "                overlap_coverage_sums[sequence] = 0\n",
    "            overlap_coverage_sums[sequence] += coverage\n",
    "\n",
    "    overlap_coverage_ratios = {}\n",
    "    for sequence, coverage_sum in overlap_coverage_sums.items():\n",
    "        overlap_coverage_ratios[(uniprot_id, sequence)] = coverage_sum / peptide_sample_count[sequence]\n",
    "        \n",
    "    return overlap_coverage_ratios"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "10f4ba75-194c-4abc-923f-5351d29ff675",
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_prot_pep_coverage_ratio(all_uniprot_entries: Dict[str, List[str]], uniprot_sample_peptide_dict: Dict[str, Dict[str, List[str]]]) -> pd.DataFrame:\n",
    "    \"\"\"\n",
    "    main function for calculating the peptide coverage ratio, calls helper functions\n",
    "    \"\"\"                                                                                                     \n",
    "    protein_peptide_coverage_ratio = {}\n",
    "    for uniprot_id, samples_dict in tqdm(uniprot_sample_peptide_dict.items()):\n",
    "        protein_sequence = all_uniprot_entries[uniprot_id]\n",
    "        peptide_coverage_all_samples = []\n",
    "\n",
    "        # count samples containing each peptides\n",
    "        peptide_sample_count = {} # \"peptide_seq\" : number of samples in which this peptide was measured\n",
    "\n",
    "        for sample_name, peptide_list in samples_dict.items():\n",
    "            peptide_list = list(set(peptide_list))\n",
    "            peptide_coverage_dict = compute_peptide_coverage(peptide_list, protein_sequence, uniprot_id)\n",
    "            peptide_coverage_all_samples.append(peptide_coverage_dict)\n",
    "\n",
    "            for sequence in set(peptide_list):\n",
    "                if sequence not in peptide_sample_count.keys():\n",
    "                    peptide_sample_count[sequence] = 0\n",
    "                peptide_sample_count[sequence] += 1\n",
    "\n",
    "        coverage_ratio_protein = average_coverage_ratio(peptide_coverage_all_samples, peptide_sample_count, uniprot_id)\n",
    "\n",
    "        protein_peptide_coverage_ratio |= coverage_ratio_protein\n",
    "    \n",
    "    data = []\n",
    "    for protein_sequence_tuple, peptide_coverage_ratio in protein_peptide_coverage_ratio.items():\n",
    "        uniprot = protein_sequence_tuple[0]\n",
    "        sequence = protein_sequence_tuple[1]\n",
    "        data.append([uniprot, sequence, peptide_coverage_ratio])\n",
    "    df = pd.DataFrame(data=data, columns=[\"Uniprot\", \"sequence\", \"peptide_coverage_ratio\"])\n",
    "    \n",
    "    return df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "d05dfb2b-6fbb-4520-89ad-3ba7515aff9f",
   "metadata": {},
   "outputs": [],
   "source": [
    "# prepare nested dict for computation of the score\n",
    "# Dict[str, Dict[str, List]]\n",
    "uniprot_sample_peptide_dict_all = generate_uniprot_pept_dict(df_peptides_protcov_propensity, \"\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "f9adaefd-a61e-4a3e-9aa1-3f7d11e5c84e",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 12197/12197 [00:06<00:00, 2017.79it/s]\n"
     ]
    }
   ],
   "source": [
    "# compute the peptide coverage ratio\n",
    "protein_peptide_coverage_ratio_all = get_prot_pep_coverage_ratio(uniprot_ref_dict, uniprot_sample_peptide_dict_all)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "bcf18c33-30fa-4d07-bebc-cb55e59f06aa",
   "metadata": {},
   "outputs": [],
   "source": [
    "# merge the peptide coverage to the results dataframe\n",
    "df_peptides_protcov_propensity_peptcov = pd.merge(df_peptides_protcov_propensity, protein_peptide_coverage_ratio_all, on=[\"Uniprot\", \"sequence\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "8feccf85-641c-4f9e-8c53-09e571185364",
   "metadata": {},
   "outputs": [],
   "source": [
    "# write file to csv\n",
    "df_peptides_protcov_propensity_peptcov.to_csv(\"../data/test_out.csv\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6a2f5c57-80e5-4256-a7e1-8a8730de61f4",
   "metadata": {},
   "source": [
    "# Expectation maximization\n",
    "This approach can be used to distinguish two mixture gaussian distributions from each other that partially overlap. The discrimimation of those two distributions can be done by the EM algorithm, which performs a soft clustering approach maximizing the log-likelihood that the discriminanted clusters belong to the right cluster\n",
    "\n",
    "See also: https://towardsdatascience.com/gaussian-mixture-models-explained-6986aaf5a95"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d176171b-6e5e-4de0-8588-c7d0a391e085",
   "metadata": {},
   "source": [
    "### Threshold finding for the peptide coverage ratio score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "062b785a-d317-445d-ae5b-83a37ea28210",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_peptcov_em = df_peptides_protcov_propensity_peptcov[[\"sequence\", \"Uniprot\", \"peptide_coverage_ratio\"]]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "3c518f63-2ce4-444d-aafe-4e487c129aac",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "all_coverage_ratio_values = [x for x in df_peptcov_em['peptide_coverage_ratio'] if x != 1.0] # drop peptides without any overlaps"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "615f030a-f758-427a-893c-33607c822b78",
   "metadata": {},
   "outputs": [],
   "source": [
    "all_coverage_ratio_values_reshaped = np.array(all_coverage_ratio_values).reshape(-1, 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "851e2834-d2e5-4812-bd83-fb28adc1e3a0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=(10, 10))\n",
    "\n",
    "n_optimal = 2\n",
    "\n",
    "# create GMM model object\n",
    "gmm = GMM(n_components = n_optimal, max_iter=1000, random_state=100, covariance_type = 'full')\n",
    "\n",
    "# find useful parameters\n",
    "mean = gmm.fit(all_coverage_ratio_values_reshaped).means_  \n",
    "covs  = gmm.fit(all_coverage_ratio_values_reshaped).covariances_\n",
    "weights = gmm.fit(all_coverage_ratio_values_reshaped).weights_\n",
    "\n",
    "# create necessary things to plot\n",
    "x_axis = np.arange(-1, 15, 0.001)\n",
    "y_axis0 = norm.pdf(x_axis, float(mean[0][0]), np.sqrt(float(covs[0][0][0])))*weights[0] # 1st gaussian\n",
    "y_axis1 = norm.pdf(x_axis, float(mean[1][0]), np.sqrt(float(covs[1][0][0])))*weights[1] # 2nd gaussian\n",
    "\n",
    "plt.hist(all_coverage_ratio_values_reshaped, density=True, color='black', bins=np.arange(1.0, 15, 0.2))\n",
    "plt.plot(x_axis, y_axis0, lw=3, c='C0', ls='dashed')\n",
    "plt.plot(x_axis, y_axis1, lw=3, c='C1')\n",
    "plt.xlim(1, 15)\n",
    "plt.xlabel(r\"X\", fontsize=20)\n",
    "plt.ylabel(r\"Density\", fontsize=20)\n",
    "\n",
    "plt.subplots_adjust(wspace=0.3)\n",
    "plt.show()\n",
    "plt.close('all')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "84f1e6e9-d10c-4109-b4d7-04136b64cea2",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Threshold value peptide coverage ratio: 4.310000000000005\n"
     ]
    }
   ],
   "source": [
    "fdr = 0.05\n",
    "\n",
    "y_true_sum = 0\n",
    "y_false_sum = 0\n",
    "left_probabilities = []\n",
    "x_values = []\n",
    "\n",
    "threshold = 1 - fdr\n",
    "previous_over_threshold = False\n",
    "\n",
    "for y_true, y_false, x in zip(y_axis0, y_axis1, x_axis):\n",
    "    y_true_sum += y_true\n",
    "    y_false_sum += y_false\n",
    "    \n",
    "    if (y_true_sum + y_false_sum) != 0:\n",
    "        if x >= 1:\n",
    "            left_probability = y_true_sum / (y_true_sum + y_false_sum)\n",
    "            left_probabilities.append(y_true_sum / (y_true_sum + y_false_sum))\n",
    "            x_values.append(x)\n",
    "            \n",
    "            if previous_over_threshold and threshold > left_probability:\n",
    "                pept_cov_threshold = x\n",
    "                print(f'Threshold value peptide coverage ratio: {pept_cov_threshold}')\n",
    "            \n",
    "            if left_probability >= threshold:\n",
    "                previous_over_threshold = True\n",
    "            else:\n",
    "                previous_over_threshold = False"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "d3438acd-db1f-4e49-8c7e-f45309ee1d51",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x161298df0>]"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(x_values, left_probabilities)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "5dd97087-635e-4255-8deb-210b921da29c",
   "metadata": {},
   "outputs": [],
   "source": [
    "def estimation_maximation(coverage_ratio: List[float], fdr: float) -> float: \n",
    "    x = np.asarray(coverage_ratio).reshape(-1, 1)\n",
    "\n",
    "    gmm = GMM(n_components = 2, max_iter=500, random_state=10, covariance_type = 'full')\n",
    "\n",
    "    # find useful parameters\n",
    "    mean = gmm.fit(x).means_  \n",
    "    covs  = gmm.fit(x).covariances_\n",
    "    weights = gmm.fit(x).weights_\n",
    "\n",
    "    # create necessary things to plot\n",
    "    x_axis = np.arange(min(coverage_ratio), max(coverage_ratio), 0.00001)\n",
    "    y_axis0 = norm.pdf(x_axis, float(mean[0][0]), np.sqrt(float(covs[0][0][0])))*weights[0] # 1st gaussian\n",
    "    y_axis1 = norm.pdf(x_axis, float(mean[1][0]), np.sqrt(float(covs[1][0][0])))*weights[1] # 2nd gaussian\n",
    "\n",
    "    # Plot 2\n",
    "    plt.figure(figsize=(10, 8))\n",
    "    plt.hist(x, density=True, color='black', bins=np.linspace(0.1,2,500))\n",
    "    plt.plot(x_axis, y_axis0, lw=3, c='C0')\n",
    "    plt.plot(x_axis, y_axis1, lw=3, c='C1')\n",
    "    plt.plot(x_axis, y_axis0+y_axis1, lw=3, c='C2', ls='dashed')\n",
    "    plt.xlim(0, 4)\n",
    "\n",
    "    plt.xlabel(r\"X\", fontsize=20)\n",
    "    plt.ylabel(r\"Density\", fontsize=20)\n",
    "\n",
    "    plt.show()\n",
    "    plt.close('all')\n",
    "    \n",
    "    y_true_sum = 0\n",
    "    y_false_sum = 0\n",
    "    left_probability = []\n",
    "    x_values = []\n",
    "    for y_true, y_false, x in zip(y_axis0, y_axis1, x_axis):\n",
    "        y_true_sum += y_true\n",
    "        y_false_sum += y_false\n",
    "    \n",
    "        if (y_true_sum + y_false_sum) != 0:\n",
    "            left_probability.append(y_true_sum / (y_true_sum + y_false_sum))\n",
    "            x_values.append(x)\n",
    "\n",
    "    plt.plot(x_values, left_probability)\n",
    "    \n",
    "    y_max = 0\n",
    "    x_max = 0\n",
    "    for x, y in zip(x_values, left_probability):\n",
    "        if 1 - y > 0.05:\n",
    "            break\n",
    "        else:\n",
    "            y_max = y\n",
    "            x_max = x\n",
    "    threshold = x_max\n",
    "    print(f'Threshold: {x_max}')\n",
    "    return x_max"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e3378deb-1ed3-4c5f-8282-3185fccd6142",
   "metadata": {},
   "source": [
    "### Protein coverage ratio"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "a5289f0f-3fa0-4007-9b7a-3fb466e93109",
   "metadata": {},
   "outputs": [],
   "source": [
    "fdr = 0.05 # false discovery rate"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "b5e4d224-cc0c-40ca-b7bb-3d4dc78a0da3",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Threshold: 2.5447527480916095\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "threshold_protein_coverage_ratio = estimation_maximation(df_peptides_protcov_propensity_peptcov[\"protein_coverage_ratio\"], fdr)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "10a8eb81-126c-4609-b256-2ef2ac2d21b1",
   "metadata": {},
   "source": [
    "### Propensity"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "d0a77210-0546-4ad8-aa4e-4cb922932d98",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Threshold: 11.705029999999999\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "threshold_propensity = estimation_maximation(df_peptides_protcov_propensity_peptcov[\"propensity\"], fdr)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d10139ab-5f66-4ce7-9c72-eed48e9b85ce",
   "metadata": {},
   "source": [
    "## Proteolytic Contamination Count (PCC)\n",
    "\n",
    "$ \n",
    "\\textit{PCC} = \n",
    "\\left\\{ \n",
    "  \\begin{array}{ c l }\n",
    "    1 & \\quad \\textrm{if } \\textrm{protein coverage ratio} > \\textrm{threshold value} \\\\\n",
    "    0                 & \\quad \\textrm{otherwise}\n",
    "  \\end{array}\n",
    "\\right\\}\n",
    "+\n",
    "\\left\\{ \n",
    "  \\begin{array}{ c l }\n",
    "    1 & \\quad \\textrm{if } \\textrm{peptide coverage ratio} > \\textrm{threshold value} \\\\\n",
    "    0                 & \\quad \\textrm{otherwise}\n",
    "  \\end{array}\n",
    "\\right\\}\n",
    "+\n",
    "\\left\\{ \n",
    "  \\begin{array}{ c l }\n",
    "    1 & \\quad \\textrm{if } \\textrm{HLA ligand propensity} > \\textrm{threshold value} \\\\\n",
    "    0                 & \\quad \\textrm{otherwise}\n",
    "  \\end{array}\n",
    "\\right\\}\n",
    "$"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "77598267-344e-4b57-8694-4e9093ae264f",
   "metadata": {},
   "outputs": [],
   "source": [
    "def predict_proteolytic_peptides(row):\n",
    "    return sum([row[\"pcc_prot_cov\"], row[\"pcc_pept_cov\"], row[\"pcc_prop\"]]) >= 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "9c4b9d18-74a7-437d-b321-d953598b3e16",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_peptides_protcov_propensity_peptcov[\"pcc_prot_cov\"] = np.where(df_peptides_protcov_propensity_peptcov['protein_coverage_ratio'] > threshold_protein_coverage_ratio, 1, 0)\n",
    "df_peptides_protcov_propensity_peptcov[\"pcc_pept_cov\"] = np.where(df_peptides_protcov_propensity_peptcov['peptide_coverage_ratio'] > pept_cov_threshold, 1, 0)\n",
    "df_peptides_protcov_propensity_peptcov[\"pcc_prop\"] = np.where(df_peptides_protcov_propensity_peptcov['propensity'] > threshold_propensity, 1, 0)\n",
    "df_peptides_protcov_propensity_peptcov[\"is_proteolytic\"] = df_peptides_protcov_propensity_peptcov.apply(predict_proteolytic_peptides, axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "af04306b-3373-48b4-a11f-1383eac89f51",
   "metadata": {},
   "outputs": [],
   "source": [
    "timstof_liegands = df_peptides_protcov_propensity_peptcov[df_peptides_protcov_propensity_peptcov[\"Sample\"].str.contains(\"timsTOF\")]\n",
    "orbitrap_liegands = df_peptides_protcov_propensity_peptcov[df_peptides_protcov_propensity_peptcov[\"Sample\"].str.contains(\"orbitrap\")]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "3aba15bc-e6b1-4c90-a064-6bafdbc73885",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "labels = ['timsTOF', 'orbitrap']\n",
    "proteolytic  = [sum(timstof_liegands[\"is_proteolytic\"]), sum(orbitrap_liegands[\"is_proteolytic\"])]\n",
    "not_proteolytic = [sum(timstof_liegands[\"is_proteolytic\"] == False), sum(orbitrap_liegands[\"is_proteolytic\"] == False)]\n",
    "\n",
    "width = 0.35   \n",
    "fig, ax = plt.subplots(figsize=(10,10))\n",
    "\n",
    "ax.bar(labels, not_proteolytic, width, label='not_proteolytic')\n",
    "ax.bar(labels, proteolytic, width, bottom=not_proteolytic,\n",
    "       label='proteolytic')\n",
    "\n",
    "ax.set_ylabel('Peptide count')\n",
    "ax.set_title('Proteolytic peptide discovery')\n",
    "ax.legend()\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "id": "b8c92ef1-4c0e-4260-a9f1-b405c69c4e70",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "timsTOF proteolytic: 1.9651943072044833 %\n"
     ]
    }
   ],
   "source": [
    "print(\"timsTOF proteolytic: {0} %\".format((proteolytic[0] / (proteolytic[0] + not_proteolytic[0])) * 100)) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "98d7a068-5fe2-4b8c-a1fc-15ba035f228c",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "orbitrap proteolytic: 0.869966752862948 %\n"
     ]
    }
   ],
   "source": [
    "print(\"orbitrap proteolytic: {0} %\".format((proteolytic[1] / (proteolytic[1] + not_proteolytic[1])) * 100)) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "72a87e6f-58c8-4f6b-83a2-cadc57ed6854",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x576 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=(8, 8))\n",
    "\n",
    "pcc = from_contents({'Protein coverage ratio': timstof_liegands[timstof_liegands['pcc_prot_cov'] == 1].index.tolist(),\n",
    "                     'Peptide coverage ratio': timstof_liegands[timstof_liegands['pcc_pept_cov'] == 1].index.tolist(),\n",
    "                     'HLA ligand propensity': timstof_liegands[timstof_liegands['pcc_prop'] == 1].index.tolist()\n",
    "                    })\n",
    "\n",
    "plot(pcc, fig=fig, element_size=None, show_counts=True, sort_by='cardinality')\n",
    "plt.suptitle('timsTOF - PCC categories')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "id": "0a9d4d82-e629-49da-a486-b4018c542da4",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x576 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=(8, 8))\n",
    "\n",
    "pcc = from_contents({'Protein coverage ratio': orbitrap_liegands[orbitrap_liegands['pcc_prot_cov'] == 1].index.tolist(),\n",
    "                     'Peptide coverage ratio': orbitrap_liegands[orbitrap_liegands['pcc_pept_cov'] == 1].index.tolist(),\n",
    "                     'HLA ligand propensity': orbitrap_liegands[orbitrap_liegands['pcc_prop'] == 1].index.tolist()\n",
    "                    })\n",
    "\n",
    "plot(pcc, fig=fig, element_size=None, show_counts=True, sort_by='cardinality')\n",
    "plt.suptitle('orbitrap - PCC categories')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "id": "17b14fba-dec1-49f9-bce2-4d21bee2fb7e",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_peptides_protcov_propensity_peptcov.to_csv(\"../data/proteolytic_degradation_output_061222.tsv\", sep=\"\\t\", index=False)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.10.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}