grafimo

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Searching modulated *br* TFBS on DM3 genome using GRAFIMO"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In this notebook we will provide a brief showcase of a complete GRAFIMO analysis, starting from genome Variation Graph (VG) construction to VG scan and simple results analysis.\n",
    "\n",
    "We begin our example by constructing the VG using the genetic variants dataset provided in the shared Google drive folder. To construct the VG we will use the dedicated GRAFIMO functionality. We, then, continue by scanning the previously created VGs searching for the occurrences of *br* (JASPAR ID MA0010.1) TFBS. We analyze the obtained results to recover how many binding sites are created, disrupted and modulated by the genetic variants used to construct the VG."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "from typing import List, Tuple\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import pandas as pd  \n",
    "import numpy as np\n",
    "\n",
    "import time\n",
    "import sys\n",
    "import os"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# define path to data directories\n",
    "data = \"data\"  # data root directory\n",
    "genome = os.path.join(data, \"genome\")  # genome directory\n",
    "vcf = os.path.join(data, \"vcf\")  # genetic variants directory\n",
    "peaks = os.path.join(data, \"peaks\")  # ATAC-seq peaks directory"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Genome Variation Graph (VG) data structure construction"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We begin our toy analysis by building the genome Variation Graph for *drosophila melanogaster* (dm3 genome version). To construct the VG data structure, we need the reference genome sequence stored in a FASTA file and the genetic variants to enrich the linear reference sequence, stored in a VCF file. \n",
    "\n",
    "This operation can be done directly using the VG software suite, or employing the functionality offered by GRAFIMO called ```buildvg```. While constructing the genome variation graph data structure with VG requires several precise steps and commands to be run on the Shell, GRAFIMO provides a simple and intuitive command. Moreover, GRAFIMO's ```buildvg``` command offers many different options to the user, such as constructing the VG for a subset of chromosomes.\n",
    "\n",
    "Once built the VGs, to allow efficient and fast queries of the data structure, the graphs are indexed (```*.xg``` files). Similarly the ```*.gbwt``` files are computed to keep track of the haplotypes (paths) encoded in the graph and count how many samples show the binding site during the TFBS scanning procedure.\n",
    "\n",
    "Let's now build the VG with GRAFIMO and store the resulting ```*.xg``` and ```*.gbwt``` files within the directory ```vg``` in ```data``` directory. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "# construct drosophila VGs\n",
    "vg = os.path.join(data, \"vg\")\n",
    "if not os.path.isdir(vg):\n",
    "    os.mkdir(vg)  # create ```vg``` directory\n",
    "\n",
    "# construct VGs starting from d.melanogaster reference, custom variants and store XGs and GBWTs in ```vg``` directory\n",
    "!grafimo buildvg -l {os.path.join(genome, \"dm3.fa\")} -v {os.path.join(vcf, \"Dsec.to.dm3.RNA.ATAC.filtered.vcf.gz\")} -o {vg}"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "vscode": {
     "languageId": "plaintext"
    }
   },
   "source": [
    "Now we should have built the genome variation graph for each drosophila chromosome. We recall that the VGs are indexed in the ```*.xg``` format and the haplotype information are stored within the ```*.gbwt``` files.\n",
    "\n",
    "Now that we built the VGs we can scan them to find potential TFBS occurrences and assess the consequences that genetic variants could have on TF binding affinities. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Scanning VG data structure to recover potential TFBS occurrences in reference and alternative genomes"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To sacn the previously computed genome Variation Graph, we need a TFBS motif PWM and a set of genomic regions where to look for potential occurrences of the investigated binding site.\n",
    "\n",
    "To scan already computed VGs, GRAFIMO provides the ```findmotif``` functionality that scans the given genome Variation Graph for the occurrences of the input TFBS motif. While scanning the graph, GRAFIMO search for TFBS occurrences within all the genomes encoded in the graph. Importantly, GRAFIMO exploits the information stored within the ```GBWT``` files, traversing only the paths in the graph corresponding to the haplotypes observed in the ```VCF``` used to construct the scanned VG. Therefore, GRAFIMO reports potential binding sites which are actually observed in the input data and does not create recombinants, unless explicitely required by the user, while calling grafimo through the command line.\n",
    "\n",
    "In our toy example, let's scan dm3 VG for the occurrences of *br* TFBS (JASPAR ID [MA0010.1](https://jaspar.genereg.net/matrix/MA0010.1/)) within the shared ATAC-seq peak regions.\n",
    "\n",
    "Let's begin by downloading the previously metioned TF motif from JASPAR database."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "--2022-06-29 16:17:16--  https://jaspar.genereg.net/api/v1/matrix/MA0010.1.meme\n",
      "Resolving jaspar.genereg.net (jaspar.genereg.net)... 193.60.222.202\n",
      "Connecting to jaspar.genereg.net (jaspar.genereg.net)|193.60.222.202|:443... connected.\n",
      "HTTP request sent, awaiting response... 200 OK\n",
      "Length: 789 [text/meme]\n",
      "Saving to: ‘MA0010.1.meme’\n",
      "\n",
      "MA0010.1.meme       100%[===================>]     789  --.-KB/s    in 0s      \n",
      "\n",
      "2022-06-29 16:17:16 (212 MB/s) - ‘MA0010.1.meme’ saved [789/789]\n",
      "\n"
     ]
    }
   ],
   "source": [
    "# download br TFBS motif from JASPAR\n",
    "motif = os.path.join(data, \"motif\")\n",
    "if not os.path.isdir(motif):\n",
    "    os.mkdir(motif)\n",
    "!wget https://jaspar.genereg.net/api/v1/matrix/MA0010.1.meme\n",
    "assert os.stat(\"MA0010.1.meme\").st_size > 0  # assess if file is empty\n",
    "!mv MA0010.1.meme {motif}"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now, we have all the data required to scan the genome Variation Graph we search the potential occurrences of *br* binding site.\n",
    "\n",
    "Since in our toy example we are interested in assessing the potential consequences of genetic variation on TFBS binding affinity, we skip the thresholding procedure on statistical significance. In fact, since genetic variants could create/disrupt TFBS not observed/observed in the reference sequence, we keep all potential TFBS motif occurremces, in order to capture the effects of genetic variation. Therefore, we use a threshold of 1 on the *P*-value. Othrewise, if we were just interested in recovering the potential motif occurrences within the sequences of the genomes encoded in the input VG, we would have used the threshold on *P*-values to filter the non significant motif occurrences, similarly to other tools working on the linear reference genome, like FIMO.\n",
    "\n",
    "Let's now run GRAFIMO to search *br* motif occurrences on the previously computed VG. We will store the results in a directory called ```results```.\n",
    "\n",
    "For the sake of simplicity, in our example we computed only the TSV report and skipped the GFF3 and HTML reports creation."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "***************************************************************************\n",
      "\n",
      "\tWelcome to GRAFIMO v1.1.5\n",
      "\n",
      "***************************************************************************\n",
      "\n",
      "Read 1 motifs in data/motif/MA0010.1.meme\n",
      "\n",
      "Processing motifs\n",
      "\n",
      "\n",
      "Extracting regions defined in data/peaks/reference.peakSet.10-12hATAC.names.bed.\n",
      "\n",
      "Progress: [==================================================] 100.0% Complete\n",
      "Scoring hits for motif +MA0010.1.\n",
      "Scoring hits for motif -MA0010.1.\n",
      "\n",
      "Progress: [==================================================] 100.0% Complete\n",
      "\n",
      "Computing q-values...\n",
      "\n",
      "Scanned sequences:\t31176531\n",
      "Scanned nucleotides:\t436471434\n",
      "\n",
      "Writing results in results.\n",
      "\n",
      "Elapsed time 648.95s.\n",
      "\u001b[0m"
     ]
    }
   ],
   "source": [
    "results = \"results\"\n",
    "if not os.path.isdir(results):\n",
    "    os.mkdir(results)\n",
    "# run grafimo\n",
    "!grafimo findmotif -d {vg} -b {os.path.join(peaks, \"reference.peakSet.10-12hATAC.names.bed\")} -m {os.path.join(motif, \"MA0010.1.meme\")} --chroms-prefix-find chr -o {results} -t 1 --debug"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Assessing genetic variants effects on *br* binding sites."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "GRAFIMO results are reported in three files (stored in output directory):\n",
    "\n",
    "- tab-delimited report (TSV report)\n",
    "- HTML report\n",
    "- GFF3 report\n",
    "\n",
    "The TSV report contains all the statistically significant potential motif occurrence found by GRAFIMO (according to the applied threshold). Each retrieved motif occurrence has a log-likelihood score, a P-value, a q-value, its DNA sequence, a flag value stating if a sequence is part of the reference or has been found in the haplotypes and the number of haplotype sequences where the motif candidate sequence occurs. An example of TSV report is the following.\n",
    "\n",
    "This report can be easily processed for a downstream analysis.\n",
    "\n",
    "The HTML report has the same content of the TSV, but it can be loaded and viewed on the most commonly used web browsers.\n",
    "\n",
    "The GFF3 report can be loaded on the UCSC Genome Browser as a custom track. For example, this allows a fast linking between the genomic variants used to build the VG and those present in annotated databases like dbSNP or ClinVar."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "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>motif_id</th>\n",
       "      <th>motif_alt_id</th>\n",
       "      <th>sequence_name</th>\n",
       "      <th>start</th>\n",
       "      <th>stop</th>\n",
       "      <th>strand</th>\n",
       "      <th>score</th>\n",
       "      <th>p-value</th>\n",
       "      <th>q-value</th>\n",
       "      <th>matched_sequence</th>\n",
       "      <th>haplotype_frequency</th>\n",
       "      <th>reference</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>MA0010.1</td>\n",
       "      <td>br</td>\n",
       "      <td>chr3L:8333115-8333767</td>\n",
       "      <td>8333346</td>\n",
       "      <td>8333360</td>\n",
       "      <td>+</td>\n",
       "      <td>18.594595</td>\n",
       "      <td>7.450581e-09</td>\n",
       "      <td>0.139370</td>\n",
       "      <td>GTAATAAACAAATC</td>\n",
       "      <td>2</td>\n",
       "      <td>ref</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>MA0010.1</td>\n",
       "      <td>br</td>\n",
       "      <td>chr2R:17367986-17368592</td>\n",
       "      <td>17368077</td>\n",
       "      <td>17368091</td>\n",
       "      <td>+</td>\n",
       "      <td>18.270270</td>\n",
       "      <td>1.490116e-08</td>\n",
       "      <td>0.139370</td>\n",
       "      <td>GTAATAAACAAAAC</td>\n",
       "      <td>2</td>\n",
       "      <td>ref</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>MA0010.1</td>\n",
       "      <td>br</td>\n",
       "      <td>chr2R:7873123-7873580</td>\n",
       "      <td>7873526</td>\n",
       "      <td>7873540</td>\n",
       "      <td>+</td>\n",
       "      <td>18.189189</td>\n",
       "      <td>2.235174e-08</td>\n",
       "      <td>0.139370</td>\n",
       "      <td>ATAATAAACAAATC</td>\n",
       "      <td>4</td>\n",
       "      <td>ref</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>MA0010.1</td>\n",
       "      <td>br</td>\n",
       "      <td>chr2L:21756732-21758293</td>\n",
       "      <td>21757363</td>\n",
       "      <td>21757377</td>\n",
       "      <td>+</td>\n",
       "      <td>18.189189</td>\n",
       "      <td>2.235174e-08</td>\n",
       "      <td>0.139370</td>\n",
       "      <td>ATAATAAACAAATC</td>\n",
       "      <td>2</td>\n",
       "      <td>ref</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>MA0010.1</td>\n",
       "      <td>br</td>\n",
       "      <td>chr3L:4349944-4350445</td>\n",
       "      <td>4350200</td>\n",
       "      <td>4350186</td>\n",
       "      <td>-</td>\n",
       "      <td>18.009009</td>\n",
       "      <td>4.470348e-08</td>\n",
       "      <td>0.143027</td>\n",
       "      <td>GTAATAAACAAATG</td>\n",
       "      <td>2</td>\n",
       "      <td>ref</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   motif_id motif_alt_id            sequence_name     start      stop strand  \\\n",
       "0  MA0010.1           br    chr3L:8333115-8333767   8333346   8333360      +   \n",
       "1  MA0010.1           br  chr2R:17367986-17368592  17368077  17368091      +   \n",
       "2  MA0010.1           br    chr2R:7873123-7873580   7873526   7873540      +   \n",
       "3  MA0010.1           br  chr2L:21756732-21758293  21757363  21757377      +   \n",
       "4  MA0010.1           br    chr3L:4349944-4350445   4350200   4350186      -   \n",
       "\n",
       "       score       p-value   q-value matched_sequence  haplotype_frequency  \\\n",
       "0  18.594595  7.450581e-09  0.139370   GTAATAAACAAATC                    2   \n",
       "1  18.270270  1.490116e-08  0.139370   GTAATAAACAAAAC                    2   \n",
       "2  18.189189  2.235174e-08  0.139370   ATAATAAACAAATC                    4   \n",
       "3  18.189189  2.235174e-08  0.139370   ATAATAAACAAATC                    2   \n",
       "4  18.009009  4.470348e-08  0.143027   GTAATAAACAAATG                    2   \n",
       "\n",
       "  reference  \n",
       "0       ref  \n",
       "1       ref  \n",
       "2       ref  \n",
       "3       ref  \n",
       "4       ref  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results = \"results\"\n",
    "report = pd.read_csv(\n",
    "    os.path.join(results, \"grafimo_out.tsv\"), sep=\"\\t\", index_col=0\n",
    ")\n",
    "report.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now that we computed our results we can assess how many potential binding sites for *br* factor show their binding affinity modulated (enhanced or weakened) by genetic variation, or have been disrupted or created by the presence of mutations.\n",
    "\n",
    "To perform such analysis we require the scores for all the potential binding site occurrences, recovered by removing the threshold on statistical significance.\n",
    "\n",
    "To show how genetic variation modulates binding affinity scores in TFBS, we use pie charts."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's begin by defining some functions that will be used throughout the analysis."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_coords(seqname: str, start: int, stop: int, strand: str) -> str:\n",
    "    chrom = seqname.split(\":\")[0]\n",
    "    coord = \"_\".join([chrom, start, stop, strand])\n",
    "    return coord"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "def assign_tfbs_class(df: pd.DataFrame) -> Tuple:\n",
    "    colnames = df.columns.tolist()\n",
    "    assert \"p-value\" in colnames\n",
    "    assert \"coords\" in colnames\n",
    "    # sort results by p-value\n",
    "    df = df.sort_values(\"p-value\", ascending=True)\n",
    "    df.reset_index(drop=True, inplace=True)\n",
    "    coords_set = set(df[\"coords\"].tolist())  # recover TFBS coordinates\n",
    "    hap_bs = []  # used to study population specific binding sites\n",
    "    ref_bs = []  # used to find disrupted binding sites\n",
    "    enha_bs = []\n",
    "    weak_bs = []\n",
    "    only_ref = 0\n",
    "    only_hap = 0\n",
    "    hap_enha = 0\n",
    "    hap_weak = 0\n",
    "    start = time.time()\n",
    "    print(f\"Classifying {df.shape[0]} motif occurrences...\")\n",
    "    for c in coords_set:\n",
    "        wdf = df[df[\"coords\"] == c]\n",
    "        wdf.reset_index(drop=True, inplace=True)\n",
    "        if np.all(np.array(wdf.reference) == \"ref\"):\n",
    "            only_ref += 1\n",
    "            ref_bs.append(c)\n",
    "        elif np.all(np.array(wdf.reference) == \"non.ref\"):\n",
    "            only_hap += 1\n",
    "            hap_bs.append(c)\n",
    "        elif (\n",
    "            np.any(np.array(wdf.reference) == \"ref\") and \n",
    "            np.any(np.array(wdf.reference) == \"non.ref\")\n",
    "        ):\n",
    "            ref_rank = np.where(np.array(wdf.reference) == \"ref\")[0][0]\n",
    "            if ref_rank == 0: \n",
    "                hap_weak += 1  # ref has highest binding affinity\n",
    "                weak_bs.append(c)\n",
    "            else:  # there are occurrences with higher \n",
    "                enhanced = np.array(wdf.iloc[:(ref_rank),11].tolist())\n",
    "                if np.all(enhanced == \"non.ref\"): \n",
    "                    hap_enha += 1\n",
    "                    enha_bs.append(c)\n",
    "                else:\n",
    "                    raise ValueError(\n",
    "                        f\"A problem occurred classifying binding site at {c}.\"\n",
    "                    )\n",
    "        else:\n",
    "            raise ValueError(\n",
    "                        f\"A problem occurred classifying binding site at {c}.\"\n",
    "                    )  \n",
    "    stop = time.time()\n",
    "    print(\"Classified %d motif occurrences in:\\t%.2fs\" % (len(df), (stop - start)))\n",
    "    results = [only_ref, only_hap, hap_enha, hap_weak]\n",
    "    return results, ref_bs, hap_bs, enha_bs, weak_bs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_disrupted_tfbs(df: pd.DataFrame, bs_list: List) -> Tuple:\n",
    "    colnames = df.columns.tolist()\n",
    "    assert \"p-value\" in colnames\n",
    "    assert \"coords\" in colnames\n",
    "    assert \"reference\" in colnames\n",
    "    coords_list = df[\"coords\"].tolist()\n",
    "    pvalue_list = df[\"p-value\"].tolist()\n",
    "    ref_list = df['reference'].tolist() \n",
    "    assert len(coords_list) == len(pvalue_list)\n",
    "    assert len(coords_list) == len(ref_list)\n",
    "    coords_set = set(coords_list)\n",
    "    disrupted_ref_bs = 0\n",
    "    disr_bs = []\n",
    "    bs_dict = {}\n",
    "    for c in coords_set: bs_dict.update({c:{\"p-value\":[], \"ref\":[]}})        \n",
    "    start = time.time()\n",
    "    print(f\"Analyzing {len(bs_list)} binding sites...\")        \n",
    "    for i in range(len(df)):\n",
    "        bs_dict[coords_list[i]][\"p-value\"].append(pvalue_list[i])\n",
    "        bs_dict[coords_list[i]][\"ref\"].append(ref_list[i])\n",
    "    for bs in bs_list:  # make sure each binding site appears exactly once in the list\n",
    "        pvals = np.array(bs_dict[bs][\"p-value\"])\n",
    "        references = np.array(bs_dict[bs][\"ref\"])\n",
    "        if np.any(references == \"non.ref\"):\n",
    "            notsig_idxs = np.where(pvals >= 1e-4)[0]\n",
    "            for i in notsig_idxs:  # found at least one motif occurrence with variants non significant\n",
    "                if references[i] == \"non.ref\": \n",
    "                    disrupted_ref_bs += 1\n",
    "                    disr_bs.append(c)\n",
    "                    break\n",
    "    stop = time.time()\n",
    "    print(\"Analyzed %d binding sites in:\\t%.2fs\" % (len(set(bs_list)), (stop - start)))\n",
    "    return disrupted_ref_bs, disr_bs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_pie(pie_chart_data: List) -> None:\n",
    "    assert isinstance(pie_chart_data, list)\n",
    "    assert len(pie_chart_data) == 5\n",
    "    #slice labels\n",
    "    labels = [\n",
    "        \"Reference binding sites only\", \n",
    "        \"Reference binding sites disrupted in haplotypes\",\n",
    "        \"Haplotypes binding sites only\", \n",
    "        \"Haplotype binding sites with enhanced binding affinity\", \n",
    "        \"Haplotype binding sites with weakened binding affinity\"\n",
    "    ]\n",
    "    # slice colors\n",
    "    colors = [\n",
    "        \"#5f8dd3ff\", \"#0055d4ff\", \"#ff6600ff\", \"#ffcc00ff\", \"#ffeeaaff\"\n",
    "    ]\n",
    "    # slice explode\n",
    "    explode = (0, 0.05, 0.05, 0.05, 0.05)\n",
    "    # plot pie chart\n",
    "    plt.figure(figsize=(15, 10))\n",
    "    plt.pie(\n",
    "        pie_chart_data, explode=explode, colors=colors, autopct='%1.2f%%', \n",
    "        shadow=False, startangle=140, textprops={'fontsize': 20}\n",
    "    )\n",
    "    plt.legend(labels, loc=(0.8,0.85), prop={'size': 22})\n",
    "    plt.axis('equal')\n",
    "    plt.rc('axes', titlesize=22)\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's now analyse our data to assess genetic variants effects on TFBS binding affinity."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "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>motif_id</th>\n",
       "      <th>motif_alt_id</th>\n",
       "      <th>sequence_name</th>\n",
       "      <th>start</th>\n",
       "      <th>stop</th>\n",
       "      <th>strand</th>\n",
       "      <th>score</th>\n",
       "      <th>p-value</th>\n",
       "      <th>q-value</th>\n",
       "      <th>matched_sequence</th>\n",
       "      <th>haplotype_frequency</th>\n",
       "      <th>reference</th>\n",
       "      <th>coords</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>MA0010.1</td>\n",
       "      <td>br</td>\n",
       "      <td>chr3L:8333115-8333767</td>\n",
       "      <td>8333346</td>\n",
       "      <td>8333360</td>\n",
       "      <td>+</td>\n",
       "      <td>18.594595</td>\n",
       "      <td>7.450581e-09</td>\n",
       "      <td>0.139370</td>\n",
       "      <td>GTAATAAACAAATC</td>\n",
       "      <td>2</td>\n",
       "      <td>ref</td>\n",
       "      <td>chr3L_8333346_8333360_+</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>MA0010.1</td>\n",
       "      <td>br</td>\n",
       "      <td>chr2R:17367986-17368592</td>\n",
       "      <td>17368077</td>\n",
       "      <td>17368091</td>\n",
       "      <td>+</td>\n",
       "      <td>18.270270</td>\n",
       "      <td>1.490116e-08</td>\n",
       "      <td>0.139370</td>\n",
       "      <td>GTAATAAACAAAAC</td>\n",
       "      <td>2</td>\n",
       "      <td>ref</td>\n",
       "      <td>chr2R_17368077_17368091_+</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>MA0010.1</td>\n",
       "      <td>br</td>\n",
       "      <td>chr2R:7873123-7873580</td>\n",
       "      <td>7873526</td>\n",
       "      <td>7873540</td>\n",
       "      <td>+</td>\n",
       "      <td>18.189189</td>\n",
       "      <td>2.235174e-08</td>\n",
       "      <td>0.139370</td>\n",
       "      <td>ATAATAAACAAATC</td>\n",
       "      <td>4</td>\n",
       "      <td>ref</td>\n",
       "      <td>chr2R_7873526_7873540_+</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>MA0010.1</td>\n",
       "      <td>br</td>\n",
       "      <td>chr2L:21756732-21758293</td>\n",
       "      <td>21757363</td>\n",
       "      <td>21757377</td>\n",
       "      <td>+</td>\n",
       "      <td>18.189189</td>\n",
       "      <td>2.235174e-08</td>\n",
       "      <td>0.139370</td>\n",
       "      <td>ATAATAAACAAATC</td>\n",
       "      <td>2</td>\n",
       "      <td>ref</td>\n",
       "      <td>chr2L_21757363_21757377_+</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>MA0010.1</td>\n",
       "      <td>br</td>\n",
       "      <td>chr3L:4349944-4350445</td>\n",
       "      <td>4350200</td>\n",
       "      <td>4350186</td>\n",
       "      <td>-</td>\n",
       "      <td>18.009009</td>\n",
       "      <td>4.470348e-08</td>\n",
       "      <td>0.143027</td>\n",
       "      <td>GTAATAAACAAATG</td>\n",
       "      <td>2</td>\n",
       "      <td>ref</td>\n",
       "      <td>chr3L_4350200_4350186_-</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   motif_id motif_alt_id            sequence_name     start      stop strand  \\\n",
       "0  MA0010.1           br    chr3L:8333115-8333767   8333346   8333360      +   \n",
       "1  MA0010.1           br  chr2R:17367986-17368592  17368077  17368091      +   \n",
       "2  MA0010.1           br    chr2R:7873123-7873580   7873526   7873540      +   \n",
       "3  MA0010.1           br  chr2L:21756732-21758293  21757363  21757377      +   \n",
       "4  MA0010.1           br    chr3L:4349944-4350445   4350200   4350186      -   \n",
       "\n",
       "       score       p-value   q-value matched_sequence  haplotype_frequency  \\\n",
       "0  18.594595  7.450581e-09  0.139370   GTAATAAACAAATC                    2   \n",
       "1  18.270270  1.490116e-08  0.139370   GTAATAAACAAAAC                    2   \n",
       "2  18.189189  2.235174e-08  0.139370   ATAATAAACAAATC                    4   \n",
       "3  18.189189  2.235174e-08  0.139370   ATAATAAACAAATC                    2   \n",
       "4  18.009009  4.470348e-08  0.143027   GTAATAAACAAATG                    2   \n",
       "\n",
       "  reference                     coords  \n",
       "0       ref    chr3L_8333346_8333360_+  \n",
       "1       ref  chr2R_17368077_17368091_+  \n",
       "2       ref    chr2R_7873526_7873540_+  \n",
       "3       ref  chr2L_21757363_21757377_+  \n",
       "4       ref    chr3L_4350200_4350186_-  "
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# add coordinates \n",
    "report[\"coords\"] = report.apply(\n",
    "    lambda x : get_coords(x[2], str(x[3]), str(x[4]), x[5]), axis=1\n",
    ")\n",
    "# keep only significant motif occurrences (p-value < 1e-4)\n",
    "report_significant = report[report[\"p-value\"] < 1e-4]\n",
    "report_significant.reset_index(drop=True, inplace=True)\n",
    "report_significant.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Classifying 15688 motif occurrences...\n",
      "Classified 15688 motif occurrences in:\t14.19s\n",
      "Analyzing 14623 binding sites...\n",
      "Analyzed 14623 binding sites in:\t24.87s\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1080x720 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "pie_chart_data, ref_tfbs, hap_bs, enha_bs, weak_bs = assign_tfbs_class(report_significant)\n",
    "disr_refbs, disr_bs = get_disrupted_tfbs(report, ref_tfbs)\n",
    "pie_chart_data = [pie_chart_data[0]] + [disr_refbs] + pie_chart_data[1:]\n",
    "plot_pie(pie_chart_data)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "By looking at the results, in our example we can observe that 93.10 % of *br* binding sites are not modulated by genetic variants in our datasets. However, 6.67% of potential TFBS are can be found only on samples genomes. Therefore, these binding sites are created by the presence of variants. Similarly, 0.18% of the binding sites recovered on the reference genome sequence are disrupted by variants (their *P*-value is no more significant). The remaining binding sites are enhanced or weakened by genetic variants, 0.04% and 0.02% respectively. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3.8.5 ('base')",
   "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": "3.8.5"
  },
  "orig_nbformat": 4,
  "vscode": {
   "interpreter": {
    "hash": "4e208e320e1e0d80f8282c33c7f52dcec975daea0a03c3bb395073eef6c393a2"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}