pileup_CTCF.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Pileups and average features"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Welcome to the cooltools pileups notebook!\n",
    " \n",
    "Averaging Hi-C/Micro-C maps allows the quantification of general patterns observed in the maps. Averaging comes in various forms: contact-vs-distance plots, saddle plots, and pileup plots. **Pileup plots** are the averaged local Hi-C map over the 2D windows (i.e. **snippets**). These are also referred to as \"average Hi-C maps\". Pileups can be useful for determining the relationship between features (e.g. CTCF and TAD boundaries). Pileups can also be beneficial for reliably observing features in low-coverage Hi-C or single-cell HiC maps.\n",
    "\n",
    "For pileups, we retrieve local windows that are centered at the **anchors**. We call this procedure **snipping**. Anchors can be ChIP-Seq binding sites, anchors of dots, or any other genomic features. Pileups come in two varieties: \n",
    "\n",
    "- **On-diagonal pileup**. Each window is centered at the pixel located at the anchor position, at the main diagonal. Both coordinates of the window center are equivalent to the bin of the anchor. \n",
    "- **Off-diagonal pileup**. Each window is centered at the pixel with one anchor as a left coordinate and another anchor as a right coordinate. \n",
    "\n",
    "Typically, the sizes of windows are equivalent. After the selection of windows, we average them elementwise. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Content:\n",
    "\n",
    "1. Download data\n",
    "2. Load data\n",
    "    - Load genomic regions\n",
    "    - Load features for anchors\n",
    "3. On-diagonal pipeup of CTCF\n",
    "    - On-diagonal pileup of ICed Hi-C interactions\n",
    "    - On-diagonal pileup of observed over expected interactions\n",
    "    - Inspect the snips\n",
    "4. Off-diagonal pileup of CTCF"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "tags": [
     "remove_cell"
    ]
   },
   "outputs": [],
   "source": [
    "# If you are a developer, you may want to reload the packages on a fly. \n",
    "# Jupyter has a magic for this particular purpose:\n",
    "%load_ext autoreload\n",
    "%autoreload 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# import standard python libraries\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import pandas as pd\n",
    "import seaborn as sns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# import libraries for biological data analysis\n",
    "import cooler\n",
    "import bioframe\n",
    "\n",
    "import cooltools\n",
    "\n",
    "from packaging import version\n",
    "if version.parse(cooltools.__version__) < version.parse('0.5.2'):\n",
    "    raise AssertionError(\"tutorials rely on cooltools version 0.5.2 or higher,\"+\n",
    "                         \"please check your cooltools version and update to the latest\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Download data\n",
    "\n",
    "For this example notebook, we collected the data from [immortalized human foreskin fibroblast cell line HFFc6](https://data.4dnucleome.org/ontology-terms/EFO:0009318/):\n",
    "\n",
    "- Micro-C data from [Krietenstein et al. 2020](https://pubmed.ncbi.nlm.nih.gov/32213324/)\n",
    "- ChIP-Seq for CTCF from [ENCODE ENCSR000DWQ](https://www.encodeproject.org/experiments/ENCSR000DWQ/)\n",
    "\n",
    "You can automatically download test datasets with cooltools. More information on the files and how they were obtained is available from the datasets description. \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1) HFF_MicroC : Micro-C data from HFF human cells for two chromosomes (hg38) in a multi-resolution mcool format. \n",
      "\tDownloaded from https://osf.io/3h9js/download \n",
      "\tStored as test.mcool \n",
      "\tOriginal md5sum: e4a0fc25c8dc3d38e9065fd74c565dd1\n",
      "\n",
      "2) HFF_CTCF_fc : ChIP-Seq fold change over input with CTCF antibodies in HFF cells (hg38). Downloaded from ENCODE ENCSR000DWQ, ENCFF761RHS.bigWig file \n",
      "\tDownloaded from https://osf.io/w92u3/download \n",
      "\tStored as test_CTCF.bigWig \n",
      "\tOriginal md5sum: 62429de974b5b4a379578cc85adc65a3\n",
      "\n",
      "3) HFF_CTCF_binding : Binding sites called from CTCF ChIP-Seq peaks for HFF cells (hg38). Peaks are from ENCODE ENCSR000DWQ, ENCFF498QCT.bed file. The motifs are called with gimmemotifs (options --nreport 1 --cutoff 0), with JASPAR pwm MA0139. \n",
      "\tDownloaded from https://osf.io/c9pwe/download \n",
      "\tStored as test_CTCF.bed.gz \n",
      "\tOriginal md5sum: 61ecfdfa821571a8e0ea362e8fd48f63\n",
      "\n"
     ]
    }
   ],
   "source": [
    "# Print available datasets for download\n",
    "cooltools.print_available_datasets()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Downloading test data for pileups\n",
    "# cache = True will doanload the data only if it was not previously downloaded\n",
    "data_dir = './data/'\n",
    "cool_file = cooltools.download_data(\"HFF_MicroC\", cache=True, data_dir=data_dir)\n",
    "ctcf_peaks_file = cooltools.download_data(\"HFF_CTCF_binding\", cache=True, data_dir=data_dir)\n",
    "ctcf_fc_file = cooltools.download_data(\"HFF_CTCF_fc\", cache=True, data_dir=data_dir)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Load data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Load genomic regions\n",
    "\n",
    "The pileup function needs **genomic regions**. Why?\n",
    "\n",
    "- First, pileup uses regions for parallelization of snipping. Different genomic regions are loaded simultaneously by different processes, and the snipping can be done in parallel. \n",
    "- Second, the observed over expected pileup requires calculating expected interactions before snipping (P(s), in other words). Typically, P(s) is calculated separately for each chromosome arm as inter-arms interactions might be affected by strong insulation of centromeres or Rabl configuration. \n",
    "\n",
    "For species that do not have information on chromosome arms, or have *telocentric chromosomes* (e.g., mouse), you may want to use full chromosomes instead. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Open cool file with Micro-C data:\n",
    "clr = cooler.Cooler(data_dir+'/test.mcool::/resolutions/10000')\n",
    "# Set up selected data resolution:\n",
    "resolution = clr.binsize"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Use bioframe to fetch the genomic features from the UCSC.\n",
    "hg38_chromsizes = bioframe.fetch_chromsizes('hg38')\n",
    "hg38_cens = bioframe.fetch_centromeres('hg38')\n",
    "hg38_arms = bioframe.make_chromarms(hg38_chromsizes, hg38_cens)\n",
    "\n",
    "# Select only chromosomes that are present in the cooler. \n",
    "# This step is typically not required! we call it only because the test data are reduced. \n",
    "hg38_arms = hg38_arms.set_index(\"chrom\").loc[clr.chromnames].reset_index()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Load features for anchors\n",
    "\n",
    "Construction of the pileup requires  genomic **features** that will be used for centering of the **snippets**. In this example, we will use *positions of motifs in CTCF peaks* as features. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "scrolled": true
   },
   "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>chrom</th>\n",
       "      <th>start</th>\n",
       "      <th>end</th>\n",
       "      <th>name</th>\n",
       "      <th>score</th>\n",
       "      <th>strand</th>\n",
       "      <th>mid</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>17271</th>\n",
       "      <td>chr17</td>\n",
       "      <td>118485</td>\n",
       "      <td>118504</td>\n",
       "      <td>MA0139.1_CTCF_human</td>\n",
       "      <td>12.384042</td>\n",
       "      <td>-</td>\n",
       "      <td>118494</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17272</th>\n",
       "      <td>chr17</td>\n",
       "      <td>144002</td>\n",
       "      <td>144021</td>\n",
       "      <td>MA0139.1_CTCF_human</td>\n",
       "      <td>11.542617</td>\n",
       "      <td>+</td>\n",
       "      <td>144011</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17273</th>\n",
       "      <td>chr17</td>\n",
       "      <td>163676</td>\n",
       "      <td>163695</td>\n",
       "      <td>MA0139.1_CTCF_human</td>\n",
       "      <td>5.294219</td>\n",
       "      <td>-</td>\n",
       "      <td>163685</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17274</th>\n",
       "      <td>chr17</td>\n",
       "      <td>164711</td>\n",
       "      <td>164730</td>\n",
       "      <td>MA0139.1_CTCF_human</td>\n",
       "      <td>11.889376</td>\n",
       "      <td>+</td>\n",
       "      <td>164720</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17275</th>\n",
       "      <td>chr17</td>\n",
       "      <td>309416</td>\n",
       "      <td>309435</td>\n",
       "      <td>MA0139.1_CTCF_human</td>\n",
       "      <td>7.879575</td>\n",
       "      <td>-</td>\n",
       "      <td>309425</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       chrom   start     end                 name      score strand     mid\n",
       "17271  chr17  118485  118504  MA0139.1_CTCF_human  12.384042      -  118494\n",
       "17272  chr17  144002  144021  MA0139.1_CTCF_human  11.542617      +  144011\n",
       "17273  chr17  163676  163695  MA0139.1_CTCF_human   5.294219      -  163685\n",
       "17274  chr17  164711  164730  MA0139.1_CTCF_human  11.889376      +  164720\n",
       "17275  chr17  309416  309435  MA0139.1_CTCF_human   7.879575      -  309425"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Read CTCF peaks data and select only chromosomes present in cooler:\n",
    "ctcf = bioframe.read_table(ctcf_peaks_file, schema='bed').query(f'chrom in {clr.chromnames}')\n",
    "ctcf['mid'] = (ctcf.end+ctcf.start)//2\n",
    "ctcf.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Feature inspection and filtering\n",
    "\n",
    "Since we have both the list of strongest motifs of CTCF located in CTCF ChIP-Seq and the fold change over input for the genome, we have two characteristics of each feature: \n",
    "\n",
    "- score of the motif\n",
    "- CTCF ChIP-Seq fold-change over input\n",
    "\n",
    "Let's take a look at joint distribution of these scores: "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "import bbi\n",
    "from scipy.stats import linregress"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Get CTCF ChIP-Seq fold-change over input for genomic regions centered at the positions of the motifs\n",
    "\n",
    "flank = 250 # Length of flank to one side from the boundary, in basepairs\n",
    "ctcf_chip_signal = bbi.stackup(\n",
    "    ctcf_fc_file, \n",
    "    ctcf.chrom, \n",
    "    ctcf.mid-flank, \n",
    "    ctcf.mid+flank, \n",
    "    bins=1)\n",
    "\n",
    "ctcf['FC_score'] = ctcf_chip_signal"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "ctcf['quartile_score']    = pd.qcut(ctcf['score'], 4, labels=False) + 1\n",
    "ctcf['quartile_FC_score'] = pd.qcut(ctcf['FC_score'], 4, labels=False) + 1\n",
    "ctcf['peaks_importance'] = ctcf.apply(\n",
    "    lambda x: 'Top by both scores' if x.quartile_score==4 and x.quartile_FC_score==4 else\n",
    "                'Top by Motif score' if x.quartile_score==4 else\n",
    "                'Top by FC score' if x.quartile_FC_score==4 else 'Ordinary peaks', axis=1\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/geofffudenberg/anaconda3/envs/open2c/lib/python3.8/site-packages/seaborn/_decorators.py:36: FutureWarning: Pass the following variables as keyword args: x, y. From version 0.12, the only valid positional argument will be `data`, and passing other arguments without an explicit keyword will result in an error or misinterpretation.\n",
      "  warnings.warn(\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": [
    "x = ctcf['score']\n",
    "y = np.log(ctcf['FC_score'])\n",
    "\n",
    "fig, ax = plt.subplots()\n",
    "\n",
    "sns.scatterplot(x=x, y=y, hue=ctcf['peaks_importance'],\n",
    "    s=2,\n",
    "    alpha=0.5,\n",
    "    label='All peaks', \n",
    "    ax=ax\n",
    ")\n",
    "\n",
    "slope, intercept, r, p, se = linregress(x, y)\n",
    "\n",
    "ax.plot([-6, 19], [intercept-6*slope, intercept+19*slope], \n",
    "        alpha=0.5,\n",
    "        color='black',\n",
    "        label=f\"Regression line, R value: {r:.2f}\")\n",
    "\n",
    "ax.set(\n",
    "    xlabel='Motif score',\n",
    "    ylabel='ChIP-Seq fold-change over input')\n",
    "\n",
    "ax.legend(bbox_to_anchor=(1.01,1), loc=\"upper left\")\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "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>chrom</th>\n",
       "      <th>start</th>\n",
       "      <th>end</th>\n",
       "      <th>name</th>\n",
       "      <th>score</th>\n",
       "      <th>strand</th>\n",
       "      <th>mid</th>\n",
       "      <th>FC_score</th>\n",
       "      <th>quartile_score</th>\n",
       "      <th>quartile_FC_score</th>\n",
       "      <th>peaks_importance</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>659</th>\n",
       "      <td>chr17</td>\n",
       "      <td>8158938</td>\n",
       "      <td>8158957</td>\n",
       "      <td>MA0139.1_CTCF_human</td>\n",
       "      <td>13.276979</td>\n",
       "      <td>-</td>\n",
       "      <td>8158947</td>\n",
       "      <td>25.056849</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>Top by both scores</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>660</th>\n",
       "      <td>chr2</td>\n",
       "      <td>176127201</td>\n",
       "      <td>176127220</td>\n",
       "      <td>MA0139.1_CTCF_human</td>\n",
       "      <td>12.820343</td>\n",
       "      <td>+</td>\n",
       "      <td>176127210</td>\n",
       "      <td>25.027294</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>Top by both scores</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>661</th>\n",
       "      <td>chr17</td>\n",
       "      <td>38322364</td>\n",
       "      <td>38322383</td>\n",
       "      <td>MA0139.1_CTCF_human</td>\n",
       "      <td>13.534864</td>\n",
       "      <td>-</td>\n",
       "      <td>38322373</td>\n",
       "      <td>25.010430</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>Top by both scores</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>662</th>\n",
       "      <td>chr2</td>\n",
       "      <td>119265336</td>\n",
       "      <td>119265355</td>\n",
       "      <td>MA0139.1_CTCF_human</td>\n",
       "      <td>13.739862</td>\n",
       "      <td>-</td>\n",
       "      <td>119265345</td>\n",
       "      <td>24.980141</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>Top by both scores</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>663</th>\n",
       "      <td>chr2</td>\n",
       "      <td>118003514</td>\n",
       "      <td>118003533</td>\n",
       "      <td>MA0139.1_CTCF_human</td>\n",
       "      <td>12.646685</td>\n",
       "      <td>-</td>\n",
       "      <td>118003523</td>\n",
       "      <td>24.957502</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>Top by both scores</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     chrom      start        end                 name      score strand  \\\n",
       "659  chr17    8158938    8158957  MA0139.1_CTCF_human  13.276979      -   \n",
       "660   chr2  176127201  176127220  MA0139.1_CTCF_human  12.820343      +   \n",
       "661  chr17   38322364   38322383  MA0139.1_CTCF_human  13.534864      -   \n",
       "662   chr2  119265336  119265355  MA0139.1_CTCF_human  13.739862      -   \n",
       "663   chr2  118003514  118003533  MA0139.1_CTCF_human  12.646685      -   \n",
       "\n",
       "           mid   FC_score  quartile_score  quartile_FC_score  \\\n",
       "659    8158947  25.056849               4                  4   \n",
       "660  176127210  25.027294               4                  4   \n",
       "661   38322373  25.010430               4                  4   \n",
       "662  119265345  24.980141               4                  4   \n",
       "663  118003523  24.957502               4                  4   \n",
       "\n",
       "       peaks_importance  \n",
       "659  Top by both scores  \n",
       "660  Top by both scores  \n",
       "661  Top by both scores  \n",
       "662  Top by both scores  \n",
       "663  Top by both scores  "
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Select the CTCF sites that are in top quartile by both the ChIP-Seq data and motif score\n",
    "\n",
    "sites = ctcf[ctcf['peaks_importance']=='Top by both scores']\\\n",
    "    .sort_values('FC_score', ascending=False)\\\n",
    "    .reset_index(drop=True)\n",
    "sites.tail()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "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>chrom</th>\n",
       "      <th>start</th>\n",
       "      <th>end</th>\n",
       "      <th>name</th>\n",
       "      <th>score</th>\n",
       "      <th>strand</th>\n",
       "      <th>mid</th>\n",
       "      <th>FC_score</th>\n",
       "      <th>quartile_score</th>\n",
       "      <th>quartile_FC_score</th>\n",
       "      <th>peaks_importance</th>\n",
       "      <th>cluster</th>\n",
       "      <th>cluster_start</th>\n",
       "      <th>cluster_end</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>608</th>\n",
       "      <td>chr17</td>\n",
       "      <td>8158938</td>\n",
       "      <td>8158957</td>\n",
       "      <td>MA0139.1_CTCF_human</td>\n",
       "      <td>13.276979</td>\n",
       "      <td>-</td>\n",
       "      <td>8158947</td>\n",
       "      <td>25.056849</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>Top by both scores</td>\n",
       "      <td>34</td>\n",
       "      <td>8158938</td>\n",
       "      <td>8158957</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>609</th>\n",
       "      <td>chr2</td>\n",
       "      <td>176127201</td>\n",
       "      <td>176127220</td>\n",
       "      <td>MA0139.1_CTCF_human</td>\n",
       "      <td>12.820343</td>\n",
       "      <td>+</td>\n",
       "      <td>176127210</td>\n",
       "      <td>25.027294</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>Top by both scores</td>\n",
       "      <td>515</td>\n",
       "      <td>176127201</td>\n",
       "      <td>176127220</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>610</th>\n",
       "      <td>chr17</td>\n",
       "      <td>38322364</td>\n",
       "      <td>38322383</td>\n",
       "      <td>MA0139.1_CTCF_human</td>\n",
       "      <td>13.534864</td>\n",
       "      <td>-</td>\n",
       "      <td>38322373</td>\n",
       "      <td>25.010430</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>Top by both scores</td>\n",
       "      <td>104</td>\n",
       "      <td>38322364</td>\n",
       "      <td>38322383</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>611</th>\n",
       "      <td>chr2</td>\n",
       "      <td>119265336</td>\n",
       "      <td>119265355</td>\n",
       "      <td>MA0139.1_CTCF_human</td>\n",
       "      <td>13.739862</td>\n",
       "      <td>-</td>\n",
       "      <td>119265345</td>\n",
       "      <td>24.980141</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>Top by both scores</td>\n",
       "      <td>465</td>\n",
       "      <td>119265336</td>\n",
       "      <td>119265355</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>612</th>\n",
       "      <td>chr2</td>\n",
       "      <td>118003514</td>\n",
       "      <td>118003533</td>\n",
       "      <td>MA0139.1_CTCF_human</td>\n",
       "      <td>12.646685</td>\n",
       "      <td>-</td>\n",
       "      <td>118003523</td>\n",
       "      <td>24.957502</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>Top by both scores</td>\n",
       "      <td>462</td>\n",
       "      <td>118003514</td>\n",
       "      <td>118003533</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     chrom      start        end                 name      score strand  \\\n",
       "608  chr17    8158938    8158957  MA0139.1_CTCF_human  13.276979      -   \n",
       "609   chr2  176127201  176127220  MA0139.1_CTCF_human  12.820343      +   \n",
       "610  chr17   38322364   38322383  MA0139.1_CTCF_human  13.534864      -   \n",
       "611   chr2  119265336  119265355  MA0139.1_CTCF_human  13.739862      -   \n",
       "612   chr2  118003514  118003533  MA0139.1_CTCF_human  12.646685      -   \n",
       "\n",
       "           mid   FC_score  quartile_score  quartile_FC_score  \\\n",
       "608    8158947  25.056849               4                  4   \n",
       "609  176127210  25.027294               4                  4   \n",
       "610   38322373  25.010430               4                  4   \n",
       "611  119265345  24.980141               4                  4   \n",
       "612  118003523  24.957502               4                  4   \n",
       "\n",
       "       peaks_importance  cluster  cluster_start  cluster_end  \n",
       "608  Top by both scores       34        8158938      8158957  \n",
       "609  Top by both scores      515      176127201    176127220  \n",
       "610  Top by both scores      104       38322364     38322383  \n",
       "611  Top by both scores      465      119265336    119265355  \n",
       "612  Top by both scores      462      118003514    118003533  "
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Some CTCF sites might be located too close in the genome and interfere with analysis. \n",
    "# We will collapse the sites falling into the same size genomic bins as the resolution of our micro-C data:\n",
    "sites = bioframe.cluster(sites, min_dist=resolution)\\\n",
    "    .drop_duplicates('cluster')\\\n",
    "    .reset_index(drop=True)\n",
    "sites.tail()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## On-diagonal pileup\n",
    "\n",
    "On-diagonal pileup is the simplest, you need the positions of **features** (middlepoints of CTCF motifs) and the size of flanks aroung each motif. cooltools will create a snippet of Hi-C map for each feature. Then you can combine them into a single 2D pileup. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### On-diagonal pileup of ICed Hi-C interactions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "stack = cooltools.pileup(clr, sites, view_df=hg38_arms, flank=300_000)\n",
    "# Mirror reflect snippets when the feature is on the opposite strand\n",
    "mask = np.array(sites.strand == '-', dtype=bool)\n",
    "stack[:, :, mask] = stack[::-1, ::-1, mask]\n",
    "\n",
    "# Aggregate. Note that some pixels might be converted to NaNs after IC, thus we aggregate by nanmean: \n",
    "mtx = np.nanmean(stack, axis=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Load colormap with large number of distinguishable intermediary tones,\n",
    "# The \"fall\" colormap in cooltools is exactly for this purpose.\n",
    "# After this step, you can use \"fall\" as cmap parameter in matplotlib:\n",
    "import cooltools.lib.plotting"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/var/folders/4s/d866wm3s4zbc9m41334fxfwr0000gp/T/ipykernel_33615/2426526626.py:2: RuntimeWarning: divide by zero encountered in log10\n",
      "  np.log10(mtx),\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(\n",
    "    np.log10(mtx),\n",
    "    vmin = -3,\n",
    "    vmax = -1,\n",
    "    cmap='fall',\n",
    "    interpolation='none')\n",
    "\n",
    "plt.colorbar(label = 'log10 mean ICed Hi-C')\n",
    "ticks_pixels = np.linspace(0, flank*2//resolution,5)\n",
    "ticks_kbp = ((ticks_pixels-ticks_pixels[-1]/2)*resolution//1000).astype(int)\n",
    "plt.xticks(ticks_pixels, ticks_kbp)\n",
    "plt.yticks(ticks_pixels, ticks_kbp)\n",
    "plt.xlabel('relative position, kbp')\n",
    "plt.ylabel('relative position, kbp')\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### On-diagonal pileup of observed over expected interactions\n",
    "\n",
    "Sometimes you don't want to include the **distance decay** P(s) in your pileups. For example, when you make comparison of pileups between experiments and they have different P(s). Even if these differences are slight, they might affect the pileup of raw ICed Hi-C interactions. \n",
    "\n",
    "In this case, the observed over expected pileup is your choice. Prior to running the pileup function, you need to calculate expected interactions for chromosome arms."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "expected = cooltools.expected_cis(clr, view_df=hg38_arms, nproc=2, chunksize=1_000_000)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "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>region1</th>\n",
       "      <th>region2</th>\n",
       "      <th>dist</th>\n",
       "      <th>n_valid</th>\n",
       "      <th>count.sum</th>\n",
       "      <th>balanced.sum</th>\n",
       "      <th>count.avg</th>\n",
       "      <th>balanced.avg</th>\n",
       "      <th>balanced.avg.smoothed</th>\n",
       "      <th>balanced.avg.smoothed.agg</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>chr2_p</td>\n",
       "      <td>chr2_p</td>\n",
       "      <td>0</td>\n",
       "      <td>8771</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>chr2_p</td>\n",
       "      <td>chr2_p</td>\n",
       "      <td>1</td>\n",
       "      <td>8753</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.000495</td>\n",
       "      <td>0.000520</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>chr2_p</td>\n",
       "      <td>chr2_p</td>\n",
       "      <td>2</td>\n",
       "      <td>8745</td>\n",
       "      <td>2656738.0</td>\n",
       "      <td>406.013088</td>\n",
       "      <td>303.800800</td>\n",
       "      <td>0.046428</td>\n",
       "      <td>0.042469</td>\n",
       "      <td>0.044728</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>chr2_p</td>\n",
       "      <td>chr2_p</td>\n",
       "      <td>3</td>\n",
       "      <td>8741</td>\n",
       "      <td>1563363.0</td>\n",
       "      <td>237.585271</td>\n",
       "      <td>178.854021</td>\n",
       "      <td>0.027181</td>\n",
       "      <td>0.026796</td>\n",
       "      <td>0.028226</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>chr2_p</td>\n",
       "      <td>chr2_p</td>\n",
       "      <td>4</td>\n",
       "      <td>8738</td>\n",
       "      <td>1125674.0</td>\n",
       "      <td>169.308714</td>\n",
       "      <td>128.825132</td>\n",
       "      <td>0.019376</td>\n",
       "      <td>0.019097</td>\n",
       "      <td>0.020152</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32543</th>\n",
       "      <td>chr17_q</td>\n",
       "      <td>chr17_q</td>\n",
       "      <td>5850</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.000010</td>\n",
       "      <td>0.000006</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32544</th>\n",
       "      <td>chr17_q</td>\n",
       "      <td>chr17_q</td>\n",
       "      <td>5851</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.000010</td>\n",
       "      <td>0.000006</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32545</th>\n",
       "      <td>chr17_q</td>\n",
       "      <td>chr17_q</td>\n",
       "      <td>5852</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.000010</td>\n",
       "      <td>0.000006</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32546</th>\n",
       "      <td>chr17_q</td>\n",
       "      <td>chr17_q</td>\n",
       "      <td>5853</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.000010</td>\n",
       "      <td>0.000006</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32547</th>\n",
       "      <td>chr17_q</td>\n",
       "      <td>chr17_q</td>\n",
       "      <td>5854</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.000010</td>\n",
       "      <td>0.000006</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>32548 rows × 10 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       region1  region2  dist  n_valid  count.sum  balanced.sum   count.avg  \\\n",
       "0       chr2_p   chr2_p     0     8771        NaN           NaN         NaN   \n",
       "1       chr2_p   chr2_p     1     8753        NaN           NaN         NaN   \n",
       "2       chr2_p   chr2_p     2     8745  2656738.0    406.013088  303.800800   \n",
       "3       chr2_p   chr2_p     3     8741  1563363.0    237.585271  178.854021   \n",
       "4       chr2_p   chr2_p     4     8738  1125674.0    169.308714  128.825132   \n",
       "...        ...      ...   ...      ...        ...           ...         ...   \n",
       "32543  chr17_q  chr17_q  5850        0        0.0      0.000000         NaN   \n",
       "32544  chr17_q  chr17_q  5851        0        0.0      0.000000         NaN   \n",
       "32545  chr17_q  chr17_q  5852        0        0.0      0.000000         NaN   \n",
       "32546  chr17_q  chr17_q  5853        0        0.0      0.000000         NaN   \n",
       "32547  chr17_q  chr17_q  5854        0        0.0      0.000000         NaN   \n",
       "\n",
       "       balanced.avg  balanced.avg.smoothed  balanced.avg.smoothed.agg  \n",
       "0               NaN                    NaN                        NaN  \n",
       "1               NaN               0.000495                   0.000520  \n",
       "2          0.046428               0.042469                   0.044728  \n",
       "3          0.027181               0.026796                   0.028226  \n",
       "4          0.019376               0.019097                   0.020152  \n",
       "...             ...                    ...                        ...  \n",
       "32543           NaN               0.000010                   0.000006  \n",
       "32544           NaN               0.000010                   0.000006  \n",
       "32545           NaN               0.000010                   0.000006  \n",
       "32546           NaN               0.000010                   0.000006  \n",
       "32547           NaN               0.000010                   0.000006  \n",
       "\n",
       "[32548 rows x 10 columns]"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "expected"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create the stack of snips:\n",
    "stack = cooltools.pileup(clr, sites, view_df=hg38_arms, expected_df=expected, flank=300_000) \n",
    "\n",
    "# Mirror reflect snippets when the feature is on the opposite strand\n",
    "mask = np.array(sites.strand == '-', dtype=bool)\n",
    "stack[:, :, mask] = stack[::-1, ::-1, mask]\n",
    "    \n",
    "mtx = np.nanmean(stack, axis=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/var/folders/4s/d866wm3s4zbc9m41334fxfwr0000gp/T/ipykernel_33615/2557000624.py:2: RuntimeWarning: divide by zero encountered in log2\n",
      "  np.log2(mtx),\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(\n",
    "    np.log2(mtx),\n",
    "    vmax = 1.0,\n",
    "    vmin = -1.0,\n",
    "    cmap='coolwarm',\n",
    "    interpolation='none')\n",
    "\n",
    "plt.colorbar(label = 'log2 mean obs/exp')\n",
    "ticks_pixels = np.linspace(0, flank*2//resolution,5)\n",
    "ticks_kbp = ((ticks_pixels-ticks_pixels[-1]/2)*resolution//1000).astype(int)\n",
    "plt.xticks(ticks_pixels, ticks_kbp)\n",
    "plt.yticks(ticks_pixels, ticks_kbp)\n",
    "plt.xlabel('relative position, kbp')\n",
    "plt.ylabel('relative position, kbp')\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Inspect the snips\n",
    "\n",
    "Aggregation is a convenient though dangerous step. It averages your data so that you cannot distinguish whether the signal is indeed average, or there is a single dataset that introduces a bias to your analysis. To make sure there are no outliers, you may want to use inspection of individual snippets.\n",
    "\n",
    "The cell below shows one way to interactively investigate snippets contributing to a pileup. Note that this is not interactive on readthedocs, but can be run if the notebook is obtained from [open2c_examples](https://github.com/open2c/open2c_examples). This widget sorts the dataframe with CTCF motifs by the strength of binding. This allows us to inspect the Micro-C maps at the positions of the strongest and weakest CTCF sites. Run the cell below and try to compare snippets with the lowest score to the snippets with the largest score. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "1cadd6adde5843ac9876bcdc6dda20ff",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "interactive(children=(IntSlider(value=306, description='i', max=612), Output()), _dom_classes=('widget-interac…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from ipywidgets import interact\n",
    "from matplotlib.gridspec import GridSpec\n",
    "\n",
    "n_examples = len(sites)\n",
    "\n",
    "@interact(i=(0, n_examples-1))\n",
    "def f(i):\n",
    "    fig, ax = plt.subplots(figsize=[5,5])\n",
    "    img = ax.matshow(\n",
    "        np.log2(stack[:, :, i]),  \n",
    "        vmin=-1,\n",
    "        vmax=1,\n",
    "        extent=[-flank//1000, flank//1000, -flank//1000, flank//1000],\n",
    "        cmap='coolwarm'\n",
    "    )\n",
    "    ax.xaxis.tick_bottom()\n",
    "    if i > 0:\n",
    "        ax.yaxis.set_visible(False)\n",
    "    plt.title(f'{i+1}-th snippet from top \\n FC score: {sites.loc[i, \"FC_score\"]:.2f}\\n and motif score: {sites.loc[i, \"score\"]:.2f}')\n",
    "    plt.axvline(0, c='g', ls=':')\n",
    "    plt.axhline(0, c='g', ls=':')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Compare top strongest peaks with others\n",
    "\n",
    "Compare the top peaks with both motif score and FC score to the rest of the peaks:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create the stack of snips:\n",
    "stack = cooltools.pileup(clr, ctcf, view_df=hg38_arms, expected_df=expected, flank=300_000\n",
    "            ) \n",
    "\n",
    "# Mirror reflect snippets where the feature is on the opposite strand\n",
    "mask = np.array(ctcf.strand == '-', dtype=bool)\n",
    "stack[:, :, mask] = stack[::-1, ::-1, mask]\n",
    "    \n",
    "mtx = np.nanmean(stack, axis=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/var/folders/4s/d866wm3s4zbc9m41334fxfwr0000gp/T/ipykernel_33615/2210978640.py:14: RuntimeWarning: divide by zero encountered in log2\n",
      "  np.log2(mtx),\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# TODO: add some strength of insulation for the pileup? \n",
    "\n",
    "groups = ['Top by both scores', 'Ordinary peaks']\n",
    "n_groups = len(groups)\n",
    "\n",
    "ticks_pixels = np.linspace(0, flank*2//resolution,5)\n",
    "ticks_kbp = ((ticks_pixels-ticks_pixels[-1]/2)*resolution//1000).astype(int)\n",
    "\n",
    "fig, axs = plt.subplots(1, n_groups, sharex=True, sharey=True, figsize=(4*n_groups, 4))\n",
    "for i in range(n_groups):\n",
    "    mtx = np.nanmean( stack[:, :, ctcf['peaks_importance']==groups[i]], axis=2)\n",
    "    ax = axs[i]\n",
    "    ax.imshow(\n",
    "        np.log2(mtx),\n",
    "        vmax = 1.0,\n",
    "        vmin = -1.0,\n",
    "        cmap='coolwarm',\n",
    "        interpolation='none')\n",
    "    \n",
    "    ax.set(title=f'{groups[i]} group', \n",
    "           xticks=ticks_pixels, \n",
    "           xticklabels=ticks_kbp, \n",
    "           xlabel='relative position, kbp')\n",
    "\n",
    "axs[0].set(yticks=ticks_pixels, \n",
    "       yticklabels=ticks_kbp, \n",
    "       ylabel='relative position, kbp')\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Off-diagonal pileup\n",
    "\n",
    "**Off-diagonal pileups** are the averaged Hi-C maps around double anchors. In this case, the anchors are CTCF sites in the genome. \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "paired_sites = bioframe.pair_by_distance(sites, min_sep=200000, max_sep=1000000, suffixes=('1', '2'))\n",
    "paired_sites.loc[:, 'mid1'] = (paired_sites['start1'] + paired_sites['end1'])//2\n",
    "paired_sites.loc[:, 'mid2'] = (paired_sites['start2'] + paired_sites['end2'])//2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1634\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>chrom1</th>\n",
       "      <th>start1</th>\n",
       "      <th>end1</th>\n",
       "      <th>name1</th>\n",
       "      <th>score1</th>\n",
       "      <th>strand1</th>\n",
       "      <th>mid1</th>\n",
       "      <th>FC_score1</th>\n",
       "      <th>quartile_score1</th>\n",
       "      <th>quartile_FC_score1</th>\n",
       "      <th>...</th>\n",
       "      <th>score2</th>\n",
       "      <th>strand2</th>\n",
       "      <th>mid2</th>\n",
       "      <th>FC_score2</th>\n",
       "      <th>quartile_score2</th>\n",
       "      <th>quartile_FC_score2</th>\n",
       "      <th>peaks_importance2</th>\n",
       "      <th>cluster2</th>\n",
       "      <th>cluster_start2</th>\n",
       "      <th>cluster_end2</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>chr17</td>\n",
       "      <td>412407</td>\n",
       "      <td>412426</td>\n",
       "      <td>MA0139.1_CTCF_human</td>\n",
       "      <td>12.212548</td>\n",
       "      <td>+</td>\n",
       "      <td>412416</td>\n",
       "      <td>41.645123</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>...</td>\n",
       "      <td>13.272118</td>\n",
       "      <td>-</td>\n",
       "      <td>1056231</td>\n",
       "      <td>35.072572</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>Top by both scores</td>\n",
       "      <td>1</td>\n",
       "      <td>1056222</td>\n",
       "      <td>1056241</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>chr17</td>\n",
       "      <td>412407</td>\n",
       "      <td>412426</td>\n",
       "      <td>MA0139.1_CTCF_human</td>\n",
       "      <td>12.212548</td>\n",
       "      <td>+</td>\n",
       "      <td>412416</td>\n",
       "      <td>41.645123</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>...</td>\n",
       "      <td>13.996208</td>\n",
       "      <td>-</td>\n",
       "      <td>1187374</td>\n",
       "      <td>69.994562</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>Top by both scores</td>\n",
       "      <td>2</td>\n",
       "      <td>1187365</td>\n",
       "      <td>1187384</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>chr17</td>\n",
       "      <td>412407</td>\n",
       "      <td>412426</td>\n",
       "      <td>MA0139.1_CTCF_human</td>\n",
       "      <td>12.212548</td>\n",
       "      <td>+</td>\n",
       "      <td>412416</td>\n",
       "      <td>41.645123</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>...</td>\n",
       "      <td>14.735101</td>\n",
       "      <td>+</td>\n",
       "      <td>1259280</td>\n",
       "      <td>31.643758</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>Top by both scores</td>\n",
       "      <td>3</td>\n",
       "      <td>1259271</td>\n",
       "      <td>1259290</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>chr17</td>\n",
       "      <td>412407</td>\n",
       "      <td>412426</td>\n",
       "      <td>MA0139.1_CTCF_human</td>\n",
       "      <td>12.212548</td>\n",
       "      <td>+</td>\n",
       "      <td>412416</td>\n",
       "      <td>41.645123</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>...</td>\n",
       "      <td>13.983562</td>\n",
       "      <td>+</td>\n",
       "      <td>1276338</td>\n",
       "      <td>35.440247</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>Top by both scores</td>\n",
       "      <td>4</td>\n",
       "      <td>1276329</td>\n",
       "      <td>1276348</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>chr17</td>\n",
       "      <td>412407</td>\n",
       "      <td>412426</td>\n",
       "      <td>MA0139.1_CTCF_human</td>\n",
       "      <td>12.212548</td>\n",
       "      <td>+</td>\n",
       "      <td>412416</td>\n",
       "      <td>41.645123</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>...</td>\n",
       "      <td>12.045221</td>\n",
       "      <td>+</td>\n",
       "      <td>1365609</td>\n",
       "      <td>36.746886</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>Top by both scores</td>\n",
       "      <td>5</td>\n",
       "      <td>1365600</td>\n",
       "      <td>1365619</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 28 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "  chrom1  start1    end1                name1     score1 strand1    mid1  \\\n",
       "0  chr17  412407  412426  MA0139.1_CTCF_human  12.212548       +  412416   \n",
       "1  chr17  412407  412426  MA0139.1_CTCF_human  12.212548       +  412416   \n",
       "2  chr17  412407  412426  MA0139.1_CTCF_human  12.212548       +  412416   \n",
       "3  chr17  412407  412426  MA0139.1_CTCF_human  12.212548       +  412416   \n",
       "4  chr17  412407  412426  MA0139.1_CTCF_human  12.212548       +  412416   \n",
       "\n",
       "   FC_score1  quartile_score1  quartile_FC_score1  ...     score2  strand2  \\\n",
       "0  41.645123                4                   4  ...  13.272118        -   \n",
       "1  41.645123                4                   4  ...  13.996208        -   \n",
       "2  41.645123                4                   4  ...  14.735101        +   \n",
       "3  41.645123                4                   4  ...  13.983562        +   \n",
       "4  41.645123                4                   4  ...  12.045221        +   \n",
       "\n",
       "      mid2  FC_score2 quartile_score2  quartile_FC_score2   peaks_importance2  \\\n",
       "0  1056231  35.072572               4                   4  Top by both scores   \n",
       "1  1187374  69.994562               4                   4  Top by both scores   \n",
       "2  1259280  31.643758               4                   4  Top by both scores   \n",
       "3  1276338  35.440247               4                   4  Top by both scores   \n",
       "4  1365609  36.746886               4                   4  Top by both scores   \n",
       "\n",
       "  cluster2  cluster_start2 cluster_end2  \n",
       "0        1         1056222      1056241  \n",
       "1        2         1187365      1187384  \n",
       "2        3         1259271      1259290  \n",
       "3        4         1276329      1276348  \n",
       "4        5         1365600      1365619  \n",
       "\n",
       "[5 rows x 28 columns]"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "print(len(paired_sites))\n",
    "paired_sites.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "For pileup, we will use the expected calculated above:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "# create the stack of snips:\n",
    "stack = cooltools.pileup(clr, paired_sites, view_df=hg38_arms, expected_df=expected, flank=100_000)\n",
    "    \n",
    "mtx = np.nanmean(stack, axis=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(\n",
    "    np.log2(mtx),\n",
    "    vmax = 1,\n",
    "    vmin = -1,\n",
    "    cmap='coolwarm')\n",
    "\n",
    "plt.colorbar(label = 'log2 mean obs/exp')\n",
    "ticks_pixels = np.linspace(0, flank*2//resolution,5)\n",
    "ticks_kbp = ((ticks_pixels-ticks_pixels[-1]/2)*resolution//1000).astype(int)\n",
    "plt.xticks(ticks_pixels, ticks_kbp)\n",
    "plt.yticks(ticks_pixels, ticks_kbp)\n",
    "plt.xlabel('relative position, kbp')\n",
    "plt.ylabel('relative position, kbp')\n",
    "\n",
    "plt.show()"
   ]
  }
 ],
 "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": "3.9.13"
  },
  "widgets": {
   "application/vnd.jupyter.widget-state+json": {
    "state": {
     "2e316a71e99b44ae928ad85c2895594f": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "VBoxModel",
      "state": {
       "_dom_classes": [
        "widget-interact"
       ],
       "children": [
        "IPY_MODEL_717d842dd95742b4bebbdbb16b9f20b7",
        "IPY_MODEL_a2a3b893a4ee4af29d080f6018601618"
       ],
       "layout": "IPY_MODEL_3ae93699e2d0458c946f4f8839aafaea"
      }
     },
     "2f5f06f8e5d242338daf3b033f34cd30": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "3ae93699e2d0458c946f4f8839aafaea": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "5369386b2fd94cc29204c705b9738675": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "717d842dd95742b4bebbdbb16b9f20b7": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "IntSliderModel",
      "state": {
       "description": "i",
       "layout": "IPY_MODEL_2f5f06f8e5d242338daf3b033f34cd30",
       "max": 612,
       "style": "IPY_MODEL_b52bc16c6963418694c2121f75b66024",
       "value": 61
      }
     },
     "a2a3b893a4ee4af29d080f6018601618": {
      "model_module": "@jupyter-widgets/output",
      "model_module_version": "1.0.0",
      "model_name": "OutputModel",
      "state": {
       "layout": "IPY_MODEL_5369386b2fd94cc29204c705b9738675",
       "outputs": [
        {
         "name": "stderr",
         "output_type": "stream",
         "text": "/var/folders/4s/d866wm3s4zbc9m41334fxfwr0000gp/T/ipykernel_47165/1864502587.py:10: RuntimeWarning: divide by zero encountered in log2\n  np.log2(stack[:, :, i]),\n"
        },
        {
         "data": {
          "image/png": "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\n",
          "text/plain": "<Figure size 360x360 with 1 Axes>"
         },
         "metadata": {
          "needs_background": "light"
         },
         "output_type": "display_data"
        }
       ]
      }
     },
     "b52bc16c6963418694c2121f75b66024": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "SliderStyleModel",
      "state": {
       "description_width": ""
      }
     }
    },
    "version_major": 2,
    "version_minor": 0
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}