├── BalancedRowsumsCalculation.ipynb
├── CalculatingFilledBinFractionByDistance.ipynb
├── ContactMapVisualizationExampleNotebook.ipynb
├── LICENSE
├── LoopStrengthRCMC.ipynb
├── MicrocompartmentLoops_PlusMin1kb.bedpe
├── PileupsRCMC.ipynb
├── README.md
├── captureprobes_mm10.bed
├── captureprobes_mm39.bed
├── loopFeatureOverlap.R
├── microcbowtie2.py
└── spikeinChIP_PE_alignment.py


/BalancedRowsumsCalculation.ipynb:
--------------------------------------------------------------------------------
  1 | {
  2 |  "cells": [
  3 |   {
  4 |    "cell_type": "code",
  5 |    "execution_count": null,
  6 |    "metadata": {},
  7 |    "outputs": [],
  8 |    "source": [
  9 |     "# import libraries\n",
 10 |     "import numpy as np\n",
 11 |     "import matplotlib.pyplot as plt\n",
 12 |     "import pandas as pd\n",
 13 |     "import os\n",
 14 |     "import cooltools\n",
 15 |     "import cooler"
 16 |    ]
 17 |   },
 18 |   {
 19 |    "cell_type": "markdown",
 20 |    "metadata": {},
 21 |    "source": [
 22 |     "## Load in microcompartment loop anchors"
 23 |    ]
 24 |   },
 25 |   {
 26 |    "cell_type": "code",
 27 |    "execution_count": 10,
 28 |    "metadata": {},
 29 |    "outputs": [
 30 |     {
 31 |      "data": {
 32 |       "text/html": [
 33 |        "<div>\n",
 34 |        "<style scoped>\n",
 35 |        "    .dataframe tbody tr th:only-of-type {\n",
 36 |        "        vertical-align: middle;\n",
 37 |        "    }\n",
 38 |        "\n",
 39 |        "    .dataframe tbody tr th {\n",
 40 |        "        vertical-align: top;\n",
 41 |        "    }\n",
 42 |        "\n",
 43 |        "    .dataframe thead th {\n",
 44 |        "        text-align: right;\n",
 45 |        "    }\n",
 46 |        "</style>\n",
 47 |        "<table border=\"1\" class=\"dataframe\">\n",
 48 |        "  <thead>\n",
 49 |        "    <tr style=\"text-align: right;\">\n",
 50 |        "      <th></th>\n",
 51 |        "      <th>chr</th>\n",
 52 |        "      <th>start</th>\n",
 53 |        "      <th>end</th>\n",
 54 |        "    </tr>\n",
 55 |        "  </thead>\n",
 56 |        "  <tbody>\n",
 57 |        "    <tr>\n",
 58 |        "      <th>0</th>\n",
 59 |        "      <td>chr5</td>\n",
 60 |        "      <td>31273000</td>\n",
 61 |        "      <td>31275000</td>\n",
 62 |        "    </tr>\n",
 63 |        "    <tr>\n",
 64 |        "      <th>1</th>\n",
 65 |        "      <td>chr5</td>\n",
 66 |        "      <td>31295500</td>\n",
 67 |        "      <td>31297500</td>\n",
 68 |        "    </tr>\n",
 69 |        "    <tr>\n",
 70 |        "      <th>2</th>\n",
 71 |        "      <td>chr5</td>\n",
 72 |        "      <td>31306700</td>\n",
 73 |        "      <td>31308700</td>\n",
 74 |        "    </tr>\n",
 75 |        "    <tr>\n",
 76 |        "      <th>3</th>\n",
 77 |        "      <td>chr5</td>\n",
 78 |        "      <td>31310500</td>\n",
 79 |        "      <td>31312500</td>\n",
 80 |        "    </tr>\n",
 81 |        "    <tr>\n",
 82 |        "      <th>4</th>\n",
 83 |        "      <td>chr5</td>\n",
 84 |        "      <td>31336200</td>\n",
 85 |        "      <td>31338200</td>\n",
 86 |        "    </tr>\n",
 87 |        "    <tr>\n",
 88 |        "      <th>...</th>\n",
 89 |        "      <td>...</td>\n",
 90 |        "      <td>...</td>\n",
 91 |        "      <td>...</td>\n",
 92 |        "    </tr>\n",
 93 |        "    <tr>\n",
 94 |        "      <th>127</th>\n",
 95 |        "      <td>chr8</td>\n",
 96 |        "      <td>85785750</td>\n",
 97 |        "      <td>85787750</td>\n",
 98 |        "    </tr>\n",
 99 |        "    <tr>\n",
100 |        "      <th>128</th>\n",
101 |        "      <td>chr8</td>\n",
102 |        "      <td>85797300</td>\n",
103 |        "      <td>85799300</td>\n",
104 |        "    </tr>\n",
105 |        "    <tr>\n",
106 |        "      <th>129</th>\n",
107 |        "      <td>chr8</td>\n",
108 |        "      <td>85802000</td>\n",
109 |        "      <td>85804000</td>\n",
110 |        "    </tr>\n",
111 |        "    <tr>\n",
112 |        "      <th>130</th>\n",
113 |        "      <td>chr8</td>\n",
114 |        "      <td>85806500</td>\n",
115 |        "      <td>85808500</td>\n",
116 |        "    </tr>\n",
117 |        "    <tr>\n",
118 |        "      <th>131</th>\n",
119 |        "      <td>chr8</td>\n",
120 |        "      <td>85809000</td>\n",
121 |        "      <td>85811000</td>\n",
122 |        "    </tr>\n",
123 |        "  </tbody>\n",
124 |        "</table>\n",
125 |        "<p>132 rows × 3 columns</p>\n",
126 |        "</div>"
127 |       ],
128 |       "text/plain": [
129 |        "      chr     start       end\n",
130 |        "0    chr5  31273000  31275000\n",
131 |        "1    chr5  31295500  31297500\n",
132 |        "2    chr5  31306700  31308700\n",
133 |        "3    chr5  31310500  31312500\n",
134 |        "4    chr5  31336200  31338200\n",
135 |        "..    ...       ...       ...\n",
136 |        "127  chr8  85785750  85787750\n",
137 |        "128  chr8  85797300  85799300\n",
138 |        "129  chr8  85802000  85804000\n",
139 |        "130  chr8  85806500  85808500\n",
140 |        "131  chr8  85809000  85811000\n",
141 |        "\n",
142 |        "[132 rows x 3 columns]"
143 |       ]
144 |      },
145 |      "execution_count": 10,
146 |      "metadata": {},
147 |      "output_type": "execute_result"
148 |     }
149 |    ],
150 |    "source": [
151 |     "# Load in the locations of all unique microcompartment anchors in BED format with chromosome, start (anchor coordinate minus 1 kb) and end (anchor coordinate plus 1 kb) columns\n",
152 |     "anchors = pd.read_csv(r'FILE_PATH_TO_ANCHOR_LIST.bed', sep='\\t', header=None, names=['chr','start','end'])\n",
153 |     "anchors"
154 |    ]
155 |   },
156 |   {
157 |    "cell_type": "code",
158 |    "execution_count": 11,
159 |    "metadata": {},
160 |    "outputs": [],
161 |    "source": [
162 |     "# Calculate the anchor coordinate by averaging the start & end coordinates, and create a DataFrame containing just the chromosome and anchor coordinate\n",
163 |     "anchors['midpoint'] = (anchors['end'] + anchors['start']) // 2\n",
164 |     "anchorList = anchors[['chr','midpoint']]\n",
165 |     "\n",
166 |     "# Subset the DataFrame of anchors to separate out the microcompartment anchors in Klf1 (on chr8) and in Ppm1g (on chr5)\n",
167 |     "anchorDataframeKlf1 = anchorList[anchorList['chr'] == 'chr8']\n",
168 |     "anchorDataframePpm1g = anchorList[anchorList['chr'] == 'chr5']\n",
169 |     "\n",
170 |     "# Convert the region-separated anchor lists from DataFrames to lists for ease of downstream use\n",
171 |     "anchorListKlf1 = anchorDataframeKlf1['midpoint'].tolist()\n",
172 |     "anchorListPpm1g = anchorDataframePpm1g['midpoint'].tolist()"
173 |    ]
174 |   },
175 |   {
176 |    "cell_type": "markdown",
177 |    "metadata": {},
178 |    "source": [
179 |     "## Calculate rowsums and plot their distribution\n",
180 |     "### Whole-region balancing calculations"
181 |    ]
182 |   },
183 |   {
184 |    "cell_type": "code",
185 |    "execution_count": 47,
186 |    "metadata": {},
187 |    "outputs": [
188 |     {
189 |      "data": {
190 |       "image/png": "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\n",
191 |       "text/plain": [
192 |        "<Figure size 432x288 with 1 Axes>"
193 |       ]
194 |      },
195 |      "metadata": {
196 |       "needs_background": "light"
197 |      },
198 |      "output_type": "display_data"
199 |     }
200 |    ],
201 |    "source": [
202 |     "# Usage: Plot different datasets & regions by updating the filename and \"region = regionsRCMC[x]\" attributes, respectively \n",
203 |     "\n",
204 |     "saveDir = f'DIRECTORY_PATH'\n",
205 |     "fileName = f'FILE_NAME.svg' # swap out .svg for the desired output file format\n",
206 |     "\n",
207 |     "# Specify the data resolution at which balanced rowsums will be calculated\n",
208 |     "resolution = 250\n",
209 |     "\n",
210 |     "# Load in the desired dataset\n",
211 |     "clr_Data = cooler.Cooler('FILE_PATH_TO_MCOOL::resolutions/'+str(resolution))\n",
212 |     "\n",
213 |     "# List the Capture regions for region-by-region calculations\n",
214 |     "## For an apples-to-apples comparison, balanced rowsums are calculated at the RCMC regions for both the RCMC dataset as well as the genome-wide Micro-C (Hsieh 2020) and Hi-C (Bonev 2017) datasets\n",
215 |     "regionsRCMC = ['chr3:33804149-35704149','chr5:31257344-32382344','chr6:122451959-122876959','chr8:84846629-85856629','chr18:58032072-59034072']\n",
216 |     "\n",
217 |     "# Specify which specific region you would like to calculate rowsums for\n",
218 |     "region = regionsRCMC[3]\n",
219 |     "\n",
220 |     "# Load in the balanced contact matrix for the specified region\n",
221 |     "regionMat = clr_Data.matrix(balance=True).fetch(region)\n",
222 |     "\n",
223 |     "# Create arrays of zeroes corresponding to the lengths (rows & columns) of the loaded in contact matrices\n",
224 |     "## The code here calculates both the rowsums and columnsums; they should be the same distribution, so you only need 1 of the 2\n",
225 |     "rowSum = np.zeros(len(regionMat))\n",
226 |     "columnSum = np.zeros(len(regionMat[0]))\n",
227 |     "\n",
228 |     "\n",
229 |     "# Iterate through the matrix bin-by-bin, summing up the balanced contact bin values for each row & column\n",
230 |     "## Only need to iterate through one half of the contact matrix (i.e., the half above the diagonal or the half below it) to capture it all due to reflected symmetry\n",
231 |     "i = 0\n",
232 |     "while i < len(regionMat):\n",
233 |     "    j = i # Ensures that we keep our iteration to one side of the diagonal instead of spanning the whole matrix \n",
234 |     "    while j < len(regionMat[0]):\n",
235 |     "        val = regionMat[i][j] # Extracts the balanced value of the contact bin\n",
236 |     "        if np.isnan(val):\n",
237 |     "            val = 0 # NaN values can cause calculation errors, so this sets them to 0\n",
238 |     "        rowSum[i] += val # Adds the balanced value to the corresponding rowsum\n",
239 |     "        columnSum[j] += val # Adds the balanced value to the corresponding columnsum\n",
240 |     "        \n",
241 |     "        if j != i: # If j = i, then we're on the diagonal and there's no need to add in the reflected twin of the contact bin\n",
242 |     "            rowSum[j] += val # Adds the balanced value of the reflected twin to the corresponding rowsum\n",
243 |     "            columnSum[i] += val # Adds the balanced value of the reflected twin to the corresponding columnsum\n",
244 |     "        j += 1\n",
245 |     "    i += 1\n",
246 |     "\n",
247 |     "# Plotting the rowsums and columnsums yields the same distribution, so this just runs ahead with the rowsums\n",
248 |     "## This filters out the rowsums that are 0 (i.e., those that correspond to empty stripes, either due to probe coverage or hard-to-map sites)\n",
249 |     "rowSumFilt = []\n",
250 |     "for i in rowSum:\n",
251 |     "    if i > 0:\n",
252 |     "        rowSumFilt.append(i)\n",
253 |     "\n",
254 |     "\n",
255 |     "# Plot a histogram of the rowsums\n",
256 |     "## Change the number of histogram bins as desired (100 used here)\n",
257 |     "## Change the range of visualization to capture and visually center the entire distribution\n",
258 |     "plt.hist(rowSumFilt,100,range=[3.75,8.25],align='mid',color='blue') # Klf1\n",
259 |     "# plt.hist(rowSumFilt,100,range=[1.1,3.4],align='mid',color='blue') # Ppm1g\n",
260 |     "# plt.hist(rowSumFilt,100,range=[0.3,1.0],align='mid',color='blue') # Fbn2\n",
261 |     "# plt.hist(rowSumFilt,100,range=[0.5,1.3],align='mid',color='blue') # Sox2\n",
262 |     "\n",
263 |     "# plt.savefig(saveDir + fileName, dpi=1200)"
264 |    ]
265 |   },
266 |   {
267 |    "cell_type": "markdown",
268 |    "metadata": {},
269 |    "source": [
270 |     "### Microcompartment-containing bins balancing calculations"
271 |    ]
272 |   },
273 |   {
274 |    "cell_type": "code",
275 |    "execution_count": 37,
276 |    "metadata": {},
277 |    "outputs": [
278 |     {
279 |      "data": {
280 |       "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXAAAAD4CAYAAAD1jb0+AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/YYfK9AAAACXBIWXMAAAsTAAALEwEAmpwYAAAKvklEQVR4nO3d0avk91nH8c/T7EqLCeRiDxqSHNcLKcRCm7DESKDUUCStYm96kYIRRFkqFlIQRL0w5B8oogi62GDF2lJoIyU0asCGUrCpSUxr4rYSSsXQwBJLmwRFSXi8mElY192d39k9c3afPa8XHHbOzu/MPPnulze//c3Mpro7AMzztis9AACXRsABhhJwgKEEHGAoAQcY6sg2HvTYsWN9/PjxbTw0wDXp6aeffrm7d/byM1sJ+PHjx/PUU09t46EBrklV9W97/RmXUACGEnCAoQQcYCgBBxhKwAGGEnCAoRa9jbCqvpvk1SRvJHm9u09scygANtvL+8B/rrtf3tokAOyJSygAQy09A+8kf1dVneRPu/vUuQdU1ckkJ5Nkd3d3/ybk0Kp66K3b3Q9e8jFwrVp6Bn53d9+R5ANJfrOq3nvuAd19qrtPdPeJnZ09fZwfgEuwKODd/b31r2eSPJLkzm0OBcBmGwNeVT9aVTe8eTvJzyd5btuDAXBxS66B/1iSR6rqzeP/qrv/ZqtTAbDRxoB393eSvPsAZgFgD7yNEGAoAQcYSsABhhJwgKEEHGAoAQcYSsABhhJwgKEEHGAoAQcYSsABhhJwgKEEHGAoAQcYSsABhhJwgKEEHGAoAQcYSsABhhJwgKEEHGAoAQcYSsABhhJwgKEEHGAoAQcYSsABhhJwgKEEHGAoAQcYSsABhhJwgKEWB7yqrquqf6qqR7c5EADL7OUM/IEkp7c1CAB7syjgVXVLkl9I8mfbHQeApY4sPO4Pkvx2khsudEBVnUxyMkl2d3cvezA4W9VDl3x894P7PQ5cFTaegVfVLyY5091PX+y47j7V3Se6+8TOzs6+DQjA+S25hHJ3kl+qqu8m+WySe6rqL7c6FQAbbQx4d/9ud9/S3ceT3Jfk77v7l7c+GQAX5X3gAEMtfREzSdLdTyR5YiuTALAnzsABhhJwgKEEHGAoAQcYSsABhhJwgKEEHGAoAQcYSsABhhJwgKEEHGAoAQcYSsABhhJwgKEEHGAoAQcYSsABhhJwgKEEHGAoAQcYSsABhhJwgKEEHGAoAQcYSsABhhJwgKEEHGAoAQcYSsABhhJwgKEEHGAoAQcYamPAq+rtVfX1qvpGVT1fVQ8dxGAAXNyRBcf8d5J7uvu1qjqa5KtV9Vh3f23LswFwERsD3t2d5LX1t0fXX73NoQDYbNE18Kq6rqqeTXImyePd/eRWpwJgoyWXUNLdbyR5T1XdmOSRqnpXdz939jFVdTLJySTZ3d3d7znhqnL2S0HdD17BSTjM9vQulO7+QZInktx7nvtOdfeJ7j6xs7OzP9MBcEFL3oWysz7zTlW9I8n7k3xry3MBsMGSSyg3JflUVV2XVfA/192PbncsADZZ8i6Ubya5/QBmAWAPfBITYCgBBxhKwAGGEnCAoQQcYCgBBxhKwAGGEnCAoQQcYCgBBxhKwAGGEnCAoQQcYCgBBxhKwAGGEnCAoQQcYCgBBxhKwAGGEnCAoQQcYCgBBxhKwAGGEnCAoQQcYCgBBxhKwAGGEnCAoQQcYCgBBxhKwAGGEnCAoQQcYKiNAa+qW6vqy1V1uqqer6oHDmIwAC7uyIJjXk/yW939TFXdkOTpqnq8u/9ly7MBcBEbz8C7+6XufmZ9+9Ukp5PcvO3BALi4JWfgb6mq40luT/Lkee47meRkkuzu7u7HbLDvqh5663b3gxuPOduFjocrZfGLmFV1fZLPJ/l4d79y7v3dfaq7T3T3iZ2dnf2cEYDzWBTwqjqaVbw/3d1f2O5IACyx5F0oleSTSU539ye2PxIASyw5A787yf1J7qmqZ9dfH9zyXABssPFFzO7+apI6gFkA2AOfxAQYSsABhhJwgKEEHGAoAQcYSsABhhJwgKEEHGAoAQcYSsABhhJwgKEEHGAoAQcYSsABhhJwgKEEHGAoAQcYSsABhhJwgKEEHGAoAQcYSsABhhJwgKEEHGAoAQcYSsABhhJwgKEEHGAoAQcYSsABhhJwgKEEHGCojQGvqoer6kxVPXcQAwGwzJIz8D9Pcu+W5wBgjzYGvLu/kuT7BzALAHtwZL8eqKpOJjmZJLu7u/v1sBwyVQ8d2GPu13Od/TjdD+7pmHNnuNDPw/ns24uY3X2qu09094mdnZ39elgALsC7UACGEnCAoZa8jfAzSf4hyTur6sWq+rXtjwXAJhtfxOzujxzEIADsjUsoAEMJOMBQAg4wlIADDCXgAEMJOMBQAg4wlIADDCXgAEMJOMBQAg4wlIADDCXgAEMJOMBQAg4wlIADDCXgAEMJOMBQAg4wlIADDCXgAEMJOMBQAg4wlIADDCXgAEMJOMBQAg4wlIADDCXgAEMJOMBQAg4wlIADDCXgAEMtCnhV3VtV366qF6rqd7Y9FACbbQx4VV2X5I+TfCDJbUk+UlW3bXswAC5uyRn4nUle6O7vdPf/JPlskg9tdywANqnuvvgBVR9Ocm93//r6+/uT/Ex3f+yc404mObn+9p1Jvp3kWJKX93vogazDinVYsQ4r1mHlzXX4ie7e2csPHllwTJ3n9/5f9bv7VJJT/+cHq57q7hN7GehaZB1WrMOKdVixDiuXsw5LLqG8mOTWs76/Jcn3LuXJANg/SwL+j0l+qqp+sqp+JMl9Sb643bEA2GTjJZTufr2qPpbkb5Ncl+Th7n5+4eOf2nzIoWAdVqzDinVYsQ4rl7wOG1/EBODq5JOYAEMJOMBQlx3wqnq4qs5U1XMXuL+q6g/XH8P/ZlXdcbnPeTVasA7vq6ofVtWz66/fP+gZD0JV3VpVX66q01X1fFU9cJ5jrvk9sXAdrvk9UVVvr6qvV9U31uvw0HmOOQz7Yck67H0/dPdlfSV5b5I7kjx3gfs/mOSxrN5PfleSJy/3Oa/GrwXr8L4kj17pOQ9gHW5Kcsf69g1J/jXJbYdtTyxch2t+T6z/jK9f3z6a5Mkkdx3C/bBkHfa8Hy77DLy7v5Lk+xc55ENJ/qJXvpbkxqq66XKf92qzYB0Ohe5+qbufWd9+NcnpJDefc9g1vycWrsM1b/1n/Nr626Prr3PfOXEY9sOSddizg7gGfnOSfz/r+xdzCDfy2s+u/wr1WFX99JUeZtuq6niS27M62zjbodoTF1mH5BDsiaq6rqqeTXImyePdfSj3w4J1SPa4Hw4i4Is+in8IPJPVv3Xw7iR/lOSvr+w421VV1yf5fJKPd/cr5959nh+5JvfEhnU4FHuiu9/o7vdk9SnuO6vqXecccij2w4J12PN+OIiA+yh+ku5+5c2/QnX3l5IcrapjV3israiqo1lF69Pd/YXzHHIo9sSmdThMeyJJuvsHSZ5Icu85dx2K/fCmC63DpeyHgwj4F5P8yvqV5ruS/LC7XzqA572qVNWPV1Wtb9+Z1dr/x5Wdav+t/xs/meR0d3/iAodd83tiyTochj1RVTtVdeP69juSvD/Jt8457DDsh43rcCn7Ycm/RrhpsM9k9erpsap6McmDWV2gT3f/SZIvZfUq8wtJ/jPJr17uc16NFqzDh5P8RlW9nuS/ktzX65eerzF3J7k/yT+vr/clye8l2U0O1Z5Ysg6HYU/clORTtfofw7wtyee6+9Gq+mhyqPbDknXY837wUXqAoXwSE2AoAQcYSsABhhJwgKEEHGAoAQcYSsABhvpfsvur3aL9XpsAAAAASUVORK5CYII=\n",
281 |       "text/plain": [
282 |        "<Figure size 432x288 with 1 Axes>"
283 |       ]
284 |      },
285 |      "metadata": {
286 |       "needs_background": "light"
287 |      },
288 |      "output_type": "display_data"
289 |     }
290 |    ],
291 |    "source": [
292 |     "# Same approach as above, though now limited to only those contact bins containing microcompartment anchors\n",
293 |     "# Usage: Plot different datasets & regions by updating the filename, \"region = regionsRCMC[x]\", and \"anchorReg = x\" attributes \n",
294 |     "\n",
295 |     "saveDir = f'DIRECTORY_PATH'\n",
296 |     "fileName = f'FILE_NAME.svg' # swap out .svg for the desired output file format\n",
297 |     "\n",
298 |     "# Specify the data resolution at which balanced rowsums will be calculated\n",
299 |     "resolution = 250\n",
300 |     "\n",
301 |     "# Load in the desired dataset\n",
302 |     "clr_Data = cooler.Cooler('FILE_PATH_TO_MCOOL::resolutions/'+str(resolution))\n",
303 |     "\n",
304 |     "# List the Capture regions for region-by-region calculations\n",
305 |     "## For an apples-to-apples comparison, balanced rowsums are calculated at the RCMC regions for both the RCMC dataset as well as the genome-wide Micro-C (Hsieh 2020) and Hi-C (Bonev 2017) datasets\n",
306 |     "regionsRCMC = ['chr3:33804149-35704149','chr5:31257344-32382344','chr6:122451959-122876959','chr8:84846629-85856629','chr18:58032072-59034072']\n",
307 |     "\n",
308 |     "# Specify which specific region you would like to calculate rowsums for\n",
309 |     "region = regionsRCMC[1]\n",
310 |     "\n",
311 |     "# Load in the balanced contact matrix for the specified region\n",
312 |     "regionMat = clr_Data.matrix(balance=True).fetch(region)\n",
313 |     "\n",
314 |     "# Specify the region of interest (either Klf1 or Ppm1g) for which to load in microcompartment anchors\n",
315 |     "anchorReg = 'Ppm1g'\n",
316 |     "if anchorReg == 'Klf1':\n",
317 |     "    anchors = anchorListKlf1\n",
318 |     "    offset = 84846629 # This offset is the start coordinate of the Klf1 region, allowing genomic coordinate conversion to bins\n",
319 |     "if anchorReg == 'Ppm1g':\n",
320 |     "    anchors = anchorListPpm1g\n",
321 |     "    offset = 31257344 # This offset is the start coordinate of the Ppm1g region, allowing genomic coordinate conversion to bins\n",
322 |     "\n",
323 |     "# Create a list of bins which contain microcompartment anchors\n",
324 |     "bins = []\n",
325 |     "for i in anchors:\n",
326 |     "    bins.append((i - offset) // resolution)\n",
327 |     "\n",
328 |     "# Create arrays of zeroes corresponding to the lengths (rows & columns) of the loaded in contact matrices\n",
329 |     "## The code here calculates both the rowsums and columnsums; they should be the same distribution, so you only need 1 of the 2\n",
330 |     "rowSum = np.zeros(len(regionMat))\n",
331 |     "columnSum = np.zeros(len(regionMat[0]))\n",
332 |     "\n",
333 |     "\n",
334 |     "# Iterate through the matrix bin-by-bin, summing up the balanced contact bin values for each row & column that contains a microcompartment anchor\n",
335 |     "## Only need to iterate through one half of the contact matrix (i.e., the half above the diagonal or the half below it) to capture it all due to reflected symmetry\n",
336 |     "i = 0\n",
337 |     "while i < len(regionMat):\n",
338 |     "    j = i # Ensures that we keep our iteration to one side of the diagonal instead of spanning the whole matrix \n",
339 |     "    while j < len(regionMat[0]):\n",
340 |     "        if((i in bins) or (j in bins)): # The only difference from the calculation for the whole-region calculation, this if statement ensures that rowsums and columnsumes are only calculated for bins containing a microcompartment anchor            \n",
341 |     "            val = regionMat[i][j] # Extracts the balanced value of the contact bin\n",
342 |     "            if np.isnan(val):\n",
343 |     "                val = 0 # NaN values can cause calculation errors, so this sets them to 0\n",
344 |     "            rowSum[i] += val # Adds the balanced value to the corresponding rowsum\n",
345 |     "            columnSum[j] += val # Adds the balanced value to the corresponding columnsum\n",
346 |     "\n",
347 |     "            if j != i: # If j = i, then we're on the diagonal & there's no need to add in the reflected twin of the contact bin\n",
348 |     "                rowSum[j] += val # Adds the balanced value of the reflected twin to the corresponding rowsum\n",
349 |     "                columnSum[i] += val # Adds the balanced value of the reflected twin to the corresponding columnsum\n",
350 |     "        j += 1\n",
351 |     "    i += 1\n",
352 |     "\n",
353 |     "\n",
354 |     "# Plotting the rowsums and columnsums yields the same distribution, so this just runs ahead with the rowsums\n",
355 |     "## This selects for bins containing microcompartment anchors and filters out the rowsums that are 0\n",
356 |     "## Note: Filtering for non-zero bins likely changes nothing since microcompartment anchor-containing bins are guaranteed not empty. Keeping it nonetheless for homogeneity with the whole-region calculation.\n",
357 |     "rowSumFilt = []\n",
358 |     "for x in bins:\n",
359 |     "    microcompRowSum = rowSum[x] # Extract the rowsum value of a microcompartment anchor-containing bin\n",
360 |     "    if microcompRowSum > 0:\n",
361 |     "        rowSumFilt.append(microcompRowSum) # Add the rowsum to the list of microcompartment rowsums if its greater than 0\n",
362 |     "    rowSumFilt.append()\n",
363 |     "\n",
364 |     "    \n",
365 |     "# Plot a histogram of the rowsums\n",
366 |     "## Change the number of histogram bins as desired (100 used here)\n",
367 |     "## Change the range of visualization to capture and visually center the entire distribution\n",
368 |     "# plt.hist(rowSumFilt,100,range=[3.75,8.25],align='mid',color='navy') # Klf1\n",
369 |     "plt.hist(rowSumFilt,100,range=[1.1,3.4],align='mid',color='navy') # Ppm1g\n",
370 |     "\n",
371 |     "# plt.savefig(saveDir + fileName, dpi=1200)"
372 |    ]
373 |   },
374 |   {
375 |    "cell_type": "code",
376 |    "execution_count": null,
377 |    "metadata": {},
378 |    "outputs": [],
379 |    "source": []
380 |   }
381 |  ],
382 |  "metadata": {
383 |   "kernelspec": {
384 |    "display_name": "Python [conda env:coolToolsEnv]",
385 |    "language": "python",
386 |    "name": "conda-env-coolToolsEnv-py"
387 |   },
388 |   "language_info": {
389 |    "codemirror_mode": {
390 |     "name": "ipython",
391 |     "version": 3
392 |    },
393 |    "file_extension": ".py",
394 |    "mimetype": "text/x-python",
395 |    "name": "python",
396 |    "nbconvert_exporter": "python",
397 |    "pygments_lexer": "ipython3",
398 |    "version": "3.8.12"
399 |   }
400 |  },
401 |  "nbformat": 4,
402 |  "nbformat_minor": 4
403 | }
404 | 


--------------------------------------------------------------------------------
/CalculatingFilledBinFractionByDistance.ipynb:
--------------------------------------------------------------------------------
  1 | {
  2 |  "cells": [
  3 |   {
  4 |    "cell_type": "markdown",
  5 |    "metadata": {},
  6 |    "source": [
  7 |     "# Counting fraction of filled bins across contact distances"
  8 |    ]
  9 |   },
 10 |   {
 11 |    "cell_type": "code",
 12 |    "execution_count": 2,
 13 |    "metadata": {},
 14 |    "outputs": [],
 15 |    "source": [
 16 |     "# import core packages\n",
 17 |     "import warnings\n",
 18 |     "warnings.filterwarnings(\"ignore\")\n",
 19 |     "from itertools import combinations\n",
 20 |     "\n",
 21 |     "import matplotlib.pyplot as plt\n",
 22 |     "from matplotlib import colors\n",
 23 |     "%matplotlib inline\n",
 24 |     "plt.style.use('seaborn-poster')\n",
 25 |     "import numpy as np\n",
 26 |     "import pandas as pd\n",
 27 |     "\n",
 28 |     "# import open2c libraries\n",
 29 |     "import bioframe\n",
 30 |     "import cooler\n",
 31 |     "import cooltools\n",
 32 |     "\n",
 33 |     "from packaging import version\n",
 34 |     "if version.parse(cooltools.__version__) < version.parse('0.5.0'):\n",
 35 |     "    raise AssertionError(\"tutorials rely on cooltools version 0.5.0 or higher,\"+\n",
 36 |     "                         \"please check your cooltools version and update to the latest\")\n",
 37 |     "\n",
 38 |     "    \n",
 39 |     "resolution = 100 # input the desired data resolution for analysis\n",
 40 |     "distances = [1000,2500,5000,10000,25000,50000,100000,250000,500000,1000000] # input the desired contact distance diagonals in bp\n",
 41 |     "\n",
 42 |     "# Load in each of the datasets at the specified resolution\n",
 43 |     "clr_RCMC = cooler.Cooler('FILE_PATH_TO_MCOOL::resolutions/' + str(resolution))\n",
 44 |     "clr_TMCC = cooler.Cooler('FILE_PATH_TO_MCOOL::resolutions/' + str(resolution))\n",
 45 |     "clr_Hsieh = cooler.Cooler('FILE_PATH_TO_MCOOL::resolutions/' + str(resolution))\n",
 46 |     "clr_Bonev = cooler.Cooler('FILE_PATH_TO_MCOOL::resolutions/' + str(resolution))\n",
 47 |     "\n",
 48 |     "# List the Capture regions for RCMC & TMCC\n",
 49 |     "## To ensure an apples-to-apples comparison, the genome-wide Micro-C (Hsieh 2020) and Hi-C (Bonev 2017) datasets will be analyzed for the same regions as RCMC instead of across the entire genome\n",
 50 |     "regionsRCMC = ['chr3:33804149-35704149','chr5:31257344-32382344','chr6:122451959-122876959','chr8:84846629-85856629','chr18:58032072-59034072']\n",
 51 |     "regionsTMCC = ['chr3:34365200-35640000','chr6:122606805-122856450']\n",
 52 |     "\n",
 53 |     "\n",
 54 |     "# Variables for storing fill fraction data\n",
 55 |     "countsAcrossDistances = [] # Stores the fraction of all non-empty bins for each dataset (across all regions) for each of the specified contact distances\n",
 56 |     "nonZeroFracsByRegionAcrossDistances = [] # Stores the fraction of all non-empty bins for each dataset (region-separated) for each of the specified contact distances\n",
 57 |     "\n",
 58 |     "for distance in distances:\n",
 59 |     "    compiledCounts = [] # Stores the total number of bins and number of non-empty bins for each region of each dataset at the given contact distance\n",
 60 |     "\n",
 61 |     "    coolers = [clr_RCMC,clr_TMCC,clr_Hsieh,clr_Bonev]\n",
 62 |     "    coolerIt = 0 # An iterating variable for going through each of the datasets\n",
 63 |     "    while coolerIt < len(coolers):\n",
 64 |     "        if coolerIt == 1:\n",
 65 |     "            regions = regionsTMCC # Use the TMCC regions for calculation for the TMCC dataset\n",
 66 |     "        else:\n",
 67 |     "            regions = regionsRCMC # Use the RCMC regions for calculation for the RCMC, Hsieh, & Bonev datasets\n",
 68 |     "\n",
 69 |     "        totBins = [] # Counts the number of total contact bins for at the given contact distance\n",
 70 |     "        nonZeroBins = [] # Counts the number of contact bins that have at least one read in them at the given contact distance\n",
 71 |     "\n",
 72 |     "        for i in regions:        \n",
 73 |     "            # Iterating region by region for the given dataset, load in the contact matrix\n",
 74 |     "            # Keep balancing off to count raw read counts\n",
 75 |     "            regionMat = coolers[coolerIt].matrix(balance=False).fetch(i)\n",
 76 |     "            \n",
 77 |     "            bins = len(regionMat) # Get the number of genomic bins across the matrix\n",
 78 |     "            offset = distance // resolution - 1 # Calculate the number of fewer genomic bins at the given contact distance\n",
 79 |     "            binsAtDiag = bins - offset # Calculate the total number of contact bins at the contact distance\n",
 80 |     "            totBins.append(binsAtDiag)\n",
 81 |     "\n",
 82 |     "            # Set the stage for iterating across the contact diagonal at the given distance to count the number of empty bins\n",
 83 |     "            x = offset\n",
 84 |     "            y = 0\n",
 85 |     "            nonZeroCount = 0\n",
 86 |     "            \n",
 87 |     "            # Iterate through the contact diagonal tallying non-empty bins until the end of the diagonal\n",
 88 |     "            while x < len(regionMat):\n",
 89 |     "                val = regionMat[y][x]\n",
 90 |     "                if val > 0:\n",
 91 |     "                    nonZeroCount += 1\n",
 92 |     "                x += 1\n",
 93 |     "                y += 1\n",
 94 |     "            nonZeroBins.append(nonZeroCount)\n",
 95 |     "\n",
 96 |     "        binCounts = [totBins,nonZeroBins] # Combine the total bin & non-empty bin counts into a tuple\n",
 97 |     "        compiledCounts.append(binCounts) # Add the combined tuple for the given dataset & region to compiledCounts\n",
 98 |     "\n",
 99 |     "        coolerIt += 1 # Onwards to the next cooler in the dataset list\n",
100 |     "\n",
101 |     "    # print(compiledCounts) # Sanity check that the code is working \n",
102 |     "\n",
103 |     "    totNonZeroFrac = [] # Stores the across-regions fraction of non-empty bins for each dataset at a given contact distance\n",
104 |     "    nonZeroFracsByRegion = [] # The region-separated version of totNonZeroFrac\n",
105 |     "\n",
106 |     "    # Iterate through compiledCounts to extract filled bin fractions\n",
107 |     "    for i in compiledCounts:\n",
108 |     "        sumPossible = sum(i[0]) # Sums the total number of bins across all regions for a dataset\n",
109 |     "        sumNonZero = sum(i[1]) # Sums the total number of non-empty bins across all regions for a dataset\n",
110 |     "        totNonZeroFrac.append(sumNonZero / sumPossible) # Adds the non-empty bin fraction to totNonZeroFrac\n",
111 |     "\n",
112 |     "        regionFracs = [] # An array to hold the non-empty bin fraction by region for a given dataset\n",
113 |     "        for j in range(len(i[0])): # Iterate across the regions\n",
114 |     "            fracNonZero = i[1][j] / i[0][j] # Calculate non-empty bin fraction for the given region\n",
115 |     "            regionFracs.append(fracNonZero) # Add the calculated fraction to regionFracs\n",
116 |     "        nonZeroFracsByRegion.append(regionFracs) # Add the array of non-empty bin fractions by region to nonZeroFracsByRegion\n",
117 |     "\n",
118 |     "    # print(totNonZeroFrac) # Double-check that totNonZeroFrac is calculating as expected\n",
119 |     "    # print(nonZeroFracsByRegion) # Double-check that nonZeroFracsByRegion is calculating as expected\n",
120 |     "    \n",
121 |     "    countsAcrossDistances.append(totNonZeroFrac) # Add totNonZeroFrac, which contains the non-zero bin fraction for each dataset at the given contact distance, to countsAcrossDistances\n",
122 |     "    nonZeroFracsByRegionAcrossDistances.append(nonZeroFracsByRegion) # Add nonZeroFracsByRegion, which contains the region-separated non-zero bin fraction for each dataset at the given contact distance, to nonZeroFracsByRegionAcrossDistances\n",
123 |     "    \n",
124 |     "countsAcrossDistances # Print the output of countsAcrossDistances to confirm that it produces a logical result"
125 |    ]
126 |   },
127 |   {
128 |    "cell_type": "code",
129 |    "execution_count": 3,
130 |    "metadata": {},
131 |    "outputs": [],
132 |    "source": [
133 |     "# Separate the nested array of countsAcrossDistance by dataset for ease of calculation & plotting\n",
134 |     "RCMCcounts = []\n",
135 |     "TMCCcounts = []\n",
136 |     "Hsiehcounts = []\n",
137 |     "Bonevcounts = []\n",
138 |     "\n",
139 |     "for i in countsAcrossDistances:\n",
140 |     "    RCMCcounts.append(i[0])\n",
141 |     "    TMCCcounts.append(i[1])\n",
142 |     "    Hsiehcounts.append(i[2])\n",
143 |     "    Bonevcounts.append(i[3])"
144 |    ]
145 |   },
146 |   {
147 |    "cell_type": "code",
148 |    "execution_count": 6,
149 |    "metadata": {},
150 |    "outputs": [],
151 |    "source": [
152 |     "# Separate out the nested array of nonZeroFracsByRegionAcrossDistances to extract the RCMC & TMCC region counts for ease of calculation & plotting\n",
153 |     "RCMCRegionCounts = []\n",
154 |     "TMCCRegionCounts = []\n",
155 |     "\n",
156 |     "for distanceSet in nonZeroFracsByRegionAcrossDistances:\n",
157 |     "    RCMCRegionCounts.append(distanceSet[0])\n",
158 |     "    TMCCRegionCounts.append(distanceSet[1])"
159 |    ]
160 |   },
161 |   {
162 |    "cell_type": "code",
163 |    "execution_count": 7,
164 |    "metadata": {},
165 |    "outputs": [],
166 |    "source": [
167 |     "# Separate out the nested array of region counts by individual region for RCMC & TMCC for ease of plotting\n",
168 |     "RCMC_Sox2 = []\n",
169 |     "RCMC_Ppm1g = []\n",
170 |     "RCMC_Nanog = []\n",
171 |     "RCMC_Klf1 = []\n",
172 |     "RCMC_Fbn2 = []\n",
173 |     "TMCC_Sox2 = []\n",
174 |     "TMCC_Nanog = []\n",
175 |     "\n",
176 |     "for i in RCMCRegionCounts:\n",
177 |     "    RCMC_Sox2.append(i[0])\n",
178 |     "    RCMC_Ppm1g.append(i[1])\n",
179 |     "    RCMC_Nanog.append(i[2])\n",
180 |     "    RCMC_Klf1.append(i[3])\n",
181 |     "    RCMC_Fbn2.append(i[4])\n",
182 |     "    \n",
183 |     "for j in TMCCRegionCounts:\n",
184 |     "    TMCC_Sox2.append(j[0])\n",
185 |     "    TMCC_Nanog.append(j[1])"
186 |    ]
187 |   },
188 |   {
189 |    "cell_type": "code",
190 |    "execution_count": 21,
191 |    "metadata": {},
192 |    "outputs": [
193 |     {
194 |      "data": {
195 |       "image/png": "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\n",
196 |       "text/plain": [
197 |        "<Figure size 921.6x633.6 with 1 Axes>"
198 |       ]
199 |      },
200 |      "metadata": {
201 |       "needs_background": "light"
202 |      },
203 |      "output_type": "display_data"
204 |     }
205 |    ],
206 |    "source": [
207 |     "# Plot the fraction of filled bins across contact distances both by region and averaged across regions\n",
208 |     "\n",
209 |     "import matplotlib.pyplot as plt\n",
210 |     "\n",
211 |     "saveDir = f'DIRECTORY_PATH'\n",
212 |     "fileName = f'FILE_NAME.svg' # swap out .svg for the desired output file format\n",
213 |     "\n",
214 |     "plt.scatter(distances, RCMCcounts, marker = 'D', c = 'blue', label='RCMC')\n",
215 |     "plt.scatter(distances, RCMC_Sox2, marker = 'x', c = 'green', label='RCMC_Sox2')\n",
216 |     "plt.scatter(distances, RCMC_Ppm1g, marker = 'x', c = 'skyblue', label='RCMC_Ppm1g')\n",
217 |     "plt.scatter(distances, RCMC_Nanog, marker = 'x', c = 'teal', label='RCMC_Nanog')\n",
218 |     "plt.scatter(distances, RCMC_Klf1, marker = 'x', c = 'black', label='RCMC_Klf1')\n",
219 |     "plt.scatter(distances, RCMC_Fbn2, marker = 'x', c = 'cyan', label='RCMC_Fbn2')\n",
220 |     "plt.scatter(distances, TMCCcounts, marker = 's', c = 'red', label='TMCC')\n",
221 |     "plt.scatter(distances, TMCC_Sox2, marker = '+', c = 'purple', label='TMCC_Sox2')\n",
222 |     "plt.scatter(distances, TMCC_Nanog, marker = '+', c = 'pink', label='TMCC_Nanog')\n",
223 |     "plt.scatter(distances, Hsiehcounts, marker = 'v', c = 'orange', label='Hsieh')\n",
224 |     "plt.scatter(distances, Bonevcounts, marker = 'o', c = 'gray', label='Bonev')\n",
225 |     "plt.xscale(\"log\")\n",
226 |     "plt.xlabel(\"Contact Distance (bp) at 100 bp Resolution\")\n",
227 |     "plt.ylabel(\"Fraction of Nonzero Bins\")\n",
228 |     "plt.legend()\n",
229 |     "\n",
230 |     "# plt.savefig(saveDir + fileName, dpi=1200)\n",
231 |     "plt.show()"
232 |    ]
233 |   },
234 |   {
235 |    "cell_type": "code",
236 |    "execution_count": null,
237 |    "metadata": {},
238 |    "outputs": [],
239 |    "source": []
240 |   }
241 |  ],
242 |  "metadata": {
243 |   "kernelspec": {
244 |    "display_name": "Python [conda env:coolToolsEnv]",
245 |    "language": "python",
246 |    "name": "conda-env-coolToolsEnv-py"
247 |   },
248 |   "language_info": {
249 |    "codemirror_mode": {
250 |     "name": "ipython",
251 |     "version": 3
252 |    },
253 |    "file_extension": ".py",
254 |    "mimetype": "text/x-python",
255 |    "name": "python",
256 |    "nbconvert_exporter": "python",
257 |    "pygments_lexer": "ipython3",
258 |    "version": "3.8.12"
259 |   }
260 |  },
261 |  "nbformat": 4,
262 |  "nbformat_minor": 5
263 | }
264 | 


--------------------------------------------------------------------------------
/LICENSE:
--------------------------------------------------------------------------------
 1 | MIT License
 2 | 
 3 | Copyright (c) 2023 Hansen lab at MIT
 4 | 
 5 | Permission is hereby granted, free of charge, to any person obtaining a copy
 6 | of this software and associated documentation files (the "Software"), to deal
 7 | in the Software without restriction, including without limitation the rights
 8 | to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
 9 | copies of the Software, and to permit persons to whom the Software is
10 | furnished to do so, subject to the following conditions:
11 | 
12 | The above copyright notice and this permission notice shall be included in all
13 | copies or substantial portions of the Software.
14 | 
15 | THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
16 | IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
17 | FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
18 | AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
19 | LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
20 | OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
21 | SOFTWARE.
22 | 


--------------------------------------------------------------------------------
/LoopStrengthRCMC.ipynb:
--------------------------------------------------------------------------------
  1 | {
  2 |  "cells": [
  3 |   {
  4 |    "cell_type": "markdown",
  5 |    "id": "abd813be",
  6 |    "metadata": {},
  7 |    "source": [
  8 |     "# Loop strength calculation"
  9 |    ]
 10 |   },
 11 |   {
 12 |    "cell_type": "markdown",
 13 |    "id": "dfcad04e",
 14 |    "metadata": {},
 15 |    "source": [
 16 |     "## Imports"
 17 |    ]
 18 |   },
 19 |   {
 20 |    "cell_type": "code",
 21 |    "execution_count": null,
 22 |    "id": "73bb3370",
 23 |    "metadata": {},
 24 |    "outputs": [],
 25 |    "source": [
 26 |     "# import standard python libraries\n",
 27 |     "import matplotlib as mpl\n",
 28 |     "%matplotlib inline\n",
 29 |     "mpl.rcParams['figure.dpi'] = 96\n",
 30 |     "import numpy as np\n",
 31 |     "import matplotlib.pyplot as plt\n",
 32 |     "import pandas as pd\n",
 33 |     "import seaborn as sns\n",
 34 |     "\n",
 35 |     "# import libraries for biological data analysis\n",
 36 |     "from coolpuppy import coolpup\n",
 37 |     "from plotpuppy import plotpup\n",
 38 |     "import cooler\n",
 39 |     "import bioframe\n",
 40 |     "import cooltools\n",
 41 |     "from cooltools import expected_cis\n",
 42 |     "from cooltools.lib import plotting\n",
 43 |     "\n",
 44 |     "import bbi"
 45 |    ]
 46 |   },
 47 |   {
 48 |    "cell_type": "markdown",
 49 |    "id": "e71fbaa6",
 50 |    "metadata": {},
 51 |    "source": [
 52 |     "## Inputs"
 53 |    ]
 54 |   },
 55 |   {
 56 |    "cell_type": "markdown",
 57 |    "id": "70c700df",
 58 |    "metadata": {},
 59 |    "source": [
 60 |     "First, get the loops and mcools to analyse, set up variables, etc"
 61 |    ]
 62 |   },
 63 |   {
 64 |    "cell_type": "code",
 65 |    "execution_count": null,
 66 |    "id": "068a8eac",
 67 |    "metadata": {},
 68 |    "outputs": [],
 69 |    "source": [
 70 |     "#mcool resolution to read\n",
 71 |     "resolution = 250\n",
 72 |     "#List of mcool locations as strings\n",
 73 |     "conditions = [\"mcoollocation1\", \"mcoollocation2\", \"mcoollocation3\"]\n",
 74 |     "#List of loop types as strings\n",
 75 |     "loopTypesNames = [\"loop\", \"type\", \"names\"]\n",
 76 |     "#List of loop file locations (bedpe)\n",
 77 |     "loopFiles = [\"looplocation1\", \"looplocation2\", \"looplocation3\"]\n",
 78 |     "\n",
 79 |     "#Specify the RCMC regions of the mcools to look at (format: chromosome (string), start (number), end (number), name of region (string))\n",
 80 |     "regions = pd.DataFrame([['chrA',1,100,'regionname1'],['chrB',1,100,'regionname2'],['chrC',1,100,'regionname3']],\n",
 81 |     "                  columns=['chrom', 'start', 'end', 'name'])\n",
 82 |     "#Cis expected file locations from cooltools - .tsv file - one for each mcool\n",
 83 |     "expectedFiles = [\"expectedlocation1\", \"expectedlocation2\", \"expectedlocation3\"]\n",
 84 |     "#Set save directory\n",
 85 |     "saveDir = '/a/directory/on/your/system/'\n",
 86 |     "\n",
 87 |     "#Set the size of the area flanking the dot\n",
 88 |     "flankDist = 10000\n",
 89 |     "#Don't set this to be even... This is the size of the area to measure around the dot \n",
 90 |     "#(and by extension the size of the boxes at the edges of the region too)\n",
 91 |     "#For this reason, it needs to be odd to have integer box sizes on each side.\n",
 92 |     "dotWindow = 5\n"
 93 |    ]
 94 |   },
 95 |   {
 96 |    "cell_type": "markdown",
 97 |    "id": "e7cf57de",
 98 |    "metadata": {},
 99 |    "source": [
100 |     "Run the imports"
101 |    ]
102 |   },
103 |   {
104 |    "cell_type": "code",
105 |    "execution_count": null,
106 |    "id": "d76f2cee",
107 |    "metadata": {},
108 |    "outputs": [],
109 |    "source": [
110 |     "#######Don't change this section#######\n",
111 |     "#Creat an empty list to store the imported loop locations\n",
112 |     "loopTypes = []\n",
113 |     "#List of column names to use for imported loops (this is constant - do not change)\n",
114 |     "colNames = ['chrom1', 'start1', 'end1', 'chrom2', 'start2', 'end2']\n",
115 |     "#Read in files, put them in loopTypes\n",
116 |     "for file in loopFiles:\n",
117 |     "    temploops = pd.read_csv(file, sep='\\t', names=colNames, header=None)\n",
118 |     "    loopTypes.append(temploops)"
119 |    ]
120 |   },
121 |   {
122 |    "cell_type": "markdown",
123 |    "id": "3ba156f7",
124 |    "metadata": {},
125 |    "source": [
126 |     "## Enrichment calculation"
127 |    ]
128 |   },
129 |   {
130 |    "cell_type": "code",
131 |    "execution_count": null,
132 |    "id": "e4a3696a",
133 |    "metadata": {},
134 |    "outputs": [],
135 |    "source": [
136 |     "## Enrichment calculation\n",
137 |     "\n",
138 |     "#Viraat's new calculation\n",
139 |     "#Modified 2022/10/04 by Miles to try to avoid NaN values and correct an issue with the background sum, \n",
140 |     "#and generally make the code a little more streamlined\n",
141 |     "def enrichmentCalc(mtx, dotWindow):\n",
142 |     "    #Dimension of array side (should be square)\n",
143 |     "    sideLength = len(mtx)\n",
144 |     "    #Middle of side length\n",
145 |     "    midPoint = (sideLength - 1) // 2\n",
146 |     "    #Half size of box around centre pixel (one pixel smaller if even-sized dot window - don't do this)\n",
147 |     "    buffer = (dotWindow - 1) // 2\n",
148 |     "    \n",
149 |     "    #Get sum of pixels around dot\n",
150 |     "    dotSum = np.nansum(mtx[midPoint-buffer:midPoint+buffer+1, midPoint-buffer:midPoint+buffer+1])\n",
151 |     "    \n",
152 |     "    #Subset the matrix and calculate the mean without NaN values\n",
153 |     "    backgroundSum1 = np.nansum(mtx[0:dotWindow, 0:dotWindow])\n",
154 |     "    backgroundSum2 = np.nansum(mtx[sideLength-dotWindow:sideLength, sideLength-dotWindow:sideLength])\n",
155 |     "    \n",
156 |     "    #Calculate enrichment (NB this assumes all boxes are the same size.\n",
157 |     "    #If you set an even dotWindow value, they won't be)\n",
158 |     "    enrichment = dotSum / ((backgroundSum1 + backgroundSum2)/2)\n",
159 |     "    \n",
160 |     "    return enrichment"
161 |    ]
162 |   },
163 |   {
164 |    "cell_type": "markdown",
165 |    "id": "5e1e9fad",
166 |    "metadata": {},
167 |    "source": [
168 |     "# Get the strengths"
169 |    ]
170 |   },
171 |   {
172 |    "cell_type": "markdown",
173 |    "id": "666fd62a",
174 |    "metadata": {},
175 |    "source": [
176 |     "Function for getting strength of each loop (uses the pileup function from cooltools to do observed/expected)"
177 |    ]
178 |   },
179 |   {
180 |    "cell_type": "code",
181 |    "execution_count": null,
182 |    "id": "afd62abd",
183 |    "metadata": {},
184 |    "outputs": [],
185 |    "source": [
186 |     "def loopStrengthGet(loop, flankDist, clr, regions, expected, dotWindow):\n",
187 |     "    loopf = loop.to_frame().T\n",
188 |     "    loopf = loopf.astype({'start1':'int64','end1':'int64','start2':'int64','end2':'int64'})\n",
189 |     "    stack = cooltools.pileup(clr, loopf, view_df=regions, expected_df=expected, flank=flankDist)\n",
190 |     "    mtx = np.nanmean(stack, axis=2)\n",
191 |     "    enrichment = enrichmentCalc(mtx, dotWindow)\n",
192 |     "    \n",
193 |     "    return enrichment"
194 |    ]
195 |   },
196 |   {
197 |    "cell_type": "code",
198 |    "execution_count": null,
199 |    "id": "01ca3113",
200 |    "metadata": {
201 |     "scrolled": true
202 |    },
203 |    "outputs": [],
204 |    "source": [
205 |     "#Zip the names and loop info into a dictionary for easier referencing\n",
206 |     "loopDict = dict(zip(loopTypesNames, loopTypes))\n",
207 |     "#Stop the code if you used an even value for dotWindow, since it won't work\n",
208 |     "if dotWindow % 2 == 0:\n",
209 |     "    print(\"You need to use an odd number for dotWindow in the inputs section\")\n",
210 |     "else:\n",
211 |     "    #Loop through the conditions\n",
212 |     "    for i, condition in enumerate(conditions):\n",
213 |     "        #Get the cooler data\n",
214 |     "        clr = cooler.Cooler(condition+'::/resolutions/'+str(resolution))\n",
215 |     "        #Get the corresponding expected data\n",
216 |     "        expected = pd.read_csv(expectedFiles[i], sep='\\t')\n",
217 |     "\n",
218 |     "        #Loop through loopDict\n",
219 |     "        for loopsName in loopDict:\n",
220 |     "            #Read out the loops\n",
221 |     "            loops = loopDict[loopsName]\n",
222 |     "            #For each row (ie loop), do pileup, get enrichment, write to new column [condition]_strength\n",
223 |     "            loops[f'{condition}_strength'] = loops.apply(loopStrengthGet, axis = 1, flankDist = flankDist, clr = clr, regions = regions, expected = expected, dotWindow = dotWindow)\n",
224 |     "\n",
225 |     "            loopDict[loopsName] = loops\n",
226 |     "        "
227 |    ]
228 |   },
229 |   {
230 |    "cell_type": "markdown",
231 |    "id": "bdc3359f",
232 |    "metadata": {},
233 |    "source": [
234 |     "## Output files - one for each loop type"
235 |    ]
236 |   },
237 |   {
238 |    "cell_type": "code",
239 |    "execution_count": null,
240 |    "id": "50c922d5",
241 |    "metadata": {},
242 |    "outputs": [],
243 |    "source": [
244 |     "for name, df in loopDict.items():\n",
245 |     "    df.to_csv(saveDir + name + '.bedpe', sep = '\\t', index = False, header = True)"
246 |    ]
247 |   }
248 |  ],
249 |  "metadata": {
250 |   "kernelspec": {
251 |    "display_name": "Python [conda env:coolpuppy]",
252 |    "language": "python",
253 |    "name": "conda-env-coolpuppy-py"
254 |   },
255 |   "language_info": {
256 |    "codemirror_mode": {
257 |     "name": "ipython",
258 |     "version": 3
259 |    },
260 |    "file_extension": ".py",
261 |    "mimetype": "text/x-python",
262 |    "name": "python",
263 |    "nbconvert_exporter": "python",
264 |    "pygments_lexer": "ipython3",
265 |    "version": "3.7.12"
266 |   }
267 |  },
268 |  "nbformat": 4,
269 |  "nbformat_minor": 5
270 | }
271 | 


--------------------------------------------------------------------------------
/MicrocompartmentLoops_PlusMin1kb.bedpe:
--------------------------------------------------------------------------------
   1 | chr5	31273000	31275000	chr5	31336200	31338200
   2 | chr5	31273000	31275000	chr5	31349200	31351200
   3 | chr5	31273000	31275000	chr5	31358700	31360700
   4 | chr5	31273000	31275000	chr5	31376700	31378700
   5 | chr5	31273000	31275000	chr5	31397100	31399100
   6 | chr5	31273000	31275000	chr5	31408000	31410000
   7 | chr5	31273000	31275000	chr5	31447500	31449500
   8 | chr5	31273000	31275000	chr5	31452200	31454200
   9 | chr5	31273000	31275000	chr5	31608600	31610600
  10 | chr5	31273000	31275000	chr5	31645300	31647300
  11 | chr5	31273000	31275000	chr5	31651000	31653000
  12 | chr5	31273000	31275000	chr5	31787500	31789500
  13 | chr5	31273000	31275000	chr5	31791000	31793000
  14 | chr5	31273000	31275000	chr5	31854000	31856000
  15 | chr5	31273000	31275000	chr5	32015400	32017400
  16 | chr5	31273000	31275000	chr5	32291000	32293000
  17 | chr5	31295500	31297500	chr5	31336200	31338200
  18 | chr5	31295500	31297500	chr5	31349200	31351200
  19 | chr5	31295500	31297500	chr5	31358700	31360700
  20 | chr5	31295500	31297500	chr5	31376700	31378700
  21 | chr5	31295500	31297500	chr5	31397100	31399100
  22 | chr5	31295500	31297500	chr5	31408000	31410000
  23 | chr5	31295500	31297500	chr5	31447500	31449500
  24 | chr5	31295500	31297500	chr5	31452200	31454200
  25 | chr5	31295500	31297500	chr5	31483000	31485000
  26 | chr5	31295500	31297500	chr5	31608600	31610600
  27 | chr5	31295500	31297500	chr5	32291000	32293000
  28 | chr5	31306700	31308700	chr5	31336200	31338200
  29 | chr5	31306700	31308700	chr5	31349200	31351200
  30 | chr5	31306700	31308700	chr5	31358700	31360700
  31 | chr5	31306700	31308700	chr5	31376700	31378700
  32 | chr5	31306700	31308700	chr5	31397100	31399100
  33 | chr5	31306700	31308700	chr5	31408000	31410000
  34 | chr5	31306700	31308700	chr5	31447500	31449500
  35 | chr5	31306700	31308700	chr5	31452200	31454200
  36 | chr5	31306700	31308700	chr5	31483000	31485000
  37 | chr5	31306700	31308700	chr5	31608600	31610600
  38 | chr5	31310500	31312500	chr5	31336200	31338200
  39 | chr5	31310500	31312500	chr5	31349200	31351200
  40 | chr5	31310500	31312500	chr5	31358700	31360700
  41 | chr5	31310500	31312500	chr5	31376700	31378700
  42 | chr5	31310500	31312500	chr5	31397100	31399100
  43 | chr5	31310500	31312500	chr5	31408000	31410000
  44 | chr5	31310500	31312500	chr5	31447500	31449500
  45 | chr5	31310500	31312500	chr5	31452200	31454200
  46 | chr5	31310500	31312500	chr5	31608600	31610600
  47 | chr5	31310500	31312500	chr5	31683000	31685000
  48 | chr5	31336200	31338200	chr5	31349200	31351200
  49 | chr5	31336200	31338200	chr5	31358700	31360700
  50 | chr5	31336200	31338200	chr5	31376700	31378700
  51 | chr5	31336200	31338200	chr5	31397100	31399100
  52 | chr5	31336200	31338200	chr5	31408000	31410000
  53 | chr5	31336200	31338200	chr5	31447500	31449500
  54 | chr5	31336200	31338200	chr5	31452200	31454200
  55 | chr5	31336200	31338200	chr5	31608600	31610600
  56 | chr5	31336200	31338200	chr5	31645300	31647300
  57 | chr5	31336200	31338200	chr5	31651000	31653000
  58 | chr5	31336200	31338200	chr5	31683000	31685000
  59 | chr5	31336200	31338200	chr5	31854000	31856000
  60 | chr5	31336200	31338200	chr5	32015400	32017400
  61 | chr5	31336200	31338200	chr5	32291000	32293000
  62 | chr5	31344650	31346650	chr5	31483000	31485000
  63 | chr5	31349200	31351200	chr5	31358700	31360700
  64 | chr5	31349200	31351200	chr5	31376700	31378700
  65 | chr5	31349200	31351200	chr5	31397100	31399100
  66 | chr5	31349200	31351200	chr5	31408000	31410000
  67 | chr5	31349200	31351200	chr5	31447500	31449500
  68 | chr5	31349200	31351200	chr5	31452200	31454200
  69 | chr5	31349200	31351200	chr5	31608600	31610600
  70 | chr5	31349200	31351200	chr5	31645300	31647300
  71 | chr5	31349200	31351200	chr5	31651000	31653000
  72 | chr5	31349200	31351200	chr5	31683000	31685000
  73 | chr5	31349200	31351200	chr5	31854000	31856000
  74 | chr5	31349200	31351200	chr5	32015400	32017400
  75 | chr5	31349200	31351200	chr5	32291000	32293000
  76 | chr5	31351450	31353450	chr5	31376700	31378700
  77 | chr5	31351450	31353450	chr5	31397100	31399100
  78 | chr5	31351450	31353450	chr5	31447500	31449500
  79 | chr5	31358700	31360700	chr5	31376700	31378700
  80 | chr5	31358700	31360700	chr5	31397100	31399100
  81 | chr5	31358700	31360700	chr5	31408000	31410000
  82 | chr5	31358700	31360700	chr5	31447500	31449500
  83 | chr5	31358700	31360700	chr5	31452200	31454200
  84 | chr5	31358700	31360700	chr5	31608600	31610600
  85 | chr5	31358700	31360700	chr5	31645300	31647300
  86 | chr5	31358700	31360700	chr5	31651000	31653000
  87 | chr5	31358700	31360700	chr5	31683000	31685000
  88 | chr5	31358700	31360700	chr5	31854000	31856000
  89 | chr5	31358700	31360700	chr5	32291000	32293000
  90 | chr5	31369650	31371650	chr5	31483000	31485000
  91 | chr5	31376700	31378700	chr5	31397100	31399100
  92 | chr5	31376700	31378700	chr5	31408000	31410000
  93 | chr5	31376700	31378700	chr5	31447500	31449500
  94 | chr5	31376700	31378700	chr5	31452200	31454200
  95 | chr5	31376700	31378700	chr5	31483000	31485000
  96 | chr5	31376700	31378700	chr5	31608600	31610600
  97 | chr5	31376700	31378700	chr5	31645300	31647300
  98 | chr5	31376700	31378700	chr5	31651000	31653000
  99 | chr5	31376700	31378700	chr5	31683000	31685000
 100 | chr5	31376700	31378700	chr5	31753200	31755200
 101 | chr5	31376700	31378700	chr5	31770300	31772300
 102 | chr5	31376700	31378700	chr5	31787500	31789500
 103 | chr5	31376700	31378700	chr5	31791000	31793000
 104 | chr5	31376700	31378700	chr5	31854000	31856000
 105 | chr5	31376700	31378700	chr5	32015400	32017400
 106 | chr5	31376700	31378700	chr5	32291000	32293000
 107 | chr5	31397100	31399100	chr5	31408000	31410000
 108 | chr5	31397100	31399100	chr5	31447500	31449500
 109 | chr5	31397100	31399100	chr5	31452200	31454200
 110 | chr5	31397100	31399100	chr5	31483000	31485000
 111 | chr5	31397100	31399100	chr5	31608600	31610600
 112 | chr5	31397100	31399100	chr5	31645300	31647300
 113 | chr5	31397100	31399100	chr5	31651000	31653000
 114 | chr5	31397100	31399100	chr5	31683000	31685000
 115 | chr5	31397100	31399100	chr5	31753200	31755200
 116 | chr5	31397100	31399100	chr5	31770300	31772300
 117 | chr5	31397100	31399100	chr5	31787500	31789500
 118 | chr5	31397100	31399100	chr5	31791000	31793000
 119 | chr5	31397100	31399100	chr5	31854000	31856000
 120 | chr5	31397100	31399100	chr5	32291000	32293000
 121 | chr5	31397100	31399100	chr5	32339700	32341700
 122 | chr5	31408000	31410000	chr5	31447500	31449500
 123 | chr5	31408000	31410000	chr5	31452200	31454200
 124 | chr5	31408000	31410000	chr5	31483000	31485000
 125 | chr5	31408000	31410000	chr5	31608600	31610600
 126 | chr5	31408000	31410000	chr5	31645300	31647300
 127 | chr5	31408000	31410000	chr5	31651000	31653000
 128 | chr5	31408000	31410000	chr5	31683000	31685000
 129 | chr5	31408000	31410000	chr5	31854000	31856000
 130 | chr5	31408000	31410000	chr5	32291000	32293000
 131 | chr5	31408000	31410000	chr5	32339700	32341700
 132 | chr5	31447500	31449500	chr5	31608600	31610600
 133 | chr5	31447500	31449500	chr5	31645300	31647300
 134 | chr5	31447500	31449500	chr5	31651000	31653000
 135 | chr5	31447500	31449500	chr5	31683000	31685000
 136 | chr5	31447500	31449500	chr5	31770300	31772300
 137 | chr5	31447500	31449500	chr5	31791000	31793000
 138 | chr5	31447500	31449500	chr5	31854000	31856000
 139 | chr5	31447500	31449500	chr5	32015400	32017400
 140 | chr5	31447500	31449500	chr5	32291000	32293000
 141 | chr5	31447500	31449500	chr5	32339700	32341700
 142 | chr5	31452200	31454200	chr5	31608600	31610600
 143 | chr5	31452200	31454200	chr5	31645300	31647300
 144 | chr5	31452200	31454200	chr5	31651000	31653000
 145 | chr5	31452200	31454200	chr5	31753200	31755200
 146 | chr5	31452200	31454200	chr5	31791000	31793000
 147 | chr5	31452200	31454200	chr5	31854000	31856000
 148 | chr5	31452200	31454200	chr5	32291000	32293000
 149 | chr5	31452200	31454200	chr5	32339700	32341700
 150 | chr5	31483000	31485000	chr5	31626700	31628700
 151 | chr5	31608600	31610600	chr5	31645300	31647300
 152 | chr5	31608600	31610600	chr5	31651000	31653000
 153 | chr5	31608600	31610600	chr5	31683000	31685000
 154 | chr5	31608600	31610600	chr5	31753200	31755200
 155 | chr5	31608600	31610600	chr5	31770300	31772300
 156 | chr5	31608600	31610600	chr5	31787500	31789500
 157 | chr5	31608600	31610600	chr5	31791000	31793000
 158 | chr5	31608600	31610600	chr5	31854000	31856000
 159 | chr5	31608600	31610600	chr5	32015400	32017400
 160 | chr5	31608600	31610600	chr5	32291000	32293000
 161 | chr5	31608600	31610600	chr5	32339700	32341700
 162 | chr5	31645300	31647300	chr5	31651000	31653000
 163 | chr5	31645300	31647300	chr5	31683000	31685000
 164 | chr5	31645300	31647300	chr5	31753200	31755200
 165 | chr5	31645300	31647300	chr5	31791000	31793000
 166 | chr5	31645300	31647300	chr5	31854000	31856000
 167 | chr5	31645300	31647300	chr5	32015400	32017400
 168 | chr5	31645300	31647300	chr5	32291000	32293000
 169 | chr5	31645300	31647300	chr5	32339700	32341700
 170 | chr5	31651000	31653000	chr5	31683000	31685000
 171 | chr5	31651000	31653000	chr5	31753200	31755200
 172 | chr5	31651000	31653000	chr5	31791000	31793000
 173 | chr5	31651000	31653000	chr5	31854000	31856000
 174 | chr5	31651000	31653000	chr5	32015400	32017400
 175 | chr5	31651000	31653000	chr5	32291000	32293000
 176 | chr5	31651000	31653000	chr5	32339700	32341700
 177 | chr5	31683000	31685000	chr5	31753200	31755200
 178 | chr5	31683000	31685000	chr5	31770300	31772300
 179 | chr5	31683000	31685000	chr5	31787500	31789500
 180 | chr5	31683000	31685000	chr5	31791000	31793000
 181 | chr5	31683000	31685000	chr5	31854000	31856000
 182 | chr5	31683000	31685000	chr5	32291000	32293000
 183 | chr5	31683000	31685000	chr5	32339700	32341700
 184 | chr5	31720000	31722000	chr5	31753200	31755200
 185 | chr5	31720000	31722000	chr5	31770300	31772300
 186 | chr5	31720000	31722000	chr5	31854000	31856000
 187 | chr5	31720000	31722000	chr5	32083000	32085000
 188 | chr5	31720000	31722000	chr5	32291000	32293000
 189 | chr5	31720000	31722000	chr5	32339700	32341700
 190 | chr5	31720000	31722000	chr5	32354200	32356200
 191 | chr5	31720000	31722000	chr5	32372000	32374000
 192 | chr5	31753200	31755200	chr5	31770300	31772300
 193 | chr5	31753200	31755200	chr5	31787500	31789500
 194 | chr5	31753200	31755200	chr5	31791000	31793000
 195 | chr5	31753200	31755200	chr5	31854000	31856000
 196 | chr5	31753200	31755200	chr5	32015400	32017400
 197 | chr5	31753200	31755200	chr5	32083000	32085000
 198 | chr5	31753200	31755200	chr5	32291000	32293000
 199 | chr5	31753200	31755200	chr5	32339700	32341700
 200 | chr5	31753200	31755200	chr5	32354200	32356200
 201 | chr5	31753200	31755200	chr5	32372000	32374000
 202 | chr5	31759000	31761000	chr5	31770300	31772300
 203 | chr5	31759000	31761000	chr5	31787500	31789500
 204 | chr5	31759000	31761000	chr5	31791000	31793000
 205 | chr5	31759000	31761000	chr5	31854000	31856000
 206 | chr5	31759000	31761000	chr5	32291000	32293000
 207 | chr5	31759000	31761000	chr5	32339700	32341700
 208 | chr5	31759000	31761000	chr5	32354200	32356200
 209 | chr5	31759000	31761000	chr5	32372000	32374000
 210 | chr5	31767000	31769000	chr5	31791000	31793000
 211 | chr5	31767000	31769000	chr5	31854000	31856000
 212 | chr5	31767000	31769000	chr5	32291000	32293000
 213 | chr5	31767000	31769000	chr5	32339700	32341700
 214 | chr5	31770300	31772300	chr5	31791000	31793000
 215 | chr5	31770300	31772300	chr5	31854000	31856000
 216 | chr5	31770300	31772300	chr5	32015400	32017400
 217 | chr5	31770300	31772300	chr5	32291000	32293000
 218 | chr5	31770300	31772300	chr5	32339700	32341700
 219 | chr5	31770300	31772300	chr5	32354200	32356200
 220 | chr5	31787500	31789500	chr5	31854000	31856000
 221 | chr5	31787500	31789500	chr5	32015400	32017400
 222 | chr5	31787500	31789500	chr5	32291000	32293000
 223 | chr5	31787500	31789500	chr5	32339700	32341700
 224 | chr5	31787500	31789500	chr5	32354200	32356200
 225 | chr5	31791000	31793000	chr5	31854000	31856000
 226 | chr5	31791000	31793000	chr5	32015400	32017400
 227 | chr5	31791000	31793000	chr5	32291000	32293000
 228 | chr5	31791000	31793000	chr5	32339700	32341700
 229 | chr5	31791000	31793000	chr5	32354200	32356200
 230 | chr5	31854000	31856000	chr5	32015400	32017400
 231 | chr5	31854000	31856000	chr5	32291000	32293000
 232 | chr5	31854000	31856000	chr5	32339700	32341700
 233 | chr5	31854000	31856000	chr5	32354200	32356200
 234 | chr5	31854000	31856000	chr5	32372000	32374000
 235 | chr5	31865800	31867800	chr5	32296200	32298200
 236 | chr5	31865800	31867800	chr5	32354200	32356200
 237 | chr5	32015400	32017400	chr5	32291000	32293000
 238 | chr5	32083000	32085000	chr5	32291000	32293000
 239 | chr5	32083000	32085000	chr5	32296200	32298200
 240 | chr5	32083000	32085000	chr5	32354200	32356200
 241 | chr5	32174000	32176000	chr5	32291000	32293000
 242 | chr5	32174000	32176000	chr5	32339700	32341700
 243 | chr5	32291000	32293000	chr5	32339700	32341700
 244 | chr5	32291000	32293000	chr5	32354200	32356200
 245 | chr5	32291000	32293000	chr5	32372000	32374000
 246 | chr5	32296200	32298200	chr5	32354200	32356200
 247 | chr5	32339700	32341700	chr5	32354200	32356200
 248 | chr5	32339700	32341700	chr5	32372000	32374000
 249 | chr5	32354200	32356200	chr5	32372000	32374000
 250 | chr8	84856500	84858500	chr8	84873500	84875500
 251 | chr8	84873500	84875500	chr8	84906300	84908300
 252 | chr8	84873500	84875500	chr8	84912250	84914250
 253 | chr8	84873500	84875500	chr8	84917700	84919700
 254 | chr8	84873500	84875500	chr8	84924200	84926200
 255 | chr8	84873500	84875500	chr8	84936250	84938250
 256 | chr8	84873500	84875500	chr8	84949150	84951150
 257 | chr8	84873500	84875500	chr8	84962800	84964800
 258 | chr8	84873500	84875500	chr8	84975300	84977300
 259 | chr8	84873500	84875500	chr8	84983000	84985000
 260 | chr8	84873500	84875500	chr8	84996000	84998000
 261 | chr8	84873500	84875500	chr8	85011500	85013500
 262 | chr8	84873500	84875500	chr8	85083500	85085500
 263 | chr8	84873500	84875500	chr8	85141000	85143000
 264 | chr8	84873500	84875500	chr8	85227800	85229800
 265 | chr8	84873500	84875500	chr8	85285000	85287000
 266 | chr8	84873500	84875500	chr8	85327000	85329000
 267 | chr8	84873500	84875500	chr8	85359000	85361000
 268 | chr8	84873500	84875500	chr8	85365000	85367000
 269 | chr8	84873500	84875500	chr8	85373100	85375100
 270 | chr8	84873500	84875500	chr8	85375800	85377800
 271 | chr8	84873500	84875500	chr8	85378600	85380600
 272 | chr8	84873500	84875500	chr8	85386000	85388000
 273 | chr8	84873500	84875500	chr8	85388300	85390300
 274 | chr8	84873500	84875500	chr8	85389700	85391700
 275 | chr8	84873500	84875500	chr8	85412800	85414800
 276 | chr8	84873500	84875500	chr8	85414700	85416700
 277 | chr8	84873500	84875500	chr8	85448750	85450750
 278 | chr8	84873500	84875500	chr8	85467800	85469800
 279 | chr8	84873500	84875500	chr8	85526000	85528000
 280 | chr8	84873500	84875500	chr8	85566300	85568300
 281 | chr8	84873500	84875500	chr8	85572600	85574600
 282 | chr8	84873500	84875500	chr8	85582500	85584500
 283 | chr8	84873500	84875500	chr8	85620000	85622000
 284 | chr8	84873500	84875500	chr8	85629000	85631000
 285 | chr8	84873500	84875500	chr8	85641000	85643000
 286 | chr8	84873500	84875500	chr8	85751500	85753500
 287 | chr8	84873500	84875500	chr8	85762400	85764400
 288 | chr8	84906300	84908300	chr8	84917700	84919700
 289 | chr8	84906300	84908300	chr8	84924200	84926200
 290 | chr8	84906300	84908300	chr8	84936250	84938250
 291 | chr8	84906300	84908300	chr8	84949150	84951150
 292 | chr8	84906300	84908300	chr8	84962800	84964800
 293 | chr8	84906300	84908300	chr8	84983000	84985000
 294 | chr8	84906300	84908300	chr8	84996000	84998000
 295 | chr8	84906300	84908300	chr8	85083500	85085500
 296 | chr8	84906300	84908300	chr8	85141000	85143000
 297 | chr8	84906300	84908300	chr8	85227800	85229800
 298 | chr8	84906300	84908300	chr8	85265000	85267000
 299 | chr8	84906300	84908300	chr8	85285000	85287000
 300 | chr8	84906300	84908300	chr8	85327000	85329000
 301 | chr8	84906300	84908300	chr8	85365000	85367000
 302 | chr8	84906300	84908300	chr8	85375800	85377800
 303 | chr8	84906300	84908300	chr8	85378600	85380600
 304 | chr8	84906300	84908300	chr8	85389700	85391700
 305 | chr8	84906300	84908300	chr8	85412800	85414800
 306 | chr8	84906300	84908300	chr8	85526000	85528000
 307 | chr8	84906300	84908300	chr8	85572600	85574600
 308 | chr8	84906300	84908300	chr8	85702000	85704000
 309 | chr8	84912250	84914250	chr8	84936250	84938250
 310 | chr8	84912250	84914250	chr8	84949150	84951150
 311 | chr8	84912250	84914250	chr8	84956550	84958550
 312 | chr8	84912250	84914250	chr8	84962800	84964800
 313 | chr8	84912250	84914250	chr8	84983000	84985000
 314 | chr8	84912250	84914250	chr8	84996000	84998000
 315 | chr8	84912250	84914250	chr8	85011500	85013500
 316 | chr8	84912250	84914250	chr8	85083500	85085500
 317 | chr8	84912250	84914250	chr8	85122000	85124000
 318 | chr8	84912250	84914250	chr8	85141000	85143000
 319 | chr8	84912250	84914250	chr8	85227800	85229800
 320 | chr8	84912250	84914250	chr8	85327000	85329000
 321 | chr8	84912250	84914250	chr8	85365000	85367000
 322 | chr8	84912250	84914250	chr8	85373100	85375100
 323 | chr8	84912250	84914250	chr8	85375800	85377800
 324 | chr8	84912250	84914250	chr8	85378600	85380600
 325 | chr8	84912250	84914250	chr8	85389700	85391700
 326 | chr8	84912250	84914250	chr8	85412800	85414800
 327 | chr8	84912250	84914250	chr8	85526000	85528000
 328 | chr8	84912250	84914250	chr8	85572600	85574600
 329 | chr8	84912250	84914250	chr8	85702000	85704000
 330 | chr8	84917700	84919700	chr8	84962800	84964800
 331 | chr8	84917700	84919700	chr8	84983000	84985000
 332 | chr8	84917700	84919700	chr8	84996000	84998000
 333 | chr8	84917700	84919700	chr8	85011500	85013500
 334 | chr8	84917700	84919700	chr8	85083500	85085500
 335 | chr8	84917700	84919700	chr8	85122000	85124000
 336 | chr8	84917700	84919700	chr8	85141000	85143000
 337 | chr8	84917700	84919700	chr8	85227800	85229800
 338 | chr8	84917700	84919700	chr8	85327000	85329000
 339 | chr8	84917700	84919700	chr8	85365000	85367000
 340 | chr8	84917700	84919700	chr8	85373100	85375100
 341 | chr8	84917700	84919700	chr8	85375800	85377800
 342 | chr8	84917700	84919700	chr8	85378600	85380600
 343 | chr8	84917700	84919700	chr8	85386000	85388000
 344 | chr8	84917700	84919700	chr8	85388300	85390300
 345 | chr8	84917700	84919700	chr8	85389700	85391700
 346 | chr8	84917700	84919700	chr8	85404000	85406000
 347 | chr8	84917700	84919700	chr8	85412800	85414800
 348 | chr8	84917700	84919700	chr8	85526000	85528000
 349 | chr8	84917700	84919700	chr8	85536500	85538500
 350 | chr8	84917700	84919700	chr8	85557500	85559500
 351 | chr8	84917700	84919700	chr8	85572600	85574600
 352 | chr8	84917700	84919700	chr8	85597000	85599000
 353 | chr8	84917700	84919700	chr8	85629000	85631000
 354 | chr8	84917700	84919700	chr8	85641000	85643000
 355 | chr8	84917700	84919700	chr8	85702000	85704000
 356 | chr8	84917700	84919700	chr8	85715600	85717600
 357 | chr8	84917700	84919700	chr8	85751500	85753500
 358 | chr8	84917700	84919700	chr8	85762400	85764400
 359 | chr8	84917700	84919700	chr8	85797300	85799300
 360 | chr8	84924200	84926200	chr8	84962800	84964800
 361 | chr8	84924200	84926200	chr8	84983000	84985000
 362 | chr8	84924200	84926200	chr8	84996000	84998000
 363 | chr8	84924200	84926200	chr8	85083500	85085500
 364 | chr8	84924200	84926200	chr8	85122000	85124000
 365 | chr8	84924200	84926200	chr8	85141000	85143000
 366 | chr8	84924200	84926200	chr8	85227800	85229800
 367 | chr8	84924200	84926200	chr8	85365000	85367000
 368 | chr8	84924200	84926200	chr8	85373100	85375100
 369 | chr8	84924200	84926200	chr8	85375800	85377800
 370 | chr8	84924200	84926200	chr8	85378600	85380600
 371 | chr8	84924200	84926200	chr8	85386000	85388000
 372 | chr8	84924200	84926200	chr8	85388300	85390300
 373 | chr8	84924200	84926200	chr8	85389700	85391700
 374 | chr8	84924200	84926200	chr8	85412800	85414800
 375 | chr8	84924200	84926200	chr8	85526000	85528000
 376 | chr8	84924200	84926200	chr8	85557500	85559500
 377 | chr8	84924200	84926200	chr8	85572600	85574600
 378 | chr8	84924200	84926200	chr8	85629000	85631000
 379 | chr8	84924200	84926200	chr8	85641000	85643000
 380 | chr8	84924200	84926200	chr8	85702000	85704000
 381 | chr8	84924200	84926200	chr8	85715600	85717600
 382 | chr8	84924200	84926200	chr8	85751500	85753500
 383 | chr8	84936250	84938250	chr8	84962800	84964800
 384 | chr8	84936250	84938250	chr8	84983000	84985000
 385 | chr8	84936250	84938250	chr8	84996000	84998000
 386 | chr8	84936250	84938250	chr8	85011500	85013500
 387 | chr8	84936250	84938250	chr8	85141000	85143000
 388 | chr8	84936250	84938250	chr8	85227800	85229800
 389 | chr8	84949150	84951150	chr8	84962800	84964800
 390 | chr8	84949150	84951150	chr8	84983000	84985000
 391 | chr8	84949150	84951150	chr8	84996000	84998000
 392 | chr8	84949150	84951150	chr8	85141000	85143000
 393 | chr8	84949150	84951150	chr8	85227800	85229800
 394 | chr8	84956550	84958550	chr8	84983000	84985000
 395 | chr8	84962800	84964800	chr8	84975300	84977300
 396 | chr8	84962800	84964800	chr8	84983000	84985000
 397 | chr8	84962800	84964800	chr8	84996000	84998000
 398 | chr8	84962800	84964800	chr8	85083500	85085500
 399 | chr8	84962800	84964800	chr8	85141000	85143000
 400 | chr8	84962800	84964800	chr8	85227800	85229800
 401 | chr8	84962800	84964800	chr8	85285000	85287000
 402 | chr8	84962800	84964800	chr8	85327000	85329000
 403 | chr8	84962800	84964800	chr8	85365000	85367000
 404 | chr8	84962800	84964800	chr8	85373100	85375100
 405 | chr8	84962800	84964800	chr8	85378600	85380600
 406 | chr8	84962800	84964800	chr8	85388300	85390300
 407 | chr8	84962800	84964800	chr8	85412800	85414800
 408 | chr8	84962800	84964800	chr8	85526000	85528000
 409 | chr8	84962800	84964800	chr8	85536500	85538500
 410 | chr8	84962800	84964800	chr8	85566300	85568300
 411 | chr8	84962800	84964800	chr8	85572600	85574600
 412 | chr8	84962800	84964800	chr8	85629000	85631000
 413 | chr8	84962800	84964800	chr8	85633600	85635600
 414 | chr8	84962800	84964800	chr8	85641000	85643000
 415 | chr8	84975300	84977300	chr8	84983000	84985000
 416 | chr8	84975300	84977300	chr8	84996000	84998000
 417 | chr8	84975300	84977300	chr8	85083500	85085500
 418 | chr8	84975300	84977300	chr8	85141000	85143000
 419 | chr8	84975300	84977300	chr8	85227800	85229800
 420 | chr8	84975300	84977300	chr8	85327000	85329000
 421 | chr8	84975300	84977300	chr8	85365000	85367000
 422 | chr8	84975300	84977300	chr8	85373100	85375100
 423 | chr8	84975300	84977300	chr8	85375800	85377800
 424 | chr8	84975300	84977300	chr8	85386000	85388000
 425 | chr8	84975300	84977300	chr8	85388300	85390300
 426 | chr8	84975300	84977300	chr8	85389700	85391700
 427 | chr8	84975300	84977300	chr8	85412800	85414800
 428 | chr8	84975300	84977300	chr8	85566300	85568300
 429 | chr8	84975300	84977300	chr8	85572600	85574600
 430 | chr8	84975300	84977300	chr8	85629000	85631000
 431 | chr8	84975300	84977300	chr8	85641000	85643000
 432 | chr8	84983000	84985000	chr8	84996000	84998000
 433 | chr8	84983000	84985000	chr8	85083500	85085500
 434 | chr8	84983000	84985000	chr8	85141000	85143000
 435 | chr8	84983000	84985000	chr8	85227800	85229800
 436 | chr8	84983000	84985000	chr8	85285000	85287000
 437 | chr8	84983000	84985000	chr8	85296000	85298000
 438 | chr8	84983000	84985000	chr8	85305400	85307400
 439 | chr8	84983000	84985000	chr8	85317600	85319600
 440 | chr8	84983000	84985000	chr8	85323000	85325000
 441 | chr8	84983000	84985000	chr8	85327000	85329000
 442 | chr8	84983000	84985000	chr8	85365000	85367000
 443 | chr8	84983000	84985000	chr8	85373100	85375100
 444 | chr8	84983000	84985000	chr8	85375800	85377800
 445 | chr8	84983000	84985000	chr8	85378600	85380600
 446 | chr8	84983000	84985000	chr8	85386000	85388000
 447 | chr8	84983000	84985000	chr8	85388300	85390300
 448 | chr8	84983000	84985000	chr8	85389700	85391700
 449 | chr8	84983000	84985000	chr8	85412800	85414800
 450 | chr8	84983000	84985000	chr8	85430200	85432200
 451 | chr8	84983000	84985000	chr8	85433000	85435000
 452 | chr8	84983000	84985000	chr8	85526000	85528000
 453 | chr8	84983000	84985000	chr8	85566300	85568300
 454 | chr8	84983000	84985000	chr8	85572600	85574600
 455 | chr8	84983000	84985000	chr8	85597000	85599000
 456 | chr8	84983000	84985000	chr8	85629000	85631000
 457 | chr8	84983000	84985000	chr8	85633600	85635600
 458 | chr8	84983000	84985000	chr8	85641000	85643000
 459 | chr8	84996000	84998000	chr8	85083500	85085500
 460 | chr8	84996000	84998000	chr8	85122000	85124000
 461 | chr8	84996000	84998000	chr8	85141000	85143000
 462 | chr8	84996000	84998000	chr8	85227800	85229800
 463 | chr8	84996000	84998000	chr8	85285000	85287000
 464 | chr8	84996000	84998000	chr8	85305400	85307400
 465 | chr8	84996000	84998000	chr8	85323000	85325000
 466 | chr8	84996000	84998000	chr8	85327000	85329000
 467 | chr8	84996000	84998000	chr8	85365000	85367000
 468 | chr8	84996000	84998000	chr8	85373100	85375100
 469 | chr8	84996000	84998000	chr8	85375800	85377800
 470 | chr8	84996000	84998000	chr8	85378600	85380600
 471 | chr8	84996000	84998000	chr8	85386000	85388000
 472 | chr8	84996000	84998000	chr8	85388300	85390300
 473 | chr8	84996000	84998000	chr8	85389700	85391700
 474 | chr8	84996000	84998000	chr8	85412800	85414800
 475 | chr8	84996000	84998000	chr8	85430200	85432200
 476 | chr8	84996000	84998000	chr8	85433000	85435000
 477 | chr8	84996000	84998000	chr8	85448750	85450750
 478 | chr8	84996000	84998000	chr8	85455750	85457750
 479 | chr8	84996000	84998000	chr8	85526000	85528000
 480 | chr8	84996000	84998000	chr8	85536500	85538500
 481 | chr8	84996000	84998000	chr8	85566300	85568300
 482 | chr8	84996000	84998000	chr8	85572600	85574600
 483 | chr8	84996000	84998000	chr8	85597000	85599000
 484 | chr8	84996000	84998000	chr8	85620000	85622000
 485 | chr8	84996000	84998000	chr8	85629000	85631000
 486 | chr8	84996000	84998000	chr8	85633600	85635600
 487 | chr8	84996000	84998000	chr8	85641000	85643000
 488 | chr8	84996000	84998000	chr8	85751500	85753500
 489 | chr8	85011500	85013500	chr8	85083500	85085500
 490 | chr8	85014650	85016650	chr8	85083500	85085500
 491 | chr8	85017500	85019500	chr8	85083500	85085500
 492 | chr8	85017500	85019500	chr8	85296000	85298000
 493 | chr8	85017500	85019500	chr8	85320000	85322000
 494 | chr8	85083500	85085500	chr8	85141000	85143000
 495 | chr8	85083500	85085500	chr8	85227800	85229800
 496 | chr8	85083500	85085500	chr8	85285000	85287000
 497 | chr8	85083500	85085500	chr8	85296000	85298000
 498 | chr8	85083500	85085500	chr8	85320000	85322000
 499 | chr8	85083500	85085500	chr8	85365000	85367000
 500 | chr8	85083500	85085500	chr8	85373100	85375100
 501 | chr8	85083500	85085500	chr8	85375800	85377800
 502 | chr8	85089500	85091500	chr8	85141000	85143000
 503 | chr8	85089500	85091500	chr8	85171500	85173500
 504 | chr8	85089500	85091500	chr8	85179500	85181500
 505 | chr8	85089500	85091500	chr8	85207000	85209000
 506 | chr8	85089500	85091500	chr8	85227800	85229800
 507 | chr8	85089500	85091500	chr8	85235250	85237250
 508 | chr8	85089500	85091500	chr8	85239700	85241700
 509 | chr8	85089500	85091500	chr8	85262700	85264700
 510 | chr8	85089500	85091500	chr8	85279800	85281800
 511 | chr8	85089500	85091500	chr8	85293400	85295400
 512 | chr8	85089500	85091500	chr8	85296000	85298000
 513 | chr8	85089500	85091500	chr8	85321200	85323200
 514 | chr8	85089500	85091500	chr8	85365000	85367000
 515 | chr8	85089500	85091500	chr8	85373100	85375100
 516 | chr8	85122000	85124000	chr8	85141000	85143000
 517 | chr8	85122000	85124000	chr8	85227800	85229800
 518 | chr8	85122000	85124000	chr8	85365000	85367000
 519 | chr8	85122000	85124000	chr8	85373100	85375100
 520 | chr8	85122000	85124000	chr8	85375800	85377800
 521 | chr8	85122000	85124000	chr8	85378600	85380600
 522 | chr8	85122000	85124000	chr8	85414700	85416700
 523 | chr8	85141000	85143000	chr8	85227800	85229800
 524 | chr8	85141000	85143000	chr8	85265000	85267000
 525 | chr8	85141000	85143000	chr8	85285000	85287000
 526 | chr8	85141000	85143000	chr8	85296000	85298000
 527 | chr8	85141000	85143000	chr8	85305400	85307400
 528 | chr8	85141000	85143000	chr8	85327000	85329000
 529 | chr8	85141000	85143000	chr8	85365000	85367000
 530 | chr8	85141000	85143000	chr8	85373100	85375100
 531 | chr8	85141000	85143000	chr8	85375800	85377800
 532 | chr8	85141000	85143000	chr8	85378600	85380600
 533 | chr8	85141000	85143000	chr8	85386000	85388000
 534 | chr8	85141000	85143000	chr8	85388300	85390300
 535 | chr8	85141000	85143000	chr8	85389700	85391700
 536 | chr8	85141000	85143000	chr8	85412800	85414800
 537 | chr8	85141000	85143000	chr8	85526000	85528000
 538 | chr8	85141000	85143000	chr8	85572600	85574600
 539 | chr8	85141000	85143000	chr8	85597000	85599000
 540 | chr8	85141000	85143000	chr8	85629000	85631000
 541 | chr8	85141000	85143000	chr8	85641000	85643000
 542 | chr8	85227800	85229800	chr8	85265000	85267000
 543 | chr8	85227800	85229800	chr8	85285000	85287000
 544 | chr8	85227800	85229800	chr8	85296000	85298000
 545 | chr8	85227800	85229800	chr8	85327000	85329000
 546 | chr8	85227800	85229800	chr8	85365000	85367000
 547 | chr8	85227800	85229800	chr8	85373100	85375100
 548 | chr8	85227800	85229800	chr8	85375800	85377800
 549 | chr8	85227800	85229800	chr8	85378600	85380600
 550 | chr8	85227800	85229800	chr8	85386000	85388000
 551 | chr8	85227800	85229800	chr8	85388300	85390300
 552 | chr8	85227800	85229800	chr8	85389700	85391700
 553 | chr8	85227800	85229800	chr8	85412800	85414800
 554 | chr8	85227800	85229800	chr8	85430200	85432200
 555 | chr8	85227800	85229800	chr8	85448750	85450750
 556 | chr8	85227800	85229800	chr8	85526000	85528000
 557 | chr8	85227800	85229800	chr8	85536500	85538500
 558 | chr8	85227800	85229800	chr8	85557500	85559500
 559 | chr8	85227800	85229800	chr8	85566300	85568300
 560 | chr8	85227800	85229800	chr8	85572600	85574600
 561 | chr8	85227800	85229800	chr8	85597000	85599000
 562 | chr8	85227800	85229800	chr8	85620000	85622000
 563 | chr8	85227800	85229800	chr8	85629000	85631000
 564 | chr8	85227800	85229800	chr8	85633600	85635600
 565 | chr8	85227800	85229800	chr8	85641000	85643000
 566 | chr8	85227800	85229800	chr8	85715600	85717600
 567 | chr8	85227800	85229800	chr8	85751500	85753500
 568 | chr8	85227800	85229800	chr8	85762400	85764400
 569 | chr8	85227800	85229800	chr8	85797300	85799300
 570 | chr8	85285000	85287000	chr8	85365000	85367000
 571 | chr8	85285000	85287000	chr8	85526000	85528000
 572 | chr8	85285000	85287000	chr8	85536500	85538500
 573 | chr8	85285000	85287000	chr8	85557500	85559500
 574 | chr8	85305400	85307400	chr8	85320000	85322000
 575 | chr8	85305400	85307400	chr8	85365000	85367000
 576 | chr8	85305400	85307400	chr8	85536500	85538500
 577 | chr8	85305400	85307400	chr8	85538700	85540700
 578 | chr8	85317600	85319600	chr8	85538700	85540700
 579 | chr8	85323000	85325000	chr8	85538700	85540700
 580 | chr8	85327000	85329000	chr8	85365000	85367000
 581 | chr8	85327000	85329000	chr8	85373100	85375100
 582 | chr8	85327000	85329000	chr8	85375800	85377800
 583 | chr8	85327000	85329000	chr8	85378600	85380600
 584 | chr8	85327000	85329000	chr8	85386000	85388000
 585 | chr8	85327000	85329000	chr8	85389700	85391700
 586 | chr8	85327000	85329000	chr8	85412800	85414800
 587 | chr8	85327000	85329000	chr8	85526000	85528000
 588 | chr8	85327000	85329000	chr8	85572600	85574600
 589 | chr8	85327000	85329000	chr8	85629000	85631000
 590 | chr8	85327000	85329000	chr8	85641000	85643000
 591 | chr8	85365000	85367000	chr8	85412800	85414800
 592 | chr8	85365000	85367000	chr8	85414700	85416700
 593 | chr8	85365000	85367000	chr8	85430200	85432200
 594 | chr8	85365000	85367000	chr8	85435000	85437000
 595 | chr8	85365000	85367000	chr8	85448750	85450750
 596 | chr8	85365000	85367000	chr8	85455750	85457750
 597 | chr8	85365000	85367000	chr8	85463300	85465300
 598 | chr8	85365000	85367000	chr8	85467800	85469800
 599 | chr8	85365000	85367000	chr8	85526000	85528000
 600 | chr8	85365000	85367000	chr8	85536500	85538500
 601 | chr8	85365000	85367000	chr8	85538700	85540700
 602 | chr8	85365000	85367000	chr8	85541200	85543200
 603 | chr8	85365000	85367000	chr8	85557500	85559500
 604 | chr8	85365000	85367000	chr8	85572600	85574600
 605 | chr8	85365000	85367000	chr8	85597000	85599000
 606 | chr8	85365000	85367000	chr8	85620000	85622000
 607 | chr8	85365000	85367000	chr8	85629000	85631000
 608 | chr8	85365000	85367000	chr8	85633600	85635600
 609 | chr8	85365000	85367000	chr8	85641000	85643000
 610 | chr8	85365000	85367000	chr8	85691000	85693000
 611 | chr8	85365000	85367000	chr8	85695000	85697000
 612 | chr8	85365000	85367000	chr8	85702000	85704000
 613 | chr8	85365000	85367000	chr8	85711500	85713500
 614 | chr8	85365000	85367000	chr8	85715600	85717600
 615 | chr8	85365000	85367000	chr8	85751500	85753500
 616 | chr8	85365000	85367000	chr8	85762400	85764400
 617 | chr8	85365000	85367000	chr8	85797300	85799300
 618 | chr8	85373100	85375100	chr8	85412800	85414800
 619 | chr8	85373100	85375100	chr8	85414700	85416700
 620 | chr8	85373100	85375100	chr8	85448750	85450750
 621 | chr8	85373100	85375100	chr8	85455750	85457750
 622 | chr8	85373100	85375100	chr8	85463300	85465300
 623 | chr8	85373100	85375100	chr8	85467800	85469800
 624 | chr8	85373100	85375100	chr8	85526000	85528000
 625 | chr8	85373100	85375100	chr8	85557500	85559500
 626 | chr8	85373100	85375100	chr8	85566300	85568300
 627 | chr8	85373100	85375100	chr8	85572600	85574600
 628 | chr8	85373100	85375100	chr8	85597000	85599000
 629 | chr8	85373100	85375100	chr8	85620000	85622000
 630 | chr8	85373100	85375100	chr8	85629000	85631000
 631 | chr8	85373100	85375100	chr8	85633600	85635600
 632 | chr8	85373100	85375100	chr8	85641000	85643000
 633 | chr8	85373100	85375100	chr8	85691000	85693000
 634 | chr8	85373100	85375100	chr8	85695000	85697000
 635 | chr8	85373100	85375100	chr8	85702000	85704000
 636 | chr8	85373100	85375100	chr8	85711500	85713500
 637 | chr8	85373100	85375100	chr8	85715600	85717600
 638 | chr8	85373100	85375100	chr8	85751500	85753500
 639 | chr8	85373100	85375100	chr8	85762400	85764400
 640 | chr8	85373100	85375100	chr8	85785750	85787750
 641 | chr8	85373100	85375100	chr8	85797300	85799300
 642 | chr8	85375800	85377800	chr8	85412800	85414800
 643 | chr8	85375800	85377800	chr8	85414700	85416700
 644 | chr8	85375800	85377800	chr8	85448750	85450750
 645 | chr8	85375800	85377800	chr8	85455750	85457750
 646 | chr8	85375800	85377800	chr8	85463300	85465300
 647 | chr8	85375800	85377800	chr8	85467800	85469800
 648 | chr8	85375800	85377800	chr8	85526000	85528000
 649 | chr8	85375800	85377800	chr8	85557500	85559500
 650 | chr8	85375800	85377800	chr8	85566300	85568300
 651 | chr8	85375800	85377800	chr8	85572600	85574600
 652 | chr8	85375800	85377800	chr8	85597000	85599000
 653 | chr8	85375800	85377800	chr8	85629000	85631000
 654 | chr8	85375800	85377800	chr8	85633600	85635600
 655 | chr8	85375800	85377800	chr8	85641000	85643000
 656 | chr8	85375800	85377800	chr8	85691000	85693000
 657 | chr8	85375800	85377800	chr8	85702000	85704000
 658 | chr8	85375800	85377800	chr8	85711500	85713500
 659 | chr8	85375800	85377800	chr8	85715600	85717600
 660 | chr8	85375800	85377800	chr8	85751500	85753500
 661 | chr8	85375800	85377800	chr8	85762400	85764400
 662 | chr8	85375800	85377800	chr8	85797300	85799300
 663 | chr8	85378600	85380600	chr8	85412800	85414800
 664 | chr8	85378600	85380600	chr8	85414700	85416700
 665 | chr8	85378600	85380600	chr8	85430200	85432200
 666 | chr8	85378600	85380600	chr8	85448750	85450750
 667 | chr8	85378600	85380600	chr8	85455750	85457750
 668 | chr8	85378600	85380600	chr8	85463300	85465300
 669 | chr8	85378600	85380600	chr8	85467800	85469800
 670 | chr8	85378600	85380600	chr8	85557500	85559500
 671 | chr8	85378600	85380600	chr8	85566300	85568300
 672 | chr8	85378600	85380600	chr8	85572600	85574600
 673 | chr8	85378600	85380600	chr8	85597000	85599000
 674 | chr8	85378600	85380600	chr8	85620000	85622000
 675 | chr8	85378600	85380600	chr8	85629000	85631000
 676 | chr8	85378600	85380600	chr8	85633600	85635600
 677 | chr8	85378600	85380600	chr8	85641000	85643000
 678 | chr8	85378600	85380600	chr8	85691000	85693000
 679 | chr8	85378600	85380600	chr8	85695000	85697000
 680 | chr8	85378600	85380600	chr8	85702000	85704000
 681 | chr8	85378600	85380600	chr8	85711500	85713500
 682 | chr8	85378600	85380600	chr8	85715600	85717600
 683 | chr8	85378600	85380600	chr8	85751500	85753500
 684 | chr8	85378600	85380600	chr8	85762400	85764400
 685 | chr8	85378600	85380600	chr8	85797300	85799300
 686 | chr8	85386000	85388000	chr8	85412800	85414800
 687 | chr8	85386000	85388000	chr8	85430200	85432200
 688 | chr8	85386000	85388000	chr8	85448750	85450750
 689 | chr8	85386000	85388000	chr8	85455750	85457750
 690 | chr8	85386000	85388000	chr8	85463300	85465300
 691 | chr8	85386000	85388000	chr8	85467800	85469800
 692 | chr8	85386000	85388000	chr8	85526000	85528000
 693 | chr8	85386000	85388000	chr8	85557500	85559500
 694 | chr8	85386000	85388000	chr8	85566300	85568300
 695 | chr8	85386000	85388000	chr8	85572600	85574600
 696 | chr8	85386000	85388000	chr8	85597000	85599000
 697 | chr8	85386000	85388000	chr8	85629000	85631000
 698 | chr8	85386000	85388000	chr8	85633600	85635600
 699 | chr8	85386000	85388000	chr8	85641000	85643000
 700 | chr8	85386000	85388000	chr8	85691000	85693000
 701 | chr8	85386000	85388000	chr8	85695000	85697000
 702 | chr8	85386000	85388000	chr8	85702000	85704000
 703 | chr8	85386000	85388000	chr8	85711500	85713500
 704 | chr8	85386000	85388000	chr8	85715600	85717600
 705 | chr8	85386000	85388000	chr8	85751500	85753500
 706 | chr8	85386000	85388000	chr8	85762400	85764400
 707 | chr8	85386000	85388000	chr8	85797300	85799300
 708 | chr8	85388300	85390300	chr8	85412800	85414800
 709 | chr8	85388300	85390300	chr8	85448750	85450750
 710 | chr8	85388300	85390300	chr8	85526000	85528000
 711 | chr8	85388300	85390300	chr8	85557500	85559500
 712 | chr8	85388300	85390300	chr8	85566300	85568300
 713 | chr8	85388300	85390300	chr8	85572600	85574600
 714 | chr8	85388300	85390300	chr8	85597000	85599000
 715 | chr8	85388300	85390300	chr8	85629000	85631000
 716 | chr8	85388300	85390300	chr8	85633600	85635600
 717 | chr8	85388300	85390300	chr8	85641000	85643000
 718 | chr8	85388300	85390300	chr8	85691000	85693000
 719 | chr8	85388300	85390300	chr8	85695000	85697000
 720 | chr8	85388300	85390300	chr8	85702000	85704000
 721 | chr8	85388300	85390300	chr8	85711500	85713500
 722 | chr8	85388300	85390300	chr8	85715600	85717600
 723 | chr8	85388300	85390300	chr8	85751500	85753500
 724 | chr8	85388300	85390300	chr8	85762400	85764400
 725 | chr8	85388300	85390300	chr8	85797300	85799300
 726 | chr8	85389700	85391700	chr8	85412800	85414800
 727 | chr8	85389700	85391700	chr8	85414700	85416700
 728 | chr8	85389700	85391700	chr8	85430200	85432200
 729 | chr8	85389700	85391700	chr8	85435000	85437000
 730 | chr8	85389700	85391700	chr8	85448750	85450750
 731 | chr8	85389700	85391700	chr8	85455750	85457750
 732 | chr8	85389700	85391700	chr8	85463300	85465300
 733 | chr8	85389700	85391700	chr8	85467800	85469800
 734 | chr8	85389700	85391700	chr8	85526000	85528000
 735 | chr8	85389700	85391700	chr8	85549300	85551300
 736 | chr8	85389700	85391700	chr8	85557500	85559500
 737 | chr8	85389700	85391700	chr8	85566300	85568300
 738 | chr8	85389700	85391700	chr8	85572600	85574600
 739 | chr8	85389700	85391700	chr8	85597000	85599000
 740 | chr8	85389700	85391700	chr8	85620000	85622000
 741 | chr8	85389700	85391700	chr8	85629000	85631000
 742 | chr8	85389700	85391700	chr8	85633600	85635600
 743 | chr8	85389700	85391700	chr8	85641000	85643000
 744 | chr8	85389700	85391700	chr8	85691000	85693000
 745 | chr8	85389700	85391700	chr8	85695000	85697000
 746 | chr8	85389700	85391700	chr8	85702000	85704000
 747 | chr8	85389700	85391700	chr8	85711500	85713500
 748 | chr8	85389700	85391700	chr8	85715600	85717600
 749 | chr8	85389700	85391700	chr8	85751500	85753500
 750 | chr8	85389700	85391700	chr8	85762400	85764400
 751 | chr8	85389700	85391700	chr8	85785750	85787750
 752 | chr8	85389700	85391700	chr8	85797300	85799300
 753 | chr8	85412800	85414800	chr8	85448750	85450750
 754 | chr8	85412800	85414800	chr8	85463300	85465300
 755 | chr8	85412800	85414800	chr8	85467800	85469800
 756 | chr8	85412800	85414800	chr8	85526000	85528000
 757 | chr8	85412800	85414800	chr8	85557500	85559500
 758 | chr8	85412800	85414800	chr8	85566300	85568300
 759 | chr8	85412800	85414800	chr8	85572600	85574600
 760 | chr8	85412800	85414800	chr8	85582500	85584500
 761 | chr8	85412800	85414800	chr8	85597000	85599000
 762 | chr8	85412800	85414800	chr8	85620000	85622000
 763 | chr8	85412800	85414800	chr8	85629000	85631000
 764 | chr8	85412800	85414800	chr8	85633600	85635600
 765 | chr8	85412800	85414800	chr8	85639000	85641000
 766 | chr8	85412800	85414800	chr8	85641000	85643000
 767 | chr8	85412800	85414800	chr8	85663000	85665000
 768 | chr8	85412800	85414800	chr8	85671400	85673400
 769 | chr8	85412800	85414800	chr8	85691000	85693000
 770 | chr8	85412800	85414800	chr8	85695000	85697000
 771 | chr8	85412800	85414800	chr8	85702000	85704000
 772 | chr8	85412800	85414800	chr8	85711500	85713500
 773 | chr8	85412800	85414800	chr8	85715600	85717600
 774 | chr8	85412800	85414800	chr8	85751500	85753500
 775 | chr8	85412800	85414800	chr8	85762400	85764400
 776 | chr8	85412800	85414800	chr8	85797300	85799300
 777 | chr8	85414700	85416700	chr8	85448750	85450750
 778 | chr8	85414700	85416700	chr8	85455750	85457750
 779 | chr8	85414700	85416700	chr8	85463300	85465300
 780 | chr8	85414700	85416700	chr8	85467800	85469800
 781 | chr8	85414700	85416700	chr8	85526000	85528000
 782 | chr8	85414700	85416700	chr8	85557500	85559500
 783 | chr8	85414700	85416700	chr8	85566300	85568300
 784 | chr8	85414700	85416700	chr8	85572600	85574600
 785 | chr8	85414700	85416700	chr8	85597000	85599000
 786 | chr8	85414700	85416700	chr8	85620000	85622000
 787 | chr8	85414700	85416700	chr8	85629000	85631000
 788 | chr8	85414700	85416700	chr8	85633600	85635600
 789 | chr8	85414700	85416700	chr8	85639000	85641000
 790 | chr8	85414700	85416700	chr8	85641000	85643000
 791 | chr8	85414700	85416700	chr8	85715600	85717600
 792 | chr8	85414700	85416700	chr8	85751500	85753500
 793 | chr8	85414700	85416700	chr8	85762400	85764400
 794 | chr8	85414700	85416700	chr8	85785750	85787750
 795 | chr8	85414700	85416700	chr8	85797300	85799300
 796 | chr8	85423500	85425500	chr8	85536500	85538500
 797 | chr8	85423500	85425500	chr8	85538700	85540700
 798 | chr8	85423500	85425500	chr8	85541200	85543200
 799 | chr8	85425500	85427500	chr8	85536500	85538500
 800 | chr8	85425500	85427500	chr8	85538700	85540700
 801 | chr8	85425500	85427500	chr8	85541200	85543200
 802 | chr8	85425500	85427500	chr8	85597000	85599000
 803 | chr8	85430200	85432200	chr8	85526000	85528000
 804 | chr8	85430200	85432200	chr8	85536500	85538500
 805 | chr8	85430200	85432200	chr8	85541200	85543200
 806 | chr8	85430200	85432200	chr8	85597000	85599000
 807 | chr8	85430200	85432200	chr8	85629000	85631000
 808 | chr8	85430200	85432200	chr8	85633600	85635600
 809 | chr8	85430200	85432200	chr8	85641000	85643000
 810 | chr8	85430200	85432200	chr8	85715600	85717600
 811 | chr8	85433000	85435000	chr8	85526000	85528000
 812 | chr8	85433000	85435000	chr8	85597000	85599000
 813 | chr8	85433000	85435000	chr8	85629000	85631000
 814 | chr8	85433000	85435000	chr8	85633600	85635600
 815 | chr8	85433000	85435000	chr8	85641000	85643000
 816 | chr8	85433000	85435000	chr8	85715600	85717600
 817 | chr8	85435000	85437000	chr8	85463300	85465300
 818 | chr8	85435000	85437000	chr8	85526000	85528000
 819 | chr8	85435000	85437000	chr8	85536500	85538500
 820 | chr8	85435000	85437000	chr8	85538700	85540700
 821 | chr8	85435000	85437000	chr8	85541200	85543200
 822 | chr8	85435000	85437000	chr8	85549300	85551300
 823 | chr8	85435000	85437000	chr8	85597000	85599000
 824 | chr8	85435000	85437000	chr8	85629000	85631000
 825 | chr8	85435000	85437000	chr8	85633600	85635600
 826 | chr8	85435000	85437000	chr8	85715600	85717600
 827 | chr8	85448750	85450750	chr8	85526000	85528000
 828 | chr8	85448750	85450750	chr8	85536500	85538500
 829 | chr8	85448750	85450750	chr8	85557500	85559500
 830 | chr8	85448750	85450750	chr8	85566300	85568300
 831 | chr8	85448750	85450750	chr8	85572600	85574600
 832 | chr8	85448750	85450750	chr8	85597000	85599000
 833 | chr8	85448750	85450750	chr8	85629000	85631000
 834 | chr8	85448750	85450750	chr8	85633600	85635600
 835 | chr8	85448750	85450750	chr8	85639000	85641000
 836 | chr8	85448750	85450750	chr8	85641000	85643000
 837 | chr8	85455750	85457750	chr8	85526000	85528000
 838 | chr8	85455750	85457750	chr8	85536500	85538500
 839 | chr8	85455750	85457750	chr8	85557500	85559500
 840 | chr8	85455750	85457750	chr8	85566300	85568300
 841 | chr8	85455750	85457750	chr8	85572600	85574600
 842 | chr8	85455750	85457750	chr8	85597000	85599000
 843 | chr8	85455750	85457750	chr8	85629000	85631000
 844 | chr8	85455750	85457750	chr8	85633600	85635600
 845 | chr8	85455750	85457750	chr8	85639000	85641000
 846 | chr8	85455750	85457750	chr8	85641000	85643000
 847 | chr8	85463300	85465300	chr8	85526000	85528000
 848 | chr8	85463300	85465300	chr8	85536500	85538500
 849 | chr8	85463300	85465300	chr8	85538700	85540700
 850 | chr8	85463300	85465300	chr8	85541200	85543200
 851 | chr8	85463300	85465300	chr8	85549300	85551300
 852 | chr8	85463300	85465300	chr8	85597000	85599000
 853 | chr8	85463300	85465300	chr8	85620000	85622000
 854 | chr8	85463300	85465300	chr8	85629000	85631000
 855 | chr8	85463300	85465300	chr8	85633600	85635600
 856 | chr8	85463300	85465300	chr8	85639000	85641000
 857 | chr8	85463300	85465300	chr8	85641000	85643000
 858 | chr8	85463300	85465300	chr8	85715600	85717600
 859 | chr8	85467800	85469800	chr8	85526000	85528000
 860 | chr8	85467800	85469800	chr8	85536500	85538500
 861 | chr8	85467800	85469800	chr8	85538700	85540700
 862 | chr8	85467800	85469800	chr8	85557500	85559500
 863 | chr8	85467800	85469800	chr8	85566300	85568300
 864 | chr8	85467800	85469800	chr8	85572600	85574600
 865 | chr8	85467800	85469800	chr8	85597000	85599000
 866 | chr8	85467800	85469800	chr8	85620000	85622000
 867 | chr8	85467800	85469800	chr8	85629000	85631000
 868 | chr8	85467800	85469800	chr8	85633600	85635600
 869 | chr8	85467800	85469800	chr8	85639000	85641000
 870 | chr8	85467800	85469800	chr8	85641000	85643000
 871 | chr8	85467800	85469800	chr8	85715600	85717600
 872 | chr8	85497500	85499500	chr8	85526000	85528000
 873 | chr8	85499300	85501300	chr8	85526000	85528000
 874 | chr8	85511000	85513000	chr8	85526000	85528000
 875 | chr8	85511000	85513000	chr8	85597000	85599000
 876 | chr8	85511000	85513000	chr8	85620000	85622000
 877 | chr8	85511000	85513000	chr8	85629000	85631000
 878 | chr8	85511000	85513000	chr8	85641000	85643000
 879 | chr8	85526000	85528000	chr8	85536500	85538500
 880 | chr8	85526000	85528000	chr8	85549300	85551300
 881 | chr8	85526000	85528000	chr8	85557500	85559500
 882 | chr8	85526000	85528000	chr8	85566300	85568300
 883 | chr8	85526000	85528000	chr8	85572600	85574600
 884 | chr8	85526000	85528000	chr8	85597000	85599000
 885 | chr8	85526000	85528000	chr8	85620000	85622000
 886 | chr8	85526000	85528000	chr8	85629000	85631000
 887 | chr8	85526000	85528000	chr8	85633600	85635600
 888 | chr8	85526000	85528000	chr8	85639000	85641000
 889 | chr8	85526000	85528000	chr8	85641000	85643000
 890 | chr8	85526000	85528000	chr8	85663000	85665000
 891 | chr8	85526000	85528000	chr8	85671400	85673400
 892 | chr8	85526000	85528000	chr8	85691000	85693000
 893 | chr8	85526000	85528000	chr8	85695000	85697000
 894 | chr8	85526000	85528000	chr8	85702000	85704000
 895 | chr8	85526000	85528000	chr8	85705600	85707600
 896 | chr8	85526000	85528000	chr8	85711500	85713500
 897 | chr8	85526000	85528000	chr8	85715600	85717600
 898 | chr8	85526000	85528000	chr8	85751500	85753500
 899 | chr8	85526000	85528000	chr8	85762400	85764400
 900 | chr8	85536500	85538500	chr8	85572600	85574600
 901 | chr8	85536500	85538500	chr8	85597000	85599000
 902 | chr8	85536500	85538500	chr8	85629000	85631000
 903 | chr8	85536500	85538500	chr8	85633600	85635600
 904 | chr8	85536500	85538500	chr8	85715600	85717600
 905 | chr8	85536500	85538500	chr8	85751500	85753500
 906 | chr8	85536500	85538500	chr8	85762400	85764400
 907 | chr8	85538700	85540700	chr8	85597000	85599000
 908 | chr8	85541200	85543200	chr8	85597000	85599000
 909 | chr8	85549300	85551300	chr8	85597000	85599000
 910 | chr8	85549300	85551300	chr8	85629000	85631000
 911 | chr8	85549300	85551300	chr8	85797300	85799300
 912 | chr8	85557500	85559500	chr8	85572600	85574600
 913 | chr8	85557500	85559500	chr8	85582500	85584500
 914 | chr8	85557500	85559500	chr8	85597000	85599000
 915 | chr8	85557500	85559500	chr8	85620000	85622000
 916 | chr8	85557500	85559500	chr8	85629000	85631000
 917 | chr8	85557500	85559500	chr8	85633600	85635600
 918 | chr8	85557500	85559500	chr8	85641000	85643000
 919 | chr8	85557500	85559500	chr8	85691000	85693000
 920 | chr8	85557500	85559500	chr8	85702000	85704000
 921 | chr8	85557500	85559500	chr8	85705600	85707600
 922 | chr8	85557500	85559500	chr8	85715600	85717600
 923 | chr8	85557500	85559500	chr8	85751500	85753500
 924 | chr8	85557500	85559500	chr8	85762400	85764400
 925 | chr8	85557500	85559500	chr8	85785750	85787750
 926 | chr8	85557500	85559500	chr8	85797300	85799300
 927 | chr8	85566300	85568300	chr8	85572600	85574600
 928 | chr8	85566300	85568300	chr8	85582500	85584500
 929 | chr8	85566300	85568300	chr8	85597000	85599000
 930 | chr8	85566300	85568300	chr8	85620000	85622000
 931 | chr8	85566300	85568300	chr8	85629000	85631000
 932 | chr8	85566300	85568300	chr8	85633600	85635600
 933 | chr8	85566300	85568300	chr8	85641000	85643000
 934 | chr8	85566300	85568300	chr8	85691000	85693000
 935 | chr8	85566300	85568300	chr8	85702000	85704000
 936 | chr8	85566300	85568300	chr8	85705600	85707600
 937 | chr8	85566300	85568300	chr8	85715600	85717600
 938 | chr8	85566300	85568300	chr8	85751500	85753500
 939 | chr8	85566300	85568300	chr8	85762400	85764400
 940 | chr8	85566300	85568300	chr8	85785750	85787750
 941 | chr8	85566300	85568300	chr8	85797300	85799300
 942 | chr8	85572600	85574600	chr8	85582500	85584500
 943 | chr8	85572600	85574600	chr8	85597000	85599000
 944 | chr8	85572600	85574600	chr8	85620000	85622000
 945 | chr8	85572600	85574600	chr8	85629000	85631000
 946 | chr8	85572600	85574600	chr8	85633600	85635600
 947 | chr8	85572600	85574600	chr8	85641000	85643000
 948 | chr8	85572600	85574600	chr8	85691000	85693000
 949 | chr8	85572600	85574600	chr8	85695000	85697000
 950 | chr8	85572600	85574600	chr8	85702000	85704000
 951 | chr8	85572600	85574600	chr8	85705600	85707600
 952 | chr8	85572600	85574600	chr8	85711500	85713500
 953 | chr8	85572600	85574600	chr8	85715600	85717600
 954 | chr8	85572600	85574600	chr8	85751500	85753500
 955 | chr8	85572600	85574600	chr8	85762400	85764400
 956 | chr8	85572600	85574600	chr8	85785750	85787750
 957 | chr8	85572600	85574600	chr8	85797300	85799300
 958 | chr8	85582500	85584500	chr8	85597000	85599000
 959 | chr8	85582500	85584500	chr8	85620000	85622000
 960 | chr8	85582500	85584500	chr8	85629000	85631000
 961 | chr8	85582500	85584500	chr8	85633600	85635600
 962 | chr8	85582500	85584500	chr8	85641000	85643000
 963 | chr8	85582500	85584500	chr8	85691000	85693000
 964 | chr8	85582500	85584500	chr8	85715600	85717600
 965 | chr8	85582500	85584500	chr8	85751500	85753500
 966 | chr8	85582500	85584500	chr8	85762400	85764400
 967 | chr8	85582500	85584500	chr8	85797300	85799300
 968 | chr8	85597000	85599000	chr8	85620000	85622000
 969 | chr8	85597000	85599000	chr8	85629000	85631000
 970 | chr8	85597000	85599000	chr8	85633600	85635600
 971 | chr8	85597000	85599000	chr8	85641000	85643000
 972 | chr8	85597000	85599000	chr8	85663000	85665000
 973 | chr8	85597000	85599000	chr8	85671400	85673400
 974 | chr8	85597000	85599000	chr8	85691000	85693000
 975 | chr8	85597000	85599000	chr8	85695000	85697000
 976 | chr8	85597000	85599000	chr8	85715600	85717600
 977 | chr8	85597000	85599000	chr8	85751500	85753500
 978 | chr8	85597000	85599000	chr8	85762400	85764400
 979 | chr8	85597000	85599000	chr8	85785750	85787750
 980 | chr8	85597000	85599000	chr8	85797300	85799300
 981 | chr8	85620000	85622000	chr8	85629000	85631000
 982 | chr8	85620000	85622000	chr8	85633600	85635600
 983 | chr8	85620000	85622000	chr8	85641000	85643000
 984 | chr8	85620000	85622000	chr8	85663000	85665000
 985 | chr8	85620000	85622000	chr8	85671400	85673400
 986 | chr8	85620000	85622000	chr8	85691000	85693000
 987 | chr8	85620000	85622000	chr8	85695000	85697000
 988 | chr8	85620000	85622000	chr8	85715600	85717600
 989 | chr8	85620000	85622000	chr8	85751500	85753500
 990 | chr8	85620000	85622000	chr8	85762400	85764400
 991 | chr8	85620000	85622000	chr8	85797300	85799300
 992 | chr8	85629000	85631000	chr8	85633600	85635600
 993 | chr8	85629000	85631000	chr8	85641000	85643000
 994 | chr8	85629000	85631000	chr8	85663000	85665000
 995 | chr8	85629000	85631000	chr8	85671400	85673400
 996 | chr8	85629000	85631000	chr8	85691000	85693000
 997 | chr8	85629000	85631000	chr8	85695000	85697000
 998 | chr8	85629000	85631000	chr8	85702000	85704000
 999 | chr8	85629000	85631000	chr8	85711500	85713500
1000 | chr8	85629000	85631000	chr8	85715600	85717600
1001 | chr8	85629000	85631000	chr8	85751500	85753500
1002 | chr8	85629000	85631000	chr8	85762400	85764400
1003 | chr8	85629000	85631000	chr8	85797300	85799300
1004 | chr8	85633600	85635600	chr8	85641000	85643000
1005 | chr8	85633600	85635600	chr8	85663000	85665000
1006 | chr8	85633600	85635600	chr8	85671400	85673400
1007 | chr8	85633600	85635600	chr8	85691000	85693000
1008 | chr8	85633600	85635600	chr8	85695000	85697000
1009 | chr8	85633600	85635600	chr8	85702000	85704000
1010 | chr8	85633600	85635600	chr8	85711500	85713500
1011 | chr8	85633600	85635600	chr8	85715600	85717600
1012 | chr8	85633600	85635600	chr8	85751500	85753500
1013 | chr8	85633600	85635600	chr8	85762400	85764400
1014 | chr8	85633600	85635600	chr8	85797300	85799300
1015 | chr8	85641000	85643000	chr8	85663000	85665000
1016 | chr8	85641000	85643000	chr8	85671400	85673400
1017 | chr8	85641000	85643000	chr8	85691000	85693000
1018 | chr8	85641000	85643000	chr8	85695000	85697000
1019 | chr8	85641000	85643000	chr8	85702000	85704000
1020 | chr8	85641000	85643000	chr8	85711500	85713500
1021 | chr8	85641000	85643000	chr8	85715600	85717600
1022 | chr8	85641000	85643000	chr8	85751500	85753500
1023 | chr8	85641000	85643000	chr8	85762400	85764400
1024 | chr8	85641000	85643000	chr8	85797300	85799300
1025 | chr8	85654000	85656000	chr8	85682700	85684700
1026 | chr8	85654000	85656000	chr8	85715600	85717600
1027 | chr8	85663000	85665000	chr8	85682700	85684700
1028 | chr8	85663000	85665000	chr8	85691000	85693000
1029 | chr8	85663000	85665000	chr8	85715600	85717600
1030 | chr8	85663000	85665000	chr8	85751500	85753500
1031 | chr8	85663000	85665000	chr8	85762400	85764400
1032 | chr8	85663000	85665000	chr8	85797300	85799300
1033 | chr8	85671400	85673400	chr8	85691000	85693000
1034 | chr8	85671400	85673400	chr8	85715600	85717600
1035 | chr8	85671400	85673400	chr8	85751500	85753500
1036 | chr8	85671400	85673400	chr8	85762400	85764400
1037 | chr8	85671400	85673400	chr8	85797300	85799300
1038 | chr8	85682700	85684700	chr8	85715600	85717600
1039 | chr8	85691000	85693000	chr8	85702000	85704000
1040 | chr8	85691000	85693000	chr8	85711500	85713500
1041 | chr8	85691000	85693000	chr8	85715600	85717600
1042 | chr8	85691000	85693000	chr8	85751500	85753500
1043 | chr8	85691000	85693000	chr8	85762400	85764400
1044 | chr8	85691000	85693000	chr8	85785750	85787750
1045 | chr8	85691000	85693000	chr8	85797300	85799300
1046 | chr8	85691000	85693000	chr8	85806500	85808500
1047 | chr8	85691000	85693000	chr8	85809000	85811000
1048 | chr8	85695000	85697000	chr8	85702000	85704000
1049 | chr8	85695000	85697000	chr8	85711500	85713500
1050 | chr8	85695000	85697000	chr8	85715600	85717600
1051 | chr8	85695000	85697000	chr8	85751500	85753500
1052 | chr8	85695000	85697000	chr8	85762400	85764400
1053 | chr8	85695000	85697000	chr8	85785750	85787750
1054 | chr8	85695000	85697000	chr8	85797300	85799300
1055 | chr8	85695000	85697000	chr8	85806500	85808500
1056 | chr8	85695000	85697000	chr8	85809000	85811000
1057 | chr8	85702000	85704000	chr8	85711500	85713500
1058 | chr8	85702000	85704000	chr8	85715600	85717600
1059 | chr8	85702000	85704000	chr8	85723500	85725500
1060 | chr8	85702000	85704000	chr8	85751500	85753500
1061 | chr8	85702000	85704000	chr8	85762400	85764400
1062 | chr8	85702000	85704000	chr8	85785750	85787750
1063 | chr8	85702000	85704000	chr8	85797300	85799300
1064 | chr8	85702000	85704000	chr8	85802000	85804000
1065 | chr8	85702000	85704000	chr8	85806500	85808500
1066 | chr8	85702000	85704000	chr8	85809000	85811000
1067 | chr8	85705600	85707600	chr8	85751500	85753500
1068 | chr8	85705600	85707600	chr8	85762400	85764400
1069 | chr8	85705600	85707600	chr8	85797300	85799300
1070 | chr8	85711500	85713500	chr8	85751500	85753500
1071 | chr8	85711500	85713500	chr8	85762400	85764400
1072 | chr8	85711500	85713500	chr8	85797300	85799300
1073 | chr8	85715600	85717600	chr8	85751500	85753500
1074 | chr8	85715600	85717600	chr8	85762400	85764400
1075 | chr8	85715600	85717600	chr8	85785750	85787750
1076 | chr8	85715600	85717600	chr8	85797300	85799300
1077 | chr8	85715600	85717600	chr8	85806500	85808500
1078 | chr8	85715600	85717600	chr8	85809000	85811000
1079 | chr8	85751500	85753500	chr8	85762400	85764400
1080 | chr8	85751500	85753500	chr8	85785750	85787750
1081 | chr8	85751500	85753500	chr8	85797300	85799300
1082 | chr8	85751500	85753500	chr8	85806500	85808500
1083 | chr8	85751500	85753500	chr8	85809000	85811000
1084 | chr8	85762400	85764400	chr8	85785750	85787750
1085 | chr8	85762400	85764400	chr8	85797300	85799300
1086 | chr8	85762400	85764400	chr8	85806500	85808500
1087 | chr8	85762400	85764400	chr8	85809000	85811000
1088 | chr8	85785750	85787750	chr8	85797300	85799300
1089 | chr8	85785750	85787750	chr8	85806500	85808500
1090 | chr8	85797300	85799300	chr8	85806500	85808500
1091 | chr8	85797300	85799300	chr8	85809000	85811000
1092 | 


--------------------------------------------------------------------------------
/PileupsRCMC.ipynb:
--------------------------------------------------------------------------------
  1 | {
  2 |  "cells": [
  3 |   {
  4 |    "cell_type": "markdown",
  5 |    "id": "a2f31b9f",
  6 |    "metadata": {},
  7 |    "source": [
  8 |     "# Plotting pileups"
  9 |    ]
 10 |   },
 11 |   {
 12 |    "cell_type": "markdown",
 13 |    "id": "717f5fcc",
 14 |    "metadata": {},
 15 |    "source": [
 16 |     "## Imports"
 17 |    ]
 18 |   },
 19 |   {
 20 |    "cell_type": "code",
 21 |    "execution_count": null,
 22 |    "id": "da763af9",
 23 |    "metadata": {},
 24 |    "outputs": [],
 25 |    "source": [
 26 |     "# import standard python libraries\n",
 27 |     "import matplotlib as mpl\n",
 28 |     "%matplotlib inline\n",
 29 |     "mpl.rcParams['figure.dpi'] = 96\n",
 30 |     "import numpy as np\n",
 31 |     "import matplotlib.pyplot as plt\n",
 32 |     "import pandas as pd\n",
 33 |     "import seaborn as sns\n",
 34 |     "\n",
 35 |     "# import libraries for biological data analysis\n",
 36 |     "from coolpuppy import coolpup\n",
 37 |     "from plotpuppy import plotpup\n",
 38 |     "import cooler\n",
 39 |     "import bioframe\n",
 40 |     "import cooltools\n",
 41 |     "from cooltools import expected_cis\n",
 42 |     "from cooltools.lib import plotting\n",
 43 |     "\n",
 44 |     "import bbi"
 45 |    ]
 46 |   },
 47 |   {
 48 |    "cell_type": "markdown",
 49 |    "id": "98de4e75",
 50 |    "metadata": {},
 51 |    "source": [
 52 |     "## Enrichment calculation"
 53 |    ]
 54 |   },
 55 |   {
 56 |    "cell_type": "code",
 57 |    "execution_count": null,
 58 |    "id": "3a79fc85",
 59 |    "metadata": {},
 60 |    "outputs": [],
 61 |    "source": [
 62 |     "#Viraat's new calculation\n",
 63 |     "#Modified 2022/10/04 by Miles to try to avoid NaN values and correct an issue with the background sum, \n",
 64 |     "#and generally make the code a little more streamlined\n",
 65 |     "def enrichmentCalc(mtx, dotWindow):\n",
 66 |     "    #Dimension of array side (should be square)\n",
 67 |     "    sideLength = len(mtx)\n",
 68 |     "    #Middle of side length\n",
 69 |     "    midPoint = (sideLength - 1) // 2\n",
 70 |     "    #Half size of box around centre pixel (one pixel smaller if even-sized dot window - don't do this)\n",
 71 |     "    buffer = (dotWindow - 1) // 2\n",
 72 |     "    \n",
 73 |     "    #Get sum of pixels around dot\n",
 74 |     "    dotSum = np.nansum(mtx[midPoint-buffer:midPoint+buffer+1, midPoint-buffer:midPoint+buffer+1])\n",
 75 |     "    \n",
 76 |     "    #Subset the matrix and calculate the mean without NaN values\n",
 77 |     "    backgroundSum1 = np.nansum(mtx[0:dotWindow, 0:dotWindow])\n",
 78 |     "    backgroundSum2 = np.nansum(mtx[sideLength-dotWindow:sideLength, sideLength-dotWindow:sideLength])\n",
 79 |     "    \n",
 80 |     "    #Calculate enrichment (NB this assumes all boxes are the same size.\n",
 81 |     "    #If you set an even dotWindow value, they won't be)\n",
 82 |     "    enrichment = dotSum / ((backgroundSum1 + backgroundSum2)/2)\n",
 83 |     "    \n",
 84 |     "    return enrichment"
 85 |    ]
 86 |   },
 87 |   {
 88 |    "cell_type": "markdown",
 89 |    "id": "a19e8ae8",
 90 |    "metadata": {},
 91 |    "source": [
 92 |     "## Inputs"
 93 |    ]
 94 |   },
 95 |   {
 96 |    "cell_type": "code",
 97 |    "execution_count": null,
 98 |    "id": "295f6354",
 99 |    "metadata": {},
100 |    "outputs": [],
101 |    "source": [
102 |     "#mcool resolution to read\n",
103 |     "resolution = 250\n",
104 |     "#List of mcool locations as strings\n",
105 |     "clrfiles = [\"mcoollocation1\", \"mcoollocation2\", \"mcoollocation3\"]\n",
106 |     "#List of mcool conditions as strings\n",
107 |     "conditions = [\"condition1\", \"condition2\", \"condition3\"]\n",
108 |     "#List of loop types as strings\n",
109 |     "loopTypesNames = [\"loop\", \"type\", \"names\"]\n",
110 |     "#List of loop file locations (bedpe)\n",
111 |     "loopFiles = [\"looplocation1\", \"looplocation2\", \"looplocation3\"]\n",
112 |     "\n",
113 |     "#Specify the RCMC regions of the mcools to look at (format: chromosome (string), start (number), end (number), name of region (string))\n",
114 |     "regions = pd.DataFrame([['chrA',1,100,'regionname1'],['chrB',1,100,'regionname2'],['chrC',1,100,'regionname3']],\n",
115 |     "                  columns=['chrom', 'start', 'end', 'name'])\n",
116 |     "#Cis expected file locations from cooltools - .tsv file - one for each mcool\n",
117 |     "expectedFiles = [\"expectedlocation1\", \"expectedlocation2\", \"expectedlocation3\"]\n",
118 |     "#Set save directory\n",
119 |     "saveDir = '/a/directory/on/your/system/'\n",
120 |     "\n",
121 |     "#Set the size of the area flanking the dot\n",
122 |     "flankDist = 10000\n",
123 |     "#Don't set this to be even... This is the size of the area to measure around the dot \n",
124 |     "#(and by extension the size of the boxes at the edges of the region too)\n",
125 |     "#For this reason, it needs to be odd to have integer box sizes on each side.\n",
126 |     "dotWindow = 5\n",
127 |     "\n"
128 |    ]
129 |   },
130 |   {
131 |    "cell_type": "markdown",
132 |    "id": "aaa6af6c",
133 |    "metadata": {},
134 |    "source": [
135 |     "Read in the loops"
136 |    ]
137 |   },
138 |   {
139 |    "cell_type": "code",
140 |    "execution_count": null,
141 |    "id": "e9f08a9b",
142 |    "metadata": {},
143 |    "outputs": [],
144 |    "source": [
145 |     "#######Don't change this section#######\n",
146 |     "#Creat an empty list to store the imported loop locations\n",
147 |     "loopTypes = []\n",
148 |     "#List of column names to use for imported loops (this is constant - do not change)\n",
149 |     "colNames = ['chrom1', 'start1', 'end1', 'chrom2', 'start2', 'end2']\n",
150 |     "#Read in files, put them in loopTypes\n",
151 |     "for file in loopFiles:\n",
152 |     "    temploops = pd.read_csv(file, sep='\\t', names=colNames, header=None)\n",
153 |     "    loopTypes.append(temploops)"
154 |    ]
155 |   },
156 |   {
157 |    "cell_type": "markdown",
158 |    "id": "c6de3c38",
159 |    "metadata": {},
160 |    "source": [
161 |     "## Run the script"
162 |    ]
163 |   },
164 |   {
165 |    "cell_type": "code",
166 |    "execution_count": null,
167 |    "id": "f27175dd",
168 |    "metadata": {},
169 |    "outputs": [],
170 |    "source": [
171 |     "#Loop through each cooler\n",
172 |     "for i, clrfile in enumerate(clrfiles):\n",
173 |     "    #Get condition name\n",
174 |     "    condition = conditions[i]\n",
175 |     "    #Get expected file\n",
176 |     "    expected = pd.read_csv(expectedFiles[i], sep='\\t')\n",
177 |     "    #Read in cooler\n",
178 |     "    clr = cooler.Cooler(clrfile+'::/resolutions/'+str(resolution))\n",
179 |     "    #Loop through different loop types\n",
180 |     "    for j in range(len(loopTypes)):\n",
181 |     "        loops = loopTypes[j]\n",
182 |     "        loopsName = loopTypesNames[j]\n",
183 |     "        #Calculate pileups\n",
184 |     "        stack = cooltools.pileup(clr, loops, view_df=regions, expected_df=expected, flank=flankDist)\n",
185 |     "        #Flatten stack by calculating means\n",
186 |     "        mtx = np.nanmean(stack, axis=2)\n",
187 |     "        #Calculate enrichment\n",
188 |     "        enrichment = enrichmentCalc(mtx, dotWindow)\n",
189 |     "        #Plot figure\n",
190 |     "        plt.imshow(\n",
191 |     "            np.log2(mtx),\n",
192 |     "            vmax = 2.5,\n",
193 |     "            vmin = -2.5,\n",
194 |     "            cmap='coolwarm')\n",
195 |     "        \n",
196 |     "        plt.colorbar(label = 'log2 mean obs/exp')\n",
197 |     "        ticks_pixels = np.linspace(0, flankDist*2//resolution,5)\n",
198 |     "        ticks_kbp = ((ticks_pixels-ticks_pixels[-1]/2)*resolution//1000).astype(int)\n",
199 |     "        plt.xticks(ticks_pixels, ticks_kbp)\n",
200 |     "        plt.yticks(ticks_pixels, ticks_kbp)\n",
201 |     "        plt.xlabel('relative position, kbp')\n",
202 |     "        plt.ylabel('relative position, kbp')\n",
203 |     "        plt.text(1, 1, round(enrichment, 2))\n",
204 |     "        plt.savefig(saveDir+'LoopPileups_'+condition+'_'+loopsName+'_'+str(resolution)+'bp_'+str(flankDist)+'bp.pdf', dpi=1200)\n",
205 |     "        plt.clf()\n",
206 |     "\n"
207 |    ]
208 |   }
209 |  ],
210 |  "metadata": {
211 |   "kernelspec": {
212 |    "display_name": "Python [conda env:coolpuppy]",
213 |    "language": "python",
214 |    "name": "conda-env-coolpuppy-py"
215 |   },
216 |   "language_info": {
217 |    "codemirror_mode": {
218 |     "name": "ipython",
219 |     "version": 3
220 |    },
221 |    "file_extension": ".py",
222 |    "mimetype": "text/x-python",
223 |    "name": "python",
224 |    "nbconvert_exporter": "python",
225 |    "pygments_lexer": "ipython3",
226 |    "version": "3.7.12"
227 |   }
228 |  },
229 |  "nbformat": 4,
230 |  "nbformat_minor": 5
231 | }
232 | 


--------------------------------------------------------------------------------
/README.md:
--------------------------------------------------------------------------------
  1 | # RCMC Analysis Code
  2 | This repository contains source code for the article [Region Capture Micro-C reveals coalescence of enhancers and promoters into nested microcompartments](https://www.nature.com/articles/s41588-023-01391-1), used in the analysis of RCMC data. 
  3 | 
  4 | Code is provided either in the form of Python/R scripts or as Jupyter notebooks to be run in conda environments containing the required packages. Additionally, genomic positions of microcompartments identified in the paper are included in bedpe format.
  5 | 
  6 | ## Code summary
  7 | ### Micro-C alignment (microcbowtie2.py)
  8 | Required packages:
  9 | -	bowtie2
 10 | -	samtools
 11 | -	sambamba
 12 | -	pairtools
 13 | -	cooler
 14 | -	pairix
 15 | 
 16 | Python script used to align reads in .fastq format from paired-end sequencing of Micro-C experiments and produces as output .pairs, .cool and .mcool files compatible with downstream applications such as HiGlass.
 17 | 
 18 | Example usage:
 19 | 
 20 | ```
 21 | python /path/to/script/microcbowtie2.py --file_1 pair1.fastq --file_2 pair2.fasq -g mm39 -t 36 -o exampleoutput
 22 | ```
 23 | 
 24 | ### ChIP-seq alignment (spikeinChIP_PE_alignment.py)
 25 | Required packages:
 26 | -	bowtie2
 27 | -	samtools
 28 | -	sambamba
 29 | 
 30 | Python script used to align reads in .fastq format from paired-end sequencing of ChIP-seq experiments and produces aligned .bam files. .fastq information is input as a .tsv with each line containing the path to the first pair .fastq, the path to the second pair .fastq, and the desired output name of the aligned file.
 31 | 
 32 | Example usage:
 33 | 
 34 | ```
 35 | python /path/to/script/spikeinChIP_PE_alignment.py -f list_of_fastqs.tsv -g mm39 -t 36 -o alignmentcountsout
 36 | ```
 37 | 
 38 | ### Finding chromatin features overlapping microcompartment anchors (loopFeatureOverlap.R)
 39 | Required packages:
 40 | -	plyr
 41 | -	dplyr
 42 | -	reshape2
 43 | -	purrr
 44 | -	grid
 45 | -	IRanges
 46 | -	GenomicRanges
 47 | -	arrangements
 48 | -	foreach
 49 | 
 50 | R script used to classify microcompartment interactions by finding overlap between identified microcompartments (.bedpe) and chromatin features (.bed) such as promoters, enhancers, CTCF binding sites, etc. It outputs individual .bedpe files of interactions according to combinatorial classification of chromatin features (e.g for enhancer (E) and promoter (P): P-P, E-E, E-P, E-null, P-null, null-null), including interactions which have no overlap (null category). Classification can be mutually exclusive (E-P cannot also be P-P) or inclusive.
 51 | 
 52 | Example usage:
 53 | 
 54 | ```
 55 | Rscript /path/to/script/loopFeatureOverlap.R -l interactions.bedpe -b promoter.bed,enhancer.bed -i P,E -o outputdirectory/
 56 | ```
 57 | 
 58 | ### Calculating strength of individual interactions (LoopStrengthRCMC.ipynb) or aggregate pileup analysis (PileupsRCMC.ipynb)
 59 | Required packages:
 60 | -	seaborn
 61 | -	coolpuppy
 62 | -	cooltools
 63 | -	cooler
 64 | 
 65 | Jupyter notebooks used to calculate strengths of individual microcompartments (LoopStrengthRCMC.ipynb) or generate aggregate pileup analysis figures (PileupsRCMC.ipynb). Each one takes .mcool files of contacts from RCMC, a list of interactions to calculate for (.bedpe format), expected files generated by cooltools for each .mcool, and the captured region, and calculates background corrected observed/expected interaction strengths, either for each interaction individually (output as a .bedpe with additional columns for strengths for each .mcool) or as a pileup (output as a .pdf of the aggregate interaction annotated with the calculated strength).
 66 | 
 67 | ### Visualization of contact maps and genomic tracks (ContactMapVisualizationExampleNotebook.ipynb)
 68 | Required packages:
 69 | -	cooltools
 70 | -	cooler
 71 | -	coolbox
 72 | -	matplotlib
 73 | 
 74 | Jupyter notebook used to generate visualizations of contact maps and genomics tracks for figures. Contact map visualization is accomplished using cooltools and requires a .mcool file of contacts from RCMC or a comparable method. Genomic track visualization is accomplished using coolbox and requires a .mcool file of contacts, gene annotations (.gtf format or similar), and ChIP-seq, RNA-seq, and ATAC-seq datasets (.bw format).
 75 | 
 76 | ### Calculation of read-containing bin fraction by contact distance (CalculatingFilledBinFractionByDistance.ipynb)
 77 | Required packages:
 78 | -	cooltools
 79 | -	cooler
 80 | -	matplotlib
 81 | 
 82 | Jupyter notebook used to calculate the fraction of bins in .mcool-derived contact maps which contain at least one read pair at a given resolution and contact distance from the diagonal. The notebook takes an unbalanced .mcool of contacts from RCMC or a comparable method, tabulates the occupied contact bin fraction at specified contact distances, and generates a plot of occupied bin fraction by contact distance.
 83 | 
 84 | ### Calculation of row sums in ICE-balanced .mcools (BalancedRowsumsCalculation.ipynb)
 85 | Required packages:
 86 | -	cooltools
 87 | -	cooler
 88 | -	matplotlib
 89 | 
 90 | Jupyter notebook used to confirm successful ICE balancing of .mcool files by calculating and plotting row sums. Two variations of the calculation are provided in the notebook – one for calculating row sums across an entire region, and one for calculating row sums only for bins containing contact anchor sites. Both variations take ICE-balanced .mcool files of contacts from RCMC or a comparable method, and the latter additionally takes a list of contact anchor sites (.bed format). The distributions of calculated row sums in either variation are plotted as histograms.
 91 | 
 92 | ### List of manually-annotated microcompartment loops (MicrocompartmentLoops_PlusMin1kb.bedpe)
 93 | BEDPE format file listing all 1091 manually-annotated microcompartment loops across the Ppm1g (chr5) and Klf1 (chr8) regions used in the microcompartment analysis scripts above. Coordinates are provided for the mm39 reference genome, and loop anchors are listed as plus-and-minus 1kb from each anchor’s point coordinate. Columns in the file are as follows: the first is the chromosome of the left loop anchor, the second is the coordinate of the left loop anchor minus 1 kb, the third is the coordinate of the left loop anchor plus 1 kb, and the remaining three columns are the same for the right loop anchor.
 94 | 
 95 | ### Lists of probes used for capturing regions of interest (captureprobes_mm10.bed, captureprobes_mm39.bed)
 96 | BED format file listing the genomic locations of all probes used for capturing the Sox2 (chr3), Ppm1g (chr5), Nanog (chr6), Klf1 (chr8), and Fbn2 (chr18) regions used in capture. Coordinates are provided for both the mm10 and mm39 reference genomes, and loop anchors are listed as plus-and-minus 1kb from each anchor’s point coordinate. Columns in the file are as follows: the first is the chromosome the region is located on, the second is the start coordinate of the probe, and the third is the end coordinate of the probe.
 97 | 
 98 | ## How to cite
 99 | This work is shared under an MIT license. If you make use of analysis scripts or data from this work, please cite as follows:
100 | 
101 | Goel, V.Y., Huseyin, M.K. & Hansen, A.S. Region Capture Micro-C reveals coalescence of enhancers and promoters into nested microcompartments. *Nat Genet* (2023). https://doi.org/10.1038/s41588-023-01391-1
102 | 
103 | Also refer to our deposited and citable code on Zenodo:
104 | 
105 | Goel, Viraat Y, Huseyin, Miles K, & Hansen, Anders S. (2023). Code supporting Region Capture Micro-C reveals coalescence of enhancers and promoters into nested microcompartments (1.0). Zenodo. https://doi.org/10.5281/zenodo.7641852
106 | 


--------------------------------------------------------------------------------
/loopFeatureOverlap.R:
--------------------------------------------------------------------------------
  1 | #New version of loopFeatureOverlap.py, written in R using Granges instead of Python-based Pyranges version, which had an output issue
  2 | 
  3 | #Imports
  4 | library('plyr')
  5 | library('dplyr')
  6 | require('reshape2')
  7 | library('purrr')
  8 | library('grid')
  9 | # library('ChIPpeakAnno')
 10 | library('IRanges')
 11 | library('GenomicRanges')
 12 | library('arrangements')
 13 | library('foreach')
 14 | 
 15 | #Get args
 16 | if(!require(optparse)) {
 17 |   stop("Please install the optparse package and try again!", call. = FALSE)
 18 | }
 19 | 
 20 | library(optparse)
 21 | 
 22 | parser <- OptionParser(add_help_option = TRUE)
 23 | parser <- add_option(parser, c("-o", "--outdir"), action = "store", type = "character", help = "Directory to output bed files", default = "na")
 24 | parser <- add_option(parser, c("-l", "--loops"), action = "store", type = "character", help = "Input bedpe or tsv file containing loops in bedpe format", default = "na")
 25 | parser <- add_option(parser, c("-b", "--bed"), action = "store", type = "character", help = "Input bed file or files - if multiple, separate with commas", default = "na")
 26 | parser <- add_option(parser, c("-i", "--id"), action = "store", type = "character", help = "Feature names to use for each bed file - will be used to name output files. If multiple, separate with commas", default = "na")
 27 | parser <- add_option(parser, c("-e", "--exclusive"), action = "store_true", help = "When set, defines loop anchors with only a single feature - overlap with multiple features is not allowed", default = FALSE)
 28 | args <- parse_args(parser)
 29 | 
 30 | #########################################################################
 31 | #Convert args to variables
 32 | path.loops <- args$loops
 33 | outdir <- args$outdir
 34 | features <- args$bed
 35 | ids <- args$id
 36 | 
 37 | #########################################################################
 38 | #Read in loops
 39 | loops <- read.delim(path.loops, header = FALSE, col.names = c("chr1", "start1", "end1", "chr2", "start2", "end2"), sep = "", dec = ".") #opens up the file
 40 | 
 41 | #Read in features
 42 | featureslist <- unlist(strsplit(features, ","))
 43 | idlist <- unlist(strsplit(ids, ","))
 44 | 
 45 | #Check features and ids are same length
 46 | if (length(featureslist) != length(idlist)) {
 47 |   stop("Make sure the same numbers of bed files and IDs are provided")
 48 | }
 49 | 
 50 | #We only want and know the names of the first three columns, so write a little import function:
 51 | import_feature_data <- function(filename) {
 52 |   #Get number of columns
 53 |   num.cols.to.blank <- max(count.fields(filename, sep = "\t")) - 3
 54 |   df <- read.delim(filename, header = FALSE, sep = "", dec = ".", colClasses = c('character', rep('numeric', 2), rep("NULL", num.cols.to.blank)))
 55 |   colnames(df) <- c("chr", "start", "end")
 56 |   return(df)
 57 | }
 58 | 
 59 | features.data <- lapply(featureslist, import_feature_data)
 60 | 
 61 | #Ensure output directory ends with a /
 62 | if(!endsWith(outdir, '/')) {
 63 |   outdir <- paste0(outdir, '/')
 64 | }
 65 | 
 66 | if (outdir == "na") {
 67 |   stop("Please provide an option for --out", call. = FALSE)
 68 | }
 69 | 
 70 | #########################################################################
 71 | #Make Granges objects
 72 | #For features
 73 | features.data.granges <- lapply(features.data, makeGRangesFromDataFrame, keep.extra.columns = FALSE)
 74 | #For loops
 75 | #First make separate dfs for each anchor
 76 | #Make anchors function
 77 | make_anchors_separate <- function(loops) {
 78 |   #Add loop_id column for merging
 79 |   loops$loop_id <- seq.int(nrow(loops))
 80 |   #Split loops into two bed-like files
 81 |   loops.1 <- data.frame(chr = loops$chr1, start = loops$start1, end = loops$end1, loop_id = loops$loop_id)
 82 |   loops.2 <- data.frame(chr = loops$chr2, start = loops$start2, end = loops$end2, loop_id = loops$loop_id)
 83 |   #Merge with original to add back the lost info
 84 |   return(list(loops.1, loops.2))
 85 | }
 86 | #Then make Granges
 87 | loops.anchors.list <- make_anchors_separate(loops)
 88 | #Make separate dfs (maybe don't need to do this?)
 89 | loops.anchors.1 <- loops.anchors.list[[1]]
 90 | loops.anchors.2 <- loops.anchors.list[[2]]
 91 | #Make granges
 92 | loops.anchors.list.granges <- lapply(loops.anchors.list, makeGRangesFromDataFrame, keep.extra.columns = FALSE)
 93 | 
 94 | #########################################################################
 95 | #Compare Granges objects
 96 | #Start counter
 97 | feature.count <- 1
 98 | 
 99 | #Loop through the features and count overlaps
100 | for (feature in features.data.granges) {
101 |     loops.anchors.1[[idlist[feature.count]]] <- countOverlaps(loops.anchors.list.granges[[1]], feature)
102 |     loops.anchors.2[[idlist[feature.count]]] <- countOverlaps(loops.anchors.list.granges[[2]], feature)
103 |     feature.count <- feature.count + 1
104 | }
105 | 
106 | 
107 | #Change column names for loops.anchors.2 so they don't match
108 | colnames(loops.anchors.2) <- paste0(colnames(loops.anchors.2), '2')
109 | 
110 | #Next, merge based on loop_id(2) columns
111 | loops.anchors.remerge <- merge(loops.anchors.1, loops.anchors.2, by.x = "loop_id", by.y = "loop_id2")
112 | 
113 | ########################################################################
114 | #Determine loop classes
115 | 
116 | #First need to generate loop classes based on ids
117 | #Add null to the class list (for anchors with no features)
118 | idlist.null <- append(idlist, "null")
119 | 
120 | #Then make all combinations (this is combinations with replacement)
121 | position <- 1
122 | combination.object <- icombinations(idlist.null, k = 2, replace = TRUE)
123 | 
124 | id.combined.list <- vector("list", length(combination.object$collect())/2)
125 | 
126 | foreach(x = icombinations(idlist.null, k = 2, replace = TRUE), .combine = c) %do% {
127 |   id.combined.list[[position]] <- paste(idlist.null[x[1]], idlist.null[x[2]], sep = "-")
128 |   position <- position + 1
129 | }
130 | 
131 | #Classify loops. Which version is run depends on whether inclusive or exclusive loops are desired.
132 | if (args$exclusive) {
133 |   #Make named list of dfs
134 |   output.list <- setNames(replicate(length(id.combined.list), data.frame()), id.combined.list)
135 | 
136 |   #Exclusive version
137 |   for (i1 in 1:length(idlist)) {
138 |     item1 <- idlist[[i1]]
139 |     #print(item1)
140 |     for (i2 in 1:length(idlist)) {
141 |       item2 <- idlist[[i2]]
142 |       #print(item2)
143 |       if (i1 == i2 | i1 < i2) {
144 |         #Get the loop type from the indices
145 |         looptype <- paste(idlist[[i1]], idlist[[i2]], sep = '-')
146 |         #Get the relevant loops
147 |         temp.df <- loops.anchors.remerge
148 |         #Loop through all ids, check they're 0 except the ones matching the requirements
149 |         for (id in idlist) {
150 |           if (id == idlist[[i1]]) {
151 |             temp.df <- temp.df[which(temp.df[[id]] > 0),]
152 |           } else if (id != idlist[[i1]]) {
153 |             temp.df <- temp.df[which(temp.df[[id]] == 0),]
154 |           }
155 |           if (id == idlist[[i2]]) {
156 |             temp.df <- temp.df[which(temp.df[[paste0(id, '2')]] > 0),]
157 |           } else if (id != idlist[[i2]]) {
158 |             temp.df <- temp.df[which(temp.df[[paste0(id, '2')]] == 0),]
159 |           }
160 |         }
161 |         output.list[[looptype]] <- rbind(output.list[[looptype]], temp.df)
162 | 
163 |       } else if (i1 > i2) {
164 |         #Get the loop type from the indices - need to invert here so that E-P2 and P-E2 loops are both put into
165 |         #the same category
166 |         looptype <- paste(idlist[[i2]], idlist[[i1]], sep = '-')
167 |         #Get the relevant loops
168 |         temp.df <- loops.anchors.remerge
169 | 
170 |         for (id in idlist) {
171 |           if (id == idlist[[i1]]) {
172 |             temp.df <- temp.df[which(temp.df[[id]] > 0),]
173 |           } else if (id != idlist[[i1]]) {
174 |             temp.df <- temp.df[which(temp.df[[id]] == 0),]
175 |           }
176 |           if (id == idlist[[i2]]) {
177 |             temp.df <- temp.df[which(temp.df[[paste0(id, '2')]] > 0),]
178 |           } else if (id != idlist[[i2]]) {
179 |             temp.df <- temp.df[which(temp.df[[paste0(id, '2')]] == 0),]
180 |           }
181 |         }
182 |         output.list[[looptype]] <- rbind(output.list[[looptype]], temp.df)
183 |       }
184 |     }
185 |     looptype <- paste(idlist[[i1]], "null", sep = '-')
186 |     #Generate X-null
187 |     temp.df.1 <- loops.anchors.remerge
188 |     #Loop through the ID columns of the other anchor and select only rows with 0 for each column
189 |     for (id in idlist) {
190 |       if (id == idlist[[i1]]) {
191 |         temp.df.1 <- temp.df.1[which(temp.df.1[[id]] > 0),]
192 |       } else if (id != idlist[[i1]]) {
193 |         temp.df.1 <- temp.df.1[which(temp.df.1[[id]] == 0),]
194 |       }
195 |       temp.df.1 <- temp.df.1[which(temp.df.1[[paste0(id, '2')]] == 0),]
196 |     }
197 |     #Generate null-X
198 |     temp.df.2 <- loops.anchors.remerge[which(loops.anchors.remerge[[paste0(idlist[[i1]], '2')]] > 0),]
199 | 
200 |     for (id in idlist) {
201 |       temp.df.2 <- temp.df.2[which(temp.df.2[[id]] == 0),]
202 |       if (id == idlist[[i1]]) {
203 |         temp.df.2 <- temp.df.2[which(temp.df.2[[paste0(id, '2')]] > 0),]
204 |       } else if (id != idlist[[i1]]) {
205 |         temp.df.2 <- temp.df.2[which(temp.df.2[[paste0(id, '2')]] == 0),]
206 |       }
207 |     }
208 |     #Combine them
209 |     temp.df <- rbind(temp.df.1, temp.df.2)
210 |     output.list[[looptype]] <- rbind(output.list[[looptype]], temp.df)
211 | 
212 |   }
213 |   looptype <- "null-null"
214 |   temp.df <- loops.anchors.remerge
215 | 
216 |   for (id in idlist) {
217 |     temp.df <- temp.df[which(temp.df[[id]] == 0 & temp.df[[paste0(id, '2')]] == 0),]
218 |   }
219 |   output.list[[looptype]] <- rbind(output.list[[looptype]], temp.df)
220 | } else {
221 |   #Make named list of dfs
222 |   output.list <- setNames(replicate(length(id.combined.list), data.frame()), id.combined.list)
223 | 
224 |   #Inclusive version
225 |   for (i1 in 1:length(idlist)) {
226 |     item1 <- idlist[[i1]]
227 |     #print(item1)
228 |     for (i2 in 1:length(idlist)) {
229 |       item2 <- idlist[[i2]]
230 |       #print(item2)
231 |       if (i1 == i2 | i1 < i2) {
232 |         #Get the loop type from the indices
233 |         looptype <- paste(idlist[[i1]], idlist[[i2]], sep = '-')
234 |         #Get the relevant loops
235 |         temp.df <- loops.anchors.remerge[which(loops.anchors.remerge[[idlist[[i1]]]] > 0 & loops.anchors.remerge[[paste0(idlist[[i2]], "2")]] > 0), ]
236 |         output.list[[looptype]] <- rbind(output.list[[looptype]], temp.df)
237 |       } else if (i1 > i2) {
238 |         #Get the loop type from the indices - need to invert here so that E-P2 and P-E2 loops are both put into
239 |         #the same category
240 |         looptype <- paste(idlist[[i2]], idlist[[i1]], sep = '-')
241 |         #Get the relevant loops
242 |         temp.df <- loops.anchors.remerge[which(loops.anchors.remerge[[idlist[[i1]]]] > 0 & loops.anchors.remerge[[paste0(idlist[[i2]], "2")]] > 0), ]
243 |         output.list[[looptype]] <- rbind(output.list[[looptype]], temp.df)
244 |       }
245 |     }
246 |     looptype <- paste(idlist[[i1]], "null", sep = '-')
247 |     #Generate X-null
248 |     temp.df.1 <- loops.anchors.remerge[which(loops.anchors.remerge[[idlist[[i1]]]] > 0),]
249 |     #Loop through the ID columns of the other anchor and select only rows with 0 for each column
250 |     for (id in idlist) {
251 |       temp.df.1 <- temp.df.1[which(temp.df.1[[paste0(id, '2')]] == 0),]
252 |     }
253 |     #Generate null-X
254 |     temp.df.2 <- loops.anchors.remerge[which(loops.anchors.remerge[[paste0(idlist[[i1]], '2')]] > 0),]
255 | 
256 |     for (id in idlist) {
257 |       temp.df.2 <- temp.df.2[which(temp.df.2[[id]] == 0),]
258 |     }
259 |     #Combine them
260 |     temp.df <- rbind(temp.df.1, temp.df.2)
261 |     output.list[[looptype]] <- rbind(output.list[[looptype]], temp.df)
262 | 
263 |   }
264 |   looptype <- "null-null"
265 |   temp.df <- loops.anchors.remerge
266 | 
267 |   for (id in idlist) {
268 |     temp.df <- temp.df[which(temp.df[[id]] == 0 & temp.df[[paste0(id, '2')]] == 0),]
269 |   }
270 |   output.list[[looptype]] <- rbind(output.list[[looptype]], temp.df)
271 | }
272 | 
273 | #Remove any duplicates from the dfs (inclusive calling can result in many)
274 | output.list.nodups <- lapply(output.list, distinct)
275 | 
276 | #Print the lengs of each df to give the result:
277 | sapply(output.list.nodups, nrow)
278 | 
279 | #Output the dataframes
280 | for (i in 1:length(output.list.nodups)) {
281 |   temp.df <- output.list.nodups[[i]]
282 |   #Get only the relevant columns
283 |   output.df <- data.frame(chr = temp.df$chr, start = temp.df$start, end = temp.df$end, chr2 = temp.df$chr2, start2 = temp.df$start2, end2 = temp.df$end2)
284 |   write.table(output.df, file = paste0(outdir, names(output.list)[[i]], '.bedpe'), row.names = FALSE, col.names = FALSE, quote = FALSE, sep = "\t")
285 | }
286 | 


--------------------------------------------------------------------------------
/microcbowtie2.py:
--------------------------------------------------------------------------------
  1 | #Aligning and processing a single fastq file through a pipeline similar to distiller but instead using bowtie2
  2 | 
  3 | from sys import exit
  4 | import subprocess as sp
  5 | import argparse
  6 | import multiprocessing
  7 | import uuid
  8 | 
  9 | parser = argparse.ArgumentParser(description = "run bowtie2 and pairtools on fastq files to produce pairsam files")
 10 | bamopts = parser.add_mutually_exclusive_group()
 11 | parser.add_argument("--file_1", "-1", help = "first demuxed fastq of paired end reads - required", nargs = "*")
 12 | parser.add_argument("--file_2", "-2", help = "second demuxed fastq of paired end reads - required", nargs = "*")
 13 | parser.add_argument("--genome", "-g", help = "genome to align to - mouse or human - required")
 14 | parser.add_argument("--genometype", "-y", help = "genome type - use if your genome is a modified version of a standard genome - should be one of hg19, hg38, mm10, or mm39")
 15 | parser.add_argument("--threads", "-t", help = "number of threads to use for bowtie2 - default is 1", default = "1")
 16 | parser.add_argument("--resolutions", "-r", help = "list of resolutions to output in decreasing order - all resolutions must be a multiple of the smallest resolution - default: 10000000 5000000 2500000 1000000 500000 250000 100000 50000 25000 10000 5000 2000 1000", nargs = "*")
 17 | parser.add_argument("--out", "-o", help = "name for output files - defaults to name of first file")
 18 | parser.add_argument("--outdir", help = "a directory to store output files - default is current directory", default = "./")
 19 | bamopts.add_argument("--bowtieonly", "-b", help = "only run bowtie2 and make bams - can be useful for post initial analysis QC steps", action = "store_true")
 20 | bamopts.add_argument("--keepbams", "-k", help = "keep bam files while doing a normal full analysis", action = "store_true")
 21 | args = parser.parse_args()
 22 | 
 23 | file1 = args.file_1
 24 | file2 = args.file_2
 25 | genome = args.genome
 26 | gentype = args.genometype
 27 | threads = args.threads
 28 | outname = args.out
 29 | outdir = args.outdir
 30 | reslist = args.resolutions
 31 | bowtieonly = args.bowtieonly
 32 | keepbams = args.keepbams
 33 | 
 34 | #Check requirements are fulfilled:
 35 | condapacks = sp.run("conda list".split(), capture_output=True)
 36 | condapacksstr = str(condapacks.stdout)
 37 | if bowtieonly:
 38 |     if "bowtie2" not in condapacksstr or "samtools" not in condapacksstr or "sambamba" not in condapacksstr:
 39 |         print("Please make sure bowtie2, samtools and sambamba are installed in your current conda environment (check conda list)")
 40 |         exit()
 41 | elif not bowtieonly:
 42 |     if "bowtie2" not in condapacksstr or "pairtools" not in condapacksstr or "cooler" not in condapacksstr or "pairix" not in condapacksstr:
 43 |         print("Please make sure bowtie2, pairtools, pairix and cooler are installed in your current conda environment (check with 'conda list')")
 44 |         exit()
 45 | 
 46 | #Check that outdir ends with a /, add one if it doesn't
 47 | if args.outdir is not None and not outdir.endswith("/"):
 48 |     outdir = outdir + "/"
 49 | 
 50 | #Check a genome was specified
 51 | if args.genome is None:
 52 |     print("Genome not specified - check help for formatting")
 53 |     parser.print_usage()
 54 |     exit()
 55 | 
 56 | #Check if file1 and file2 are single files or multiple
 57 | if file1 is None or file2 is None:
 58 |     print("Input files not specified - check help for formatting")
 59 |     parser.print_usage()
 60 |     exit()
 61 | elif len(file1) > 1 and type(file1) == list and type(file2) == list and len(file1) == len(file2):
 62 |     multifile = 1
 63 |     if args.out is None:
 64 |         outlist = [fname + "_" + genome for fname in file1]
 65 |         outname = outlist[0]
 66 |     else:
 67 |         outlist = list()
 68 |         for i in range(len(file1)):
 69 |             outlist.append(outname + "_" + str(i + 1))
 70 |     #Input to pair merging step needs all of the outputs together
 71 |     pairnamelist = [outdir + oname + ".pairs" for oname in outlist]
 72 |     pairnamest = " ".join(pairnamelist)
 73 |     bamlist = [outdir + oname + ".sorted.bam" for oname in outlist]
 74 |     bamst = " ".join(bamlist)
 75 |     bailist = [outdir + oname + ".sorted.bam.bai" for oname in outlist]
 76 |     baist = " ".join(bailist)
 77 | elif len(file1) == 1 and len(file2) == 1:
 78 |     multifile = 0
 79 |     #If nargs = *, always makes a list, even if only one element
 80 |     file1 = "".join(file1)
 81 |     file2 = "".join(file2)
 82 |     if args.out is None:
 83 |         outname = file1 + "_" + genome
 84 |     pairnamest = outdir + outname + ".pairs"
 85 | else:
 86 |     print("Mismatch in number of input files, check arguments")
 87 |     exit()
 88 | 
 89 | #Check that a sensible number of threads has been requested - more protections here are possible - at the moment users are trusted to be sensible
 90 | cpucount = multiprocessing.cpu_count()
 91 | if args.threads is None:
 92 |     print("Defaulting to one thread")
 93 |     threads = 1
 94 | elif int(args.threads) >= cpucount:
 95 |     print("Too many threads requested, resetting to default")
 96 |     threads = 1
 97 | 
 98 | #Check that the user has entered a valid genome to align to
 99 | if args.genometype is None:
100 |     gentype = args.genome
101 | 
102 | if gentype == "mm10" or gentype == "mm39":
103 |     toprint = "Aligning to mouse genome {}".format(genome)
104 |     print(toprint)
105 | elif gentype == "hg19" or gentype == "hg38":
106 |     toprint = "Aligning to human genome {}".format(genome)
107 |     print(toprint)
108 | else:
109 |     if gentype == genome:
110 |         print("Genome option not recognised or not entered. Please use mm10/39 or hg19/38 or ask Miles to change the script to accommodate your new organism/genome. If you are using a modified version of base genome, use the -g option to indicate the base genome name.")
111 |         exit()
112 |     else: #If they're using a modified genome, make sure the base genome exists so that the files are redirected properly
113 |         print("Genome/base genome option not recognised. Please use mm10/39 or hg19/38 or ask Miles to change the script to accommodate your new organism/genome.")
114 |         exit()
115 | 
116 | #Set up resolutions as needed
117 | if args.resolutions is None:
118 |     reslist = ["10000000", "5000000", "2500000", "1000000", "500000", "250000", "100000", "50000", "25000", "10000", "5000", "2000", "1000"]
119 | resst = ",".join(reslist)
120 | #Extract minimum resolution
121 | minres = reslist[-1]
122 | 
123 | #Process ID (used to make unique sorttemp, so these are not overlapping for multiple processes in the same outdir)
124 | uniqueid = str(uuid.uuid4())
125 | 
126 | # commands as strings
127 | line1 = "mkdir {0}{10}sorttemp -p"
128 | line2 = "bowtie2 -x /mnt/md0/DataRepository/genomes/{1}/{2} --threads {3} -1 {4} -2 {5} --reorder --local --very-sensitive-local {11}| pairtools parse --add-columns mapq --walks-policy mask -c /mnt/md0/DataRepository/chromsizes/{1}/{2}.sorted.chrom.sizes --assembly {2} --min-mapq 2 --drop-sam --drop-readid --nproc-in {3} | pairtools sort --tmpdir {0}{10}sorttemp --nproc {3} -o {0}{6}.pairs | cat" #Can add drop-sam and drop-readid options later
129 | line3 = "pairtools merge --tmpdir {0}{10}sorttemp --nproc {3} {7} | pairtools dedup --max-mismatch 1 --mark-dups --output {0}{6}.nodups.pairs.gz --output-unmapped {0}{6}.unmapped.pairs.gz --output-dups {0}{6}.dups.pairs.gz --output-stats {0}{6}.dedup.stats | cat"
130 | line4 = "pairtools dedup --max-mismatch 1 --mark-dups --output {0}{6}.nodups.pairs.gz --output-unmapped {0}{6}.unmapped.pairs.gz --output-dups {0}{6}.dups.pairs.gz --output-stats {0}{6}.dedup.stats {0}{7}"
131 | line5 = "pairix {0}{6}.nodups.pairs.gz"
132 | line6 = "bgzip -cd -@ 3 {0}{6}.nodups.pairs.gz | cooler cload pairs -c1 2 -p1 3 -c2 4 -p2 5 --assembly {2} /mnt/md0/DataRepository/chromsizes/{1}/{2}.sorted.chrom.sizes:{8} - {0}{6}.{8}.cool"
133 | line7 = "cooler zoomify --nproc {3} --balance --out {0}{6}.{8}.mcool --resolutions {9} {0}{6}.{8}.cool"
134 | line8 = "rmdir {0}{10}sorttemp"
135 | 
136 | #For running only bowtie2 and making bams
137 | bline1 = "mkdir {0}{6}temp -p"
138 | bline2 = "bowtie2 -x /mnt/md0/DataRepository/genomes/{1}/{2} --threads {3} -1 {4} -2 {5} --reorder --local --very-sensitive-local | samtools view -bS -o {0}{6}.bam"
139 | bline3 = "sambamba sort -t {3} -m 6GB --tmpdir {0}{7}temp {0}{6}.bam {0}{6}.sorted.bam && rm {0}{6}.bam"
140 | bline4 = "sambamba merge -t {3} {0}{6}.sorted.merged.bam {7} && rm {7} {8}"
141 | bline5 = "sambamba markdup -t {3} --tmpdir {0}{6}temp --overflow-list-size 10000000 -r {0}{6}.sorted.merged.bam {0}{6}.nodups.sorted.merged.bam && rm {0}{6}.sorted.merged.ba*"
142 | bline6 = "sambamba markdup -t {3} --tmpdir {0}{6}temp -r {0}{6}.sorted.bam {0}{6}.nodups.sorted.bam && rm {0}{6}.sorted.ba*"
143 | bline7 = "rmdir {0}{6}temp"
144 | 
145 | multilines = [line1, line2, line3, line5, line6, line7, line8]
146 | lines = [line1, line2, line4, line5, line6, line7, line8]
147 | 
148 | multiblines = [bline1, bline2, bline3, bline4, bline5, bline7]
149 | blines = [bline1, bline2, bline3, bline6, bline7]
150 | 
151 | truncmultiblines = [bline1, bline3, bline4, bline5, bline7]
152 | truncblines = [bline1, bline3, bline6, bline7]
153 | 
154 | #Process the files depending on the run mode
155 | if not multifile and not bowtieonly:
156 |     #Include command if bams are wanted
157 |     if keepbams:
158 |         keepbamcmd = "| tee >(samtools view -bS > {0}{1}.bam) ".format(outdir, outname)
159 |     else:
160 |         keepbamcmd = ""
161 |     for line in lines:
162 |         # add file name and split by whitespace
163 |         tokenized_line = line.format(outdir, gentype, genome, threads, file1, file2, outname, pairnamest, minres, resst, uniqueid, keepbamcmd)
164 |         print(tokenized_line)
165 |         # run
166 |         sp.run(tokenized_line, shell=True, executable="/bin/bash")
167 | elif multifile and not bowtieonly:
168 |     for line in multilines:
169 |         if line == line2:
170 |             for x in range(len(file1)):
171 |                 #Include command if bams are wanted
172 |                 if keepbams:
173 |                     keepbamcmd = "| tee >(samtools view -bS > {0}{1}.bam) ".format(outdir, outlist[x])
174 |                 else:
175 |                     keepbamcmd = ""
176 |                 # add file name and split by whitespace
177 |                 tokenized_line = line.format(outdir, gentype, genome, threads, file1[x], file2[x], outlist[x], pairnamest, minres, resst, uniqueid, keepbamcmd)
178 |                 print(tokenized_line)
179 |                 # run
180 |                 sp.run(tokenized_line, shell=True, executable="/bin/bash")
181 |         else:
182 |             # add file name and split by whitespace
183 |             tokenized_line = line.format(outdir, gentype, genome, threads, file1[1], file2[1], outname, pairnamest, minres, resst, uniqueid)
184 |             print(tokenized_line)
185 |             # run
186 |             sp.run(tokenized_line, shell=True)
187 | elif bowtieonly and multifile:
188 |     for line in multiblines:
189 |         if line == bline2 or line == bline3:
190 |             for x in range(len(file1)):
191 |                 tokenized_line = line.format(outdir, gentype, genome, threads, file1[x], file2[x], outlist[x], outname)
192 |                 print(tokenized_line)
193 |                 sp.run(tokenized_line, shell=True)
194 |         else:
195 |             tokenized_line = line.format(outdir, gentype, genome, threads, file1[1], file2[1], outname, bamst, baist)
196 |             print(tokenized_line)
197 |             sp.run(tokenized_line, shell=True)
198 | elif bowtieonly and not multifile:
199 |     for line in blines:
200 |         tokenized_line = line.format(outdir, gentype, genome, threads, file1, file2, outname, outname)
201 |         print(tokenized_line)
202 |         sp.run(tokenized_line, shell=True)
203 | 
204 | #After everything finishes, merge and process bams as required if doing full analysis
205 | if keepbams and not bowtieonly:
206 |     if not multifile:
207 |         for line in truncblines:
208 |             tokenized_line = line.format(outdir, gentype, genome, threads, file1, file2, outname, outname)
209 |             print(tokenized_line)
210 |             sp.run(tokenized_line, shell=True)
211 |     elif multifile:
212 |         for line in truncmultiblines:
213 |             if line == bline3:
214 |                 for x in range(len(file1)):
215 |                     tokenized_line = line.format(outdir, gentype, genome, threads, file1[x], file2[x], outlist[x], outname)
216 |                     print(tokenized_line)
217 |                     sp.run(tokenized_line, shell=True)
218 |             else:
219 |                 tokenized_line = line.format(outdir, gentype, genome, threads, file1[1], file2[1], outname, bamst, baist)
220 |                 print(tokenized_line)
221 |                 sp.run(tokenized_line, shell=True)
222 | 


--------------------------------------------------------------------------------
/spikeinChIP_PE_alignment.py:
--------------------------------------------------------------------------------
  1 | #Aligning and processing paired end fastq files for spike-in ChIP-seq using bowtie2
  2 | 
  3 | from sys import exit
  4 | import subprocess as sp
  5 | import argparse
  6 | import multiprocessing
  7 | import uuid
  8 | import pandas as pd
  9 | 
 10 | parser = argparse.ArgumentParser(description = "run bowtie2 on paired end fastq files with spikein to produce aligned bam files")
 11 | parser.add_argument("--filename", "-f", help = "a tab-separated file containing one each line the path to the fastq with the first ends of pairs, the path to the fastq with the second ends of pairs, and the desired output name for the aligned file")
 12 | # parser.add_argument("--file_1", "-1", help = "first demuxed fastq of paired end reads - required", nargs = "*")
 13 | # parser.add_argument("--file_2", "-2", help = "second demuxed fastq of paired end reads - required", nargs = "*")
 14 | parser.add_argument("--genome", "-g", help = "genome build to align to - mouse or human - required")
 15 | parser.add_argument("--spikegenome", "-s", help = "spikein genome build to align to - mouse or human - required")
 16 | parser.add_argument("--threads", "-t", help = "number of threads to use for bowtie2 - default is 1", default = "1")
 17 | parser.add_argument("--outname", "-o", help = "name for the table to store counts")
 18 | parser.add_argument("--outdir", help = "a directory to store output files - default is current directory", default = "./")
 19 | args = parser.parse_args()
 20 | 
 21 | # file1 = args.file_1
 22 | # file2 = args.file_2
 23 | genome = args.genome
 24 | spikegenome = args.spikegenome
 25 | threads = args.threads
 26 | outname = args.outname
 27 | outdir = args.outdir
 28 | 
 29 | #Read in file to determine what to process
 30 | files = pd.read_csv(args.filename, sep='\t', header=None, names = ['end1', 'end2', 'name'])
 31 | file1 = files.end1.values.tolist()
 32 | file2 = files.end2.values.tolist()
 33 | outnames = files.name.values.tolist()
 34 | 
 35 | #Check requirements are fulfilled:
 36 | condapacks = sp.run("conda list".split(), capture_output=True)
 37 | condapacksstr = str(condapacks.stdout)
 38 | 
 39 | if "bowtie2" not in condapacksstr or "samtools" not in condapacksstr or "sambamba" not in condapacksstr:
 40 |     print("Please make sure bowtie2, samtools and sambamba are installed in your current conda environment (check conda list)")
 41 |     exit()
 42 | 
 43 | #Check that outdir ends with a /, add one if it doesn't
 44 | if args.outdir is not None and not outdir.endswith("/"):
 45 |     outdir = outdir + "/"
 46 | 
 47 | #Check a genome was specified
 48 | if args.genome is None or args.spikegenome is None:
 49 |     print("Genomes not specified - check help for formatting")
 50 |     parser.print_usage()
 51 |     exit()
 52 | 
 53 | #Check if file1, file2, and outnames are present and the same lengths
 54 | if file1 is None or file2 is None or outnames is None:
 55 |     print("Input files not specified - check help for formatting")
 56 |     parser.print_usage()
 57 |     exit()
 58 | elif type(file1) == list and type(file1) == list and type(outnames) == list and len(file1) > 1:
 59 |     if len(file1) != len(file2) or len(file1) != len(outnames) or len(file2) != len(outnames):
 60 |         print("Mismatch in number of input files or output names, check arguments")
 61 |         exit()
 62 | 
 63 | #Check that a sensible number of threads has been requested - more protections here are possible - at the moment users are trusted to be sensible
 64 | cpucount = multiprocessing.cpu_count()
 65 | if args.threads is None:
 66 |     print("Defaulting to one thread")
 67 |     threads = 1
 68 | elif int(args.threads) >= cpucount:
 69 |     print("Too many threads requested, resetting to default")
 70 |     threads = 1
 71 | else:
 72 |     print(f"Running alignment with {threads} threads...")
 73 | 
 74 | #Check that the user has entered a valid genome to align to
 75 | if genome == "mm10" or genome == "mm39":
 76 |     print(f"Aligning to mouse genome {genome}")
 77 | elif genome == "hg19" or genome == "hg38":
 78 |     print(f"Aligning to human genome {genome}")
 79 | else:
 80 |     print("Genome option not recognised or not entered. Please use mm10/39 or hg19/38 or ask Miles to change the script to accommodate your new organism/genome")
 81 |     exit()
 82 | 
 83 | #Process ID (used to make unique sorttemp, so these are not overlapping for multiple processes in the same outdir)
 84 | uniqueid = str(uuid.uuid4())
 85 | 
 86 | #Create a place to store counts
 87 | allcountslist = []
 88 | 
 89 | #Process the files
 90 | for i, fastq1 in enumerate(file1):
 91 |     fastq2 = file2[i]
 92 |     name = outnames[i]
 93 |     print(f"Aligning {name} to {genome} and {spikegenome}")
 94 |     #Explanation: align to genome and spikein genome, then remove multiply aligned reads with grep (XS: indicates the score of the next best aligning read, if it exists), then make a bam of all mapped reads (-F4 removes reads with a SAM flag of 4, which means unmapped)
 95 |     sp.run(f"bowtie2 -p {threads} --no-mixed --no-discordant -1 {fastq1} -2 {fastq2} -x /mnt/md0/DataRepository/genomes/{genome}.{spikegenome}/{genome}.{spikegenome} | grep -v XS: - | samtools view -bh -F4 - > {outdir}{name}_UniqMapped.bam", shell=True, executable="/bin/bash")
 96 |     #Sort aligned reads
 97 |     sp.run(f"sambamba sort --tmpdir {outdir}{uniqueid}/ -t {threads} -m 30G -o {outdir}{name}_UniqMapped_sorted.bam {outdir}{name}_UniqMapped.bam", shell=True, executable="/bin/bash")
 98 |     #Remove duplicates
 99 |     sp.run(f"sambamba markdup --tmpdir {outdir}{uniqueid}/ -r -t {threads} {outdir}{name}_UniqMapped_sorted.bam {outdir}{name}_UniqMapped_sorted_rmdup.bam", shell=True, executable="/bin/bash")
100 | 
101 |     #Next, need to separate out reads from each genome:
102 |     print(f"Extracting reads aligning uniquely to {genome}")
103 |     #Use samtools view to open the file, grep to remove those with the spikein genome name in them, then change the chromosomes with the genome name to just the chromosome numbers as normally used, then put the file back into bam.
104 |     sp.run(f"samtools view -h {outdir}{name}_UniqMapped_sorted_rmdup.bam | grep -v {spikegenome} | sed s/{genome}_chr/chr/g | samtools view -bhS - > {outdir}{name}_{genome}.UniqMapped_sorted_rmdup.bam", shell=True, executable="/bin/bash")
105 |     #Do the same thing for the spikein
106 |     print(f"Extracting reads aligning uniquely to {spikegenome}.")
107 |     sp.run(f"samtools view -h {outdir}{name}_UniqMapped_sorted_rmdup.bam | grep -v {genome} | sed s/{spikegenome}_chr/chr/g | samtools view -bhS - > {outdir}{name}_{spikegenome}.UniqMapped_sorted_rmdup.bam", shell=True, executable="/bin/bash")
108 |     #Index outputs
109 |     sp.run(f"sambamba index -t {threads} {outdir}{name}_UniqMapped_sorted_rmdup.bam", shell=True, executable="/bin/bash")
110 |     sp.run(f"sambamba index -t {threads} {outdir}{name}_{genome}.UniqMapped_sorted_rmdup.bam", shell=True, executable="/bin/bash")
111 |     sp.run(f"sambamba index -t {threads} {outdir}{name}_{spikegenome}.UniqMapped_sorted_rmdup.bam", shell=True, executable="/bin/bash")
112 |     #Now clean up
113 |     sp.run(f"rm {outdir}{name}_UniqMapped.bam", shell=True, executable="/bin/bash")
114 |     sp.run(f"rm {outdir}{name}_UniqMapped_sorted.bam", shell=True, executable="/bin/bash")
115 | 
116 |     #Finally, count reads in each file:
117 |     totalCount = sp.run(f"sambamba view -c -t {threads} {outdir}{name}_UniqMapped_sorted_rmdup.bam", capture_output=True, shell=True, executable="/bin/bash")
118 |     genomeCount = sp.run(f"sambamba view -c -t {threads} {outdir}{name}_{genome}.UniqMapped_sorted_rmdup.bam", capture_output=True, shell=True, executable="/bin/bash")
119 |     spikegenomeCount = sp.run(f"sambamba view -c -t {threads} {outdir}{name}_{spikegenome}.UniqMapped_sorted_rmdup.bam", capture_output=True, shell=True, executable="/bin/bash")
120 |     countsList = [totalCount.stdout.decode('ascii').strip(), genomeCount.stdout.decode('ascii').strip(), spikegenomeCount.stdout.decode('ascii').strip()]
121 |     allcountslist.append(countsList)
122 | 
123 | sp.run(f"rm {outdir}{uniqueid}/ -r", shell=True, executable="/bin/bash")
124 | countstable = pd.DataFrame(allcountslist, columns = ['allcounts', 'genomecounts', 'spikecounts'])
125 | outtable = pd.concat([files, countstable], axis=1)
126 | outtable.to_csv(outdir + outname + '.tsv', sep = '\t', index = False, header = True)
127 | 


--------------------------------------------------------------------------------