OligoFISSEQ pipeline.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## OligoFISSEQ analysis walkthrough\n",
    "\n",
    "Please check the README.md for the installation of the requirements"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Download Fiji or use you existing installation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Added new repo: scijava.public\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "method": "display_data"
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "5f388668-6442-47e1-b504-0fc5583fab77",
       "version_major": 2,
       "version_minor": 0
      },
      "method": "display_data"
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "null"
      ]
     },
     "execution_count": 1,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "// The download from maven can take some minutes\n",
    "%classpath config resolver scijava.public https://maven.scijava.org/content/groups/public\n",
    "%classpath add mvn sc.fiji fiji 2.0.0-pre-10\n",
    "// or use and existing Fiji installation \n",
    "// %classpath add jar \"/home/dcastillo/Fiji.app/jars/*\"\n",
    "// %classpath add jar \"/home/dcastillo/Fiji.app/jars/bio-formats/*\"\n",
    "// %classpath add jar \"/home/dcastillo/Fiji.app/plugins/*\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2.0.0-rc-71/1.52i"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "New dimensions: 526 522 22\n"
     ]
    }
   ],
   "source": [
    "ij = new net.imagej.ImageJ()\n",
    "ij.getVersion()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Check input data\n",
    "\n",
    "The OligoFISSEQ images are stored in the input/ folder.\n",
    "\n",
    "The input folder should contain one folder per OligoFISSEQ round. The folders should be named round_ folowed by the round order (i.e. round_1, round_2, ...)\n",
    "\n",
    "Images indide each round folder that need to be aligned should have the same name.\n",
    "\n",
    "In this example we are going to process an image of the 36plex-5k dataset with 4 rounds of OligoFISSEQ"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "/home/dcastillo/scratch/workspace/oligoFISSEQ/Notebooks/test/input"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "File input_images_folder = new File('input/');\n",
    "input_path = input_images_folder.getAbsolutePath();"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Align the images of 4 rounds of OligoFISSEQ\n",
    "\n",
    "Run the macro Fiji macro \"create_hyperstack_from_rounds.ijm\" in the input folder to align and segment the images of the 4 rounds of OligoFISSEQ. Resulting hyperstacks will be saved under the merged/ folder"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "null"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "align_macro = new File(\"scripts/create_hyperstack_from_rounds.ijm\");\n",
    "result = ij.IJ.runMacro(\"res=runMacro('\"+align_macro.getAbsolutePath()+\"','\"+input_path+\"');\");"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Detect barcodes\n",
    "\n",
    "The script detect_barcode_patches.py will be used to detect the barcodes in the hyperstacks. We will need to specify the number of rounds and a csv file with the designed oligopaints barcodes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "/home/dcastillo/scratch/workspace/oligoFISSEQ/Notebooks/test/input/probes_36plex_sharing.csv"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import java.nio.file.Path; \n",
    "import java.nio.file.Paths;\n",
    "\n",
    "File merged_images_folder = new File('input/merged/');\n",
    "merged_path = merged_images_folder.getAbsolutePath();\n",
    "barcode_file = Paths.get(input_path, \"probes_36plex_sharing.csv\").toString();"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The format of the barcode file should be the following:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "  Probe Chr.  Chr. Arm Size      Start        End   Length  Oligos  Probes/Kb  \\\n",
      "0  2pR1    2           93.3    1002895    1660898   658003    5000   7.598750   \n",
      "1  2pR2    2           93.3   47000022   47664134   664112    5000   7.528851   \n",
      "2  2pR3    2           93.3   85003257   85751873   748616    5000   6.678992   \n",
      "3  2qR1    2          149.7   98013273   99071153  1057880    5000   4.726434   \n",
      "4  2qR2    2          149.7  175000457  175884274   883817    5000   5.657280   \n",
      "\n",
      "   Readout  \n",
      "0    11243  \n",
      "1    43131  \n",
      "2    31341  \n",
      "3    13222  \n",
      "4    42144  \n"
     ]
    }
   ],
   "source": [
    "%%python\n",
    "import pandas as pd\n",
    "\n",
    "barcodes = pd.read_csv(\"input/probes_36plex_sharing.csv\", sep='\\t')\n",
    "print(barcodes.head())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "java.util.concurrent.FutureTask@900b3b6"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "/home/dcastillo/scratch/workspace/oligoFISSEQ/Notebooks/test/input/merged/OFQv61_w3_3_decon/roi_1.tif\n",
      "Using existing roi file /home/dcastillo/scratch/workspace/oligoFISSEQ/Notebooks/test/input/merged/OFQv61_w3_3_decon/roi_1_roi.zip\n",
      "Compiling round 0\n",
      "\n",
      "Compiling slice 0\n",
      "\n",
      "Compiling slice 1\n",
      "\n",
      "Compiling slice 2\n",
      "\n",
      "Compiling slice 3\n",
      "\n",
      "Compiling slice 4\n",
      "\n",
      "Compiling slice 5\n",
      "\n",
      "Compiling slice 6\n",
      "\n",
      "Compiling slice 7\n",
      "\n",
      "Compiling slice 8\n",
      "\n",
      "Compiling slice 9\n",
      "\n",
      "Compiling slice 10\n",
      "\n",
      "Compiling slice 11\n",
      "\n",
      "Compiling slice 12\n",
      "\n",
      "Compiling slice 13\n",
      "\n",
      "Compiling slice 14\n",
      "\n",
      "Compiling slice 15\n",
      "\n",
      "Compiling round 1\n",
      "\n",
      "Compiling slice 0\n",
      "\n",
      "Compiling slice 1\n",
      "\n",
      "Compiling slice 2\n",
      "\n",
      "Compiling slice 3\n",
      "\n",
      "Compiling slice 4\n",
      "\n",
      "Compiling slice 5\n",
      "\n",
      "Compiling slice 6\n",
      "\n",
      "Compiling slice 7\n",
      "\n",
      "Compiling slice 8\n",
      "\n",
      "Compiling slice 9\n",
      "\n",
      "Compiling slice 10\n",
      "\n",
      "Compiling slice 11\n",
      "\n",
      "Compiling slice 12\n",
      "\n",
      "Compiling slice 13\n",
      "\n",
      "Compiling slice 14\n",
      "\n",
      "Compiling slice 15\n",
      "\n",
      "Compiling round 2\n",
      "\n",
      "Compiling slice 0\n",
      "\n",
      "Compiling slice 1\n",
      "\n",
      "Compiling slice 2\n",
      "\n",
      "Compiling slice 3\n",
      "\n",
      "Compiling slice 4\n",
      "\n",
      "Compiling slice 5\n",
      "\n",
      "Compiling slice 6\n",
      "\n",
      "Compiling slice 7\n",
      "\n",
      "Compiling slice 8\n",
      "\n",
      "Compiling slice 9\n",
      "\n",
      "Compiling slice 10\n",
      "\n",
      "Compiling slice 11\n",
      "\n",
      "Compiling slice 12\n",
      "\n",
      "Compiling slice 13\n",
      "\n",
      "Compiling slice 14\n",
      "\n",
      "Compiling slice 15\n",
      "\n",
      "Compiling round 3\n",
      "\n",
      "Compiling slice 0\n",
      "\n",
      "Compiling slice 1\n",
      "\n",
      "Compiling slice 2\n",
      "\n",
      "Compiling slice 3\n",
      "\n",
      "Compiling slice 4\n",
      "\n",
      "Compiling slice 5\n",
      "\n",
      "Compiling slice 6\n",
      "\n",
      "Compiling slice 7\n",
      "\n",
      "Compiling slice 8\n",
      "\n",
      "Compiling slice 9\n",
      "\n",
      "Compiling slice 10\n",
      "\n",
      "Compiling slice 11\n",
      "\n",
      "Compiling slice 12\n",
      "\n",
      "Compiling slice 13\n",
      "\n",
      "Compiling slice 14\n",
      "\n",
      "Compiling slice 15\n",
      "\n",
      "Barcodind slices 0\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 1\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 2\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 3\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 4\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 5\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 6\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 7\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 8\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 9\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 10\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 11\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 12\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 13\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 14\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 15\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "/home/dcastillo/scratch/workspace/oligoFISSEQ/Notebooks/test/input/merged/OFQv61_w3_3_decon/roi_8.tif\n",
      "Using existing roi file /home/dcastillo/scratch/workspace/oligoFISSEQ/Notebooks/test/input/merged/OFQv61_w3_3_decon/roi_8_roi.zip\n",
      "Compiling round 0\n",
      "\n",
      "Compiling slice 0\n",
      "\n",
      "Compiling slice 1\n",
      "\n",
      "Compiling slice 2\n",
      "\n",
      "Compiling slice 3\n",
      "\n",
      "Compiling slice 4\n",
      "\n",
      "Compiling slice 5\n",
      "\n",
      "Compiling slice 6\n",
      "\n",
      "Compiling slice 7\n",
      "\n",
      "Compiling slice 8\n",
      "\n",
      "Compiling slice 9\n",
      "\n",
      "Compiling slice 10\n",
      "\n",
      "Compiling slice 11\n",
      "\n",
      "Compiling slice 12\n",
      "\n",
      "Compiling slice 13\n",
      "\n",
      "Compiling slice 14\n",
      "\n",
      "Compiling slice 15\n",
      "\n",
      "Compiling round 1\n",
      "\n",
      "Compiling slice 0\n",
      "\n",
      "Compiling slice 1\n",
      "\n",
      "Compiling slice 2\n",
      "\n",
      "Compiling slice 3\n",
      "\n",
      "Compiling slice 4\n",
      "\n",
      "Compiling slice 5\n",
      "\n",
      "Compiling slice 6\n",
      "\n",
      "Compiling slice 7\n",
      "\n",
      "Compiling slice 8\n",
      "\n",
      "Compiling slice 9\n",
      "\n",
      "Compiling slice 10\n",
      "\n",
      "Compiling slice 11\n",
      "\n",
      "Compiling slice 12\n",
      "\n",
      "Compiling slice 13\n",
      "\n",
      "Compiling slice 14\n",
      "\n",
      "Compiling slice 15\n",
      "\n",
      "Compiling round 2\n",
      "\n",
      "Compiling slice 0\n",
      "\n",
      "Compiling slice 1\n",
      "\n",
      "Compiling slice 2\n",
      "\n",
      "Compiling slice 3\n",
      "\n",
      "Compiling slice 4\n",
      "\n",
      "Compiling slice 5\n",
      "\n",
      "Compiling slice 6\n",
      "\n",
      "Compiling slice 7\n",
      "\n",
      "Compiling slice 8\n",
      "\n",
      "Compiling slice 9\n",
      "\n",
      "Compiling slice 10\n",
      "\n",
      "Compiling slice 11\n",
      "\n",
      "Compiling slice 12\n",
      "\n",
      "Compiling slice 13\n",
      "\n",
      "Compiling slice 14\n",
      "\n",
      "Compiling slice 15\n",
      "\n",
      "Compiling round 3\n",
      "\n",
      "Compiling slice 0\n",
      "\n",
      "Compiling slice 1\n",
      "\n",
      "Compiling slice 2\n",
      "\n",
      "Compiling slice 3\n",
      "\n",
      "Compiling slice 4\n",
      "\n",
      "Compiling slice 5\n",
      "\n",
      "Compiling slice 6\n",
      "\n",
      "Compiling slice 7\n",
      "\n",
      "Compiling slice 8\n",
      "\n",
      "Compiling slice 9\n",
      "\n",
      "Compiling slice 10\n",
      "\n",
      "Compiling slice 11\n",
      "\n",
      "Compiling slice 12\n",
      "\n",
      "Compiling slice 13\n",
      "\n",
      "Compiling slice 14\n",
      "\n",
      "Compiling slice 15\n",
      "\n",
      "Barcodind slices 0\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 1\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 2\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 3\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 4\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 5\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 6\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 7\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 8\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 9\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 10\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 11\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 12\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 13\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 14\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 15\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "/home/dcastillo/scratch/workspace/oligoFISSEQ/Notebooks/test/input/merged/OFQv61_w3_3_decon/roi_6.tif\n",
      "Using existing roi file /home/dcastillo/scratch/workspace/oligoFISSEQ/Notebooks/test/input/merged/OFQv61_w3_3_decon/roi_6_roi.zip\n",
      "Compiling round 0\n",
      "\n",
      "Compiling slice 0\n",
      "\n",
      "Compiling slice 1\n",
      "\n",
      "Compiling slice 2\n",
      "\n",
      "Compiling slice 3\n",
      "\n",
      "Compiling slice 4\n",
      "\n",
      "Compiling slice 5\n",
      "\n",
      "Compiling slice 6\n",
      "\n",
      "Compiling slice 7\n",
      "\n",
      "Compiling slice 8\n",
      "\n",
      "Compiling slice 9\n",
      "\n",
      "Compiling slice 10\n",
      "\n",
      "Compiling slice 11\n",
      "\n",
      "Compiling slice 12\n",
      "\n",
      "Compiling slice 13\n",
      "\n",
      "Compiling slice 14\n",
      "\n",
      "Compiling slice 15\n",
      "\n",
      "Compiling round 1\n",
      "\n",
      "Compiling slice 0\n",
      "\n",
      "Compiling slice 1\n",
      "\n",
      "Compiling slice 2\n",
      "\n",
      "Compiling slice 3\n",
      "\n",
      "Compiling slice 4\n",
      "\n",
      "Compiling slice 5\n",
      "\n",
      "Compiling slice 6\n",
      "\n",
      "Compiling slice 7\n",
      "\n",
      "Compiling slice 8\n",
      "\n",
      "Compiling slice 9\n",
      "\n",
      "Compiling slice 10\n",
      "\n",
      "Compiling slice 11\n",
      "\n",
      "Compiling slice 12\n",
      "\n",
      "Compiling slice 13\n",
      "\n",
      "Compiling slice 14\n",
      "\n",
      "Compiling slice 15\n",
      "\n",
      "Compiling round 2\n",
      "\n",
      "Compiling slice 0\n",
      "\n",
      "Compiling slice 1\n",
      "\n",
      "Compiling slice 2\n",
      "\n",
      "Compiling slice 3\n",
      "\n",
      "Compiling slice 4\n",
      "\n",
      "Compiling slice 5\n",
      "\n",
      "Compiling slice 6\n",
      "\n",
      "Compiling slice 7\n",
      "\n",
      "Compiling slice 8\n",
      "\n",
      "Compiling slice 9\n",
      "\n",
      "Compiling slice 10\n",
      "\n",
      "Compiling slice 11\n",
      "\n",
      "Compiling slice 12\n",
      "\n",
      "Compiling slice 13\n",
      "\n",
      "Compiling slice 14\n",
      "\n",
      "Compiling slice 15\n",
      "\n",
      "Compiling round 3\n",
      "\n",
      "Compiling slice 0\n",
      "\n",
      "Compiling slice 1\n",
      "\n",
      "Compiling slice 2\n",
      "\n",
      "Compiling slice 3\n",
      "\n",
      "Compiling slice 4\n",
      "\n",
      "Compiling slice 5\n",
      "\n",
      "Compiling slice 6\n",
      "\n",
      "Compiling slice 7\n",
      "\n",
      "Compiling slice 8\n",
      "\n",
      "Compiling slice 9\n",
      "\n",
      "Compiling slice 10\n",
      "\n",
      "Compiling slice 11\n",
      "\n",
      "Compiling slice 12\n",
      "\n",
      "Compiling slice 13\n",
      "\n",
      "Compiling slice 14\n",
      "\n",
      "Compiling slice 15\n",
      "\n",
      "Barcodind slices 0\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 1\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 2\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 3\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 4\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 5\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 6\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 7\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 8\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 9\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 10\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 11\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 12\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 13\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 14\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 15\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "/home/dcastillo/scratch/workspace/oligoFISSEQ/Notebooks/test/input/merged/OFQv61_w3_3_decon/roi_5.tif\n",
      "Using existing roi file /home/dcastillo/scratch/workspace/oligoFISSEQ/Notebooks/test/input/merged/OFQv61_w3_3_decon/roi_5_roi.zip\n",
      "Compiling round 0\n",
      "\n",
      "Compiling slice 0\n",
      "\n",
      "Compiling slice 1\n",
      "\n",
      "Compiling slice 2\n",
      "\n",
      "Compiling slice 3\n",
      "\n",
      "Compiling slice 4\n",
      "\n",
      "Compiling slice 5\n",
      "\n",
      "Compiling slice 6\n",
      "\n",
      "Compiling slice 7\n",
      "\n",
      "Compiling slice 8\n",
      "\n",
      "Compiling slice 9\n",
      "\n",
      "Compiling slice 10\n",
      "\n",
      "Compiling slice 11\n",
      "\n",
      "Compiling slice 12\n",
      "\n",
      "Compiling slice 13\n",
      "\n",
      "Compiling slice 14\n",
      "\n",
      "Compiling slice 15\n",
      "\n",
      "Compiling round 1\n",
      "\n",
      "Compiling slice 0\n",
      "\n",
      "Compiling slice 1\n",
      "\n",
      "Compiling slice 2\n",
      "\n",
      "Compiling slice 3\n",
      "\n",
      "Compiling slice 4\n",
      "\n",
      "Compiling slice 5\n",
      "\n",
      "Compiling slice 6\n",
      "\n",
      "Compiling slice 7\n",
      "\n",
      "Compiling slice 8\n",
      "\n",
      "Compiling slice 9\n",
      "\n",
      "Compiling slice 10\n",
      "\n",
      "Compiling slice 11\n",
      "\n",
      "Compiling slice 12\n",
      "\n",
      "Compiling slice 13\n",
      "\n",
      "Compiling slice 14\n",
      "\n",
      "Compiling slice 15\n",
      "\n",
      "Compiling round 2\n",
      "\n",
      "Compiling slice 0\n",
      "\n",
      "Compiling slice 1\n",
      "\n",
      "Compiling slice 2\n",
      "\n",
      "Compiling slice 3\n",
      "\n",
      "Compiling slice 4\n",
      "\n",
      "Compiling slice 5\n",
      "\n",
      "Compiling slice 6\n",
      "\n",
      "Compiling slice 7\n",
      "\n",
      "Compiling slice 8\n",
      "\n",
      "Compiling slice 9\n",
      "\n",
      "Compiling slice 10\n",
      "\n",
      "Compiling slice 11\n",
      "\n",
      "Compiling slice 12\n",
      "\n",
      "Compiling slice 13\n",
      "\n",
      "Compiling slice 14\n",
      "\n",
      "Compiling slice 15\n",
      "\n",
      "Compiling round 3\n",
      "\n",
      "Compiling slice 0\n",
      "\n",
      "Compiling slice 1\n",
      "\n",
      "Compiling slice 2\n",
      "\n",
      "Compiling slice 3\n",
      "\n",
      "Compiling slice 4\n",
      "\n",
      "Compiling slice 5\n",
      "\n",
      "Compiling slice 6\n",
      "\n",
      "Compiling slice 7\n",
      "\n",
      "Compiling slice 8\n",
      "\n",
      "Compiling slice 9\n",
      "\n",
      "Compiling slice 10\n",
      "\n",
      "Compiling slice 11\n",
      "\n",
      "Compiling slice 12\n",
      "\n",
      "Compiling slice 13\n",
      "\n",
      "Compiling slice 14\n",
      "\n",
      "Compiling slice 15\n",
      "\n",
      "Barcodind slices 0\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 1\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 2\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 3\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 4\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 5\n",
      "\n",
      "Barcoding roi 0\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Barcodind slices 6\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 7\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 8\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 9\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 10\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 11\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 12\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 13\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 14\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 15\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "/home/dcastillo/scratch/workspace/oligoFISSEQ/Notebooks/test/input/merged/OFQv61_w3_3_decon/roi_7.tif\n",
      "Using existing roi file /home/dcastillo/scratch/workspace/oligoFISSEQ/Notebooks/test/input/merged/OFQv61_w3_3_decon/roi_7_roi.zip\n",
      "Compiling round 0\n",
      "\n",
      "Compiling slice 0\n",
      "\n",
      "Compiling slice 1\n",
      "\n",
      "Compiling slice 2\n",
      "\n",
      "Compiling slice 3\n",
      "\n",
      "Compiling slice 4\n",
      "\n",
      "Compiling slice 5\n",
      "\n",
      "Compiling slice 6\n",
      "\n",
      "Compiling slice 7\n",
      "\n",
      "Compiling slice 8\n",
      "\n",
      "Compiling slice 9\n",
      "\n",
      "Compiling slice 10\n",
      "\n",
      "Compiling slice 11\n",
      "\n",
      "Compiling slice 12\n",
      "\n",
      "Compiling slice 13\n",
      "\n",
      "Compiling slice 14\n",
      "\n",
      "Compiling slice 15\n",
      "\n",
      "Compiling round 1\n",
      "\n",
      "Compiling slice 0\n",
      "\n",
      "Compiling slice 1\n",
      "\n",
      "Compiling slice 2\n",
      "\n",
      "Compiling slice 3\n",
      "\n",
      "Compiling slice 4\n",
      "\n",
      "Compiling slice 5\n",
      "\n",
      "Compiling slice 6\n",
      "\n",
      "Compiling slice 7\n",
      "\n",
      "Compiling slice 8\n",
      "\n",
      "Compiling slice 9\n",
      "\n",
      "Compiling slice 10\n",
      "\n",
      "Compiling slice 11\n",
      "\n",
      "Compiling slice 12\n",
      "\n",
      "Compiling slice 13\n",
      "\n",
      "Compiling slice 14\n",
      "\n",
      "Compiling slice 15\n",
      "\n",
      "Compiling round 2\n",
      "\n",
      "Compiling slice 0\n",
      "\n",
      "Compiling slice 1\n",
      "\n",
      "Compiling slice 2\n",
      "\n",
      "Compiling slice 3\n",
      "\n",
      "Compiling slice 4\n",
      "\n",
      "Compiling slice 5\n",
      "\n",
      "Compiling slice 6\n",
      "\n",
      "Compiling slice 7\n",
      "\n",
      "Compiling slice 8\n",
      "\n",
      "Compiling slice 9\n",
      "\n",
      "Compiling slice 10\n",
      "\n",
      "Compiling slice 11\n",
      "\n",
      "Compiling slice 12\n",
      "\n",
      "Compiling slice 13\n",
      "\n",
      "Compiling slice 14\n",
      "\n",
      "Compiling slice 15\n",
      "\n",
      "Compiling round 3\n",
      "\n",
      "Compiling slice 0\n",
      "\n",
      "Compiling slice 1\n",
      "\n",
      "Compiling slice 2\n",
      "\n",
      "Compiling slice 3\n",
      "\n",
      "Compiling slice 4\n",
      "\n",
      "Compiling slice 5\n",
      "\n",
      "Compiling slice 6\n",
      "\n",
      "Compiling slice 7\n",
      "\n",
      "Compiling slice 8\n",
      "\n",
      "Compiling slice 9\n",
      "\n",
      "Compiling slice 10\n",
      "\n",
      "Compiling slice 11\n",
      "\n",
      "Compiling slice 12\n",
      "\n",
      "Compiling slice 13\n",
      "\n",
      "Compiling slice 14\n",
      "\n",
      "Compiling slice 15\n",
      "\n",
      "Barcodind slices 0\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 1\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 2\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 3\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 4\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 5\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 6\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 7\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 8\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 9\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 10\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 11\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 12\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 13\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 14\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 15\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "/home/dcastillo/scratch/workspace/oligoFISSEQ/Notebooks/test/input/merged/OFQv61_w3_3_decon/roi_3.tif\n",
      "Using existing roi file /home/dcastillo/scratch/workspace/oligoFISSEQ/Notebooks/test/input/merged/OFQv61_w3_3_decon/roi_3_roi.zip\n",
      "Compiling round 0\n",
      "\n",
      "Compiling slice 0\n",
      "\n",
      "Compiling slice 1\n",
      "\n",
      "Compiling slice 2\n",
      "\n",
      "Compiling slice 3\n",
      "\n",
      "Compiling slice 4\n",
      "\n",
      "Compiling slice 5\n",
      "\n",
      "Compiling slice 6\n",
      "\n",
      "Compiling slice 7\n",
      "\n",
      "Compiling slice 8\n",
      "\n",
      "Compiling slice 9\n",
      "\n",
      "Compiling slice 10\n",
      "\n",
      "Compiling slice 11\n",
      "\n",
      "Compiling slice 12\n",
      "\n",
      "Compiling slice 13\n",
      "\n",
      "Compiling slice 14\n",
      "\n",
      "Compiling slice 15\n",
      "\n",
      "Compiling round 1\n",
      "\n",
      "Compiling slice 0\n",
      "\n",
      "Compiling slice 1\n",
      "\n",
      "Compiling slice 2\n",
      "\n",
      "Compiling slice 3\n",
      "\n",
      "Compiling slice 4\n",
      "\n",
      "Compiling slice 5\n",
      "\n",
      "Compiling slice 6\n",
      "\n",
      "Compiling slice 7\n",
      "\n",
      "Compiling slice 8\n",
      "\n",
      "Compiling slice 9\n",
      "\n",
      "Compiling slice 10\n",
      "\n",
      "Compiling slice 11\n",
      "\n",
      "Compiling slice 12\n",
      "\n",
      "Compiling slice 13\n",
      "\n",
      "Compiling slice 14\n",
      "\n",
      "Compiling slice 15\n",
      "\n",
      "Compiling round 2\n",
      "\n",
      "Compiling slice 0\n",
      "\n",
      "Compiling slice 1\n",
      "\n",
      "Compiling slice 2\n",
      "\n",
      "Compiling slice 3\n",
      "\n",
      "Compiling slice 4\n",
      "\n",
      "Compiling slice 5\n",
      "\n",
      "Compiling slice 6\n",
      "\n",
      "Compiling slice 7\n",
      "\n",
      "Compiling slice 8\n",
      "\n",
      "Compiling slice 9\n",
      "\n",
      "Compiling slice 10\n",
      "\n",
      "Compiling slice 11\n",
      "\n",
      "Compiling slice 12\n",
      "\n",
      "Compiling slice 13\n",
      "\n",
      "Compiling slice 14\n",
      "\n",
      "Compiling slice 15\n",
      "\n",
      "Compiling round 3\n",
      "\n",
      "Compiling slice 0\n",
      "\n",
      "Compiling slice 1\n",
      "\n",
      "Compiling slice 2\n",
      "\n",
      "Compiling slice 3\n",
      "\n",
      "Compiling slice 4\n",
      "\n",
      "Compiling slice 5\n",
      "\n",
      "Compiling slice 6\n",
      "\n",
      "Compiling slice 7\n",
      "\n",
      "Compiling slice 8\n",
      "\n",
      "Compiling slice 9\n",
      "\n",
      "Compiling slice 10\n",
      "\n",
      "Compiling slice 11\n",
      "\n",
      "Compiling slice 12\n",
      "\n",
      "Compiling slice 13\n",
      "\n",
      "Compiling slice 14\n",
      "\n",
      "Compiling slice 15\n",
      "\n",
      "Barcodind slices 0\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 1\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 2\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 3\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 4\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 5\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 6\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 7\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 8\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 9\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 10\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 11\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 12\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 13\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 14\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 15\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "/home/dcastillo/scratch/workspace/oligoFISSEQ/Notebooks/test/input/merged/OFQv61_w3_3_decon/roi_2.tif\n",
      "Using existing roi file /home/dcastillo/scratch/workspace/oligoFISSEQ/Notebooks/test/input/merged/OFQv61_w3_3_decon/roi_2_roi.zip\n",
      "Compiling round 0\n",
      "\n",
      "Compiling slice 0\n",
      "\n",
      "Compiling slice 1\n",
      "\n",
      "Compiling slice 2\n",
      "\n",
      "Compiling slice 3\n",
      "\n",
      "Compiling slice 4\n",
      "\n",
      "Compiling slice 5\n",
      "\n",
      "Compiling slice 6\n",
      "\n",
      "Compiling slice 7\n",
      "\n",
      "Compiling slice 8\n",
      "\n",
      "Compiling slice 9\n",
      "\n",
      "Compiling slice 10\n",
      "\n",
      "Compiling slice 11\n",
      "\n",
      "Compiling slice 12\n",
      "\n",
      "Compiling slice 13\n",
      "\n",
      "Compiling slice 14\n",
      "\n",
      "Compiling slice 15\n",
      "\n",
      "Compiling round 1\n",
      "\n",
      "Compiling slice 0\n",
      "\n",
      "Compiling slice 1\n",
      "\n",
      "Compiling slice 2\n",
      "\n",
      "Compiling slice 3\n",
      "\n",
      "Compiling slice 4\n",
      "\n",
      "Compiling slice 5\n",
      "\n",
      "Compiling slice 6\n",
      "\n",
      "Compiling slice 7\n",
      "\n",
      "Compiling slice 8\n",
      "\n",
      "Compiling slice 9\n",
      "\n",
      "Compiling slice 10\n",
      "\n",
      "Compiling slice 11\n",
      "\n",
      "Compiling slice 12\n",
      "\n",
      "Compiling slice 13\n",
      "\n",
      "Compiling slice 14\n",
      "\n",
      "Compiling slice 15\n",
      "\n",
      "Compiling round 2\n",
      "\n",
      "Compiling slice 0\n",
      "\n",
      "Compiling slice 1\n",
      "\n",
      "Compiling slice 2\n",
      "\n",
      "Compiling slice 3\n",
      "\n",
      "Compiling slice 4\n",
      "\n",
      "Compiling slice 5\n",
      "\n",
      "Compiling slice 6\n",
      "\n",
      "Compiling slice 7\n",
      "\n",
      "Compiling slice 8\n",
      "\n",
      "Compiling slice 9\n",
      "\n",
      "Compiling slice 10\n",
      "\n",
      "Compiling slice 11\n",
      "\n",
      "Compiling slice 12\n",
      "\n",
      "Compiling slice 13\n",
      "\n",
      "Compiling slice 14\n",
      "\n",
      "Compiling slice 15\n",
      "\n",
      "Compiling round 3\n",
      "\n",
      "Compiling slice 0\n",
      "\n",
      "Compiling slice 1\n",
      "\n",
      "Compiling slice 2\n",
      "\n",
      "Compiling slice 3\n",
      "\n",
      "Compiling slice 4\n",
      "\n",
      "Compiling slice 5\n",
      "\n",
      "Compiling slice 6\n",
      "\n",
      "Compiling slice 7\n",
      "\n",
      "Compiling slice 8\n",
      "\n",
      "Compiling slice 9\n",
      "\n",
      "Compiling slice 10\n",
      "\n",
      "Compiling slice 11\n",
      "\n",
      "Compiling slice 12\n",
      "\n",
      "Compiling slice 13\n",
      "\n",
      "Compiling slice 14\n",
      "\n",
      "Compiling slice 15\n",
      "\n",
      "Barcodind slices 0\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 1\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 2\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 3\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 4\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 5\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 6\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 7\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 8\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 9\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 10\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 11\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 12\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 13\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 14\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 15\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "/home/dcastillo/scratch/workspace/oligoFISSEQ/Notebooks/test/input/merged/OFQv61_w3_3_decon/roi_10.tif\n",
      "Using existing roi file /home/dcastillo/scratch/workspace/oligoFISSEQ/Notebooks/test/input/merged/OFQv61_w3_3_decon/roi_10_roi.zip\n",
      "Compiling round 0\n",
      "\n",
      "Compiling slice 0\n",
      "\n",
      "Compiling slice 1\n",
      "\n",
      "Compiling slice 2\n",
      "\n",
      "Compiling slice 3\n",
      "\n",
      "Compiling slice 4\n",
      "\n",
      "Compiling slice 5\n",
      "\n",
      "Compiling slice 6\n",
      "\n",
      "Compiling slice 7\n",
      "\n",
      "Compiling slice 8\n",
      "\n",
      "Compiling slice 9\n",
      "\n",
      "Compiling slice 10\n",
      "\n",
      "Compiling slice 11\n",
      "\n",
      "Compiling slice 12\n",
      "\n",
      "Compiling slice 13\n",
      "\n",
      "Compiling slice 14\n",
      "\n",
      "Compiling slice 15\n",
      "\n",
      "Compiling round 1\n",
      "\n",
      "Compiling slice 0\n",
      "\n",
      "Compiling slice 1\n",
      "\n",
      "Compiling slice 2\n",
      "\n",
      "Compiling slice 3\n",
      "\n",
      "Compiling slice 4\n",
      "\n",
      "Compiling slice 5\n",
      "\n",
      "Compiling slice 6\n",
      "\n",
      "Compiling slice 7\n",
      "\n",
      "Compiling slice 8\n",
      "\n",
      "Compiling slice 9\n",
      "\n",
      "Compiling slice 10\n",
      "\n",
      "Compiling slice 11\n",
      "\n",
      "Compiling slice 12\n",
      "\n",
      "Compiling slice 13\n",
      "\n",
      "Compiling slice 14\n",
      "\n",
      "Compiling slice 15\n",
      "\n",
      "Compiling round 2\n",
      "\n",
      "Compiling slice 0\n",
      "\n",
      "Compiling slice 1\n",
      "\n",
      "Compiling slice 2\n",
      "\n",
      "Compiling slice 3\n",
      "\n",
      "Compiling slice 4\n",
      "\n",
      "Compiling slice 5\n",
      "\n",
      "Compiling slice 6\n",
      "\n",
      "Compiling slice 7\n",
      "\n",
      "Compiling slice 8\n",
      "\n",
      "Compiling slice 9\n",
      "\n",
      "Compiling slice 10\n",
      "\n",
      "Compiling slice 11\n",
      "\n",
      "Compiling slice 12\n",
      "\n",
      "Compiling slice 13\n",
      "\n",
      "Compiling slice 14\n",
      "\n",
      "Compiling slice 15\n",
      "\n",
      "Compiling round 3\n",
      "\n",
      "Compiling slice 0\n",
      "\n",
      "Compiling slice 1\n",
      "\n",
      "Compiling slice 2\n",
      "\n",
      "Compiling slice 3\n",
      "\n",
      "Compiling slice 4\n",
      "\n",
      "Compiling slice 5\n",
      "\n",
      "Compiling slice 6\n",
      "\n",
      "Compiling slice 7\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Compiling slice 8\n",
      "\n",
      "Compiling slice 9\n",
      "\n",
      "Compiling slice 10\n",
      "\n",
      "Compiling slice 11\n",
      "\n",
      "Compiling slice 12\n",
      "\n",
      "Compiling slice 13\n",
      "\n",
      "Compiling slice 14\n",
      "\n",
      "Compiling slice 15\n",
      "\n",
      "Barcodind slices 0\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 1\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 2\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 3\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 4\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 5\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 6\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 7\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 8\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 9\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 10\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 11\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 12\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 13\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 14\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 15\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "/home/dcastillo/scratch/workspace/oligoFISSEQ/Notebooks/test/input/merged/OFQv61_w3_3_decon/roi_4.tif\n",
      "Using existing roi file /home/dcastillo/scratch/workspace/oligoFISSEQ/Notebooks/test/input/merged/OFQv61_w3_3_decon/roi_4_roi.zip\n",
      "Compiling round 0\n",
      "\n",
      "Compiling slice 0\n",
      "\n",
      "Compiling slice 1\n",
      "\n",
      "Compiling slice 2\n",
      "\n",
      "Compiling slice 3\n",
      "\n",
      "Compiling slice 4\n",
      "\n",
      "Compiling slice 5\n",
      "\n",
      "Compiling slice 6\n",
      "\n",
      "Compiling slice 7\n",
      "\n",
      "Compiling slice 8\n",
      "\n",
      "Compiling slice 9\n",
      "\n",
      "Compiling slice 10\n",
      "\n",
      "Compiling slice 11\n",
      "\n",
      "Compiling slice 12\n",
      "\n",
      "Compiling slice 13\n",
      "\n",
      "Compiling slice 14\n",
      "\n",
      "Compiling slice 15\n",
      "\n",
      "Compiling round 1\n",
      "\n",
      "Compiling slice 0\n",
      "\n",
      "Compiling slice 1\n",
      "\n",
      "Compiling slice 2\n",
      "\n",
      "Compiling slice 3\n",
      "\n",
      "Compiling slice 4\n",
      "\n",
      "Compiling slice 5\n",
      "\n",
      "Compiling slice 6\n",
      "\n",
      "Compiling slice 7\n",
      "\n",
      "Compiling slice 8\n",
      "\n",
      "Compiling slice 9\n",
      "\n",
      "Compiling slice 10\n",
      "\n",
      "Compiling slice 11\n",
      "\n",
      "Compiling slice 12\n",
      "\n",
      "Compiling slice 13\n",
      "\n",
      "Compiling slice 14\n",
      "\n",
      "Compiling slice 15\n",
      "\n",
      "Compiling round 2\n",
      "\n",
      "Compiling slice 0\n",
      "\n",
      "Compiling slice 1\n",
      "\n",
      "Compiling slice 2\n",
      "\n",
      "Compiling slice 3\n",
      "\n",
      "Compiling slice 4\n",
      "\n",
      "Compiling slice 5\n",
      "\n",
      "Compiling slice 6\n",
      "\n",
      "Compiling slice 7\n",
      "\n",
      "Compiling slice 8\n",
      "\n",
      "Compiling slice 9\n",
      "\n",
      "Compiling slice 10\n",
      "\n",
      "Compiling slice 11\n",
      "\n",
      "Compiling slice 12\n",
      "\n",
      "Compiling slice 13\n",
      "\n",
      "Compiling slice 14\n",
      "\n",
      "Compiling slice 15\n",
      "\n",
      "Compiling round 3\n",
      "\n",
      "Compiling slice 0\n",
      "\n",
      "Compiling slice 1\n",
      "\n",
      "Compiling slice 2\n",
      "\n",
      "Compiling slice 3\n",
      "\n",
      "Compiling slice 4\n",
      "\n",
      "Compiling slice 5\n",
      "\n",
      "Compiling slice 6\n",
      "\n",
      "Compiling slice 7\n",
      "\n",
      "Compiling slice 8\n",
      "\n",
      "Compiling slice 9\n",
      "\n",
      "Compiling slice 10\n",
      "\n",
      "Compiling slice 11\n",
      "\n",
      "Compiling slice 12\n",
      "\n",
      "Compiling slice 13\n",
      "\n",
      "Compiling slice 14\n",
      "\n",
      "Compiling slice 15\n",
      "\n",
      "Barcodind slices 0\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 1\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 2\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 3\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 4\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 5\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 6\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 7\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 8\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 9\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 10\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 11\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 12\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 13\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 14\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 15\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "/home/dcastillo/scratch/workspace/oligoFISSEQ/Notebooks/test/input/merged/OFQv61_w3_3_decon/roi_9.tif\n",
      "Using existing roi file /home/dcastillo/scratch/workspace/oligoFISSEQ/Notebooks/test/input/merged/OFQv61_w3_3_decon/roi_9_roi.zip\n",
      "Compiling round 0\n",
      "\n",
      "Compiling slice 0\n",
      "\n",
      "Compiling slice 1\n",
      "\n",
      "Compiling slice 2\n",
      "\n",
      "Compiling slice 3\n",
      "\n",
      "Compiling slice 4\n",
      "\n",
      "Compiling slice 5\n",
      "\n",
      "Compiling slice 6\n",
      "\n",
      "Compiling slice 7\n",
      "\n",
      "Compiling slice 8\n",
      "\n",
      "Compiling slice 9\n",
      "\n",
      "Compiling slice 10\n",
      "\n",
      "Compiling slice 11\n",
      "\n",
      "Compiling slice 12\n",
      "\n",
      "Compiling slice 13\n",
      "\n",
      "Compiling slice 14\n",
      "\n",
      "Compiling slice 15\n",
      "\n",
      "Compiling round 1\n",
      "\n",
      "Compiling slice 0\n",
      "\n",
      "Compiling slice 1\n",
      "\n",
      "Compiling slice 2\n",
      "\n",
      "Compiling slice 3\n",
      "\n",
      "Compiling slice 4\n",
      "\n",
      "Compiling slice 5\n",
      "\n",
      "Compiling slice 6\n",
      "\n",
      "Compiling slice 7\n",
      "\n",
      "Compiling slice 8\n",
      "\n",
      "Compiling slice 9\n",
      "\n",
      "Compiling slice 10\n",
      "\n",
      "Compiling slice 11\n",
      "\n",
      "Compiling slice 12\n",
      "\n",
      "Compiling slice 13\n",
      "\n",
      "Compiling slice 14\n",
      "\n",
      "Compiling slice 15\n",
      "\n",
      "Compiling round 2\n",
      "\n",
      "Compiling slice 0\n",
      "\n",
      "Compiling slice 1\n",
      "\n",
      "Compiling slice 2\n",
      "\n",
      "Compiling slice 3\n",
      "\n",
      "Compiling slice 4\n",
      "\n",
      "Compiling slice 5\n",
      "\n",
      "Compiling slice 6\n",
      "\n",
      "Compiling slice 7\n",
      "\n",
      "Compiling slice 8\n",
      "\n",
      "Compiling slice 9\n",
      "\n",
      "Compiling slice 10\n",
      "\n",
      "Compiling slice 11\n",
      "\n",
      "Compiling slice 12\n",
      "\n",
      "Compiling slice 13\n",
      "\n",
      "Compiling slice 14\n",
      "\n",
      "Compiling slice 15\n",
      "\n",
      "Compiling round 3\n",
      "\n",
      "Compiling slice 0\n",
      "\n",
      "Compiling slice 1\n",
      "\n",
      "Compiling slice 2\n",
      "\n",
      "Compiling slice 3\n",
      "\n",
      "Compiling slice 4\n",
      "\n",
      "Compiling slice 5\n",
      "\n",
      "Compiling slice 6\n",
      "\n",
      "Compiling slice 7\n",
      "\n",
      "Compiling slice 8\n",
      "\n",
      "Compiling slice 9\n",
      "\n",
      "Compiling slice 10\n",
      "\n",
      "Compiling slice 11\n",
      "\n",
      "Compiling slice 12\n",
      "\n",
      "Compiling slice 13\n",
      "\n",
      "Compiling slice 14\n",
      "\n",
      "Compiling slice 15\n",
      "\n",
      "Barcodind slices 0\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 1\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 2\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 3\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 4\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 5\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 6\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 7\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 8\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 9\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 10\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 11\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 12\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 13\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 14\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Barcodind slices 15\n",
      "\n",
      "Barcoding roi 0\n",
      "\n",
      "Finished!\n"
     ]
    }
   ],
   "source": [
    "detect_barcodes_script = new File(\"scripts/detect_barcode_patches.py\");\n",
    "inputs = [\"input_folder\": merged_path,\n",
    "          \"barcode_file\": barcode_file,\n",
    "          \"nbr_rounds\"  : 4]\n",
    "result = ij.script().run(detect_barcodes_script, true, inputs);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We should wait for a \"Finished!\" message before continuing.\n",
    "\n",
    "The merged/ folder should have been populated with the following files per nuclei:\n",
    "\n",
    "* roi_{nucleus_nbr}_barcoded.tif          : image of the nucleus where every matched barcode is colored\n",
    "* roi_{nucleus_nbr}_barcodes_splitted.tif : image where every matched barcode is shown in a separate channel\n",
    "* roi_{nucleus_nbr}_max_intensity.tif     : maximum intensity projection over all channels over all rounds\n",
    "* roi_{nucleus_nbr}_punctas.tsv           : tsv file with the information of the matched barcodes for the analysis \n",
    "\n",
    "Let's have a look at the barcoded image"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Figure size 720x288 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x288 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Lossy conversion from uint16 to uint8. Losing 8 bits of resolution. Convert image to uint8 prior to saving to suppress this warning.\n"
     ]
    }
   ],
   "source": [
    "%%python\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "from skimage.external import tifffile\n",
    "from skimage.io import imsave\n",
    "from skimage.color import gray2rgb\n",
    "from skimage import exposure\n",
    "from mpl_toolkits.axes_grid1 import make_axes_locatable\n",
    "\n",
    "roi_barcode_image_file = \"input/merged/OFQv61_w3_3_decon/roi_2_barcoded.tif\"\n",
    "roi_max_intensity_image_file = \"input/merged/OFQv61_w3_3_decon/roi_2_max_intensity.tif\"\n",
    "\n",
    "roi_barcode_slice = tifffile.imread(roi_barcode_image_file)\n",
    "roi_max_intensity_slice = tifffile.imread(roi_max_intensity_image_file)\n",
    "\n",
    "roi_max_intensity_proj = np.max(roi_max_intensity_slice, axis=0)\n",
    "\n",
    "plt.figure(figsize=(10, 4))\n",
    "fig, (ax1, ax2) = plt.subplots(1, 2, sharex=True, sharey=True, figsize=(10,4))\n",
    "\n",
    "tifffile.imshow(roi_barcode_slice[11],figure=fig,subplot=121,aspect='auto')\n",
    "f2,s2,a2=tifffile.imshow(roi_max_intensity_proj,figure=fig,cmap=plt.cm.gray,subplot=122,\n",
    "               vmin=26000, vmax=None, aspect='auto')\n",
    "\n",
    "\n",
    "\n",
    "plt.show()\n",
    "roi_max_intensity_slice_rgb = exposure.rescale_intensity(roi_max_intensity_proj,\n",
    "                                                           in_range=(26000, roi_max_intensity_slice.max()))\n",
    "imsave('input/roi_max_intensity.png',gray2rgb(roi_max_intensity_slice_rgb))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In the roi_{nucleus_nbr}_punctas.tsv we find information for the analysis of the positions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "   barcode  roi  max_intensity_pixel_x  max_intensity_pixel_y  \\\n",
      "0    21314    0               9.334501               7.200901   \n",
      "1    21314    0              10.668001               5.600700   \n",
      "2    21314    0               9.067801               7.734301   \n",
      "3    21314    0               7.467601               7.734301   \n",
      "4    21314    0              14.401801              10.134601   \n",
      "\n",
      "   max_intensity_pixel_z  max_intensity  nbr_pixels  match_barcodes  \\\n",
      "0                    0.0       0.679586           1               4   \n",
      "1                    0.0       0.651315           1               4   \n",
      "2                    0.5       0.741655           4               4   \n",
      "3                    1.0       0.733699           6               4   \n",
      "4                    0.5       0.583588           1               4   \n",
      "\n",
      "   center_pixel_x  center_pixel_y  center_pixel_z  max_intensity_0  \\\n",
      "0        9.334501        7.200901        0.000000         0.668890   \n",
      "1       10.668001        5.600700        0.000000         0.624359   \n",
      "2        9.134476        7.667626        0.500000         0.732429   \n",
      "3        7.734301        7.512051        0.666667         0.710903   \n",
      "4       14.401801       10.134601        0.500000         0.643075   \n",
      "\n",
      "   second_intensity_0  max_intensity_1  second_intensity_1  max_intensity_2  \\\n",
      "0            0.643845         0.675176            0.654761         0.701073   \n",
      "1            0.613720         0.691651            0.689833         0.652607   \n",
      "2            0.684066         0.761673            0.756509         0.740581   \n",
      "3            0.687295         0.759841            0.722705         0.759198   \n",
      "4            0.612103         0.595841            0.567424         0.519388   \n",
      "\n",
      "   second_intensity_2  max_intensity_3  second_intensity_3  \n",
      "0            0.686126         0.673206            0.652107  \n",
      "1            0.629533         0.636643            0.628491  \n",
      "2            0.726123         0.731936            0.703046  \n",
      "3            0.754594         0.704853            0.695542  \n",
      "4            0.519002         0.576050            0.519322  \n"
     ]
    }
   ],
   "source": [
    "%%python\n",
    "roi_puncta_file = \"input/merged/OFQv61_w3_3_decon/roi_2_punctas.tsv\"\n",
    "roi_punctas = pd.read_csv(roi_puncta_file, sep='\\t')\n",
    "print(roi_punctas.head())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Tier 1 detection\n",
    "\n",
    "In the tsv file we have all the possible patches in the images. We can apply our filters to the data to obtain the strong positives"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "%%python\n",
    "thresholds = {('SOLiD',4,'MSBS',False):(64000./(2**16-1),63500./(2**16-1),0)}\n",
    "\n",
    "roi_punctas['avg_intensity3'] = roi_punctas[['max_intensity_0','max_intensity_1','max_intensity_2']].mean(axis=1)\n",
    "use_threshold = ('SOLiD',4,'MSBS',False)\n",
    "\n",
    "roi_punctas = roi_punctas.drop(roi_punctas[(roi_punctas['max_intensity_0']<thresholds[use_threshold][0])\n",
    "                                             & (roi_punctas['max_intensity_1']<thresholds[use_threshold][0])\n",
    "                                             & (roi_punctas['max_intensity_2']<thresholds[use_threshold][0])\n",
    "                                             & (roi_punctas['max_intensity_3']<thresholds[use_threshold][0])].index)\n",
    "    \n",
    "roi_punctas = roi_punctas.drop(roi_punctas[(roi_punctas['avg_intensity3']<thresholds[use_threshold][1])].index)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can assign our barcodes to the targeted loci"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "     barcode  roi  max_intensity_pixel_x  max_intensity_pixel_y  \\\n",
      "11     21314    0              12.534901               8.801101   \n",
      "32     21314    0              15.201901               4.000500   \n",
      "143    13314    0              10.401301              10.134601   \n",
      "152    13314    0              14.135101               8.267701   \n",
      "171    32133    0              14.401801               6.134101   \n",
      "\n",
      "     max_intensity_pixel_z  max_intensity  nbr_pixels  match_barcodes  \\\n",
      "11                     3.0       0.999332         110               4   \n",
      "32                     4.5       0.976947         132               4   \n",
      "143                    6.0       0.989582          81               4   \n",
      "152                    6.0       0.990281          49               4   \n",
      "171                    3.5       0.995818         138               4   \n",
      "\n",
      "     center_pixel_x  center_pixel_y  ...  max_intensity_0  second_intensity_0  \\\n",
      "11        12.440344        8.885960  ...         0.999618            0.932588   \n",
      "32        15.264535        3.673187  ...         0.994596            0.442778   \n",
      "143       10.384838        9.792171  ...         0.988706            0.892163   \n",
      "152       14.254844        8.256815  ...         0.992063            0.824053   \n",
      "171       14.649175        5.940840  ...         0.994140            0.854731   \n",
      "\n",
      "     max_intensity_1  second_intensity_1  max_intensity_2  second_intensity_2  \\\n",
      "11          0.998198            0.914263         0.999664            0.958718   \n",
      "32          0.955538            0.673229         0.991653            0.713950   \n",
      "143         0.986110            0.933568         0.996826            0.939328   \n",
      "152         0.996428            0.861720         0.985885            0.815690   \n",
      "171         0.997161            0.904511         0.992018            0.898048   \n",
      "\n",
      "     max_intensity_3  second_intensity_3  avg_intensity3  loci  \n",
      "11          0.999847            0.920504        0.999160  3pR3  \n",
      "32          0.966002            0.707450        0.980595  3pR3  \n",
      "143         0.986688            0.944775        0.990547  5qR1  \n",
      "152         0.986749            0.926748        0.991459  5qR1  \n",
      "171         0.999954            0.905470        0.994439  3qR3  \n",
      "\n",
      "[5 rows x 21 columns]\n"
     ]
    }
   ],
   "source": [
    "%%python\n",
    "barcodes = [11243,43131,31341,13222,42144,32311,12321,42212,21314,14232,43232,32133,11313,43341,32143,13314,41143,34324,\n",
    "            32442,41234,22433,33332,12112,31224,42314,14323,22344,42413,32232,12434,41323,33414,13133,24442,44142,31444]\n",
    "locis = ['2pR1', '2pR2', '2pR3', '2qR1', '2qR2', '2qR3', '3pR1', '3pR2', '3pR3', '3qR1', '3qR2', '3qR3', '5pR1', '5pR2',\n",
    "         '5pR3', '5qR1', '5qR2', '5qR3', '16pR1', '16pR2', '16pR3', '16qR1', '16qR2', '16qR3',\n",
    "         '19pR1', '19pR2', '19pR3', '19qR1', '19qR2', '19qR3', 'XpR1', 'XpR2', 'XpR3', 'XqR1', 'XqR2', 'XqR3']\n",
    "chrs = list(set([(loci.split('p')[0]).split('q')[0] for loci in locis]))\n",
    "\n",
    "barcode_loci = dict(zip(barcodes,locis))\n",
    "roi_punctas['loci'] = roi_punctas.barcode.map(barcode_loci)\n",
    "print(roi_punctas.head())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "And check the position of the detected loci in the images"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x7f86730c1590>"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "%%python\n",
    "import matplotlib.pyplot as plt\n",
    "from skimage.io import imsave\n",
    "import numpy as np\n",
    "from PIL import Image, ImageDraw, ImageFont\n",
    "\n",
    "colors_chrs = {'2':[0,255,0],'3':[255,255,0],'5':[0,0,255],\n",
    "               '16':[128,0,128],'19':[255,0,0],'X':[128,128,128]}\n",
    "pixelWidth = 0.265\n",
    "im = Image.open('input/roi_max_intensity.png')\n",
    "\n",
    "draw = ImageDraw.Draw(im)\n",
    "for index,row in roi_punctas.iterrows():\n",
    "    chrom = (row.loci.split('p')[0]).split('q')[0]\n",
    "    draw.ellipse((row.max_intensity_pixel_x/pixelWidth, row.max_intensity_pixel_y/pixelWidth,\n",
    "                  row.max_intensity_pixel_x/pixelWidth+1, row.max_intensity_pixel_y/pixelWidth+1),\n",
    "                 fill = tuple(colors_chrs[chrom]), outline = tuple(colors_chrs[chrom]))\n",
    "\n",
    "plt.imshow(np.asarray(im),cmap=plt.cm.gray, vmin=29000, vmax=None, aspect='auto')\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Tier 2 detection\n",
    "\n",
    "From the tsv file containing all the possible patches we trace we use the script \"trace_OligoFISSEQ.py\" to trace the chromosomes in our sample image"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "input/merged/OFQv61_w3_3_decon/roi_2_punctas.tsv ['3', 'X', '16', '5', '19', '2'] None input/probes_36plex_sharing.csv input/merged/OFQv61_w3_3_decon ['2pR1', '2pR2', '2pR3', '2qR1', '2qR2', '2qR3', '3pR1', '3pR2', '3pR3', '3qR1', '3qR2', '3qR3', '5pR1', '5pR2', '5pR3', '5qR1', '5qR2', '5qR3', '16pR1', '16pR2', '16pR3', '16qR1', '16qR2', '16qR3', 'XpR1', 'XpR2', 'XpR3', 'XqR1', 'XqR2', 'XqR3', '19pR1', '19pR2', '19pR3', '19qR1', '19qR2', '19qR3'] ['3', 'X', '16', '5', '19', '2']\n",
      "Using filter by minimum intensity: 0.49972528792207777\n",
      "Processing roi 1\n",
      "Centers for chromosome  3 [array([12.14547705,  8.62385482,  3.83207989]), array([14.80347393,  4.64374186,  4.36058502])]\n",
      "Centers for chromosome  X [array([ 6.74667354, 12.10066285,  3.39937497])]\n",
      "Centers for chromosome  16 [array([4.54825752, 8.532579  , 3.88600102]), array([8.39957203, 9.90295191, 4.19243958])]\n",
      "Centers for chromosome  5 [array([10.61718325, 11.11498288,  5.56876794]), array([14.74218192,  6.32917321,  4.74479167])]\n",
      "Centers for chromosome  19 [array([13.61590196,  7.95041687,  4.40821156]), array([11.43251448,  4.88500249,  4.45535714])]\n",
      "Centers for chromosome  2 [array([4.37394069, 8.67487815, 4.99047962]), array([9.47563253, 3.81498715, 4.44897259])]\n",
      "getting solutions\n",
      " There are 200 possible solutions\n",
      "=================> BEST SOLUTION <==============\n",
      "prob 10.515703568560363\n",
      "\"3pR1-1\" (1, array([13.86840114,  8.5344007 ,  3.5       ]), 1.169237132685704)\n",
      "\"3pR2-1\" (0, array([12.53490103,  8.5344007 ,  4.        ]), 1.141052288608902)\n",
      "\"3pR3-1\" (0, array([12.53490103,  8.80110072,  3.        ]), 1.7601869955213436)\n",
      "\"3qR3-1\" (1, array([ 9.33450077, 10.66800088,  6.        ]), 1.437326481230551)\n",
      "\"3pR1-2\" (0, array([15.20190125,  5.06730042,  3.        ]), 1.1511211119021119)\n",
      "\"3pR3-2\" (1, array([15.20190125,  4.00050033,  4.5       ]), 1.8972099187338811)\n",
      "\"3qR3-2\" (0, array([14.40180119,  6.1341005 ,  3.5       ]), 1.95956963987787)\n",
      "7/7\n",
      "getting solutions\n",
      " There are 1 possible solutions\n",
      "=================> BEST SOLUTION <==============\n",
      "prob 10.515703568560363\n",
      "7/7\n",
      "getting solutions\n",
      " There are 600 possible solutions\n",
      "=================> BEST SOLUTION <==============\n",
      "prob 16.335845966159813\n",
      "\"3qR1-1\" (0, array([12.00150099,  6.93420057,  3.        ]), 1.5009148250557272)\n",
      "\"3qR2-1\" (6, array([10.9347009 , 10.13460083,  3.5       ]), 0.8360150631446713)\n",
      "\"3pR2-2\" (1, array([14.40180119,  4.00050033,  4.5       ]), 1.1682742427834876)\n",
      "\"3qR1-2\" (1, array([14.66850121,  4.8006004 ,  5.        ]), 1.3575131015753357)\n",
      "\"3qR2-2\" (0, array([13.60170112,  6.40080053,  4.        ]), 0.9574251650402267)\n",
      "getting solutions\n",
      " There are 12 possible solutions\n",
      "=================> BEST SOLUTION <==============\n",
      "prob 3.8629960523129436\n",
      "\"XpR3-1\" (0, array([ 6.1341005, 10.9347009,  3.5      ]), 1.3956238587144456)\n",
      "\"XqR1-1\" (0, array([ 6.66750055, 11.73480097,  3.5       ]), 1.1296781562083846)\n",
      "\"XqR2-1\" (0, array([ 7.20090059, 13.60170112,  3.        ]), 1.3376940373901136)\n",
      "4/4\n",
      "getting solutions\n",
      " There are 1 possible solutions\n",
      "=================> BEST SOLUTION <==============\n",
      "prob 3.8629960523129436\n",
      "4/4\n",
      "getting solutions\n",
      " There are 66 possible solutions\n",
      "=================> BEST SOLUTION <==============\n",
      "prob 6.66880447088936\n",
      "\"XpR1-1\" (0, array([ 8.00100066, 10.13460083,  3.        ]), 0.9149887281663889)\n",
      "\"XpR2-1\" (1, array([ 6.93420057, 10.13460083,  4.        ]), 1.1437699281454246)\n",
      "\"XqR3-1\" (10, array([ 7.20090059, 13.06830108,  7.        ]), 0.7470497622646027)\n",
      "getting solutions\n",
      " There are 32281 possible solutions\n",
      "=================> BEST SOLUTION <==============\n",
      "prob 11.397338806951876\n",
      "\"16pR1-1\" (0, array([4.8006004 , 8.80110072, 3.5       ]), 1.016070897002887)\n",
      "\"16pR2-1\" (0, array([4.8006004 , 7.46760061, 3.        ]), 1.1201241950711212)\n",
      "\"16pR3-1\" (0, array([4.53390037, 9.06780075, 4.        ]), 0.9766475078373416)\n",
      "\"16qR1-1\" (1, array([5.06730042, 8.00100066, 5.        ]), 1.054482108278244)\n",
      "\"16qR2-1\" (1, array([4.26720035, 9.86790081, 4.        ]), 1.1803697139838223)\n",
      "\"16qR3-1\" (0, array([3.73380031, 9.06780075, 4.        ]), 1.289405599744422)\n",
      "\"16pR3-2\" (1, array([ 7.73430064, 10.13460083,  4.5       ]), 1.1127726183598663)\n",
      "\"16qR1-2\" (0, array([8.5344007 , 9.60120079, 3.5       ]), 1.2364160473224364)\n",
      "\"16qR2-2\" (0, array([8.80110072, 9.86790081, 4.5       ]), 1.0446764542615372)\n",
      "\"16qR3-2\" (1, array([ 8.5344007 , 10.40130086,  3.5       ]), 1.3663736650901974)\n",
      "11/11\n",
      "getting solutions\n",
      " There are 1 possible solutions\n",
      "=================> BEST SOLUTION <==============\n",
      "prob 11.397338806951876\n",
      "11/11\n",
      "getting solutions\n",
      " There are 6 possible solutions\n",
      "=================> BEST SOLUTION <==============\n",
      "prob 13.252057320230428\n",
      "\"16pR1-2\" (1, array([8.5344007 , 9.60120079, 4.5       ]), 0.9447709488556953)\n",
      "\"16pR2-2\" (1, array([6.93420057, 9.86790081, 5.        ]), 0.9099475644228563)\n",
      "getting solutions\n",
      " There are 32 possible solutions\n",
      "=================> BEST SOLUTION <==============\n",
      "prob 6.960849029171688\n",
      "\"5pR3-1\" (1, array([ 9.33450077, 10.66800088,  6.        ]), 1.437326481230551)\n",
      "\"5qR2-1\" (0, array([10.9347009 , 12.26820101,  4.5       ]), 1.3024453757191359)\n",
      "\"5qR3-1\" (0, array([10.13460083, 11.73480097,  4.5       ]), 1.1975056732169373)\n",
      "\"5pR3-2\" (0, array([14.40180119,  6.1341005 ,  3.5       ]), 1.95956963987787)\n",
      "\"5qR3-2\" (1, array([14.66850121,  6.66750055,  4.5       ]), 1.064001859127193)\n",
      "5/5\n",
      "getting solutions\n",
      " There are 1 possible solutions\n",
      "=================> BEST SOLUTION <==============\n",
      "prob 6.960849029171688\n",
      "5/5\n",
      "getting solutions\n",
      " There are 152 possible solutions\n",
      "=================> BEST SOLUTION <==============\n",
      "prob 13.380994581689789\n",
      "\"5pR1-1\" (0, array([10.13460083, 10.9347009 ,  4.5       ]), 0.9874932496036946)\n",
      "\"5pR2-1\" (0, array([12.00150099,  9.86790081,  5.5       ]), 1.064981933647946)\n",
      "\"5qR1-1\" (4, array([10.9347009 ,  9.60120079,  3.5       ]), 0.8801897517781134)\n",
      "\"5pR1-2\" (1, array([13.3350011,  8.5344007,  3.       ]), 1.225045713194616)\n",
      "\"5pR2-2\" (1, array([14.93520123,  7.20090059,  4.        ]), 0.859273453360049)\n",
      "\"5qR2-2\" (1, array([15.20190125,  7.73430064,  5.        ]), 1.4031614509336796)\n",
      "66/66\n",
      "getting solutions\n",
      " There are 202176 possible solutions\n",
      "=================> BEST SOLUTION <==============\n",
      "prob 10.93860275484358\n",
      "\"19pR1-1\" (0, array([12.26820101,  8.5344007 ,  3.5       ]), 0.9539769496290531)\n",
      "\"19pR2-1\" (0, array([13.60170112,  8.5344007 ,  4.        ]), 1.155569245890045)\n",
      "\"19qR1-1\" (0, array([12.80160105,  8.26770068,  4.        ]), 1.0415113040256458)\n",
      "\"19qR2-1\" (0, array([13.86840114,  7.73430064,  4.        ]), 1.1495096302810632)\n",
      "\"19qR3-1\" (0, array([13.60170112,  7.20090059,  5.5       ]), 1.165795581888734)\n",
      "\"19pR1-2\" (1, array([11.20140092,  5.60070046,  4.5       ]), 1.0510872306871548)\n",
      "\"19pR2-2\" (1, array([10.66800088,  4.53390037,  4.5       ]), 1.1239429216717973)\n",
      "\"19pR3-2\" (0, array([11.73480097,  6.1341005 ,  4.5       ]), 0.9841130131359713)\n",
      "\"19qR1-2\" (1, array([12.00150099,  5.33400044,  4.        ]), 1.101006147414209)\n",
      "\"19qR2-2\" (1, array([11.73480097,  4.00050033,  4.        ]), 1.2120907302199053)\n",
      "12/12\n",
      "getting solutions\n",
      " There are 1 possible solutions\n",
      "=================> BEST SOLUTION <==============\n",
      "prob 10.93860275484358\n",
      "12/12\n",
      "getting solutions\n",
      " There are 3 possible solutions\n",
      "=================> BEST SOLUTION <==============\n",
      "prob 11.872237886171634\n",
      "\"19qR3-2\" (2, array([11.46810094,  5.06730042,  4.        ]), 0.933635131328055)\n",
      "24/24\n",
      "getting solutions\n",
      " There are 512 possible solutions\n",
      "=================> BEST SOLUTION <==============\n",
      "prob 11.328272090742104\n",
      "\"2pR1-1\" (1, array([4.00050033, 6.93420057, 4.5       ]), 1.1929611982557866)\n",
      "\"2pR2-1\" (0, array([4.8006004 , 8.26770068, 4.        ]), 1.0395457373909973)\n",
      "\"2pR3-1\" (1, array([5.33400044, 7.73430064, 5.        ]), 1.4583503700693552)\n",
      "\"2qR1-1\" (0, array([4.53390037, 8.80110072, 5.        ]), 1.2458049679081609)\n",
      "\"2qR2-1\" (1, array([3.46710029, 9.86790081, 5.        ]), 1.2249560041978718)\n",
      "\"2pR1-2\" (0, array([7.73430064, 4.00050033, 3.5       ]), 1.2056902021807472)\n",
      "\"2pR2-2\" (1, array([9.06780075, 4.26720035, 4.5       ]), 1.4910001372068886)\n",
      "\"2pR3-2\" (0, array([10.40130086,  3.73380031,  5.        ]), 1.4108415367819636)\n",
      "\"2qR2-2\" (0, array([10.66800088,  3.46710029,  4.        ]), 1.059121936750333)\n",
      "9/9\n",
      "getting solutions\n",
      " There are 1 possible solutions\n",
      "=================> BEST SOLUTION <==============\n",
      "prob 11.328272090742104\n",
      "9/9\n",
      "getting solutions\n",
      " There are 27 possible solutions\n",
      "=================> BEST SOLUTION <==============\n",
      "prob 14.329598671283504\n",
      "\"2qR3-1\" (0, array([ 4.00050033, 10.40130086,  4.        ]), 1.0804513145133279)\n",
      "\"2qR1-2\" (3, array([8.80110072, 6.40080053, 5.        ]), 0.8771078165407084)\n",
      "\"2qR3-2\" (1, array([11.46810094,  3.73380031,  3.5       ]), 1.0437674494873648)\n",
      "Percentage of detected loci for roi 0, 96.969697\n",
      "Percentage of detected loci, 96.969697\n"
     ]
    }
   ],
   "source": [
    "%%python\n",
    "import os\n",
    "\n",
    "output_dir = os.path.dirname(roi_puncta_file)\n",
    "barcode_file = \"input/probes_36plex_sharing.csv\"\n",
    "                         \n",
    "%run \"scripts/trace_OligoFISSEQ\" --puncta_file $roi_puncta_file \\\n",
    "    --barcode_file $barcode_file --threshold 'SOLiD,4,MSBS,False' -n 4 --outdir $output_dir"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The resulting chromosome traces are stored in \"roi_{roi_number}_punctas.out\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "  roi chr  loci      pos_x      pos_y  pos_z     score  nbr_pixels  \\\n",
      "0   0   3  3pR1  13.868401   8.534401    3.5  1.169237          29   \n",
      "1   0   3  3pR2  12.534901   8.534401    4.0  1.141052          26   \n",
      "2   0   3  3pR3  12.534901   8.801101    3.0  1.760187         110   \n",
      "3   0   3  3qR1  12.001501   6.934201    3.0  1.500915          81   \n",
      "4   0   3  3qR2  10.934701  10.134601    3.5  0.836015           1   \n",
      "\n",
      "   max_intensity  match_barcodes ploid  \n",
      "0       0.995341               4     1  \n",
      "1       0.988899               4     1  \n",
      "2       0.999332               4     1  \n",
      "3       0.950232               4     1  \n",
      "4       0.865089               4     1  \n"
     ]
    }
   ],
   "source": [
    "%%python\n",
    "roi_traced_file = \"input/merged/OFQv61_w3_3_decon/roi_2_punctas_out.tsv\"\n",
    "roi_traced = pd.read_csv(roi_traced_file, sep='\\t')\n",
    "tmp_col = roi_traced['loci'].str.split(\"-\", n = 1, expand = True)\n",
    "roi_traced['chr'] = roi_traced['chr'].astype(str)\n",
    "roi_traced['loci'] = tmp_col[0]\n",
    "roi_traced['ploid'] = tmp_col[1]\n",
    "print(roi_traced.head())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We check on the image the traces"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x7f8673195b50>"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "%%python\n",
    "im = Image.open('input/roi_max_intensity.png')\n",
    "double_size = (im.size[0]*10, im.size[1]*10)\n",
    "im = im.resize(double_size,resample=0)\n",
    "draw = ImageDraw.Draw(im)\n",
    "for chrom in colors_chrs.keys():\n",
    "    for pl in range(2):\n",
    "        chr_locis = roi_traced[(\n",
    "                           (roi_traced['chr']==chrom)\n",
    "                           &(roi_traced['ploid']==str(pl+1))      \n",
    "                           )].sort_values('loci')[['pos_x','pos_y','loci']].values.tolist()\n",
    "        if len(chr_locis) == 0:\n",
    "            continue\n",
    "        for i in range(0,len(chr_locis)-1):\n",
    "            draw.line((10*chr_locis[i][0]/pixelWidth, 10*chr_locis[i][1]/pixelWidth,\n",
    "                       10*chr_locis[i+1][0]/pixelWidth+1, 10*chr_locis[i+1][1]/pixelWidth+1),\n",
    "                       fill = tuple(colors_chrs[chrom]), width=3)\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(12, 10)) \n",
    "ax.imshow(np.asarray(im),cmap=plt.cm.gray, vmin=29000, vmax=None, aspect='auto')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Groovy",
   "language": "groovy",
   "name": "groovy"
  },
  "language_info": {
   "codemirror_mode": "groovy",
   "file_extension": ".groovy",
   "mimetype": "",
   "name": "Groovy",
   "nbconverter_exporter": "",
   "version": "2.5.6"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": false,
   "sideBar": false,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {},
   "toc_section_display": false,
   "toc_window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}