{
 "metadata": {
  "name": "",
  "signature": "sha256:6beabf4f7341b77308ce7652d05a0bc127bac90cd89859a4b47e64a632def4db"
 },
 "nbformat": 3,
 "nbformat_minor": 0,
 "worksheets": [
  {
   "cells": [
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "# Measuring chromatin fluorescence\n",
      "\n",
      "Goal: we want to quantify the amount of a particular protein (red fluorescence) localized on the centromeres (green) versus the rest of the chromosome (blue).\n",
      "\n",
      "![Chromosomes](images/chromosomes.jpg)\n",
      "\n",
      "The main challenge here is the uneven illumination, which makes isolating the chromosomes a struggle."
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "%matplotlib inline\n",
      "import numpy as np\n",
      "from matplotlib import cm, pyplot as plt\n",
      "import skdemo\n",
      "plt.rcParams['image.cmap'] = 'cubehelix'\n",
      "plt.rcParams['image.interpolation'] = 'none'"
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "from skimage import io\n",
      "image = io.imread('images/chromosomes.tif')\n",
      "skdemo.imshow_with_histogram(image)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "Let's separate the channels so we can work on each individually."
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "protein, centromeres, chromosomes = image.transpose((2, 0, 1))"
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "Getting the centromeres is easy because the signal is so clean:"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "from skimage.filter import threshold_otsu\n",
      "centromeres_binary = centromeres > threshold_otsu(centromeres)\n",
      "skdemo.imshow_all(centromeres, centromeres_binary)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "But getting the chromosomes is not so easy:"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "chromosomes_binary = chromosomes > threshold_otsu(chromosomes)\n",
      "skdemo.imshow_all(chromosomes, chromosomes_binary)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "Let's try using an adaptive threshold:"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "from skimage.filter import threshold_adaptive\n",
      "chromosomes_adapt = threshold_adaptive(chromosomes, block_size=51)\n",
      "# Question: how did I choose this block size?\n",
      "skdemo.imshow_all(chromosomes, chromosomes_adapt)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "Not only is the uneven illumination a problem, but there seem to be some artifacts due to the illumination pattern!\n",
      "\n",
      "**Exercise:** Can you think of a way to fix this?\n",
      "\n",
      "(Hint: in addition to everything you've learned so far, check out [`skimage.morphology.remove_small_objects`](http://scikit-image.org/docs/dev/api/skimage.morphology.html#skimage.morphology.remove_small_objects))"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "Now that we have the centromeres and the chromosomes, it's time to do the science: get the distribution of intensities in the red channel using both centromere and chromosome locations."
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "# Replace \"None\" below with the right expressions!\n",
      "centromere_intensities = None\n",
      "chromosome_intensities = None\n",
      "all_intensities = np.concatenate((centromere_intensities,\n",
      "                                  chromosome_intensities))\n",
      "minint = np.min(all_intensities)\n",
      "maxint = np.max(all_intensities)\n",
      "bins = np.linspace(minint, maxint, 100)\n",
      "plt.hist(centromere_intensities, bins=bins, color='blue',\n",
      "         alpha=0.5, label='centromeres')\n",
      "plt.hist(chromosome_intensities, bins=bins, color='orange',\n",
      "         alpha=0.5, label='chromosomes')\n",
      "plt.legend(loc='upper right')\n",
      "plt.show()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    }
   ],
   "metadata": {}
  }
 ]
}