calcium+visual_stimuli=nwb

calcium+to+nwb.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### build dataset object from lims ID for the session"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "lims_id = 672584839"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/local1/miniconda2/envs/ophys_nwb/lib/python2.7/site-packages/h5py/__init__.py:36: FutureWarning: Conversion of the second argument of issubdtype from `float` to `np.floating` is deprecated. In future, it will be treated as `np.float64 == np.dtype(float).type`.\n",
      "  from ._conv import register_converters as _register_converters\n"
     ]
    }
   ],
   "source": [
    "from visual_behavior_ophys.dataset.visual_behavior_scientifica_dataset import VisualBehaviorScientificaDataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "sync\n",
      "calculating monitor delay\n",
      "Visual frames detected in sync: 215800\n",
      "2P frames detected in sync: 111682\n",
      "180313\n",
      "visual frames in pkl file: 215800\n",
      "stim type is  images\n",
      "length of traces: 111680\n",
      "number of segmented cells: 130\n",
      "dropped 2P frames detected\n"
     ]
    }
   ],
   "source": [
    "dataset = VisualBehaviorScientificaDataset(lims_id)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## this is the data we need to get into NWB\n",
    "\n",
    "❌ - incomplete\n",
    "✔️ - complete\n",
    "\n",
    "- ❌ metadata\n",
    "- ❌ alignment\n",
    "  - ❌ Body camera\n",
    "  - ❌ Eye tracking\n",
    "  - ❌ Stimulus\n",
    "  - ❌ 2P vsync\n",
    "- ❌ acquisition\n",
    "  - ❌ body camera\n",
    "  - ❌ eye camera\n",
    "  - ❌ wheel encoder\n",
    "  - ❌ lick events\n",
    "  - ❌ calcium movie\n",
    "- ❌ stimulus\n",
    "  - ❌ visual stimulus events\n",
    "  - ❌ reward events\n",
    "  - ❌ trials\n",
    "- ❌ processing/ophys\n",
    "  - ❌ motion correction\n",
    "  - ✔️ image segmentation\n",
    "  - ✔️ ROIs\n",
    "  - ❌ ROI metadata?\n",
    "  - ✔️ dFF\n",
    "  - ❌ fluorescence\n",
    "- ❌ processing/eye\n",
    "  - ❌ pupil location\n",
    "  - ❌ pupil size\n",
    "- ❌ processing/running\n",
    "  - ❌ wheel velocity"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## let's load data using the dataset object"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "length of traces: 111680\n",
      "number of segmented cells: 130\n",
      "dropped 2P frames detected\n"
     ]
    }
   ],
   "source": [
    "# get ROI & dF/F data\n",
    "\n",
    "roi_metrics = dataset.get_roi_metrics()\n",
    "\n",
    "imaging_rate = '{:0.1f}Hz'.format(dataset.metadata['ophys_frame_rate'])\n",
    "\n",
    "roi_masks = dataset.roi_dict\n",
    "max_projection = dataset.max_projection\n",
    "\n",
    "dFF, dFF_t = dataset.get_dff_traces()\n",
    "\n",
    "cell_ids = dataset.get_cell_specimen_id_for_index(range(dFF.shape[0]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# first, we create the nwbfile object"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "NA = 'THIS REQUIRED ATTRIBUTE INTENTIONALLY LEFT BLANK.'\n",
    "\n",
    "from pynwb import NWBFile\n",
    "nwbfile = NWBFile(\n",
    "    source = NA, \n",
    "    session_description = NA, \n",
    "    identifier = str(lims_id), \n",
    "    session_start_time = dataset.lims_data['experiment_date'][0],\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## next, we create nwb objects for ophys assets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "two_p_device = 'scientifica'\n",
    "\n",
    "from pynwb import ophys\n",
    "optical_channel = ophys.OpticalChannel(\n",
    "    name = 'Optical Channel',\n",
    "    source = NA,\n",
    "    description = NA,\n",
    "    emission_lambda = NA,\n",
    ")\n",
    "\n",
    "imaging_plane = nwbfile.create_imaging_plane(\n",
    "    name='test_imaging_plane',\n",
    "    source=NA,\n",
    "    optical_channel=optical_channel,\n",
    "    description=NA,\n",
    "    device=NA,\n",
    "    excitation_lambda=NA, \n",
    "    imaging_rate=NA, \n",
    "    indicator=NA, \n",
    "    location=NA,\n",
    "    manifold=[], \n",
    "    conversion=1.0, \n",
    "    unit=NA, \n",
    "    reference_frame=NA,\n",
    ")\n",
    "\n",
    "from pynwb import image\n",
    "max_proj = image.ImageSeries(\n",
    "    name = 'Maximum Projection Image',\n",
    "    source = NA,\n",
    "    data = max_projection,\n",
    "    unit = NA,\n",
    "    format = 'raw',\n",
    "    timestamps = [0.0], # <- This is required, so let's make it at the beginning of the experiment\n",
    ")\n",
    "nwbfile.add_acquisition(max_proj)\n",
    "\n",
    "ophys_module = nwbfile.create_processing_module(\n",
    "    name=\"ophys\",\n",
    "    description=\"calcium responses\",\n",
    "    source=\"Allen Brain Observatory: Visual Behavior\",\n",
    ")\n",
    "\n",
    "from pynwb.ophys import ImageSegmentation\n",
    "img_seg = ImageSegmentation(\n",
    "    'image segmentation container'\n",
    ")\n",
    "ophys_module.add_data_interface(img_seg)\n",
    "\n",
    "plane_segmentation = img_seg.create_plane_segmentation(\n",
    "    source=NA,\n",
    "    description=NA,\n",
    "    imaging_plane=imaging_plane, \n",
    "    name='plane_0', \n",
    "    reference_images=max_proj,\n",
    ")\n",
    "\n",
    "\n",
    "for cell_id in cell_ids: # <- must iterate like this because order matters\n",
    "    img_mask = roi_masks[cell_id]\n",
    "    \n",
    "    plane_segmentation.add_roi(\n",
    "        pixel_mask=[], \n",
    "        image_mask=img_mask,\n",
    "    )\n",
    "    \n",
    "    \n",
    "from pynwb.ophys import DfOverF\n",
    "\n",
    "dff_container = DfOverF('container of fluorescence data')\n",
    "ophys_module.add_data_interface(dff_container)\n",
    "\n",
    "# get an ROI table region i.e. a subset of ROIs to create a RoiResponseSeries from\n",
    "rt_region = plane_segmentation.create_roi_table_region(\n",
    "    range(len(cell_ids)), # <- this should be names, in order\n",
    "    NA,\n",
    ")\n",
    "\n",
    "dFF_series = dff_container.create_roi_response_series(\n",
    "    name='df_over_f', \n",
    "    source=NA,\n",
    "    data=dFF, \n",
    "    unit='NA', \n",
    "    rois = rt_region,\n",
    "    timestamps=dFF_t,\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## next, let's load the record of visual stimuli from the dataset object"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "visual_stimulus = dataset.flashes[['time','image_category','image_name']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/local1/miniconda2/envs/ophys_nwb/lib/python2.7/site-packages/ipykernel_launcher.py:1: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/indexing.html#indexing-view-versus-copy\n",
      "  \"\"\"Entry point for launching an IPython kernel.\n"
     ]
    }
   ],
   "source": [
    "visual_stimulus['duration'] = 0.25"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>time</th>\n",
       "      <th>image_category</th>\n",
       "      <th>image_name</th>\n",
       "      <th>duration</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>flash_index</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>57.56506</td>\n",
       "      <td>im045</td>\n",
       "      <td>im045</td>\n",
       "      <td>0.25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>58.31548</td>\n",
       "      <td>im045</td>\n",
       "      <td>im045</td>\n",
       "      <td>0.25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>59.06612</td>\n",
       "      <td>im045</td>\n",
       "      <td>im045</td>\n",
       "      <td>0.25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>59.81668</td>\n",
       "      <td>im045</td>\n",
       "      <td>im045</td>\n",
       "      <td>0.25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>60.56731</td>\n",
       "      <td>im045</td>\n",
       "      <td>im045</td>\n",
       "      <td>0.25</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 time image_category image_name  duration\n",
       "flash_index                                              \n",
       "0            57.56506          im045      im045      0.25\n",
       "1            58.31548          im045      im045      0.25\n",
       "2            59.06612          im045      im045      0.25\n",
       "3            59.81668          im045      im045      0.25\n",
       "4            60.56731          im045      im045      0.25"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "visual_stimulus.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "images = dataset.pkl['image_dict']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(8, 918, 1174)"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "image_stack = []\n",
    "image_lookup = {}\n",
    "\n",
    "ii = 0\n",
    "for cat, cat_images in images.iteritems():\n",
    "    for name, img in cat_images.iteritems():\n",
    "        image_stack.append(img)\n",
    "        image_lookup[name] = ii \n",
    "        ii += 1\n",
    "\n",
    "image_stack = np.stack(image_stack)\n",
    "\n",
    "image_stack.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## now, we'll create the NWB objects for the visual stimulus record"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "visual_stimulus_images = image.ImageSeries(\n",
    "    name = 'natural_images',\n",
    "    source = NA,\n",
    "    data = image_stack,\n",
    "    unit = NA,\n",
    "    format = 'raw',\n",
    "    timestamps = [0.0], # <- This is required, so let's make it at the beginning of the experiment\n",
    ")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<pynwb.image.ImageSeries at 0x7ff6d08e5dd0>"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "nwbfile.add_stimulus_template(visual_stimulus_images)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([4, 6, 1, 7, 3, 2, 0, 5])"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "visual_stimulus['image_name'].map(image_lookup).unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "image_index = image.IndexSeries(\n",
    "    name='natural_images_timeseries',\n",
    "    source=NA,\n",
    "    data=visual_stimulus['image_name'].map(image_lookup).values,\n",
    "    unit=NA,\n",
    "    indexed_timeseries=visual_stimulus_images,\n",
    "    timestamps=visual_stimulus['time'].values,\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<pynwb.image.IndexSeries at 0x7ff6d3f072d0>"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "nwbfile.add_stimulus(image_index)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## finally, we'll save the file"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ophys_672584839.nwb\n"
     ]
    }
   ],
   "source": [
    "from pynwb import NWBHDF5IO\n",
    "# write data\n",
    "\n",
    "filename = 'ophys_{}.nwb'.format(lims_id)\n",
    "print filename\n",
    "with NWBHDF5IO(filename, 'w') as io:\n",
    "    io.write(nwbfile)"
   ]
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "kernelspec": {
   "display_name": "ophys_nwb",
   "language": "python",
   "name": "ophys_nwb"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}