RUN_Atlas.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## RUN: Atlas Construction\n",
    "\n",
    "This notebook runs the atlas prediction pipeline across the different datasets.\n",
    "\n",
    "Model selection and hyperparameter optimization was done in `DEV_Atlas.ipynb`."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Prep"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "### Imports\n",
    "\n",
    "# Generic\n",
    "from __future__ import division\n",
    "import os, sys, pickle\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "# Modules\n",
    "from katachi.pipelines import atlas_construction as ac"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "### Function to parse relevant IDs from IDR bulk data\n",
    "\n",
    "def parse_from_IDR(dir_path, target):\n",
    "    \n",
    "    # Get all samples\n",
    "    samples = [d for d in os.listdir(dir_path) if len(d)==10\n",
    "               and os.path.isdir(os.path.join(dir_path, d))]\n",
    "    \n",
    "    # Select relevant samples\n",
    "    relevant_samples = []\n",
    "    for d in samples:\n",
    "        \n",
    "        # Get image files\n",
    "        images = [i for i in os.listdir(os.path.join(dir_path, d))\n",
    "                  if i.startswith(d) and i.endswith('.tif')]\n",
    "        \n",
    "        # Special case for membranes only\n",
    "        if target=='membranes_only':\n",
    "            if all(['lynEGFP' in img for img in images]):\n",
    "                relevant_samples.append(d)\n",
    "            \n",
    "        # All other cases\n",
    "        else:\n",
    "            if any([img.endswith(target+'.tif') for img in images]):\n",
    "                relevant_samples.append(d)\n",
    "    \n",
    "    return relevant_samples"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Atlas Construction based on TFOR"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**tagRFPtUtrCH**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "### Predict tagRFPtUtrCH channel for all prims that do not have it\n",
    "\n",
    "# Target directories\n",
    "train_dirpath   = r'data\\experimentA\\image_data'\n",
    "predict_dirpath = r'data\\experimentA\\image_data'\n",
    "\n",
    "# Target IDs\n",
    "train_IDs = parse_from_IDR(train_dirpath, 'tagRFPtUtrCH')\n",
    "print \"Found %i training IDs!\" % len(train_IDs)\n",
    "\n",
    "# Channels\n",
    "ref_channel = ['lynEGFP_seg_LMs_TFOR_kmeansPRES_DDDS_CBEmanh',\n",
    "               'lynEGFP_linUnmix_seg_LMs_TFOR_kmeansPRES_DDDS_CBEmanh']\n",
    "sec_channel = 'tagRFPtUtrCH_LMs_TFOR_kmeansPRES_DDDS_CBEmanh'\n",
    "\n",
    "# Core settings\n",
    "outlier_removal_ref = 'isolation_forest'\n",
    "outlier_removal_sec = 'isolation_forest'\n",
    "outlier_removal_cov = 'percentile_thresh'\n",
    "covariates_to_use   = 'img.cell.'+sec_channel.split('_')[0]+'.mean_total'\n",
    "regressor           = 'MO-SVR'\n",
    "\n",
    "# Additional parameters\n",
    "outlier_params_ref = { 'isoforest_params' : {'contamination':0.05}}\n",
    "outlier_params_sec = { 'isoforest_params' : {'contamination':0.05}}\n",
    "outlier_params_cov = { 'bounds'     : 'lower',\n",
    "                       'percentile' : 33 }\n",
    "regressor_params   = { 'kernel'  : 'rbf',\n",
    "                       'C'       : 20.0,\n",
    "                       'epsilon' :  0.5,\n",
    "                       'gamma'   :  1.0 / 20.0 * 0.01 }\n",
    "atlas_params       = { 'zscore_X'       : True,\n",
    "                       'zscore_y'       : True,\n",
    "                       'pca_X'          : True,\n",
    "                       'pca_y'          : True,\n",
    "                       'rezscore_X'     : False,\n",
    "                       'rezscore_y'     : False,\n",
    "                       'subselect_X'    : 20,\n",
    "                       'subselect_y'    : 20,\n",
    "                       'add_covariates' : None }\n",
    "\n",
    "# Additional arguments\n",
    "recurse     = True\n",
    "ignore_self = False\n",
    "processes   = 14\n",
    "profiling   = True\n",
    "verbose     = True\n",
    "\n",
    "# Run prediction pipeline\n",
    "ac.atlas_construction(train_dirpath, predict_dirpath, \n",
    "                      ref_channel, sec_channel, \n",
    "                      train_IDs=train_IDs, predict_IDs=None,\n",
    "                      recurse=recurse, ignore_self=ignore_self, \n",
    "                      processes=processes, profiling=profiling, verbose=verbose,\n",
    "                      outlier_removal_ref=outlier_removal_ref,\n",
    "                      outlier_removal_sec=outlier_removal_sec,\n",
    "                      outlier_removal_cov=outlier_removal_cov,\n",
    "                      covariates_to_use=covariates_to_use,\n",
    "                      regressor=regressor, \n",
    "                      outlier_params_ref=outlier_params_ref,\n",
    "                      outlier_params_sec=outlier_params_sec,\n",
    "                      outlier_params_cov=outlier_params_cov,\n",
    "                      regressor_params=regressor_params,  \n",
    "                      atlas_params=atlas_params)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**NLStdTomato**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "### Predict NLStdTomato channel for all prims that do not have it\n",
    "\n",
    "# Target directories\n",
    "train_dirpath   = r'data\\experimentA\\image_data'\n",
    "predict_dirpath = r'data\\experimentA\\image_data'\n",
    "\n",
    "# Target IDs\n",
    "train_IDs = parse_from_IDR(train_dirpath, 'NLStdTomato')\n",
    "print \"Found %i training IDs!\" % len(train_IDs)\n",
    "\n",
    "# Channels\n",
    "ref_channel = ['lynEGFP_seg_LMs_TFOR_kmeansPRES_DDDS_CBEmanh',\n",
    "               'lynEGFP_linUnmix_seg_LMs_TFOR_kmeansPRES_DDDS_CBEmanh']\n",
    "sec_channel = 'NLStdTomato_LMs_TFOR_kmeansPRES_DDDS_CBEmanh'\n",
    "\n",
    "# Core settings\n",
    "outlier_removal_ref = 'isolation_forest'\n",
    "outlier_removal_sec = 'isolation_forest'\n",
    "outlier_removal_cov = 'percentile_thresh'\n",
    "covariates_to_use   = 'img.cell.'+sec_channel.split('_')[0]+'.mean_total'\n",
    "regressor           = 'MO-SVR'\n",
    "\n",
    "# Additional parameters\n",
    "outlier_params_ref = { 'isoforest_params' : {'contamination':0.05}}\n",
    "outlier_params_sec = { 'isoforest_params' : {'contamination':0.05}}\n",
    "outlier_params_cov = { 'bounds'     : 'lower',\n",
    "                       'percentile' : 33 }\n",
    "regressor_params   = { 'kernel'  : 'rbf',\n",
    "                       'C'       : 10.0,\n",
    "                       'epsilon' :  0.5,\n",
    "                       'gamma'   :  1.0 / 20.0 * 0.1 }\n",
    "atlas_params       = { 'zscore_X'       : True,\n",
    "                       'zscore_y'       : True,\n",
    "                       'pca_X'          : True,\n",
    "                       'pca_y'          : True,\n",
    "                       'rezscore_X'     : False,\n",
    "                       'rezscore_y'     : False,\n",
    "                       'subselect_X'    : 20,\n",
    "                       'subselect_y'    : 20,\n",
    "                       'add_covariates' : None }\n",
    "\n",
    "# Additional arguments\n",
    "recurse     = True\n",
    "ignore_self = False\n",
    "processes   = 10\n",
    "profiling   = True\n",
    "verbose     = True\n",
    "\n",
    "# Run prediction pipeline\n",
    "ac.atlas_construction(train_dirpath, predict_dirpath, \n",
    "                      ref_channel, sec_channel, \n",
    "                      train_IDs=train_IDs, predict_IDs=None,\n",
    "                      recurse=recurse, ignore_self=ignore_self, \n",
    "                      processes=processes, profiling=profiling, verbose=verbose,\n",
    "                      outlier_removal_ref=outlier_removal_ref,\n",
    "                      outlier_removal_sec=outlier_removal_sec,\n",
    "                      outlier_removal_cov=outlier_removal_cov,\n",
    "                      covariates_to_use=covariates_to_use,\n",
    "                      regressor=regressor, \n",
    "                      outlier_params_ref=outlier_params_ref,\n",
    "                      outlier_params_sec=outlier_params_sec,\n",
    "                      outlier_params_cov=outlier_params_cov,\n",
    "                      regressor_params=regressor_params,  \n",
    "                      atlas_params=atlas_params)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**b4galT1tagRFPt**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "### Predict b4galT1tagRFPt channel for all prims that do not have it\n",
    "\n",
    "# Target directories\n",
    "train_dirpath   = r'data\\experimentA\\image_data'\n",
    "predict_dirpath = r'data\\experimentA\\image_data'\n",
    "\n",
    "# Target IDs\n",
    "train_IDs = parse_from_IDR(train_dirpath, 'b4galT1tagRFPt')\n",
    "print \"Found %i training IDs!\" % len(train_IDs)\n",
    "\n",
    "# Channels\n",
    "ref_channel = ['lynEGFP_seg_LMs_TFOR_kmeansPRES_DDDS_CBEmanh',\n",
    "               'lynEGFP_linUnmix_seg_LMs_TFOR_kmeansPRES_DDDS_CBEmanh']\n",
    "sec_channel = 'b4galT1tagRFPt_LMs_TFOR_kmeansPRES_DDDS_CBEmanh'\n",
    "\n",
    "# Core settings\n",
    "outlier_removal_ref = 'isolation_forest'\n",
    "outlier_removal_sec = 'isolation_forest'\n",
    "outlier_removal_cov = 'percentile_thresh'\n",
    "covariates_to_use   = 'img.cell.'+sec_channel.split('_')[0]+'.mean_total'\n",
    "regressor           = 'MO-SVR'\n",
    "\n",
    "# Additional parameters\n",
    "outlier_params_ref = { 'isoforest_params' : {'contamination':0.05}}\n",
    "outlier_params_sec = { 'isoforest_params' : {'contamination':0.05}}\n",
    "outlier_params_cov = { 'bounds'     : 'lower',\n",
    "                       'percentile' : 33 }\n",
    "regressor_params   = { 'kernel'  : 'rbf',\n",
    "                       'C'       : 20.0,\n",
    "                       'epsilon' :  0.5,\n",
    "                       'gamma'   :  1.0 / 20.0 * 0.01 }\n",
    "atlas_params       = { 'zscore_X'       : True,\n",
    "                       'zscore_y'       : True,\n",
    "                       'pca_X'          : True,\n",
    "                       'pca_y'          : True,\n",
    "                       'rezscore_X'     : False,\n",
    "                       'rezscore_y'     : False,\n",
    "                       'subselect_X'    : 20,\n",
    "                       'subselect_y'    : 20,\n",
    "                       'add_covariates' : None }\n",
    "\n",
    "# Additional arguments\n",
    "recurse     = True\n",
    "ignore_self = False\n",
    "processes   = 10\n",
    "profiling   = True\n",
    "verbose     = True\n",
    "\n",
    "# Run prediction pipeline\n",
    "ac.atlas_construction(train_dirpath, predict_dirpath, \n",
    "                      ref_channel, sec_channel,\n",
    "                      train_IDs=train_IDs, predict_IDs=None,\n",
    "                      recurse=recurse, ignore_self=ignore_self, \n",
    "                      processes=processes, profiling=profiling, verbose=verbose,\n",
    "                      outlier_removal_ref=outlier_removal_ref,\n",
    "                      outlier_removal_sec=outlier_removal_sec,\n",
    "                      outlier_removal_cov=outlier_removal_cov,\n",
    "                      covariates_to_use=covariates_to_use,\n",
    "                      regressor=regressor, \n",
    "                      outlier_params_ref=outlier_params_ref,\n",
    "                      outlier_params_sec=outlier_params_sec,\n",
    "                      outlier_params_cov=outlier_params_cov,\n",
    "                      regressor_params=regressor_params,  \n",
    "                      atlas_params=atlas_params)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**CDMPRtagRFPt**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "### Predict CDMPRtagRFPt channel for all prims that do not have it\n",
    "\n",
    "# Target directories\n",
    "train_dirpath   = r'data\\experimentA\\image_data'\n",
    "predict_dirpath = r'data\\experimentA\\image_data'\n",
    "\n",
    "# Target IDs\n",
    "train_IDs = parse_from_IDR(train_dirpath, 'CDMPRtagRFPt')\n",
    "print \"Found %i training IDs!\" % len(train_IDs)\n",
    "\n",
    "# Channels\n",
    "ref_channel = ['lynEGFP_seg_LMs_TFOR_kmeansPRES_DDDS_CBEmanh',\n",
    "               'lynEGFP_linUnmix_seg_LMs_TFOR_kmeansPRES_DDDS_CBEmanh']\n",
    "sec_channel = 'CDMPRtagRFPt_LMs_TFOR_kmeansPRES_DDDS_CBEmanh'\n",
    "\n",
    "# Core settings\n",
    "outlier_removal_ref = 'isolation_forest'\n",
    "outlier_removal_sec = 'isolation_forest'\n",
    "outlier_removal_cov = 'percentile_thresh'\n",
    "covariates_to_use   = 'img.cell.'+sec_channel.split('_')[0]+'.mean_total'\n",
    "regressor           = 'MO-SVR'\n",
    "\n",
    "# Additional parameters\n",
    "outlier_params_ref = { 'isoforest_params' : {'contamination':0.05}}\n",
    "outlier_params_sec = { 'isoforest_params' : {'contamination':0.05}}\n",
    "outlier_params_cov = { 'bounds'     : 'lower',\n",
    "                       'percentile' : 33 }\n",
    "regressor_params   = { 'kernel'  : 'rbf',\n",
    "                       'C'       : 20.0,\n",
    "                       'epsilon' :  0.5,\n",
    "                       'gamma'   :  1.0 / 20.0 * 0.01 }\n",
    "atlas_params       = { 'zscore_X'       : True,\n",
    "                       'zscore_y'       : True,\n",
    "                       'pca_X'          : True,\n",
    "                       'pca_y'          : True,\n",
    "                       'rezscore_X'     : False,\n",
    "                       'rezscore_y'     : False,\n",
    "                       'subselect_X'    : 20,\n",
    "                       'subselect_y'    : 20,\n",
    "                       'add_covariates' : None }\n",
    "\n",
    "# Additional arguments\n",
    "recurse     = True\n",
    "ignore_self = False\n",
    "processes   = 10\n",
    "profiling   = True\n",
    "verbose     = True\n",
    "\n",
    "# Run prediction pipeline\n",
    "ac.atlas_construction(train_dirpath, predict_dirpath, \n",
    "                      ref_channel, sec_channel, \n",
    "                      train_IDs=train_IDs, predict_IDs=None,\n",
    "                      recurse=recurse, ignore_self=ignore_self, \n",
    "                      processes=processes, profiling=profiling, verbose=verbose,\n",
    "                      outlier_removal_ref=outlier_removal_ref,\n",
    "                      outlier_removal_sec=outlier_removal_sec,\n",
    "                      outlier_removal_cov=outlier_removal_cov,\n",
    "                      covariates_to_use=covariates_to_use,\n",
    "                      regressor=regressor, \n",
    "                      outlier_params_ref=outlier_params_ref,\n",
    "                      outlier_params_sec=outlier_params_sec,\n",
    "                      outlier_params_cov=outlier_params_cov,\n",
    "                      regressor_params=regressor_params,  \n",
    "                      atlas_params=atlas_params)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**mKate2GM130**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "### Predict mKate2GM130 channel for all prims that do not have it\n",
    "\n",
    "# Target directories\n",
    "train_dirpath   = r'data\\experimentA\\image_data'\n",
    "predict_dirpath = r'data\\experimentA\\image_data'\n",
    "\n",
    "# Target IDs\n",
    "train_IDs = parse_from_IDR(train_dirpath, 'mKate2GM130')\n",
    "print \"Found %i training IDs!\" % len(train_IDs)\n",
    "\n",
    "# Channels\n",
    "ref_channel = ['lynEGFP_seg_LMs_TFOR_kmeansPRES_DDDS_CBEmanh',\n",
    "               'lynEGFP_linUnmix_seg_LMs_TFOR_kmeansPRES_DDDS_CBEmanh']\n",
    "sec_channel = 'mKate2GM130_LMs_TFOR_kmeansPRES_DDDS_CBEmanh'\n",
    "\n",
    "# Core settings\n",
    "outlier_removal_ref = 'isolation_forest'\n",
    "outlier_removal_sec = 'isolation_forest'\n",
    "outlier_removal_cov = 'percentile_thresh'\n",
    "covariates_to_use   = 'img.cell.'+sec_channel.split('_')[0]+'.mean_total'\n",
    "regressor           = 'MO-SVR'\n",
    "\n",
    "# Additional parameters\n",
    "outlier_params_ref = { 'isoforest_params' : {'contamination':0.05}}\n",
    "outlier_params_sec = { 'isoforest_params' : {'contamination':0.05}}\n",
    "outlier_params_cov = { 'bounds'     : 'lower',\n",
    "                       'percentile' : 33 }\n",
    "regressor_params   = { 'kernel'  : 'rbf',\n",
    "                       'C'       : 20.0,\n",
    "                       'epsilon' :  0.5,\n",
    "                       'gamma'   :  1.0 / 20.0 * 0.01 }\n",
    "atlas_params       = { 'zscore_X'       : True,\n",
    "                       'zscore_y'       : True,\n",
    "                       'pca_X'          : True,\n",
    "                       'pca_y'          : True,\n",
    "                       'rezscore_X'     : False,\n",
    "                       'rezscore_y'     : False,\n",
    "                       'subselect_X'    : 20,\n",
    "                       'subselect_y'    : 20,\n",
    "                       'add_covariates' : None }\n",
    "\n",
    "# Additional arguments\n",
    "recurse     = True\n",
    "ignore_self = False\n",
    "processes   = 10\n",
    "profiling   = True\n",
    "verbose     = True\n",
    "\n",
    "# Run prediction pipeline\n",
    "ac.atlas_construction(train_dirpath, predict_dirpath, \n",
    "                      ref_channel, sec_channel, \n",
    "                      train_IDs=train_IDs, predict_IDs=None,\n",
    "                      recurse=recurse, ignore_self=ignore_self, \n",
    "                      processes=processes, profiling=profiling, verbose=verbose,\n",
    "                      outlier_removal_ref=outlier_removal_ref,\n",
    "                      outlier_removal_sec=outlier_removal_sec,\n",
    "                      outlier_removal_cov=outlier_removal_cov,\n",
    "                      covariates_to_use=covariates_to_use,\n",
    "                      regressor=regressor, \n",
    "                      outlier_params_ref=outlier_params_ref,\n",
    "                      outlier_params_sec=outlier_params_sec,\n",
    "                      outlier_params_cov=outlier_params_cov,\n",
    "                      regressor_params=regressor_params,  \n",
    "                      atlas_params=atlas_params)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**lysotrackerdeepred**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "### Predict lysotrackerdeepred channel for all prims that do not have it\n",
    "\n",
    "# Target directories\n",
    "train_dirpath   = r'data\\experimentA\\image_data'\n",
    "predict_dirpath = r'data\\experimentA\\image_data'\n",
    "\n",
    "# Target IDs\n",
    "train_IDs = parse_from_IDR(train_dirpath, 'lysotrackerdeepred')\n",
    "print \"Found %i training IDs!\" % len(train_IDs)\n",
    "\n",
    "# Channels\n",
    "ref_channel = ['lynEGFP_seg_LMs_TFOR_kmeansPRES_DDDS_CBEmanh',\n",
    "               'lynEGFP_linUnmix_seg_LMs_TFOR_kmeansPRES_DDDS_CBEmanh']\n",
    "sec_channel = 'lysotrackerdeepred_LMs_TFOR_kmeansPRES_DDDS_CBEmanh'\n",
    "\n",
    "# Core settings\n",
    "outlier_removal_ref = 'isolation_forest'\n",
    "outlier_removal_sec = 'isolation_forest'\n",
    "outlier_removal_cov = 'percentile_thresh'\n",
    "covariates_to_use   = 'img.cell.'+sec_channel.split('_')[0]+'.mean_total'\n",
    "regressor           = 'MO-SVR'\n",
    "\n",
    "# Additional parameters\n",
    "outlier_params_ref = { 'isoforest_params' : {'contamination':0.05}}\n",
    "outlier_params_sec = { 'isoforest_params' : {'contamination':0.05}}\n",
    "outlier_params_cov = { 'bounds'     : 'lower',\n",
    "                       'percentile' : 33 }\n",
    "regressor_params   = { 'kernel'  : 'rbf',\n",
    "                       'C'       : 20.0,\n",
    "                       'epsilon' :  0.5,\n",
    "                       'gamma'   :  1.0 / 20.0 * 0.01 }\n",
    "atlas_params       = { 'zscore_X'       : True,\n",
    "                       'zscore_y'       : True,\n",
    "                       'pca_X'          : True,\n",
    "                       'pca_y'          : True,\n",
    "                       'rezscore_X'     : False,\n",
    "                       'rezscore_y'     : False,\n",
    "                       'subselect_X'    : 20,\n",
    "                       'subselect_y'    : 20,\n",
    "                       'add_covariates' : None }\n",
    "\n",
    "# Additional arguments\n",
    "recurse     = True\n",
    "ignore_self = False\n",
    "processes   = 10\n",
    "profiling   = True\n",
    "verbose     = True\n",
    "\n",
    "# Run prediction pipeline\n",
    "ac.atlas_construction(train_dirpath, predict_dirpath, \n",
    "                      ref_channel, sec_channel, \n",
    "                      train_IDs=train_IDs, predict_IDs=None,\n",
    "                      recurse=recurse, ignore_self=ignore_self, \n",
    "                      processes=processes, profiling=profiling, verbose=verbose,\n",
    "                      outlier_removal_ref=outlier_removal_ref,\n",
    "                      outlier_removal_sec=outlier_removal_sec,\n",
    "                      outlier_removal_cov=outlier_removal_cov,\n",
    "                      covariates_to_use=covariates_to_use,\n",
    "                      regressor=regressor, \n",
    "                      outlier_params_ref=outlier_params_ref,\n",
    "                      outlier_params_sec=outlier_params_sec,\n",
    "                      outlier_params_cov=outlier_params_cov,\n",
    "                      regressor_params=regressor_params,  \n",
    "                      atlas_params=atlas_params)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**mKate2rab5**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "### Predict mKate2rab5 channel for all prims that do not have it\n",
    "\n",
    "# Target directories\n",
    "train_dirpath   = r'data\\experimentA\\image_data'\n",
    "predict_dirpath = r'data\\experimentA\\image_data'\n",
    "\n",
    "# Target IDs\n",
    "train_IDs = parse_from_IDR(train_dirpath, 'mKate2rab5')\n",
    "print \"Found %i training IDs!\" % len(train_IDs)\n",
    "\n",
    "# Channels\n",
    "ref_channel = ['lynEGFP_seg_LMs_TFOR_kmeansPRES_DDDS_CBEmanh',\n",
    "               'lynEGFP_linUnmix_seg_LMs_TFOR_kmeansPRES_DDDS_CBEmanh']\n",
    "sec_channel = 'mKate2rab5_LMs_TFOR_kmeansPRES_DDDS_CBEmanh'\n",
    "\n",
    "# Core settings\n",
    "outlier_removal_ref = 'isolation_forest'\n",
    "outlier_removal_sec = 'isolation_forest'\n",
    "outlier_removal_cov = 'percentile_thresh'\n",
    "covariates_to_use   = 'img.cell.'+sec_channel.split('_')[0]+'.mean_total'\n",
    "regressor           = 'MO-SVR'\n",
    "\n",
    "# Additional parameters\n",
    "outlier_params_ref = { 'isoforest_params' : {'contamination':0.05}}\n",
    "outlier_params_sec = { 'isoforest_params' : {'contamination':0.05}}\n",
    "outlier_params_cov = { 'bounds'     : 'lower',\n",
    "                       'percentile' : 33 }\n",
    "regressor_params   = { 'kernel'  : 'rbf',\n",
    "                       'C'       : 20.0,\n",
    "                       'epsilon' :  0.5,\n",
    "                       'gamma'   :  1.0 / 20.0 * 0.01 }\n",
    "atlas_params       = { 'zscore_X'       : True,\n",
    "                       'zscore_y'       : True,\n",
    "                       'pca_X'          : True,\n",
    "                       'pca_y'          : True,\n",
    "                       'rezscore_X'     : False,\n",
    "                       'rezscore_y'     : False,\n",
    "                       'subselect_X'    : 20,\n",
    "                       'subselect_y'    : 20,\n",
    "                       'add_covariates' : None }\n",
    "\n",
    "# Additional arguments\n",
    "recurse     = True\n",
    "ignore_self = False\n",
    "processes   = 10\n",
    "profiling   = True\n",
    "verbose     = True\n",
    "\n",
    "# Run prediction pipeline\n",
    "ac.atlas_construction(train_dirpath, predict_dirpath, \n",
    "                      ref_channel, sec_channel, \n",
    "                      train_IDs=train_IDs, predict_IDs=None,\n",
    "                      recurse=recurse, ignore_self=ignore_self, \n",
    "                      processes=processes, profiling=profiling, verbose=verbose,\n",
    "                      outlier_removal_ref=outlier_removal_ref,\n",
    "                      outlier_removal_sec=outlier_removal_sec,\n",
    "                      outlier_removal_cov=outlier_removal_cov,\n",
    "                      covariates_to_use=covariates_to_use,\n",
    "                      regressor=regressor, \n",
    "                      outlier_params_ref=outlier_params_ref,\n",
    "                      outlier_params_sec=outlier_params_sec,\n",
    "                      outlier_params_cov=outlier_params_cov,\n",
    "                      regressor_params=regressor_params,  \n",
    "                      atlas_params=atlas_params)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**mKate2rab11**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "### Predict mKate2rab11 channel for all prims that do not have it\n",
    "\n",
    "# Target directories\n",
    "train_dirpath   = r'data\\experimentA\\image_data'\n",
    "predict_dirpath = r'data\\experimentA\\image_data'\n",
    "\n",
    "# Target IDs\n",
    "train_IDs = parse_from_IDR(train_dirpath, 'mKate2rab11')\n",
    "print \"Found %i training IDs!\" % len(train_IDs)\n",
    "\n",
    "# Channels\n",
    "ref_channel = ['lynEGFP_seg_LMs_TFOR_kmeansPRES_DDDS_CBEmanh',\n",
    "               'lynEGFP_linUnmix_seg_LMs_TFOR_kmeansPRES_DDDS_CBEmanh']\n",
    "sec_channel = 'mKate2rab11_LMs_TFOR_kmeansPRES_DDDS_CBEmanh'\n",
    "\n",
    "# Core settings\n",
    "outlier_removal_ref = 'isolation_forest'\n",
    "outlier_removal_sec = 'isolation_forest'\n",
    "outlier_removal_cov = 'percentile_thresh'\n",
    "covariates_to_use   = 'img.cell.'+sec_channel.split('_')[0]+'.mean_total'\n",
    "regressor           = 'MO-SVR'\n",
    "\n",
    "# Additional parameters\n",
    "outlier_params_ref = { 'isoforest_params' : {'contamination':0.05}}\n",
    "outlier_params_sec = { 'isoforest_params' : {'contamination':0.05}}\n",
    "outlier_params_cov = { 'bounds'     : 'lower',\n",
    "                       'percentile' : 33 }\n",
    "regressor_params   = { 'kernel'  : 'rbf',\n",
    "                       'C'       : 20.0,\n",
    "                       'epsilon' :  0.5,\n",
    "                       'gamma'   :  1.0 / 20.0 * 0.01 }\n",
    "atlas_params       = { 'zscore_X'       : True,\n",
    "                       'zscore_y'       : True,\n",
    "                       'pca_X'          : True,\n",
    "                       'pca_y'          : True,\n",
    "                       'rezscore_X'     : False,\n",
    "                       'rezscore_y'     : False,\n",
    "                       'subselect_X'    : 20,\n",
    "                       'subselect_y'    : 20,\n",
    "                       'add_covariates' : None }\n",
    "\n",
    "# Additional arguments\n",
    "recurse     = True\n",
    "ignore_self = False\n",
    "processes   = 10\n",
    "profiling   = True\n",
    "verbose     = True\n",
    "\n",
    "# Run prediction pipeline\n",
    "ac.atlas_construction(train_dirpath, predict_dirpath, \n",
    "                      ref_channel, sec_channel, \n",
    "                      train_IDs=train_IDs, predict_IDs=None,\n",
    "                      recurse=recurse, ignore_self=ignore_self, \n",
    "                      processes=processes, profiling=profiling, verbose=verbose,\n",
    "                      outlier_removal_ref=outlier_removal_ref,\n",
    "                      outlier_removal_sec=outlier_removal_sec,\n",
    "                      outlier_removal_cov=outlier_removal_cov,\n",
    "                      covariates_to_use=covariates_to_use,\n",
    "                      regressor=regressor, \n",
    "                      outlier_params_ref=outlier_params_ref,\n",
    "                      outlier_params_sec=outlier_params_sec,\n",
    "                      outlier_params_cov=outlier_params_cov,\n",
    "                      regressor_params=regressor_params,  \n",
    "                      atlas_params=atlas_params)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Atlas Construction based on CFOR"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**tagRFPtUtrCH**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "### Predict tagRFPtUtrCH channel for all prims that do not have it\n",
    "\n",
    "# Target directories\n",
    "train_dirpath   = r'data\\experimentA\\image_data'\n",
    "predict_dirpath = r'data\\experimentA\\image_data'\n",
    "\n",
    "# Target IDs\n",
    "train_IDs = parse_from_IDR(train_dirpath, 'tagRFPtUtrCH')\n",
    "print \"Found %i training IDs!\" % len(train_IDs)\n",
    "\n",
    "# Channels\n",
    "ref_channel = ['lynEGFP_seg_LMs_kmeansPRES_pdCFOR_DDDS_CBEmanh',\n",
    "               'lynEGFP_linUnmix_seg_LMs_kmeansPRES_pdCFOR_DDDS_CBEmanh']\n",
    "sec_channel = 'tagRFPtUtrCH_LMs_kmeansPRES_pdCFOR_DDDS_CBEmanh'\n",
    "\n",
    "# Core settings\n",
    "outlier_removal_ref = 'isolation_forest'\n",
    "outlier_removal_sec = 'isolation_forest'\n",
    "outlier_removal_cov = 'percentile_thresh'\n",
    "covariates_to_use   = 'img.cell.'+sec_channel.split('_')[0]+'.mean_total'\n",
    "regressor           = 'MT-ENetCV'\n",
    "\n",
    "# Additional parameters\n",
    "outlier_params_ref = { 'isoforest_params' : {'contamination':0.05}}\n",
    "outlier_params_sec = { 'isoforest_params' : {'contamination':0.05}}\n",
    "outlier_params_cov = { 'bounds'     : 'lower',\n",
    "                       'percentile' : 33 }\n",
    "regressor_params   = { }\n",
    "atlas_params       = { 'zscore_X'       : True,\n",
    "                       'zscore_y'       : True,\n",
    "                       'pca_X'          : True,\n",
    "                       'pca_y'          : True,\n",
    "                       'rezscore_X'     : False,\n",
    "                       'rezscore_y'     : False,\n",
    "                       'subselect_X'    : 20,\n",
    "                       'subselect_y'    : 20,\n",
    "                       'add_covariates' : None }\n",
    "\n",
    "# Additional arguments\n",
    "recurse     = True\n",
    "ignore_self = False\n",
    "processes   = 10\n",
    "profiling   = True\n",
    "verbose     = True\n",
    "\n",
    "# Run prediction pipeline\n",
    "ac.atlas_construction(train_dirpath, predict_dirpath, \n",
    "                      ref_channel, sec_channel, \n",
    "                      train_IDs=train_IDs, predict_IDs=None,\n",
    "                      recurse=recurse, ignore_self=ignore_self, \n",
    "                      processes=processes, profiling=profiling, verbose=verbose,\n",
    "                      outlier_removal_ref=outlier_removal_ref,\n",
    "                      outlier_removal_sec=outlier_removal_sec,\n",
    "                      outlier_removal_cov=outlier_removal_cov,\n",
    "                      covariates_to_use=covariates_to_use,\n",
    "                      regressor=regressor, \n",
    "                      outlier_params_ref=outlier_params_ref,\n",
    "                      outlier_params_sec=outlier_params_sec,\n",
    "                      outlier_params_cov=outlier_params_cov,\n",
    "                      regressor_params=regressor_params,  \n",
    "                      atlas_params=atlas_params)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**NLStdTomato**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "### Predict NLStdTomato channel for all prims that do not have it\n",
    "\n",
    "# Target directories\n",
    "train_dirpath   = r'data\\experimentA\\image_data'\n",
    "predict_dirpath = r'data\\experimentA\\image_data'\n",
    "\n",
    "# Target IDs\n",
    "train_IDs = parse_from_IDR(train_dirpath, 'NLStdTomato')\n",
    "print \"Found %i training IDs!\" % len(train_IDs)\n",
    "\n",
    "# Channels\n",
    "ref_channel = ['lynEGFP_seg_LMs_kmeansPRES_pdCFOR_DDDS_CBEmanh',\n",
    "               'lynEGFP_linUnmix_seg_LMs_kmeansPRES_pdCFOR_DDDS_CBEmanh']\n",
    "sec_channel = 'NLStdTomato_LMs_kmeansPRES_pdCFOR_DDDS_CBEmanh'\n",
    "\n",
    "# Core settings\n",
    "outlier_removal_ref = 'isolation_forest'\n",
    "outlier_removal_sec = 'isolation_forest'\n",
    "outlier_removal_cov = 'percentile_thresh'\n",
    "covariates_to_use   = 'img.cell.'+sec_channel.split('_')[0]+'.mean_total'\n",
    "regressor           = 'MT-ENetCV'\n",
    "\n",
    "# Additional parameters\n",
    "outlier_params_ref = { 'isoforest_params' : {'contamination':0.05}}\n",
    "outlier_params_sec = { 'isoforest_params' : {'contamination':0.05}}\n",
    "outlier_params_cov = { 'bounds'     : 'lower',\n",
    "                       'percentile' : 33 }\n",
    "regressor_params   = { }\n",
    "atlas_params       = { 'zscore_X'       : True,\n",
    "                       'zscore_y'       : True,\n",
    "                       'pca_X'          : True,\n",
    "                       'pca_y'          : True,\n",
    "                       'rezscore_X'     : False,\n",
    "                       'rezscore_y'     : False,\n",
    "                       'subselect_X'    : 20,\n",
    "                       'subselect_y'    : 20,\n",
    "                       'add_covariates' : None }\n",
    "\n",
    "# Additional arguments\n",
    "recurse     = True\n",
    "ignore_self = False\n",
    "processes   = 10\n",
    "profiling   = True\n",
    "verbose     = True\n",
    "\n",
    "# Run prediction pipeline\n",
    "ac.atlas_construction(train_dirpath, predict_dirpath, \n",
    "                      ref_channel, sec_channel, \n",
    "                      train_IDs=train_IDs, predict_IDs=None,\n",
    "                      recurse=recurse, ignore_self=ignore_self, \n",
    "                      processes=processes, profiling=profiling, verbose=verbose,\n",
    "                      outlier_removal_ref=outlier_removal_ref,\n",
    "                      outlier_removal_sec=outlier_removal_sec,\n",
    "                      outlier_removal_cov=outlier_removal_cov,\n",
    "                      covariates_to_use=covariates_to_use,\n",
    "                      regressor=regressor, \n",
    "                      outlier_params_ref=outlier_params_ref,\n",
    "                      outlier_params_sec=outlier_params_sec,\n",
    "                      outlier_params_cov=outlier_params_cov,\n",
    "                      regressor_params=regressor_params,  \n",
    "                      atlas_params=atlas_params)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**b4galT1tagRFPt**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "### Predict b4galT1tagRFPt channel for all prims that do not have it\n",
    "\n",
    "# Target directories\n",
    "train_dirpath   = r'data\\experimentA\\image_data'\n",
    "predict_dirpath = r'data\\experimentA\\image_data'\n",
    "\n",
    "# Target IDs\n",
    "train_IDs = parse_from_IDR(train_dirpath, 'b4galT1tagRFPt')\n",
    "print \"Found %i training IDs!\" % len(train_IDs)\n",
    "\n",
    "# Channels\n",
    "ref_channel = ['lynEGFP_seg_LMs_kmeansPRES_pdCFOR_DDDS_CBEmanh',\n",
    "               'lynEGFP_linUnmix_seg_LMs_kmeansPRES_pdCFOR_DDDS_CBEmanh']\n",
    "sec_channel = 'b4galT1tagRFPt_LMs_kmeansPRES_pdCFOR_DDDS_CBEmanh'\n",
    "\n",
    "# Core settings\n",
    "outlier_removal_ref = 'isolation_forest'\n",
    "outlier_removal_sec = 'isolation_forest'\n",
    "outlier_removal_cov = 'percentile_thresh'\n",
    "covariates_to_use   = 'img.cell.'+sec_channel.split('_')[0]+'.mean_total'\n",
    "regressor           = 'MT-ENetCV'\n",
    "\n",
    "# Additional parameters\n",
    "outlier_params_ref = { 'isoforest_params' : {'contamination':0.05}}\n",
    "outlier_params_sec = { 'isoforest_params' : {'contamination':0.05}}\n",
    "outlier_params_cov = { 'bounds'     : 'lower',\n",
    "                       'percentile' : 33 }\n",
    "regressor_params   = { }\n",
    "atlas_params       = { 'zscore_X'       : True,\n",
    "                       'zscore_y'       : True,\n",
    "                       'pca_X'          : True,\n",
    "                       'pca_y'          : True,\n",
    "                       'rezscore_X'     : False,\n",
    "                       'rezscore_y'     : False,\n",
    "                       'subselect_X'    : 20,\n",
    "                       'subselect_y'    : 20,\n",
    "                       'add_covariates' : None }\n",
    "\n",
    "# Additional arguments\n",
    "recurse     = True\n",
    "ignore_self = False\n",
    "processes   = 10\n",
    "profiling   = True\n",
    "verbose     = True\n",
    "\n",
    "# Run prediction pipeline\n",
    "ac.atlas_construction(train_dirpath, predict_dirpath, \n",
    "                      ref_channel, sec_channel, \n",
    "                      train_IDs=train_IDs, predict_IDs=None,\n",
    "                      recurse=recurse, ignore_self=ignore_self, \n",
    "                      processes=processes, profiling=profiling, verbose=verbose,\n",
    "                      outlier_removal_ref=outlier_removal_ref,\n",
    "                      outlier_removal_sec=outlier_removal_sec,\n",
    "                      outlier_removal_cov=outlier_removal_cov,\n",
    "                      covariates_to_use=covariates_to_use,\n",
    "                      regressor=regressor, \n",
    "                      outlier_params_ref=outlier_params_ref,\n",
    "                      outlier_params_sec=outlier_params_sec,\n",
    "                      outlier_params_cov=outlier_params_cov,\n",
    "                      regressor_params=regressor_params,  \n",
    "                      atlas_params=atlas_params)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**CDMPRtagRFPt**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "### Predict CDMPRtagRFPt channel for all prims that do not have it\n",
    "\n",
    "# Target directories\n",
    "train_dirpath   = r'data\\experimentA\\image_data'\n",
    "predict_dirpath = r'data\\experimentA\\image_data'\n",
    "\n",
    "# Target IDs\n",
    "train_IDs = parse_from_IDR(train_dirpath, 'CDMPRtagRFPt')\n",
    "print \"Found %i training IDs!\" % len(train_IDs)\n",
    "\n",
    "# Channels\n",
    "ref_channel = ['lynEGFP_seg_LMs_kmeansPRES_pdCFOR_DDDS_CBEmanh',\n",
    "               'lynEGFP_linUnmix_seg_LMs_kmeansPRES_pdCFOR_DDDS_CBEmanh']\n",
    "sec_channel = 'CDMPRtagRFPt_LMs_kmeansPRES_pdCFOR_DDDS_CBEmanh'\n",
    "\n",
    "# Core settings\n",
    "outlier_removal_ref = 'isolation_forest'\n",
    "outlier_removal_sec = 'isolation_forest'\n",
    "outlier_removal_cov = 'percentile_thresh'\n",
    "covariates_to_use   = 'img.cell.'+sec_channel.split('_')[0]+'.mean_total'\n",
    "regressor           = 'MT-ENetCV'\n",
    "\n",
    "# Additional parameters\n",
    "outlier_params_ref = { 'isoforest_params' : {'contamination':0.05}}\n",
    "outlier_params_sec = { 'isoforest_params' : {'contamination':0.05}}\n",
    "outlier_params_cov = { 'bounds'     : 'lower',\n",
    "                       'percentile' : 33 }\n",
    "regressor_params   = { }\n",
    "atlas_params       = { 'zscore_X'       : True,\n",
    "                       'zscore_y'       : True,\n",
    "                       'pca_X'          : True,\n",
    "                       'pca_y'          : True,\n",
    "                       'rezscore_X'     : False,\n",
    "                       'rezscore_y'     : False,\n",
    "                       'subselect_X'    : 20,\n",
    "                       'subselect_y'    : 20,\n",
    "                       'add_covariates' : None }\n",
    "\n",
    "# Additional arguments\n",
    "recurse     = True\n",
    "ignore_self = False\n",
    "processes   = 10\n",
    "profiling   = True\n",
    "verbose     = True\n",
    "\n",
    "# Run prediction pipeline\n",
    "ac.atlas_construction(train_dirpath, predict_dirpath, \n",
    "                      ref_channel, sec_channel, \n",
    "                      train_IDs=train_IDs, predict_IDs=None,\n",
    "                      recurse=recurse, ignore_self=ignore_self, \n",
    "                      processes=processes, profiling=profiling, verbose=verbose,\n",
    "                      outlier_removal_ref=outlier_removal_ref,\n",
    "                      outlier_removal_sec=outlier_removal_sec,\n",
    "                      outlier_removal_cov=outlier_removal_cov,\n",
    "                      covariates_to_use=covariates_to_use,\n",
    "                      regressor=regressor, \n",
    "                      outlier_params_ref=outlier_params_ref,\n",
    "                      outlier_params_sec=outlier_params_sec,\n",
    "                      outlier_params_cov=outlier_params_cov,\n",
    "                      regressor_params=regressor_params,  \n",
    "                      atlas_params=atlas_params)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**mKate2GM130**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "### Predict mKate2GM130 channel for all prims that do not have it\n",
    "\n",
    "# Target directories\n",
    "train_dirpath   = r'data\\experimentA\\image_data'\n",
    "predict_dirpath = r'data\\experimentA\\image_data'\n",
    "\n",
    "# Target IDs\n",
    "train_IDs = parse_from_IDR(train_dirpath, 'mKate2GM130')\n",
    "print \"Found %i training IDs!\" % len(train_IDs)\n",
    "\n",
    "# Channels\n",
    "ref_channel = ['lynEGFP_seg_LMs_kmeansPRES_pdCFOR_DDDS_CBEmanh',\n",
    "               'lynEGFP_linUnmix_seg_LMs_kmeansPRES_pdCFOR_DDDS_CBEmanh']\n",
    "sec_channel = 'mKate2GM130_LMs_kmeansPRES_pdCFOR_DDDS_CBEmanh'\n",
    "\n",
    "# Core settings\n",
    "outlier_removal_ref = 'isolation_forest'\n",
    "outlier_removal_sec = 'isolation_forest'\n",
    "outlier_removal_cov = 'percentile_thresh'\n",
    "covariates_to_use   = 'img.cell.'+sec_channel.split('_')[0]+'.mean_total'\n",
    "regressor           = 'MT-ENetCV'\n",
    "\n",
    "# Additional parameters\n",
    "outlier_params_ref = { 'isoforest_params' : {'contamination':0.05}}\n",
    "outlier_params_sec = { 'isoforest_params' : {'contamination':0.05}}\n",
    "outlier_params_cov = { 'bounds'     : 'lower',\n",
    "                       'percentile' : 33 }\n",
    "regressor_params   = { }\n",
    "atlas_params       = { 'zscore_X'       : True,\n",
    "                       'zscore_y'       : True,\n",
    "                       'pca_X'          : True,\n",
    "                       'pca_y'          : True,\n",
    "                       'rezscore_X'     : False,\n",
    "                       'rezscore_y'     : False,\n",
    "                       'subselect_X'    : 20,\n",
    "                       'subselect_y'    : 20,\n",
    "                       'add_covariates' : None }\n",
    "\n",
    "# Additional arguments\n",
    "recurse     = True\n",
    "ignore_self = False\n",
    "processes   = 10\n",
    "profiling   = True\n",
    "verbose     = True\n",
    "\n",
    "# Run prediction pipeline\n",
    "ac.atlas_construction(train_dirpath, predict_dirpath, \n",
    "                      ref_channel, sec_channel, \n",
    "                      train_IDs=train_IDs, predict_IDs=None,\n",
    "                      recurse=recurse, ignore_self=ignore_self, \n",
    "                      processes=processes, profiling=profiling, verbose=verbose,\n",
    "                      outlier_removal_ref=outlier_removal_ref,\n",
    "                      outlier_removal_sec=outlier_removal_sec,\n",
    "                      outlier_removal_cov=outlier_removal_cov,\n",
    "                      covariates_to_use=covariates_to_use,\n",
    "                      regressor=regressor, \n",
    "                      outlier_params_ref=outlier_params_ref,\n",
    "                      outlier_params_sec=outlier_params_sec,\n",
    "                      outlier_params_cov=outlier_params_cov,\n",
    "                      regressor_params=regressor_params,  \n",
    "                      atlas_params=atlas_params)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**lysotrackerdeepred**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "### Predict lysotrackerdeepred channel for all prims that do not have it\n",
    "\n",
    "# Target directories\n",
    "train_dirpath   = r'data\\experimentA\\image_data'\n",
    "predict_dirpath = r'data\\experimentA\\image_data'\n",
    "\n",
    "# Target IDs\n",
    "train_IDs = parse_from_IDR(train_dirpath, 'lysotrackerdeepred')\n",
    "print \"Found %i training IDs!\" % len(train_IDs)\n",
    "\n",
    "# Channels\n",
    "ref_channel = ['lynEGFP_seg_LMs_kmeansPRES_pdCFOR_DDDS_CBEmanh',\n",
    "               'lynEGFP_linUnmix_seg_LMs_kmeansPRES_pdCFOR_DDDS_CBEmanh']\n",
    "sec_channel = 'lysotrackerdeepred_LMs_kmeansPRES_pdCFOR_DDDS_CBEmanh'\n",
    "\n",
    "# Core settings\n",
    "outlier_removal_ref = 'isolation_forest'\n",
    "outlier_removal_sec = 'isolation_forest'\n",
    "outlier_removal_cov = 'percentile_thresh'\n",
    "covariates_to_use   = 'img.cell.'+sec_channel.split('_')[0]+'.mean_total'\n",
    "regressor           = 'MT-ENetCV'\n",
    "\n",
    "# Additional parameters\n",
    "outlier_params_ref = { 'isoforest_params' : {'contamination':0.05}}\n",
    "outlier_params_sec = { 'isoforest_params' : {'contamination':0.05}}\n",
    "outlier_params_cov = { 'bounds'     : 'lower',\n",
    "                       'percentile' : 33 }\n",
    "regressor_params   = { }\n",
    "atlas_params       = { 'zscore_X'       : True,\n",
    "                       'zscore_y'       : True,\n",
    "                       'pca_X'          : True,\n",
    "                       'pca_y'          : True,\n",
    "                       'rezscore_X'     : False,\n",
    "                       'rezscore_y'     : False,\n",
    "                       'subselect_X'    : 20,\n",
    "                       'subselect_y'    : 20,\n",
    "                       'add_covariates' : None }\n",
    "\n",
    "# Additional arguments\n",
    "recurse     = True\n",
    "ignore_self = False\n",
    "processes   = 10\n",
    "profiling   = True\n",
    "verbose     = True\n",
    "\n",
    "# Run prediction pipeline\n",
    "ac.atlas_construction(train_dirpath, predict_dirpath, \n",
    "                      ref_channel, sec_channel, \n",
    "                      train_IDs=train_IDs, predict_IDs=None,\n",
    "                      recurse=recurse, ignore_self=ignore_self, \n",
    "                      processes=processes, profiling=profiling, verbose=verbose,\n",
    "                      outlier_removal_ref=outlier_removal_ref,\n",
    "                      outlier_removal_sec=outlier_removal_sec,\n",
    "                      outlier_removal_cov=outlier_removal_cov,\n",
    "                      covariates_to_use=covariates_to_use,\n",
    "                      regressor=regressor, \n",
    "                      outlier_params_ref=outlier_params_ref,\n",
    "                      outlier_params_sec=outlier_params_sec,\n",
    "                      outlier_params_cov=outlier_params_cov,\n",
    "                      regressor_params=regressor_params,  \n",
    "                      atlas_params=atlas_params)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**mKate2rab5**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "### Predict mKate2rab5 channel for all prims that do not have it\n",
    "\n",
    "# Target directories\n",
    "train_dirpath   = r'data\\experimentA\\image_data'\n",
    "predict_dirpath = r'data\\experimentA\\image_data'\n",
    "\n",
    "# Target IDs\n",
    "train_IDs = parse_from_IDR(train_dirpath, 'mKate2rab5')\n",
    "print \"Found %i training IDs!\" % len(train_IDs)\n",
    "\n",
    "# Channels\n",
    "ref_channel = ['lynEGFP_seg_LMs_kmeansPRES_pdCFOR_DDDS_CBEmanh',\n",
    "               'lynEGFP_linUnmix_seg_LMs_kmeansPRES_pdCFOR_DDDS_CBEmanh']\n",
    "sec_channel = 'mKate2rab5_LMs_kmeansPRES_pdCFOR_DDDS_CBEmanh'\n",
    "\n",
    "# Core settings\n",
    "outlier_removal_ref = 'isolation_forest'\n",
    "outlier_removal_sec = 'isolation_forest'\n",
    "outlier_removal_cov = 'percentile_thresh'\n",
    "covariates_to_use   = 'img.cell.'+sec_channel.split('_')[0]+'.mean_total'\n",
    "regressor           = 'MT-ENetCV'\n",
    "\n",
    "# Additional parameters\n",
    "outlier_params_ref = { 'isoforest_params' : {'contamination':0.05}}\n",
    "outlier_params_sec = { 'isoforest_params' : {'contamination':0.05}}\n",
    "outlier_params_cov = { 'bounds'     : 'lower',\n",
    "                       'percentile' : 33 }\n",
    "regressor_params   = { }\n",
    "atlas_params       = { 'zscore_X'       : True,\n",
    "                       'zscore_y'       : True,\n",
    "                       'pca_X'          : True,\n",
    "                       'pca_y'          : True,\n",
    "                       'rezscore_X'     : False,\n",
    "                       'rezscore_y'     : False,\n",
    "                       'subselect_X'    : 20,\n",
    "                       'subselect_y'    : 20,\n",
    "                       'add_covariates' : None }\n",
    "\n",
    "# Additional arguments\n",
    "recurse     = True\n",
    "ignore_self = False\n",
    "processes   = 10\n",
    "profiling   = True\n",
    "verbose     = True\n",
    "\n",
    "# Run prediction pipeline\n",
    "ac.atlas_construction(train_dirpath, predict_dirpath, \n",
    "                      ref_channel, sec_channel, \n",
    "                      train_IDs=train_IDs, predict_IDs=None,\n",
    "                      recurse=recurse, ignore_self=ignore_self, \n",
    "                      processes=processes, profiling=profiling, verbose=verbose,\n",
    "                      outlier_removal_ref=outlier_removal_ref,\n",
    "                      outlier_removal_sec=outlier_removal_sec,\n",
    "                      outlier_removal_cov=outlier_removal_cov,\n",
    "                      covariates_to_use=covariates_to_use,\n",
    "                      regressor=regressor, \n",
    "                      outlier_params_ref=outlier_params_ref,\n",
    "                      outlier_params_sec=outlier_params_sec,\n",
    "                      outlier_params_cov=outlier_params_cov,\n",
    "                      regressor_params=regressor_params,  \n",
    "                      atlas_params=atlas_params)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**mKate2rab11**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "### Predict mKate2rab11 channel for all prims that do not have it\n",
    "\n",
    "# Target directories\n",
    "train_dirpath   = r'data\\experimentA\\image_data'\n",
    "predict_dirpath = r'data\\experimentA\\image_data'\n",
    "\n",
    "# Target IDs\n",
    "train_IDs = parse_from_IDR(train_dirpath, 'mKate2rab11')\n",
    "print \"Found %i training IDs!\" % len(train_IDs)\n",
    "\n",
    "# Channels\n",
    "ref_channel = ['lynEGFP_seg_LMs_kmeansPRES_pdCFOR_DDDS_CBEmanh',\n",
    "               'lynEGFP_linUnmix_seg_LMs_kmeansPRES_pdCFOR_DDDS_CBEmanh']\n",
    "sec_channel = 'mKate2rab11_LMs_kmeansPRES_pdCFOR_DDDS_CBEmanh'\n",
    "\n",
    "# Core settings\n",
    "outlier_removal_ref = 'isolation_forest'\n",
    "outlier_removal_sec = 'isolation_forest'\n",
    "outlier_removal_cov = 'percentile_thresh'\n",
    "covariates_to_use   = 'img.cell.'+sec_channel.split('_')[0]+'.mean_total'\n",
    "regressor           = 'MT-ENetCV'\n",
    "\n",
    "# Additional parameters\n",
    "outlier_params_ref = { 'isoforest_params' : {'contamination':0.05}}\n",
    "outlier_params_sec = { 'isoforest_params' : {'contamination':0.05}}\n",
    "outlier_params_cov = { 'bounds'     : 'lower',\n",
    "                       'percentile' : 33 }\n",
    "regressor_params   = { }\n",
    "atlas_params       = { 'zscore_X'       : True,\n",
    "                       'zscore_y'       : True,\n",
    "                       'pca_X'          : True,\n",
    "                       'pca_y'          : True,\n",
    "                       'rezscore_X'     : False,\n",
    "                       'rezscore_y'     : False,\n",
    "                       'subselect_X'    : 20,\n",
    "                       'subselect_y'    : 20,\n",
    "                       'add_covariates' : None }\n",
    "\n",
    "# Additional arguments\n",
    "recurse     = True\n",
    "ignore_self = False\n",
    "processes   = 10\n",
    "profiling   = True\n",
    "verbose     = True\n",
    "\n",
    "# Run prediction pipeline\n",
    "ac.atlas_construction(train_dirpath, predict_dirpath, \n",
    "                      ref_channel, sec_channel, \n",
    "                      train_IDs=train_IDs, predict_IDs=None,\n",
    "                      recurse=recurse, ignore_self=ignore_self, \n",
    "                      processes=processes, profiling=profiling, verbose=verbose,\n",
    "                      outlier_removal_ref=outlier_removal_ref,\n",
    "                      outlier_removal_sec=outlier_removal_sec,\n",
    "                      outlier_removal_cov=outlier_removal_cov,\n",
    "                      covariates_to_use=covariates_to_use,\n",
    "                      regressor=regressor, \n",
    "                      outlier_params_ref=outlier_params_ref,\n",
    "                      outlier_params_sec=outlier_params_sec,\n",
    "                      outlier_params_cov=outlier_params_cov,\n",
    "                      regressor_params=regressor_params,  \n",
    "                      atlas_params=atlas_params)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "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
}