wp_all_cor_cd_dcm.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Author: Davide Aloi - PhD student - University of Birmingham\n",
    "# Description: the script correlates current density values within M1 and Th for each \n",
    "# participant with their respective DCM values.\n",
    "\n",
    "# Imports\n",
    "import warnings\n",
    "warnings.filterwarnings(\"ignore\")\n",
    "import os\n",
    "import glob\n",
    "import numpy as np\n",
    "from nilearn import image, plotting\n",
    "from nilearn.image import new_img_like\n",
    "from scipy import ndimage\n",
    "import scipy.io\n",
    "import time\n",
    "from tqdm import tqdm\n",
    "\n",
    "# For each voxel-wise analysis I will store the peak fdr-corrected p-value to then perform\n",
    "# another fdr correction \n",
    "fdr_pvals = []"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "# FDR adjusted P value \n",
    "def fdr(p_vals):\n",
    "\n",
    "    from scipy.stats import rankdata\n",
    "    ranked_p_values = rankdata(p_vals)\n",
    "    fdr = p_vals * len(p_vals) / ranked_p_values\n",
    "    fdr[fdr > 1] = 1\n",
    "\n",
    "    return fdr\n",
    "\n",
    "# From matrix space to MNI space\n",
    "def matrix_to_mni(matrix_coord, T):\n",
    "    second_arg = np.array([matrix_coord[0],matrix_coord[1],matrix_coord[2], 1])\n",
    "    second_arg = np.reshape(second_arg,[-1,1])\n",
    "    mni_coord = np.dot(T,second_arg)\n",
    "\n",
    "    return np.reshape(mni_coord[0:3],[1,-1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "## Parameters and variables: \n",
    "results_folder = 'D:\\\\roast-chapter3\\\\wp_all_results\\\\' # Folder with results\n",
    "main_folder = 'C:\\\\Users\\\\davide\\\\Documents\\\\GitHub\\\\wp1_2_roast\\\\' # Project folder\n",
    "\n",
    "# Datasets names and subjects lists\n",
    "db_names = ['wp2a', 'wp2a','wp2a', 'wp1a', 'wp1b']\n",
    "\n",
    "# 08/09/2022 wp2a day1pre vs day5 post added\n",
    "# dataset names for the dcm results (pairwise int)\n",
    "db_names_dcm = ['wp2a_day1_pairwise','wp2a_day5_pairwise', 'wp2a_day1pre_VSday5post_pairwise', 'wp1a_pairwise', 'wp1b_pairwise']\n",
    "dcm_results_folder = 'C:\\\\Users\\\\davide\\\\Documents\\\\GitHub\\\\wp1_2_roast\\\\all_dcm_results\\\\' # Folder with dcm results\n",
    "\n",
    "db_subjects = [['01','02','03','04','06','07','08','09','10','11','12','13','14','15','16','17','18','19','20','22','23','24'], # Wp2a\n",
    "               ['03','04','05','07','09','10','11','12','13','15','16','17','18','19','20','21','22','23','24','25','26'], # Wp1a\n",
    "               ['01','02','03','04','05','06','07','08','09','10','11','12','13','15','16','17','18','19','21','22','23']] # Wp1b                              \n",
    "\n",
    "## Loading AAL3 atlas and extracting M1 / Thalamus ROIs (regions of interest)\n",
    "# AAL3 atlas paper: https://www.oxcns.org/papers/607%20Rolls%20Huang%20Lin%20Feng%20Joliot%202020%20AAL3.pdf \n",
    "AAl3_path = os.path.join(main_folder, 'rois', 'AAL3v1_1mm.nii')\n",
    "AAL3_atlas = image.load_img(AAl3_path)\n",
    "\n",
    "## Creating M1, Th and cerebellar masks from the AAL3 atlas. Load MNI template.\n",
    "# AAL3 index for left M1 = 1\n",
    "m1 = image.math_img(\"np.where(img == 1, 1, 0)\", img = AAL3_atlas) \n",
    "# AAL3 index for TH = 121 - 149 (odd values only (left thalamus)) --> I'm not convinced about this one, ask Davinia\n",
    "th = image.math_img(\"np.where(np.isin(img, np.arange(121, 150, 2)), 1, 0)\", img = AAL3_atlas)\n",
    "# AAL3 index for right Cerebellum (cb) (102 and 108: cerebellar lobes IV-V and VIII)\n",
    "cb = image.math_img(\"np.where(np.isin(img, np.array([102, 108])), 1, 0)\", img = AAL3_atlas)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Analysing dataset wp2a - N. subjects: 22\n",
      "Correlating  m1  with connection:  m1m1\n",
      "Number of significant voxels (unc):  542\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  0\n",
      "Correlating  m1  with connection:  m1th\n",
      "Number of significant voxels (unc):  585\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  0\n",
      "Correlating  m1  with connection:  thm1\n",
      "Number of significant voxels (unc):  2121\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  0\n",
      "Correlating  m1  with connection:  thth\n",
      "Number of significant voxels (unc):  1156\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  0\n",
      "Correlating  th  with connection:  m1m1\n",
      "Number of significant voxels (unc):  5828\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  5530\n",
      "peak r\n",
      "-0.7288484047532908\n",
      "p-val 0.019821192080968005\n",
      "coords\n",
      "[[-7. -9. -4.]]\n",
      "Single subject median in the roi vs DCM values\n",
      "r -0.5732470893494526\n",
      "p 0.005288327957003156\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1152x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Correlating  th  with connection:  m1th\n",
      "Number of significant voxels (unc):  12\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  0\n",
      "Correlating  th  with connection:  thm1\n",
      "Number of significant voxels (unc):  3473\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  0\n",
      "Correlating  th  with connection:  thth\n",
      "Number of significant voxels (unc):  2\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  0\n",
      "Correlating  cb  with connection:  m1m1\n",
      "Number of significant voxels (unc):  2286\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  0\n",
      "Correlating  cb  with connection:  m1th\n",
      "Number of significant voxels (unc):  774\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  0\n",
      "Correlating  cb  with connection:  thm1\n",
      "Number of significant voxels (unc):  1929\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  0\n",
      "Correlating  cb  with connection:  thth\n",
      "Number of significant voxels (unc):  2318\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  0\n",
      "\n",
      "Analysing dataset wp2a - N. subjects: 22\n",
      "Correlating  m1  with connection:  m1m1\n",
      "Number of significant voxels (unc):  697\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  0\n",
      "Correlating  m1  with connection:  m1th\n",
      "Number of significant voxels (unc):  2009\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  0\n",
      "Correlating  m1  with connection:  thm1\n",
      "Number of significant voxels (unc):  789\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  0\n",
      "Correlating  m1  with connection:  thth\n",
      "Number of significant voxels (unc):  1063\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  0\n",
      "Correlating  th  with connection:  m1m1\n",
      "Number of significant voxels (unc):  1042\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  0\n",
      "Correlating  th  with connection:  m1th\n",
      "Number of significant voxels (unc):  32\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  0\n",
      "Correlating  th  with connection:  thm1\n",
      "Number of significant voxels (unc):  14\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  0\n",
      "Correlating  th  with connection:  thth\n",
      "Number of significant voxels (unc):  0\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  0\n",
      "Correlating  cb  with connection:  m1m1\n",
      "Number of significant voxels (unc):  495\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  0\n",
      "Correlating  cb  with connection:  m1th\n",
      "Number of significant voxels (unc):  425\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  0\n",
      "Correlating  cb  with connection:  thm1\n",
      "Number of significant voxels (unc):  1369\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  0\n",
      "Correlating  cb  with connection:  thth\n",
      "Number of significant voxels (unc):  606\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  0\n",
      "\n",
      "Analysing dataset wp2a - N. subjects: 22\n",
      "Correlating  m1  with connection:  m1m1\n",
      "Number of significant voxels (unc):  1456\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  0\n",
      "Correlating  m1  with connection:  m1th\n",
      "Number of significant voxels (unc):  321\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  0\n",
      "Correlating  m1  with connection:  thm1\n",
      "Number of significant voxels (unc):  677\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  0\n",
      "Correlating  m1  with connection:  thth\n",
      "Number of significant voxels (unc):  2281\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  0\n",
      "Correlating  th  with connection:  m1m1\n",
      "Number of significant voxels (unc):  0\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  0\n",
      "Correlating  th  with connection:  m1th\n",
      "Number of significant voxels (unc):  51\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  0\n",
      "Correlating  th  with connection:  thm1\n",
      "Number of significant voxels (unc):  0\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  0\n",
      "Correlating  th  with connection:  thth\n",
      "Number of significant voxels (unc):  2249\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  0\n",
      "Correlating  cb  with connection:  m1m1\n",
      "Number of significant voxels (unc):  561\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  0\n",
      "Correlating  cb  with connection:  m1th\n",
      "Number of significant voxels (unc):  2627\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  62\n",
      "peak r\n",
      "0.8369390965645643\n",
      "p-val 0.011384309158815982\n",
      "coords\n",
      "[[ 45. -49. -56.]]\n",
      "Single subject median in the roi vs DCM values\n",
      "r -0.19930318124507868\n",
      "p 0.37388681440864185\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1152x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Correlating  cb  with connection:  thm1\n",
      "Number of significant voxels (unc):  1426\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  0\n",
      "Correlating  cb  with connection:  thth\n",
      "Number of significant voxels (unc):  2533\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  0\n",
      "\n",
      "Analysing dataset wp1a - N. subjects: 21\n",
      "Correlating  m1  with connection:  m1m1\n",
      "Number of significant voxels (unc):  2220\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  0\n",
      "Correlating  m1  with connection:  m1th\n",
      "Number of significant voxels (unc):  1621\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  0\n",
      "Correlating  m1  with connection:  thm1\n",
      "Number of significant voxels (unc):  2391\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  0\n",
      "Correlating  m1  with connection:  thth\n",
      "Number of significant voxels (unc):  2766\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  0\n",
      "Correlating  th  with connection:  m1m1\n",
      "Number of significant voxels (unc):  213\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  0\n",
      "Correlating  th  with connection:  m1th\n",
      "Number of significant voxels (unc):  0\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  0\n",
      "Correlating  th  with connection:  thm1\n",
      "Number of significant voxels (unc):  306\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  0\n",
      "Correlating  th  with connection:  thth\n",
      "Number of significant voxels (unc):  721\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  0\n",
      "Correlating  cb  with connection:  m1m1\n",
      "Number of significant voxels (unc):  369\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  0\n",
      "Correlating  cb  with connection:  m1th\n",
      "Number of significant voxels (unc):  289\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  0\n",
      "Correlating  cb  with connection:  thm1\n",
      "Number of significant voxels (unc):  29\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  0\n",
      "Correlating  cb  with connection:  thth\n",
      "Number of significant voxels (unc):  229\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  0\n",
      "\n",
      "Analysing dataset wp1b - N. subjects: 21\n",
      "Correlating  m1  with connection:  m1m1\n",
      "Number of significant voxels (unc):  2681\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  0\n",
      "Correlating  m1  with connection:  m1th\n",
      "Number of significant voxels (unc):  611\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  0\n",
      "Correlating  m1  with connection:  thm1\n",
      "Number of significant voxels (unc):  3184\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  0\n",
      "Correlating  m1  with connection:  thth\n",
      "Number of significant voxels (unc):  605\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  0\n",
      "Correlating  th  with connection:  m1m1\n",
      "Number of significant voxels (unc):  6626\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  6373\n",
      "peak r\n",
      "-0.6534618145673995\n",
      "p-val 0.03099691424215644\n",
      "coords\n",
      "[[-12. -24.   0.]]\n",
      "Single subject median in the roi vs DCM values\n",
      "r -0.5537138992227195\n",
      "p 0.009209433238651296\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1152x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Correlating  th  with connection:  m1th\n",
      "Number of significant voxels (unc):  0\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  0\n",
      "Correlating  th  with connection:  thm1\n",
      "Number of significant voxels (unc):  937\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  0\n",
      "Correlating  th  with connection:  thth\n",
      "Number of significant voxels (unc):  127\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  0\n",
      "Correlating  cb  with connection:  m1m1\n",
      "Number of significant voxels (unc):  13041\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  11082\n",
      "peak r\n",
      "-0.8053144472645702\n",
      "p-val 0.016090082358461346\n",
      "coords\n",
      "[[ 39. -40. -53.]]\n",
      "Single subject median in the roi vs DCM values\n",
      "r -0.6345521035806387\n",
      "p 0.0020029362803773935\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1152x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Correlating  cb  with connection:  m1th\n",
      "Number of significant voxels (unc):  106\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  0\n",
      "Correlating  cb  with connection:  thm1\n",
      "Number of significant voxels (unc):  617\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  0\n",
      "Correlating  cb  with connection:  thth\n",
      "Number of significant voxels (unc):  410\n",
      "Number of significant voxels (fdr):  0\n",
      "Number of significant voxels (fdr statsmodels):  0\n",
      "Number of significant voxels (fdr statsmodels no nans):  0\n"
     ]
    }
   ],
   "source": [
    "## Iterate all three datasets\n",
    "from scipy import stats\n",
    "from statsmodels.stats import multitest as mt\n",
    "import scipy.stats as st\n",
    "from nilearn import glm\n",
    "\n",
    "for db_id, db in enumerate(db_names):\n",
    "    ## Loading results for each dataset (current density maps)\n",
    "    cd_maps = image.load_img(os.path.join(results_folder, db + '_all_cd_maps.nii'))\n",
    "    print('\\nAnalysing dataset ' + db + ' - N. subjects: ' + str(cd_maps.shape[3]))\n",
    "\n",
    "    ## Loading DCM results (nb: the unthresholded ones, for the correlation analysis)\n",
    "    DCM = np.load(os.path.join(dcm_results_folder, db_names_dcm[db_id] + '_dcm_unthresholded.npy'))\n",
    "\n",
    "    ## Conversion of diagonal values to Hz + recentering on 0\n",
    "    funcHz = lambda x: (-0.5*(np.exp(x))) + 0.5 if x != 0 else 0\n",
    "\n",
    "    # Positive values now indicate less self inhibition\n",
    "    # Negative values indicate more self inhibition\n",
    "    for pos in range(0, DCM.shape[0]): \n",
    "        diag = np.diagonal(DCM[pos,:,:])\n",
    "        diag_converted = []\n",
    "        for element in diag:\n",
    "            diag_converted.append(funcHz(element))\n",
    "        np.fill_diagonal(DCM[pos,:,:], np.array(diag_converted))\n",
    "    \n",
    "    ## Resampling masks (masks and current density maps need to have the same shape)\n",
    "    m1_resampled = image.resample_to_img(m1, cd_maps, interpolation = 'nearest').get_fdata()\n",
    "    th_resampled = image.resample_to_img(th, cd_maps, interpolation = 'nearest').get_fdata()\n",
    "    cb_resampled = image.resample_to_img(cb, cd_maps, interpolation = 'nearest').get_fdata()\n",
    "    \n",
    "    # List of Rois and connections \n",
    "    rois = [['m1', m1_resampled],\n",
    "            ['th', th_resampled],\n",
    "            ['cb', cb_resampled]]\n",
    "            \n",
    "    cons = [['m1m1', DCM[:,0,0]],\n",
    "            ['m1th', DCM[:,1,0]],\n",
    "            ['thm1', DCM[:,0,1]],\n",
    "            ['thth', DCM[:,1,1]]]\n",
    "\n",
    "    cd_maps_a = cd_maps.get_fdata() # current density data in numpy format (easier to access)\n",
    "    shape = cd_maps.shape \n",
    "    df = shape[3]-1 # degrees of freedom (N-1). Used for the calculation of T-maps.\n",
    "\n",
    "    # For each dataset, we will iterate each ROI, and for each ROI, each connection (4 connections)\n",
    "    for roi_name, roi in rois:\n",
    "         for con_name, con in cons:\n",
    "            print('Correlating ', roi_name, ' with connection: ', con_name)\n",
    "\n",
    "            # Arrays where we'll save our results (We have 1 result per con/roi/db)\n",
    "            cormap_r = np.zeros(m1_resampled.shape) # r\n",
    "            cormap_p = np.zeros(m1_resampled.shape) # p (uncorrected)\n",
    "            \n",
    "            previous_p = []\n",
    "            # Let's now iterate each voxel in the roi (but consider only those == 1)\n",
    "            for i, v in enumerate(roi.flatten()):\n",
    "                if v != 0: # if voxel is not masked in the ROI\n",
    "                    # Extract the 3d coordinates from 1d index\n",
    "                    idx, idy, idz = np.unravel_index(i, (shape[0],shape[1],shape[2]))\n",
    "                    p_v = cd_maps_a[idx,idy,idz,:] # value for that voxel for all participants (1xN array)\n",
    "                    # calculate r and p and store it in the right voxel\n",
    "                    \n",
    "                    cormap_r[idx,idy,idz], cormap_p[idx,idy,idz] = stats.pearsonr(p_v, con)\n",
    "                    # this is just a check I was doing but you can ignore it \n",
    "                    if previous_p == cormap_p[idx,idy,idz]:\n",
    "                        crash()\n",
    "                    else:\n",
    "                        previous_p = cormap_p[idx,idy,idz]\n",
    "\n",
    "            zvals = st.norm.ppf(cormap_p) # converting pval map to z scores\n",
    "            zvals = new_img_like(cd_maps, zvals) # saving z-scores map into nilearn object\n",
    "            \n",
    "            # First approach (using glm.threshold_stats_img from nilearn)\n",
    "            fdr_p = glm.threshold_stats_img(zvals, alpha = 0.05) # calculating fdr map + threshold using nilearn\n",
    "           \n",
    "            fdr_p = fdr_p[0].get_fdata() # from nilearn to numpy array \n",
    "            \n",
    "            # Second approach using statsmodels (stats.multimodels.fdrcorrection)\n",
    "            cormap_p_nans = np.where(roi == 0, np.nan, cormap_p) # Assigning nans to values outside ROI\n",
    "            fdr_p_2 = mt.multipletests(cormap_p_nans.flatten(), alpha=0.05, method='fdr_bh', is_sorted=False)\n",
    "            \n",
    "            # Third approach (this is very slow I'm sure there's a quicker way)\n",
    "            # like above but I'm using a vector with only the voxels from each roi (that is\n",
    "            # without nans). I do this to cross check results from approach 1 and 2.\n",
    "            \n",
    "            # filtering out nans\n",
    "            cormap_p_nonans = cormap_p_nans[~np.isnan(cormap_p_nans)].flatten()  # this removes NaNs\n",
    "            fdr_p_3 = mt.multipletests(cormap_p_nonans, alpha=0.05, method='fdr_bh', is_sorted=False)\n",
    "\n",
    "            # Printing results \n",
    "            print('Number of significant voxels (unc): ', ((cormap_p < 0.05) & (cormap_p > 0)).sum())\n",
    "            print('Number of significant voxels (fdr): ', ((fdr_p < 0.05) & (fdr_p > 0)).sum())\n",
    "\n",
    "            # printing results with the second approach\n",
    "            print('Number of significant voxels (fdr statsmodels): ', fdr_p_2[0].sum())\n",
    "\n",
    "            # The right one is the one below: this excludes nans from the FDR calculation, \n",
    "            # and some correlations are actually significant. Just a few\n",
    "            print('Number of significant voxels (fdr statsmodels no nans): ', fdr_p_3[0].sum())\n",
    "\n",
    "            # if more than 0 voxels survive FDR then save everything into nifti files\n",
    "            \n",
    "            fdr_p_map = np.zeros(m1_resampled.flatten().shape)\n",
    "\n",
    "            # We need to iterate the roi, when we find a value in the roi that is = 1, we take the p from the fdr corrected\n",
    "            # p values list and we increment pos_in_array to move to the next p.\n",
    "            pos_in_array = 0\n",
    "            for i, v in enumerate(roi.flatten()):\n",
    "                if v != 0:\n",
    "                    fdr_p_map[i] = fdr_p_3[1][pos_in_array]\n",
    "                    pos_in_array+=1\n",
    "\n",
    "            # I'll save a map with only the significant p values (only for rois with\n",
    "            # statistically significant correlations between current density and dcm)\n",
    "            fdr_p_map_thresholded = np.where(fdr_p_map < 0.05, fdr_p_map, 0)  \n",
    "            fdr_p_map_thresholded = np.reshape(fdr_p_map_thresholded, m1_resampled.shape)\n",
    "\n",
    "            # r maps\n",
    "            r_map = new_img_like(cd_maps, cormap_r).get_fdata()\n",
    "            r_thresholded = np.where(new_img_like(cd_maps, fdr_p_map_thresholded).get_fdata() > 0, r_map, 0)\n",
    "            \n",
    "            # Peak r and coordinates \n",
    "            peak_r_arg = np.abs(r_thresholded).argmax()  \n",
    "            peak_r_pos = np.unravel_index(peak_r_arg,r_thresholded.shape)\n",
    "            peak_r = r_thresholded[peak_r_pos]\n",
    "            peak_p = fdr_p_map_thresholded[peak_r_pos]\n",
    "            \n",
    "            # I will store the peak p value for future fdr correction\n",
    "            fdr_pvals.append(np.min(fdr_p_3[1]))\n",
    "                \n",
    "                # Let's save a map with the adjusted p values (FDR statsmodels, no nans = last approach)\n",
    "                # this isn't very straightforward because the array without nans has a different dimension than \n",
    "                # the original scan (as we took only the values inside the ROI to avoid passing NaNs to the fdr function) \n",
    "            if fdr_p_3[0].sum() > 0:\n",
    "                new_img_like(cd_maps, cormap_p).to_filename(db + '_' + roi_name + '_' + con_name + '_p_uncr.nii')\n",
    "                #new_img_like(cd_maps, fdr_p_map).to_filename(db_names_dcm[db_id] + '_' + roi_name + '_' + con_name + '_p_corFDR.nii')\n",
    "                new_img_like(cd_maps, fdr_p_map_thresholded).to_filename(db_names_dcm[db_id] + '_' + roi_name + '_' + con_name + '_p_corFDR_thresholded.nii')\n",
    "                new_img_like(cd_maps, r_thresholded).to_filename(db_names_dcm[db_id] + '_' + roi_name + '_' + con_name + '_r.nii')\n",
    "\n",
    "                mni_coords = matrix_to_mni(peak_r_pos, new_img_like(cd_maps, cormap_p).affine)\n",
    "                print('peak r')\n",
    "                print(peak_r)\n",
    "                print('p-val ' + str(peak_p))\n",
    "                print('coords')\n",
    "                print(mni_coords)\n",
    "\n",
    "                # I also want to plot correlation between the median of the current density\n",
    "                # for the roi that resulted significant and dcm value\n",
    "                #masking current density for each subject\n",
    "                import matplotlib.pyplot as plt \n",
    "                cd_maps_a_masked = cd_maps_a.copy()\n",
    "                all_roi_medians = []\n",
    "                for i in range(0, cd_maps_a.shape[3]):\n",
    "                    cd_maps_a_masked[:,:,:,i] = cd_maps_a[:,:,:,i] * roi\n",
    "                    cd_maps_a_masked[:,:,:,i] = np.where(cd_maps_a_masked[:,:,:,i] == 0, np.nan, cd_maps_a_masked[:,:,:,i])\n",
    "                    all_roi_medians.append(np.nanmedian(cd_maps_a_masked[:,:,:,i]))\n",
    "\n",
    "                # printing r and p\n",
    "                r1,p1  = scipy.stats.pearsonr(np.asarray(all_roi_medians), con)\n",
    "                print('Single subject median in the roi vs DCM values')\n",
    "                print('r ' +str(r1))\n",
    "                print('p ' +str(p1))\n",
    "                from numpy.polynomial.polynomial import polyfit\n",
    "                x, y = np.asarray(all_roi_medians), con\n",
    "                b, m = polyfit(x, y, 1)\n",
    "                plt.rcParams['font.family'] = 'sans-serif'\n",
    "                plt.rcParams['figure.figsize'] = (16,8)\n",
    "                plt.rcParams['figure.facecolor'] = 'white'\n",
    "                plt.scatter(x, y, edgecolors= \"black\")\n",
    "                plt.xlabel('Current density median ' + roi_name)\n",
    "                plt.ylabel('DCM ' + con_name)\n",
    "                plt.plot(x, b + m * x, '--',color='black', linewidth=0.8, alpha = 0.7)\n",
    "                plt.tick_params(axis='both', which='major', labelsize=11)\n",
    "\n",
    "                text = 'r = ' + str(round(r1, 5)) + ' \\np = ' + str(round(p1,4))\n",
    "\n",
    "                #plt.savefig(db + 'cor.png', dpi=300)\n",
    "                plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "correction_multiple_comp = mt.multipletests(fdr_pvals, alpha=0.05, method='fdr_bh', is_sorted=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([False, False, False, False, False, False, False, False, False,\n",
       "        False, False, False, False, False, False, False, False, False,\n",
       "        False, False, False, False, False, False, False, False, False,\n",
       "        False, False, False, False, False, False, False, False, False,\n",
       "        False, False, False, False, False, False, False, False, False,\n",
       "        False, False, False, False, False, False, False, False, False,\n",
       "        False, False, False, False, False, False]),\n",
       " array([0.99999942, 0.99999942, 0.99999942, 0.99999942, 0.39642384,\n",
       "        0.99999942, 0.99999942, 0.99999942, 0.99999942, 0.99999942,\n",
       "        0.99999942, 0.99999942, 0.99999942, 0.99999942, 0.99999942,\n",
       "        0.99999942, 0.99999942, 0.99999942, 0.99999942, 0.99999942,\n",
       "        0.99999942, 0.99999942, 0.99999942, 0.99999942, 0.99999942,\n",
       "        0.99999942, 0.99999942, 0.99999942, 0.99999942, 0.99999942,\n",
       "        0.99999942, 0.99999942, 0.99999942, 0.39642384, 0.99999942,\n",
       "        0.99999942, 0.99999942, 0.99999942, 0.99999942, 0.99999942,\n",
       "        0.99999942, 0.99999942, 0.99999942, 0.99999942, 0.99999942,\n",
       "        0.99999942, 0.99999942, 0.99999942, 0.99999942, 0.99999942,\n",
       "        0.99999942, 0.99999942, 0.46495371, 0.99999942, 0.99999942,\n",
       "        0.99999942, 0.39642384, 0.99999942, 0.99999942, 0.99999942]),\n",
       " 0.0008545229269492083,\n",
       " 0.0008333333333333334)"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "correction_multiple_comp"
   ]
  }
 ],
 "metadata": {
  "interpreter": {
   "hash": "1a04e8d3cf57eb13e3e424f0af4edd4725046eee0d61fcbf258511c525184dd3"
  },
  "kernelspec": {
   "display_name": "Python 3.9.7 ('neuroimg')",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.12"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}