downstream_analysis_2.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "9ef103dc-c1a2-4668-84c0-6b4657307121",
   "metadata": {},
   "source": [
    "### Downstream Analysis 2\n",
    "In this notebook, we will focus on the downstream analysis from a more macro scale. We will combine all CRE (regions) and look at all of them together. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "8dd90f05-f3af-4c26-bb0d-af8f31d83a78",
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "from sklearn.decomposition import PCA\n",
    "import pickle\n",
    "from scipy.stats import spearmanr, pearsonr\n",
    "import warnings\n",
    "from tqdm import tqdm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "44088849-7ba2-4c83-a4b3-ada4a4790bd0",
   "metadata": {},
   "outputs": [],
   "source": [
    "def array_stats(arr):\n",
    "    arr_min = np.min(arr)\n",
    "    arr_max = np.max(arr)\n",
    "    arr_mean = np.mean(arr)\n",
    "    arr_median = np.median(arr)\n",
    "    arr_std = np.std(arr)\n",
    "    arr_variance = np.var(arr)\n",
    "    arr_length = len(arr)\n",
    "    print(\"Minimum: \", arr_min)\n",
    "    print(\"Maximum: \", arr_max)\n",
    "    print(\"Mean: \", arr_mean)\n",
    "    print(\"Median: \", arr_median)\n",
    "    print(\"Standard Deviation: \", arr_std)\n",
    "    print(\"Variance: \", arr_variance)\n",
    "    print(\"Length: \", arr_length)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "fe94cb52-d67e-4f95-830b-5a8c6479c8dd",
   "metadata": {},
   "outputs": [],
   "source": [
    "colors = {'0': '#000000',\n",
    "             '1': '#010067',\n",
    "             '2': '#D5FF00',\n",
    "             '3': '#FF0056',\n",
    "             '4': '#9E008E',\n",
    "             '5': '#0E4CA1',\n",
    "             '6': '#FFE502',\n",
    "             '7': '#005F39',\n",
    "             '8': '#00FF00',\n",
    "             '9': '#95003A',\n",
    "             '10': '#FF937E',\n",
    "             '11': '#A42400',\n",
    "             '12': '#001544',\n",
    "             '13': '#91D0CB',\n",
    "             '14': '#620E00',\n",
    "             '15': '#6B6882',\n",
    "             '16': '#0000FF',\n",
    "             '17': '#007DB5',\n",
    "             '18': '#6A826C',\n",
    "             '19': '#00AE7E',\n",
    "             '20': '#C28C9F',\n",
    "             '21': '#BE9970',\n",
    "             '22': '#008F9C',\n",
    "             '23': '#5FAD4E',\n",
    "             '24': '#FF0000',\n",
    "             '25': '#FF00F6',\n",
    "             '26': '#FF029D',\n",
    "             '27': '#683D3B',\n",
    "             '28': '#FF74A3',\n",
    "             '29': '#968AE8',\n",
    "             '30': '#98FF52',\n",
    "             '31': '#A75740',\n",
    "             '32': '#01FFFE',\n",
    "             '33': '#FFEEE8',\n",
    "             '34': '#FE8900',\n",
    "             '35': '#BDC6FF',\n",
    "             '36': '#01D0FF',\n",
    "             '37': '#BB8800',\n",
    "             '38': '#7544B1',\n",
    "             '39': '#A5FFD2',\n",
    "             '40': '#FFA6FE',\n",
    "             '41': '#774D00',\n",
    "             '42': '#7A4782',\n",
    "             '43': '#263400',\n",
    "             '44': '#004754',\n",
    "             '45': '#43002C',\n",
    "             '46': '#B500FF',\n",
    "             '47': '#FFB167',\n",
    "             '48': '#FFDB66',\n",
    "             '49': '#90FB92',\n",
    "             '50': '#7E2DD2',\n",
    "             '51': '#BDD393',\n",
    "             '52': '#E56FFE',\n",
    "             '53': '#DEFF74',\n",
    "             '54': '#00FF78',\n",
    "             '55': '#009BFF',\n",
    "             '56': '#006401',\n",
    "             '57': '#0076FF',\n",
    "             '58': '#85A900',\n",
    "             '59': '#00B917',\n",
    "             '60': '#788231',\n",
    "             '61': '#00FFC6',\n",
    "             '62': '#FF6E41',\n",
    "             '-1': '#000000'}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "29a584d2-8e1a-4de5-930c-eb3a45452a5d",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(359263, 25, 501)\n"
     ]
    }
   ],
   "source": [
    "# # Load the .npy file\n",
    "# data = np.load('/data/peer/panb2/replicate/result_tensor.npy')\n",
    "# print(data.shape)\n",
    "\n",
    "# # the dimension of the region is not important, so we remove it to faciliate later analysis\n",
    "# concatenated_data = np.concatenate(data, axis=1)\n",
    "# print(concatenated_data.shape)\n",
    "\n",
    "# np.save('/data/peer/panb2/replicate/concatenated_result_tensor.npy', concatenated_data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "9ea1d5ce-2954-4ceb-8943-e3d8609887e5",
   "metadata": {},
   "outputs": [],
   "source": [
    "concatenated_data = np.load('/data/peer/panb2/replicate/concatenated_result_tensor.npy')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "id": "998d7753-d83f-4307-8f49-fc048b067337",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(25, 179990763)"
      ]
     },
     "execution_count": 86,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "concatenated_data.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4064d6b6-223c-4531-96db-cffb1fed031c",
   "metadata": {},
   "source": [
    "#### Perform preprocessing on the data\n",
    "We first remove all the columns that sum to 0. These columns don't provide us with any valuable information regarding how cell clusters are different from each other."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "id": "c3470d1f-7adf-480b-aada-d96c5714bebc",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Initialize a list to store valid column indices\n",
    "valid_column_indices = []\n",
    "\n",
    "# Iterate over each column index\n",
    "for i in range(concatenated_data.shape[1]):\n",
    "    # Extract the column\n",
    "    column = concatenated_data[:, i]\n",
    "    \n",
    "    # Check if the column sum is 0 or contains NaN values\n",
    "    if np.sum(column) != 0 and np.sum(column) != np.inf:\n",
    "        # If the column is valid, add its index to the list\n",
    "        valid_column_indices.append(i)\n",
    "\n",
    "# Create a new array containing only the valid columns\n",
    "filtered_data = concatenated_data[:, valid_column_indices]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "id": "c0e1d3b4-c2df-46f7-93db-d67a2747b00e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(25, 174884318)"
      ]
     },
     "execution_count": 88,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "filtered_data.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6f60a32c-2274-47c8-8ca8-93548366314f",
   "metadata": {},
   "source": [
    "174884335 columns were left after removing columns with sum of 0. \n",
    "174884318 columns were left after furthur removing columns that sums up to inf. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "id": "85715d9d-c718-42dd-bd6c-21f08eab76a7",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "289.913888276746\n"
     ]
    }
   ],
   "source": [
    "# making sure that the largest element is no longer infinity\n",
    "largest_element = np.max(filtered_data)\n",
    "print(largest_element)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "id": "73492428-76db-4eda-a2aa-d50a16ee8c66",
   "metadata": {},
   "outputs": [],
   "source": [
    "np.save('/data/peer/panb2/replicate/filtered_concatenated_result_tensor.npy', filtered_data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "f7938afa-f565-4dc2-b325-9a3157a54f10",
   "metadata": {},
   "outputs": [],
   "source": [
    "filtered_data = np.load('/data/peer/panb2/replicate/filtered_concatenated_result_tensor.npy')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5676ca96-ba19-4126-94e4-f2da36971324",
   "metadata": {},
   "source": [
    "#### Perform PCA on a subset of the data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "9765d4af-c9f0-4e98-b647-c7c8f41707f3",
   "metadata": {},
   "outputs": [],
   "source": [
    "# 'data' is the array of size (25, 174884318)\n",
    "\n",
    "# Take a random subset of data\n",
    "subset_size = 174884318\n",
    "random_indices = np.random.choice(filtered_data.shape[1], size=subset_size, replace=False)\n",
    "\n",
    "data_subset = filtered_data#[:, random_indices]\n",
    "\n",
    "# Perform PCA\n",
    "n_components = 25  # Choose the number of principal components\n",
    "pca = PCA(n_components=n_components)\n",
    "pca.fit(data_subset)\n",
    "\n",
    "# Transform the data using the learned PCA model\n",
    "transformed_data = pca.transform(data_subset)\n",
    "np.save('transformed_data_2.npy', transformed_data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "247e5780-d72c-46b9-9572-55b671aad7ae",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(25, 25)"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "transformed_data.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "72fd019a-869a-4096-b3a6-09471d1179c5",
   "metadata": {},
   "outputs": [],
   "source": [
    "transformed_data = np.load('transformed_data_2.npy')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "3bdaca69-2abd-434c-9e08-3440cee39abc",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Create labels for each data point\n",
    "original_clusters = [0,1,2,3,4,5,6,7,8,9,10,11,12,13,15,16,17,18,19,20,21,22,23,26,28]\n",
    "labels = [f'Cluster {i}' for i in original_clusters]\n",
    "\n",
    "# Create a scatter plot of the transformed data with colored labels\n",
    "plt.figure(figsize=(8, 6))\n",
    "for i in range(transformed_data.shape[0]):\n",
    "    plt.scatter(transformed_data[i, 0], transformed_data[i, 1], color=[colors[str(i)]], alpha=0.5, label=labels[i])\n",
    "\n",
    "plt.title('PCA Transformed Data with Colored Labels')\n",
    "plt.xlabel('Principal Component 1')\n",
    "plt.ylabel('Principal Component 2')\n",
    "plt.grid(True)\n",
    "plt.legend(bbox_to_anchor=(1.05, 1), loc='upper left')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1f8e7c9f-31dc-4a24-917d-cc3a2fce89f9",
   "metadata": {},
   "source": [
    "#### Feature Selection: Examine which CRE contributes to the correlation with PC1\n",
    "First load the relevant numpy arrays"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "906e07e2-a2af-4b1d-ab8a-d780f962d514",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(25, 174884318)"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "filtered_data = np.load('/data/peer/panb2/replicate/filtered_concatenated_result_tensor.npy')\n",
    "filtered_data.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "d0b2d67d-01d9-4bc4-a990-89094f4e794e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(25, 25)"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "transformed_data = np.load('transformed_data_2.npy')\n",
    "transformed_data.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "70c2038d-b0ae-4361-b6be-43c076c33034",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([-1776.03160405,  5113.13641059, -1705.2421875 ,  1875.4731345 ,\n",
       "       -1804.27766616,  -326.90456065, -1349.51725793, -1600.76801955,\n",
       "        3742.5016668 ,  -463.76532787,  2820.39029319,  5508.6286909 ,\n",
       "        -766.86782964, -1673.31864632,   -27.14293585, -1518.26323506,\n",
       "       -1723.93632789, -1822.91907442, -1732.18919954,  -513.34666809,\n",
       "       -1268.75903735,  2515.61111491,  1624.0219051 , -1796.06706955,\n",
       "       -1330.44656857])"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# extracting the first principle component\n",
    "pc1 = transformed_data[:,0]\n",
    "pc1"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fc11ceed-9228-48f4-9de6-00b2e550dcad",
   "metadata": {},
   "source": [
    "We realized that the clusters with the two largest values of PC1 are Cluster 1 and Cluster 11. (not Cluster 11 and 12 as previously believed).\n",
    "\n",
    "We now calculate the correlation coefficients of PC1 with each of the 174884318 features in the filtered_data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "9538528c-4696-4f76-aae5-c70280bcb117",
   "metadata": {},
   "outputs": [],
   "source": [
    "# filtered_data = np.random.rand(25,1000)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "bd973406-b6c3-49ec-947c-c4197f7b4856",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Processing:   0%|          | 590187/174884318 [08:02<41:44:56, 1159.67it/s]"
     ]
    }
   ],
   "source": [
    "# Pre-allocate NumPy arrays for the Spearman and Pearson coefficients\n",
    "pearson_coeff_array = np.empty(filtered_data.shape[1])\n",
    "\n",
    "total_iterations = filtered_data.shape[1]\n",
    "progress_bar = tqdm(total=total_iterations, desc=\"Processing\")\n",
    "\n",
    "for i in range(total_iterations):\n",
    "    pearson_coeff_array[i] = pearsonr(pc1, filtered_data[:,i])[0]\n",
    "    progress_bar.update(1)\n",
    "\n",
    "progress_bar.close()\n",
    "print(\"Task completed!\")\n",
    "print(pearson_coeff_array.shape)\n",
    "\n",
    "np.save('/data/peer/panb2/replicate/pearson_coeff_array.npy', pearson_coeff_array)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "06674edf-19c9-424f-b010-6c2c3e574a78",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Processing:  14%|█▍        | 24990824/174884318 [2:02:04<12:07:28, 3434.08it/s]"
     ]
    }
   ],
   "source": [
    "# Pre-allocate NumPy arrays for the Spearman and Pearson coefficients\n",
    "spearman_coeff_array = np.empty(filtered_data.shape[1])\n",
    "\n",
    "total_iterations = filtered_data.shape[1]\n",
    "progress_bar = tqdm(total=total_iterations, desc=\"Processing\")\n",
    "\n",
    "# We still need a loop for Spearman since it involves ranking\n",
    "for i in range(total_iterations):\n",
    "    spearman_coeff_array[i] = spearmanr(pc1, filtered_data[:, i])[0]\n",
    "    progress_bar.update(1)\n",
    "\n",
    "progress_bar.close()\n",
    "print(\"Task completed!\")\n",
    "print(spearman_coeff_array.shape)\n",
    "\n",
    "np.save('/data/peer/panb2/replicate/spearman_coeff_array.npy', spearman_coeff_array)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b4312ff8-75ef-4dc3-af81-a9b403934294",
   "metadata": {},
   "source": [
    "The next step is to find the CRE's with very positive or very negative correlation coefficients, and plot the nucleosome position changes across clusters."
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python (chrom_env)",
   "language": "python",
   "name": "chrom_env"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.18"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}