scFEA_tutorial1.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# scFEA Tutorial"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 1. Video introduction\n",
    "scFEA has been highlighted in recomb 2021. For more details, please see our presentation video at [link](https://www.youtube.com/watch?v=-S8ouz2F-Rk).\n",
    "\n",
    "### 2. Installation\n",
    "\n",
    "scFEA is implemented by Python 3. If you don't have Python, please download [Anaconda](https://www.anaconda.com/products/individual-d) with Python 3 version. \n",
    "\n",
    "- torch >= 0.4.1\n",
    "- numpy >= 1.15.4\n",
    "- pandas >= 0.23.4\n",
    "- matplotlib >= 3.0.2\n",
    "- magic >= 2.0.4\n",
    "\n",
    "Download scFEA:\n",
    "\n",
    "\n",
    "`git clone https://github.com/changwn/scFEA\n",
    "`\n",
    "\n",
    "Install requirements:\n",
    "\n",
    "`\n",
    "cd scFEA\n",
    "conda install --file requirements\n",
    "conda install pytorch torchvision -c pytorch\n",
    "pip install --user magic-impute\n",
    "`\n",
    "\n",
    "### 3. Verify Installation\n",
    "\n",
    "To ensure that the scFEA is installed correctly, run below code. If you can see help documentation, the software was installed successfully."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "usage: scFEA.py [-h] [--data_dir <data_directory>]\n",
      "                [--input_dir <input_directory>] [--res_dir <data_directory>]\n",
      "                [--test_file TEST_FILE] [--moduleGene_file MODULEGENE_FILE]\n",
      "                [--stoichiometry_matrix STOICHIOMETRY_MATRIX]\n",
      "                [--cName_file CNAME_FILE] [--sc_imputation {True,False}]\n",
      "                [--output_flux_file OUTPUT_FLUX_FILE]\n",
      "                [--output_balance_file OUTPUT_BALANCE_FILE]\n",
      "\n",
      "scFEA: A graph neural network model to estimate cell-wise metabolic flux using\n",
      "single cell RNA-seq data\n",
      "\n",
      "optional arguments:\n",
      "  -h, --help            show this help message and exit\n",
      "  --data_dir <data_directory>\n",
      "                        The data directory for scFEA model files.\n",
      "  --input_dir <input_directory>\n",
      "                        The data directory for single cell input data.\n",
      "  --res_dir <data_directory>\n",
      "                        The data directory for result [output]. The output of\n",
      "                        scFEA includes two matrices, predicted metabolic flux\n",
      "                        and metabolites stress at single cell resolution.\n",
      "  --test_file TEST_FILE\n",
      "                        The test SC file [input]. The input of scFEA is a\n",
      "                        single cell profile matrix, where row is gene and\n",
      "                        column is cell. Example datasets are provided in\n",
      "                        /data/ folder. The input can be raw counts or\n",
      "                        normalised counts. The logarithm would be performed if\n",
      "                        value larger than 30.\n",
      "  --moduleGene_file MODULEGENE_FILE\n",
      "                        The table contains genes for each module. We provide\n",
      "                        human and mouse two models in scFEA. For human model,\n",
      "                        please use module_gene_m168.csv which is default. All\n",
      "                        candidate moduleGene files are provided in /data/\n",
      "                        folder.\n",
      "  --stoichiometry_matrix STOICHIOMETRY_MATRIX\n",
      "                        The table describes relationship between compounds and\n",
      "                        modules. Each row is an intermediate metabolite and\n",
      "                        each column is metabolic module. For human model,\n",
      "                        please use cmMat_171.csv which is default. All\n",
      "                        candidate stoichiometry matrices are provided in\n",
      "                        /data/ folder.\n",
      "  --cName_file CNAME_FILE\n",
      "                        The name of compounds. The table contains two rows.\n",
      "                        First row is compounds name and second row is\n",
      "                        corresponding id.\n",
      "  --sc_imputation {True,False}\n",
      "                        Whether perform imputation for SC dataset (recommend\n",
      "                        set to <True> for 10x data).\n",
      "  --output_flux_file OUTPUT_FLUX_FILE\n",
      "                        User defined predicted flux file name.\n",
      "  --output_balance_file OUTPUT_BALANCE_FILE\n",
      "                        User defined predicted balance file name.\n"
     ]
    }
   ],
   "source": [
    "%%bash\n",
    "cd /Users/chang/Documents/work/flux/scFEA\n",
    "python src/scFEA.py --help"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 4. Run scFEA\n",
    "\n",
    "#### 4.1 example 1 human complete modules\n",
    "\n",
    "This example run scFEA using Melissa data on humam complete modules."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Starting load data...\n",
      "Load compound name file, the balance output will have compound name.\n",
      "Load data done.\n",
      "Starting process data...\n",
      "Process data done.\n",
      "Starting train neural network...\n",
      "Training time:  15.639593124389648\n",
      "scFEA job finished. Check result in the desired output folder.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  0%|          | 0/100 [00:00<?, ?it/s]\r",
      "  1%|          | 1/100 [00:00<00:30,  3.27it/s]\r",
      "  2%|▏         | 2/100 [00:00<00:26,  3.76it/s]\r",
      "  3%|▎         | 3/100 [00:00<00:22,  4.24it/s]\r",
      "  4%|▍         | 4/100 [00:00<00:20,  4.79it/s]\r",
      "  5%|▌         | 5/100 [00:00<00:19,  4.97it/s]\r",
      "  6%|▌         | 6/100 [00:01<00:18,  4.96it/s]\r",
      "  7%|▋         | 7/100 [00:01<00:17,  5.18it/s]\r",
      "  8%|▊         | 8/100 [00:01<00:17,  5.30it/s]\r",
      "  9%|▉         | 9/100 [00:01<00:17,  5.29it/s]\r",
      " 10%|█         | 10/100 [00:01<00:16,  5.43it/s]\r",
      " 11%|█         | 11/100 [00:02<00:16,  5.46it/s]\r",
      " 12%|█▏        | 12/100 [00:02<00:15,  5.82it/s]\r",
      " 13%|█▎        | 13/100 [00:02<00:14,  6.10it/s]\r",
      " 14%|█▍        | 14/100 [00:02<00:14,  6.13it/s]\r",
      " 15%|█▌        | 15/100 [00:02<00:13,  6.25it/s]\r",
      " 16%|█▌        | 16/100 [00:02<00:13,  6.01it/s]\r",
      " 17%|█▋        | 17/100 [00:03<00:13,  6.02it/s]\r",
      " 18%|█▊        | 18/100 [00:03<00:14,  5.85it/s]\r",
      " 19%|█▉        | 19/100 [00:03<00:13,  6.15it/s]\r",
      " 20%|██        | 20/100 [00:03<00:12,  6.21it/s]\r",
      " 21%|██        | 21/100 [00:03<00:13,  6.05it/s]\r",
      " 22%|██▏       | 22/100 [00:03<00:12,  6.05it/s]\r",
      " 23%|██▎       | 23/100 [00:04<00:13,  5.91it/s]\r",
      " 24%|██▍       | 24/100 [00:04<00:12,  6.01it/s]\r",
      " 25%|██▌       | 25/100 [00:04<00:12,  6.04it/s]\r",
      " 26%|██▌       | 26/100 [00:04<00:12,  5.90it/s]\r",
      " 27%|██▋       | 27/100 [00:04<00:12,  5.81it/s]\r",
      " 28%|██▊       | 28/100 [00:04<00:12,  5.80it/s]\r",
      " 29%|██▉       | 29/100 [00:05<00:12,  5.76it/s]\r",
      " 30%|███       | 30/100 [00:05<00:11,  5.84it/s]\r",
      " 31%|███       | 31/100 [00:05<00:11,  5.82it/s]\r",
      " 32%|███▏      | 32/100 [00:05<00:11,  5.72it/s]\r",
      " 33%|███▎      | 33/100 [00:05<00:11,  5.62it/s]\r",
      " 34%|███▍      | 34/100 [00:05<00:11,  5.66it/s]\r",
      " 35%|███▌      | 35/100 [00:06<00:11,  5.69it/s]\r",
      " 36%|███▌      | 36/100 [00:06<00:11,  5.79it/s]\r",
      " 37%|███▋      | 37/100 [00:06<00:10,  5.88it/s]\r",
      " 38%|███▊      | 38/100 [00:06<00:10,  6.10it/s]\r",
      " 39%|███▉      | 39/100 [00:06<00:09,  6.31it/s]\r",
      " 40%|████      | 40/100 [00:06<00:09,  6.11it/s]\r",
      " 41%|████      | 41/100 [00:07<00:09,  6.21it/s]\r",
      " 42%|████▏     | 42/100 [00:07<00:08,  6.46it/s]\r",
      " 43%|████▎     | 43/100 [00:07<00:08,  6.65it/s]\r",
      " 44%|████▍     | 44/100 [00:07<00:08,  6.74it/s]\r",
      " 45%|████▌     | 45/100 [00:07<00:08,  6.81it/s]\r",
      " 46%|████▌     | 46/100 [00:07<00:07,  6.91it/s]\r",
      " 47%|████▋     | 47/100 [00:07<00:07,  6.86it/s]\r",
      " 48%|████▊     | 48/100 [00:08<00:07,  6.92it/s]\r",
      " 49%|████▉     | 49/100 [00:08<00:07,  6.99it/s]\r",
      " 50%|█████     | 50/100 [00:08<00:07,  7.05it/s]\r",
      " 51%|█████     | 51/100 [00:08<00:06,  7.07it/s]\r",
      " 52%|█████▏    | 52/100 [00:08<00:06,  6.97it/s]\r",
      " 53%|█████▎    | 53/100 [00:08<00:06,  7.01it/s]\r",
      " 54%|█████▍    | 54/100 [00:08<00:06,  7.06it/s]\r",
      " 55%|█████▌    | 55/100 [00:09<00:06,  7.12it/s]\r",
      " 56%|█████▌    | 56/100 [00:09<00:06,  7.09it/s]\r",
      " 57%|█████▋    | 57/100 [00:09<00:06,  7.09it/s]\r",
      " 58%|█████▊    | 58/100 [00:09<00:05,  7.11it/s]\r",
      " 59%|█████▉    | 59/100 [00:09<00:05,  7.13it/s]\r",
      " 60%|██████    | 60/100 [00:09<00:05,  7.16it/s]\r",
      " 61%|██████    | 61/100 [00:09<00:05,  7.17it/s]\r",
      " 62%|██████▏   | 62/100 [00:10<00:05,  7.15it/s]\r",
      " 63%|██████▎   | 63/100 [00:10<00:05,  7.13it/s]\r",
      " 64%|██████▍   | 64/100 [00:10<00:05,  7.14it/s]\r",
      " 65%|██████▌   | 65/100 [00:10<00:04,  7.16it/s]\r",
      " 66%|██████▌   | 66/100 [00:10<00:04,  7.12it/s]\r",
      " 67%|██████▋   | 67/100 [00:10<00:04,  7.10it/s]\r",
      " 68%|██████▊   | 68/100 [00:10<00:04,  7.11it/s]\r",
      " 69%|██████▉   | 69/100 [00:11<00:04,  7.14it/s]\r",
      " 70%|███████   | 70/100 [00:11<00:04,  7.12it/s]\r",
      " 71%|███████   | 71/100 [00:11<00:04,  7.12it/s]\r",
      " 72%|███████▏  | 72/100 [00:11<00:03,  7.14it/s]\r",
      " 73%|███████▎  | 73/100 [00:11<00:03,  7.16it/s]\r",
      " 74%|███████▍  | 74/100 [00:11<00:03,  7.13it/s]\r",
      " 75%|███████▌  | 75/100 [00:11<00:03,  7.16it/s]\r",
      " 76%|███████▌  | 76/100 [00:11<00:03,  7.17it/s]\r",
      " 77%|███████▋  | 77/100 [00:12<00:03,  7.15it/s]\r",
      " 78%|███████▊  | 78/100 [00:12<00:03,  7.04it/s]\r",
      " 79%|███████▉  | 79/100 [00:12<00:02,  7.05it/s]\r",
      " 80%|████████  | 80/100 [00:12<00:02,  7.04it/s]\r",
      " 81%|████████  | 81/100 [00:12<00:02,  7.02it/s]\r",
      " 82%|████████▏ | 82/100 [00:12<00:02,  7.04it/s]\r",
      " 83%|████████▎ | 83/100 [00:12<00:02,  7.07it/s]\r",
      " 84%|████████▍ | 84/100 [00:13<00:02,  7.09it/s]\r",
      " 85%|████████▌ | 85/100 [00:13<00:02,  7.08it/s]\r",
      " 86%|████████▌ | 86/100 [00:13<00:02,  6.82it/s]\r",
      " 87%|████████▋ | 87/100 [00:13<00:01,  6.91it/s]\r",
      " 88%|████████▊ | 88/100 [00:13<00:01,  6.91it/s]\r",
      " 89%|████████▉ | 89/100 [00:13<00:01,  6.57it/s]\r",
      " 90%|█████████ | 90/100 [00:14<00:01,  6.28it/s]\r",
      " 91%|█████████ | 91/100 [00:14<00:01,  6.14it/s]\r",
      " 92%|█████████▏| 92/100 [00:14<00:01,  5.96it/s]\r",
      " 93%|█████████▎| 93/100 [00:14<00:01,  6.06it/s]\r",
      " 94%|█████████▍| 94/100 [00:14<00:00,  6.28it/s]\r",
      " 95%|█████████▌| 95/100 [00:14<00:00,  6.52it/s]\r",
      " 96%|█████████▌| 96/100 [00:14<00:00,  6.70it/s]\r",
      " 97%|█████████▋| 97/100 [00:15<00:00,  6.81it/s]\r",
      " 98%|█████████▊| 98/100 [00:15<00:00,  6.86it/s]\r",
      " 99%|█████████▉| 99/100 [00:15<00:00,  6.46it/s]\r",
      "100%|██████████| 100/100 [00:15<00:00,  6.18it/s]\r",
      "100%|██████████| 100/100 [00:15<00:00,  6.40it/s]\n"
     ]
    }
   ],
   "source": [
    "%%bash\n",
    "cd /Users/chang/Documents/work/flux/scFEA\n",
    "python src/scFEA.py --data_dir data --input_dir input \\\n",
    "                    --test_file Melissa_full.csv \\\n",
    "                    --moduleGene_file module_gene_m168.csv \\\n",
    "                    --stoichiometry_matrix cmMat_c70_m168.csv"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 4.2 example 2 mouse complete modules\n",
    "\n",
    "This example run scFEA with a mouse example data on mouse complete modules. Also, we set `sc_imputation` as `True` to enable imputation process. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Calculating MAGIC...\n",
      "Starting load data...\n",
      "  Running MAGIC on 5659 cells and 678 genes.\n",
      "  Calculating graph and diffusion operator...\n",
      "    Calculating PCA...\n",
      "    Calculated PCA in 0.36 seconds.\n",
      "    Calculating KNN search...\n",
      "    Calculated KNN search in 4.47 seconds.\n",
      "    Calculating affinities...\n",
      "    Calculated affinities in 4.45 seconds.\n",
      "  Calculated graph and diffusion operator in 9.30 seconds.\n",
      "  Calculating imputation...\n",
      "  Calculated imputation in 0.24 seconds.\n",
      "Calculated MAGIC in 9.56 seconds.\n",
      "Load compound name file, the balance output will have compound name.\n",
      "Load data done.\n",
      "Starting process data...\n",
      "Process data done.\n",
      "Starting train neural network...\n",
      "Training time:  1200.0908679962158\n",
      "scFEA job finished. Check result in the desired output folder.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  0%|          | 0/100 [00:00<?, ?it/s]\r",
      "  1%|          | 1/100 [00:13<21:40, 13.14s/it]\r",
      "  2%|▏         | 2/100 [00:24<20:22, 12.47s/it]\r",
      "  3%|▎         | 3/100 [00:37<20:37, 12.76s/it]\r",
      "  4%|▍         | 4/100 [00:49<20:05, 12.55s/it]\r",
      "  5%|▌         | 5/100 [01:01<19:45, 12.48s/it]\r",
      "  6%|▌         | 6/100 [01:14<19:47, 12.63s/it]\r",
      "  7%|▋         | 7/100 [01:27<19:30, 12.58s/it]\r",
      "  8%|▊         | 8/100 [01:39<19:20, 12.61s/it]\r",
      "  9%|▉         | 9/100 [01:52<19:03, 12.57s/it]\r",
      " 10%|█         | 10/100 [02:03<18:11, 12.13s/it]\r",
      " 11%|█         | 11/100 [02:14<17:38, 11.90s/it]\r",
      " 12%|█▏        | 12/100 [02:28<18:01, 12.29s/it]\r",
      " 13%|█▎        | 13/100 [02:40<17:51, 12.32s/it]\r",
      " 14%|█▍        | 14/100 [02:53<17:48, 12.42s/it]\r",
      " 15%|█▌        | 15/100 [03:05<17:42, 12.50s/it]\r",
      " 16%|█▌        | 16/100 [03:18<17:28, 12.48s/it]\r",
      " 17%|█▋        | 17/100 [03:33<18:16, 13.21s/it]\r",
      " 18%|█▊        | 18/100 [03:47<18:22, 13.44s/it]\r",
      " 19%|█▉        | 19/100 [04:00<18:00, 13.34s/it]\r",
      " 20%|██        | 20/100 [04:14<17:56, 13.46s/it]\r",
      " 21%|██        | 21/100 [04:26<17:16, 13.12s/it]\r",
      " 22%|██▏       | 22/100 [04:36<16:00, 12.32s/it]\r",
      " 23%|██▎       | 23/100 [04:47<15:20, 11.96s/it]\r",
      " 24%|██▍       | 24/100 [04:58<14:42, 11.61s/it]\r",
      " 25%|██▌       | 25/100 [05:09<14:16, 11.42s/it]\r",
      " 26%|██▌       | 26/100 [05:21<14:03, 11.40s/it]\r",
      " 27%|██▋       | 27/100 [05:32<13:49, 11.36s/it]\r",
      " 28%|██▊       | 28/100 [05:43<13:36, 11.35s/it]\r",
      " 29%|██▉       | 29/100 [05:54<13:14, 11.19s/it]\r",
      " 30%|███       | 30/100 [06:05<12:50, 11.01s/it]\r",
      " 31%|███       | 31/100 [06:15<12:24, 10.79s/it]\r",
      " 32%|███▏      | 32/100 [06:25<12:00, 10.59s/it]\r",
      " 33%|███▎      | 33/100 [06:35<11:43, 10.50s/it]\r",
      " 34%|███▍      | 34/100 [06:46<11:40, 10.61s/it]\r",
      " 35%|███▌      | 35/100 [06:56<11:25, 10.54s/it]\r",
      " 36%|███▌      | 36/100 [07:09<11:51, 11.12s/it]\r",
      " 37%|███▋      | 37/100 [07:20<11:36, 11.05s/it]\r",
      " 38%|███▊      | 38/100 [07:32<11:36, 11.24s/it]\r",
      " 39%|███▉      | 39/100 [07:42<11:20, 11.16s/it]\r",
      " 40%|████      | 40/100 [07:55<11:32, 11.54s/it]\r",
      " 41%|████      | 41/100 [08:07<11:36, 11.81s/it]\r",
      " 42%|████▏     | 42/100 [08:19<11:23, 11.78s/it]\r",
      " 43%|████▎     | 43/100 [08:30<10:53, 11.46s/it]\r",
      " 44%|████▍     | 44/100 [08:42<10:51, 11.64s/it]\r",
      " 45%|████▌     | 45/100 [08:53<10:29, 11.44s/it]\r",
      " 46%|████▌     | 46/100 [09:04<10:07, 11.26s/it]\r",
      " 47%|████▋     | 47/100 [09:15<09:57, 11.27s/it]\r",
      " 48%|████▊     | 48/100 [09:26<09:37, 11.11s/it]\r",
      " 49%|████▉     | 49/100 [09:37<09:31, 11.22s/it]\r",
      " 50%|█████     | 50/100 [09:51<10:05, 12.11s/it]\r",
      " 51%|█████     | 51/100 [10:06<10:30, 12.87s/it]\r",
      " 52%|█████▏    | 52/100 [10:20<10:37, 13.28s/it]\r",
      " 53%|█████▎    | 53/100 [10:34<10:37, 13.57s/it]\r",
      " 54%|█████▍    | 54/100 [10:48<10:22, 13.53s/it]\r",
      " 55%|█████▌    | 55/100 [11:02<10:14, 13.65s/it]\r",
      " 56%|█████▌    | 56/100 [11:17<10:14, 13.97s/it]\r",
      " 57%|█████▋    | 57/100 [11:30<09:48, 13.69s/it]\r",
      " 58%|█████▊    | 58/100 [11:41<09:11, 13.14s/it]\r",
      " 59%|█████▉    | 59/100 [11:55<09:00, 13.17s/it]\r",
      " 60%|██████    | 60/100 [12:08<08:47, 13.19s/it]\r",
      " 61%|██████    | 61/100 [12:22<08:39, 13.33s/it]\r",
      " 62%|██████▏   | 62/100 [12:35<08:25, 13.32s/it]\r",
      " 63%|██████▎   | 63/100 [12:49<08:25, 13.66s/it]\r",
      " 64%|██████▍   | 64/100 [13:03<08:08, 13.58s/it]\r",
      " 65%|██████▌   | 65/100 [13:16<07:48, 13.38s/it]\r",
      " 66%|██████▌   | 66/100 [13:28<07:21, 12.98s/it]\r",
      " 67%|██████▋   | 67/100 [13:39<06:49, 12.41s/it]\r",
      " 68%|██████▊   | 68/100 [13:50<06:24, 12.03s/it]\r",
      " 69%|██████▉   | 69/100 [14:01<06:05, 11.78s/it]\r",
      " 70%|███████   | 70/100 [14:13<05:54, 11.83s/it]\r",
      " 71%|███████   | 71/100 [14:23<05:30, 11.41s/it]\r",
      " 72%|███████▏  | 72/100 [14:38<05:42, 12.24s/it]\r",
      " 73%|███████▎  | 73/100 [14:51<05:40, 12.61s/it]\r",
      " 74%|███████▍  | 74/100 [15:04<05:26, 12.57s/it]\r",
      " 75%|███████▌  | 75/100 [15:17<05:20, 12.81s/it]\r",
      " 76%|███████▌  | 76/100 [15:30<05:08, 12.85s/it]\r",
      " 77%|███████▋  | 77/100 [15:42<04:48, 12.55s/it]\r",
      " 78%|███████▊  | 78/100 [15:53<04:29, 12.24s/it]\r",
      " 79%|███████▉  | 79/100 [16:04<04:07, 11.78s/it]\r",
      " 80%|████████  | 80/100 [16:15<03:50, 11.51s/it]\r",
      " 81%|████████  | 81/100 [16:26<03:34, 11.27s/it]\r",
      " 82%|████████▏ | 82/100 [16:36<03:17, 10.97s/it]\r",
      " 83%|████████▎ | 83/100 [16:46<03:02, 10.75s/it]\r",
      " 84%|████████▍ | 84/100 [16:58<02:56, 11.05s/it]\r",
      " 85%|████████▌ | 85/100 [17:10<02:48, 11.25s/it]\r",
      " 86%|████████▌ | 86/100 [17:20<02:34, 11.04s/it]\r",
      " 87%|████████▋ | 87/100 [17:31<02:23, 11.00s/it]\r",
      " 88%|████████▊ | 88/100 [17:42<02:13, 11.09s/it]\r",
      " 89%|████████▉ | 89/100 [17:53<02:01, 11.09s/it]\r",
      " 90%|█████████ | 90/100 [18:04<01:49, 10.95s/it]\r",
      " 91%|█████████ | 91/100 [18:14<01:36, 10.76s/it]\r",
      " 92%|█████████▏| 92/100 [18:25<01:27, 10.88s/it]\r",
      " 93%|█████████▎| 93/100 [18:37<01:18, 11.15s/it]\r",
      " 94%|█████████▍| 94/100 [18:50<01:09, 11.63s/it]\r",
      " 95%|█████████▌| 95/100 [19:03<00:59, 11.93s/it]\r",
      " 96%|█████████▌| 96/100 [19:15<00:48, 12.13s/it]\r",
      " 97%|█████████▋| 97/100 [19:27<00:35, 11.89s/it]\r",
      " 98%|█████████▊| 98/100 [19:38<00:23, 11.70s/it]\r",
      " 99%|█████████▉| 99/100 [19:49<00:11, 11.47s/it]\r",
      "100%|██████████| 100/100 [20:00<00:00, 11.25s/it]\r",
      "100%|██████████| 100/100 [20:00<00:00, 12.00s/it]\n"
     ]
    }
   ],
   "source": [
    "%%bash\n",
    "cd /Users/chang/Documents/work/flux/scFEA\n",
    "python src/scFEA.py --data_dir data --input_dir input \\\n",
    "                    --test_file mouse_example_data.csv \\\n",
    "                    --moduleGene_file module_gene_complete_mouse_m168.csv \\\n",
    "                    --stoichiometry_matrix cmMat_complete_mouse_c70_m168.csv \\\n",
    "                    --sc_imputation True \\\n",
    "                    --output_flux_file output/mouse_flux.csv \\\n",
    "                    --output_balance_file output/mouse_balance.csv"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 4.3 example 3 iron ion modules\n",
    "\n",
    "This example run scFEA using GSE72056 on small metabolic system which is iron ion modules."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Starting load data...\n",
      "Load compound name file, the balance output will have compound name.\n",
      "Load data done.\n",
      "Starting process data...\n",
      "Process data done.\n",
      "Starting train neural network...\n",
      "Training time:  151.6518177986145\n",
      "scFEA job finished. Check result in the desired output folder.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  0%|          | 0/100 [00:00<?, ?it/s]\r",
      "  1%|          | 1/100 [00:01<02:25,  1.47s/it]\r",
      "  2%|▏         | 2/100 [00:02<02:22,  1.45s/it]\r",
      "  3%|▎         | 3/100 [00:04<02:21,  1.46s/it]\r",
      "  4%|▍         | 4/100 [00:05<02:21,  1.47s/it]\r",
      "  5%|▌         | 5/100 [00:07<02:17,  1.44s/it]\r",
      "  6%|▌         | 6/100 [00:08<02:13,  1.42s/it]\r",
      "  7%|▋         | 7/100 [00:09<02:10,  1.40s/it]\r",
      "  8%|▊         | 8/100 [00:11<02:09,  1.40s/it]\r",
      "  9%|▉         | 9/100 [00:12<02:06,  1.39s/it]\r",
      " 10%|█         | 10/100 [00:14<02:05,  1.39s/it]\r",
      " 11%|█         | 11/100 [00:15<02:06,  1.43s/it]\r",
      " 12%|█▏        | 12/100 [00:17<02:06,  1.44s/it]\r",
      " 13%|█▎        | 13/100 [00:18<02:03,  1.42s/it]\r",
      " 14%|█▍        | 14/100 [00:19<02:00,  1.40s/it]\r",
      " 15%|█▌        | 15/100 [00:21<01:58,  1.39s/it]\r",
      " 16%|█▌        | 16/100 [00:22<01:57,  1.40s/it]\r",
      " 17%|█▋        | 17/100 [00:24<01:55,  1.39s/it]\r",
      " 18%|█▊        | 18/100 [00:25<01:53,  1.38s/it]\r",
      " 19%|█▉        | 19/100 [00:26<01:57,  1.45s/it]\r",
      " 20%|██        | 20/100 [00:28<01:57,  1.47s/it]\r",
      " 21%|██        | 21/100 [00:30<01:57,  1.49s/it]\r",
      " 22%|██▏       | 22/100 [00:31<01:53,  1.46s/it]\r",
      " 23%|██▎       | 23/100 [00:32<01:51,  1.45s/it]\r",
      " 24%|██▍       | 24/100 [00:34<01:48,  1.43s/it]\r",
      " 25%|██▌       | 25/100 [00:35<01:45,  1.41s/it]\r",
      " 26%|██▌       | 26/100 [00:36<01:43,  1.40s/it]\r",
      " 27%|██▋       | 27/100 [00:38<01:42,  1.41s/it]\r",
      " 28%|██▊       | 28/100 [00:39<01:40,  1.40s/it]\r",
      " 29%|██▉       | 29/100 [00:41<01:38,  1.39s/it]\r",
      " 30%|███       | 30/100 [00:42<01:36,  1.38s/it]\r",
      " 31%|███       | 31/100 [00:44<01:40,  1.46s/it]\r",
      " 32%|███▏      | 32/100 [00:45<01:42,  1.51s/it]\r",
      " 33%|███▎      | 33/100 [00:47<01:40,  1.51s/it]\r",
      " 34%|███▍      | 34/100 [00:48<01:40,  1.52s/it]\r",
      " 35%|███▌      | 35/100 [00:50<01:37,  1.50s/it]\r",
      " 36%|███▌      | 36/100 [00:51<01:35,  1.49s/it]\r",
      " 37%|███▋      | 37/100 [00:53<01:33,  1.49s/it]\r",
      " 38%|███▊      | 38/100 [00:54<01:38,  1.58s/it]\r",
      " 39%|███▉      | 39/100 [00:56<01:34,  1.55s/it]\r",
      " 40%|████      | 40/100 [00:58<01:32,  1.55s/it]\r",
      " 41%|████      | 41/100 [00:59<01:30,  1.54s/it]\r",
      " 42%|████▏     | 42/100 [01:01<01:28,  1.53s/it]\r",
      " 43%|████▎     | 43/100 [01:02<01:27,  1.53s/it]\r",
      " 44%|████▍     | 44/100 [01:04<01:27,  1.56s/it]\r",
      " 45%|████▌     | 45/100 [01:05<01:25,  1.55s/it]\r",
      " 46%|████▌     | 46/100 [01:07<01:23,  1.55s/it]\r",
      " 47%|████▋     | 47/100 [01:08<01:20,  1.52s/it]\r",
      " 48%|████▊     | 48/100 [01:10<01:20,  1.55s/it]\r",
      " 49%|████▉     | 49/100 [01:11<01:19,  1.57s/it]\r",
      " 50%|█████     | 50/100 [01:13<01:17,  1.55s/it]\r",
      " 51%|█████     | 51/100 [01:15<01:16,  1.55s/it]\r",
      " 52%|█████▏    | 52/100 [01:16<01:14,  1.54s/it]\r",
      " 53%|█████▎    | 53/100 [01:18<01:13,  1.57s/it]\r",
      " 54%|█████▍    | 54/100 [01:19<01:14,  1.61s/it]\r",
      " 55%|█████▌    | 55/100 [01:21<01:10,  1.58s/it]\r",
      " 56%|█████▌    | 56/100 [01:22<01:08,  1.56s/it]\r",
      " 57%|█████▋    | 57/100 [01:24<01:08,  1.59s/it]\r",
      " 58%|█████▊    | 58/100 [01:26<01:05,  1.56s/it]\r",
      " 59%|█████▉    | 59/100 [01:27<01:04,  1.57s/it]\r",
      " 60%|██████    | 60/100 [01:29<01:00,  1.52s/it]\r",
      " 61%|██████    | 61/100 [01:30<00:59,  1.52s/it]\r",
      " 62%|██████▏   | 62/100 [01:32<00:56,  1.49s/it]\r",
      " 63%|██████▎   | 63/100 [01:33<00:56,  1.52s/it]\r",
      " 64%|██████▍   | 64/100 [01:35<00:56,  1.58s/it]\r",
      " 65%|██████▌   | 65/100 [01:37<00:57,  1.65s/it]\r",
      " 66%|██████▌   | 66/100 [01:38<00:56,  1.67s/it]\r",
      " 67%|██████▋   | 67/100 [01:40<00:55,  1.69s/it]\r",
      " 68%|██████▊   | 68/100 [01:42<00:54,  1.69s/it]\r",
      " 69%|██████▉   | 69/100 [01:43<00:52,  1.69s/it]\r",
      " 70%|███████   | 70/100 [01:45<00:51,  1.72s/it]\r",
      " 71%|███████   | 71/100 [01:47<00:50,  1.74s/it]\r",
      " 72%|███████▏  | 72/100 [01:49<00:49,  1.77s/it]\r",
      " 73%|███████▎  | 73/100 [01:51<00:46,  1.74s/it]\r",
      " 74%|███████▍  | 74/100 [01:52<00:43,  1.69s/it]\r",
      " 75%|███████▌  | 75/100 [01:54<00:40,  1.61s/it]\r",
      " 76%|███████▌  | 76/100 [01:55<00:38,  1.61s/it]\r",
      " 77%|███████▋  | 77/100 [01:57<00:36,  1.58s/it]\r",
      " 78%|███████▊  | 78/100 [01:58<00:34,  1.56s/it]\r",
      " 79%|███████▉  | 79/100 [02:00<00:32,  1.54s/it]\r",
      " 80%|████████  | 80/100 [02:01<00:31,  1.56s/it]\r",
      " 81%|████████  | 81/100 [02:03<00:30,  1.58s/it]\r",
      " 82%|████████▏ | 82/100 [02:04<00:28,  1.56s/it]\r",
      " 83%|████████▎ | 83/100 [02:06<00:26,  1.55s/it]\r",
      " 84%|████████▍ | 84/100 [02:08<00:25,  1.57s/it]\r",
      " 85%|████████▌ | 85/100 [02:09<00:23,  1.57s/it]\r",
      " 86%|████████▌ | 86/100 [02:11<00:21,  1.54s/it]\r",
      " 87%|████████▋ | 87/100 [02:12<00:20,  1.56s/it]\r",
      " 88%|████████▊ | 88/100 [02:14<00:18,  1.57s/it]\r",
      " 89%|████████▉ | 89/100 [02:15<00:17,  1.56s/it]\r",
      " 90%|█████████ | 90/100 [02:17<00:16,  1.62s/it]\r",
      " 91%|█████████ | 91/100 [02:19<00:14,  1.60s/it]\r",
      " 92%|█████████▏| 92/100 [02:20<00:12,  1.54s/it]\r",
      " 93%|█████████▎| 93/100 [02:21<00:10,  1.49s/it]\r",
      " 94%|█████████▍| 94/100 [02:23<00:08,  1.46s/it]\r",
      " 95%|█████████▌| 95/100 [02:24<00:07,  1.45s/it]\r",
      " 96%|█████████▌| 96/100 [02:26<00:05,  1.42s/it]\r",
      " 97%|█████████▋| 97/100 [02:27<00:04,  1.41s/it]\r",
      " 98%|█████████▊| 98/100 [02:28<00:02,  1.40s/it]\r",
      " 99%|█████████▉| 99/100 [02:30<00:01,  1.40s/it]\r",
      "100%|██████████| 100/100 [02:31<00:00,  1.39s/it]\r",
      "100%|██████████| 100/100 [02:31<00:00,  1.52s/it]\n"
     ]
    }
   ],
   "source": [
    "%%bash\n",
    "cd /Users/chang/Documents/work/flux/scFEA\n",
    "python src/scFEA.py --data_dir data --input_dir input \\\n",
    "                    --test_file GSE72056_full.csv \\\n",
    "                    --moduleGene_file module_gene_iron_m15.csv \\\n",
    "                    --stoichiometry_matrix cmMat_iron_c8_m15.csv \\\n",
    "                    --cName_file cName_iron_c8_m15.csv"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 4.4 example 4 glutaminolysis modules\n",
    "\n",
    "This example run scFEA using GSE72056 on small metabolic system which is glutaminolysis modules."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Starting load data...\n",
      "Load compound name file, the balance output will have compound name.\n",
      "Load data done.\n",
      "Starting process data...\n",
      "Process data done.\n",
      "Starting train neural network...\n",
      "Training time:  154.2574908733368\n",
      "scFEA job finished. Check result in the desired output folder.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  0%|          | 0/100 [00:00<?, ?it/s]\r",
      "  1%|          | 1/100 [00:01<02:48,  1.70s/it]\r",
      "  2%|▏         | 2/100 [00:03<02:46,  1.70s/it]\r",
      "  3%|▎         | 3/100 [00:04<02:41,  1.67s/it]\r",
      "  4%|▍         | 4/100 [00:06<02:43,  1.70s/it]\r",
      "  5%|▌         | 5/100 [00:08<02:38,  1.66s/it]\r",
      "  6%|▌         | 6/100 [00:10<02:42,  1.73s/it]\r",
      "  7%|▋         | 7/100 [00:12<02:46,  1.79s/it]\r",
      "  8%|▊         | 8/100 [00:13<02:39,  1.73s/it]\r",
      "  9%|▉         | 9/100 [00:15<02:34,  1.70s/it]\r",
      " 10%|█         | 10/100 [00:17<02:31,  1.68s/it]\r",
      " 11%|█         | 11/100 [00:18<02:31,  1.70s/it]\r",
      " 12%|█▏        | 12/100 [00:20<02:24,  1.64s/it]\r",
      " 13%|█▎        | 13/100 [00:21<02:20,  1.61s/it]\r",
      " 14%|█▍        | 14/100 [00:23<02:14,  1.57s/it]\r",
      " 15%|█▌        | 15/100 [00:24<02:11,  1.55s/it]\r",
      " 16%|█▌        | 16/100 [00:26<02:09,  1.54s/it]\r",
      " 17%|█▋        | 17/100 [00:27<02:05,  1.52s/it]\r",
      " 18%|█▊        | 18/100 [00:29<02:03,  1.51s/it]\r",
      " 19%|█▉        | 19/100 [00:30<02:04,  1.53s/it]\r",
      " 20%|██        | 20/100 [00:32<02:06,  1.58s/it]\r",
      " 21%|██        | 21/100 [00:34<02:07,  1.62s/it]\r",
      " 22%|██▏       | 22/100 [00:35<02:06,  1.63s/it]\r",
      " 23%|██▎       | 23/100 [00:37<02:14,  1.75s/it]\r",
      " 24%|██▍       | 24/100 [00:39<02:14,  1.77s/it]\r",
      " 25%|██▌       | 25/100 [00:41<02:08,  1.71s/it]\r",
      " 26%|██▌       | 26/100 [00:42<02:02,  1.66s/it]\r",
      " 27%|██▋       | 27/100 [00:44<01:58,  1.63s/it]\r",
      " 28%|██▊       | 28/100 [00:45<01:53,  1.58s/it]\r",
      " 29%|██▉       | 29/100 [00:47<01:49,  1.55s/it]\r",
      " 30%|███       | 30/100 [00:48<01:46,  1.52s/it]\r",
      " 31%|███       | 31/100 [00:50<01:44,  1.52s/it]\r",
      " 32%|███▏      | 32/100 [00:51<01:42,  1.50s/it]\r",
      " 33%|███▎      | 33/100 [00:53<01:39,  1.49s/it]\r",
      " 34%|███▍      | 34/100 [00:54<01:38,  1.50s/it]\r",
      " 35%|███▌      | 35/100 [00:56<01:36,  1.48s/it]\r",
      " 36%|███▌      | 36/100 [00:57<01:34,  1.48s/it]\r",
      " 37%|███▋      | 37/100 [00:59<01:32,  1.47s/it]\r",
      " 38%|███▊      | 38/100 [01:00<01:31,  1.48s/it]\r",
      " 39%|███▉      | 39/100 [01:02<01:30,  1.48s/it]\r",
      " 40%|████      | 40/100 [01:03<01:28,  1.47s/it]\r",
      " 41%|████      | 41/100 [01:05<01:26,  1.47s/it]\r",
      " 42%|████▏     | 42/100 [01:06<01:25,  1.48s/it]\r",
      " 43%|████▎     | 43/100 [01:07<01:24,  1.48s/it]\r",
      " 44%|████▍     | 44/100 [01:09<01:22,  1.47s/it]\r",
      " 45%|████▌     | 45/100 [01:10<01:20,  1.47s/it]\r",
      " 46%|████▌     | 46/100 [01:12<01:20,  1.48s/it]\r",
      " 47%|████▋     | 47/100 [01:13<01:18,  1.48s/it]\r",
      " 48%|████▊     | 48/100 [01:15<01:16,  1.47s/it]\r",
      " 49%|████▉     | 49/100 [01:16<01:14,  1.47s/it]\r",
      " 50%|█████     | 50/100 [01:18<01:13,  1.48s/it]\r",
      " 51%|█████     | 51/100 [01:19<01:12,  1.47s/it]\r",
      " 52%|█████▏    | 52/100 [01:21<01:10,  1.47s/it]\r",
      " 53%|█████▎    | 53/100 [01:22<01:09,  1.48s/it]\r",
      " 54%|█████▍    | 54/100 [01:24<01:07,  1.48s/it]\r",
      " 55%|█████▌    | 55/100 [01:25<01:06,  1.47s/it]\r",
      " 56%|█████▌    | 56/100 [01:27<01:04,  1.47s/it]\r",
      " 57%|█████▋    | 57/100 [01:28<01:03,  1.48s/it]\r",
      " 58%|█████▊    | 58/100 [01:30<01:01,  1.47s/it]\r",
      " 59%|█████▉    | 59/100 [01:31<01:00,  1.48s/it]\r",
      " 60%|██████    | 60/100 [01:33<00:59,  1.48s/it]\r",
      " 61%|██████    | 61/100 [01:34<00:58,  1.49s/it]\r",
      " 62%|██████▏   | 62/100 [01:36<00:56,  1.48s/it]\r",
      " 63%|██████▎   | 63/100 [01:37<00:57,  1.56s/it]\r",
      " 64%|██████▍   | 64/100 [01:39<00:57,  1.60s/it]\r",
      " 65%|██████▌   | 65/100 [01:41<00:57,  1.63s/it]\r",
      " 66%|██████▌   | 66/100 [01:42<00:54,  1.60s/it]\r",
      " 67%|██████▋   | 67/100 [01:44<00:51,  1.56s/it]\r",
      " 68%|██████▊   | 68/100 [01:45<00:49,  1.55s/it]\r",
      " 69%|██████▉   | 69/100 [01:47<00:47,  1.52s/it]\r",
      " 70%|███████   | 70/100 [01:48<00:45,  1.51s/it]\r",
      " 71%|███████   | 71/100 [01:50<00:43,  1.50s/it]\r",
      " 72%|███████▏  | 72/100 [01:51<00:42,  1.52s/it]\r",
      " 73%|███████▎  | 73/100 [01:53<00:40,  1.50s/it]\r",
      " 74%|███████▍  | 74/100 [01:54<00:38,  1.49s/it]\r",
      " 75%|███████▌  | 75/100 [01:56<00:37,  1.48s/it]\r",
      " 76%|███████▌  | 76/100 [01:57<00:35,  1.49s/it]\r",
      " 77%|███████▋  | 77/100 [01:59<00:34,  1.49s/it]\r",
      " 78%|███████▊  | 78/100 [02:00<00:32,  1.48s/it]\r",
      " 79%|███████▉  | 79/100 [02:01<00:31,  1.48s/it]\r",
      " 80%|████████  | 80/100 [02:03<00:30,  1.52s/it]\r",
      " 81%|████████  | 81/100 [02:05<00:28,  1.50s/it]\r",
      " 82%|████████▏ | 82/100 [02:06<00:26,  1.49s/it]\r",
      " 83%|████████▎ | 83/100 [02:07<00:25,  1.48s/it]\r",
      " 84%|████████▍ | 84/100 [02:09<00:23,  1.50s/it]\r",
      " 85%|████████▌ | 85/100 [02:10<00:22,  1.48s/it]\r",
      " 86%|████████▌ | 86/100 [02:12<00:20,  1.47s/it]\r",
      " 87%|████████▋ | 87/100 [02:13<00:19,  1.49s/it]\r",
      " 88%|████████▊ | 88/100 [02:15<00:17,  1.48s/it]\r",
      " 89%|████████▉ | 89/100 [02:16<00:16,  1.48s/it]\r",
      " 90%|█████████ | 90/100 [02:18<00:14,  1.49s/it]\r",
      " 91%|█████████ | 91/100 [02:20<00:14,  1.57s/it]\r",
      " 92%|█████████▏| 92/100 [02:21<00:12,  1.61s/it]\r",
      " 93%|█████████▎| 93/100 [02:23<00:11,  1.64s/it]\r",
      " 94%|█████████▍| 94/100 [02:25<00:09,  1.62s/it]\r",
      " 95%|█████████▌| 95/100 [02:26<00:07,  1.59s/it]\r",
      " 96%|█████████▌| 96/100 [02:28<00:06,  1.55s/it]\r",
      " 97%|█████████▋| 97/100 [02:29<00:04,  1.52s/it]\r",
      " 98%|█████████▊| 98/100 [02:30<00:02,  1.50s/it]\r",
      " 99%|█████████▉| 99/100 [02:32<00:01,  1.53s/it]\r",
      "100%|██████████| 100/100 [02:34<00:00,  1.57s/it]\r",
      "100%|██████████| 100/100 [02:34<00:00,  1.54s/it]\n"
     ]
    }
   ],
   "source": [
    "%%bash\n",
    "cd /Users/chang/Documents/work/flux/scFEA\n",
    "python src/scFEA.py --data_dir data --input_dir input \\\n",
    "                    --test_file GSE72056_full.csv \\\n",
    "                    --moduleGene_file module_gene_glutaminolysis1_m23.csv \\\n",
    "                    --stoichiometry_matrix cmMat_glutaminolysis1_c17_m23.csv \\\n",
    "                    --cName_file cName_glutaminolysis1_c17_m23.csv \\\n",
    "                    --output_flux_file output/GSE72056_gluta_flux.csv \\\n",
    "                    --output_balance_file output/GSE72056_gluta_balance.csv"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 5. Check output files and loss function convergence\n",
    "\n",
    "We could use `ls` command to see the output files after we run glutaminolysis experiment."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "GSE72056_gluta_balance.csv\n",
      "GSE72056_gluta_flux.csv\n",
      "lossValue_20210923-112543.txt\n",
      "loss_20210923-112543.png\n",
      "time_20210923-112543.txt\n"
     ]
    }
   ],
   "source": [
    "%%bash\n",
    "cd /Users/chang/Documents/work/flux/scFEA/output\n",
    "ls"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from IPython.display import Image\n",
    "Image(filename='./scFEA/output/loss_20210923-112543.png') # make sure use correct file with time stamp"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 6. Downstream Analysis\n",
    "\n",
    "(R package scFEA coming soon which includes downstream analysis and visualization of flux)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# use rpy2 to enable R and python in one jupyter notebook\n",
    "%load_ext rpy2.ipython"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "R[write to console]: Picking joint bandwidth of 0.00152\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%%R\n",
    "library(tidyverse)\n",
    "library(rstatix)\n",
    "library(ggpubr)\n",
    "library(reshape)\n",
    "library(ggridges)\n",
    "library(ggplot2)\n",
    "\n",
    "# load predicted flux\n",
    "data_c <- read.csv('./scFEA/output/mouse_flux.csv')\n",
    "data_c0 <- as.matrix(data_c[,-1])\n",
    "rownames(data_c0) <- as.character(data_c[,1])\n",
    "data_c0 <- t(data_c0)\n",
    "ppp_all <-c()\n",
    "\n",
    "# load cell label of mouse example data\n",
    "load('./scFEA/input/mouse_example_cell_ident.RData')\n",
    "yyy <- cell_id[colnames(data_c0)]\n",
    "for(ii in 1:nrow(data_c0)){\n",
    "    xxx <- data_c0[ii,]\n",
    "    final_df <- cbind(paste('X', 1:length(xxx), sep=''), xxx, yyy)\n",
    "    final_df <- as.data.frame(final_df)\n",
    "    final_df[,2] <- as.numeric(final_df[,2])\n",
    "    colnames(final_df) <- c('var', 'flux', 'cellType')\n",
    "    pp <- sd(final_df$flux)/abs(mean(final_df$flux))\n",
    "    ppp_all <- c(ppp_all, pp)\n",
    "}\n",
    "tg_ids <- which(ppp_all > 1e-10)\n",
    "\n",
    "# load mouse module info\n",
    "load('./scFEA/input/mouse_module_info.RData')\n",
    "if(length(tg_ids) > 0){\n",
    "    jj = 1 # check module 2\n",
    "    xxx <- data_c0[tg_ids[jj], ]\n",
    "    #final_df <- cbind(paste('X', 1:length(xxx), sep=''), xxx, yyy)\n",
    "    #final_df <- as.data.frame(final_df)\n",
    "    #final_df[,2] <- as.numeric(final_df[,2])\n",
    "    #colnames(final_df) <- c('var', 'flux', 'cellType')\n",
    "    final_df <- data.frame(var = paste('X', 1:length(xxx), sep=''),\n",
    "                          flux = xxx,\n",
    "                          cellType = yyy)\n",
    "        \n",
    "    title <- mouse_module_info[rownames(data_c0)[tg_ids[jj]], 'M_name']\n",
    "    aa <- ggplot(final_df, aes(x = flux, y = cellType, fill = cellType)) +\n",
    "                  geom_density_ridges() +\n",
    "                  theme_ridges() +\n",
    "                  theme(legend.position = 'none') +\n",
    "                  ggtitle(title) +\n",
    "                  theme(plot.title = element_text(hjust = 0.5))\n",
    "    plot(aa)\n",
    "    \n",
    "}"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}