adage_training.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# ADAGE: Pan-cancer gene expression\n",
    "\n",
    "**Gregory Way 2017**\n",
    "\n",
    "This script trains a denoising autoencoder for cancer gene expression data using Keras. It modifies the framework presented by the ADAGE (Analysis using denoising autoencoders of gene expression) model published by [Tan _et al_ 2015](https://doi.org/10.1128/mSystems.00025-15).\n",
    "\n",
    "An ADAGE model learns a non-linear, reduced dimensional representation of gene expression data by bottlenecking raw features into a smaller set. The model is then trained by minimizing the information lost between input and reconstructed input.\n",
    "\n",
    "The specific model trained in this notebook consists of gene expression input (5000 most variably expressed genes by median absolute deviation) compressed down into one length 100 vector. The hidden layer is then decoded back to the original 5000 dimensions. The encoding (compression) layer has a `relu` activation and the decoding layer has a `sigmoid` activation. The weights of each layer are glorot uniform initialized. We include an l1 regularization term (see [`keras.regularizers.l1`](https://keras.io/regularizers/) for more details) to induce sparsity in the model, as well as a term controlling the probability of input feature dropout. This is only active during training and is the denoising aspect of the model. See [`keras.layers.noise.Dropout`](https://keras.io/layers/core/) for more details.\n",
    "\n",
    "We train the autoencoder with the Adadelta optimizer and MSE reconstruction loss.\n",
    "\n",
    "The pan-cancer ADAGE model is similar to tybalt, but does not constrain the features to match a Gaussian distribution. It is an active research question if the VAE learned features provide any additional benefits over ADAGE features. The VAE is a generative model and therefore permits easy generation of fake data. Additionally, we hypothesize that the VAE learns a manifold that can be interpolated to extract meaningful biological knowledge."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "import os\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "\n",
    "import pydot\n",
    "import graphviz\n",
    "from keras.utils import plot_model\n",
    "from IPython.display import SVG\n",
    "from keras.utils.vis_utils import model_to_dot\n",
    "\n",
    "from keras.layers import Input, Dense, Dropout, Activation\n",
    "from keras.layers.noise import GaussianDropout\n",
    "from keras.models import Model\n",
    "from keras.regularizers import l1\n",
    "from keras import optimizers\n",
    "import keras"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2.0.6\n"
     ]
    }
   ],
   "source": [
    "print(keras.__version__)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "plt.style.use('seaborn-notebook')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "sns.set(style=\"white\", color_codes=True)\n",
    "sns.set_context(\"paper\", rc={\"font.size\":14,\"axes.titlesize\":15,\"axes.labelsize\":20,\n",
    "                             'xtick.labelsize':14, 'ytick.labelsize':14})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(10459, 5000)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>RPS4Y1</th>\n",
       "      <th>XIST</th>\n",
       "      <th>KRT5</th>\n",
       "      <th>AGR2</th>\n",
       "      <th>CEACAM5</th>\n",
       "      <th>KRT6A</th>\n",
       "      <th>KRT14</th>\n",
       "      <th>CEACAM6</th>\n",
       "      <th>DDX3Y</th>\n",
       "      <th>KDM5D</th>\n",
       "      <th>...</th>\n",
       "      <th>FAM129A</th>\n",
       "      <th>C8orf48</th>\n",
       "      <th>CDK5R1</th>\n",
       "      <th>FAM81A</th>\n",
       "      <th>C13orf18</th>\n",
       "      <th>GDPD3</th>\n",
       "      <th>SMAGP</th>\n",
       "      <th>C2orf85</th>\n",
       "      <th>POU5F1B</th>\n",
       "      <th>CHST2</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>TCGA-02-0047-01</th>\n",
       "      <td>0.678296</td>\n",
       "      <td>0.289910</td>\n",
       "      <td>0.034230</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.084731</td>\n",
       "      <td>0.031863</td>\n",
       "      <td>0.037709</td>\n",
       "      <td>0.746797</td>\n",
       "      <td>0.687833</td>\n",
       "      <td>...</td>\n",
       "      <td>0.440610</td>\n",
       "      <td>0.428782</td>\n",
       "      <td>0.732819</td>\n",
       "      <td>0.634340</td>\n",
       "      <td>0.580662</td>\n",
       "      <td>0.294313</td>\n",
       "      <td>0.458134</td>\n",
       "      <td>0.478219</td>\n",
       "      <td>0.168263</td>\n",
       "      <td>0.638497</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>TCGA-02-0055-01</th>\n",
       "      <td>0.200633</td>\n",
       "      <td>0.654917</td>\n",
       "      <td>0.181993</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.100606</td>\n",
       "      <td>0.050011</td>\n",
       "      <td>0.092586</td>\n",
       "      <td>0.103725</td>\n",
       "      <td>0.140642</td>\n",
       "      <td>...</td>\n",
       "      <td>0.620658</td>\n",
       "      <td>0.363207</td>\n",
       "      <td>0.592269</td>\n",
       "      <td>0.602755</td>\n",
       "      <td>0.610192</td>\n",
       "      <td>0.374569</td>\n",
       "      <td>0.722420</td>\n",
       "      <td>0.271356</td>\n",
       "      <td>0.160465</td>\n",
       "      <td>0.602560</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>2 rows × 5000 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                   RPS4Y1      XIST      KRT5  AGR2  CEACAM5     KRT6A  \\\n",
       "TCGA-02-0047-01  0.678296  0.289910  0.034230   0.0      0.0  0.084731   \n",
       "TCGA-02-0055-01  0.200633  0.654917  0.181993   0.0      0.0  0.100606   \n",
       "\n",
       "                    KRT14   CEACAM6     DDX3Y     KDM5D    ...      FAM129A  \\\n",
       "TCGA-02-0047-01  0.031863  0.037709  0.746797  0.687833    ...     0.440610   \n",
       "TCGA-02-0055-01  0.050011  0.092586  0.103725  0.140642    ...     0.620658   \n",
       "\n",
       "                  C8orf48    CDK5R1    FAM81A  C13orf18     GDPD3     SMAGP  \\\n",
       "TCGA-02-0047-01  0.428782  0.732819  0.634340  0.580662  0.294313  0.458134   \n",
       "TCGA-02-0055-01  0.363207  0.592269  0.602755  0.610192  0.374569  0.722420   \n",
       "\n",
       "                  C2orf85   POU5F1B     CHST2  \n",
       "TCGA-02-0047-01  0.478219  0.168263  0.638497  \n",
       "TCGA-02-0055-01  0.271356  0.160465  0.602560  \n",
       "\n",
       "[2 rows x 5000 columns]"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Load RNAseq data\n",
    "rnaseq_file = os.path.join('data', 'pancan_scaled_zeroone_rnaseq.tsv.gz')\n",
    "rnaseq_df = pd.read_table(rnaseq_file, index_col=0)\n",
    "print(rnaseq_df.shape)\n",
    "rnaseq_df.head(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "np.random.seed(123)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Split 10% test set randomly\n",
    "test_set_percent = 0.1\n",
    "rnaseq_test_df = rnaseq_df.sample(frac=test_set_percent)\n",
    "rnaseq_train_df = rnaseq_df.drop(rnaseq_test_df.index)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Parameter Sweep Results\n",
    "\n",
    "We previously performed a parameter sweep search over a grid of potential hyperparameter values. Based on this sweep, we determined that the optimal ADAGE parameters are:\n",
    "\n",
    "| Parameter | Optimal Setting |\n",
    "| :-------: | :-------------: |\n",
    "| Learning Rate | 1.1 |\n",
    "| Sparsity | 0 |\n",
    "| Noise | 0.05 |\n",
    "| Epochs | 100 |\n",
    "| Batch Size | 50 |"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "num_features = rnaseq_df.shape[1]\n",
    "encoding_dim = 100\n",
    "sparsity = 0\n",
    "noise = 0.05\n",
    "epochs = 100\n",
    "batch_size = 50\n",
    "learning_rate = 1.1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# Build the Keras graph\n",
    "input_rnaseq = Input(shape=(num_features, ))\n",
    "encoded_rnaseq = Dropout(noise)(input_rnaseq)\n",
    "encoded_rnaseq_2 = Dense(encoding_dim,\n",
    "                         activity_regularizer=l1(sparsity))(encoded_rnaseq)\n",
    "activation = Activation('relu')(encoded_rnaseq_2)\n",
    "decoded_rnaseq = Dense(num_features, activation='sigmoid')(activation)\n",
    "\n",
    "autoencoder = Model(input_rnaseq, decoded_rnaseq)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "input_1 (InputLayer)         (None, 5000)              0         \n",
      "_________________________________________________________________\n",
      "dropout_1 (Dropout)          (None, 5000)              0         \n",
      "_________________________________________________________________\n",
      "dense_1 (Dense)              (None, 100)               500100    \n",
      "_________________________________________________________________\n",
      "activation_1 (Activation)    (None, 100)               0         \n",
      "_________________________________________________________________\n",
      "dense_2 (Dense)              (None, 5000)              505000    \n",
      "=================================================================\n",
      "Total params: 1,005,100\n",
      "Trainable params: 1,005,100\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "autoencoder.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/svg+xml": [
       "<svg height=\"337pt\" viewBox=\"0.00 0.00 156.00 337.00\" width=\"156pt\" xmlns=\"http://www.w3.org/2000/svg\" xmlns:xlink=\"http://www.w3.org/1999/xlink\">\n",
       "<g class=\"graph\" id=\"graph0\" transform=\"scale(1 1) rotate(0) translate(4 333)\">\n",
       "<title>G</title>\n",
       "<polygon fill=\"white\" points=\"-4,4 -4,-333 152,-333 152,4 -4,4\" stroke=\"none\"/>\n",
       "<!-- 139712716571032 -->\n",
       "<g class=\"node\" id=\"node1\"><title>139712716571032</title>\n",
       "<polygon fill=\"none\" points=\"11.5,-292.5 11.5,-328.5 136.5,-328.5 136.5,-292.5 11.5,-292.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"74\" y=\"-306.8\">input_1: InputLayer</text>\n",
       "</g>\n",
       "<!-- 139712716571256 -->\n",
       "<g class=\"node\" id=\"node2\"><title>139712716571256</title>\n",
       "<polygon fill=\"none\" points=\"11.5,-219.5 11.5,-255.5 136.5,-255.5 136.5,-219.5 11.5,-219.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"74\" y=\"-233.8\">dropout_1: Dropout</text>\n",
       "</g>\n",
       "<!-- 139712716571032&#45;&gt;139712716571256 -->\n",
       "<g class=\"edge\" id=\"edge1\"><title>139712716571032-&gt;139712716571256</title>\n",
       "<path d=\"M74,-292.313C74,-284.289 74,-274.547 74,-265.569\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"77.5001,-265.529 74,-255.529 70.5001,-265.529 77.5001,-265.529\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 139712716702384 -->\n",
       "<g class=\"node\" id=\"node3\"><title>139712716702384</title>\n",
       "<polygon fill=\"none\" points=\"23,-146.5 23,-182.5 125,-182.5 125,-146.5 23,-146.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"74\" y=\"-160.8\">dense_1: Dense</text>\n",
       "</g>\n",
       "<!-- 139712716571256&#45;&gt;139712716702384 -->\n",
       "<g class=\"edge\" id=\"edge2\"><title>139712716571256-&gt;139712716702384</title>\n",
       "<path d=\"M74,-219.313C74,-211.289 74,-201.547 74,-192.569\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"77.5001,-192.529 74,-182.529 70.5001,-192.529 77.5001,-192.529\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 139712716703280 -->\n",
       "<g class=\"node\" id=\"node4\"><title>139712716703280</title>\n",
       "<polygon fill=\"none\" points=\"0,-73.5 0,-109.5 148,-109.5 148,-73.5 0,-73.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"74\" y=\"-87.8\">activation_1: Activation</text>\n",
       "</g>\n",
       "<!-- 139712716702384&#45;&gt;139712716703280 -->\n",
       "<g class=\"edge\" id=\"edge3\"><title>139712716702384-&gt;139712716703280</title>\n",
       "<path d=\"M74,-146.313C74,-138.289 74,-128.547 74,-119.569\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"77.5001,-119.529 74,-109.529 70.5001,-119.529 77.5001,-119.529\" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- 139712716572712 -->\n",
       "<g class=\"node\" id=\"node5\"><title>139712716572712</title>\n",
       "<polygon fill=\"none\" points=\"23,-0.5 23,-36.5 125,-36.5 125,-0.5 23,-0.5\" stroke=\"black\"/>\n",
       "<text font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"74\" y=\"-14.8\">dense_2: Dense</text>\n",
       "</g>\n",
       "<!-- 139712716703280&#45;&gt;139712716572712 -->\n",
       "<g class=\"edge\" id=\"edge4\"><title>139712716703280-&gt;139712716572712</title>\n",
       "<path d=\"M74,-73.3129C74,-65.2895 74,-55.5475 74,-46.5691\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"77.5001,-46.5288 74,-36.5288 70.5001,-46.5289 77.5001,-46.5288\" stroke=\"black\"/>\n",
       "</g>\n",
       "</g>\n",
       "</svg>"
      ],
      "text/plain": [
       "<IPython.core.display.SVG object>"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Visualize the connections of the custom VAE model\n",
    "output_model_file = os.path.join('figures', 'adage_architecture.png')\n",
    "plot_model(autoencoder, to_file=output_model_file)\n",
    "\n",
    "SVG(model_to_dot(autoencoder).create(prog='dot', format='svg'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Separate out the encoder and decoder model\n",
    "encoder = Model(input_rnaseq, encoded_rnaseq_2)\n",
    "\n",
    "encoded_input = Input(shape=(encoding_dim, ))\n",
    "decoder_layer = autoencoder.layers[-1]\n",
    "decoder = Model(encoded_input, decoder_layer(encoded_input))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# Compile the autoencoder to prepare for training\n",
    "adadelta = optimizers.Adadelta(lr=learning_rate)\n",
    "autoencoder.compile(optimizer=adadelta, loss='mse')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 9413 samples, validate on 1046 samples\n",
      "Epoch 1/100\n",
      "9413/9413 [==============================] - 4s - loss: 0.0570 - val_loss: 0.0365\n",
      "Epoch 2/100\n",
      "9413/9413 [==============================] - 4s - loss: 0.0358 - val_loss: 0.0359\n",
      "Epoch 3/100\n",
      "9413/9413 [==============================] - 5s - loss: 0.0355 - val_loss: 0.0357\n",
      "Epoch 4/100\n",
      "9413/9413 [==============================] - 6s - loss: 0.0352 - val_loss: 0.0353\n",
      "Epoch 5/100\n",
      "9413/9413 [==============================] - 5s - loss: 0.0348 - val_loss: 0.0348\n",
      "Epoch 6/100\n",
      "9413/9413 [==============================] - 4s - loss: 0.0342 - val_loss: 0.0341\n",
      "Epoch 7/100\n",
      "9413/9413 [==============================] - 5s - loss: 0.0334 - val_loss: 0.0331\n",
      "Epoch 8/100\n",
      "9413/9413 [==============================] - 4s - loss: 0.0323 - val_loss: 0.0321\n",
      "Epoch 9/100\n",
      "9413/9413 [==============================] - 4s - loss: 0.0313 - val_loss: 0.0311\n",
      "Epoch 10/100\n",
      "9413/9413 [==============================] - 4s - loss: 0.0305 - val_loss: 0.0303\n",
      "Epoch 11/100\n",
      "9413/9413 [==============================] - 4s - loss: 0.0298 - val_loss: 0.0296\n",
      "Epoch 12/100\n",
      "9413/9413 [==============================] - 4s - loss: 0.0292 - val_loss: 0.0291\n",
      "Epoch 13/100\n",
      "9413/9413 [==============================] - 4s - loss: 0.0287 - val_loss: 0.0285\n",
      "Epoch 14/100\n",
      "9413/9413 [==============================] - 4s - loss: 0.0281 - val_loss: 0.0280\n",
      "Epoch 15/100\n",
      "9413/9413 [==============================] - 4s - loss: 0.0276 - val_loss: 0.0274\n",
      "Epoch 16/100\n",
      "9413/9413 [==============================] - 4s - loss: 0.0270 - val_loss: 0.0268\n",
      "Epoch 17/100\n",
      "9413/9413 [==============================] - 4s - loss: 0.0264 - val_loss: 0.0262\n",
      "Epoch 18/100\n",
      "9413/9413 [==============================] - 4s - loss: 0.0258 - val_loss: 0.0257\n",
      "Epoch 19/100\n",
      "9413/9413 [==============================] - 4s - loss: 0.0253 - val_loss: 0.0252\n",
      "Epoch 20/100\n",
      "9413/9413 [==============================] - 4s - loss: 0.0248 - val_loss: 0.0247\n",
      "Epoch 21/100\n",
      "9413/9413 [==============================] - 4s - loss: 0.0244 - val_loss: 0.0243\n",
      "Epoch 22/100\n",
      "9413/9413 [==============================] - 4s - loss: 0.0240 - val_loss: 0.0239\n",
      "Epoch 23/100\n",
      "9413/9413 [==============================] - 4s - loss: 0.0237 - val_loss: 0.0236\n",
      "Epoch 24/100\n",
      "9413/9413 [==============================] - 4s - loss: 0.0233 - val_loss: 0.0233\n",
      "Epoch 25/100\n",
      "9413/9413 [==============================] - 4s - loss: 0.0230 - val_loss: 0.0230\n",
      "Epoch 26/100\n",
      "9413/9413 [==============================] - 4s - loss: 0.0227 - val_loss: 0.0227\n",
      "Epoch 27/100\n",
      "9413/9413 [==============================] - 4s - loss: 0.0225 - val_loss: 0.0224\n",
      "Epoch 28/100\n",
      "9413/9413 [==============================] - 5s - loss: 0.0222 - val_loss: 0.0222\n",
      "Epoch 29/100\n",
      "9413/9413 [==============================] - 5s - loss: 0.0219 - val_loss: 0.0219\n",
      "Epoch 30/100\n",
      "9413/9413 [==============================] - 5s - loss: 0.0217 - val_loss: 0.0217\n",
      "Epoch 31/100\n",
      "9413/9413 [==============================] - 5s - loss: 0.0215 - val_loss: 0.0214\n",
      "Epoch 32/100\n",
      "9413/9413 [==============================] - 6s - loss: 0.0212 - val_loss: 0.0212\n",
      "Epoch 33/100\n",
      "9413/9413 [==============================] - 6s - loss: 0.0210 - val_loss: 0.0209\n",
      "Epoch 34/100\n",
      "9413/9413 [==============================] - 6s - loss: 0.0208 - val_loss: 0.0207\n",
      "Epoch 35/100\n",
      "9413/9413 [==============================] - 6s - loss: 0.0205 - val_loss: 0.0205\n",
      "Epoch 36/100\n",
      "9413/9413 [==============================] - 6s - loss: 0.0203 - val_loss: 0.0203\n",
      "Epoch 37/100\n",
      "9413/9413 [==============================] - 6s - loss: 0.0201 - val_loss: 0.0201\n",
      "Epoch 38/100\n",
      "9413/9413 [==============================] - 6s - loss: 0.0199 - val_loss: 0.0199\n",
      "Epoch 39/100\n",
      "9413/9413 [==============================] - 6s - loss: 0.0197 - val_loss: 0.0197\n",
      "Epoch 40/100\n",
      "9413/9413 [==============================] - 6s - loss: 0.0196 - val_loss: 0.0196\n",
      "Epoch 41/100\n",
      "9413/9413 [==============================] - 6s - loss: 0.0194 - val_loss: 0.0194\n",
      "Epoch 42/100\n",
      "9413/9413 [==============================] - 6s - loss: 0.0192 - val_loss: 0.0192\n",
      "Epoch 43/100\n",
      "9413/9413 [==============================] - 6s - loss: 0.0191 - val_loss: 0.0191\n",
      "Epoch 44/100\n",
      "9413/9413 [==============================] - 6s - loss: 0.0189 - val_loss: 0.0189\n",
      "Epoch 45/100\n",
      "9413/9413 [==============================] - 6s - loss: 0.0188 - val_loss: 0.0188\n",
      "Epoch 46/100\n",
      "9413/9413 [==============================] - 6s - loss: 0.0186 - val_loss: 0.0186\n",
      "Epoch 47/100\n",
      "9413/9413 [==============================] - 6s - loss: 0.0185 - val_loss: 0.0185\n",
      "Epoch 48/100\n",
      "9413/9413 [==============================] - 6s - loss: 0.0183 - val_loss: 0.0184\n",
      "Epoch 49/100\n",
      "9413/9413 [==============================] - 6s - loss: 0.0182 - val_loss: 0.0182\n",
      "Epoch 50/100\n",
      "9413/9413 [==============================] - 6s - loss: 0.0181 - val_loss: 0.0181\n",
      "Epoch 51/100\n",
      "9413/9413 [==============================] - 6s - loss: 0.0179 - val_loss: 0.0180\n",
      "Epoch 52/100\n",
      "9413/9413 [==============================] - 6s - loss: 0.0178 - val_loss: 0.0179\n",
      "Epoch 53/100\n",
      "9413/9413 [==============================] - 4s - loss: 0.0177 - val_loss: 0.0177\n",
      "Epoch 54/100\n",
      "9413/9413 [==============================] - 5s - loss: 0.0176 - val_loss: 0.0176\n",
      "Epoch 55/100\n",
      "9413/9413 [==============================] - 4s - loss: 0.0175 - val_loss: 0.0175\n",
      "Epoch 56/100\n",
      "9413/9413 [==============================] - 6s - loss: 0.0174 - val_loss: 0.0174\n",
      "Epoch 57/100\n",
      "9413/9413 [==============================] - 6s - loss: 0.0173 - val_loss: 0.0173\n",
      "Epoch 58/100\n",
      "9413/9413 [==============================] - 5s - loss: 0.0172 - val_loss: 0.0172\n",
      "Epoch 59/100\n",
      "9413/9413 [==============================] - 6s - loss: 0.0171 - val_loss: 0.0171\n",
      "Epoch 60/100\n",
      "9413/9413 [==============================] - 5s - loss: 0.0170 - val_loss: 0.0170\n",
      "Epoch 61/100\n",
      "9413/9413 [==============================] - 4s - loss: 0.0169 - val_loss: 0.0169\n",
      "Epoch 62/100\n",
      "9413/9413 [==============================] - 4s - loss: 0.0168 - val_loss: 0.0169\n",
      "Epoch 63/100\n",
      "9413/9413 [==============================] - 4s - loss: 0.0167 - val_loss: 0.0168\n",
      "Epoch 64/100\n",
      "9413/9413 [==============================] - 4s - loss: 0.0166 - val_loss: 0.0167\n",
      "Epoch 65/100\n",
      "9413/9413 [==============================] - 4s - loss: 0.0166 - val_loss: 0.0166\n",
      "Epoch 66/100\n",
      "9413/9413 [==============================] - 4s - loss: 0.0165 - val_loss: 0.0165\n",
      "Epoch 67/100\n",
      "9413/9413 [==============================] - 4s - loss: 0.0164 - val_loss: 0.0165\n",
      "Epoch 68/100\n",
      "9413/9413 [==============================] - 4s - loss: 0.0164 - val_loss: 0.0164\n",
      "Epoch 69/100\n",
      "9413/9413 [==============================] - 4s - loss: 0.0163 - val_loss: 0.0163\n",
      "Epoch 70/100\n",
      "9413/9413 [==============================] - 4s - loss: 0.0162 - val_loss: 0.0163\n",
      "Epoch 71/100\n",
      "9413/9413 [==============================] - 4s - loss: 0.0161 - val_loss: 0.0162\n",
      "Epoch 72/100\n",
      "9413/9413 [==============================] - 4s - loss: 0.0161 - val_loss: 0.0161\n",
      "Epoch 73/100\n",
      "9413/9413 [==============================] - 4s - loss: 0.0160 - val_loss: 0.0161\n",
      "Epoch 74/100\n",
      "9413/9413 [==============================] - 4s - loss: 0.0160 - val_loss: 0.0160\n",
      "Epoch 75/100\n",
      "9413/9413 [==============================] - 5s - loss: 0.0159 - val_loss: 0.0160\n",
      "Epoch 76/100\n",
      "9413/9413 [==============================] - 5s - loss: 0.0158 - val_loss: 0.0159\n",
      "Epoch 77/100\n",
      "9413/9413 [==============================] - 4s - loss: 0.0158 - val_loss: 0.0158\n",
      "Epoch 78/100\n",
      "9413/9413 [==============================] - 4s - loss: 0.0157 - val_loss: 0.0158\n",
      "Epoch 79/100\n",
      "9413/9413 [==============================] - 4s - loss: 0.0157 - val_loss: 0.0157\n",
      "Epoch 80/100\n",
      "9413/9413 [==============================] - 4s - loss: 0.0156 - val_loss: 0.0157\n",
      "Epoch 81/100\n",
      "9413/9413 [==============================] - 4s - loss: 0.0156 - val_loss: 0.0156\n",
      "Epoch 82/100\n",
      "9413/9413 [==============================] - 4s - loss: 0.0155 - val_loss: 0.0156\n",
      "Epoch 83/100\n",
      "9413/9413 [==============================] - 4s - loss: 0.0155 - val_loss: 0.0155\n",
      "Epoch 84/100\n",
      "9413/9413 [==============================] - 4s - loss: 0.0154 - val_loss: 0.0155\n",
      "Epoch 85/100\n",
      "9413/9413 [==============================] - 4s - loss: 0.0154 - val_loss: 0.0154\n",
      "Epoch 86/100\n",
      "9413/9413 [==============================] - 4s - loss: 0.0153 - val_loss: 0.0154\n",
      "Epoch 87/100\n",
      "9413/9413 [==============================] - 4s - loss: 0.0153 - val_loss: 0.0153\n",
      "Epoch 88/100\n",
      "9413/9413 [==============================] - 4s - loss: 0.0152 - val_loss: 0.0153\n",
      "Epoch 89/100\n",
      "9413/9413 [==============================] - 4s - loss: 0.0152 - val_loss: 0.0153\n",
      "Epoch 90/100\n",
      "9413/9413 [==============================] - 4s - loss: 0.0151 - val_loss: 0.0152\n",
      "Epoch 91/100\n",
      "9413/9413 [==============================] - 4s - loss: 0.0151 - val_loss: 0.0152\n",
      "Epoch 92/100\n",
      "9413/9413 [==============================] - 5s - loss: 0.0151 - val_loss: 0.0151\n",
      "Epoch 93/100\n",
      "9413/9413 [==============================] - 5s - loss: 0.0150 - val_loss: 0.0151\n",
      "Epoch 94/100\n",
      "9413/9413 [==============================] - 5s - loss: 0.0150 - val_loss: 0.0150\n",
      "Epoch 95/100\n",
      "9413/9413 [==============================] - 5s - loss: 0.0149 - val_loss: 0.0150\n",
      "Epoch 96/100\n",
      "9413/9413 [==============================] - 5s - loss: 0.0149 - val_loss: 0.0150\n",
      "Epoch 97/100\n",
      "9413/9413 [==============================] - 5s - loss: 0.0149 - val_loss: 0.0149\n",
      "Epoch 98/100\n",
      "9413/9413 [==============================] - 5s - loss: 0.0148 - val_loss: 0.0149\n",
      "Epoch 99/100\n",
      "9413/9413 [==============================] - 5s - loss: 0.0148 - val_loss: 0.0148\n",
      "Epoch 100/100\n",
      "9413/9413 [==============================] - 5s - loss: 0.0147 - val_loss: 0.0148\n",
      "CPU times: user 17min 56s, sys: 2min 47s, total: 20min 43s\n",
      "Wall time: 8min 46s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "hist = autoencoder.fit(np.array(rnaseq_train_df), np.array(rnaseq_train_df),\n",
    "                       shuffle=True,\n",
    "                       epochs=epochs,\n",
    "                       batch_size=batch_size,\n",
    "                       validation_data=(np.array(rnaseq_test_df), np.array(rnaseq_test_df)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAggAAAFjCAYAAABYJ/NnAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nkRQKCCIiIpJCAUFERERS/D/5kaFIw7oaNAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f116fb0a208>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Visualize training performance\n",
    "history_df = pd.DataFrame(hist.history)\n",
    "hist_plot_file = os.path.join('figures', 'adage_training.png')\n",
    "ax = history_df.plot()\n",
    "ax.set_xlabel('Epochs')\n",
    "ax.set_ylabel('ADAGE Reconstruction Loss')\n",
    "fig = ax.get_figure()\n",
    "fig.savefig(hist_plot_file)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Save Model Outputs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Encode rnaseq into the hidden/latent representation - and save output\n",
    "encoded_samples = encoder.predict(np.array(rnaseq_df))\n",
    "encoded_rnaseq_df = pd.DataFrame(encoded_samples, index=rnaseq_df.index)\n",
    "\n",
    "encoded_rnaseq_df.columns.name = 'sample_id'\n",
    "encoded_rnaseq_df.columns = encoded_rnaseq_df.columns + 1\n",
    "encoded_file = os.path.join('data', 'encoded_adage_features.tsv')\n",
    "encoded_rnaseq_df.to_csv(encoded_file, sep='\\t')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Output weight matrix of gene contributions per node\n",
    "weight_file = os.path.join('results', 'adage_gene_weights.tsv')\n",
    "\n",
    "weight_matrix = pd.DataFrame(autoencoder.get_weights()[0], index=rnaseq_df.columns,\n",
    "                             columns=range(1, 101)).T\n",
    "weight_matrix.index.name = 'encodings'\n",
    "weight_matrix.to_csv(weight_file, sep='\\t')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Observe reconstruction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# Reconstruct input RNAseq\n",
    "decoded_samples = decoder.predict(encoded_samples)\n",
    "\n",
    "reconstructed_df = pd.DataFrame(decoded_samples, index=rnaseq_df.index,\n",
    "                                columns=rnaseq_df.columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>gene mean</th>\n",
       "      <th>gene abs(sum)</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>PPAN-P2RY11</th>\n",
       "      <td>-0.001976</td>\n",
       "      <td>0.239137</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>GSTT1</th>\n",
       "      <td>0.023419</td>\n",
       "      <td>0.232442</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>GSTM1</th>\n",
       "      <td>0.005316</td>\n",
       "      <td>0.219905</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>EIF1AY</th>\n",
       "      <td>-0.048686</td>\n",
       "      <td>0.209455</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DDX3Y</th>\n",
       "      <td>-0.034089</td>\n",
       "      <td>0.207383</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             gene mean  gene abs(sum)\n",
       "PPAN-P2RY11  -0.001976       0.239137\n",
       "GSTT1         0.023419       0.232442\n",
       "GSTM1         0.005316       0.219905\n",
       "EIF1AY       -0.048686       0.209455\n",
       "DDX3Y        -0.034089       0.207383"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "reconstruction_fidelity = rnaseq_df - reconstructed_df\n",
    "\n",
    "gene_mean = reconstruction_fidelity.mean(axis=0)\n",
    "gene_abssum = reconstruction_fidelity.abs().sum(axis=0).divide(rnaseq_df.shape[0])\n",
    "gene_summary = pd.DataFrame([gene_mean, gene_abssum], index=['gene mean', 'gene abs(sum)']).T\n",
    "gene_summary.sort_values(by='gene abs(sum)', ascending=False).head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAacAAAGfCAYAAAD7xh1YAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nERHZHAYnIiKyOQxORERkc/4fJ/xwgH+BJPMAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f1167998978>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Mean of gene reconstruction vs. absolute reconstructed difference per sample\n",
    "reconstruct_fig_file = os.path.join('figures', 'adage_gene_reconstruction.png')\n",
    "g = sns.jointplot('gene mean', 'gene abs(sum)', data=gene_summary, stat_func=None);\n",
    "g.savefig(reconstruct_fig_file)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "sample_id\n",
      "32    116126.578125\n",
      "16    111204.562500\n",
      "80     97674.945312\n",
      "66     96646.546875\n",
      "40     95180.679688\n",
      "dtype: float32\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "sample_id\n",
       "25   -11347.204102\n",
       "24   -12289.624023\n",
       "87   -12757.212891\n",
       "41   -13314.549805\n",
       "49   -16946.775391\n",
       "dtype: float32"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# What are the most and least activated nodes\n",
    "sum_node_activity = encoded_rnaseq_df.sum(axis=0).sort_values(ascending=False)\n",
    "\n",
    "# Top 5 most active nodes\n",
    "print(sum_node_activity.head(5))\n",
    "\n",
    "# Bottom 10 least active nodes\n",
    "sum_node_activity.tail(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAg0AAAFjCAYAAAC+DjgjAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nALCE0AAAACwhNAAAAEv+P8iyoHQeIde1AAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f116e636ef0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Histogram of node activity for all 100 latent features\n",
    "sum_node_activity.hist()\n",
    "plt.xlabel('Activation Sum')\n",
    "plt.ylabel('Count');"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAgwAAAFjCAYAAABRzFMdAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f1167d653c8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Histogram of sample activity for all 10,000 samples\n",
    "encoded_rnaseq_df.sum(axis=1).hist()\n",
    "plt.xlabel('Sample Total Activation')\n",
    "plt.ylabel('Count');"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYIAAAF/CAYAAABE9s5OAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f1167c74ba8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Example of node activation distribution for the first two latent features\n",
    "plt.figure(figsize=(6, 6))\n",
    "plt.scatter(encoded_rnaseq_df.iloc[:, 1], encoded_rnaseq_df.iloc[:, 2])\n",
    "plt.xlabel('Latent Feature 1')\n",
    "plt.ylabel('Latent Feature 2');"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python [default]",
   "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": "3.5.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}