01_usage.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "713d9b4a",
   "metadata": {},
   "source": [
    "# Usage\n",
    "\n",
    "`decoupler` contains different statistical methods to infer enrichment scores from omics data using prior knowledge. In this notebook we showcase how to use it with some toy data."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "63e9268b-add2-465a-a6f2-0294a16f1d8a",
   "metadata": {},
   "source": [
    "decouplerには、事前知識を用いてオミックスデータからエンリッチメントスコアを推測するための様々な統計的手法が含まれている。このノートブックでは、おもちゃのデータを使った使い方を紹介する。"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2337c9d3",
   "metadata": {},
   "source": [
    "## Loading packages\n",
    "\n",
    "`decoupler` can be imported as:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "b7e99394-7575-49c7-9eee-68d8757a6c87",
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "f31c41bb-0711-422e-b17c-230bba387c39",
   "metadata": {},
   "outputs": [],
   "source": [
    "import seaborn as sns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "8dd66119-759b-40fa-b31e-e3c75d0d90b6",
   "metadata": {},
   "outputs": [],
   "source": [
    "import decoupler as dc"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3316c8ad",
   "metadata": {},
   "source": [
    "## Loading toy data\n",
    "\n",
    "`decoupler` needs a matrix (`mat`) of molecular readouts (gene expression, logFC, p-values, etc.) and a network (`net`) that relates target features (genes, proteins, etc.) to \"source\" biological entities (pathways, transcription factors, molecular processes, etc.)\n",
    "\n",
    "To load the example data-set, simply run:"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "62cf961d-96b8-4bb5-a644-5795cb3d0019",
   "metadata": {},
   "source": [
    "decouplerが必要とするのは、分子の読み出し値（遺伝子発現、logFC、p値など）のマトリックス（mat）と、ターゲットの特徴（遺伝子、タンパク質など）を「ソース」生物学的実体（パスウェイ、転写因子、分子プロセスなど）に関連付けるネットワーク（net）である。\n",
    "\n",
    "サンプルのデータセットを読み込むには、次のように実行するだけでよい："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "1a85cb79",
   "metadata": {},
   "outputs": [],
   "source": [
    "mat, net = dc.get_toy_data()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "cddc3630-4f02-4fd2-b777-1a4b37cc6a1c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>G01</th>\n",
       "      <th>G02</th>\n",
       "      <th>G03</th>\n",
       "      <th>G04</th>\n",
       "      <th>G05</th>\n",
       "      <th>G06</th>\n",
       "      <th>G07</th>\n",
       "      <th>G08</th>\n",
       "      <th>G09</th>\n",
       "      <th>G10</th>\n",
       "      <th>G11</th>\n",
       "      <th>G12</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>S01</th>\n",
       "      <td>8.304717</td>\n",
       "      <td>9.039984</td>\n",
       "      <td>8.750451</td>\n",
       "      <td>8.940565</td>\n",
       "      <td>1.951035</td>\n",
       "      <td>1.302180</td>\n",
       "      <td>0.127840</td>\n",
       "      <td>0.316243</td>\n",
       "      <td>0.016801</td>\n",
       "      <td>0.853044</td>\n",
       "      <td>0.879398</td>\n",
       "      <td>0.777792</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>S02</th>\n",
       "      <td>8.066031</td>\n",
       "      <td>9.127241</td>\n",
       "      <td>8.467509</td>\n",
       "      <td>8.859292</td>\n",
       "      <td>0.368751</td>\n",
       "      <td>0.958883</td>\n",
       "      <td>0.878450</td>\n",
       "      <td>0.049926</td>\n",
       "      <td>0.184862</td>\n",
       "      <td>0.680930</td>\n",
       "      <td>1.222541</td>\n",
       "      <td>0.154529</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>S03</th>\n",
       "      <td>8.428328</td>\n",
       "      <td>8.352134</td>\n",
       "      <td>8.532309</td>\n",
       "      <td>8.365444</td>\n",
       "      <td>0.412733</td>\n",
       "      <td>0.430821</td>\n",
       "      <td>2.141648</td>\n",
       "      <td>0.406415</td>\n",
       "      <td>0.512243</td>\n",
       "      <td>0.813773</td>\n",
       "      <td>0.615979</td>\n",
       "      <td>1.128972</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>S04</th>\n",
       "      <td>8.113947</td>\n",
       "      <td>8.840156</td>\n",
       "      <td>8.824481</td>\n",
       "      <td>8.650593</td>\n",
       "      <td>0.743254</td>\n",
       "      <td>0.543154</td>\n",
       "      <td>0.665510</td>\n",
       "      <td>0.232161</td>\n",
       "      <td>0.116686</td>\n",
       "      <td>0.218689</td>\n",
       "      <td>0.871429</td>\n",
       "      <td>0.223596</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>S05</th>\n",
       "      <td>8.678914</td>\n",
       "      <td>8.067579</td>\n",
       "      <td>8.289119</td>\n",
       "      <td>8.631288</td>\n",
       "      <td>1.457156</td>\n",
       "      <td>0.319671</td>\n",
       "      <td>0.470373</td>\n",
       "      <td>0.638878</td>\n",
       "      <td>0.275142</td>\n",
       "      <td>1.494941</td>\n",
       "      <td>0.865831</td>\n",
       "      <td>0.968278</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          G01       G02       G03       G04       G05       G06       G07  \\\n",
       "S01  8.304717  9.039984  8.750451  8.940565  1.951035  1.302180  0.127840   \n",
       "S02  8.066031  9.127241  8.467509  8.859292  0.368751  0.958883  0.878450   \n",
       "S03  8.428328  8.352134  8.532309  8.365444  0.412733  0.430821  2.141648   \n",
       "S04  8.113947  8.840156  8.824481  8.650593  0.743254  0.543154  0.665510   \n",
       "S05  8.678914  8.067579  8.289119  8.631288  1.457156  0.319671  0.470373   \n",
       "\n",
       "          G08       G09       G10       G11       G12  \n",
       "S01  0.316243  0.016801  0.853044  0.879398  0.777792  \n",
       "S02  0.049926  0.184862  0.680930  1.222541  0.154529  \n",
       "S03  0.406415  0.512243  0.813773  0.615979  1.128972  \n",
       "S04  0.232161  0.116686  0.218689  0.871429  0.223596  \n",
       "S05  0.638878  0.275142  1.494941  0.865831  0.968278  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mat.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "ae12e78b-a7ff-450f-a02d-3bee9ff98ed2",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(24, 12)"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mat.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "1ce10c00-b116-44f1-b046-883b1877beb7",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>source</th>\n",
       "      <th>target</th>\n",
       "      <th>weight</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>T1</td>\n",
       "      <td>G01</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>T1</td>\n",
       "      <td>G02</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>T1</td>\n",
       "      <td>G03</td>\n",
       "      <td>0.7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>T2</td>\n",
       "      <td>G04</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>T2</td>\n",
       "      <td>G06</td>\n",
       "      <td>-0.5</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  source target  weight\n",
       "0     T1    G01     1.0\n",
       "1     T1    G02     1.0\n",
       "2     T1    G03     0.7\n",
       "3     T2    G04     1.0\n",
       "4     T2    G06    -0.5"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "net.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "3f6ea3b6-cf0c-4af0-b76e-5ad6548935e9",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(17, 3)"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "net.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "59856674",
   "metadata": {},
   "source": [
    "This example consists of two small populations of samples (S, rows) with different gene expression patterns (G, columns):"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d521086d-016a-4cf1-9b27-a949dd8052d8",
   "metadata": {},
   "source": [
    "この例は、異なる遺伝子発現パターン（G、列）を持つ2つの小さなサンプル集団（S、行）からなる："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "80c04914-4939-48e9-80dd-b19395292b01",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 300x300 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(\n",
    "    1, 1, figsize=(3, 3), tight_layout=True, sharey=True\n",
    ")\n",
    "\n",
    "sns.heatmap(mat, cmap='viridis', ax=ax)\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6898f90c",
   "metadata": {},
   "source": [
    "Here we can see that some genes seem to be more expressed in one group of samples than in the other and vice-versa. Ideally, we would like to capture these differences in gene programs into interpretable biological entities. In this example we will do it by summarizing gene expression into transcription factor activities."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ab820efd",
   "metadata": {},
   "source": [
    "The toy data also contains a simple `net` consisting of 5 transcription factors (Ts) with specific regulation to target genes (Gs), either positive or negative:"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "605ac9f5-e57e-408a-b324-91075bce6f7d",
   "metadata": {},
   "source": [
    "ここで、ある遺伝子はあるサンプル群では他のサンプル群よりも発現量が多く、またその逆もあることがわかる。理想的には、このような遺伝子プログラムの違いを、解釈可能な生物学的実体としてとらえたい。この例では、遺伝子発現を転写因子活性に要約することでそれを実現する。\n",
    "\n",
    "このおもちゃのデータには、5つの転写因子（Ts）からなる単純なネットが含まれており、標的遺伝子（Gs）に対して正または負のどちらかの特異的な制御が行われている："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "184c02ee",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>source</th>\n",
       "      <th>target</th>\n",
       "      <th>weight</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>T1</td>\n",
       "      <td>G01</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>T1</td>\n",
       "      <td>G02</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>T1</td>\n",
       "      <td>G03</td>\n",
       "      <td>0.7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>T2</td>\n",
       "      <td>G04</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>T2</td>\n",
       "      <td>G06</td>\n",
       "      <td>-0.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>T2</td>\n",
       "      <td>G07</td>\n",
       "      <td>-3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>T2</td>\n",
       "      <td>G08</td>\n",
       "      <td>-1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>T3</td>\n",
       "      <td>G06</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>T3</td>\n",
       "      <td>G07</td>\n",
       "      <td>0.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>T3</td>\n",
       "      <td>G08</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>T4</td>\n",
       "      <td>G05</td>\n",
       "      <td>1.9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>T4</td>\n",
       "      <td>G10</td>\n",
       "      <td>-1.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>T4</td>\n",
       "      <td>G11</td>\n",
       "      <td>-2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>T4</td>\n",
       "      <td>G09</td>\n",
       "      <td>3.1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>T5</td>\n",
       "      <td>G09</td>\n",
       "      <td>0.7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>T5</td>\n",
       "      <td>G10</td>\n",
       "      <td>1.1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>T5</td>\n",
       "      <td>G11</td>\n",
       "      <td>0.1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   source target  weight\n",
       "0      T1    G01     1.0\n",
       "1      T1    G02     1.0\n",
       "2      T1    G03     0.7\n",
       "3      T2    G04     1.0\n",
       "4      T2    G06    -0.5\n",
       "5      T2    G07    -3.0\n",
       "6      T2    G08    -1.0\n",
       "7      T3    G06     1.0\n",
       "8      T3    G07     0.5\n",
       "9      T3    G08     1.0\n",
       "10     T4    G05     1.9\n",
       "11     T4    G10    -1.5\n",
       "12     T4    G11    -2.0\n",
       "13     T4    G09     3.1\n",
       "14     T5    G09     0.7\n",
       "15     T5    G10     1.1\n",
       "16     T5    G11     0.1"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "net"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2f5a37e7",
   "metadata": {},
   "source": [
    "This network can be visualized like a graph. Green edges are positive regulation (activation), red edges are negative regulation (inactivation):"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9f24fec7-1ed1-4bcc-b113-52c8b62761e2",
   "metadata": {},
   "source": [
    "このネットワークはグラフのように視覚化できる。緑のエッジは正の調節（活性化）、赤のエッジは負の調節（不活性化）である："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "b42dce4d-6725-4798-acf4-c8e365b9c3c9",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 600x600 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "dc.plot_network(\n",
    "    net,\n",
    "    node_size=50,\n",
    "    figsize=(4, 4),\n",
    "    s_cmap='white',\n",
    "    t_cmap='white',\n",
    "    c_pos_w='darkgreen',\n",
    "    c_neg_w='darkred'\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e2eba9a7",
   "metadata": {},
   "source": [
    "According to this network, the first population of samples should show high activity for T1 and T2, while the second one for T3 and T4. T5 should have no activity in all samples."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "83d3f243-6bb8-4504-9ac5-47483ee1f7b8",
   "metadata": {},
   "source": [
    "このネットワークによると、最初のサンプル集団はT1とT2に高い活性を示し、2番目のサンプル集団はT3とT4に高い活性を示すはずである。T5はすべてのサンプルで活性がないはずである。"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0adc4371",
   "metadata": {},
   "source": [
    "## Methods\n",
    "\n",
    "`decoupler` contains several methods. To check how many are available, run:"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "98c76aac-9ff3-4756-816c-631bfb6e93b1",
   "metadata": {},
   "source": [
    "decouplerにはいくつかのメソッドがある。使用可能なメソッドの数を確認するには、以下を実行する："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "8d022899",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Function</th>\n",
       "      <th>Name</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>run_aucell</td>\n",
       "      <td>AUCell.</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>run_consensus</td>\n",
       "      <td>Consensus score from top methods.</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>run_gsea</td>\n",
       "      <td>Gene Set Enrichment Analysis (GSEA).</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>run_gsva</td>\n",
       "      <td>Gene Set Variation Analysis (GSVA).</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>run_mdt</td>\n",
       "      <td>Multivariate Decision Tree (MDT).</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>run_mlm</td>\n",
       "      <td>Multivariate Linear Model (MLM).</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>run_ora</td>\n",
       "      <td>Over Representation Analysis (ORA).</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>run_udt</td>\n",
       "      <td>Univariate Decision Tree (UDT).</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>run_ulm</td>\n",
       "      <td>Univariate Linear Model (ULM).</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>run_viper</td>\n",
       "      <td>Virtual Inference of Protein-activity by Enric...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>run_wmean</td>\n",
       "      <td>Weighted mean (WMEAN).</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>run_wsum</td>\n",
       "      <td>Weighted sum (WSUM).</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         Function                                               Name\n",
       "0      run_aucell                                            AUCell.\n",
       "1   run_consensus                  Consensus score from top methods.\n",
       "2        run_gsea               Gene Set Enrichment Analysis (GSEA).\n",
       "3        run_gsva                Gene Set Variation Analysis (GSVA).\n",
       "4         run_mdt                  Multivariate Decision Tree (MDT).\n",
       "5         run_mlm                   Multivariate Linear Model (MLM).\n",
       "6         run_ora                Over Representation Analysis (ORA).\n",
       "7         run_udt                    Univariate Decision Tree (UDT).\n",
       "8         run_ulm                     Univariate Linear Model (ULM).\n",
       "9       run_viper  Virtual Inference of Protein-activity by Enric...\n",
       "10      run_wmean                             Weighted mean (WMEAN).\n",
       "11       run_wsum                               Weighted sum (WSUM)."
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dc.show_methods()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f64f1bdc",
   "metadata": {},
   "source": [
    "Each method models biological activities in a different manner, sometimes returning more than one estimate or providing significance of the estimation. To know what each method returns, please check their documentation like this `?run_mlm`."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "590e371b",
   "metadata": {},
   "source": [
    "To have a unified framework, methods inside `decoupler` have these shared arguments:\n",
    "\n",
    "* `mat` : input matrix of molecular readouts.\n",
    "* `net` : input prior knowledge information relating molecular features to biological entities.\n",
    "* `source`,`target` and `weight` : column names where to extract the information from `net`. \n",
    "    * `source` refers to the biological entities.\n",
    "    * `target` refers to the molecular features.\n",
    "    * `weight` refers to the \"strength\" of the interaction (if available, else 1s will be used). Only available for methods that can model interaction weights.\n",
    "* `min_n` : Minimum of target features per biological entity (5 by default). If less, sources are removed. This filtering prevents obtaining noisy activities from biological entities with very few matching target features in `mat`. For this example data-set we will have to keep it to 0 though.\n",
    "* `verbose` : Whether to show progress.\n",
    "* `use_raw` : When the input is an `AnnData` object, whether to use the data stored in it's `.raw` atribute or not (`True` by default)."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0f2cb87d-d0a8-442a-b35f-abc29b87cb5d",
   "metadata": {},
   "source": [
    "各手法は異なる方法で生物学的活動をモデル化し、時には複数の推定値を返したり、推定値の有意性を示したりする。各メソッドが何を返すかを知るには、このようなドキュメントをチェックしてほしい。\n",
    "\n",
    "統一されたフレームワークを持つために、decouplerの中のメソッドは以下の引数を共有している：\n",
    "\n",
    "* mat : 分子の読み取り値の入力行列。\n",
    "* net : 分子の特徴と生物学的実体を関連付ける事前知識情報の入力。\n",
    "* source、target、weight : netから情報を抽出する列名。\n",
    "    * sourceは生物学的実体を指す。\n",
    "    * targetは分子の特徴を指す。\n",
    "    * weightは相互作用の強さを示す（利用可能な場合、それ以外は1が使用される）。相互作用の重みをモデル化できるメソッドでのみ利用可能。\n",
    "* min_n : 生物学的実体ごとのtarget featuresの最小値（デフォルトは5）。それ以下の場合、ソースは削除される。このフィルタリングにより、matに一致するターゲット特徴がほとんどない生物学的実体から、ノイズの多いアクティビティが得られるのを防ぐことができる。このサンプルのデータセットでは、0にしておく必要がある。\n",
    "* verbose : 進捗を表示するかどうか。\n",
    "* use_raw : 入力が AnnData オブジェクトの場合、.raw 属性に格納されているデータを使用するかどうか（デフォルトは True）。"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0f85f343",
   "metadata": {},
   "source": [
    "## Running methods\n",
    "\n",
    "### Individual methods\n",
    "\n",
    "As an example, let's first run the Gene Set Enrichment Analysis method (`gsea`), one of the most well-known statistics:"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d36a1ac0-4717-4c78-a207-38639b146db0",
   "metadata": {},
   "source": [
    "例として、最もよく知られた統計の一つである遺伝子セット濃縮解析法（gsea）をまず実行してみよう："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "16e359a3-7804-44ef-b51a-94282918be61",
   "metadata": {},
   "outputs": [],
   "source": [
    "# For this toy data, we need to set min_n to 0 (otherwise it is better to keep at 5):\n",
    "acts, norm_acts, pvals = dc.run_gsea(mat, net, min_n=0, times=100)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "75ccd460",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th>source</th>\n",
       "      <th>T1</th>\n",
       "      <th>T2</th>\n",
       "      <th>T3</th>\n",
       "      <th>T4</th>\n",
       "      <th>T5</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>S01</th>\n",
       "      <td>0.888889</td>\n",
       "      <td>0.711597</td>\n",
       "      <td>-0.555556</td>\n",
       "      <td>-0.50</td>\n",
       "      <td>-0.666667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>S02</th>\n",
       "      <td>0.888889</td>\n",
       "      <td>0.699385</td>\n",
       "      <td>-0.555556</td>\n",
       "      <td>-0.50</td>\n",
       "      <td>-0.444444</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>S03</th>\n",
       "      <td>0.888889</td>\n",
       "      <td>0.551198</td>\n",
       "      <td>-0.444444</td>\n",
       "      <td>-0.75</td>\n",
       "      <td>-0.666667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>S04</th>\n",
       "      <td>0.888889</td>\n",
       "      <td>0.607223</td>\n",
       "      <td>-0.666667</td>\n",
       "      <td>-0.50</td>\n",
       "      <td>-0.444444</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>S05</th>\n",
       "      <td>0.888889</td>\n",
       "      <td>0.732963</td>\n",
       "      <td>-0.888889</td>\n",
       "      <td>-0.50</td>\n",
       "      <td>-0.444444</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "source        T1        T2        T3    T4        T5\n",
       "S01     0.888889  0.711597 -0.555556 -0.50 -0.666667\n",
       "S02     0.888889  0.699385 -0.555556 -0.50 -0.444444\n",
       "S03     0.888889  0.551198 -0.444444 -0.75 -0.666667\n",
       "S04     0.888889  0.607223 -0.666667 -0.50 -0.444444\n",
       "S05     0.888889  0.732963 -0.888889 -0.50 -0.444444"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "acts.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "5820c63a-cb11-4b5c-bb86-88089f557044",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(24, 5)"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "acts.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "d315e847-6d80-4631-a131-f32de07762df",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th>source</th>\n",
       "      <th>T1</th>\n",
       "      <th>T2</th>\n",
       "      <th>T3</th>\n",
       "      <th>T4</th>\n",
       "      <th>T5</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>S01</th>\n",
       "      <td>1.234678</td>\n",
       "      <td>1.011988</td>\n",
       "      <td>-1.018518</td>\n",
       "      <td>-0.825397</td>\n",
       "      <td>-1.096154</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>S02</th>\n",
       "      <td>1.202222</td>\n",
       "      <td>0.929855</td>\n",
       "      <td>-0.967742</td>\n",
       "      <td>-0.780488</td>\n",
       "      <td>-0.732824</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>S03</th>\n",
       "      <td>1.216594</td>\n",
       "      <td>0.776507</td>\n",
       "      <td>-0.711656</td>\n",
       "      <td>-1.200000</td>\n",
       "      <td>-1.086207</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>S04</th>\n",
       "      <td>1.157576</td>\n",
       "      <td>0.820708</td>\n",
       "      <td>-1.100000</td>\n",
       "      <td>-0.800000</td>\n",
       "      <td>-0.759124</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>S05</th>\n",
       "      <td>1.254691</td>\n",
       "      <td>1.049219</td>\n",
       "      <td>-1.518248</td>\n",
       "      <td>-0.876712</td>\n",
       "      <td>-0.724409</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "source        T1        T2        T3        T4        T5\n",
       "S01     1.234678  1.011988 -1.018518 -0.825397 -1.096154\n",
       "S02     1.202222  0.929855 -0.967742 -0.780488 -0.732824\n",
       "S03     1.216594  0.776507 -0.711656 -1.200000 -1.086207\n",
       "S04     1.157576  0.820708 -1.100000 -0.800000 -0.759124\n",
       "S05     1.254691  1.049219 -1.518248 -0.876712 -0.724409"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "norm_acts.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "018650c6-9c56-43b8-9e35-68c59d75092b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(24, 5)"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "norm_acts.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "e2a50fea-4f4d-42ad-a991-017b5fe437fa",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th>source</th>\n",
       "      <th>T1</th>\n",
       "      <th>T2</th>\n",
       "      <th>T3</th>\n",
       "      <th>T4</th>\n",
       "      <th>T5</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>S01</th>\n",
       "      <td>0.144737</td>\n",
       "      <td>0.426966</td>\n",
       "      <td>0.454545</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.421053</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>S02</th>\n",
       "      <td>0.144737</td>\n",
       "      <td>0.717647</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>S03</th>\n",
       "      <td>0.062500</td>\n",
       "      <td>0.850000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.272727</td>\n",
       "      <td>0.380952</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>S04</th>\n",
       "      <td>0.125000</td>\n",
       "      <td>0.873563</td>\n",
       "      <td>0.363636</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>S05</th>\n",
       "      <td>0.120000</td>\n",
       "      <td>0.360465</td>\n",
       "      <td>0.076923</td>\n",
       "      <td>0.875000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "source        T1        T2        T3        T4        T5\n",
       "S01     0.144737  0.426966  0.454545  1.000000  0.421053\n",
       "S02     0.144737  0.717647  0.666667  1.000000  1.000000\n",
       "S03     0.062500  0.850000  1.000000  0.272727  0.380952\n",
       "S04     0.125000  0.873563  0.363636  1.000000  1.000000\n",
       "S05     0.120000  0.360465  0.076923  0.875000  1.000000"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pvals.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "e0275d87-394d-4903-b8a9-cdbcd4763698",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(24, 5)"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pvals.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ffc2fd1f",
   "metadata": {},
   "source": [
    "In the case of `gsea`, it returns a simple estimate of activities (`acts`), a normalised estimate (`norm_acts`) and `pvals` data-frames after doing permutations."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4f11b1d4",
   "metadata": {},
   "source": [
    "Let us plot the obtained results:"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "93a3ecb6-3c42-4d41-b990-1e607c74e18f",
   "metadata": {},
   "source": [
    "gseaの場合、単純なアクティビティ推定値（acts）、正規化推定値（norm_acts）、並べ替え後のpvalsデータフレームを返す。\n",
    "\n",
    "得られた結果をプロットしてみよう："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "dd2b1fb4",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 900x300 with 6 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axes = plt.subplots(\n",
    "    1, 3, figsize=(9, 3), tight_layout=True, sharey=True\n",
    ")\n",
    "\n",
    "axes[0].set_title('acts')\n",
    "sns.heatmap(acts, cmap='coolwarm', vmin=-1, vmax=1, ax=axes[0])\n",
    "\n",
    "axes[1].set_title('norm_acts')\n",
    "sns.heatmap(norm_acts, cmap='coolwarm', vmin=-1, vmax=1, ax=axes[1])\n",
    "\n",
    "axes[2].set_title('pvals')\n",
    "sns.heatmap(pvals, cmap='viridis_r', ax=axes[2], vmax=1)\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c476bc3a",
   "metadata": {},
   "source": [
    "We can observe that for transcription factors T1 and T3, the obtained activities correctly distinguish the two sample populations. T2, on the other hand, should be down for the second population of samples since it is a repressor. This mislabeling of activities happens because `gsea` cannot model weights when inferring biological activities.\n",
    "\n",
    "When weights are available in the prior knowledge, we definitely recommend using any of the methods that take them into account to get better estimates, one example is the Univariate Linear Model method `ulm`:"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3f04ceb7-c163-4c88-abbc-4bb0d9177ff7",
   "metadata": {},
   "source": [
    "転写因子T1とT3については、得られた活性が2つのサンプル集団を正しく区別していることが観察できる。一方、T2は抑制因子であるため、2つ目のサンプル集団では減少しているはずである。gseaは生物学的活性を推論する際に重みをモデル化することができないため、このような誤ったラベル付けが起こります。\n",
    "\n",
    "事前知識で重みが利用可能な場合、より良い推定値を得るために、重みを考慮した方法を使用することをお勧めします："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "20525e79",
   "metadata": {},
   "outputs": [],
   "source": [
    "# For this toy data, we need to set min_n to 0 (otherwise it is better to keep at 5):\n",
    "acts, pvals = dc.run_ulm(mat, net, min_n=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "6ccd2a9a-60f6-4e08-a9f4-3c1f53571b3b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>T1</th>\n",
       "      <th>T2</th>\n",
       "      <th>T3</th>\n",
       "      <th>T4</th>\n",
       "      <th>T5</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>S01</th>\n",
       "      <td>4.020366</td>\n",
       "      <td>1.667426</td>\n",
       "      <td>-1.378734</td>\n",
       "      <td>-0.243605</td>\n",
       "      <td>-1.188727</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>S02</th>\n",
       "      <td>4.060297</td>\n",
       "      <td>1.392356</td>\n",
       "      <td>-1.313069</td>\n",
       "      <td>-0.387175</td>\n",
       "      <td>-1.102454</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>S03</th>\n",
       "      <td>4.189780</td>\n",
       "      <td>1.086979</td>\n",
       "      <td>-1.346471</td>\n",
       "      <td>-0.294281</td>\n",
       "      <td>-1.139018</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>S04</th>\n",
       "      <td>4.149536</td>\n",
       "      <td>1.395950</td>\n",
       "      <td>-1.332407</td>\n",
       "      <td>-0.273508</td>\n",
       "      <td>-1.197696</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>S05</th>\n",
       "      <td>4.052075</td>\n",
       "      <td>1.610522</td>\n",
       "      <td>-1.534533</td>\n",
       "      <td>-0.311155</td>\n",
       "      <td>-0.990895</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           T1        T2        T3        T4        T5\n",
       "S01  4.020366  1.667426 -1.378734 -0.243605 -1.188727\n",
       "S02  4.060297  1.392356 -1.313069 -0.387175 -1.102454\n",
       "S03  4.189780  1.086979 -1.346471 -0.294281 -1.139018\n",
       "S04  4.149536  1.395950 -1.332407 -0.273508 -1.197696\n",
       "S05  4.052075  1.610522 -1.534533 -0.311155 -0.990895"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "acts.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "010be201-a652-49a0-bf8c-8b4e61f411fc",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(24, 5)"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "acts.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "838477e6-fe0d-43fe-9a52-73f2c03925a3",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>T1</th>\n",
       "      <th>T2</th>\n",
       "      <th>T3</th>\n",
       "      <th>T4</th>\n",
       "      <th>T5</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>S01</th>\n",
       "      <td>0.002437</td>\n",
       "      <td>0.126394</td>\n",
       "      <td>0.198034</td>\n",
       "      <td>0.812462</td>\n",
       "      <td>0.262009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>S02</th>\n",
       "      <td>0.002285</td>\n",
       "      <td>0.193999</td>\n",
       "      <td>0.218491</td>\n",
       "      <td>0.706739</td>\n",
       "      <td>0.296089</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>S03</th>\n",
       "      <td>0.001859</td>\n",
       "      <td>0.302552</td>\n",
       "      <td>0.207875</td>\n",
       "      <td>0.774561</td>\n",
       "      <td>0.281246</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>S04</th>\n",
       "      <td>0.001981</td>\n",
       "      <td>0.192946</td>\n",
       "      <td>0.212291</td>\n",
       "      <td>0.790029</td>\n",
       "      <td>0.258653</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>S05</th>\n",
       "      <td>0.002316</td>\n",
       "      <td>0.138361</td>\n",
       "      <td>0.155908</td>\n",
       "      <td>0.762072</td>\n",
       "      <td>0.345107</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           T1        T2        T3        T4        T5\n",
       "S01  0.002437  0.126394  0.198034  0.812462  0.262009\n",
       "S02  0.002285  0.193999  0.218491  0.706739  0.296089\n",
       "S03  0.001859  0.302552  0.207875  0.774561  0.281246\n",
       "S04  0.001981  0.192946  0.212291  0.790029  0.258653\n",
       "S05  0.002316  0.138361  0.155908  0.762072  0.345107"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pvals.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "90de60e9-0bdd-4336-9969-e65312b603bc",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(24, 5)"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pvals.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5ecc8fe7",
   "metadata": {},
   "source": [
    "In this case, `ulm` only returns infered activities and their associated p-value.\n",
    "\n",
    "As before, let us plot the resulting activities:"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "eef00af4-a297-471e-9319-dbda3d394f67",
   "metadata": {},
   "source": [
    "この場合、ulmは推論されたアクティビティと関連するp値のみを返す。\n",
    "\n",
    "前回と同様に、結果のアクティビティをプロットしてみよう："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "9c61d04b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 600x300 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axes = plt.subplots(\n",
    "    1, 2, figsize=(6, 3), tight_layout=True, sharey=True\n",
    ")\n",
    "\n",
    "axes[0].set_title('acts')\n",
    "sns.heatmap(acts, cmap='coolwarm', vmin=-1, vmax=1, ax=axes[0])\n",
    "\n",
    "axes[1].set_title('pvals')\n",
    "sns.heatmap(pvals, cmap='viridis_r', ax=axes[1], vmax=1)\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4c21c032",
   "metadata": {},
   "source": [
    "Since `ulm` models weights when estimating biological activities, it correctly assigns T2 as inactive in the second population of samples."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fa90204c-da8a-447c-8b0d-4162ed9aa69c",
   "metadata": {},
   "source": [
    "ulmは生物学的活性を推定する際に重みをモデル化するため、2番目のサンプルの母集団ではT2を不活性と正しく割り当てている。"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7e00a68b",
   "metadata": {},
   "source": [
    "### Multiple methods\n",
    "\n",
    "`decoupler` also allows to run multiple methods at the same time. Moreover, it computes a consensus score based on the obtained activities across methods, called `consensus`.\n",
    "\n",
    "By default, `decouple` runs only the top performer methods in our benchmark (`mlm`, `ulm` and `wsum`), and estimates a consensus score across them. Specific arguments to specific methods can be passed using the variable `args`. For more information check `?decouple`. If we wanted to only obtain the consensus score, we could also have used `run_consensus`, check `?run_consensus` for more information."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "335ba5ad-a392-49dd-8560-0a350472f2f4",
   "metadata": {},
   "source": [
    "decouplerは、複数のメソッドを同時に実行することもできる。さらに、コンセンサスと呼ばれる、手法間で得られたアクティビティに基づくコンセンサススコアを計算する。\n",
    "\n",
    "デフォルトでは、decoupleはベンチマークのトップパフォーマーのメソッド（mlm、ulm、wsum）のみを実行し、それら全体のコンセンサススコアを推定する。特定のメソッドへの引数は、変数 args を使って渡すことができる。詳細は ?decouple を参照。コンセンサススコアだけを取得したい場合は、run_consensusを使うこともできます。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "d399c1e9-eba5-4e09-9e2e-f2b2dfb51f9a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\u001b[0;31mSignature:\u001b[0m\n",
       "\u001b[0mdc\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdecouple\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m\u001b[0m\n",
       "\u001b[0;34m\u001b[0m    \u001b[0mmat\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\n",
       "\u001b[0;34m\u001b[0m    \u001b[0mnet\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\n",
       "\u001b[0;34m\u001b[0m    \u001b[0msource\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m'source'\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\n",
       "\u001b[0;34m\u001b[0m    \u001b[0mtarget\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m'target'\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\n",
       "\u001b[0;34m\u001b[0m    \u001b[0mweight\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m'weight'\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\n",
       "\u001b[0;34m\u001b[0m    \u001b[0mmethods\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mNone\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\n",
       "\u001b[0;34m\u001b[0m    \u001b[0margs\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m{\u001b[0m\u001b[0;34m}\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\n",
       "\u001b[0;34m\u001b[0m    \u001b[0mconsensus\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mTrue\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\n",
       "\u001b[0;34m\u001b[0m    \u001b[0mcns_metds\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mNone\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\n",
       "\u001b[0;34m\u001b[0m    \u001b[0mmin_n\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m5\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\n",
       "\u001b[0;34m\u001b[0m    \u001b[0mverbose\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mTrue\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\n",
       "\u001b[0;34m\u001b[0m    \u001b[0muse_raw\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mTrue\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\n",
       "\u001b[0;34m\u001b[0m    \u001b[0mdense\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mFalse\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\n",
       "\u001b[0;34m\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
       "\u001b[0;31mDocstring:\u001b[0m\n",
       "Decouple function.\n",
       "\n",
       "Runs simultaneously several methods of biological activity inference.\n",
       "\n",
       "Parameters\n",
       "----------\n",
       "mat : list, DataFrame or AnnData\n",
       "    List of [features, matrix], dataframe (samples x features) or an AnnData instance.\n",
       "net : DataFrame\n",
       "    Network in long format.\n",
       "source : str\n",
       "    Column name in net with source nodes.\n",
       "target : str\n",
       "    Column name in net with target nodes.\n",
       "weight : str\n",
       "    Column name in net with weights.\n",
       "methods : list, str, None\n",
       "    List of methods to run. If none are provided use weighted top performers (mlm, ulm and wsum). To run all methods set to\n",
       "    \"all\".\n",
       "args : dict\n",
       "    A dict of argument-dicts.\n",
       "consensus : bool\n",
       "    Boolean whether to run a consensus score between methods.\n",
       "cns_metds : list\n",
       "    List of estimate names to use for the calculation of the consensus score. If empty it will use all the estimates\n",
       "    obtained after running the different methods. If methods is also None, it will use mlm, ulm and norm_wsum instead.\n",
       "min_n : int\n",
       "    Minimum of targets per source. If less, sources are removed.\n",
       "verbose : bool\n",
       "    Whether to show progress.\n",
       "use_raw : bool\n",
       "    If mat is AnnData, use its raw attribute.\n",
       "dense : bool\n",
       "    Whether to run methods ignoring zero values per observation. See ``decoupler.dense_run`` for more information.\n",
       "\n",
       "Returns\n",
       "-------\n",
       "results : dict\n",
       "    Dictionary of activity estimates and p-values. If `mat` is AnnData, results for each method are stored in\n",
       "    `.obsm['method_estimate']` and if available in `.obsm['method_pvals']`.\n",
       "\u001b[0;31mFile:\u001b[0m      ~/mambaforge/envs/py_scvi/lib/python3.10/site-packages/decoupler/decouple.py\n",
       "\u001b[0;31mType:\u001b[0m      function"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "?dc.decouple"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "5b1398f4-c1e9-4ef0-b948-487c6dafc289",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\u001b[0;31mSignature:\u001b[0m\n",
       "\u001b[0mdc\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrun_consensus\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m\u001b[0m\n",
       "\u001b[0;34m\u001b[0m    \u001b[0mmat\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\n",
       "\u001b[0;34m\u001b[0m    \u001b[0mnet\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\n",
       "\u001b[0;34m\u001b[0m    \u001b[0msource\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m'source'\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\n",
       "\u001b[0;34m\u001b[0m    \u001b[0mtarget\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m'target'\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\n",
       "\u001b[0;34m\u001b[0m    \u001b[0mweight\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m'weight'\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\n",
       "\u001b[0;34m\u001b[0m    \u001b[0mmin_n\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m5\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\n",
       "\u001b[0;34m\u001b[0m    \u001b[0mverbose\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mFalse\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\n",
       "\u001b[0;34m\u001b[0m    \u001b[0muse_raw\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mTrue\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\n",
       "\u001b[0;34m\u001b[0m    \u001b[0margs\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m{\u001b[0m\u001b[0;34m}\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\n",
       "\u001b[0;34m\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
       "\u001b[0;31mDocstring:\u001b[0m\n",
       "Consensus score from top methods.\n",
       "\n",
       "This consensus score is calculated from the three top performer methods: `ulm`, `mlm` and `wsum_norm`.\n",
       "For each of these methods, the obtained activities are transformed into z-scores, first for positive\n",
       "values and then for negative ones. These two sets of z-score transformed activities are computed by\n",
       "subsetting the values bigger or lower than 0, then by mirroring the selected values into their\n",
       "opposite sign and finally calculating a classic z-score. This transformation ensures that values across\n",
       "methods are comparable, and that they remain in their original sign (active or inactive). The final\n",
       "consensus score is the mean across different methods. A p-value is then estimated from these using a\n",
       "cumulative distribution function.\n",
       "\n",
       "Parameters\n",
       "----------\n",
       "mat : list, DataFrame or AnnData\n",
       "    List of [features, matrix], dataframe (samples x features) or an AnnData instance.\n",
       "net : DataFrame\n",
       "    Network in long format.\n",
       "source : str\n",
       "    Column name in net with source nodes.\n",
       "target : str\n",
       "    Column name in net with target nodes.\n",
       "weight : str\n",
       "    Column name in net with weights.\n",
       "min_n : int\n",
       "    Minimum of targets per source. If less, sources are removed.\n",
       "verbose : bool\n",
       "    Whether to show progress.\n",
       "use_raw : bool\n",
       "    Use raw attribute of mat if present.\n",
       "args : dict\n",
       "    Parameters for the ``decoupler.decouple`` function.\n",
       "\n",
       "Returns\n",
       "-------\n",
       "estimate : DataFrame\n",
       "    Consensus scores. Stored in `.obsm['consensus_estimate']` if `mat` is AnnData.\n",
       "pvals : DataFrame\n",
       "    Obtained p-values. Stored in `.obsm['consensus_pvals']` if `mat` is AnnData.\n",
       "\u001b[0;31mFile:\u001b[0m      ~/mambaforge/envs/py_scvi/lib/python3.10/site-packages/decoupler/decouple.py\n",
       "\u001b[0;31mType:\u001b[0m      function"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "?dc.run_consensus"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "e3825926",
   "metadata": {},
   "outputs": [],
   "source": [
    "# For this toy data, we need to set min_n to 0 (otherwise it is better to keep at 5):\n",
    "results = dc.decouple(mat, net, min_n=0, verbose=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "93566d48-d844-4b31-8659-81c491811b57",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "dict_keys(['mlm_estimate', 'mlm_pvals', 'ulm_estimate', 'ulm_pvals', 'wsum_estimate', 'wsum_norm', 'wsum_corr', 'wsum_pvals', 'consensus_estimate', 'consensus_pvals'])"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results.keys()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0fb3db1b",
   "metadata": {},
   "source": [
    "`decouple` either returns a dictionary of activities and p-values, or stores them in the `AnnData` instance provided. \n",
    "\n",
    "Let us see how the consensus score looks like:"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "70fe4af7-8e72-45f4-b160-b38c91ade345",
   "metadata": {},
   "source": [
    "decoupleは、アクティビティとp値の辞書を返すか、提供されたAnnDataインスタンスに格納します。\n",
    "\n",
    "コンセンサス・スコアがどのように見えるか見てみよう："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "ddedb9dd",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Extract from dictionary\n",
    "acts = results['consensus_estimate']\n",
    "pvals = results['consensus_pvals']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "ca700b0f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 600x300 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axes = plt.subplots(\n",
    "    1, 2, figsize=(6, 3), tight_layout=True, sharey=True\n",
    ")\n",
    "\n",
    "axes[0].set_title('acts')\n",
    "sns.heatmap(acts, cmap='coolwarm', vmin=-1, vmax=1, ax=axes[0])\n",
    "\n",
    "axes[1].set_title('pvals')\n",
    "sns.heatmap(pvals, cmap='viridis_r', ax=axes[1], vmax=1)\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "65e14da3",
   "metadata": {},
   "source": [
    "We can observe that the consensus score correctly predicts that T1 and T2 should be active in the first population of samples while T3 and T4 in the second one. T5 is inactive everywhere as expected."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "624a9750-25b3-4831-bb5c-b5b065203ff4",
   "metadata": {},
   "source": [
    "コンセンサス・スコアは、T1とT2が最初のサンプル集団で活性化し、T3とT4が2番目のサンプル集団で活性化することを正しく予測していることがわかる。T5は予想通りどこでも不活性である。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "85b523d4-12ce-4144-94bf-13ad809f6879",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "py_scvi",
   "language": "python",
   "name": "py_scvi"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.14"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}