{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 数据处理：SMILES信息在Gene里，CP里面没有，所以首先在这里实现映射和对齐"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "import os\n",
    "import h5py\n",
    "\n",
    "def load_from_HDF(fname):\n",
    "    \"\"\"Load data from a HDF5 file to a dictionary.\"\"\"\n",
    "    data = dict()\n",
    "    with h5py.File(fname, 'r') as f:\n",
    "        for key in f:\n",
    "            data[key] = np.asarray(f[key])\n",
    "            if isinstance(data[key][0], np.bytes_):\n",
    "                data[key] = data[key].astype(str)\n",
    "    return data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(113416, 898) (61313, 980)\n"
     ]
    },
    {
     "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>pert_dose</th>\n",
       "      <th>det_plate</th>\n",
       "      <th>CPD_SMILES</th>\n",
       "      <th>221227_x_at</th>\n",
       "      <th>212345_s_at</th>\n",
       "      <th>218597_s_at</th>\n",
       "      <th>217140_s_at</th>\n",
       "      <th>209253_at</th>\n",
       "      <th>214404_x_at</th>\n",
       "      <th>219888_at</th>\n",
       "      <th>...</th>\n",
       "      <th>218397_at</th>\n",
       "      <th>202996_at</th>\n",
       "      <th>204608_at</th>\n",
       "      <th>211071_s_at</th>\n",
       "      <th>203341_at</th>\n",
       "      <th>202801_at</th>\n",
       "      <th>206414_s_at</th>\n",
       "      <th>204978_at</th>\n",
       "      <th>205379_at</th>\n",
       "      <th>203897_at</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>50.0</td>\n",
       "      <td>PAC001_U2OS_6H_X1_B1_UNI4445L</td>\n",
       "      <td>CC(C)[C@H](CO)Nc1nc(Nc2cc(N)cc(Cl)c2)c2ncn(C(C...</td>\n",
       "      <td>-0.118505</td>\n",
       "      <td>-0.293445</td>\n",
       "      <td>-0.294088</td>\n",
       "      <td>0.292745</td>\n",
       "      <td>-0.145896</td>\n",
       "      <td>0.97358</td>\n",
       "      <td>0.247915</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.529905</td>\n",
       "      <td>0.804992</td>\n",
       "      <td>0.430269</td>\n",
       "      <td>-0.97559</td>\n",
       "      <td>-1.740605</td>\n",
       "      <td>0.340382</td>\n",
       "      <td>0.498778</td>\n",
       "      <td>0.352152</td>\n",
       "      <td>0.53591</td>\n",
       "      <td>0.52484</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>12.7</td>\n",
       "      <td>PAC001_U2OS_6H_X1_B1_UNI4445L</td>\n",
       "      <td>Ic1cccc(CSc2nnc(o2)-c2ccncc2)c1</td>\n",
       "      <td>0.043264</td>\n",
       "      <td>-0.264945</td>\n",
       "      <td>-0.000633</td>\n",
       "      <td>0.000935</td>\n",
       "      <td>0.177604</td>\n",
       "      <td>-0.50472</td>\n",
       "      <td>0.214315</td>\n",
       "      <td>...</td>\n",
       "      <td>0.228895</td>\n",
       "      <td>-0.381308</td>\n",
       "      <td>0.556969</td>\n",
       "      <td>0.25051</td>\n",
       "      <td>-0.025625</td>\n",
       "      <td>-0.024428</td>\n",
       "      <td>-0.225122</td>\n",
       "      <td>0.533052</td>\n",
       "      <td>-0.19329</td>\n",
       "      <td>-0.25436</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>50.0</td>\n",
       "      <td>PAC001_U2OS_6H_X1_B1_UNI4445L</td>\n",
       "      <td>Clc1cc(Cl)c(NC(=O)Nc2ccncc2)c(Cl)c1</td>\n",
       "      <td>-0.070805</td>\n",
       "      <td>0.195755</td>\n",
       "      <td>0.004606</td>\n",
       "      <td>-0.040855</td>\n",
       "      <td>-0.067346</td>\n",
       "      <td>-0.19872</td>\n",
       "      <td>-0.053195</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.031795</td>\n",
       "      <td>-0.051728</td>\n",
       "      <td>-0.017741</td>\n",
       "      <td>0.05681</td>\n",
       "      <td>0.130695</td>\n",
       "      <td>0.005919</td>\n",
       "      <td>-0.408722</td>\n",
       "      <td>-0.057177</td>\n",
       "      <td>-0.25699</td>\n",
       "      <td>0.31004</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>16.2</td>\n",
       "      <td>PAC001_U2OS_6H_X1_B1_UNI4445L</td>\n",
       "      <td>COc1ccc(CNC(=O)Nc2ncc(s2)[N+]([O-])=O)cc1</td>\n",
       "      <td>0.027165</td>\n",
       "      <td>-0.149545</td>\n",
       "      <td>0.173113</td>\n",
       "      <td>-0.100695</td>\n",
       "      <td>-0.309296</td>\n",
       "      <td>-0.20782</td>\n",
       "      <td>-0.323985</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.013787</td>\n",
       "      <td>-0.177608</td>\n",
       "      <td>0.219869</td>\n",
       "      <td>-0.16129</td>\n",
       "      <td>0.365195</td>\n",
       "      <td>-0.040848</td>\n",
       "      <td>-0.251122</td>\n",
       "      <td>0.040732</td>\n",
       "      <td>0.17091</td>\n",
       "      <td>-0.05144</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>12.5</td>\n",
       "      <td>PAC001_U2OS_6H_X1_B1_UNI4445L</td>\n",
       "      <td>OC(=O)c1ccc2c3nc4ccccc4n3c(=O)c3cccc1c23</td>\n",
       "      <td>0.303294</td>\n",
       "      <td>0.254455</td>\n",
       "      <td>-0.055418</td>\n",
       "      <td>-0.053635</td>\n",
       "      <td>0.000455</td>\n",
       "      <td>-0.10908</td>\n",
       "      <td>0.031425</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.052585</td>\n",
       "      <td>-0.196108</td>\n",
       "      <td>0.245469</td>\n",
       "      <td>0.00784</td>\n",
       "      <td>0.093595</td>\n",
       "      <td>-0.303418</td>\n",
       "      <td>0.474878</td>\n",
       "      <td>-0.022947</td>\n",
       "      <td>-0.28589</td>\n",
       "      <td>0.06423</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 980 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   pert_dose                      det_plate  \\\n",
       "0       50.0  PAC001_U2OS_6H_X1_B1_UNI4445L   \n",
       "1       12.7  PAC001_U2OS_6H_X1_B1_UNI4445L   \n",
       "2       50.0  PAC001_U2OS_6H_X1_B1_UNI4445L   \n",
       "3       16.2  PAC001_U2OS_6H_X1_B1_UNI4445L   \n",
       "4       12.5  PAC001_U2OS_6H_X1_B1_UNI4445L   \n",
       "\n",
       "                                          CPD_SMILES  221227_x_at  \\\n",
       "0  CC(C)[C@H](CO)Nc1nc(Nc2cc(N)cc(Cl)c2)c2ncn(C(C...    -0.118505   \n",
       "1                    Ic1cccc(CSc2nnc(o2)-c2ccncc2)c1     0.043264   \n",
       "2                Clc1cc(Cl)c(NC(=O)Nc2ccncc2)c(Cl)c1    -0.070805   \n",
       "3          COc1ccc(CNC(=O)Nc2ncc(s2)[N+]([O-])=O)cc1     0.027165   \n",
       "4           OC(=O)c1ccc2c3nc4ccccc4n3c(=O)c3cccc1c23     0.303294   \n",
       "\n",
       "   212345_s_at  218597_s_at  217140_s_at  209253_at  214404_x_at  219888_at  \\\n",
       "0    -0.293445    -0.294088     0.292745  -0.145896      0.97358   0.247915   \n",
       "1    -0.264945    -0.000633     0.000935   0.177604     -0.50472   0.214315   \n",
       "2     0.195755     0.004606    -0.040855  -0.067346     -0.19872  -0.053195   \n",
       "3    -0.149545     0.173113    -0.100695  -0.309296     -0.20782  -0.323985   \n",
       "4     0.254455    -0.055418    -0.053635   0.000455     -0.10908   0.031425   \n",
       "\n",
       "   ...  218397_at  202996_at  204608_at  211071_s_at  203341_at  202801_at  \\\n",
       "0  ...  -0.529905   0.804992   0.430269     -0.97559  -1.740605   0.340382   \n",
       "1  ...   0.228895  -0.381308   0.556969      0.25051  -0.025625  -0.024428   \n",
       "2  ...  -0.031795  -0.051728  -0.017741      0.05681   0.130695   0.005919   \n",
       "3  ...  -0.013787  -0.177608   0.219869     -0.16129   0.365195  -0.040848   \n",
       "4  ...  -0.052585  -0.196108   0.245469      0.00784   0.093595  -0.303418   \n",
       "\n",
       "   206414_s_at  204978_at  205379_at  203897_at  \n",
       "0     0.498778   0.352152    0.53591    0.52484  \n",
       "1    -0.225122   0.533052   -0.19329   -0.25436  \n",
       "2    -0.408722  -0.057177   -0.25699    0.31004  \n",
       "3    -0.251122   0.040732    0.17091   -0.05144  \n",
       "4     0.474878  -0.022947   -0.28589    0.06423  \n",
       "\n",
       "[5 rows x 980 columns]"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "CP_path = './CP_data.parquet'\n",
    "GE_path = './GE_data.parquet'\n",
    "\n",
    "CP_CSV = pd.read_parquet(CP_path)\n",
    "GE_CSV = pd.read_parquet(GE_path)\n",
    "print(CP_CSV.shape, GE_CSV.shape)\n",
    "GE_CSV.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((20342,), (20342,))"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "CP_CSV['SMILES'].unique().shape, GE_CSV['CPD_SMILES'].unique().shape  # 包含 NaN的"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 数据对齐\n",
    "\n",
    "### 对齐逻辑：数据清洗，各自生成对齐数据即可"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "# GE_CSV = GE_CSV.rename(columns={\n",
    "#     'CPD_SMILES': 'SMILES',\n",
    "#     'pert_dose': 'dose',\n",
    "#     'det_plate': 'Plate'\n",
    "# })\n",
    "# GE_CSV = GE_CSV.drop(columns=['Metadata_moa'])\n",
    "# GE_CSV.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "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>dose</th>\n",
       "      <th>Plate</th>\n",
       "      <th>SMILES</th>\n",
       "      <th>221227_x_at</th>\n",
       "      <th>212345_s_at</th>\n",
       "      <th>218597_s_at</th>\n",
       "      <th>217140_s_at</th>\n",
       "      <th>209253_at</th>\n",
       "      <th>214404_x_at</th>\n",
       "      <th>219888_at</th>\n",
       "      <th>...</th>\n",
       "      <th>218397_at</th>\n",
       "      <th>202996_at</th>\n",
       "      <th>204608_at</th>\n",
       "      <th>211071_s_at</th>\n",
       "      <th>203341_at</th>\n",
       "      <th>202801_at</th>\n",
       "      <th>206414_s_at</th>\n",
       "      <th>204978_at</th>\n",
       "      <th>205379_at</th>\n",
       "      <th>203897_at</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>50.0</td>\n",
       "      <td>PAC001_U2OS_6H_X1_B1_UNI4445L</td>\n",
       "      <td>CC(C)[C@H](CO)Nc1nc(Nc2cc(N)cc(Cl)c2)c2ncn(C(C...</td>\n",
       "      <td>-0.118505</td>\n",
       "      <td>-0.293445</td>\n",
       "      <td>-0.294088</td>\n",
       "      <td>0.292745</td>\n",
       "      <td>-0.145896</td>\n",
       "      <td>0.97358</td>\n",
       "      <td>0.247915</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.529905</td>\n",
       "      <td>0.804992</td>\n",
       "      <td>0.430269</td>\n",
       "      <td>-0.97559</td>\n",
       "      <td>-1.740605</td>\n",
       "      <td>0.340382</td>\n",
       "      <td>0.498778</td>\n",
       "      <td>0.352152</td>\n",
       "      <td>0.53591</td>\n",
       "      <td>0.52484</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>12.7</td>\n",
       "      <td>PAC001_U2OS_6H_X1_B1_UNI4445L</td>\n",
       "      <td>Ic1cccc(CSc2nnc(o2)-c2ccncc2)c1</td>\n",
       "      <td>0.043264</td>\n",
       "      <td>-0.264945</td>\n",
       "      <td>-0.000633</td>\n",
       "      <td>0.000935</td>\n",
       "      <td>0.177604</td>\n",
       "      <td>-0.50472</td>\n",
       "      <td>0.214315</td>\n",
       "      <td>...</td>\n",
       "      <td>0.228895</td>\n",
       "      <td>-0.381308</td>\n",
       "      <td>0.556969</td>\n",
       "      <td>0.25051</td>\n",
       "      <td>-0.025625</td>\n",
       "      <td>-0.024428</td>\n",
       "      <td>-0.225122</td>\n",
       "      <td>0.533052</td>\n",
       "      <td>-0.19329</td>\n",
       "      <td>-0.25436</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>50.0</td>\n",
       "      <td>PAC001_U2OS_6H_X1_B1_UNI4445L</td>\n",
       "      <td>Clc1cc(Cl)c(NC(=O)Nc2ccncc2)c(Cl)c1</td>\n",
       "      <td>-0.070805</td>\n",
       "      <td>0.195755</td>\n",
       "      <td>0.004606</td>\n",
       "      <td>-0.040855</td>\n",
       "      <td>-0.067346</td>\n",
       "      <td>-0.19872</td>\n",
       "      <td>-0.053195</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.031795</td>\n",
       "      <td>-0.051728</td>\n",
       "      <td>-0.017741</td>\n",
       "      <td>0.05681</td>\n",
       "      <td>0.130695</td>\n",
       "      <td>0.005919</td>\n",
       "      <td>-0.408722</td>\n",
       "      <td>-0.057177</td>\n",
       "      <td>-0.25699</td>\n",
       "      <td>0.31004</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>16.2</td>\n",
       "      <td>PAC001_U2OS_6H_X1_B1_UNI4445L</td>\n",
       "      <td>COc1ccc(CNC(=O)Nc2ncc(s2)[N+]([O-])=O)cc1</td>\n",
       "      <td>0.027165</td>\n",
       "      <td>-0.149545</td>\n",
       "      <td>0.173113</td>\n",
       "      <td>-0.100695</td>\n",
       "      <td>-0.309296</td>\n",
       "      <td>-0.20782</td>\n",
       "      <td>-0.323985</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.013787</td>\n",
       "      <td>-0.177608</td>\n",
       "      <td>0.219869</td>\n",
       "      <td>-0.16129</td>\n",
       "      <td>0.365195</td>\n",
       "      <td>-0.040848</td>\n",
       "      <td>-0.251122</td>\n",
       "      <td>0.040732</td>\n",
       "      <td>0.17091</td>\n",
       "      <td>-0.05144</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>12.5</td>\n",
       "      <td>PAC001_U2OS_6H_X1_B1_UNI4445L</td>\n",
       "      <td>OC(=O)c1ccc2c3nc4ccccc4n3c(=O)c3cccc1c23</td>\n",
       "      <td>0.303294</td>\n",
       "      <td>0.254455</td>\n",
       "      <td>-0.055418</td>\n",
       "      <td>-0.053635</td>\n",
       "      <td>0.000455</td>\n",
       "      <td>-0.10908</td>\n",
       "      <td>0.031425</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.052585</td>\n",
       "      <td>-0.196108</td>\n",
       "      <td>0.245469</td>\n",
       "      <td>0.00784</td>\n",
       "      <td>0.093595</td>\n",
       "      <td>-0.303418</td>\n",
       "      <td>0.474878</td>\n",
       "      <td>-0.022947</td>\n",
       "      <td>-0.28589</td>\n",
       "      <td>0.06423</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 980 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   dose                          Plate  \\\n",
       "0  50.0  PAC001_U2OS_6H_X1_B1_UNI4445L   \n",
       "1  12.7  PAC001_U2OS_6H_X1_B1_UNI4445L   \n",
       "2  50.0  PAC001_U2OS_6H_X1_B1_UNI4445L   \n",
       "3  16.2  PAC001_U2OS_6H_X1_B1_UNI4445L   \n",
       "4  12.5  PAC001_U2OS_6H_X1_B1_UNI4445L   \n",
       "\n",
       "                                              SMILES  221227_x_at  \\\n",
       "0  CC(C)[C@H](CO)Nc1nc(Nc2cc(N)cc(Cl)c2)c2ncn(C(C...    -0.118505   \n",
       "1                    Ic1cccc(CSc2nnc(o2)-c2ccncc2)c1     0.043264   \n",
       "2                Clc1cc(Cl)c(NC(=O)Nc2ccncc2)c(Cl)c1    -0.070805   \n",
       "3          COc1ccc(CNC(=O)Nc2ncc(s2)[N+]([O-])=O)cc1     0.027165   \n",
       "4           OC(=O)c1ccc2c3nc4ccccc4n3c(=O)c3cccc1c23     0.303294   \n",
       "\n",
       "   212345_s_at  218597_s_at  217140_s_at  209253_at  214404_x_at  219888_at  \\\n",
       "0    -0.293445    -0.294088     0.292745  -0.145896      0.97358   0.247915   \n",
       "1    -0.264945    -0.000633     0.000935   0.177604     -0.50472   0.214315   \n",
       "2     0.195755     0.004606    -0.040855  -0.067346     -0.19872  -0.053195   \n",
       "3    -0.149545     0.173113    -0.100695  -0.309296     -0.20782  -0.323985   \n",
       "4     0.254455    -0.055418    -0.053635   0.000455     -0.10908   0.031425   \n",
       "\n",
       "   ...  218397_at  202996_at  204608_at  211071_s_at  203341_at  202801_at  \\\n",
       "0  ...  -0.529905   0.804992   0.430269     -0.97559  -1.740605   0.340382   \n",
       "1  ...   0.228895  -0.381308   0.556969      0.25051  -0.025625  -0.024428   \n",
       "2  ...  -0.031795  -0.051728  -0.017741      0.05681   0.130695   0.005919   \n",
       "3  ...  -0.013787  -0.177608   0.219869     -0.16129   0.365195  -0.040848   \n",
       "4  ...  -0.052585  -0.196108   0.245469      0.00784   0.093595  -0.303418   \n",
       "\n",
       "   206414_s_at  204978_at  205379_at  203897_at  \n",
       "0     0.498778   0.352152    0.53591    0.52484  \n",
       "1    -0.225122   0.533052   -0.19329   -0.25436  \n",
       "2    -0.408722  -0.057177   -0.25699    0.31004  \n",
       "3    -0.251122   0.040732    0.17091   -0.05144  \n",
       "4     0.474878  -0.022947   -0.28589    0.06423  \n",
       "\n",
       "[5 rows x 980 columns]"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "GE_CSV.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "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>Metadata_mmoles_per_liter2</th>\n",
       "      <th>SMILES</th>\n",
       "      <th>Metadata_Plate</th>\n",
       "      <th>Cells_AreaShape_Compactness</th>\n",
       "      <th>Cells_AreaShape_EulerNumber</th>\n",
       "      <th>Cells_AreaShape_Extent</th>\n",
       "      <th>Cells_AreaShape_FormFactor</th>\n",
       "      <th>Cells_AreaShape_MaxFeretDiameter</th>\n",
       "      <th>Cells_AreaShape_MaximumRadius</th>\n",
       "      <th>Cells_AreaShape_Solidity</th>\n",
       "      <th>...</th>\n",
       "      <th>Nuclei_Texture_SumEntropy_DNA_5_0</th>\n",
       "      <th>Nuclei_Texture_SumEntropy_ER_3_0</th>\n",
       "      <th>Nuclei_Texture_SumEntropy_Mito_3_0</th>\n",
       "      <th>Nuclei_Texture_SumEntropy_RNA_5_0</th>\n",
       "      <th>Nuclei_Texture_SumVariance_ER_3_0</th>\n",
       "      <th>Nuclei_Texture_SumVariance_Mito_5_0</th>\n",
       "      <th>Nuclei_Texture_SumVariance_RNA_10_0</th>\n",
       "      <th>Nuclei_Texture_Variance_AGP_5_0</th>\n",
       "      <th>Nuclei_Texture_Variance_ER_5_0</th>\n",
       "      <th>Nuclei_Texture_Variance_RNA_5_0</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>6.05</td>\n",
       "      <td>CCCOc1cc(N)ccc1C(=O)OCCN(CC)CC</td>\n",
       "      <td>24277</td>\n",
       "      <td>0.256710</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.114990</td>\n",
       "      <td>-0.237090</td>\n",
       "      <td>0.615594</td>\n",
       "      <td>0.550434</td>\n",
       "      <td>0.072795</td>\n",
       "      <td>...</td>\n",
       "      <td>0.090021</td>\n",
       "      <td>0.375747</td>\n",
       "      <td>-0.010427</td>\n",
       "      <td>0.118997</td>\n",
       "      <td>-0.156442</td>\n",
       "      <td>-0.018014</td>\n",
       "      <td>-0.024573</td>\n",
       "      <td>-0.383890</td>\n",
       "      <td>-0.199967</td>\n",
       "      <td>-0.075658</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>10.00</td>\n",
       "      <td>COc1cc(C)cc(OC)c1[C@@H]1C=C(C)CC[C@H]1C(C)=C</td>\n",
       "      <td>24277</td>\n",
       "      <td>0.050962</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.051053</td>\n",
       "      <td>-0.033733</td>\n",
       "      <td>0.271953</td>\n",
       "      <td>0.172714</td>\n",
       "      <td>0.145941</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.030338</td>\n",
       "      <td>0.390687</td>\n",
       "      <td>-0.058560</td>\n",
       "      <td>0.065822</td>\n",
       "      <td>-0.113721</td>\n",
       "      <td>-0.089807</td>\n",
       "      <td>-0.041392</td>\n",
       "      <td>-0.248967</td>\n",
       "      <td>-0.126183</td>\n",
       "      <td>0.018563</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>10.00</td>\n",
       "      <td>COc1cc(O)cc(\\C=C\\c2ccccc2)c1</td>\n",
       "      <td>24277</td>\n",
       "      <td>-0.098185</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.044212</td>\n",
       "      <td>0.164668</td>\n",
       "      <td>0.011985</td>\n",
       "      <td>0.127098</td>\n",
       "      <td>0.094736</td>\n",
       "      <td>...</td>\n",
       "      <td>0.102946</td>\n",
       "      <td>0.279148</td>\n",
       "      <td>0.029891</td>\n",
       "      <td>0.079777</td>\n",
       "      <td>-0.007979</td>\n",
       "      <td>0.081155</td>\n",
       "      <td>-0.089633</td>\n",
       "      <td>0.043140</td>\n",
       "      <td>-0.034355</td>\n",
       "      <td>0.052914</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>10.00</td>\n",
       "      <td>COc1c(O)cc2C(=O)O[C@H]3[C@@H](O)[C@H](O)[C@@H]...</td>\n",
       "      <td>24277</td>\n",
       "      <td>0.410722</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.236208</td>\n",
       "      <td>-0.055084</td>\n",
       "      <td>0.694648</td>\n",
       "      <td>0.516785</td>\n",
       "      <td>0.154239</td>\n",
       "      <td>...</td>\n",
       "      <td>0.470516</td>\n",
       "      <td>0.214792</td>\n",
       "      <td>0.015408</td>\n",
       "      <td>0.339100</td>\n",
       "      <td>-0.034251</td>\n",
       "      <td>-0.047597</td>\n",
       "      <td>0.103611</td>\n",
       "      <td>-0.139579</td>\n",
       "      <td>-0.042241</td>\n",
       "      <td>0.249977</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4.39</td>\n",
       "      <td>CCOc1ccc2ccccc2c1C(=O)N[C@H]1[C@H]2SC(C)(C)[C@...</td>\n",
       "      <td>24277</td>\n",
       "      <td>0.248559</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.020058</td>\n",
       "      <td>0.085772</td>\n",
       "      <td>0.678742</td>\n",
       "      <td>0.671897</td>\n",
       "      <td>0.214587</td>\n",
       "      <td>...</td>\n",
       "      <td>0.671790</td>\n",
       "      <td>0.424186</td>\n",
       "      <td>0.118873</td>\n",
       "      <td>0.415623</td>\n",
       "      <td>0.257598</td>\n",
       "      <td>0.193370</td>\n",
       "      <td>0.238471</td>\n",
       "      <td>-0.033797</td>\n",
       "      <td>0.080018</td>\n",
       "      <td>0.335890</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 897 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   Metadata_mmoles_per_liter2  \\\n",
       "0                        6.05   \n",
       "1                       10.00   \n",
       "2                       10.00   \n",
       "3                       10.00   \n",
       "4                        4.39   \n",
       "\n",
       "                                              SMILES  Metadata_Plate  \\\n",
       "0                     CCCOc1cc(N)ccc1C(=O)OCCN(CC)CC           24277   \n",
       "1       COc1cc(C)cc(OC)c1[C@@H]1C=C(C)CC[C@H]1C(C)=C           24277   \n",
       "2                       COc1cc(O)cc(\\C=C\\c2ccccc2)c1           24277   \n",
       "3  COc1c(O)cc2C(=O)O[C@H]3[C@@H](O)[C@H](O)[C@@H]...           24277   \n",
       "4  CCOc1ccc2ccccc2c1C(=O)N[C@H]1[C@H]2SC(C)(C)[C@...           24277   \n",
       "\n",
       "   Cells_AreaShape_Compactness  Cells_AreaShape_EulerNumber  \\\n",
       "0                     0.256710                          NaN   \n",
       "1                     0.050962                          NaN   \n",
       "2                    -0.098185                          NaN   \n",
       "3                     0.410722                          NaN   \n",
       "4                     0.248559                          NaN   \n",
       "\n",
       "   Cells_AreaShape_Extent  Cells_AreaShape_FormFactor  \\\n",
       "0               -0.114990                   -0.237090   \n",
       "1                0.051053                   -0.033733   \n",
       "2                0.044212                    0.164668   \n",
       "3               -0.236208                   -0.055084   \n",
       "4               -0.020058                    0.085772   \n",
       "\n",
       "   Cells_AreaShape_MaxFeretDiameter  Cells_AreaShape_MaximumRadius  \\\n",
       "0                          0.615594                       0.550434   \n",
       "1                          0.271953                       0.172714   \n",
       "2                          0.011985                       0.127098   \n",
       "3                          0.694648                       0.516785   \n",
       "4                          0.678742                       0.671897   \n",
       "\n",
       "   Cells_AreaShape_Solidity  ...  Nuclei_Texture_SumEntropy_DNA_5_0  \\\n",
       "0                  0.072795  ...                           0.090021   \n",
       "1                  0.145941  ...                          -0.030338   \n",
       "2                  0.094736  ...                           0.102946   \n",
       "3                  0.154239  ...                           0.470516   \n",
       "4                  0.214587  ...                           0.671790   \n",
       "\n",
       "   Nuclei_Texture_SumEntropy_ER_3_0  Nuclei_Texture_SumEntropy_Mito_3_0  \\\n",
       "0                          0.375747                           -0.010427   \n",
       "1                          0.390687                           -0.058560   \n",
       "2                          0.279148                            0.029891   \n",
       "3                          0.214792                            0.015408   \n",
       "4                          0.424186                            0.118873   \n",
       "\n",
       "   Nuclei_Texture_SumEntropy_RNA_5_0  Nuclei_Texture_SumVariance_ER_3_0  \\\n",
       "0                           0.118997                          -0.156442   \n",
       "1                           0.065822                          -0.113721   \n",
       "2                           0.079777                          -0.007979   \n",
       "3                           0.339100                          -0.034251   \n",
       "4                           0.415623                           0.257598   \n",
       "\n",
       "   Nuclei_Texture_SumVariance_Mito_5_0  Nuclei_Texture_SumVariance_RNA_10_0  \\\n",
       "0                            -0.018014                            -0.024573   \n",
       "1                            -0.089807                            -0.041392   \n",
       "2                             0.081155                            -0.089633   \n",
       "3                            -0.047597                             0.103611   \n",
       "4                             0.193370                             0.238471   \n",
       "\n",
       "   Nuclei_Texture_Variance_AGP_5_0  Nuclei_Texture_Variance_ER_5_0  \\\n",
       "0                        -0.383890                       -0.199967   \n",
       "1                        -0.248967                       -0.126183   \n",
       "2                         0.043140                       -0.034355   \n",
       "3                        -0.139579                       -0.042241   \n",
       "4                        -0.033797                        0.080018   \n",
       "\n",
       "   Nuclei_Texture_Variance_RNA_5_0  \n",
       "0                        -0.075658  \n",
       "1                         0.018563  \n",
       "2                         0.052914  \n",
       "3                         0.249977  \n",
       "4                         0.335890  \n",
       "\n",
       "[5 rows x 897 columns]"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "CP_CSV.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "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>dose</th>\n",
       "      <th>SMILES</th>\n",
       "      <th>Plate</th>\n",
       "      <th>Cells_AreaShape_Compactness</th>\n",
       "      <th>Cells_AreaShape_EulerNumber</th>\n",
       "      <th>Cells_AreaShape_Extent</th>\n",
       "      <th>Cells_AreaShape_FormFactor</th>\n",
       "      <th>Cells_AreaShape_MaxFeretDiameter</th>\n",
       "      <th>Cells_AreaShape_MaximumRadius</th>\n",
       "      <th>Cells_AreaShape_Solidity</th>\n",
       "      <th>...</th>\n",
       "      <th>Nuclei_Texture_SumEntropy_DNA_5_0</th>\n",
       "      <th>Nuclei_Texture_SumEntropy_ER_3_0</th>\n",
       "      <th>Nuclei_Texture_SumEntropy_Mito_3_0</th>\n",
       "      <th>Nuclei_Texture_SumEntropy_RNA_5_0</th>\n",
       "      <th>Nuclei_Texture_SumVariance_ER_3_0</th>\n",
       "      <th>Nuclei_Texture_SumVariance_Mito_5_0</th>\n",
       "      <th>Nuclei_Texture_SumVariance_RNA_10_0</th>\n",
       "      <th>Nuclei_Texture_Variance_AGP_5_0</th>\n",
       "      <th>Nuclei_Texture_Variance_ER_5_0</th>\n",
       "      <th>Nuclei_Texture_Variance_RNA_5_0</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>6.05</td>\n",
       "      <td>CCCOc1cc(N)ccc1C(=O)OCCN(CC)CC</td>\n",
       "      <td>24277</td>\n",
       "      <td>0.256710</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.114990</td>\n",
       "      <td>-0.237090</td>\n",
       "      <td>0.615594</td>\n",
       "      <td>0.550434</td>\n",
       "      <td>0.072795</td>\n",
       "      <td>...</td>\n",
       "      <td>0.090021</td>\n",
       "      <td>0.375747</td>\n",
       "      <td>-0.010427</td>\n",
       "      <td>0.118997</td>\n",
       "      <td>-0.156442</td>\n",
       "      <td>-0.018014</td>\n",
       "      <td>-0.024573</td>\n",
       "      <td>-0.383890</td>\n",
       "      <td>-0.199967</td>\n",
       "      <td>-0.075658</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>10.00</td>\n",
       "      <td>COc1cc(C)cc(OC)c1[C@@H]1C=C(C)CC[C@H]1C(C)=C</td>\n",
       "      <td>24277</td>\n",
       "      <td>0.050962</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.051053</td>\n",
       "      <td>-0.033733</td>\n",
       "      <td>0.271953</td>\n",
       "      <td>0.172714</td>\n",
       "      <td>0.145941</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.030338</td>\n",
       "      <td>0.390687</td>\n",
       "      <td>-0.058560</td>\n",
       "      <td>0.065822</td>\n",
       "      <td>-0.113721</td>\n",
       "      <td>-0.089807</td>\n",
       "      <td>-0.041392</td>\n",
       "      <td>-0.248967</td>\n",
       "      <td>-0.126183</td>\n",
       "      <td>0.018563</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>10.00</td>\n",
       "      <td>COc1cc(O)cc(\\C=C\\c2ccccc2)c1</td>\n",
       "      <td>24277</td>\n",
       "      <td>-0.098185</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.044212</td>\n",
       "      <td>0.164668</td>\n",
       "      <td>0.011985</td>\n",
       "      <td>0.127098</td>\n",
       "      <td>0.094736</td>\n",
       "      <td>...</td>\n",
       "      <td>0.102946</td>\n",
       "      <td>0.279148</td>\n",
       "      <td>0.029891</td>\n",
       "      <td>0.079777</td>\n",
       "      <td>-0.007979</td>\n",
       "      <td>0.081155</td>\n",
       "      <td>-0.089633</td>\n",
       "      <td>0.043140</td>\n",
       "      <td>-0.034355</td>\n",
       "      <td>0.052914</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>10.00</td>\n",
       "      <td>COc1c(O)cc2C(=O)O[C@H]3[C@@H](O)[C@H](O)[C@@H]...</td>\n",
       "      <td>24277</td>\n",
       "      <td>0.410722</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.236208</td>\n",
       "      <td>-0.055084</td>\n",
       "      <td>0.694648</td>\n",
       "      <td>0.516785</td>\n",
       "      <td>0.154239</td>\n",
       "      <td>...</td>\n",
       "      <td>0.470516</td>\n",
       "      <td>0.214792</td>\n",
       "      <td>0.015408</td>\n",
       "      <td>0.339100</td>\n",
       "      <td>-0.034251</td>\n",
       "      <td>-0.047597</td>\n",
       "      <td>0.103611</td>\n",
       "      <td>-0.139579</td>\n",
       "      <td>-0.042241</td>\n",
       "      <td>0.249977</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4.39</td>\n",
       "      <td>CCOc1ccc2ccccc2c1C(=O)N[C@H]1[C@H]2SC(C)(C)[C@...</td>\n",
       "      <td>24277</td>\n",
       "      <td>0.248559</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.020058</td>\n",
       "      <td>0.085772</td>\n",
       "      <td>0.678742</td>\n",
       "      <td>0.671897</td>\n",
       "      <td>0.214587</td>\n",
       "      <td>...</td>\n",
       "      <td>0.671790</td>\n",
       "      <td>0.424186</td>\n",
       "      <td>0.118873</td>\n",
       "      <td>0.415623</td>\n",
       "      <td>0.257598</td>\n",
       "      <td>0.193370</td>\n",
       "      <td>0.238471</td>\n",
       "      <td>-0.033797</td>\n",
       "      <td>0.080018</td>\n",
       "      <td>0.335890</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 897 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "    dose                                             SMILES  Plate  \\\n",
       "0   6.05                     CCCOc1cc(N)ccc1C(=O)OCCN(CC)CC  24277   \n",
       "1  10.00       COc1cc(C)cc(OC)c1[C@@H]1C=C(C)CC[C@H]1C(C)=C  24277   \n",
       "2  10.00                       COc1cc(O)cc(\\C=C\\c2ccccc2)c1  24277   \n",
       "3  10.00  COc1c(O)cc2C(=O)O[C@H]3[C@@H](O)[C@H](O)[C@@H]...  24277   \n",
       "4   4.39  CCOc1ccc2ccccc2c1C(=O)N[C@H]1[C@H]2SC(C)(C)[C@...  24277   \n",
       "\n",
       "   Cells_AreaShape_Compactness  Cells_AreaShape_EulerNumber  \\\n",
       "0                     0.256710                          NaN   \n",
       "1                     0.050962                          NaN   \n",
       "2                    -0.098185                          NaN   \n",
       "3                     0.410722                          NaN   \n",
       "4                     0.248559                          NaN   \n",
       "\n",
       "   Cells_AreaShape_Extent  Cells_AreaShape_FormFactor  \\\n",
       "0               -0.114990                   -0.237090   \n",
       "1                0.051053                   -0.033733   \n",
       "2                0.044212                    0.164668   \n",
       "3               -0.236208                   -0.055084   \n",
       "4               -0.020058                    0.085772   \n",
       "\n",
       "   Cells_AreaShape_MaxFeretDiameter  Cells_AreaShape_MaximumRadius  \\\n",
       "0                          0.615594                       0.550434   \n",
       "1                          0.271953                       0.172714   \n",
       "2                          0.011985                       0.127098   \n",
       "3                          0.694648                       0.516785   \n",
       "4                          0.678742                       0.671897   \n",
       "\n",
       "   Cells_AreaShape_Solidity  ...  Nuclei_Texture_SumEntropy_DNA_5_0  \\\n",
       "0                  0.072795  ...                           0.090021   \n",
       "1                  0.145941  ...                          -0.030338   \n",
       "2                  0.094736  ...                           0.102946   \n",
       "3                  0.154239  ...                           0.470516   \n",
       "4                  0.214587  ...                           0.671790   \n",
       "\n",
       "   Nuclei_Texture_SumEntropy_ER_3_0  Nuclei_Texture_SumEntropy_Mito_3_0  \\\n",
       "0                          0.375747                           -0.010427   \n",
       "1                          0.390687                           -0.058560   \n",
       "2                          0.279148                            0.029891   \n",
       "3                          0.214792                            0.015408   \n",
       "4                          0.424186                            0.118873   \n",
       "\n",
       "   Nuclei_Texture_SumEntropy_RNA_5_0  Nuclei_Texture_SumVariance_ER_3_0  \\\n",
       "0                           0.118997                          -0.156442   \n",
       "1                           0.065822                          -0.113721   \n",
       "2                           0.079777                          -0.007979   \n",
       "3                           0.339100                          -0.034251   \n",
       "4                           0.415623                           0.257598   \n",
       "\n",
       "   Nuclei_Texture_SumVariance_Mito_5_0  Nuclei_Texture_SumVariance_RNA_10_0  \\\n",
       "0                            -0.018014                            -0.024573   \n",
       "1                            -0.089807                            -0.041392   \n",
       "2                             0.081155                            -0.089633   \n",
       "3                            -0.047597                             0.103611   \n",
       "4                             0.193370                             0.238471   \n",
       "\n",
       "   Nuclei_Texture_Variance_AGP_5_0  Nuclei_Texture_Variance_ER_5_0  \\\n",
       "0                        -0.383890                       -0.199967   \n",
       "1                        -0.248967                       -0.126183   \n",
       "2                         0.043140                       -0.034355   \n",
       "3                        -0.139579                       -0.042241   \n",
       "4                        -0.033797                        0.080018   \n",
       "\n",
       "   Nuclei_Texture_Variance_RNA_5_0  \n",
       "0                        -0.075658  \n",
       "1                         0.018563  \n",
       "2                         0.052914  \n",
       "3                         0.249977  \n",
       "4                         0.335890  \n",
       "\n",
       "[5 rows x 897 columns]"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "CP_CSV = CP_CSV.rename(columns={\n",
    "    # 'CPD_SMILES': 'SMILES',\n",
    "    'Metadata_mmoles_per_liter2': 'dose',\n",
    "    'Metadata_Plate': 'Plate'\n",
    "})\n",
    "# CP_CSV = CP_CSV.drop(columns=['Metadata_moa'])\n",
    "CP_CSV.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total samples: 113416\n",
      "DMSO (Negative Control): 26572\n",
      "Non-DMSO (Perturbed): 86844\n",
      "\n",
      "[Step 1] Basic Cleaning & Log1p Transformation...\n",
      "   > Dropped 55 columns containing only NaNs\n",
      "   > Feature processing done. Current feature count: 839\n",
      "\n",
      "[Step 2] Performing Plate-wise Robust Normalization (The JUMP-CP Way)...\n",
      "\n",
      "[Step 3] Feature Selection...\n",
      "Features reduced from 839 to 838\n",
      "\n",
      "[Step 4] Global Splitting by SMILES & Saving to Unified HDF5...\n",
      "Unique SMILES for splitting: 20341\n",
      "Generating paired controls (random sampling from same plate)...\n",
      "Saving 86844 samples to 'combined' group...\n",
      "\n",
      "=========== DONE ===========\n",
      "HDF5 saved to: /home/bob/boom/VCBench/data/MVC/MVC_BBBC047/BBBC047_smiles_split_CP.h5\n",
      "Final Data Shape: (86844, 838)\n",
      "Split Distribution:\n",
      "{0: 17338, 1: 17376, 2: 17414, 3: 17405, 4: 17311}\n",
      "Verification passed.\n"
     ]
    }
   ],
   "source": [
    "import os\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import h5py\n",
    "from sklearn.model_selection import KFold\n",
    "from sklearn.feature_selection import VarianceThreshold\n",
    "\n",
    "# ===========================================================\n",
    "# 1. 配置与加载 (Setup & Load)\n",
    "# ===========================================================\n",
    "output_h5 = \"./BBBC047_smiles_split_CP.h5\" \n",
    "RANDOM_SEED = 42  # 保证未来 GE 数据划分时的一致性\n",
    "\n",
    "# 假设 CP_CSV 是你环境中已经存在的 DataFrame\n",
    "# df = pd.read_csv(\"BBBC047_CP_data.csv\") \n",
    "df = CP_CSV.copy()\n",
    "\n",
    "# 分离元数据和特征\n",
    "meta_cols = [\"dose\", \"SMILES\", \"Plate\"]\n",
    "feature_cols = [c for c in df.columns if c not in meta_cols]\n",
    "df_meta = df[meta_cols].copy()\n",
    "df_features = df[feature_cols].apply(pd.to_numeric, errors='coerce')\n",
    "\n",
    "# 标记 DMSO\n",
    "dmso_mask = df_meta[\"SMILES\"].astype(str).str.contains(\"DMSO\", case=False, na=False)\n",
    "non_dmso_mask = ~dmso_mask\n",
    "\n",
    "print(f\"Total samples: {len(df)}\")\n",
    "print(f\"DMSO (Negative Control): {dmso_mask.sum()}\")\n",
    "print(f\"Non-DMSO (Perturbed): {non_dmso_mask.sum()}\")\n",
    "\n",
    "# ===========================================================\n",
    "# 2. 预处理：清洗与对数变换 (Cleaning & Log Transform)\n",
    "# ===========================================================\n",
    "print(\"\\n[Step 1] Basic Cleaning & Log1p Transformation...\")\n",
    "\n",
    "# 1. 替换 Inf 为 NaN\n",
    "df_features = df_features.replace([np.inf, -np.inf], np.nan)\n",
    "\n",
    "# 2. 删除全部为 NaN 的列\n",
    "original_n_cols = df_features.shape[1]\n",
    "df_features = df_features.dropna(axis=1, how='all')\n",
    "dropped_nan_cols = original_n_cols - df_features.shape[1]\n",
    "print(f\"   > Dropped {dropped_nan_cols} columns containing only NaNs\")\n",
    "\n",
    "# 3. 均值填充\n",
    "df_features = df_features.fillna(df_features.mean())\n",
    "\n",
    "# 4. 【兜底填充】防止均值也是 NaN 的情况 (即残留 NaN 补 0)\n",
    "if df_features.isnull().values.any():\n",
    "    print(\"   > Warning: Residual NaNs found after mean filling. Filling with 0.\")\n",
    "    df_features = df_features.fillna(0)\n",
    "\n",
    "# 5. 【移除零方差列】\n",
    "var_mask = (df_features.var() > 1e-9)\n",
    "if (~var_mask).sum() > 0:\n",
    "    print(f\"   > Dropped {(~var_mask).sum()} columns with zero/near-zero variance\")\n",
    "    df_features = df_features.loc[:, var_mask]\n",
    "\n",
    "# 6. 【关键】同步更新 feature_cols 列表\n",
    "feature_cols = df_features.columns.tolist()\n",
    "\n",
    "# 7. Log1p 变换\n",
    "cols_to_log = [c for c in feature_cols if df_features[c].max() > 50 and df_features[c].min() >= 0]\n",
    "if cols_to_log:\n",
    "    print(f\"Applying Log1p to {len(cols_to_log)} features...\")\n",
    "    df_features[cols_to_log] = np.log1p(df_features[cols_to_log])\n",
    "\n",
    "# 更新 DataFrame\n",
    "df_combined = pd.concat([df_meta, df_features], axis=1)\n",
    "print(f\"   > Feature processing done. Current feature count: {len(feature_cols)}\")\n",
    "\n",
    "# ===========================================================\n",
    "# 3. 核心步骤：按板标准化 (Plate-wise Robust Normalization)\n",
    "# ===========================================================\n",
    "print(\"\\n[Step 2] Performing Plate-wise Robust Normalization (The JUMP-CP Way)...\")\n",
    "\n",
    "normalized_data = []\n",
    "for plate_id, plate_df in df_combined.groupby(\"Plate\"):\n",
    "    \n",
    "    # 获取该板内的特征矩阵\n",
    "    X_plate = plate_df[feature_cols].values\n",
    "    is_dmso_in_plate = plate_df[\"SMILES\"].astype(str).str.contains(\"DMSO\", case=False, na=False).values\n",
    "    \n",
    "    # 真正的丢弃逻辑\n",
    "    if is_dmso_in_plate.sum() < 2:\n",
    "        print(f\"⚠️ Warning: Plate {plate_id} has < 2 DMSO samples. DROPPING this plate completely.\")\n",
    "        continue\n",
    "    \n",
    "    # 标准化计算\n",
    "    X_dmso = X_plate[is_dmso_in_plate]\n",
    "    medians = np.median(X_dmso, axis=0) # 计算每个基因维度上的中位数，更鲁棒\n",
    "    deviations = np.abs(X_dmso - medians)\n",
    "    mads = np.median(deviations, axis=0)\n",
    "    mads[mads < 1e-5] = 1.0 \n",
    "    \n",
    "    X_norm = (X_plate - medians) / (mads * 1.4826)\n",
    "    X_norm = np.clip(X_norm, -10, 10) # 截断离群值\n",
    "    \n",
    "    plate_df_norm = plate_df.copy()\n",
    "    plate_df_norm[feature_cols] = X_norm\n",
    "    normalized_data.append(plate_df_norm)\n",
    "\n",
    "if len(normalized_data) == 0:\n",
    "    raise ValueError(\"Error: No valid plates found! Check your data.\")\n",
    "\n",
    "df_norm = pd.concat(normalized_data, axis=0)\n",
    "# 重新整理索引，防止 concat 后索引混乱\n",
    "df_norm = df_norm.reset_index(drop=True)\n",
    "df_features_norm = df_norm[feature_cols]\n",
    "\n",
    "# ===========================================================\n",
    "# 4. 特征选择 (Feature Selection)\n",
    "# ===========================================================\n",
    "print(\"\\n[Step 3] Feature Selection...\")\n",
    "\n",
    "selector = VarianceThreshold(threshold=0.01)\n",
    "selector.fit(df_features_norm)\n",
    "selected_indices = selector.get_support(indices=True)\n",
    "selected_feat_cols = [feature_cols[i] for i in selected_indices]\n",
    "# 最终用于保存的特征数据\n",
    "df_features_final = df_features_norm[selected_feat_cols]\n",
    "\n",
    "print(f\"Features reduced from {len(feature_cols)} to {len(selected_feat_cols)}\")\n",
    "\n",
    "# ===========================================================\n",
    "# 5. 划分数据集与保存 (Splitting & HDF5 Saving)\n",
    "# ===========================================================\n",
    "print(\"\\n[Step 4] Global Splitting by SMILES & Saving to Unified HDF5...\")\n",
    "\n",
    "# --- 5.1 准备数据 ---\n",
    "is_dmso = df_norm[\"SMILES\"].astype(str).str.contains(\"DMSO\", case=False, na=False)\n",
    "is_drug = ~is_dmso\n",
    "\n",
    "# 提取药物数据 (Target / Post)\n",
    "# 这里必须保留原始 df_norm 的索引，方便后续 mapping\n",
    "df_drugs = df_norm[is_drug].copy()\n",
    "X_drugs = df_features_final.loc[is_drug].values.astype(np.float32)\n",
    "\n",
    "# --- 5.2 构建 DMSO 资源池 (用于随机抽样 Control) ---\n",
    "df_dmso = df_norm[is_dmso].copy()\n",
    "dmso_pool = {}\n",
    "for pid, group in df_dmso.groupby(\"Plate\"):\n",
    "    dmso_pool[str(pid)] = group[selected_feat_cols].values.astype(np.float32)\n",
    "\n",
    "# --- 5.3 计算 Fold ID (关键修改点) ---\n",
    "# 获取所有唯一的药物 SMILES\n",
    "unique_drug_smiles = sorted(df_drugs[\"SMILES\"].unique().astype(str))\n",
    "print(f\"Unique SMILES for splitting: {len(unique_drug_smiles)}\")\n",
    "\n",
    "# 建立 SMILES -> Fold ID 的映射\n",
    "smiles_to_fold = {}\n",
    "kf = KFold(n_splits=5, shuffle=True, random_state=RANDOM_SEED)\n",
    "\n",
    "# 对 Unique SMILES 进行划分\n",
    "for fold_idx, (_, test_idx) in enumerate(kf.split(unique_drug_smiles)):\n",
    "    # test_idx 里的 SMILES 被分配给当前 fold (通常作为验证集/测试集标记)\n",
    "    # 但在 HDF5 里我们只需要标记它属于哪个 \"Partition\"\n",
    "    for idx in test_idx:\n",
    "        s = unique_drug_smiles[idx]\n",
    "        smiles_to_fold[s] = fold_idx\n",
    "\n",
    "# --- 5.4 生成最终数据列表 ---\n",
    "# 我们不再按 fold 循环写入，而是生成整个数据集，带上 split_id\n",
    "final_smiles = df_drugs[\"SMILES\"].astype(str).values\n",
    "final_dose = df_drugs[\"dose\"].values\n",
    "final_plate = df_drugs[\"Plate\"].astype(str).values\n",
    "final_target = X_drugs\n",
    "final_split_id = np.array([smiles_to_fold[s] for s in final_smiles], dtype=np.int32)\n",
    "\n",
    "# 生成对应的 Control (Pre) 数据\n",
    "final_control = []\n",
    "print(\"Generating paired controls (random sampling from same plate)...\")\n",
    "for i, pid in enumerate(final_plate):\n",
    "    if pid in dmso_pool and len(dmso_pool[pid]) > 0:\n",
    "        # 随机抽取一个同板的 DMSO\n",
    "        rand_idx = np.random.randint(len(dmso_pool[pid]))\n",
    "        chosen_dmso = dmso_pool[pid][rand_idx]\n",
    "        final_control.append(chosen_dmso)\n",
    "    else:\n",
    "        # 理论上不应发生，因为之前过滤过 DMSO<2 的板\n",
    "        final_control.append(np.zeros(X_drugs.shape[1], dtype=np.float32))\n",
    "\n",
    "final_control = np.array(final_control, dtype=np.float32)\n",
    "\n",
    "# --- 5.5 写入 HDF5 (Unified Structure) ---\n",
    "dt_str = h5py.string_dtype(encoding='utf-8')\n",
    "\n",
    "with h5py.File(output_h5, \"w\") as f:\n",
    "    g_all = f.create_group(\"combined\")\n",
    "    \n",
    "    print(f\"Saving {len(final_smiles)} samples to 'combined' group...\")\n",
    "    \n",
    "    # 核心数据\n",
    "    g_all.create_dataset(\"Target\", data=final_target, compression=\"gzip\")\n",
    "    g_all.create_dataset(\"Control\", data=final_control, compression=\"gzip\")\n",
    "    \n",
    "    # 元数据\n",
    "    g_all.create_dataset(\"smiles\", data=final_smiles.astype(object), dtype=dt_str, compression=\"gzip\")\n",
    "    g_all.create_dataset(\"dose\", data=final_dose, compression=\"gzip\")\n",
    "    g_all.create_dataset(\"plate_id\", data=final_plate.astype(object), dtype=dt_str, compression=\"gzip\")\n",
    "    \n",
    "    # 【最关键的元数据】用于后续动态 Dataset 划分\n",
    "    g_all.create_dataset(\"split_id\", data=final_split_id, compression=\"gzip\")\n",
    "\n",
    "# ===========================================================\n",
    "# 6. 验证 (Validation)\n",
    "# ===========================================================\n",
    "print(\"\\n=========== DONE ===========\")\n",
    "print(f\"HDF5 saved to: {os.path.abspath(output_h5)}\")\n",
    "print(f\"Final Data Shape: {final_target.shape}\")\n",
    "print(\"Split Distribution:\")\n",
    "unique, counts = np.unique(final_split_id, return_counts=True)\n",
    "print(dict(zip(unique, counts)))\n",
    "\n",
    "assert not np.isnan(final_target).any(), \"Error: NaN found in final output!\"\n",
    "print(\"Verification passed.\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total samples: 61313\n",
      "DMSO (Negative Control): 3478\n",
      "Non-DMSO (Perturbed): 57835\n",
      "\n",
      "[Step 1] Basic Cleaning & Log1p Transformation...\n",
      "   > Dropped 0 columns containing only NaNs\n",
      "   > Feature processing done. Current feature count: 977\n",
      "\n",
      "[Step 2] Performing Plate-wise Robust Normalization (The JUMP-CP Way)...\n",
      "⚠️ Warning: Plate PAC062_U2OS_6H_X1_B1_UNI4445L has < 2 DMSO samples. DROPPING this plate completely.\n",
      "⚠️ Warning: Plate PAC062_U2OS_6H_X2_B1_UNI4445L has < 2 DMSO samples. DROPPING this plate completely.\n",
      "⚠️ Warning: Plate PAC062_U2OS_6H_X3_B1_UNI4445L has < 2 DMSO samples. DROPPING this plate completely.\n",
      "\n",
      "[Step 3] Feature Selection...\n",
      "Features reduced from 977 to 977\n",
      "\n",
      "[Step 4] Global Splitting by SMILES & Saving to Unified HDF5...\n",
      "Unique SMILES for splitting: 20294\n",
      "Generating paired controls (random sampling from same plate)...\n",
      "Saving 57692 samples to 'combined' group...\n",
      "\n",
      "=========== DONE ===========\n",
      "HDF5 saved to: /home/bob/boom/VCBench/data/MVC/MVC_BBBC047/BBBC047_smiles_split_GE.h5\n",
      "Final Data Shape: (57692, 977)\n",
      "Split Distribution:\n",
      "{0: 11742, 1: 11549, 2: 11478, 3: 11493, 4: 11430}\n",
      "Verification passed.\n"
     ]
    }
   ],
   "source": [
    "import os\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import h5py\n",
    "from sklearn.model_selection import KFold\n",
    "from sklearn.feature_selection import VarianceThreshold\n",
    "\n",
    "# ===========================================================\n",
    "# 1. 配置与加载 (Setup & Load)\n",
    "# ===========================================================\n",
    "output_h5 = \"./BBBC047_smiles_split_GE.h5\" \n",
    "RANDOM_SEED = 42  # 保证未来 GE 数据划分时的一致性\n",
    "\n",
    "# 假设 CP_CSV 是你环境中已经存在的 DataFrame\n",
    "# df = pd.read_csv(\"BBBC047_CP_data.csv\") \n",
    "df = GE_CSV.copy()\n",
    "\n",
    "# 分离元数据和特征\n",
    "meta_cols = [\"dose\", \"SMILES\", \"Plate\"]\n",
    "feature_cols = [c for c in df.columns if c not in meta_cols]\n",
    "df_meta = df[meta_cols].copy()\n",
    "df_features = df[feature_cols].apply(pd.to_numeric, errors='coerce')\n",
    "\n",
    "# 标记 DMSO\n",
    "dmso_mask = df_meta[\"SMILES\"].astype(str).str.contains(\"DMSO\", case=False, na=False)\n",
    "non_dmso_mask = ~dmso_mask\n",
    "\n",
    "print(f\"Total samples: {len(df)}\")\n",
    "print(f\"DMSO (Negative Control): {dmso_mask.sum()}\")\n",
    "print(f\"Non-DMSO (Perturbed): {non_dmso_mask.sum()}\")\n",
    "\n",
    "# ===========================================================\n",
    "# 2. 预处理：清洗与对数变换 (Cleaning & Log Transform)\n",
    "# ===========================================================\n",
    "print(\"\\n[Step 1] Basic Cleaning & Log1p Transformation...\")\n",
    "\n",
    "# 1. 替换 Inf 为 NaN\n",
    "df_features = df_features.replace([np.inf, -np.inf], np.nan)\n",
    "\n",
    "# 2. 删除全部为 NaN 的列\n",
    "original_n_cols = df_features.shape[1]\n",
    "df_features = df_features.dropna(axis=1, how='all')\n",
    "dropped_nan_cols = original_n_cols - df_features.shape[1]\n",
    "print(f\"   > Dropped {dropped_nan_cols} columns containing only NaNs\")\n",
    "\n",
    "# 3. 均值填充\n",
    "df_features = df_features.fillna(df_features.mean())\n",
    "\n",
    "# 4. 【兜底填充】防止均值也是 NaN 的情况 (即残留 NaN 补 0)\n",
    "if df_features.isnull().values.any():\n",
    "    print(\"   > Warning: Residual NaNs found after mean filling. Filling with 0.\")\n",
    "    df_features = df_features.fillna(0)\n",
    "\n",
    "# 5. 【移除零方差列】\n",
    "var_mask = (df_features.var() > 1e-9)\n",
    "if (~var_mask).sum() > 0:\n",
    "    print(f\"   > Dropped {(~var_mask).sum()} columns with zero/near-zero variance\")\n",
    "    df_features = df_features.loc[:, var_mask]\n",
    "\n",
    "# 6. 【关键】同步更新 feature_cols 列表\n",
    "feature_cols = df_features.columns.tolist()\n",
    "\n",
    "# 7. Log1p 变换\n",
    "cols_to_log = [c for c in feature_cols if df_features[c].max() > 50 and df_features[c].min() >= 0]\n",
    "if cols_to_log:\n",
    "    print(f\"Applying Log1p to {len(cols_to_log)} features...\")\n",
    "    df_features[cols_to_log] = np.log1p(df_features[cols_to_log])\n",
    "\n",
    "# 更新 DataFrame\n",
    "df_combined = pd.concat([df_meta, df_features], axis=1)\n",
    "print(f\"   > Feature processing done. Current feature count: {len(feature_cols)}\")\n",
    "\n",
    "# ===========================================================\n",
    "# 3. 核心步骤：按板标准化 (Plate-wise Robust Normalization)\n",
    "# ===========================================================\n",
    "print(\"\\n[Step 2] Performing Plate-wise Robust Normalization (The JUMP-CP Way)...\")\n",
    "\n",
    "normalized_data = []\n",
    "for plate_id, plate_df in df_combined.groupby(\"Plate\"):\n",
    "    \n",
    "    # 获取该板内的特征矩阵\n",
    "    X_plate = plate_df[feature_cols].values\n",
    "    is_dmso_in_plate = plate_df[\"SMILES\"].astype(str).str.contains(\"DMSO\", case=False, na=False).values\n",
    "    \n",
    "    # 真正的丢弃逻辑\n",
    "    if is_dmso_in_plate.sum() < 2:\n",
    "        print(f\"⚠️ Warning: Plate {plate_id} has < 2 DMSO samples. DROPPING this plate completely.\")\n",
    "        continue\n",
    "    \n",
    "    # 标准化计算\n",
    "    X_dmso = X_plate[is_dmso_in_plate]\n",
    "    medians = np.median(X_dmso, axis=0)\n",
    "    deviations = np.abs(X_dmso - medians)\n",
    "    mads = np.median(deviations, axis=0)\n",
    "    mads[mads < 1e-5] = 1.0 \n",
    "    \n",
    "    X_norm = (X_plate - medians) / (mads * 1.4826)\n",
    "    X_norm = np.clip(X_norm, -10, 10) # 截断离群值\n",
    "    \n",
    "    plate_df_norm = plate_df.copy()\n",
    "    plate_df_norm[feature_cols] = X_norm\n",
    "    normalized_data.append(plate_df_norm)\n",
    "\n",
    "if len(normalized_data) == 0:\n",
    "    raise ValueError(\"Error: No valid plates found! Check your data.\")\n",
    "\n",
    "df_norm = pd.concat(normalized_data, axis=0)\n",
    "# 重新整理索引，防止 concat 后索引混乱\n",
    "df_norm = df_norm.reset_index(drop=True)\n",
    "df_features_norm = df_norm[feature_cols]\n",
    "\n",
    "# ===========================================================\n",
    "# 4. 特征选择 (Feature Selection)\n",
    "# ===========================================================\n",
    "print(\"\\n[Step 3] Feature Selection...\")\n",
    "\n",
    "selector = VarianceThreshold(threshold=0.01)\n",
    "selector.fit(df_features_norm)\n",
    "selected_indices = selector.get_support(indices=True)\n",
    "selected_feat_cols = [feature_cols[i] for i in selected_indices]\n",
    "# 最终用于保存的特征数据\n",
    "df_features_final = df_features_norm[selected_feat_cols]\n",
    "\n",
    "print(f\"Features reduced from {len(feature_cols)} to {len(selected_feat_cols)}\")\n",
    "\n",
    "# ===========================================================\n",
    "# 5. 划分数据集与保存 (Splitting & HDF5 Saving)\n",
    "# ===========================================================\n",
    "print(\"\\n[Step 4] Global Splitting by SMILES & Saving to Unified HDF5...\")\n",
    "\n",
    "# --- 5.1 准备数据 ---\n",
    "is_dmso = df_norm[\"SMILES\"].astype(str).str.contains(\"DMSO\", case=False, na=False)\n",
    "is_drug = ~is_dmso\n",
    "\n",
    "# 提取药物数据 (Target / Post)\n",
    "# 这里必须保留原始 df_norm 的索引，方便后续 mapping\n",
    "df_drugs = df_norm[is_drug].copy()\n",
    "X_drugs = df_features_final.loc[is_drug].values.astype(np.float32)\n",
    "\n",
    "# --- 5.2 构建 DMSO 资源池 (用于随机抽样 Control) ---\n",
    "df_dmso = df_norm[is_dmso].copy()\n",
    "dmso_pool = {}\n",
    "for pid, group in df_dmso.groupby(\"Plate\"):\n",
    "    dmso_pool[str(pid)] = group[selected_feat_cols].values.astype(np.float32)\n",
    "\n",
    "# --- 5.3 计算 Fold ID (关键修改点) ---\n",
    "# 获取所有唯一的药物 SMILES\n",
    "unique_drug_smiles = sorted(df_drugs[\"SMILES\"].unique().astype(str))\n",
    "print(f\"Unique SMILES for splitting: {len(unique_drug_smiles)}\")\n",
    "\n",
    "# 建立 SMILES -> Fold ID 的映射\n",
    "smiles_to_fold = {}\n",
    "kf = KFold(n_splits=5, shuffle=True, random_state=RANDOM_SEED)\n",
    "\n",
    "# 对 Unique SMILES 进行划分\n",
    "for fold_idx, (_, test_idx) in enumerate(kf.split(unique_drug_smiles)):\n",
    "    # test_idx 里的 SMILES 被分配给当前 fold (通常作为验证集/测试集标记)\n",
    "    # 但在 HDF5 里我们只需要标记它属于哪个 \"Partition\"\n",
    "    for idx in test_idx:\n",
    "        s = unique_drug_smiles[idx]\n",
    "        smiles_to_fold[s] = fold_idx\n",
    "\n",
    "# --- 5.4 生成最终数据列表 ---\n",
    "# 我们不再按 fold 循环写入，而是生成整个数据集，带上 split_id\n",
    "final_smiles = df_drugs[\"SMILES\"].astype(str).values\n",
    "final_dose = df_drugs[\"dose\"].values\n",
    "final_plate = df_drugs[\"Plate\"].astype(str).values\n",
    "final_target = X_drugs\n",
    "final_split_id = np.array([smiles_to_fold[s] for s in final_smiles], dtype=np.int32)\n",
    "\n",
    "# 生成对应的 Control (Pre) 数据\n",
    "final_control = []\n",
    "print(\"Generating paired controls (random sampling from same plate)...\")\n",
    "for i, pid in enumerate(final_plate):\n",
    "    if pid in dmso_pool and len(dmso_pool[pid]) > 0:\n",
    "        # 随机抽取一个同板的 DMSO\n",
    "        rand_idx = np.random.randint(len(dmso_pool[pid]))\n",
    "        chosen_dmso = dmso_pool[pid][rand_idx]\n",
    "        final_control.append(chosen_dmso)\n",
    "    else:\n",
    "        # 理论上不应发生，因为之前过滤过 DMSO<2 的板\n",
    "        final_control.append(np.zeros(X_drugs.shape[1], dtype=np.float32))\n",
    "\n",
    "final_control = np.array(final_control, dtype=np.float32)\n",
    "\n",
    "# --- 5.5 写入 HDF5 (Unified Structure) ---\n",
    "dt_str = h5py.string_dtype(encoding='utf-8')\n",
    "\n",
    "with h5py.File(output_h5, \"w\") as f:\n",
    "    g_all = f.create_group(\"combined\")\n",
    "    \n",
    "    print(f\"Saving {len(final_smiles)} samples to 'combined' group...\")\n",
    "    \n",
    "    # 核心数据\n",
    "    g_all.create_dataset(\"Target\", data=final_target, compression=\"gzip\")\n",
    "    g_all.create_dataset(\"Control\", data=final_control, compression=\"gzip\")\n",
    "    \n",
    "    # 元数据\n",
    "    g_all.create_dataset(\"smiles\", data=final_smiles.astype(object), dtype=dt_str, compression=\"gzip\")\n",
    "    g_all.create_dataset(\"dose\", data=final_dose, compression=\"gzip\")\n",
    "    g_all.create_dataset(\"plate_id\", data=final_plate.astype(object), dtype=dt_str, compression=\"gzip\")\n",
    "    \n",
    "    # 【最关键的元数据】用于后续动态 Dataset 划分\n",
    "    g_all.create_dataset(\"split_id\", data=final_split_id, compression=\"gzip\")\n",
    "\n",
    "# ===========================================================\n",
    "# 6. 验证 (Validation)\n",
    "# ===========================================================\n",
    "print(\"\\n=========== DONE ===========\")\n",
    "print(f\"HDF5 saved to: {os.path.abspath(output_h5)}\")\n",
    "print(f\"Final Data Shape: {final_target.shape}\")\n",
    "print(\"Split Distribution:\")\n",
    "unique, counts = np.unique(final_split_id, return_counts=True)\n",
    "print(dict(zip(unique, counts)))\n",
    "\n",
    "assert not np.isnan(final_target).any(), \"Error: NaN found in final output!\"\n",
    "print(\"Verification passed.\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "===== Control Statistics =====\n",
      "Shape: (86844, 838)\n",
      "Min: -10.0\n",
      "Max: 10.0\n",
      "Mean: 0.12683004\n",
      "Median: 0.0\n",
      "Std: 1.5478053\n",
      "\n",
      "===== Target Statistics =====\n",
      "Shape: (86844, 838)\n",
      "Min: -10.0\n",
      "Max: 10.0\n",
      "Mean: 0.1153677\n",
      "Median: 0.0\n",
      "Std: 1.8243449\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import h5py\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "h5_path = \"BBBC047_smiles_split_CP.h5\"\n",
    "\n",
    "with h5py.File(h5_path, \"r\") as f:\n",
    "    morph_pre = f[\"combined/Control\"][:]\n",
    "    morph_post = f[\"combined/Target\"][:]\n",
    "\n",
    "def summarize(arr, name):\n",
    "    print(f\"\\n===== {name} Statistics =====\")\n",
    "    print(\"Shape:\", arr.shape)\n",
    "    print(\"Min:\", np.min(arr))\n",
    "    print(\"Max:\", np.max(arr))\n",
    "    print(\"Mean:\", np.mean(arr))\n",
    "    print(\"Median:\", np.median(arr))\n",
    "    print(\"Std:\", np.std(arr))\n",
    "\n",
    "summarize(morph_pre, \"Control\")\n",
    "summarize(morph_post, \"Target\")\n",
    "\n",
    "# -------- Plot hist distribution --------\n",
    "plt.figure(figsize=(10,5))\n",
    "plt.hist(morph_pre.flatten(), bins=200, alpha=0.6, label=\"morph_pre\")\n",
    "plt.hist(morph_post.flatten(), bins=200, alpha=0.6, label=\"morph_post\")\n",
    "plt.title(\"Distribution of Control & Target\")\n",
    "plt.xlabel(\"Value\")\n",
    "plt.ylabel(\"Frequency\")\n",
    "plt.legend()\n",
    "plt.grid(True)\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "===== Control Statistics =====\n",
      "Shape: (57692, 977)\n",
      "Min: -10.0\n",
      "Max: 10.0\n",
      "Mean: -0.008263751\n",
      "Median: 0.0\n",
      "Std: 1.3541602\n",
      "\n",
      "===== Target Statistics =====\n",
      "Shape: (57692, 977)\n",
      "Min: -10.0\n",
      "Max: 10.0\n",
      "Mean: -0.016291842\n",
      "Median: -0.0020430526\n",
      "Std: 1.9742315\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import h5py\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "h5_path = \"BBBC047_smiles_split_GE.h5\"\n",
    "\n",
    "with h5py.File(h5_path, \"r\") as f:\n",
    "    morph_pre = f[\"combined/Control\"][:]\n",
    "    morph_post = f[\"combined/Target\"][:]\n",
    "\n",
    "def summarize(arr, name):\n",
    "    print(f\"\\n===== {name} Statistics =====\")\n",
    "    print(\"Shape:\", arr.shape)\n",
    "    print(\"Min:\", np.min(arr))\n",
    "    print(\"Max:\", np.max(arr))\n",
    "    print(\"Mean:\", np.mean(arr))\n",
    "    print(\"Median:\", np.median(arr))\n",
    "    print(\"Std:\", np.std(arr))\n",
    "\n",
    "summarize(morph_pre, \"Control\")\n",
    "summarize(morph_post, \"Target\")\n",
    "\n",
    "# -------- Plot hist distribution --------\n",
    "plt.figure(figsize=(10,5))\n",
    "plt.hist(morph_pre.flatten(), bins=200, alpha=0.6, label=\"morph_pre\")\n",
    "plt.hist(morph_post.flatten(), bins=200, alpha=0.6, label=\"morph_post\")\n",
    "plt.title(\"Distribution of Control & Target\")\n",
    "plt.xlabel(\"Value\")\n",
    "plt.ylabel(\"Frequency\")\n",
    "plt.legend()\n",
    "plt.grid(True)\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "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>dose</th>\n",
       "      <th>Plate</th>\n",
       "      <th>SMILES</th>\n",
       "      <th>221227_x_at</th>\n",
       "      <th>212345_s_at</th>\n",
       "      <th>218597_s_at</th>\n",
       "      <th>217140_s_at</th>\n",
       "      <th>209253_at</th>\n",
       "      <th>214404_x_at</th>\n",
       "      <th>219888_at</th>\n",
       "      <th>...</th>\n",
       "      <th>218397_at</th>\n",
       "      <th>202996_at</th>\n",
       "      <th>204608_at</th>\n",
       "      <th>211071_s_at</th>\n",
       "      <th>203341_at</th>\n",
       "      <th>202801_at</th>\n",
       "      <th>206414_s_at</th>\n",
       "      <th>204978_at</th>\n",
       "      <th>205379_at</th>\n",
       "      <th>203897_at</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>50.0</td>\n",
       "      <td>PAC001_U2OS_6H_X1_B1_UNI4445L</td>\n",
       "      <td>CC(C)[C@H](CO)Nc1nc(Nc2cc(N)cc(Cl)c2)c2ncn(C(C...</td>\n",
       "      <td>-0.118505</td>\n",
       "      <td>-0.293445</td>\n",
       "      <td>-0.294088</td>\n",
       "      <td>0.292745</td>\n",
       "      <td>-0.145896</td>\n",
       "      <td>0.97358</td>\n",
       "      <td>0.247915</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.529905</td>\n",
       "      <td>0.804992</td>\n",
       "      <td>0.430269</td>\n",
       "      <td>-0.97559</td>\n",
       "      <td>-1.740605</td>\n",
       "      <td>0.340382</td>\n",
       "      <td>0.498778</td>\n",
       "      <td>0.352152</td>\n",
       "      <td>0.53591</td>\n",
       "      <td>0.52484</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>12.7</td>\n",
       "      <td>PAC001_U2OS_6H_X1_B1_UNI4445L</td>\n",
       "      <td>Ic1cccc(CSc2nnc(o2)-c2ccncc2)c1</td>\n",
       "      <td>0.043264</td>\n",
       "      <td>-0.264945</td>\n",
       "      <td>-0.000633</td>\n",
       "      <td>0.000935</td>\n",
       "      <td>0.177604</td>\n",
       "      <td>-0.50472</td>\n",
       "      <td>0.214315</td>\n",
       "      <td>...</td>\n",
       "      <td>0.228895</td>\n",
       "      <td>-0.381308</td>\n",
       "      <td>0.556969</td>\n",
       "      <td>0.25051</td>\n",
       "      <td>-0.025625</td>\n",
       "      <td>-0.024428</td>\n",
       "      <td>-0.225122</td>\n",
       "      <td>0.533052</td>\n",
       "      <td>-0.19329</td>\n",
       "      <td>-0.25436</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>50.0</td>\n",
       "      <td>PAC001_U2OS_6H_X1_B1_UNI4445L</td>\n",
       "      <td>Clc1cc(Cl)c(NC(=O)Nc2ccncc2)c(Cl)c1</td>\n",
       "      <td>-0.070805</td>\n",
       "      <td>0.195755</td>\n",
       "      <td>0.004606</td>\n",
       "      <td>-0.040855</td>\n",
       "      <td>-0.067346</td>\n",
       "      <td>-0.19872</td>\n",
       "      <td>-0.053195</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.031795</td>\n",
       "      <td>-0.051728</td>\n",
       "      <td>-0.017741</td>\n",
       "      <td>0.05681</td>\n",
       "      <td>0.130695</td>\n",
       "      <td>0.005919</td>\n",
       "      <td>-0.408722</td>\n",
       "      <td>-0.057177</td>\n",
       "      <td>-0.25699</td>\n",
       "      <td>0.31004</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>16.2</td>\n",
       "      <td>PAC001_U2OS_6H_X1_B1_UNI4445L</td>\n",
       "      <td>COc1ccc(CNC(=O)Nc2ncc(s2)[N+]([O-])=O)cc1</td>\n",
       "      <td>0.027165</td>\n",
       "      <td>-0.149545</td>\n",
       "      <td>0.173113</td>\n",
       "      <td>-0.100695</td>\n",
       "      <td>-0.309296</td>\n",
       "      <td>-0.20782</td>\n",
       "      <td>-0.323985</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.013787</td>\n",
       "      <td>-0.177608</td>\n",
       "      <td>0.219869</td>\n",
       "      <td>-0.16129</td>\n",
       "      <td>0.365195</td>\n",
       "      <td>-0.040848</td>\n",
       "      <td>-0.251122</td>\n",
       "      <td>0.040732</td>\n",
       "      <td>0.17091</td>\n",
       "      <td>-0.05144</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>12.5</td>\n",
       "      <td>PAC001_U2OS_6H_X1_B1_UNI4445L</td>\n",
       "      <td>OC(=O)c1ccc2c3nc4ccccc4n3c(=O)c3cccc1c23</td>\n",
       "      <td>0.303294</td>\n",
       "      <td>0.254455</td>\n",
       "      <td>-0.055418</td>\n",
       "      <td>-0.053635</td>\n",
       "      <td>0.000455</td>\n",
       "      <td>-0.10908</td>\n",
       "      <td>0.031425</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.052585</td>\n",
       "      <td>-0.196108</td>\n",
       "      <td>0.245469</td>\n",
       "      <td>0.00784</td>\n",
       "      <td>0.093595</td>\n",
       "      <td>-0.303418</td>\n",
       "      <td>0.474878</td>\n",
       "      <td>-0.022947</td>\n",
       "      <td>-0.28589</td>\n",
       "      <td>0.06423</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 980 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   dose                          Plate  \\\n",
       "0  50.0  PAC001_U2OS_6H_X1_B1_UNI4445L   \n",
       "1  12.7  PAC001_U2OS_6H_X1_B1_UNI4445L   \n",
       "2  50.0  PAC001_U2OS_6H_X1_B1_UNI4445L   \n",
       "3  16.2  PAC001_U2OS_6H_X1_B1_UNI4445L   \n",
       "4  12.5  PAC001_U2OS_6H_X1_B1_UNI4445L   \n",
       "\n",
       "                                              SMILES  221227_x_at  \\\n",
       "0  CC(C)[C@H](CO)Nc1nc(Nc2cc(N)cc(Cl)c2)c2ncn(C(C...    -0.118505   \n",
       "1                    Ic1cccc(CSc2nnc(o2)-c2ccncc2)c1     0.043264   \n",
       "2                Clc1cc(Cl)c(NC(=O)Nc2ccncc2)c(Cl)c1    -0.070805   \n",
       "3          COc1ccc(CNC(=O)Nc2ncc(s2)[N+]([O-])=O)cc1     0.027165   \n",
       "4           OC(=O)c1ccc2c3nc4ccccc4n3c(=O)c3cccc1c23     0.303294   \n",
       "\n",
       "   212345_s_at  218597_s_at  217140_s_at  209253_at  214404_x_at  219888_at  \\\n",
       "0    -0.293445    -0.294088     0.292745  -0.145896      0.97358   0.247915   \n",
       "1    -0.264945    -0.000633     0.000935   0.177604     -0.50472   0.214315   \n",
       "2     0.195755     0.004606    -0.040855  -0.067346     -0.19872  -0.053195   \n",
       "3    -0.149545     0.173113    -0.100695  -0.309296     -0.20782  -0.323985   \n",
       "4     0.254455    -0.055418    -0.053635   0.000455     -0.10908   0.031425   \n",
       "\n",
       "   ...  218397_at  202996_at  204608_at  211071_s_at  203341_at  202801_at  \\\n",
       "0  ...  -0.529905   0.804992   0.430269     -0.97559  -1.740605   0.340382   \n",
       "1  ...   0.228895  -0.381308   0.556969      0.25051  -0.025625  -0.024428   \n",
       "2  ...  -0.031795  -0.051728  -0.017741      0.05681   0.130695   0.005919   \n",
       "3  ...  -0.013787  -0.177608   0.219869     -0.16129   0.365195  -0.040848   \n",
       "4  ...  -0.052585  -0.196108   0.245469      0.00784   0.093595  -0.303418   \n",
       "\n",
       "   206414_s_at  204978_at  205379_at  203897_at  \n",
       "0     0.498778   0.352152    0.53591    0.52484  \n",
       "1    -0.225122   0.533052   -0.19329   -0.25436  \n",
       "2    -0.408722  -0.057177   -0.25699    0.31004  \n",
       "3    -0.251122   0.040732    0.17091   -0.05144  \n",
       "4     0.474878  -0.022947   -0.28589    0.06423  \n",
       "\n",
       "[5 rows x 980 columns]"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "GE_CSV.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "检测到 977 个基因特征。\n",
      "Step 2: 执行 Log2 变换 (Log2(x+1))...\n",
      "警告：数据中包含负值，跳过 Log 变换（假设已经是 Log 后的数据）。\n",
      "Step 3: 执行基于 DMSO 的 Robust Z-score...\n",
      "Step 4: Clipping to [-10, 10]...\n"
     ]
    }
   ],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "\n",
    "def process_gene_expression_data(df):\n",
    "    \n",
    "    # 1. 分离 Metadata 和 Genes\n",
    "    # 假设前 4 列是 Metadata (dose, SMILES, Plate, etc.)\n",
    "    # 实际情况请根据你的 CSV 调整\n",
    "    metadata_cols = df.columns[:3]\n",
    "    gene_cols = df.columns[3:]\n",
    "    \n",
    "    df_meta = df[metadata_cols]\n",
    "    df_genes = df[gene_cols]\n",
    "    \n",
    "    print(f\"检测到 {len(gene_cols)} 个基因特征。\")\n",
    "\n",
    "    # ==========================================\n",
    "    # Step 1: 检查是否需要 CPM/TPM 标准化\n",
    "    # ==========================================\n",
    "    # 简单的启发式检查：如果最大值非常大 (>10000) 且是整数，可能是 Raw Counts\n",
    "    if df_genes.max().max() > 10000 and df_genes.dtypes[0] == 'int64':\n",
    "        print(\"检测到 Raw Counts，正在执行 CPM 标准化...\")\n",
    "        # CPM = (counts / total_counts) * 1e6\n",
    "        library_sizes = df_genes.sum(axis=1)\n",
    "        df_genes = df_genes.div(library_sizes, axis=0) * 1e6\n",
    "    \n",
    "    # ==========================================\n",
    "    # Step 2: Log2 变换 (Log2(x+1))\n",
    "    # ==========================================\n",
    "    print(\"Step 2: 执行 Log2 变换 (Log2(x+1))...\")\n",
    "    # 确保没有负数 (基因表达量不应为负，除了已经标准化过的数据)\n",
    "    if df_genes.min().min() >= 0:\n",
    "        df_genes_log = np.log2(df_genes + 1)\n",
    "    else:\n",
    "        print(\"警告：数据中包含负值，跳过 Log 变换（假设已经是 Log 后的数据）。\")\n",
    "        df_genes_log = df_genes\n",
    "\n",
    "    # ==========================================\n",
    "    # Step 3: Robust Z-score based on DMSO\n",
    "    # ==========================================\n",
    "    print(\"Step 3: 执行基于 DMSO 的 Robust Z-score...\")\n",
    "    \n",
    "    # 寻找 DMSO\n",
    "    dmso_mask = df_meta['SMILES'].astype(str).str.contains('DMSO', case=False, na=False)\n",
    "    \n",
    "    # 如果是 L1000 数据，通常是按 Plate 进行标准化的。\n",
    "    # 为了简化，这里演示全局 DMSO 标准化。\n",
    "    # 进阶建议：如果 Metadata_Plate 存在，最好做一个 groupby('Metadata_Plate') 的循环处理。\n",
    "    \n",
    "    if not dmso_mask.any():\n",
    "        raise ValueError(\"未找到 DMSO 对照组！\")\n",
    "        \n",
    "    dmso_data = df_genes_log.loc[dmso_mask]\n",
    "    \n",
    "    # 计算统计量\n",
    "    medians = dmso_data.median()\n",
    "    mads = (dmso_data - medians).abs().median()\n",
    "    \n",
    "    # 安全处理 MAD=0\n",
    "    mads_safe = mads.replace(0, 1e-6)\n",
    "    \n",
    "    # 标准化\n",
    "    df_genes_scaled = (df_genes_log - medians) / (mads_safe * 1.4826)\n",
    "    \n",
    "    # ==========================================\n",
    "    # Step 4: 极值截断 (Clipping)\n",
    "    # ==========================================\n",
    "    # 基因表达通常不会像图像特征那样爆炸，但为了 Transformer 的稳定性，截断依然是好习惯\n",
    "    # Z-score > 10 意味着极其离谱的表达量变化，通常是噪音\n",
    "    print(\"Step 4: Clipping to [-10, 10]...\")\n",
    "    df_genes_final = df_genes_scaled.clip(-10, 10)\n",
    "\n",
    "    # 合并保存\n",
    "    df_output = pd.concat([df_meta, df_genes_final], axis=1)\n",
    "    return df_output\n",
    "    # df_output.to_csv(output_file, index=False)\n",
    "    # print(f\"处理完成，保存至: {output_file}\")\n",
    "\n",
    "# --- 执行 ---\n",
    "if __name__ == \"__main__\":\n",
    "    df_GE = process_gene_expression_data(GE_CSV)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "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>SMILES</th>\n",
       "      <th>Metadata_moa</th>\n",
       "      <th>Metadata_mmoles_per_liter2</th>\n",
       "      <th>Metadata_Plate</th>\n",
       "      <th>Cells_AreaShape_Compactness</th>\n",
       "      <th>Cells_AreaShape_Extent</th>\n",
       "      <th>Cells_AreaShape_FormFactor</th>\n",
       "      <th>Cells_AreaShape_MaxFeretDiameter</th>\n",
       "      <th>Cells_AreaShape_MaximumRadius</th>\n",
       "      <th>Cells_AreaShape_Solidity</th>\n",
       "      <th>...</th>\n",
       "      <th>Nuclei_Texture_SumEntropy_DNA_5_0</th>\n",
       "      <th>Nuclei_Texture_SumEntropy_ER_3_0</th>\n",
       "      <th>Nuclei_Texture_SumEntropy_Mito_3_0</th>\n",
       "      <th>Nuclei_Texture_SumEntropy_RNA_5_0</th>\n",
       "      <th>Nuclei_Texture_SumVariance_ER_3_0</th>\n",
       "      <th>Nuclei_Texture_SumVariance_Mito_5_0</th>\n",
       "      <th>Nuclei_Texture_SumVariance_RNA_10_0</th>\n",
       "      <th>Nuclei_Texture_Variance_AGP_5_0</th>\n",
       "      <th>Nuclei_Texture_Variance_ER_5_0</th>\n",
       "      <th>Nuclei_Texture_Variance_RNA_5_0</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>CCCOc1cc(N)ccc1C(=O)OCCN(CC)CC</td>\n",
       "      <td>local anesthetic</td>\n",
       "      <td>6.05</td>\n",
       "      <td>24277</td>\n",
       "      <td>1.798728</td>\n",
       "      <td>-1.067827</td>\n",
       "      <td>-1.390869</td>\n",
       "      <td>2.342674</td>\n",
       "      <td>2.191147</td>\n",
       "      <td>0.603093</td>\n",
       "      <td>...</td>\n",
       "      <td>0.554060</td>\n",
       "      <td>2.349977</td>\n",
       "      <td>0.060778</td>\n",
       "      <td>0.611735</td>\n",
       "      <td>-0.527389</td>\n",
       "      <td>0.040127</td>\n",
       "      <td>-0.110624</td>\n",
       "      <td>-2.284495</td>\n",
       "      <td>-1.438048</td>\n",
       "      <td>-0.403544</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>COc1cc(C)cc(OC)c1[C@@H]1C=C(C)CC[C@H]1C(C)=C</td>\n",
       "      <td>cannabinoid receptor antagonist</td>\n",
       "      <td>10.00</td>\n",
       "      <td>24277</td>\n",
       "      <td>0.303087</td>\n",
       "      <td>0.574654</td>\n",
       "      <td>-0.184627</td>\n",
       "      <td>0.967908</td>\n",
       "      <td>0.606741</td>\n",
       "      <td>1.175119</td>\n",
       "      <td>...</td>\n",
       "      <td>0.158181</td>\n",
       "      <td>2.434973</td>\n",
       "      <td>-0.133689</td>\n",
       "      <td>0.383268</td>\n",
       "      <td>-0.312258</td>\n",
       "      <td>-0.275593</td>\n",
       "      <td>-0.232614</td>\n",
       "      <td>-1.437120</td>\n",
       "      <td>-0.858000</td>\n",
       "      <td>0.206142</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>COc1cc(O)cc(\\C=C\\c2ccccc2)c1</td>\n",
       "      <td>None</td>\n",
       "      <td>10.00</td>\n",
       "      <td>24277</td>\n",
       "      <td>-0.781105</td>\n",
       "      <td>0.506980</td>\n",
       "      <td>0.992217</td>\n",
       "      <td>-0.072119</td>\n",
       "      <td>0.415398</td>\n",
       "      <td>0.774680</td>\n",
       "      <td>...</td>\n",
       "      <td>0.596572</td>\n",
       "      <td>1.800434</td>\n",
       "      <td>0.223677</td>\n",
       "      <td>0.443226</td>\n",
       "      <td>0.220224</td>\n",
       "      <td>0.476239</td>\n",
       "      <td>-0.582515</td>\n",
       "      <td>0.397441</td>\n",
       "      <td>-0.136095</td>\n",
       "      <td>0.428414</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>COc1c(O)cc2C(=O)O[C@H]3[C@@H](O)[C@H](O)[C@@H]...</td>\n",
       "      <td>interleukin inhibitor</td>\n",
       "      <td>10.00</td>\n",
       "      <td>24277</td>\n",
       "      <td>2.918283</td>\n",
       "      <td>-2.266905</td>\n",
       "      <td>-0.311273</td>\n",
       "      <td>2.658938</td>\n",
       "      <td>2.050004</td>\n",
       "      <td>1.240017</td>\n",
       "      <td>...</td>\n",
       "      <td>1.805571</td>\n",
       "      <td>1.434323</td>\n",
       "      <td>0.165160</td>\n",
       "      <td>1.557404</td>\n",
       "      <td>0.087928</td>\n",
       "      <td>-0.089972</td>\n",
       "      <td>0.819127</td>\n",
       "      <td>-0.750117</td>\n",
       "      <td>-0.198089</td>\n",
       "      <td>1.703560</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>CCOc1ccc2ccccc2c1C(=O)N[C@H]1[C@H]2SC(C)(C)[C@...</td>\n",
       "      <td>bacterial cell wall synthesis inhibitor</td>\n",
       "      <td>4.39</td>\n",
       "      <td>24277</td>\n",
       "      <td>1.739474</td>\n",
       "      <td>-0.128771</td>\n",
       "      <td>0.524232</td>\n",
       "      <td>2.595305</td>\n",
       "      <td>2.700646</td>\n",
       "      <td>1.711957</td>\n",
       "      <td>...</td>\n",
       "      <td>2.467593</td>\n",
       "      <td>2.625543</td>\n",
       "      <td>0.583185</td>\n",
       "      <td>1.886184</td>\n",
       "      <td>1.557583</td>\n",
       "      <td>0.969722</td>\n",
       "      <td>1.797292</td>\n",
       "      <td>-0.085759</td>\n",
       "      <td>0.763047</td>\n",
       "      <td>2.259479</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 745 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                              SMILES  \\\n",
       "0                     CCCOc1cc(N)ccc1C(=O)OCCN(CC)CC   \n",
       "1       COc1cc(C)cc(OC)c1[C@@H]1C=C(C)CC[C@H]1C(C)=C   \n",
       "2                       COc1cc(O)cc(\\C=C\\c2ccccc2)c1   \n",
       "3  COc1c(O)cc2C(=O)O[C@H]3[C@@H](O)[C@H](O)[C@@H]...   \n",
       "4  CCOc1ccc2ccccc2c1C(=O)N[C@H]1[C@H]2SC(C)(C)[C@...   \n",
       "\n",
       "                              Metadata_moa  Metadata_mmoles_per_liter2  \\\n",
       "0                         local anesthetic                        6.05   \n",
       "1          cannabinoid receptor antagonist                       10.00   \n",
       "2                                     None                       10.00   \n",
       "3                    interleukin inhibitor                       10.00   \n",
       "4  bacterial cell wall synthesis inhibitor                        4.39   \n",
       "\n",
       "   Metadata_Plate  Cells_AreaShape_Compactness  Cells_AreaShape_Extent  \\\n",
       "0           24277                     1.798728               -1.067827   \n",
       "1           24277                     0.303087                0.574654   \n",
       "2           24277                    -0.781105                0.506980   \n",
       "3           24277                     2.918283               -2.266905   \n",
       "4           24277                     1.739474               -0.128771   \n",
       "\n",
       "   Cells_AreaShape_FormFactor  Cells_AreaShape_MaxFeretDiameter  \\\n",
       "0                   -1.390869                          2.342674   \n",
       "1                   -0.184627                          0.967908   \n",
       "2                    0.992217                         -0.072119   \n",
       "3                   -0.311273                          2.658938   \n",
       "4                    0.524232                          2.595305   \n",
       "\n",
       "   Cells_AreaShape_MaximumRadius  Cells_AreaShape_Solidity  ...  \\\n",
       "0                       2.191147                  0.603093  ...   \n",
       "1                       0.606741                  1.175119  ...   \n",
       "2                       0.415398                  0.774680  ...   \n",
       "3                       2.050004                  1.240017  ...   \n",
       "4                       2.700646                  1.711957  ...   \n",
       "\n",
       "   Nuclei_Texture_SumEntropy_DNA_5_0  Nuclei_Texture_SumEntropy_ER_3_0  \\\n",
       "0                           0.554060                          2.349977   \n",
       "1                           0.158181                          2.434973   \n",
       "2                           0.596572                          1.800434   \n",
       "3                           1.805571                          1.434323   \n",
       "4                           2.467593                          2.625543   \n",
       "\n",
       "   Nuclei_Texture_SumEntropy_Mito_3_0  Nuclei_Texture_SumEntropy_RNA_5_0  \\\n",
       "0                            0.060778                           0.611735   \n",
       "1                           -0.133689                           0.383268   \n",
       "2                            0.223677                           0.443226   \n",
       "3                            0.165160                           1.557404   \n",
       "4                            0.583185                           1.886184   \n",
       "\n",
       "   Nuclei_Texture_SumVariance_ER_3_0  Nuclei_Texture_SumVariance_Mito_5_0  \\\n",
       "0                          -0.527389                             0.040127   \n",
       "1                          -0.312258                            -0.275593   \n",
       "2                           0.220224                             0.476239   \n",
       "3                           0.087928                            -0.089972   \n",
       "4                           1.557583                             0.969722   \n",
       "\n",
       "   Nuclei_Texture_SumVariance_RNA_10_0  Nuclei_Texture_Variance_AGP_5_0  \\\n",
       "0                            -0.110624                        -2.284495   \n",
       "1                            -0.232614                        -1.437120   \n",
       "2                            -0.582515                         0.397441   \n",
       "3                             0.819127                        -0.750117   \n",
       "4                             1.797292                        -0.085759   \n",
       "\n",
       "   Nuclei_Texture_Variance_ER_5_0  Nuclei_Texture_Variance_RNA_5_0  \n",
       "0                       -1.438048                        -0.403544  \n",
       "1                       -0.858000                         0.206142  \n",
       "2                       -0.136095                         0.428414  \n",
       "3                       -0.198089                         1.703560  \n",
       "4                        0.763047                         2.259479  \n",
       "\n",
       "[5 rows x 745 columns]"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# SMILES列移动到第一列\n",
    "\n",
    "if 'SMILES' in df_CP.columns:\n",
    "    cols = df_CP.columns.tolist()\n",
    "    cols.insert(0, cols.pop(cols.index('SMILES')))\n",
    "    df_CP = df_CP[cols]\n",
    "\n",
    "df_CP.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 规范化SMILES"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "最终 SMILES 唯一值数量: 20342\n"
     ]
    }
   ],
   "source": [
    "final_smiles = df_CP['SMILES'].unique()\n",
    "print(f\"最终 SMILES 唯一值数量: {len(final_smiles)}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "最终 SMILES 唯一值数量: 20342\n"
     ]
    }
   ],
   "source": [
    "import pandas as pd\n",
    "from rdkit import Chem\n",
    "\n",
    "def canonicalize_smiles(smiles):\n",
    "    \"\"\"尝试将一个 SMILES 转换为规范化形式，忽略 DMSO\"\"\"\n",
    "    if pd.isna(smiles):  \n",
    "        return smiles\n",
    "    if str(smiles).upper() == \"DMSO\":  # 特殊情况，直接跳过\n",
    "        return \"DMSO\"\n",
    "    try:\n",
    "        mol = Chem.MolFromSmiles(smiles)\n",
    "        if mol is not None:\n",
    "            return Chem.MolToSmiles(mol)\n",
    "        else:\n",
    "            return smiles\n",
    "    except:\n",
    "        return smiles\n",
    "\n",
    "# 应用到整列\n",
    "df_CP['SMILES'] = df_CP['SMILES'].apply(canonicalize_smiles)\n",
    "\n",
    "# 查看结果\n",
    "final_smiles = df_CP['SMILES'].unique()\n",
    "print(f\"最终 SMILES 唯一值数量: {len(final_smiles)}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 假设你想把 'old_name1' 改成 'new_name1'，'old_name2' 改成 'new_name2'\n",
    "df_CP = df_CP.rename(columns={\n",
    "    'Metadata_mmoles_per_liter2': 'dose',\n",
    "    'Metadata_Plate': 'Plate'\n",
    "})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "最终 SMILES 唯一值数量: 20342\n"
     ]
    },
    {
     "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>dose</th>\n",
       "      <th>Plate</th>\n",
       "      <th>SMILES</th>\n",
       "      <th>221227_x_at</th>\n",
       "      <th>212345_s_at</th>\n",
       "      <th>218597_s_at</th>\n",
       "      <th>217140_s_at</th>\n",
       "      <th>209253_at</th>\n",
       "      <th>214404_x_at</th>\n",
       "      <th>219888_at</th>\n",
       "      <th>...</th>\n",
       "      <th>218397_at</th>\n",
       "      <th>202996_at</th>\n",
       "      <th>204608_at</th>\n",
       "      <th>211071_s_at</th>\n",
       "      <th>203341_at</th>\n",
       "      <th>202801_at</th>\n",
       "      <th>206414_s_at</th>\n",
       "      <th>204978_at</th>\n",
       "      <th>205379_at</th>\n",
       "      <th>203897_at</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>50.0</td>\n",
       "      <td>PAC001_U2OS_6H_X1_B1_UNI4445L</td>\n",
       "      <td>CC(C)[C@H](CO)Nc1nc(Nc2cc(N)cc(Cl)c2)c2ncn(C(C...</td>\n",
       "      <td>-0.718409</td>\n",
       "      <td>-1.412662</td>\n",
       "      <td>-1.777799</td>\n",
       "      <td>1.890969</td>\n",
       "      <td>-0.686329</td>\n",
       "      <td>3.356343</td>\n",
       "      <td>1.075315</td>\n",
       "      <td>...</td>\n",
       "      <td>-1.717976</td>\n",
       "      <td>3.902234</td>\n",
       "      <td>1.456811</td>\n",
       "      <td>-3.779831</td>\n",
       "      <td>-10.000000</td>\n",
       "      <td>1.271066</td>\n",
       "      <td>2.485811</td>\n",
       "      <td>1.176754</td>\n",
       "      <td>2.000332</td>\n",
       "      <td>2.497287</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>12.7</td>\n",
       "      <td>PAC001_U2OS_6H_X1_B1_UNI4445L</td>\n",
       "      <td>Ic1cccc(CSc2nnc(-c3ccncc3)o2)c1</td>\n",
       "      <td>0.253430</td>\n",
       "      <td>-1.271711</td>\n",
       "      <td>-0.036673</td>\n",
       "      <td>0.028366</td>\n",
       "      <td>0.739463</td>\n",
       "      <td>-1.707844</td>\n",
       "      <td>0.935368</td>\n",
       "      <td>...</td>\n",
       "      <td>0.671731</td>\n",
       "      <td>-1.828476</td>\n",
       "      <td>1.888788</td>\n",
       "      <td>0.915792</td>\n",
       "      <td>-0.224897</td>\n",
       "      <td>-0.052488</td>\n",
       "      <td>-1.118856</td>\n",
       "      <td>1.812681</td>\n",
       "      <td>-0.634178</td>\n",
       "      <td>-1.210292</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>50.0</td>\n",
       "      <td>PAC001_U2OS_6H_X1_B1_UNI4445L</td>\n",
       "      <td>O=C(Nc1ccncc1)Nc1c(Cl)cc(Cl)cc1Cl</td>\n",
       "      <td>-0.431849</td>\n",
       "      <td>1.006752</td>\n",
       "      <td>-0.005589</td>\n",
       "      <td>-0.238377</td>\n",
       "      <td>-0.340128</td>\n",
       "      <td>-0.659585</td>\n",
       "      <td>-0.178832</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.149266</td>\n",
       "      <td>-0.236360</td>\n",
       "      <td>-0.070655</td>\n",
       "      <td>0.173975</td>\n",
       "      <td>0.885048</td>\n",
       "      <td>0.057613</td>\n",
       "      <td>-2.033094</td>\n",
       "      <td>-0.262186</td>\n",
       "      <td>-0.864318</td>\n",
       "      <td>1.475228</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>16.2</td>\n",
       "      <td>PAC001_U2OS_6H_X1_B1_UNI4445L</td>\n",
       "      <td>COc1ccc(CNC(=O)Nc2ncc([N+](=O)[O-])s2)cc1</td>\n",
       "      <td>0.156709</td>\n",
       "      <td>-0.700982</td>\n",
       "      <td>0.994199</td>\n",
       "      <td>-0.620332</td>\n",
       "      <td>-1.406496</td>\n",
       "      <td>-0.690759</td>\n",
       "      <td>-1.306693</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.092553</td>\n",
       "      <td>-0.844454</td>\n",
       "      <td>0.739464</td>\n",
       "      <td>-0.661288</td>\n",
       "      <td>2.550107</td>\n",
       "      <td>-0.112061</td>\n",
       "      <td>-1.248323</td>\n",
       "      <td>0.082003</td>\n",
       "      <td>0.681632</td>\n",
       "      <td>-0.244761</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>12.5</td>\n",
       "      <td>PAC001_U2OS_6H_X1_B1_UNI4445L</td>\n",
       "      <td>O=C(O)c1ccc2c3c1cccc3c(=O)n1c3ccccc3nc21</td>\n",
       "      <td>1.815569</td>\n",
       "      <td>1.297062</td>\n",
       "      <td>-0.361718</td>\n",
       "      <td>-0.319951</td>\n",
       "      <td>-0.041302</td>\n",
       "      <td>-0.352507</td>\n",
       "      <td>0.173617</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.214741</td>\n",
       "      <td>-0.933822</td>\n",
       "      <td>0.826746</td>\n",
       "      <td>-0.013567</td>\n",
       "      <td>0.621621</td>\n",
       "      <td>-1.064682</td>\n",
       "      <td>2.366801</td>\n",
       "      <td>-0.141855</td>\n",
       "      <td>-0.968730</td>\n",
       "      <td>0.305618</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 980 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   dose                          Plate  \\\n",
       "0  50.0  PAC001_U2OS_6H_X1_B1_UNI4445L   \n",
       "1  12.7  PAC001_U2OS_6H_X1_B1_UNI4445L   \n",
       "2  50.0  PAC001_U2OS_6H_X1_B1_UNI4445L   \n",
       "3  16.2  PAC001_U2OS_6H_X1_B1_UNI4445L   \n",
       "4  12.5  PAC001_U2OS_6H_X1_B1_UNI4445L   \n",
       "\n",
       "                                              SMILES  221227_x_at  \\\n",
       "0  CC(C)[C@H](CO)Nc1nc(Nc2cc(N)cc(Cl)c2)c2ncn(C(C...    -0.718409   \n",
       "1                    Ic1cccc(CSc2nnc(-c3ccncc3)o2)c1     0.253430   \n",
       "2                  O=C(Nc1ccncc1)Nc1c(Cl)cc(Cl)cc1Cl    -0.431849   \n",
       "3          COc1ccc(CNC(=O)Nc2ncc([N+](=O)[O-])s2)cc1     0.156709   \n",
       "4           O=C(O)c1ccc2c3c1cccc3c(=O)n1c3ccccc3nc21     1.815569   \n",
       "\n",
       "   212345_s_at  218597_s_at  217140_s_at  209253_at  214404_x_at  219888_at  \\\n",
       "0    -1.412662    -1.777799     1.890969  -0.686329     3.356343   1.075315   \n",
       "1    -1.271711    -0.036673     0.028366   0.739463    -1.707844   0.935368   \n",
       "2     1.006752    -0.005589    -0.238377  -0.340128    -0.659585  -0.178832   \n",
       "3    -0.700982     0.994199    -0.620332  -1.406496    -0.690759  -1.306693   \n",
       "4     1.297062    -0.361718    -0.319951  -0.041302    -0.352507   0.173617   \n",
       "\n",
       "   ...  218397_at  202996_at  204608_at  211071_s_at  203341_at  202801_at  \\\n",
       "0  ...  -1.717976   3.902234   1.456811    -3.779831 -10.000000   1.271066   \n",
       "1  ...   0.671731  -1.828476   1.888788     0.915792  -0.224897  -0.052488   \n",
       "2  ...  -0.149266  -0.236360  -0.070655     0.173975   0.885048   0.057613   \n",
       "3  ...  -0.092553  -0.844454   0.739464    -0.661288   2.550107  -0.112061   \n",
       "4  ...  -0.214741  -0.933822   0.826746    -0.013567   0.621621  -1.064682   \n",
       "\n",
       "   206414_s_at  204978_at  205379_at  203897_at  \n",
       "0     2.485811   1.176754   2.000332   2.497287  \n",
       "1    -1.118856   1.812681  -0.634178  -1.210292  \n",
       "2    -2.033094  -0.262186  -0.864318   1.475228  \n",
       "3    -1.248323   0.082003   0.681632  -0.244761  \n",
       "4     2.366801  -0.141855  -0.968730   0.305618  \n",
       "\n",
       "[5 rows x 980 columns]"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "final_smiles = df_GE['SMILES'].unique()\n",
    "\n",
    "# 应用到整列\n",
    "df_GE['SMILES'] = df_GE['SMILES'].apply(canonicalize_smiles)\n",
    "\n",
    "# 查看结果\n",
    "final_smiles = df_GE['SMILES'].unique()\n",
    "print(f\"最终 SMILES 唯一值数量: {len(final_smiles)}\")\n",
    "\n",
    "df_GE.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "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>SMILES</th>\n",
       "      <th>dose</th>\n",
       "      <th>Plate</th>\n",
       "      <th>Cells_AreaShape_Compactness</th>\n",
       "      <th>Cells_AreaShape_Extent</th>\n",
       "      <th>Cells_AreaShape_FormFactor</th>\n",
       "      <th>Cells_AreaShape_MaxFeretDiameter</th>\n",
       "      <th>Cells_AreaShape_MaximumRadius</th>\n",
       "      <th>Cells_AreaShape_Solidity</th>\n",
       "      <th>Cells_AreaShape_Zernike_0_0</th>\n",
       "      <th>...</th>\n",
       "      <th>Nuclei_Texture_SumEntropy_DNA_5_0</th>\n",
       "      <th>Nuclei_Texture_SumEntropy_ER_3_0</th>\n",
       "      <th>Nuclei_Texture_SumEntropy_Mito_3_0</th>\n",
       "      <th>Nuclei_Texture_SumEntropy_RNA_5_0</th>\n",
       "      <th>Nuclei_Texture_SumVariance_ER_3_0</th>\n",
       "      <th>Nuclei_Texture_SumVariance_Mito_5_0</th>\n",
       "      <th>Nuclei_Texture_SumVariance_RNA_10_0</th>\n",
       "      <th>Nuclei_Texture_Variance_AGP_5_0</th>\n",
       "      <th>Nuclei_Texture_Variance_ER_5_0</th>\n",
       "      <th>Nuclei_Texture_Variance_RNA_5_0</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>CCCOc1cc(N)ccc1C(=O)OCCN(CC)CC</td>\n",
       "      <td>6.05</td>\n",
       "      <td>24277</td>\n",
       "      <td>1.798728</td>\n",
       "      <td>-1.067827</td>\n",
       "      <td>-1.390869</td>\n",
       "      <td>2.342674</td>\n",
       "      <td>2.191147</td>\n",
       "      <td>0.603093</td>\n",
       "      <td>-1.579134</td>\n",
       "      <td>...</td>\n",
       "      <td>0.554060</td>\n",
       "      <td>2.349977</td>\n",
       "      <td>0.060778</td>\n",
       "      <td>0.611735</td>\n",
       "      <td>-0.527389</td>\n",
       "      <td>0.040127</td>\n",
       "      <td>-0.110624</td>\n",
       "      <td>-2.284495</td>\n",
       "      <td>-1.438048</td>\n",
       "      <td>-0.403544</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>C=C(C)[C@@H]1CCC(C)=C[C@H]1c1c(OC)cc(C)cc1OC</td>\n",
       "      <td>10.00</td>\n",
       "      <td>24277</td>\n",
       "      <td>0.303087</td>\n",
       "      <td>0.574654</td>\n",
       "      <td>-0.184627</td>\n",
       "      <td>0.967908</td>\n",
       "      <td>0.606741</td>\n",
       "      <td>1.175119</td>\n",
       "      <td>-0.233026</td>\n",
       "      <td>...</td>\n",
       "      <td>0.158181</td>\n",
       "      <td>2.434973</td>\n",
       "      <td>-0.133689</td>\n",
       "      <td>0.383268</td>\n",
       "      <td>-0.312258</td>\n",
       "      <td>-0.275593</td>\n",
       "      <td>-0.232614</td>\n",
       "      <td>-1.437120</td>\n",
       "      <td>-0.858000</td>\n",
       "      <td>0.206142</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>COc1cc(O)cc(/C=C/c2ccccc2)c1</td>\n",
       "      <td>10.00</td>\n",
       "      <td>24277</td>\n",
       "      <td>-0.781105</td>\n",
       "      <td>0.506980</td>\n",
       "      <td>0.992217</td>\n",
       "      <td>-0.072119</td>\n",
       "      <td>0.415398</td>\n",
       "      <td>0.774680</td>\n",
       "      <td>0.685892</td>\n",
       "      <td>...</td>\n",
       "      <td>0.596572</td>\n",
       "      <td>1.800434</td>\n",
       "      <td>0.223677</td>\n",
       "      <td>0.443226</td>\n",
       "      <td>0.220224</td>\n",
       "      <td>0.476239</td>\n",
       "      <td>-0.582515</td>\n",
       "      <td>0.397441</td>\n",
       "      <td>-0.136095</td>\n",
       "      <td>0.428414</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>COc1c(O)cc2c(c1O)[C@H]1O[C@H](CO)[C@@H](O)[C@H...</td>\n",
       "      <td>10.00</td>\n",
       "      <td>24277</td>\n",
       "      <td>2.918283</td>\n",
       "      <td>-2.266905</td>\n",
       "      <td>-0.311273</td>\n",
       "      <td>2.658938</td>\n",
       "      <td>2.050004</td>\n",
       "      <td>1.240017</td>\n",
       "      <td>-2.649614</td>\n",
       "      <td>...</td>\n",
       "      <td>1.805571</td>\n",
       "      <td>1.434323</td>\n",
       "      <td>0.165160</td>\n",
       "      <td>1.557404</td>\n",
       "      <td>0.087928</td>\n",
       "      <td>-0.089972</td>\n",
       "      <td>0.819127</td>\n",
       "      <td>-0.750117</td>\n",
       "      <td>-0.198089</td>\n",
       "      <td>1.703560</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>CCOc1ccc2ccccc2c1C(=O)N[C@@H]1C(=O)N2[C@@H]1SC...</td>\n",
       "      <td>4.39</td>\n",
       "      <td>24277</td>\n",
       "      <td>1.739474</td>\n",
       "      <td>-0.128771</td>\n",
       "      <td>0.524232</td>\n",
       "      <td>2.595305</td>\n",
       "      <td>2.700646</td>\n",
       "      <td>1.711957</td>\n",
       "      <td>-1.661281</td>\n",
       "      <td>...</td>\n",
       "      <td>2.467593</td>\n",
       "      <td>2.625543</td>\n",
       "      <td>0.583185</td>\n",
       "      <td>1.886184</td>\n",
       "      <td>1.557583</td>\n",
       "      <td>0.969722</td>\n",
       "      <td>1.797292</td>\n",
       "      <td>-0.085759</td>\n",
       "      <td>0.763047</td>\n",
       "      <td>2.259479</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 744 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                              SMILES   dose  Plate  \\\n",
       "0                     CCCOc1cc(N)ccc1C(=O)OCCN(CC)CC   6.05  24277   \n",
       "1       C=C(C)[C@@H]1CCC(C)=C[C@H]1c1c(OC)cc(C)cc1OC  10.00  24277   \n",
       "2                       COc1cc(O)cc(/C=C/c2ccccc2)c1  10.00  24277   \n",
       "3  COc1c(O)cc2c(c1O)[C@H]1O[C@H](CO)[C@@H](O)[C@H...  10.00  24277   \n",
       "4  CCOc1ccc2ccccc2c1C(=O)N[C@@H]1C(=O)N2[C@@H]1SC...   4.39  24277   \n",
       "\n",
       "   Cells_AreaShape_Compactness  Cells_AreaShape_Extent  \\\n",
       "0                     1.798728               -1.067827   \n",
       "1                     0.303087                0.574654   \n",
       "2                    -0.781105                0.506980   \n",
       "3                     2.918283               -2.266905   \n",
       "4                     1.739474               -0.128771   \n",
       "\n",
       "   Cells_AreaShape_FormFactor  Cells_AreaShape_MaxFeretDiameter  \\\n",
       "0                   -1.390869                          2.342674   \n",
       "1                   -0.184627                          0.967908   \n",
       "2                    0.992217                         -0.072119   \n",
       "3                   -0.311273                          2.658938   \n",
       "4                    0.524232                          2.595305   \n",
       "\n",
       "   Cells_AreaShape_MaximumRadius  Cells_AreaShape_Solidity  \\\n",
       "0                       2.191147                  0.603093   \n",
       "1                       0.606741                  1.175119   \n",
       "2                       0.415398                  0.774680   \n",
       "3                       2.050004                  1.240017   \n",
       "4                       2.700646                  1.711957   \n",
       "\n",
       "   Cells_AreaShape_Zernike_0_0  ...  Nuclei_Texture_SumEntropy_DNA_5_0  \\\n",
       "0                    -1.579134  ...                           0.554060   \n",
       "1                    -0.233026  ...                           0.158181   \n",
       "2                     0.685892  ...                           0.596572   \n",
       "3                    -2.649614  ...                           1.805571   \n",
       "4                    -1.661281  ...                           2.467593   \n",
       "\n",
       "   Nuclei_Texture_SumEntropy_ER_3_0  Nuclei_Texture_SumEntropy_Mito_3_0  \\\n",
       "0                          2.349977                            0.060778   \n",
       "1                          2.434973                           -0.133689   \n",
       "2                          1.800434                            0.223677   \n",
       "3                          1.434323                            0.165160   \n",
       "4                          2.625543                            0.583185   \n",
       "\n",
       "   Nuclei_Texture_SumEntropy_RNA_5_0  Nuclei_Texture_SumVariance_ER_3_0  \\\n",
       "0                           0.611735                          -0.527389   \n",
       "1                           0.383268                          -0.312258   \n",
       "2                           0.443226                           0.220224   \n",
       "3                           1.557404                           0.087928   \n",
       "4                           1.886184                           1.557583   \n",
       "\n",
       "   Nuclei_Texture_SumVariance_Mito_5_0  Nuclei_Texture_SumVariance_RNA_10_0  \\\n",
       "0                             0.040127                            -0.110624   \n",
       "1                            -0.275593                            -0.232614   \n",
       "2                             0.476239                            -0.582515   \n",
       "3                            -0.089972                             0.819127   \n",
       "4                             0.969722                             1.797292   \n",
       "\n",
       "   Nuclei_Texture_Variance_AGP_5_0  Nuclei_Texture_Variance_ER_5_0  \\\n",
       "0                        -2.284495                       -1.438048   \n",
       "1                        -1.437120                       -0.858000   \n",
       "2                         0.397441                       -0.136095   \n",
       "3                        -0.750117                       -0.198089   \n",
       "4                        -0.085759                        0.763047   \n",
       "\n",
       "   Nuclei_Texture_Variance_RNA_5_0  \n",
       "0                        -0.403544  \n",
       "1                         0.206142  \n",
       "2                         0.428414  \n",
       "3                         1.703560  \n",
       "4                         2.259479  \n",
       "\n",
       "[5 rows x 744 columns]"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# df_CP 删除 Metadata_moa 这列\n",
    "df_CP.drop(columns=['Metadata_moa'], inplace=True)\n",
    "df_CP.head()    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "# save to parquet\n",
    "\n",
    "df_CP.to_parquet('Step2_CP_cleaned.parquet', index=False)\n",
    "df_GE.to_parquet('Step2_GE_cleaned.parquet', index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "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>dose</th>\n",
       "      <th>Plate</th>\n",
       "      <th>SMILES</th>\n",
       "      <th>221227_x_at</th>\n",
       "      <th>212345_s_at</th>\n",
       "      <th>218597_s_at</th>\n",
       "      <th>217140_s_at</th>\n",
       "      <th>209253_at</th>\n",
       "      <th>214404_x_at</th>\n",
       "      <th>219888_at</th>\n",
       "      <th>...</th>\n",
       "      <th>218397_at</th>\n",
       "      <th>202996_at</th>\n",
       "      <th>204608_at</th>\n",
       "      <th>211071_s_at</th>\n",
       "      <th>203341_at</th>\n",
       "      <th>202801_at</th>\n",
       "      <th>206414_s_at</th>\n",
       "      <th>204978_at</th>\n",
       "      <th>205379_at</th>\n",
       "      <th>203897_at</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>50.0</td>\n",
       "      <td>PAC001_U2OS_6H_X1_B1_UNI4445L</td>\n",
       "      <td>CC(C)[C@H](CO)Nc1nc(Nc2cc(N)cc(Cl)c2)c2ncn(C(C...</td>\n",
       "      <td>-0.718409</td>\n",
       "      <td>-1.412662</td>\n",
       "      <td>-1.777799</td>\n",
       "      <td>1.890969</td>\n",
       "      <td>-0.686329</td>\n",
       "      <td>3.356343</td>\n",
       "      <td>1.075315</td>\n",
       "      <td>...</td>\n",
       "      <td>-1.717976</td>\n",
       "      <td>3.902234</td>\n",
       "      <td>1.456811</td>\n",
       "      <td>-3.779831</td>\n",
       "      <td>-10.000000</td>\n",
       "      <td>1.271066</td>\n",
       "      <td>2.485811</td>\n",
       "      <td>1.176754</td>\n",
       "      <td>2.000332</td>\n",
       "      <td>2.497287</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>12.7</td>\n",
       "      <td>PAC001_U2OS_6H_X1_B1_UNI4445L</td>\n",
       "      <td>Ic1cccc(CSc2nnc(-c3ccncc3)o2)c1</td>\n",
       "      <td>0.253430</td>\n",
       "      <td>-1.271711</td>\n",
       "      <td>-0.036673</td>\n",
       "      <td>0.028366</td>\n",
       "      <td>0.739463</td>\n",
       "      <td>-1.707844</td>\n",
       "      <td>0.935368</td>\n",
       "      <td>...</td>\n",
       "      <td>0.671731</td>\n",
       "      <td>-1.828476</td>\n",
       "      <td>1.888788</td>\n",
       "      <td>0.915792</td>\n",
       "      <td>-0.224897</td>\n",
       "      <td>-0.052488</td>\n",
       "      <td>-1.118856</td>\n",
       "      <td>1.812681</td>\n",
       "      <td>-0.634178</td>\n",
       "      <td>-1.210292</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>50.0</td>\n",
       "      <td>PAC001_U2OS_6H_X1_B1_UNI4445L</td>\n",
       "      <td>O=C(Nc1ccncc1)Nc1c(Cl)cc(Cl)cc1Cl</td>\n",
       "      <td>-0.431849</td>\n",
       "      <td>1.006752</td>\n",
       "      <td>-0.005589</td>\n",
       "      <td>-0.238377</td>\n",
       "      <td>-0.340128</td>\n",
       "      <td>-0.659585</td>\n",
       "      <td>-0.178832</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.149266</td>\n",
       "      <td>-0.236360</td>\n",
       "      <td>-0.070655</td>\n",
       "      <td>0.173975</td>\n",
       "      <td>0.885048</td>\n",
       "      <td>0.057613</td>\n",
       "      <td>-2.033094</td>\n",
       "      <td>-0.262186</td>\n",
       "      <td>-0.864318</td>\n",
       "      <td>1.475228</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>16.2</td>\n",
       "      <td>PAC001_U2OS_6H_X1_B1_UNI4445L</td>\n",
       "      <td>COc1ccc(CNC(=O)Nc2ncc([N+](=O)[O-])s2)cc1</td>\n",
       "      <td>0.156709</td>\n",
       "      <td>-0.700982</td>\n",
       "      <td>0.994199</td>\n",
       "      <td>-0.620332</td>\n",
       "      <td>-1.406496</td>\n",
       "      <td>-0.690759</td>\n",
       "      <td>-1.306693</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.092553</td>\n",
       "      <td>-0.844454</td>\n",
       "      <td>0.739464</td>\n",
       "      <td>-0.661288</td>\n",
       "      <td>2.550107</td>\n",
       "      <td>-0.112061</td>\n",
       "      <td>-1.248323</td>\n",
       "      <td>0.082003</td>\n",
       "      <td>0.681632</td>\n",
       "      <td>-0.244761</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>12.5</td>\n",
       "      <td>PAC001_U2OS_6H_X1_B1_UNI4445L</td>\n",
       "      <td>O=C(O)c1ccc2c3c1cccc3c(=O)n1c3ccccc3nc21</td>\n",
       "      <td>1.815569</td>\n",
       "      <td>1.297062</td>\n",
       "      <td>-0.361718</td>\n",
       "      <td>-0.319951</td>\n",
       "      <td>-0.041302</td>\n",
       "      <td>-0.352507</td>\n",
       "      <td>0.173617</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.214741</td>\n",
       "      <td>-0.933822</td>\n",
       "      <td>0.826746</td>\n",
       "      <td>-0.013567</td>\n",
       "      <td>0.621621</td>\n",
       "      <td>-1.064682</td>\n",
       "      <td>2.366801</td>\n",
       "      <td>-0.141855</td>\n",
       "      <td>-0.968730</td>\n",
       "      <td>0.305618</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 980 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   dose                          Plate  \\\n",
       "0  50.0  PAC001_U2OS_6H_X1_B1_UNI4445L   \n",
       "1  12.7  PAC001_U2OS_6H_X1_B1_UNI4445L   \n",
       "2  50.0  PAC001_U2OS_6H_X1_B1_UNI4445L   \n",
       "3  16.2  PAC001_U2OS_6H_X1_B1_UNI4445L   \n",
       "4  12.5  PAC001_U2OS_6H_X1_B1_UNI4445L   \n",
       "\n",
       "                                              SMILES  221227_x_at  \\\n",
       "0  CC(C)[C@H](CO)Nc1nc(Nc2cc(N)cc(Cl)c2)c2ncn(C(C...    -0.718409   \n",
       "1                    Ic1cccc(CSc2nnc(-c3ccncc3)o2)c1     0.253430   \n",
       "2                  O=C(Nc1ccncc1)Nc1c(Cl)cc(Cl)cc1Cl    -0.431849   \n",
       "3          COc1ccc(CNC(=O)Nc2ncc([N+](=O)[O-])s2)cc1     0.156709   \n",
       "4           O=C(O)c1ccc2c3c1cccc3c(=O)n1c3ccccc3nc21     1.815569   \n",
       "\n",
       "   212345_s_at  218597_s_at  217140_s_at  209253_at  214404_x_at  219888_at  \\\n",
       "0    -1.412662    -1.777799     1.890969  -0.686329     3.356343   1.075315   \n",
       "1    -1.271711    -0.036673     0.028366   0.739463    -1.707844   0.935368   \n",
       "2     1.006752    -0.005589    -0.238377  -0.340128    -0.659585  -0.178832   \n",
       "3    -0.700982     0.994199    -0.620332  -1.406496    -0.690759  -1.306693   \n",
       "4     1.297062    -0.361718    -0.319951  -0.041302    -0.352507   0.173617   \n",
       "\n",
       "   ...  218397_at  202996_at  204608_at  211071_s_at  203341_at  202801_at  \\\n",
       "0  ...  -1.717976   3.902234   1.456811    -3.779831 -10.000000   1.271066   \n",
       "1  ...   0.671731  -1.828476   1.888788     0.915792  -0.224897  -0.052488   \n",
       "2  ...  -0.149266  -0.236360  -0.070655     0.173975   0.885048   0.057613   \n",
       "3  ...  -0.092553  -0.844454   0.739464    -0.661288   2.550107  -0.112061   \n",
       "4  ...  -0.214741  -0.933822   0.826746    -0.013567   0.621621  -1.064682   \n",
       "\n",
       "   206414_s_at  204978_at  205379_at  203897_at  \n",
       "0     2.485811   1.176754   2.000332   2.497287  \n",
       "1    -1.118856   1.812681  -0.634178  -1.210292  \n",
       "2    -2.033094  -0.262186  -0.864318   1.475228  \n",
       "3    -1.248323   0.082003   0.681632  -0.244761  \n",
       "4     2.366801  -0.141855  -0.968730   0.305618  \n",
       "\n",
       "[5 rows x 980 columns]"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_GE.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "正在处理 Cell Painting 数据...\n",
      "原始形状: (113416, 744)\n",
      "Cell Painting 处理完毕。最终样本对数量: 86844\n",
      "CP 数据已保存到 ./Processed_Paired_CP_KeepRep.h5 (包含 dose)\n",
      "\n",
      "正在处理 Gene Expression 数据...\n",
      "原始形状: (61313, 980)\n",
      "警告: 发现 143 个样本所在的 Plate 没有 DMSO。使用全局 DMSO 中位数填充。\n",
      "Gene Expression 处理完毕。最终样本对数量: 57835\n",
      "GE 数据已保存到 ./Processed_Paired_GE_KeepRep.h5 (包含 dose)\n",
      "------------------------------\n",
      "最终数据形状检查:\n",
      "CP Smiles: (86844,), Control: (86844, 741), Target: (86844, 741), Dose: (86844,)\n",
      "GE Smiles: (57835,), Control: (57835, 977), Target: (57835, 977), Dose: (57835,)\n"
     ]
    },
    {
     "ename": "",
     "evalue": "",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m在当前单元格或上一个单元格中执行代码时 Kernel 崩溃。\n",
      "\u001b[1;31m请查看单元格中的代码，以确定故障的可能原因。\n",
      "\u001b[1;31m单击<a href='https://aka.ms/vscodeJupyterKernelCrash'>此处</a>了解详细信息。\n",
      "\u001b[1;31m有关更多详细信息，请查看 Jupyter <a href='command:jupyter.viewOutput'>log</a>。"
     ]
    }
   ],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import h5py\n",
    "\n",
    "# ==========================================\n",
    "# 通用函数：处理对齐逻辑 (方案 1: 保留 Target 重复 + Control 中位数)\n",
    "# ==========================================\n",
    "def process_alignment_keep_replicates(df, data_cols, modality_name):\n",
    "    print(f\"正在处理 {modality_name} 数据...\")\n",
    "    print(f\"原始形状: {df.shape}\")\n",
    "    \n",
    "    # 1. 类型转换 (省内存)\n",
    "    # 也就是除了 Plate, SMILES, pert_dose 之外的所有列\n",
    "    df[data_cols] = df[data_cols].astype(np.float32)\n",
    "\n",
    "    # 2. 分离 Control (DMSO) 和 Target (非 DMSO)\n",
    "    df_dmso = df[df['SMILES'] == 'DMSO'].copy()\n",
    "    df_target = df[df['SMILES'] != 'DMSO'].copy()\n",
    "    \n",
    "    # 3. 计算 Control 基准向量 (按 Plate 分组取中位数)\n",
    "    # 使用中位数 (median) 比均值 (mean) 更抗干扰\n",
    "    plate_controls = df_dmso.groupby('Plate')[data_cols].median().reset_index()\n",
    "    \n",
    "    # 计算全局 Control 中位数 (用于填补那些没有 DMSO 的板)\n",
    "    if not df_dmso.empty:\n",
    "        global_control = df_dmso[data_cols].median().values.astype(np.float32)\n",
    "    else:\n",
    "        print(f\"警告: {modality_name} 数据中完全没有 DMSO，将使用全 0 填充！\")\n",
    "        global_control = np.zeros(len(data_cols), dtype=np.float32)\n",
    "\n",
    "    # 4. 执行对齐 (Merge 操作)\n",
    "    # 将计算好的 Plate Control 匹配回每一个 Target 样本\n",
    "    # suffixes=['', '_ctrl'] 表示 Target 数据列名不变，Control 数据列名加后缀\n",
    "    merged_df = pd.merge(\n",
    "        df_target, \n",
    "        plate_controls, \n",
    "        on='Plate', \n",
    "        how='left', \n",
    "        suffixes=('', '_ctrl')\n",
    "    )\n",
    "    \n",
    "    # 5. 处理缺失的 Control (即某些 Plate 没有 DMSO 的情况)\n",
    "    # 找出 Merge 后 Control 列为 NaN 的行\n",
    "    ctrl_cols = [f\"{col}_ctrl\" for col in data_cols]\n",
    "    \n",
    "    # 检查是否有缺失 Control 的行\n",
    "    missing_mask = merged_df[ctrl_cols[0]].isna()\n",
    "    missing_count = missing_mask.sum()\n",
    "    \n",
    "    if missing_count > 0:\n",
    "        print(f\"警告: 发现 {missing_count} 个样本所在的 Plate 没有 DMSO。使用全局 DMSO 中位数填充。\")\n",
    "        # 填充 NaN\n",
    "        # 构建一个临时的 DataFrame 来填充，列名要对应\n",
    "        # global_fill_values = pd.DataFrame([global_control], columns=ctrl_cols, index=[0]) \n",
    "        # 优化：直接定位赋值\n",
    "        for i, col in enumerate(ctrl_cols):\n",
    "             merged_df.loc[missing_mask, col] = global_control[i]\n",
    "\n",
    "    # 6. 提取最终数组\n",
    "    # Target 数据 (原始特征)\n",
    "    target_data = merged_df[data_cols].values.astype(np.float32)\n",
    "    # Control 数据 (对应的板内中位数)\n",
    "    control_data = merged_df[ctrl_cols].values.astype(np.float32)\n",
    "    # SMILES\n",
    "    smiles_data = merged_df['SMILES'].values.astype('S') # 转为字节串\n",
    "    \n",
    "    # --- [新增] 提取 Dose ---\n",
    "    # 假设 'dose' 列在 df_target 中本身就是数值型。如果含字符串需另行处理。\n",
    "    if 'dose' in merged_df.columns:\n",
    "        dose_data = merged_df['dose'].values.astype(np.float32)\n",
    "    else:\n",
    "        raise ValueError(f\"错误: 在 {modality_name} 数据中找不到 'dose' 列！\")\n",
    "    \n",
    "    print(f\"{modality_name} 处理完毕。最终样本对数量: {len(smiles_data)}\")\n",
    "    \n",
    "    # --- [修改] 返回值增加了 dose_data ---\n",
    "    return smiles_data, control_data, target_data, dose_data\n",
    "\n",
    "# ==========================================\n",
    "# 1. 处理 Cell Painting (CP)\n",
    "# ==========================================\n",
    "# 读取清洗后的数据\n",
    "cp_path = 'Step2_CP_cleaned.parquet'\n",
    "df_cp = pd.read_parquet(cp_path)\n",
    "\n",
    "# 识别特征列 (排除元数据)\n",
    "cp_meta_cols = ['SMILES', 'Plate', 'dose']\n",
    "cp_features = [c for c in df_cp.columns if c not in cp_meta_cols]\n",
    "\n",
    "# 执行处理 (接收 4 个返回值)\n",
    "cp_smiles, cp_control, cp_target, cp_dose = process_alignment_keep_replicates(df_cp, cp_features, \"Cell Painting\")\n",
    "\n",
    "# 保存 CP 结果\n",
    "with h5py.File('./Processed_Paired_CP_KeepRep.h5', 'w') as f:\n",
    "    f.create_dataset('canonical_smiles', data=cp_smiles)\n",
    "    f.create_dataset('control', data=cp_control, dtype=np.float32)\n",
    "    f.create_dataset('target', data=cp_target, dtype=np.float32)\n",
    "    f.create_dataset('dose', data=cp_dose, dtype=np.float32) # [新增] 保存 Dose\n",
    "print(\"CP 数据已保存到 ./Processed_Paired_CP_KeepRep.h5 (包含 dose)\\n\")\n",
    "\n",
    "# ==========================================\n",
    "# 2. 处理 Gene Expression (GE)\n",
    "# ==========================================\n",
    "# 读取清洗后的数据\n",
    "ge_path = './Step2_GE_cleaned.parquet'\n",
    "df_ge = pd.read_parquet(ge_path)\n",
    "\n",
    "# 识别特征列\n",
    "ge_meta_cols = ['SMILES', 'Plate', 'dose'] \n",
    "ge_features = [c for c in df_ge.columns if c not in ge_meta_cols]\n",
    "\n",
    "# 执行处理 (接收 4 个返回值)\n",
    "ge_smiles, ge_control, ge_target, ge_dose = process_alignment_keep_replicates(df_ge, ge_features, \"Gene Expression\")\n",
    "\n",
    "# 保存 GE 结果\n",
    "with h5py.File('./Processed_Paired_GE_KeepRep.h5', 'w') as f:\n",
    "    f.create_dataset('canonical_smiles', data=ge_smiles)\n",
    "    f.create_dataset('control', data=ge_control, dtype=np.float32)\n",
    "    f.create_dataset('target', data=ge_target, dtype=np.float32)\n",
    "    f.create_dataset('dose', data=ge_dose, dtype=np.float32) # [新增] 保存 Dose\n",
    "print(\"GE 数据已保存到 ./Processed_Paired_GE_KeepRep.h5 (包含 dose)\")\n",
    "\n",
    "# ==========================================\n",
    "# 3. 验证数据形状\n",
    "# ==========================================\n",
    "print(\"-\" * 30)\n",
    "print(\"最终数据形状检查:\")\n",
    "print(f\"CP Smiles: {cp_smiles.shape}, Control: {cp_control.shape}, Target: {cp_target.shape}, Dose: {cp_dose.shape}\")\n",
    "print(f\"GE Smiles: {ge_smiles.shape}, Control: {ge_control.shape}, Target: {ge_target.shape}, Dose: {ge_dose.shape}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "boom",
   "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": "3.11.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}