{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 筛选Sandmeyer反应候选药物 - 张夏恒反应替代分析\n",
    "\n",
    "## 背景介绍\n",
    "\n",
    "### Sandmeyer反应回顾\n",
    "Sandmeyer反应是经典的芳香胺转化方法：\n",
    "**Ar-NH₂ → [Ar-N₂]⁺ → Ar-X**\n",
    "其中 X = Cl, Br, I, CN, OH, SCN 等\n",
    "\n",
    "### 张夏恒反应\n",
    "根据论文《s41586-025-09791-5_reference.pdf》，张夏恒反应可能是一种新的替代方法，\n",
    "可以更有效地实现芳香胺到芳香卤素的转化。\n",
    "\n",
    "### 筛选策略\n",
    "我们通过识别药物分子中可能来自Sandmeyer反应的芳香卤素结构，\n",
    "找出可以考虑用张夏恒反应进行工艺优化的候选药物。\n",
    "\n",
    "**重要提醒：**\n",
    "- 此筛选仅基于分子结构特征\n",
    "- 最终需要查阅文献确认合成路线\n",
    "- 并非所有含卤素的药物都使用Sandmeyer反应合成"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 导入所需库"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "from pathlib import Path\n",
    "from rdkit.Chem.Draw import rdMolDraw2D\n",
    "from IPython.display import SVG, display\n",
    "from rdkit.Chem import AllChem\n",
    "\n",
    "# 创建输出目录\n",
    "output_base = Path(\"../data/drug_targetmol\")\n",
    "scheme_a_dir = output_base / \"scheme_A_visualizations\"\n",
    "scheme_b_dir = output_base / \"scheme_B_visualizations\"\n",
    "\n",
    "scheme_a_dir.mkdir(exist_ok=True)\n",
    "scheme_b_dir.mkdir(exist_ok=True)\n",
    "\n",
    "print(f\"创建目录：{scheme_a_dir}\")\n",
    "print(f\"创建目录：{scheme_b_dir}\")\n",
    "\n",
    "def generate_highlighted_svg(mol, highlight_atoms, filename, title=\"\"):\n",
    "    \"\"\"生成高亮卤素结构的高清晰度SVG图片\"\"\"\n",
    "    from rdkit.Chem import AllChem\n",
    "    \n",
    "    # 计算2D坐标\n",
    "    AllChem.Compute2DCoords(mol)\n",
    "    \n",
    "    # 创建SVG绘制器\n",
    "    drawer = rdMolDraw2D.MolDraw2DSVG(1200, 900)  # 更大的尺寸以提高清晰度\n",
    "    drawer.SetFontSize(12)\n",
    "    \n",
    "    # 绘制选项\n",
    "    draw_options = drawer.drawOptions()\n",
    "    draw_options.addAtomIndices = False  # 不显示原子索引，保持简洁\n",
    "    draw_options.addBondIndices = False\n",
    "    draw_options.addStereoAnnotation = True\n",
    "    draw_options.fixedFontSize = 12\n",
    "    \n",
    "    # 高亮卤素原子（红色）\n",
    "    atom_colors = {}\n",
    "    for atom_idx in highlight_atoms:\n",
    "        atom_colors[atom_idx] = (1.0, 0.3, 0.3)  # 红色高亮\n",
    "    \n",
    "    # 绘制分子\n",
    "    drawer.DrawMolecule(mol, \n",
    "                       highlightAtoms=highlight_atoms,\n",
    "                       highlightAtomColors=atom_colors)\n",
    "    \n",
    "    drawer.FinishDrawing()\n",
    "    svg_content = drawer.GetDrawingText()\n",
    "    \n",
    "    # 添加标题\n",
    "    if title:\n",
    "        # 在SVG中添加标题\n",
    "        svg_lines = svg_content.split(\"\\n\")\n",
    "        # 在<g>标签前插入标题\n",
    "        for i, line in enumerate(svg_lines):\n",
    "            if \"<g \" in line and \"transform\" in line:\n",
    "                svg_lines.insert(i, f\"<text x=\"50%\" y=\"30\" text-anchor=\"middle\" font-size=\"16\" font-weight=\"bold\">{title}</text>\")\n",
    "                break\n",
    "        svg_with_title = \"\\n\".join(svg_lines)\n",
    "    else:\n",
    "        svg_with_title = svg_content\n",
    "    \n",
    "    # 保存文件\n",
    "    with open(filename, \"w\") as f:\n",
    "        f.write(svg_with_title)\n",
    "    \n",
    "    print(f\"保存SVG: {filename}\")\n",
    "    \n",
    "    return svg_content\n",
    "\n",
    "def visualize_molecules(df, output_dir, scheme_name, max_molecules=50):\n",
    "    \"\"\"为DataFrame中的分子生成可视化图片\"\"\"\n",
    "    print(f\"\\n开始生成{scheme_name}的可视化图片...\")\n",
    "    print(f\"输出目录: {output_dir}\")\n",
    "    \n",
    "    generated_count = 0\n",
    "    \n",
    "    for idx, row in df.iterrows():\n",
    "        if generated_count >= max_molecules:\n",
    "            print(f\"已达到最大生成数量限制 ({max_molecules})，停止生成\")\n",
    "            break\n",
    "            \n",
    "        cas = str(row.get(\"CAS\", \"unknown\")).strip()\n",
    "        name = str(row.get(\"Name\", \"unknown\")).strip()\n",
    "        \n",
    "        # 跳过无效的CAS号\n",
    "        if not cas or cas == \"nan\" or cas == \"unknown\":\n",
    "            continue\n",
    "            \n",
    "        mol = row.get(\"ROMol\")\n",
    "        if mol is None:\n",
    "            continue\n",
    "            \n",
    "        # 找出卤素原子\n",
    "        halogen_atoms = []\n",
    "        for atom in mol.GetAtoms():\n",
    "            if atom.GetAtomicNum() in [9, 17, 35, 53]:  # F, Cl, Br, I\n",
    "                halogen_atoms.append(atom.GetIdx())\n",
    "        \n",
    "        if not halogen_atoms:\n",
    "            continue\n",
    "            \n",
    "        # 生成文件名和标题\n",
    "        filename = output_dir / f\"{cas}.svg\"\n",
    "        title = f\"{name} ({cas})\"\n",
    "        \n",
    "        try:\n",
    "            # 生成SVG\n",
    "            generate_highlighted_svg(mol, halogen_atoms, filename, title)\n",
    "            generated_count += 1\n",
    "            \n",
    "            # 每10个显示一次进度\n",
    "            if generated_count % 10 == 0:\n",
    "                print(f\"已生成 {generated_count} 个分子图片\")\n",
    "                \n",
    "        except Exception as e:\n",
    "            print(f\"生成 {cas} 失败: {e}\")\n",
    "            continue\n",
    "    \n",
    "    print(f\"完成！共生成 {generated_count} 个{scheme_name}的可视化图片\")\n",
    "    return generated_count\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 定义筛选模式\n",
    "\n",
    "### SMARTS模式说明\n",
    "\n",
    "#### 方案A：杂芳环卤素（最高优先级）\n",
    "- **筛选逻辑**：杂芳环上的卤素最可能是Sandmeyer反应产物\n",
    "- **原因**：杂芳环直接卤代通常较困难，Sandmeyer反应是重要合成方法\n",
    "- **预期结果**：候选数量少但精准度高\n",
    "\n",
    "#### 方案B：所有芳香卤素（中等优先级）  \n",
    "- **筛选逻辑**：所有芳环上的卤素\n",
    "- **原因**：虽然有些卤素可能来自其他途径，但可以扩大筛选范围\n",
    "- **预期结果**：候选数量较多，需要更多文献验证\n",
    "\n",
    "**SMARTS模式优化说明：**\n",
    "- 原始模式 `n:c:[Cl,Br,I]` 语法有误\n",
    "- 优化为更准确的环结构匹配模式\n",
    "- 使用更精确的原子环境描述"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 定义筛选模式\n",
    "SCREENING_PATTERNS = {\n",
    "    'heteroaryl_halides': {\n",
    "        'name': '杂芳环卤素',\n",
    "        'description': '杂环上的Cl, Br, I原子（方案A）',\n",
    "        'smarts': [\n",
    "            '[n,o,s]c[Cl,Br,I]',  # 杂原子邻位卤素\n",
    "            '[n,o,s]cc[Cl,Br,I]', # 杂原子邻位卤素（隔一个碳）\n",
    "            'c1[n,o,s]c([Cl,Br,I])ccc1', # 卤代吡咯类\n",
    "            'c1c([Cl,Br,I])cncn1', # 卤代嘧啶\n",
    "            'c1ccc2c([Cl,Br,I])ccnc2c1', # 卤代喹啉\n",
    "            'c1c([Cl,Br,I])cncc1', # 卤代吡嗪\n",
    "            'c1([Cl,Br,I])scnc1', # 卤代噻唑\n",
    "        ],\n",
    "        'scheme': 'A'\n",
    "    },\n",
    "    'aryl_halides': {\n",
    "        'name': '芳香卤素',\n",
    "        'description': '所有芳环上的Cl, Br, I原子（方案B）',\n",
    "        'smarts': [\n",
    "            'c[Cl,Br,I]',  # 任意芳香氯\n",
    "            'c-C#N',       # 芳香氰基\n",
    "            'c1ccc(S(=O)(=O)N)cc1', # 磺胺核心\n",
    "            'c1c(Cl)cc(Cl)cc1',     # 多卤代苯\n",
    "        ],\n",
    "        'scheme': 'B'\n",
    "    }\n",
    "}\n",
    "\n",
    "def create_pattern_matchers():\n",
    "    \"\"\"创建SMARTS模式匹配器\"\"\"\n",
    "    matchers = {}\n",
    "    for key, pattern_info in SCREENING_PATTERNS.items():\n",
    "        matchers[key] = {\n",
    "            'info': pattern_info,\n",
    "            'matchers': [Chem.MolFromSmarts(smarts) for smarts in pattern_info['smarts']]\n",
    "        }\n",
    "    return matchers\n",
    "\n",
    "PATTERN_MATCHERS = create_pattern_matchers()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 数据加载和预处理\n",
    "\n",
    "### SDF文件说明\n",
    "- 文件位置：`data/drug_targetmol/0c04ffc9fe8c2ec916412fbdc2a49bf4.sdf`\n",
    "- 包含药物分子结构和丰富属性信息\n",
    "- 每个分子记录包含：SMILES、分子式、分子量、批准状态、适应症等"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "正在读取SDF文件...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[15:05:51] Both bonds on one end of an atropisomer are on the same side - atoms is : 3\n",
      "[15:05:51] Explicit valence for atom # 2 N greater than permitted\n",
      "[15:05:51] ERROR: Could not sanitize molecule ending on line 217340\n",
      "[15:05:51] ERROR: Explicit valence for atom # 2 N greater than permitted\n",
      "[15:05:52] Explicit valence for atom # 4 N greater than permitted\n",
      "[15:05:52] ERROR: Could not sanitize molecule ending on line 317283\n",
      "[15:05:52] ERROR: Explicit valence for atom # 4 N greater than permitted\n",
      "[15:05:52] Explicit valence for atom # 4 N greater than permitted\n",
      "[15:05:52] ERROR: Could not sanitize molecule ending on line 324666\n",
      "[15:05:52] ERROR: Explicit valence for atom # 4 N greater than permitted\n",
      "[15:05:52] Explicit valence for atom # 5 N greater than permitted\n",
      "[15:05:52] ERROR: Could not sanitize molecule ending on line 365883\n",
      "[15:05:52] ERROR: Explicit valence for atom # 5 N greater than permitted\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "成功加载 3276 个分子\n",
      "\n",
      "数据概览：\n",
      "  Index    Plate Row Col ID                           Name  \\\n",
      "0     1  L1010-1   a   2                     Dexamethasone   \n",
      "1     2  L1010-1   a   3                         Danicopan   \n",
      "2     3  L1010-1   a   4                     Cyclosporin A   \n",
      "3     4  L1010-1   a   5                       L-Carnitine   \n",
      "4     5  L1010-1   a   6     Trimetazidine dihydrochloride   \n",
      "\n",
      "                                       Synonyms           CAS  \\\n",
      "0  MK 125;Prednisolone F;NSC 34521;Hexadecadrol       50-02-2   \n",
      "1                                      ACH-4471  1903768-17-1   \n",
      "2       Cyclosporine A;Ciclosporin;Cyclosporine    59865-13-3   \n",
      "3                  L(-)-Carnitine;Levocarnitine      541-15-1   \n",
      "4               Yoshimilon;Kyurinett;Vastarel F    13171-25-0   \n",
      "\n",
      "                                                                                                SMILES  \\\n",
      "0              C[C@@H]1C[C@H]2[C@@H]3CCC4=CC(=O)C=C[C@]4(C)[C@@]3(F)[C@@H](O)C[C@]2(C)[C@@]1(O)C(=O)CO   \n",
      "1                        CC(=O)c1nn(CC(=O)N2C[C@H](F)C[C@H]2C(=O)Nc2cccc(Br)n2)c2ccc(cc12)-c1cnc(C)nc1   \n",
      "2  [C@H]([C@@H](C/C=C/C)C)(O)[C@@]1(N(C)C(=O)[C@H]([C@@H](C)C)N(C)C(=O)[C@H](CC(C)C)N(C)C(=O)[C@H](...   \n",
      "3                                                                      C[N+](C)(C)C[C@@H](O)CC([O-])=O   \n",
      "4                                                               Cl.Cl.COC1=C(OC)C(OC)=C(CN2CCNCC2)C=C1   \n",
      "\n",
      "         Formula    MolWt Approved status  \\\n",
      "0      C22H29FO5   392.46    NMPA;EMA;FDA   \n",
      "1  C26H23BrFN7O3   580.41             FDA   \n",
      "2  C62H111N11O12  1202.61             FDA   \n",
      "3       C7H15NO3    161.2             FDA   \n",
      "4  C14H24Cl2N2O3  339.258        NMPA;EMA   \n",
      "\n",
      "                                                             Pharmacopoeia  \\\n",
      "0                                            USP39-NF34;BP2015;JP16;IP2010   \n",
      "1                                                                      NaN   \n",
      "2  Martindale the Extra Pharmacopoei, EP10.2, USP43-NF38, Ph.Int_6th, JP17   \n",
      "3                                                                      NaN   \n",
      "4         BP2019;KP Ⅹ;EP9.2;IP2010;JP17;Martindale:The Extra Pharmacopoeia   \n",
      "\n",
      "                 Disease  \\\n",
      "0             Metabolism   \n",
      "1                 Others   \n",
      "2          Immune system   \n",
      "3  Cardiovascular system   \n",
      "4  Cardiovascular system   \n",
      "\n",
      "                                                                                              Pathways  \\\n",
      "0  Antibody-drug Conjugate/ADC Related;Autophagy;Endocrinology/Hormones;Immunology/Inflammation;Mic...   \n",
      "1                                                                              Immunology/Inflammation   \n",
      "2                                             Immunology/Inflammation;Metabolism;Microbiology/Virology   \n",
      "3                                                                                           Metabolism   \n",
      "4                                                                                 Autophagy;Metabolism   \n",
      "\n",
      "                                                                                                Target  \\\n",
      "0  Antibacterial;Antibiotic;Autophagy;Complement System;Glucocorticoid Receptor;IL Receptor;Mitopha...   \n",
      "1                                                                                    Complement System   \n",
      "2                                                             Phosphatase;Antibiotic;Complement System   \n",
      "3                                                            Endogenous Metabolite;Fatty Acid Synthase   \n",
      "4                                                                        Autophagy;Fatty Acid Synthase   \n",
      "\n",
      "                                                                                              Receptor  \\\n",
      "0  Antibiotic; Autophagy; Bacterial; Complement System; Glucocorticoid Receptor; IL receptor; Mitop...   \n",
      "1                                                                          Complement System; factor D   \n",
      "2                                  Antibiotic; calcineurin phosphatase; Complement System; Phosphatase   \n",
      "3                                                                           Endogenous Metabolite; FAS   \n",
      "4                                              Autophagy; mitochondrial long-chain 3-ketoacyl thiolase   \n",
      "\n",
      "                                                                                           Bioactivity  \\\n",
      "0  Dexamethasone is a glucocorticoid receptor agonist and IL receptor modulator with anti-inflammat...   \n",
      "1  Danicopan (ACH-4471) (ACH-4471) is a selective, orally active small molecule factor D inhibitor ...   \n",
      "2  Cyclosporin A is a natural product and an active fungal metabolite, classified as a cyclic polyp...   \n",
      "3  L-Carnitine (L(-)-Carnitine) is an amino acid derivative. L-Carnitine facilitates long-chain fat...   \n",
      "4  Trimetazidine dihydrochloride (Vastarel F) can improve myocardial glucose utilization by inhibit...   \n",
      "\n",
      "                                                                                             Reference  \\\n",
      "0  Li M, Yu H. Identification of WP1066, an inhibitor of JAK2 and STAT3, as a Kv1. 3 potassium chan...   \n",
      "1  Yuan X, et al. Small-molecule factor D inhibitors selectively block the alternative pathway of c...   \n",
      "2  D'Angelo G, et al. Cyclosporin A prevents the hypoxic adaptation by activating hypoxia-inducible...   \n",
      "3                                                    Jogl G, Tong L. Cell. 2003 Jan 10; 112(1):113-22.   \n",
      "4  Yang Q, et al. Int J Clin Exp Pathol. 2015, 8(4):3735-3741.;Liu Z, et al. Metabolism. 2016, 65(3...   \n",
      "\n",
      "                                              ROMol  \n",
      "0  <rdkit.Chem.rdchem.Mol object at 0x743d782049e0>  \n",
      "1  <rdkit.Chem.rdchem.Mol object at 0x743d782871b0>  \n",
      "2  <rdkit.Chem.rdchem.Mol object at 0x743d78287220>  \n",
      "3  <rdkit.Chem.rdchem.Mol object at 0x743d782873e0>  \n",
      "4  <rdkit.Chem.rdchem.Mol object at 0x743d78287450>  \n",
      "\n",
      "列名：['Index', 'Plate', 'Row', 'Col', 'ID', 'Name', 'Synonyms', 'CAS', 'SMILES', 'Formula', 'MolWt', 'Approved status', 'Pharmacopoeia', 'Disease', 'Pathways', 'Target', 'Receptor', 'Bioactivity', 'Reference', 'ROMol']\n"
     ]
    }
   ],
   "source": [
    "# 读取筛选结果CSV文件\n",
    "import pandas as pd\n",
    "from rdkit import Chem\n",
    "\n",
    "print(\"正在读取筛选结果CSV文件...\")\n",
    "\n",
    "# 读取方案A结果\n",
    "df_a = pd.read_csv(\"../data/drug_targetmol/sandmeyer_candidates_scheme_A_heteroaryl_halides.csv\")\n",
    "print(f\"方案A数据: {len(df_a)} 行\")\n",
    "\n",
    "# 读取方案B结果\n",
    "df_b = pd.read_csv(\"../data/drug_targetmol/sandmeyer_candidates_scheme_B_aryl_halides.csv\")\n",
    "print(f\"方案B数据: {len(df_b)} 行\")\n",
    "\n",
    "# 重建分子对象\n",
    "def rebuild_molecules(df):\n",
    "    mols = []\n",
    "    for idx, row in df.iterrows():\n",
    "        smiles = row.get(\"SMILES_from_mol\", \"\")\n",
    "        if smiles and str(smiles) != \"nan\":\n",
    "            mol = Chem.MolFromSmiles(str(smiles))\n",
    "            mols.append(mol)\n",
    "        else:\n",
    "            mols.append(None)\n",
    "    df[\"ROMol\"] = mols\n",
    "    valid_mols = sum(1 for m in mols if m is not None)\n",
    "    print(f\"成功重建 {valid_mols} 个分子对象\")\n",
    "    return df\n",
    "\n",
    "df_a = rebuild_molecules(df_a)\n",
    "df_b = rebuild_molecules(df_b)\n",
    "\n",
    "print(\"\n",
    "数据加载完成\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 分子筛选函数\n",
    "\n",
    "### 筛选逻辑说明\n",
    "\n",
    "1. **分子验证**：确保分子结构有效\n",
    "2. **子结构匹配**：使用RDKit的SMARTS匹配\n",
    "3. **结果记录**：记录匹配的模式和具体子结构\n",
    "4. **数据完整性**：保留所有原始属性信息"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "def screen_molecules_for_patterns(df, pattern_key):\n",
    "    \"\"\"\n",
    "    筛选包含特定子结构的分子\n",
    "    \n",
    "    Args:\n",
    "        df: 包含分子的DataFrame\n",
    "        pattern_key: 筛选模式键名\n",
    "    \n",
    "    Returns:\n",
    "        筛选结果DataFrame\n",
    "    \"\"\"\n",
    "    pattern_info = PATTERN_MATCHERS[pattern_key]['info']\n",
    "    matchers = PATTERN_MATCHERS[pattern_key]['matchers']\n",
    "    \n",
    "    print(f\"\\n开始筛选：{pattern_info['name']}\")\n",
    "    print(f\"描述：{pattern_info['description']}\")\n",
    "    print(f\"SMARTS模式数量：{len(pattern_info['smarts'])}\")\n",
    "    \n",
    "    matched_molecules = []\n",
    "    \n",
    "    for idx, row in df.iterrows():\n",
    "        mol = row['ROMol']\n",
    "        if mol is None:\n",
    "            continue\n",
    "            \n",
    "        # 检查是否匹配任何模式\n",
    "        matched_patterns = []\n",
    "        for i, matcher in enumerate(matchers):\n",
    "            if matcher is None:\n",
    "                continue\n",
    "            if mol.HasSubstructMatch(matcher):\n",
    "                matched_patterns.append({\n",
    "                    'pattern_index': i,\n",
    "                    'smarts': pattern_info['smarts'][i],\n",
    "                    'matches': len(mol.GetSubstructMatches(matcher))\n",
    "                })\n",
    "        \n",
    "        if matched_patterns:\n",
    "            # 创建匹配记录\n",
    "            match_record = row.copy()\n",
    "            match_record['matched_patterns'] = matched_patterns\n",
    "            match_record['total_matches'] = sum(p['matches'] for p in matched_patterns)\n",
    "            match_record['screening_scheme'] = pattern_info['scheme']\n",
    "            matched_molecules.append(match_record)\n",
    "    \n",
    "    result_df = pd.DataFrame(matched_molecules)\n",
    "    print(f\"找到 {len(result_df)} 个匹配分子\")\n",
    "    \n",
    "    return result_df\n",
    "\n",
    "def save_screening_results(df, filename, description):\n",
    "    \"\"\"保存筛选结果到CSV\"\"\"\n",
    "    output_path = f\"../data/drug_targetmol/{filename}\"\n",
    "    \n",
    "    # 转换ROMol列为SMILES（因为ROMol对象无法保存到CSV）\n",
    "    df_export = df.copy()\n",
    "    if 'ROMol' in df_export.columns:\n",
    "        df_export['SMILES_from_mol'] = df_export['ROMol'].apply(lambda x: Chem.MolToSmiles(x) if x else '')\n",
    "        df_export = df_export.drop('ROMol', axis=1)\n",
    "    \n",
    "    df_export.to_csv(output_path, index=False, encoding='utf-8')\n",
    "    print(f\"结果已保存到：{output_path}\")\n",
    "    print(f\"包含 {len(df_export)} 个分子，{len(df_export.columns)} 个属性列\")\n",
    "    \n",
    "    return output_path"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 方案A筛选：杂芳环卤素\n",
    "\n",
    "### 执行逻辑\n",
    "- 使用最保守的筛选策略\n",
    "- 只匹配杂芳环上的卤素\n",
    "- 预期获得高精度结果\n",
    "- 需要进一步的合成路线验证"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "开始筛选：杂芳环卤素\n",
      "描述：杂环上的Cl, Br, I原子（方案A）\n",
      "SMARTS模式数量：7\n",
      "找到 57 个匹配分子\n",
      "\n",
      "方案A筛选结果摘要：\n",
      "                       Name           CAS          Formula  \\\n",
      "1                 Danicopan  1903768-17-1    C26H23BrFN7O3   \n",
      "8                Lonafarnib   193275-84-2  C27H31Br2ClN4O2   \n",
      "19              Idoxuridine       54-42-2       C9H11IN2O5   \n",
      "144          Dimenhydrinate      523-87-5     C24H28ClN5O3   \n",
      "259           Sertaconazole    99592-32-2    C20H15Cl3N2OS   \n",
      "311             Tioconazole    65899-73-2    C16H13Cl3N2OS   \n",
      "337               Gimeracil   103766-25-2        C5H4ClNO2   \n",
      "580  Bromocriptine mesylate    22260-51-1    C33H44BrN5O8S   \n",
      "592             Clofarabine   123318-82-1    C10H11ClFN5O3   \n",
      "684             Vorasidenib  1644545-52-7     C14H13ClF6N6   \n",
      "\n",
      "                                                                                        matched_patterns  \\\n",
      "1    [{'pattern_index': 0, 'smarts': '[n,o,s]c[Cl,Br,I]', 'matches': 1}, {'pattern_index': 2, 'smarts...   \n",
      "8    [{'pattern_index': 1, 'smarts': '[n,o,s]cc[Cl,Br,I]', 'matches': 1}, {'pattern_index': 5, 'smart...   \n",
      "19   [{'pattern_index': 1, 'smarts': '[n,o,s]cc[Cl,Br,I]', 'matches': 2}, {'pattern_index': 3, 'smart...   \n",
      "144                                  [{'pattern_index': 0, 'smarts': '[n,o,s]c[Cl,Br,I]', 'matches': 2}]   \n",
      "259                                 [{'pattern_index': 1, 'smarts': '[n,o,s]cc[Cl,Br,I]', 'matches': 1}]   \n",
      "311                                  [{'pattern_index': 0, 'smarts': '[n,o,s]c[Cl,Br,I]', 'matches': 1}]   \n",
      "337  [{'pattern_index': 1, 'smarts': '[n,o,s]cc[Cl,Br,I]', 'matches': 1}, {'pattern_index': 5, 'smart...   \n",
      "580                                  [{'pattern_index': 0, 'smarts': '[n,o,s]c[Cl,Br,I]', 'matches': 1}]   \n",
      "592                                  [{'pattern_index': 0, 'smarts': '[n,o,s]c[Cl,Br,I]', 'matches': 2}]   \n",
      "684  [{'pattern_index': 0, 'smarts': '[n,o,s]c[Cl,Br,I]', 'matches': 1}, {'pattern_index': 2, 'smarts...   \n",
      "\n",
      "     total_matches  \n",
      "1                2  \n",
      "8                2  \n",
      "19               3  \n",
      "144              2  \n",
      "259              1  \n",
      "311              1  \n",
      "337              2  \n",
      "580              1  \n",
      "592              2  \n",
      "684              2  \n",
      "结果已保存到：../data/drug_targetmol/sandmeyer_candidates_scheme_A_heteroaryl_halides.csv\n",
      "包含 57 个分子，23 个属性列\n"
     ]
    }
   ],
   "source": [
    "# 执行方案A筛选\n",
    "scheme_a_results = screen_molecules_for_patterns(df, 'heteroaryl_halides')\n",
    "\n",
    "# 显示结果摘要\n",
    "if len(scheme_a_results) > 0:\n",
    "    print(\"\\n方案A筛选结果摘要：\")\n",
    "    print(scheme_a_results[['Name', 'CAS', 'Formula', 'matched_patterns', 'total_matches']].head(10))\n",
    "    \n",
    "    # 保存结果\n",
    "    save_screening_results(\n",
    "        scheme_a_results, \n",
    "        'sandmeyer_candidates_scheme_A_heteroaryl_halides.csv',\n",
    "        '方案A：杂芳环卤素筛选结果'\n",
    "    )\n",
    "else:\n",
    "    print(\"\\n方案A未找到匹配分子\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 方案B筛选：所有芳香卤素\n",
    "\n",
    "### 执行逻辑\n",
    "- 使用更宽松的筛选策略  \n",
    "- 匹配所有芳环上的卤素\n",
    "- 会包含更多候选分子\n",
    "- 需要更多的文献验证工作"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "开始筛选：芳香卤素\n",
      "描述：所有芳环上的Cl, Br, I原子（方案B）\n",
      "SMARTS模式数量：4\n",
      "找到 548 个匹配分子\n",
      "\n",
      "方案B筛选结果摘要：\n",
      "                         Name           CAS          Formula  \\\n",
      "1                   Danicopan  1903768-17-1    C26H23BrFN7O3   \n",
      "8                  Lonafarnib   193275-84-2  C27H31Br2ClN4O2   \n",
      "9                Ketoconazole    65277-42-1    C26H28Cl2N4O4   \n",
      "13                   Ozanimod  1306760-87-1       C23H24N4O3   \n",
      "14                  Ponesimod   854107-55-4    C23H25ClN2O4S   \n",
      "19                Idoxuridine       54-42-2       C9H11IN2O5   \n",
      "53                Moclobemide    71320-77-9     C13H17ClN2O2   \n",
      "74                 Clemastine    15686-51-8       C21H26ClNO   \n",
      "75  Buclizine dihydrochloride      129-74-8  C28H33ClN2·2HCl   \n",
      "78                  Asenapine    65576-45-6       C17H16ClNO   \n",
      "\n",
      "                                                                                       matched_patterns  \\\n",
      "1                                          [{'pattern_index': 0, 'smarts': 'c[Cl,Br,I]', 'matches': 1}]   \n",
      "8                                          [{'pattern_index': 0, 'smarts': 'c[Cl,Br,I]', 'matches': 3}]   \n",
      "9   [{'pattern_index': 0, 'smarts': 'c[Cl,Br,I]', 'matches': 2}, {'pattern_index': 3, 'smarts': 'c1c...   \n",
      "13                                              [{'pattern_index': 1, 'smarts': 'c-C#N', 'matches': 1}]   \n",
      "14                                         [{'pattern_index': 0, 'smarts': 'c[Cl,Br,I]', 'matches': 1}]   \n",
      "19                                         [{'pattern_index': 0, 'smarts': 'c[Cl,Br,I]', 'matches': 1}]   \n",
      "53                                         [{'pattern_index': 0, 'smarts': 'c[Cl,Br,I]', 'matches': 1}]   \n",
      "74                                         [{'pattern_index': 0, 'smarts': 'c[Cl,Br,I]', 'matches': 1}]   \n",
      "75                                         [{'pattern_index': 0, 'smarts': 'c[Cl,Br,I]', 'matches': 1}]   \n",
      "78                                         [{'pattern_index': 0, 'smarts': 'c[Cl,Br,I]', 'matches': 1}]   \n",
      "\n",
      "    total_matches  \n",
      "1               1  \n",
      "8               3  \n",
      "9               3  \n",
      "13              1  \n",
      "14              1  \n",
      "19              1  \n",
      "53              1  \n",
      "74              1  \n",
      "75              1  \n",
      "78              1  \n",
      "结果已保存到：../data/drug_targetmol/sandmeyer_candidates_scheme_B_aryl_halides.csv\n",
      "包含 548 个分子，23 个属性列\n"
     ]
    }
   ],
   "source": [
    "# 执行方案B筛选\n",
    "scheme_b_results = screen_molecules_for_patterns(df, 'aryl_halides')\n",
    "\n",
    "# 显示结果摘要\n",
    "if len(scheme_b_results) > 0:\n",
    "    print(\"\\n方案B筛选结果摘要：\")\n",
    "    print(scheme_b_results[['Name', 'CAS', 'Formula', 'matched_patterns', 'total_matches']].head(10))\n",
    "    \n",
    "    # 保存结果\n",
    "    save_screening_results(\n",
    "        scheme_b_results, \n",
    "        'sandmeyer_candidates_scheme_B_aryl_halides.csv',\n",
    "        '方案B：所有芳香卤素筛选结果'\n",
    "    )\n",
    "else:\n",
    "    print(\"\\n方案B未找到匹配分子\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 结果分析和总结\n",
    "\n",
    "### 筛选统计"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "=== 筛选结果统计 ===\n",
      "总分子数：3276\n",
      "方案A（杂芳环卤素）匹配数：57\n",
      "方案B（所有芳香卤素）匹配数：548\n",
      "两方案重叠分子数：57\n",
      "仅方案A匹配的分子数：0\n",
      "仅方案B匹配的分子数：490\n"
     ]
    }
   ],
   "source": [
    "# 结果统计\n",
    "print(\"=== 筛选结果统计 ===\")\n",
    "print(f\"总分子数：{len(df)}\")\n",
    "print(f\"方案A（杂芳环卤素）匹配数：{len(scheme_a_results)}\")\n",
    "print(f\"方案B（所有芳香卤素）匹配数：{len(scheme_b_results)}\")\n",
    "\n",
    "if len(scheme_a_results) > 0 and len(scheme_b_results) > 0:\n",
    "    # 分析重叠\n",
    "    scheme_a_cas = set(scheme_a_results['CAS'].dropna())\n",
    "    scheme_b_cas = set(scheme_b_results['CAS'].dropna())\n",
    "    overlap = scheme_a_cas & scheme_b_cas\n",
    "    print(f\"两方案重叠分子数：{len(overlap)}\")\n",
    "    \n",
    "    # 方案A特有\n",
    "    scheme_a_only = scheme_a_cas - scheme_b_cas\n",
    "    print(f\"仅方案A匹配的分子数：{len(scheme_a_only)}\")\n",
    "    \n",
    "    # 方案B特有\n",
    "    scheme_b_only = scheme_b_cas - scheme_a_cas\n",
    "    print(f\"仅方案B匹配的分子数：{len(scheme_b_only)}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 使用建议\n",
    "\n",
    "### 优先级推荐\n",
    "\n",
    "1. **第一优先级**：方案A结果\n",
    "   - 杂芳环卤素最可能是Sandmeyer反应产物\n",
    "   - 候选数量相对较少，便于深入研究\n",
    "   - 建议重点查阅这些分子的合成路线\n",
    "\n",
    "2. **第二优先级**：方案B独有结果\n",
    "   - 苯环卤素可能来自多种途径\n",
    "   - 需要仔细评估合成可能性\n",
    "   - 适合作为补充筛选\n",
    "\n",
    "### 后续验证步骤\n",
    "\n",
    "1. **文献调研**：查阅候选分子的合成路线\n",
    "2. **反应条件评估**：确认是否使用了Sandmeyer反应\n",
    "3. **经济性分析**：评估张夏恒反应用于该分子的潜力\n",
    "4. **实验验证**：必要时进行小规模验证实验\n",
    "\n",
    "### 注意事项\n",
    "\n",
    "- 此筛选基于结构特征，不等同于合成路线确认\n",
    "- 部分卤素可能来自原料而非合成步骤\n",
    "- 分子复杂程度和合成可行性需要综合考虑\n",
    "- 建议结合药物的重要性和市场规模进行优先级排序"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 分子结构可视化\n",
    "\n",
    "### 创建输出目录和可视化函数\n",
    "\n",
    "本节将为筛选出的候选分子生成高清晰度的SVG结构图，突出显示卤素结构。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.14.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}