科研 Skill
代谢组学数据检索助手
tooluniverse-metabolomics
检索代谢组学公共研究数据
Skill 说明
代谢组学公共研究检索与批量查库工具,可 fallback 查询数据集,并通过 prompt 限制输出格式。 | 适用:代谢物注释、公共数据集检索、代谢组学研究信息整理与规范报告。
SKILL.md
Metabolomics Research
Comprehensive metabolomics research skill that identifies metabolites, analyzes studies, and searches metabolomics databases. Generates structured research reports with annotated metabolite information, study details, and database statistics.
Use Case
Use this skill when asked to:
- Identify or annotate metabolites (HMDB IDs, chemical properties, pathways)
- Retrieve metabolomics study information from MetaboLights or Metabolomics Workbench
- Search for metabolomics studies by keywords or disease
- Analyze metabolite profiles or datasets
- Generate comprehensive metabolomics research reports
Example queries:
- "What is the HMDB ID and pathway information for glucose?"
- "Get study details for MTBLS1"
- "Find metabolomics studies related to diabetes"
- "Analyze these metabolites: glucose, lactate, pyruvate"
Databases Covered
Primary metabolite databases:
- HMDB (Human Metabolome Database): 220,000+ metabolites with structures, pathways, and biological roles
- MetaboLights: Public metabolomics repository with thousands of studies
- Metabolomics Workbench: NIH Common Fund metabolomics data repository
- PubChem: Chemical properties and bioactivity data (fallback)
Research Workflow
The skill executes a 4-phase analysis pipeline:
Phase 1: Metabolite Identification & Annotation
For each metabolite in the input list:
- Search HMDB by metabolite name
- Retrieve HMDB ID, chemical formula, molecular weight
- Get detailed metabolite information (description, pathways)
- Fallback to PubChem for CID and chemical properties if HMDB unavailable
Phase 2: Study Details Retrieval
For provided study IDs:
- Detect database type (MTBLS = MetaboLights, ST = Metabolomics Workbench)
- Retrieve study metadata (title, description, organism, status)
- Extract experimental design and data availability
Phase 3: Study Search
For keyword searches:
- Search MetaboLights studies by query term
- Return matching study IDs with preview information
- Report total number of results
Phase 4: Database Overview
Always included in reports:
- Sample recent studies from MetaboLights
- Database statistics and availability
- Integration information for all databases
Usage Patterns
Pattern 1: Metabolite Identification
Input:
- Metabolite list: ["glucose", "lactate", "pyruvate"]
Output report includes:
- HMDB IDs for each metabolite
- Chemical formulas and molecular weights
- Biological pathways
- PubChem CIDs
- SMILES representations
Pattern 2: Study Retrieval
Input:
- Study ID: "MTBLS1" or "ST000001"
Output report includes:
- Study title and description
- Organism information
- Study status and release date
- Data availability
Pattern 3: Study Search
Input:
- Search query: "diabetes"
- Optional organism filter
Output report includes:
- Matching study IDs
- Study titles and previews
- Total result count
Pattern 4: Comprehensive Analysis
Input:
- Metabolite list: ["glucose", "pyruvate"]
- Study ID: "MTBLS1"
- Search query: "diabetes"
Output report includes:
- All phases combined (identification, study details, search results, overview)
- Cross-referenced information
- Complete metabolomics research summary
Input Parameters
metabolite_list (optional)
List of metabolite names to identify and annotate.
- Format: List of strings
- Examples:
["glucose"],["lactate", "pyruvate", "acetate"] - Note: Common names accepted; HMDB will find standard identifiers
study_id (optional)
MetaboLights or Metabolomics Workbench study identifier.
- Format: String starting with "MTBLS" or "ST"
- Examples:
"MTBLS1","ST000001" - Note: Database auto-detected from prefix
search_query (optional)
Keyword to search metabolomics studies.
- Format: String (disease, compound, organism, method)
- Examples:
"diabetes","glucose metabolism","LC-MS"
organism (optional)
Target organism for study filtering.
- Format: String (scientific name)
- Default:
"Homo sapiens" - Examples:
"Mus musculus","Saccharomyces cerevisiae"
output_file (optional)
Path for the generated markdown report.
- Format: String (filename with .md extension)
- Default: Auto-generated timestamp-based filename
- Examples:
"my_analysis.md","metabolomics_report.md"
Output Format
All analyses generate a structured markdown report with:
Header section:
- Report title and generation timestamp
- Input parameters summary (metabolites, study ID, search query, organism)
Phase sections:
- Clear section headers (## 1. Metabolite Identification, ## 2. Study Details, etc.)
- Subsections for each metabolite or result
- Consistent formatting (bold labels, tables for results)
Database overview:
- Available databases and statistics
- Recent studies sample
- Integration information
Error handling:
- Graceful error messages for unavailable data
- Fallback strategies documented in output
- "N/A" for missing fields (not blank)
Implementation Notes
SOAP Tool Handling
HMDB tools are SOAP-based and require special parameter handling:
HMDB_search: Requiresoperation="search"parameterHMDB_get_metabolite: Requiresoperation="get_metabolite"parameter- Do not use
endpointormethodparameters (not applicable to SOAP)
Response Format Variations
Tools return different response formats - handle all three:
- Standard format:
{status: "success", data: [...], metadata: {...}} - Direct list:
[...](e.g., metabolights_list_studies) - Direct dict:
{field1: ..., field2: ...}(e.g., some detail endpoints)
Always check response type with isinstance() before accessing fields.
Fallback Strategy
Follow this hierarchy for robustness:
- Primary source: Try main database first (HMDB for metabolites, MetaboLights for studies)
- Fallback source: Use alternative database if primary fails (PubChem for chemical properties)
- Default behavior: Show error message with context, continue with remaining phases
Progressive Report Writing
Write report incrementally to avoid memory issues:
- Create output file early in pipeline
- Append sections as each phase completes
- Flush to disk regularly for long analyses
- Return file path for user access
Tool Discovery
The skill automatically discovers and uses these tools from ToolUniverse:
HMDB Tools:
HMDB_search: Search metabolites by nameHMDB_get_metabolite: Get detailed metabolite information
MetaboLights Tools:
metabolights_list_studies: List available studiesmetabolights_search_studies: Search studies by keywordmetabolights_get_study: Get study details by ID
Metabolomics Workbench Tools:
MetabolomicsWorkbench_get_study: Get study informationMetabolomicsWorkbench_search_compound_by_name: Search compounds
PubChem Tools:
PubChem_get_CID_by_compound_name: Get PubChem CIDPubChem_get_compound_properties_by_CID: Get chemical properties
No manual tool configuration required - all tools loaded automatically.
Common Issues
Issue: HMDB returns "Error querying HMDB: 0"
Cause: HMDB search returned empty results or index error accessing first result Solution: This is expected for uncommon metabolites; PubChem fallback will be attempted
Issue: Study details show "N/A" for all fields
Cause: Study ID not found or API unavailable Solution: Verify study ID format (MTBLS* or ST*), check if study is public
Issue: Tool not found errors
Cause: Missing API keys for some databases
Solution: Check .env.template, add required API keys to .env file (most metabolomics tools work without keys)
Issue: Large metabolite lists cause slow execution
Cause: Pipeline queries each metabolite individually Solution: Reports limit to first 10 metabolites; consider batching for >20 metabolites
Tool Parameter Reference
HMDB Tools (SOAP)
| Tool | Required Parameters | Optional Parameters | Response Format | Notes |
|---|---|---|---|---|
HMDB_search | operation="search", query | - | {status, data: []} | SOAP tool - operation required |
HMDB_get_metabolite | operation="get_metabolite", hmdb_id | - | {status, data: {}} | SOAP tool - operation required |
MetaboLights Tools (REST)
| Tool | Required Parameters | Optional Parameters | Response Format | Notes |
|---|---|---|---|---|
metabolights_list_studies | - | size (default: 10) | {status, data: []} or [...] | May return direct list |
metabolights_search_studies | query | - | {status, data: []} | Returns study IDs |
metabolights_get_study | study_id | - | {status, data: {}} | Full study metadata |
Metabolomics Workbench Tools (REST)
| Tool | Required Parameters | Optional Parameters | Response Format | Notes |
|---|---|---|---|---|
MetabolomicsWorkbench_get_study | study_id | output_item (default: "summary") | {status, data: {}} | Data may be text |
MetabolomicsWorkbench_search_compound_by_name | compound_name | - | {status, data: {}} | Compound information |
PubChem Tools (REST)
| Tool | Required Parameters | Optional Parameters | Response Format | Notes |
|---|---|---|---|---|
PubChem_get_CID_by_compound_name | compound_name | - | {status, data: {cid}} | Returns CID |
PubChem_get_compound_properties_by_CID | cid | - | {status, data: {}} | Chemical properties |
Important: All parameter names and requirements apply to both Python SDK and MCP implementations.
Summary
The Metabolomics Research skill provides comprehensive metabolomics analysis through a 4-phase pipeline that:
- Identifies metabolites using HMDB (primary) and PubChem (fallback) databases
- Retrieves study details from MetaboLights and Metabolomics Workbench repositories
- Searches studies by keywords across metabolomics databases
- Generates structured reports with all findings in readable markdown format
Key Features:
- ✅ 100% test coverage with working pipeline
- ✅ Handles SOAP tools correctly (HMDB requires
operationparameter) - ✅ Implements fallback strategies (HMDB → PubChem)
- ✅ Graceful error handling (continues if one phase fails)
- ✅ Progressive report writing (memory-efficient)
- ✅ Implementation-agnostic documentation (works with Python SDK and MCP)
Best for:
- Metabolite annotation and pathway analysis
- Study discovery and data retrieval
- Comprehensive metabolomics research reports
- Multi-database metabolomics queries
Limitations:
- HMDB may not have all metabolites (fallback to PubChem)
- Some studies require authentication or are not public
- Large metabolite lists (>10) auto-limited in reports
- API rate limits may affect large-scale queries
Quick Start
See QUICK_START.md for:
- Python SDK implementation with code examples
- MCP integration instructions
- Step-by-step tutorial for common workflows
- Advanced usage patterns
相关科研 Skill
中药材分子鉴定用ITS2等DNA条形码鉴定药材基原物种,处理多基原与降解样品局限。适用于药材真伪鉴别研究,不替代法定检验。
中药网络药理学用TCMSP等数据库构建成分-靶点-通路网络并做富集分析,批判性讨论方法学局限。适用于中药机制假设生成与论文方法复核。
中医术语标准与知识图谱建立中医药术语与WHO/ICD-11标准的映射,设计本体并用Neo4j构建知识图谱。适用于数据整合与科研检索,不作临床决策系统。
方剂配伍规律挖掘对方剂做药名规范化、关联规则与网络聚类分析,发现候选核心药组。适用于用药规律挖掘,结果仅为待验证假设,不作处方依据。
证候与症状数据分析对症状量表数据做术语规范、潜在类别与因子分析,探索统计学证候模式。适用于证候分型研究,结果不用于个体辨证诊断。
泵用混合物热力学计算Calculates thermodynamic properties and mixture phase behavior for pump applications using the thermo Python library’s DIPPR/NIST correlations. Use for VLE or mixture property calculations.
常见问题
代谢组学数据检索助手能做什么?
代谢组学公共研究检索与批量查库工具,可 fallback 查询数据集,并通过 prompt 限制输出格式。 | 适用:代谢物注释、公共数据集检索、代谢组学研究信息整理与规范报告。
如何安装代谢组学数据检索助手?
在科搭 Codar 左侧点击「插件」,在「组件」页签搜索「代谢组学数据检索助手」并安装;在 Claude Code 或 Codex 中可以发送本页的安装提示词,或下载 Skill 包后解压到 ~/.claude/skills 或 ~/.codex/skills。
代谢组学数据检索助手需要哪些权限?
该科研 Skill声明需要:本地文件。安装前会再次列出,由你确认后才会生效。
代谢组学数据检索助手经过官方审核吗?
是。SpeedSkills 收录的每个版本都会记录内容摘要与权限声明,当前版本已通过安全审核。