pubmearch

# PubMed Analysis MCP Server > **Struggling medical students' own MCP server**: This is a newly developed project with features still being refined. Suggestions and improvements are welcome! > > **Note**: This is a newly developed project with features still being refined. Suggestions and improvements are welcome! A professional MCP server for analyzing PubMed medical literature to help researchers quickly gain insights into medical research dynamics. ## Features - **Literature Retrieval**: Supports PubMed advanced search syntax with date filtering. / 支持PubMed高级检索语法，可设置日期范围和结果数量。 - **Hotspot Analysis**: Analyzes keyword frequencies to identify popular research areas. / 统计关键词频率，识别热门研究方向，汇总相关文献。 - **Trend Tracking**: Tracks keyword changes over time to reveal evolving research trends. / 追踪关键词随时间的频率变化，揭示研究趋势演变。 - **Publication Count**: Analyzes publication volume changes with customizable time periods. / 提供灵活的时间周期设置，分析文献数量变化。 - **Comprehensive Reports**: Generates complete reports with customizable parameters. / 一键生成包含热点、趋势和统计的分析报告。 ## MCP Tools ### 1. search_pubmed Search PubMed and save results. / 搜索PubMed并保存结果。 Key parameters: - `email`: Your email (required) / 您的电子邮件（必填） - `advanced_search`: PubMed search query (required, same as advanced search syntax) / PubMed搜索查询（必填，与高级检索语法相同） - `max_results`: Maximum results (default: 1000) / 最大结果数（默认：1000） ### 2. list_result_files List available result files. / 列出可用的结果文件。 ### 3. analyze_research_hotspots Analyze research hotspots. / 分析研究热点。 Key parameters: - `filename`: Result filename (required) / 结果文件名（必填） - `top_n`: Number of keywords (default: 20) / 分析的关键词数量（默认：20） ### 4. analyze_research_trends Analyze research trends. / 分析研究趋势。 ### 5. analyze_publication_count Analyze publication counts. / 分析发文数量。 ### 6. generate_comprehensive_analysis Generate comprehensive analysis. / 生成全面分析报告。 ## Example for Cursor ```bash # Install dependencies pip install -r requirements.txt # Or use uv uv pip install -r requirements.txt ``` ### Write mcp.json Since I prefer using the uv virtual environment, I will directly run the Python file using the Python path here. ```json // Add the following configuration in mcp.json (for Windows) "PubMed": { "command": "cmd", "args": [ "/c", "path/to/python.exe", "path/to/server.py" ] } // For example, my mcp.json file looks like this { "mcpServers": { "fetch":{ "command": "cmd", "args": [ "/c", "uvx", "mcp-server-fetch" ] }, "PubMed": { "command": "cmd", "args": [ "/c", "path/to/python.exe", "path/to/server.py" ] } } } ``` ### LLM prompt (Agent mode) /your_mcp_name (note: it is PubMed here) Help me analyze the research hotspots on prostate cancer immunotherapy in the past three months. Set top_n to 50 and max_results to 5000. My email address is ... /your_mcp_name（注：比如我的mcp.json里面是PubMed）帮我分析一下近三个月前列腺癌免疫治疗的研究热点。top_n设置为50，max_results设置为5000。我的电子邮箱是... ## Notes - Follow NCBI usage policies. / 请遵循NCBI的API使用政策。 - Results saved in `results` directory, logs in `pubmed_server.log`. / 结果文件保存在`results`目录，日志位于`pubmed_server.log`。 - Project in development, contributions welcome via Issues or Pull Requests. / 项目处于开发阶段，欢迎通过Issue或Pull Request贡献改进。