Orchestra：基于多源佐证的调控候选发现 MCP 智能体
原标题：Orchestra: Corroboration-Based Regulatory Candidate Discovery via Composed Bioinformatics MCP Agents
AI 摘要
该论文提出 Orchestra，将两个独立构建的生物信息学 MCP 服务器 RegNetAgents（推断调控网络拓扑）与 CASCADE（提供 LINCS 敲低、DepMap 必需性、超级增强子、DoRothEA 等独立实验证据）组合为基于 LangGraph 的多智能体工作流，并通过 MCP 协议暴露。核心主张是：要求网络拓扑证据与独立实验证据在候选调控因子上达成一致，比任一系统单独使用更可信。作者在 TCGA 肿瘤获得性调控因子层级及 BRCA/COAD、STAD 面板上测试，发现多源一致可预测 OncoKB 状态，且该效应在独立真实标签（COSMIC）和 MI 边权重之外仍显著。
正文节选
Orchestra: Corroboration-Based Regulatory Candidate Discovery via Composed Bioinformatics MCP Agents Abstract Orchestra composes two independently-built bioinformatics MCP servers – RegNetAgents (Bird, 2026c), which infers regulatory network topology, and CASCADE (Bird, 2026a), which provides independent experimental evidence (LINCS knockdown, DepMap essentiality, super-enhancer status, DoRothEA transcription-factor confidence) – into a single multi-agent LangGraph workflow (LangChain, Inc., 202