发表机构
Massachusetts General Hospital; Harvard University; Northeastern University（麻省总医院; 哈佛大学; 东北大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
该研究针对生物医学因果分析中构建因果DAG多为手动过程的问题，提出DAGForge系统。它能基于研究概念描述创作可审计的因果DAG，经评估在召回率等方面表现良好，减轻了DAG策划负担，支持生物医学研究相关工作。
AI 中文摘要
构建因果有向无环图（DAG）是生物医学因果分析的核心步骤，但很大程度上仍是手动过程。分析师需连接研究变量与先前文献、评估不确定因果关系并保留足够出处以供专家审查。我们提出了DAGForge，一个基于浏览器的系统，用于创作可审计、证据关联的因果DAG工件。给定研究概念的自由文本描述，DAGForge创建可重现的文献快照，使用基于大语言模型的推理模块生成基于逐字证据摘录的结构化成对因果判断，并将这些判断组装成经过约束检查的图。每个提议的边都包括置信度估计、出处和可审查的理由。该界面支持研究规范、进度监控、证据审查、图比较、调整集计算和导出。在针对紧凑基准DAG和从已发表文献中导出的参考DAG的评估中，DAGForge在基于文献的队列上实现了高边召回率，同时保留了仅基于大语言模型的基线中不存在的可验证证据线索。因此，DAGForge减轻了因果DAG策划的负担，同时使所得假设可审计，支持生物医学研究的设计、分析和解释。
英文摘要
Constructing causal directed acyclic graphs (DAGs) is a core step in biomedical causal analysis, yet it remains a largely manual process. Analysts must connect study variables to prior literature, evaluate uncertain causal claims, and preserve sufficient provenance for expert review. We present EviDAG, a browser-based system for authoring causal DAGs as auditable, evidence-linked artifacts from biomedical literature. Given free-text descriptions of study concepts, EviDAG creates a reproducible literature snapshot, uses an LLM-based reasoning module to generate structured pairwise causal judgments, links literature-supported judgments to verbatim evidence excerpts, and assembles the judgments into a constraint-checked graph. Each proposed edge includes confidence estimates, provenance, and a reviewable rationale. The interface supports study specification, progress monitoring, evidence review, graph comparison, adjustment-set computation, and export. In evaluations against both compact benchmark DAGs and reference DAGs derived from published literature, EviDAG achieves high edge recall on the literature-based cohort while retaining verifiable evidence trails absent from LLM-only baselines. EviDAG thus reduces the burden of causal DAG curation while making the resulting assumptions auditable, supporting the design, analysis, and interpretation of biomedical studies.
Comments10 pages, 2 figures
