ADMET-EvO：一个面向异构任务持续研究的自进化科学智能体
ADMET-EvO: a self-evolving scientific agent for sustained research across heterogeneous tasks
另 1 家 · 查看机构详情
- Valhalla Technology（华亚科技）
- The Chinese University of Hong Kong（香港中文大学）
- Zhejiang University（浙江大学）
- Stanford University（斯坦福大学）
机构由 AI 辅助整理，请以论文原文为准。
浏览论文内容
中文总结 AI 辅助
ADMET-EvO是一种自进化科学智能体，通过证据门控机制在异构ADMET任务中持续修正策略，在TDC基准上取得最高得分96.77，并显著减少拟合时间。
中文摘要 AI 辅助
科学智能体可以超越自动建模，通过利用累积的证据来修正其研究问题和实验策略。挑战在于如何在异构任务中维持这种适应性，同时避免对内部验证的过拟合决策。吸收、分布、代谢、排泄和毒性（ADMET）预测在多种测定、数据集和化学领域中提供了一个要求严苛的环境。因此，我们开发了ADMET-EvO，一个证据门控的智能体，它形式化端点、生成可证伪的假设，并在数据、特征和模型轴上进行干预测试。它携带支持、拒绝和不确定的结果，以指导每个新周期。在22任务治疗数据共享（TDC）ADMET基准测试中，ADMET-EvO取得了最高的任务归一化得分96.77。证据引导的选择在预定义的非劣效性范围内将累积拟合时间减少了72.2%。它还形式化了43个毒性相关任务，并构建了端点特异性预测器。总之，这些结果表明ADMET-EvO如何能够随时间累积证据、修订策略并扩展其研究范围。
英文摘要
Scientific agents can move beyond automated model building by using accumulated evidence to revise both their questions and experimental strategies. The challenge is sustaining this adaptation across heterogeneous tasks without overfitting decisions to internal validation. Absorption, distribution, metabolism, excretion and toxicity (ADMET) prediction provides a demanding setting across diverse assays, datasets and chemical domains. We therefore developed ADMET-EvO, an evidence-gated agent that formalizes endpoints, generates falsifiable hypotheses and tests interventions across data, feature and model axes. It carries supported, rejected and inconclusive outcomes forward to guide each new cycle. Across the 22-task Therapeutics Data Commons (TDC) ADMET benchmark, ADMET-EvO achieved the highest task-normalized score of 96.77. Evidence-guided selection reduced cumulative fitting time by 72.2% within a predefined non-inferiority margin. It also formalized 43 toxicity-related tasks and constructed endpoint-specific predictors. Together, these results show how ADMET-EvO can accumulate evidence, revise its strategy and expand its research scope over time.
