发表机构
Science and Technology Facilities Council; Diamond Light Source; University of Southampton（科学与技术设施委员会; 钻石光源; 南安普顿大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
该研究提出\textbf{\textit{OPAL}}框架，验证决策时间信息能否为自适应实验提供正当性，其在细胞绘画数据集上实现了低误激活率，是六种方法中唯一结合非零激活与风险控制的方案。
AI 中文摘要
自适应实验室会在实验过程中选择测量方式，但多数方法在允许自适应后才启动。我们提出实验室机会感知策略授权框架（\textbf{\textit{OPAL}}），用于决定是否启用自适应。\textbf{\textit{OPAL}}采用预先承诺的合约要求非平凡自适应、受控目标风险及成本后正执行价值。我们确立了一个不可能边界：在无约束条件结果偏移下，源域结果与未标记目标协变量无法一致支持非平凡授权，并推导了目标校准恢复方法。应用于含11265个化合物的未见过的细胞绘画分区时，冻结门选择了595个化合物，捕获384个正机会，在最不利完成下实现严格正执行价值，其5.18%的误激活上限低于7.5%的限制。在六种方法中，仅\textbf{\textit{OPAL}}结合了非零激活与该风险控制。锁定药物基因组学和有限实验研究将策略错位与不可认证区分开，确立授权为安全自适应科学的独立层。
英文摘要
Machine learning determines which follow-up measurements biological screens collect. In a six-rule Cell Painting battery, the highest-value rule would re-image 96.01% of the library and had a 97.14% false-activation upper bound, showing why predicted value alone cannot justify replacing a fixed plan. We developed OPAL, a held-out decision test that freezes a rule and judges unnecessary measurement, coverage and value after cost against archive-specific criteria fixed before final evaluation. A development-selected sparse Cell Painting rule had 18.2-fold lower added-well burden, but its false-discovery bound exceeded 35%, so the fixed plan remained. LINCS--LJP favored broad acquisition under point-estimate criteria set during development, not selective saving. CTRP required fallback because its frozen score missed measured opportunity. OPAL separates optimization from evidence sufficient to change an experiment.
