如何使用你的专家预算：蛋白质结构预测模型的实用指南
How to Spend Your Oracle Budget: Practical Guidance for Protein Structure Prediction Models
浏览论文内容
中文总结 AI 辅助
本研究针对蛋白质结构预测模型的专家预算约束，通过基准测试FK-steering、DPO、Best K-of-N采样及O3方法，明确不同预算下的最优方法并给出实用选择建议。
中文摘要 AI 辅助
蛋白质结构预测的基础模型在某些靶点上仍不可靠。外部专家可标记并纠正这些缺陷，但生物专家成本高昂，使得专家预算成为关键约束。现有指导方法（如FK-steering、DPO和Best K-of-N采样）在预算使用方式上存在差异，但尚无系统比较来指导方法选择。为填补这一空白，我们将这些方法与近期提出的Optimisation Over Outputs（O3，在生成模型的潜在子空间内应用现成优化器）一同进行基准测试，并将O3的使用扩展至蛋白质结构预测模型。总体而言，本研究为感知专家预算的指导提供了首个实用参考。我们对两个蛋白质靶点（钙调蛋白1CLL和大肠杆菌天冬氨酸转氨甲酰酶9EEH）的评估显示，没有任何一种方法在所有预算和专家条件下均持续占优。具体而言，O3在低专家预算下最有效，而FK-steering和DPO的性能随预算增加而提升。我们将这些发现提炼为在现实世界专家预算约束下操作的从业者可执行建议。
英文摘要
Foundation models for protein structure prediction remain unreliable on certain targets. External oracles can flag and correct these failures, but biological oracles are expensive, making oracle budget a critical constraint. Existing guidance methods, such as FK-steering, DPO, and Best K-of-N sampling, differ in how they spend this budget, yet no systematic comparison exists to guide method selection. To bridge this gap, we benchmark these methods alongside the recently proposed Optimisation Over Outputs (O3), which applies off-the-shelf optimisers within a generative model's latent subspace. We extend the usage of O3 to protein structure prediction models. Overall, our work provides the first practical reference for oracle budget-aware guidance. Our evaluation on two protein targets, calmodulin (1CLL) and E. coli aspartate transcarbamoylase (9EEH), reveals that no single method consistently dominates across all budgets and oracles. Specifically, O3 proves most effective at low oracle budgets, while FK-steering and DPO demonstrate improved performance as the budget increases. We distil these findings into actionable recommendations for practitioners operating under real-world oracle-budget constraints.
发表机构
- InstaDeep Ltd（InstaDeep公司）
- University of Oxford（牛津大学）
- Lancaster University（兰卡斯特大学）
机构由 AI 辅助整理，请以论文原文为准。
