Search papers, labs, and topics across Lattice.
This study benchmarks various oracle budget allocation methods for protein structure prediction, including FK-steering, DPO, and the newly applied Optimisation Over Outputs (O3). The findings reveal that while O3 excels at low oracle budgets, FK-steering and DPO outperform it at higher budgets, indicating that no single method is universally superior. This work serves as the first systematic comparison and provides practical guidance for researchers facing budget constraints in protein structure prediction tasks.
O3 outperforms traditional methods at low oracle budgets, but FK-steering and DPO take the lead as budgets increase鈥攕howing that context matters in model guidance.
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.
