Protein optimization adapts using feedback to improve high-fitness variants
Feedback-Calibrated Protein Optimization with Batch-Aligned Tail Arbitration
Machine Learning
Summary
Finding proteins that work really well can be hard and expensive because there are many possible versions to test. The authors developed a new way to pick which proteins to test next by combining different guesses and adjusting based on past results. This helps focus on the best candidates more accurately as experiments go on. Their method performed well on tests with real protein data, showing it can better find top-performing proteins.
What this means in practice
- •For biotech engineers: Select candidate protein variants more accurately during iterative lab testing to speed up development of high-performance proteins.
- •For drug discovery teams: Improve selection of optimized protein targets for therapeutic development by dynamically incorporating feedback from experimental results.
Authors
Zefeng Lin, Xianyong Fang, Tianfan Fu, Xiaohua Xu
Abstract
Protein optimization aims to discover high-fitness sequences under a limited experimental budget. Existing machine-learning methods use task-specific predictors, biological priors, or ranking-aware objectives to guide which variants are tested in the next experimental round. However, these methods cannot adapt to shifts in the reliability of predictive evidence as measurements accumulate and ensure the correct ranking of key high-fitness candidates. To address these challenges, we propose Batch-Aligned Tail Arbitration (BATA), which uses experimental feedback to adaptively combine prior-informed and task-specific rankings for next-batch selection, with calibration focused on the batch-aligned high-fitness region. Across measured GB1, PABP, and TrpB landscapes, BATA achieves the best mean task rank (1.67) in final best fitness after 480 measurements. Controlled comparisons further show task-dependent gains from high-fitness calibration and batch alignment. Our work introduces feedback-calibrated predictor arbitration, where experimental feedback dynamically determines how predictive evidence guides next-batch selection, opening a new direction for protein optimization.