# Natural-Language-Guided Generator-Agnostic Shortlisting for Protein Binder Design

arXiv: https://arxiv.org/abs/2608.20755 · HTML: https://mlsanity.com/paper/2608.20755/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 2 of its 29 references that are in the map.
- First posted on arXiv 21 Aug 2026.

## Builds on (most cited first, 2 of 2)

- [XGBoost: A Scalable Tree Boosting System](https://mlsanity.com/paper/1603.02754/) (2016)
- [ProtAgents: Protein discovery via large language model multi-agent collaborations combining physics and machine learning](https://mlsanity.com/paper/2402.04268/) (2024)

## Abstract (arXiv)

Modern de novo design workflows generate many candidate protein binders, but wet-lab validation capacity remains limited, making shortlisting a major bottleneck. We study whether LLMs can generate multi-metric ranking policies from precomputed structural-confidence and interface-quality proxy scores. Rather than proposing a new protein binder design pipeline, we focus on post-generation binder shortlisting: selecting the final top-K candidates from already generated binder pools using a shared panel of precomputed proxy scores. On the 10-target held-out split, averaging performance over five sampled global iterative gpt-4o policies reaches 0.589 Recall@10, modestly improving over the strongest single-feature fixed baseline, Protenix binder ipTM, which reaches 0.571 Recall@10. On the 3-target held-out subset comprising Nipah, RBX1, and TREM2, target-conditioned iterative gpt-5.4 policies reach the strongest LLM performance, with 0.519 Recall@10 and 0.583 NDCG@10. These results suggest that LLM-generated ranking policies can act as an interpretable post-generation decision layer for combining heterogeneous proxy metrics to prioritize binders from large candidate pools.