Search, Edit, and Fold: LLM-Guided MSA Optimization for Protein Conformation Prediction

Accurately identifying alternative protein conformations remains a fundamental challenge, particularly when functionally relevant states are encoded by sparse evolutionary signals within large multiple sequence alignments (MSAs). In this work, we reformulate protein conformation prediction as a combinatorial search problem in MSA space, shifting the focus from structure divergence to evolutionary information discovery. We introduce MSA-Evolver, a optimization framework that enabling LLM with direct manipulation and iterative exploration of MSAs. Specifically, MSA-Evolver introduces a unified action space for MSA editing together with a feedback-guided multi-step reasoning strategy that allows the language model to progressively explore, evaluate, and refine candidate MSAs based on historical search trajectories. Our framework efficiently identifies informative sub-MSAs under limited folding budgets and substantially improves the prediction accuracy of alternative conformations, including open-closed, inward-outward, apo-holo, fold-switching, and intrinsically disordered proteins.