Evolution-Aware MSA Reasoning for Subsampling via Factor Graphs

2026-07-24

AI Summary

This work addresses the challenge of multiple sequence alignment (MSA) subsampling under limited token budgets, where existing methods struggle to simultaneously preserve evolutionary signals and ensure controllability. The study formalizes MSA subsampling as a constrained optimization problem with explicit control objectives and introduces a novel framework based on factor graph inference. By modeling evolutionary awareness through unary factors and exemplar consistency factors, and integrating affinity propagation with message-passing algorithms, the approach explicitly balances query identity and diversity in the subsampled output. Evaluated on long-range contact prediction and conformational ensemble prediction tasks, the method significantly outperforms baseline approaches and effectively recovers alternative protein conformations.

Abstract

Multiple Sequence Alignments (MSAs) provide protein language models with explicit evolutionary context, but their large depth makes subsampling unavoidable under limited token budgets. Existing strategies, including random selection, identity-based filtering, and diversity-driven sampling, are effective heuristics, yet provide limited control over the evolutionary signals retained in the subset. In this work, we recast MSA subsampling as an explicit optimization problem, where key evolutionary measures, including query identity and diversity, are treated as controllable objectives. Building on this view, we introduce AP-REASONER, an Affinity-Propagation-based factor-graph approach. With evolution-aware unary factors, exemplar-consistency factors, and two control knobs, AP-REASONER performs factor-graph reasoning through message passing to infer a fixed-budget MSA subset. Experiments on long-range contact prediction and conformational ensemble prediction show that AP-REASONER outperforms baseline subsamplers on structure-sensitive downstream tasks and enables controllable recovery of alternative protein conformations. These results highlight the value of modeling MSA subsampling as a controllable optimization problem, where factor-graph reasoning offers an effective alternative to heuristic selection.

Problem

• MSA subsampling
• evolutionary signals
• token budget
• protein language models
• sequence alignment

Innovation

• MSA subsampling
• factor graph
• evolution-aware optimization
• affinity propagation
• controllable sampling