AReUReDi: Annealed Rectified Updates for Refining Discrete Flows with Multi-Objective Guidance

2025-09-30

AI Summary

Addressing the challenge of designing biomolecular sequences under conflicting objectives—such as affinity, solubility, hemolytic activity, half-life, and antifouling capability—this paper introduces the first discrete flow model optimization framework with theoretical guarantees of Pareto front convergence. Methodologically, we embed the Tchebycheff scalarization into a modified discrete flow (ReDi) architecture, augmented with a locally balanced proposal distribution and an annealed Metropolis–Hastings update mechanism, enabling efficient multi-objective-guided discrete sequence optimization. Compared to evolutionary algorithms and diffusion-based baselines, our approach achieves superior Pareto-optimal solution quality and search efficiency. Empirical evaluation on peptide and SMILES sequence generation tasks demonstrates enhanced capability in balancing multiple biochemical properties and improved practical utility.

Technology Category

• Search and Optimization: Mixed Discrete/Continuous Search
• Constraint Satisfaction and Optimization: Mixed Discrete/Continuous Optimization
• Machine Learning: Optimization

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

Designing sequences that satisfy multiple, often conflicting, objectives is a central challenge in therapeutic and biomolecular engineering. Existing generative frameworks largely operate in continuous spaces with single-objective guidance, while discrete approaches lack guarantees for multi-objective Pareto optimality. We introduce AReUReDi (Annealed Rectified Updates for Refining Discrete Flows), a discrete optimization algorithm with theoretical guarantees of convergence to the Pareto front. Building on Rectified Discrete Flows (ReDi), AReUReDi combines Tchebycheff scalarization, locally balanced proposals, and annealed Metropolis-Hastings updates to bias sampling toward Pareto-optimal states while preserving distributional invariance. Applied to peptide and SMILES sequence design, AReUReDi simultaneously optimizes up to five therapeutic properties (including affinity, solubility, hemolysis, half-life, and non-fouling) and outperforms both evolutionary and diffusion-based baselines. These results establish AReUReDi as a powerful, sequence-based framework for multi-property biomolecule generation.

Problem

• Optimizing sequences for conflicting therapeutic objectives
• Guaranteeing Pareto optimality in discrete biomolecule design
• Simultaneously improving multiple drug properties like affinity

Innovation

• Discrete optimization algorithm for multi-objective Pareto front
• Combines Tchebycheff scalarization with annealed Metropolis-Hastings updates
• Biases sampling toward Pareto-optimal states preserving invariance