MARBLE: Multi-Agent Reasoning for Bioinformatics Learning and Evolution

2026-01-20

AI Summary

This work proposes an autonomous and stable framework for bioinformatics model optimization, addressing the inefficiency and poor reproducibility of conventional trial-and-error approaches. By integrating role-specialized multi-agent systems engaged in structured debate, literature-aware evidence selection, and performance-driven memory updating, the framework enables continuous, iterative model evolution. Evaluated on spatial transcriptomics segmentation, drug–target interaction prediction, and drug response prediction tasks, the method significantly outperforms strong baselines, demonstrating high robustness and an exceptionally low rate of performance regression. These results establish a novel paradigm for automated bioinformatics modeling that enhances both reliability and scalability in computational biology applications.

Technology Category

• Search and Optimization: Evolutionary Computation
• Multiagent Systems: Modeling other Agents
• Machine Learning: Bio-inspired Learning

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

Motivation: Developing high-performing bioinformatics models typically requires repeated cycles of hypothesis formulation, architectural redesign, and empirical validation, making progress slow, labor-intensive, and difficult to reproduce. Although recent LLM-based assistants can automate isolated steps, they lack performance-grounded reasoning and stability-aware mechanisms required for reliable, iterative model improvement in bioinformatics workflows. Results: We introduce MARBLE, an execution-stable autonomous model refinement framework for bioinformatics models. MARBLE couples literature-aware reference selection with structured, debate-driven architectural reasoning among role-specialized agents, followed by autonomous execution, evaluation, and memory updates explicitly grounded in empirical performance. Across spatial transcriptomics domain segmentation, drug-target interaction prediction, and drug response prediction, MARBLE consistently achieves sustained performance improvements over strong baselines across multiple refinement cycles, while maintaining high execution robustness and low regression rates. Framework-level analyses demonstrate that structured debate, balanced evidence selection, and performance-grounded memory are critical for stable, repeatable model evolution, rather than single-run or brittle gains. Availability: Source code, data and Supplementary Information are available at https://github.com/PRISM-DGU/MARBLE.

Problem

• bioinformatics
• model refinement
• iterative improvement
• reproducibility
• LLM-based assistants

Innovation

• multi-agent reasoning
• autonomous model refinement
• performance-grounded memory
• structured debate
• bioinformatics learning