Purely Agentic Black-Box Optimization for Biological Design

2026-01-29

AI Summary

This work addresses the challenge of effectively integrating scientific literature knowledge into black-box optimization for biological design—such as small-molecule drug discovery, antimicrobial peptide development, and protein engineering—by proposing the first end-to-end language-agent-driven framework. The approach models the optimization process as hierarchical, language-based agent reasoning, leveraging a large scientific language model pretrained on chemistry and biology literature. It integrates retrieval-augmented generation, semantic task descriptions, and complex constraint handling to transcend traditional structure-centric paradigms. Evaluated on GuacaMol molecular generation and antimicrobial peptide optimization benchmarks, the method achieves state-of-the-art performance, substantially improving sample efficiency and target metrics. In vitro experiments further validate that the optimized peptides exhibit potent activity against drug-resistant pathogens.

Technology Category

• Search and Optimization: Sampling/Simulation-based Search
• Reasoning under Uncertainty: Stochastic Optimization
• Machine Learning: Optimization

Application Category

• Search and Retrieval-Augmented AI: Agentic search
• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

Many key challenges in biological design-such as small-molecule drug discovery, antimicrobial peptide development, and protein engineering-can be framed as black-box optimization over vast, complex structured spaces. Existing methods rely mainly on raw structural data and struggle to exploit the rich scientific literature. While large language models (LLMs) have been added to these pipelines, they have been confined to narrow roles within structure-centered optimizers. We instead cast biological black-box optimization as a fully agentic, language-based reasoning process. We introduce Purely Agentic BLack-box Optimization (PABLO), a hierarchical agentic system that uses scientific LLMs pretrained on chemistry and biology literature to generate and iteratively refine biological candidates. On both the standard GuacaMol molecular design and antimicrobial peptide optimization tasks, PABLO achieves state-of-the-art performance, substantially improving sample efficiency and final objective values over established baselines. Compared to prior optimization methods that incorporate LLMs, PABLO achieves competitive token usage per run despite relying on LLMs throughout the optimization loop. Beyond raw performance, the agentic formulation offers key advantages for realistic design: it naturally incorporates semantic task descriptions, retrieval-augmented domain knowledge, and complex constraints. In follow-up in vitro validation, PABLO-optimized peptides showed strong activity against drug-resistant pathogens, underscoring the practical potential of PABLO for therapeutic discovery.

Problem

• black-box optimization
• biological design
• large language models
• antimicrobial peptide
• molecular design

Innovation

• agentic optimization
• black-box optimization
• large language models
• biological design
• retrieval-augmented reasoning