DrugAgent: Automating AI-aided Drug Discovery Programming through LLM Multi-Agent Collaboration

2024-11-24

arXiv.org

AI Summary

To address the critical bottleneck in AI-driven pharmaceutical applications—where theoretical concepts frequently fail to translate into reliable, executable code—this paper introduces the first multi-agent framework synergistically driven by an LLM Planner and an LLM Instructor, enabling automated machine learning programming for drug discovery. The framework integrates pharmacology-informed prompt engineering, domain-adaptive task planning, dynamic knowledge injection, and a rigorous code validation mechanism. Evaluated on three core tasks—including drug–target interaction prediction—it significantly outperforms baselines such as ReAct: achieving an average ROC-AUC improvement of 4.92% and generating executable code at a rate of 92.7%. The implementation is publicly open-sourced, establishing a reproducible, scalable, and domain-aware programming paradigm for AI-powered drug discovery.

Technology Category

• Humans and AI: Human-Aware Planning and Behavior Prediction
• Machine Learning: Large Multimodal Models (LMMs)
• Multiagent Systems: Multiagent Planning

Application Category

• Search and Retrieval-Augmented AI: Search Tool Learning with LLM: Teaching LLMs to invoke search and make use of retrieved information
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems

Abstract

Recent progress in Large Language Models (LLMs) has drawn attention to their potential for accelerating drug discovery. However, a central problem remains: translating theoretical ideas into robust implementations in the highly specialized context of pharmaceutical research. This limitation prevents practitioners from making full use of the latest AI developments in drug discovery. To address this challenge, we introduce DrugAgent, a multi-agent framework that automates machine learning (ML) programming for drug discovery tasks. DrugAgent employs an LLM Planner that formulates high-level ideas and an LLM Instructor that identifies and integrates domain knowledge when implementing those ideas. We present case studies on three representative drug discovery tasks. Our results show that DrugAgent consistently outperforms leading baselines, including a relative improvement of 4.92% in ROC-AUC compared to ReAct for drug-target interaction (DTI). DrugAgent is publicly available at https://anonymous.4open.science/r/drugagent-5C42/.

Problem

• Automating AI-aided drug discovery programming
• Translating theoretical ideas into robust implementations
• Improving drug discovery task performance using LLM collaboration

Innovation

• Multi-agent framework automates drug discovery programming
• LLM Planner formulates high-level drug discovery ideas
• LLM Instructor integrates domain knowledge for implementation