# Nature meets machine: the AI renaissance in natural product drug discovery

## TL;DR

This paper explores how AI can enhance natural product drug discovery by addressing traditional challenges and enabling more efficient exploration of nature-derived compounds.

## Contribution

The paper introduces a framework for integrating AI across natural product discovery pipelines, highlighting novel applications like machine learning-driven dereplication and genome mining.

## Key findings

- AI tools like GNPS2 demonstrate scalable progress in natural product discovery workflows.
- AI integration accelerates early-stage discovery and improves translational relevance in antibiotic and anticancer research.
- Emerging AI paradigms, such as quantum machine learning, may further expand natural product research.

## Abstract

Natural products (NPs) have long served as a cornerstone of drug discovery, yielding landmark therapeutics such as paclitaxel and artemisinin and providing sustained access to biologically relevant chemical space. Despite this legacy, NP-based discovery has gradually declined with the rise of synthetic chemistry and high-throughput screening, even as many contemporary “synthetic” drugs remain structurally inspired by natural scaffolds. Classical NP workflows—centered on phenotypic screening and bioassay-guided fractionation—continue to face persistent bottlenecks, including structural complexity, low bioactive yield, frequent rediscovery, and limited scalability. Rather than competing with NP research, artificial intelligence (AI) offers a complementary methodological framework to address these longstanding challenges. This review critically examines the bottlenecks inherent to traditional NP discovery and outlines how AI can be systematically integrated across the pipeline. We discuss AI-enabled advances ranging from natural language processing for mining ethnopharmacological knowledge to machine learning–driven dereplication, cheminformatics, and genome mining, with platforms such as GNPS2 exemplifying scalable progress. Case studies in antibiotic and anticancer discovery, as well as the modernization of traditional medicine, illustrate how AI–NP integration can accelerate early-stage discovery while enhancing translational relevance. Looking ahead, we examine emerging paradigms—including quantum machine learning, federated data ecosystems, and AI-assisted molecular design—that may further expand the scope of NP-based research. Collectively, this review presents a forward-looking framework in which AI functions not as a replacement for NP science, but as a synergistic discipline that enables more efficient, scalable, and informed exploration of nature-derived chemical diversity.

The online version contains supplementary material available at 10.1007/s13659-025-00589-6.