Mapping of specialized metabolite terms onto a plant phylogeny using text mining and large language models

• Published in: The Plant journal : for cell and molecular biology Vol. 120; no. 1; pp. 406 - 419
• Main Authors: Busta, Lucas, Hall, Drew, Johnson, Braidon, Schaut, Madelyn, Hanson, Caroline M., Gupta, Anika, Gundrum, Megan, Wang, Yuer, A. Maeda, Hiroshi
• Format: Journal Article
• Language: English
• Published: England Blackwell Publishing Ltd 01-10-2024
• Subjects: artificial intelligence; Betalains - metabolism; biochemical evolution; Caryophyllales - genetics; Caryophyllales - metabolism; Chemical compounds; Data Mining; Dopamine; Gene mapping; Gene sequencing; Genome, Plant - genetics; Language; Large language models; Medicinal plants; Metabolites; Natural products; Nutrients; phylochemistry; Phylogeny; Pigments; Plant layout; plant natural products; Plants - classification; Plants - genetics; Plants - metabolism; technical advance; Transcriptomes; Tyrosine; Workflow; phylochemistry; plant natural products; tyrosine; biochemical evolution; technical advance; artificial intelligence
• ISSN: 0960-7412; 1365-313X; 1365-313X
• DOI: 10.1111/tpj.16906

SUMMARY
Plants produce a staggering array of chemicals that are the basis for organismal function and important human nutrients and medicines. However, it is poorly defined how these compounds evolved and are distributed across the plant kingdom, hindering a systematic view and understanding of plant chemical diversity. Recent advances in plant genome/transcriptome sequencing have provided a well‐defined molecular phylogeny of plants, on which the presence of diverse natural products can be mapped to systematically determine their phylogenetic distribution. Here, we built a proof‐of‐concept workflow where previously reported diverse tyrosine‐derived plant natural products were mapped onto the plant tree of life. Plant chemical‐species associations were mined from literature, filtered, evaluated through manual inspection of over 2500 scientific articles, and mapped onto the plant phylogeny. The resulting “phylochemical” map confirmed several highly lineage‐specific compound class distributions, such as betalain pigments and Amaryllidaceae alkaloids. The map also highlighted several lineages enriched in dopamine‐derived compounds, including the orders Caryophyllales, Liliales, and Fabales. Additionally, the application of large language models, using our manually curated data as a ground truth set, showed that post‐mining processing can largely be automated with a low false‐positive rate, critical for generating a reliable phylochemical map. Although a high false‐negative rate remains a challenge, our study demonstrates that combining text mining with language model‐based processing can generate broader phylochemical maps, which will serve as a valuable community resource to uncover key evolutionary events that underlie plant chemical diversity and enable system‐level views of nature's millions of years of chemical experimentation.
Significance Statement
This study introduces a novel workflow to map the distribution of natural products onto a phylogeny. It also demonstrates the effectiveness of combining text mining of the literature with large language model evaluation for constructing a “phylochemical” database that can organize vast portions of the chemical diversity present across the plant tree of life.