Decoding emergent properties of microbial community functions through subcommunity observations and interpretable machine learning
The ISME Journal · 02.01.2025 · 19 (1)
DOI: 10.1093/ismejo/wraf236 ↗Kurz gefasst
Fachartikel in der Zeitschrift The ISME Journal, online erschienen am 23.10.2025. Verfasst von Hidehiro Ishizawa und 5 weiteren Autor:innen. Beteiligte Einrichtung: Department of Applied Chemistry, Graduate School of Engineering, University of Hyogo, Himeji, Hyogo 671-2280, Japan. Hauptthemen laut MEDLINE: Machine Learning, Microbiota, Microbial Consortia, Microbial Interactions. Die Arbeit wurde bisher 5-mal zitiert (Crossref) und nennt 49 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Kurita Water and Environment Foundation.
Abstract (Originalsprache)
The functions of microbial communities, including substrate conversion and pathogen suppression, arise not as a simple sum of individual species’ capabilities but through complex interspecies interactions. Understanding how such functions arise from individual species and their interactions remains a major challenge, limiting efforts to rationally understand microbial roles in both natural and engineered ecosystems. Because current holistic (meta-omics) and reductionist (isolation- or single-cell-based) approaches struggle to capture these emergent microbial community functions, this study explores an intermediate strategy: analyzing simple subcommunity combinations to enable a bottom-up understanding of community-level functions. To examine the validity of this approach, we used a nine-member synthetic microbial community capable of degrading the environmental pollutant aniline, and systematically generated a dataset of 256 subcommunity combinations and their associated functions. Analyses using random forest models revealed that the subcommunity combinations of just three to four species enabled the quantitative prediction of functions in larger communities (5–9-member; Pearson’s r = 0.78–0.80). Prediction performance remained robust even with limited subcommunity data, suggesting applicability to more diverse microbial communities where exhaustive subcommunity observation is infeasible. Moreover, interpreting models trained on these simple subcommunity combinations enabled the identification of key species and interspecies interactions that strongly influence the overall community function. These findings provide a methodological framework for mechanistically dissecting complex microbial community functions through subcommunity-based analysis.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Substanzen
Autor:innen und Institutionen
- Department of Applied Chemistry, Graduate School of Engineering, University of Hyogo, Himeji, Hyogo 671-2280, Japan.
- Department of Applied Chemistry and Chemical Engineering, School of Engineering, University of Hyogo, Himeji, Hyogo 671-2280, Japan.
ORCID-Profile: Hidehiro Ishizawa
Förderung
- Kurita Water and Environment Foundation
Zitieren
APA
Ishizawa, H., Noguchi, S., Kito, M., Nomura, Y., Kimura, K., & Takeo, M. (2025). Decoding emergent properties of microbial community functions through subcommunity observations and interpretable machine learning. The ISME Journal, 19(1). https://doi.org/10.1093/ismejo/wraf236
Vancouver
Ishizawa H, Noguchi S, Kito M, Nomura Y, Kimura K, Takeo M. Decoding emergent properties of microbial community functions through subcommunity observations and interpretable machine learning. The ISME Journal. 2025;19(1). doi:10.1093/ismejo/wraf236
BibTeX
@article{Ishizawa2025,
author = {Ishizawa, Hidehiro and Noguchi, Sunao and Kito, Miku and Nomura, Yui and Kimura, Kodai and Takeo, Masahiro},
title = {Decoding emergent properties of microbial community functions through subcommunity observations and interpretable machine learning},
journal = {The ISME Journal},
year = {2025},
volume = {19},
number = {1},
doi = {10.1093/ismejo/wraf236}
}
EUVIMED