Machine learning reveals biocontrol agents shaping disease outcome in natural Arabidopsis populations

Abstract

Plants recruit antagonistic microbes to defend against phytopathogens, offering a route to rational biocontrol beyond empirical screening. Here, using six generations of leaf-microbiome data from natural Arabidopsis populations infected by the oomycete Albugo laibachii , we show that microbial diversity is driven by infection, site, and host genotype, and that infected plants form modular networks with increased inter-kingdom antagonism. We train four machine-learning models to discriminate infected from uninfected plants by microbiota composition and identify microbes enriched in diseased (disease-associated) or healthy (health-associated) plants. Testing the most predictive bacteria, fungi, and cercozoa in planta , we find all confer varying protection against Albugo , with health-associated microbes outperforming disease-associated taxa. The best candidate, a Cystofilobasidium fungus, is validated in a synthetic community, where genomic and community assays indicate biocontrol acts mainly through microbe-microbe interactions rather than plant immune activation. This work shows that pairing microbiome data with machine learning identifies effective biocontrol agents.

Genomic and metabolomic characterization of Bacillus velezensis CMC-6 reveals biocontrol potential against potato Fusarium Wilt
Chao Zhang, Yuanzheng Zhao, Dong Wang +5
BMC Genomics13 Oct 2026Peer-reviewedCC BY-NC-ND 4.0Plant biologyCrossref
Leveraging mRNA cMH and machine learning for the deconvolution of mixed body fluid stains in East Asian populations.
Qinglin Liu, Xiaolian Wu, Fanzhang Lei +6
Briefings in bioinformatics1 Sep 2026Peer-reviewedCC BYFull textGenetics & genomicsEurope PMC
Machine‐Learning Models Trained on Genomic Summary Statistics Assess Extinction Risk and Recovery Potential of Bottlenecked Populations
Johanna C. Winder, Jack N. Marsters, Thomas Birley +5
Evolutionary Applications28 Oct 2026Peer-reviewedCC BYFull textGenetics & genomicsDOAJ
Machine learning-assisted Repli-Histo labeling reveals distinct transcription-dependent constraints on chromatin motion in living cells
Katsuhiko Minami, Kako Nakazato, Kai Shimizu +4
Cell Reports30 Oct 2026Peer-reviewedCC BY 4.0General biologyCrossref
Biotic response networks are an important organizer of the transcriptome in wild Arabidopsis thaliana populations
Ana Paula Leite Montalvão, Kevin D. Murray, Ilja Bezrukov +10
Proceedings of the National Academy of Sciences of the United States of America19 Oct 2026Peer-reviewedCC BY 4.0Full textPlant biologyPMC

Cite

BibTeX RIS

BibTeX

@article{mahmoudi2026machine,
  title = {{Machine learning reveals biocontrol agents shaping disease outcome in natural Arabidopsis populations}},
  author = {Maryam Mahmoudi and Yiheng Hu and Juliana Almario and Paolo Stincone and Lynn-Marie Tenzer and Vasvi Chaudhry and Lukas Braun and Samuel Quinzer and Kay Nieselt and Eric Kemen},
  journal = {Nature Communications},
  year = {2026},
  month = oct,
  doi = {10.1038/s41467-026-75789-w},
  url = {https://hal.science/hal-05828875v1},
  note = {CC BY 4.0},
}

RIS

TY  - JOUR
TI  - Machine learning reveals biocontrol agents shaping disease outcome in natural Arabidopsis populations
AU  - Maryam Mahmoudi
AU  - Yiheng Hu
AU  - Juliana Almario
AU  - Paolo Stincone
AU  - Lynn-Marie Tenzer
AU  - Vasvi Chaudhry
AU  - Lukas Braun
AU  - Samuel Quinzer
AU  - Kay Nieselt
AU  - Eric Kemen
PY  - 2026
DA  - 2026/10/21
JO  - Nature Communications
DO  - 10.1038/s41467-026-75789-w
UR  - https://hal.science/hal-05828875v1
AB  - Plants recruit antagonistic microbes to defend against phytopathogens, offering a route to rational biocontrol beyond empirical screening. Here, using six generations of leaf-microbiome data from natural Arabidopsis populations infected by the oomycete Albugo laibachii , we show that microbial diversity is driven by infection, site, and host genotype, and that infected plants form modular networks with increased inter-kingdom antagonism. We train four machine-learning models to discriminate infected from uninfected plants by microbiota composition and identify microbes enriched in diseased (disease-associated) or healthy (health-associated) plants. Testing the most predictive bacteria, fungi, and cercozoa in planta , we find all confer varying protection against Albugo , with health-associated microbes outperforming disease-associated taxa. The best candidate, a Cystofilobasidium fungus, is validated in a synthetic community, where genomic and community assays indicate biocontrol acts mainly through microbe-microbe interactions rather than plant immune activation. This work shows that pairing microbiome data with machine learning identifies effective biocontrol agents.
SN  - 2041-1723
SN  - 2041-1723
ER  -

Abstract

Related

Cite