bioRxiv Science⌕ Search
Home/Archive/bioRxiv

bioRxiv · 10.1101/2025.09.03.673959

Predicting Antifouling Paint Particle Contamination based on 16S rRNA Gene Sequencing Data using Random Forest-Based Machine Learning

Tagg, A. S.·Sperlea, T.·Labrenz, M.·Schenk, A.·Kreikemeyer, B.
View original source ↗Export metadata ↓

Abstract

Antifouling paints often contain biocides designed to inhibit biological growth, and antifouling paint particles (APPs) have been previously shown to affect microbial communities in sediment. Given typical methods for monitoring for APP presence can be specialised and challenging, alternative methods using simple, standardised and universal approaches, such as 16S rRNA amplicon sequencing, would be highly valuable. This study uses a field-based mesocosm approach to train a random forest-based (supervised) machine learning model to predict APPs presence and concentration in sediment based on 16S microbial community data. The model correctly predicted 100% APP-presence samples and 83.3% APP-absence samples in the incubation testset, although the model could not correctly predict APP concentration with sufficient accuracy. To determine real-world applicability of the model, samples from 14 different sites along the Baltic Sea coastline and Warnow estuary in NE Germany were collected and APP-presence was pre-determined using SEM-EDX spectroscopy. The model correctly assigned APP-absence status to all APP-absent sites, and correctly assigned 3 of 5 APP-contaminated sites as having APP presence. As such, these results confirm it is possible to predict APPs in sediment based on the microbial community, and serves as a proof-of-concept for the further development of machine learning-based predictive tools for environmental monitoring.

Source connections

Explore related subjects

microbiologyPreprint

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Tagg, A. S., Sperlea, T., Labrenz, M., Schenk, A., Kreikemeyer, B.. 2025-09-03. Predicting Antifouling Paint Particle Contamination based on 16S rRNA Gene Sequencing Data using Random Forest-Based Machine Learning. https://doi.org/10.1101/2025.09.03.673959

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related preprints

bioRxiv2026

Matrix-controlled emergence of biofilm architecture shapes antimicrobial survival

Biofilms are structured microbial communities whose extracellular matrix is widely regarded as a basis of their protection against antimicrobial compounds. Yet how matrix production by individual bacteria gives rise to collective architecture and antimicrobial protection remains poorly understood. Here, we systematically varied expression of the master biofilm regulator csgD in Salmonella enterica and found that increasing matrix production reorganizes biofilms from dense, isotropic packings into sparse, nematically aligned communities by altering cell-cell interactions. By combining experimentally measured biofilm architectures with reaction-diffusion modeling, we show that these structural changes produce distinct patterns of antimicrobial killing, ranging from preferential killing near the liquid-biofilm interface to more uniform killing throughout the community. Consequently, increasing matrix production unexpectedly reduces antimicrobial survival by shifting the biofilm into different transport regimes, while strain-specific physiological differences further modulate antimicrobial depletion. Rather than acting as a passive barrier, EPS therefore shapes antimicrobial susceptibility by reorganizing biofilm architecture and its transport properties. EPS thus provides a physical link between molecular regulation, collective architecture and antimicrobial survival, providing a quantitative framework for understanding how cellular matrix production generates emergent biofilm function.

microbiology↗
bioRxiv2026

Mapping virulence-associated protein interaction networks reveals regulators of thermotolerance in Cryptococcus neoformans

Protein-protein interactions (PPIs) influence critical biological processes in pathogenic microorganisms, such as the human fungal pathogen, Cryptococcus neoformans. Fungal thermotolerance and stress response pathways are key virulence determinants that directly impact pathogen adaptation and survival and the infection process. To establish a comprehensive baseline of PPIs in C. neoformans and explore these interactions to infer functional roles for uncharacterized proteins, we applied size exclusion chromatography coupled with mass spectrometry to the secreted and cellular proteomes of the fungi. As a result, 216 and 1699 unique proteins were identified across 24 secretome and proteome fractions, respectively. The predicted secretome networks included expected proteins associated with vesicles and virulence, indicating a role in extracellular defense. Whereas the cryptococcal proteome highlighted interactions among proteins with defined roles in fungal virulence for protein stability and thermotolerance, including two previously uncharacterized proteins, CNAG_00287 and CNAG_05199, putatively involved in complex formation with heat-shock proteins (HSP). Based on sequence and structure homology, we propose that CNAG_00287 is a tetratricopeptide repeat-containing co-chaperone that modulates Hsp 70 activity and CNAG_05199 functions as a Hsp70. We validated the thermotolerance role of CNAG_00287 in heat-related stress, as its absence significantly impaired fungal growth in nutrient-limited media at 37 {degrees}C. Together, this work resolves virulence-associated PPIs within C. neoformans and reveals new molecular regulators of thermotolerance that underpin fungal pathogenicity.

microbiology↗
bioRxiv2026

Environmental filtering and host identity collectively shape root-associated microbiomes of Ericaceae and ectomycorrhizal plants in fumarole fields

Background Symbiosis with microbes is a key strategy that has enabled plants to colonize extreme environments. Since the benefits conferred by root-associated microbes depend on both environmental conditions and host-microbe combinations, plant adaptation to harsh environments is closely linked to the assembly of root microbial communities. Understanding how environmental and host filtering jointly shape these communities is therefore fundamental to elucidating the mechanisms underlying plant adaptation to extreme environments. Results In this study, we investigated the differentiation of root-associated prokaryotic and fungal communities and individual operational taxonomic units (OTUs) across two contrasting habitats surrounding fumaroles, solfatara-field and forest-edge habitats, and six dominant Ericaceae and ectomycorrhizal plant taxa. Prokaryotic and fungal OTUs rarely exhibited strong preferences for both habitat and host identity. Instead, many of prokaryotic and fungal OTUs specialized to one of these niches, collectively generating root microbial communities differentiated by both factors. Nonetheless, striking specializations in habitat and host niches were observed in the fungal family Hyaloscyphaceae (Helotiales). To gain insight into the evolutionary basis of microbial specialization, we examined phylogenetic signals in preference phenotypes. The resulting weak phylogenetic signals in these preference phenotypes further suggest that this fungal clade has undergone substantial ecological divergence. Conclusion Overall, our findings indicate that root-associated microbial communities in extreme environments are assembled through the accumulation of microbial taxa specialized to either habitat or host, and that strong ecological specialization in fungi can arise with little phylogenetic constraint.

microbiology↗