Skip to main content
Scientific Data logoLink to Scientific Data
. 2026 Jun 29;13:1453. doi: 10.1038/s41597-026-07716-z

Recovery of 178 metagenome-assembled genomes from sediments in subterranean estuary

Jiapeng Wu 1,#, Baoshan Zhang 1,#, Yuexi Ma 1, Chunyi Kuang 1, Yiguo Hong 1,✉
PMCID: PMC13680465  PMID: 42374043

Abstract

Subterranean estuaries (STEs), the mixing zones between terrestrial groundwater and seawater, function as critical biogeochemical reactors that buffer anthropogenic pollutants from entering the open ocean. To date, microbial diversity and community structure within STEs remain poorly characterized. Here, we reconstructed 178 metagenome-assembled genomes (MAGs) exclusively from bacteria (no archaeal MAGs identified). All MAGs met medium-quality standards (>70% completeness, <10% contamination), including 59 near-complete (>90%), 47 with completeness over 80%, and 23 over 75% complete genomes. These MAGs spanned 17 bacterial phyla, with Pseudomonadota dominating (30.9%). Crucially, 157 MAGs (88%) are unclassified at the species level based on GTDB assessment, potentially representing novel taxa, including 1 candidate family, 28 candidate genera, and 128 candidate species. This study provides a genomic resource for studying the functional roles of these unclassified taxa in STEs.

Background & Summary

Subterranean estuaries (STEs), first defined by Moore1 (1999) as the transition zone within coastal aquifers where terrestrial freshwater and seawater mix, typically form in sandy, permeable sediments. Such sediments cover >70% of ice-free continental shelves2. Tidal forces generate steep geochemical gradients across STEs, creating dynamic biogeochemical hotspots. Within these critical interfaces, microbial communities mediate the fate of pollutants entering coastal waters. Consequently, comprehensive characterization of microbial diversity and spatial distribution patterns in STEs is essential for understanding coastal ecosystem resilience.

Compared to inland groundwater microbiology, STEs remain significantly understudied. Currently, fewer than 30 publications characterize microbial diversity and distribution in STEs3–6. Most prior studies relied on high-throughput sequencing of 16S rRNA gene or functional genes (e.g., nirK, nirS, amoA). However, amplicon-based approaches exhibit three critical limitations for studying STEs: (1) PCR primer inherently restricts detection scope, (2) database dependency impedes novel lineage discovery, and (3) functional inference remains indirect. Shotgun metagenomics overcomes these constraints by providing primer-unbiased genomic data, enabling assembly-based reconstruction of uncultured novel microbes. Despite this advantage, metagenomic applications in STEs remain scarce.

We reconstructed 178 metagenome-assembled genomes (MAGs) from sediment samples collected along the transect in STEs. All MAGs exhibited completeness >70% and contamination <10%, meeting the medium-quality standard defined by the MIMAG guidelines7. Among these, 59 MAGs (33%) were near-complete (completeness >90%), 47 MAGs (26%) had completeness > 80%, and 23 MAGs (13%) had completeness >75% (Table S1). Additionally, 153 MAGs (86%) showed contamination <5%, including 7 MAGs (4%) with no detectable contamination (Table S1). Overall, 121 MAGs (68%) had an N50 length exceeding 10,000 bp (Table S1), indicative of high assembly quality. MAGs were distributed across samples HM1_45, HM2_12, and HM2_67, with HM3_01 yielding the highest number (69 MAGs). Classification using the Genome Taxonomy Database (GTDB) assigned all 178 MAGs to Bacteria; no Archaea were identified (Fig. 1e). These MAGs represented 17 phyla (Fig. 2). Pseudomonadota was the most abundant phylum (30.9%), followed by Actinomycetota (16.3%), Bacteroidota (11.2%), and Acidobacteriota (10.7%) (Figs. 2 and 3a). Notably, only 21 MAGs (12%) could be assigned to named species in GTDB, suggesting that the majority (88%) represent potentially novel taxa (Table S1). We identified 1 potentially novel family (0.6%), 28 potentially novel genera (15.7%), and 128 potentially novel species (71.9%) (Fig. 3b). MAG genome sizes ranged from 1.47 to 6.86 Mbp (mean = 3.36 Mbp; Table S1 and Fig. 3c). Planctomycetota MAGs exhibited the largest average genome size (5.91 Mbp), while Desulfobacterota_E MAGs had the smallest (1.64 Mbp). Genomic GC content ranged from 31.22% to 72.61% (mean = 61.00%; Table S1 and Fig. 3d), with Actinomycetota showing the highest average GC content (68.61%) and Bacteroidota the lowest (40.57%). These findings provide a valuable genomic resource for investigating the functional potential of these novel taxa.

Fig. 1.

Fig. 1

Sample collection and data analysis. (a) Geographical location of Houmen Beach. (b) Electronic map of the sampling transect in Houmen Beach. (c) Picture of sampling transect with three sampling wells. Black dots represent sediment samples. Red triangles represent the selected sediment sample for metagenomic sequencing. (d) Picture of sampling Well 1. We collected the fine and blackish sandy sediment near the water table. (e) Workflow used for assembled genomes from metagenomes in this study.

Fig. 2.

Fig. 2

A phylogenetic tree of all species-level microbial MAGs constructed from 120 conserved marker genes.

Fig. 3.

Fig. 3

(a) Community composition at the phylum level. (b) Potential novel MAGs at different taxonomic levels. (c) The genomic size of MAGs at the phylum level. (d) The GC content of MAGs at the phylum level.

Methods

Site location and sampling

Sediment cores were collected from a subterranean estuary (115.15 °E, 22.81 °N) adjacent to the South China Sea near Houmen Town, Shanwei City, China (Fig. 1a, b). Along a 40 m transect, three representative groundwater monitoring wells (W1, W2, W3) were sampled on 21 August 2020 (Fig. 1c). Using a vibracore system (Rhino S1, USA), near-water-table sediments were collected from each well (designated HM1_45, HM2_12, HM3_01; Fig. 1d), with an additional deep sediment sample (1 m depth, HM2_67) retrieved from W2. All sediments were immediately sealed in sterile polyethylene bags, flash-frozen in liquid nitrogen, and stored at −80 °C during transport to the laboratory. Groundwater was co-sampled from corresponding depths using peristaltic pumps (GeoTech, USA) with dedicated groundwater collection units (GVPKIT, ASM, USA). In situ physicochemical parameters (pH, salinity, temperature) were measured using a handheld multiparameter meter (SL1000, Hach, USA). Dissolved oxygen (DO) concentrations were determined via flow-through cell analysis with a calibrated probe (HQ40d, Hach, USA). In addition, each layer of groundwater was filtered by 0.22 μm membrane and subsequent stored at −20 °C in 25 mL polyethylene bottles for the determinations of dissolved inorganic nitrogen (DIN), including NO3−, NO2− and NH4+8. Detailed sampling information and physicochemical parameters are provided in Table 1.

Table 1.

Sampling details and physicochemical parameters of the samples.

Sample Sample types Depth (cm) Salinity (psu) Dissolved oxygen (mg L−1) pH Nitrite (μmol L−1) Ammonium (μmol L−1) Nitrate (μmol L−1)
HM1_45 groundwater 50 5 0.35 7.65 16.33 18.01 2.55
HM2_12 groundwater 20 10 0.45 7.85 0.06 25.21 1.62
HM2_67 groundwater 100 24 0.4 7.7 0.04 17.56 5.33
HM3_01 groundwater 10 26 1.69 8.2 0.01 14.12 1.27

Metagenome assembly and binning

DNA extraction, metagenome sequencing and analysis were following the procedures described in Fig. 1e. In brief, DNA was extracted from 0.3 g sediment collected in this study using DNeasy PowerSoil Kits (Qiagen, USA). Total genomic DNA of sediment were sequenced with sequencing depth of 90 Gb per sample using the NovaSeq Illumina platform with 150 bp paired-end reads in the GENEWIZ Co. Ltd. A total of 2,400,000,000 raw reads were retrieved with an average of 600,000,000 for each sample.

Raw reads were trimmed to remove the adapters and low-quality reads by Trimmomatic9. The quality of reads was checked using FastQC10. The quality-controlled reads were then de novo assembled into contigs using METASPAdes11 with different k-mer lengths (21, 33, 55, 77, 99 and 127). Genomic binning was performed using an association of MetaBAT212, CONCOCT13 and MaxBin214 with the default settings. The resulting bins were consolidated and refined with DAS Tool15 to produce a non-redundant, high-quality set of MAGs. Each assembled bin was evaluated for genome quality with CheckM16 software to obtain information on genomic quality, such as contamination, completeness, strain heterogeneity. The GC content, N50, N90, genome size, and gene number of each assembled bin were quantified with in-house scripts. The relative abundances of these bins were estimated with Bowtie217 and CoverM (https://github.com/wwood/CoverM) software. Open reading frames (ORFs) in these genomes were predicted with Prodigal18. The high-quality MAGs were taxonomically classified using GTDB-tk19 with the GTDB RS226 taxonomy. A phylogenetic tree of newly recovered MAGs was constructed by 120 single-copy genes in IQ-tree2 with the default settings20. 1000 bootstrap replications were selected. All phylogenetic trees were visualized using iTOL21 (https://itol.embl.de).

Data Records

The raw shotgun metagenome reads generated in this study are publicly available on the NCBI Sequence Reads Archive (SRA) under the BioProject PRJNA1279602 and SRA accession SRP59362022. The 178 MAGs of bacteria described in this study have been deposited in NCBI under the BioProject PRJNA129138923, with BioSample IDs ranging from SAMN49962124 to SAMN49962301. Annotation files for these 178 MAGs are available via Figshare24: https://doi.org/10.6084/m9.figshare.32213223.

Technical Validation

All sampling tools and containers have been sterilized to avoid contamination. After sampling, sediments were immediately sealed in plastic bags, placed on −80 °C and transported back to the laboratory. The MAGs quality was assessed using CheckM to validate the contamination and the completeness of the bins.

Supplementary information

Supplementary Table S1 (40.2KB, xlsx)

Author contributions

J.W. and Y.H. designed the study and collected the samples. J.W. and B.Z. extracted DNA and performed assembly and binning. J.W., B.Z., Y.M. and C.K., performed bin selection, bin reassembly, phylogenetic analysis, abundance and community composition analysis, and MAGs annotation. J.W. and B.Z. drew the figures and drafted the manuscript. J.W., B.Z. and Y.H. revised and edited the manuscript. All authors reviewed the results and approved the manuscript.

Funding

This work was funded by the National Natural Science Foundation of China (No. 42476141, 42276130), Basic and Applied Basic Research Foundation of Guangdong Province (No. 2023B1515120029), Young Scientific and Technological Talents Cultivation Program of the Guangdong Association for Science and Technology (No. SKXRC2026616), Science and Technology Projects in Guangzhou (No. 2025A03J3103), Open Fund of Hainan Xisha Marine Environment National Observation and Research Station (No. 2025001), and Graduate Innovative Research Grant Program of Guangzhou Education Bureau (No. 2024312281).

Data availability

The raw shotgun metagenome reads have been deposited in the NCBI database under BioProject accession number PRJNA1279602 and SRA accession SRP593620, with BioSample accession numbers SAMN49525810 to SAMN49525813. The 178 constructed MAGs are available under BioProject accession number PRJNA1291389 and BioSample accession numbers SAMN49962124 to SAMN49962301. Annotation files for all 178 MAGs have been deposited to Figshare (https://doi.org/10.6084/m9.figshare.32213223). All datasets are publicly accessible.

Code availability

All versions of third-party software and scripts used in this study are described and referenced accordingly in the Methods section. The source code used for MAG generation is available at https://github.com/ZhangBaoshan668/binning/MAG_script.sh.

Competing interests

The authors declare no competing interests.

Footnotes

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

These authors contributed equally: Jiapeng Wu, Baoshan Zhang.

Supplementary information

The online version contains supplementary material available at https://doi.org/10.1038/s41597-026-07716-z.

References

  • 1.Moore, W. S. The subterranean estuary: a reaction zone of ground water and sea water. Marine Chemistry65, 111–125 (1999). [Google Scholar]
  • 2.Luijendijk, A. et al. The State of the World’s Beaches. Sci Rep8, 6641 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Archana, A., Francis, C. A. & Boehm, A. B. The Beach Aquifer Microbiome: Research Gaps and Data Needs. Frontiers in Environmental Science9 (2021).
  • 4.Ruiz-Gonzalez, C., Rodellas, V. & Garcia-Orellana, J. The microbial dimension of submarine groundwater discharge: current challenges and future directions. FEMS Microbiol Rev45 (2021). [DOI] [PMC free article] [PubMed]
  • 5.Euler, S. et al. Microbiome mediating methane and nitrogen transformations in a subterranean estuary. Environ Microbiol26, e16558 (2024). [DOI] [PubMed] [Google Scholar]
  • 6.Wilson, S. J., Song, B. & Anderson, I. C. Geochemical factors impacting nitrifying communities in sandy sediments. Environmental Microbiology25, 3180–3191 (2023). [DOI] [PubMed] [Google Scholar]
  • 7.Bowers, R. M. et al. Minimum information about a single amplified genome (MISAG) and a metagenome-assembled genome (MIMAG) of bacteria and archaea. Nat Biotechnol35, 725–731 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Wu, J. et al. A rapid and high-throughput microplate spectrophotometric method for field measurement of nitrate in seawater and freshwater. Scientific Reports6, 20165 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Bolger, A. M., Lohse, M. & Usadel, B. Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinformatics30, 2114–2120 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Brown, J., Pirrung, M. & McCue, L. A. FQC Dashboard: integrates FastQC results into a web-based, interactive, and extensible FASTQ quality control tool. Bioinformatics33, 3137–3139 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Nurk, S., Meleshko, D., Korobeynikov, A. & Pevzner, P. A. metaSPAdes: a new versatile metagenomic assembler. Genome Res27, 824–834 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Kang, D. D. et al. MetaBAT 2: an adaptive binning algorithm for robust and efficient genome reconstruction from metagenome assemblies. PeerJ7, e7359 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Alneberg, J. et al. Binning metagenomic contigs by coverage and composition. Nat Methods11, 1144–1146 (2014). [DOI] [PubMed] [Google Scholar]
  • 14.Wu, Y. W., Simmons, B. A. & Singer, S. W. MaxBin 2.0: an automated binning algorithm to recover genomes from multiple metagenomic datasets. Bioinformatics32, 605–607 (2016). [DOI] [PubMed] [Google Scholar]
  • 15.Sieber, C. M. K. et al. Recovery of genomes from metagenomes via a dereplication, aggregation and scoring strategy. Nat Microbiol3, 836–843 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Parks, D. H., Imelfort, M., Skennerton, C. T., Hugenholtz, P. & Tyson, G. W. CheckM: assessing the quality of microbial genomes recovered from isolates, single cells, and metagenomes. Genome Res25, 1043–1055 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Langmead, B. & Salzberg, S. L. Fast gapped-read alignment with Bowtie 2. Nat Methods9, 357–359 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Hyatt, D., Chen, G., Locascio, P. F., Land, M. L., Larimer, F. W. & Hauser, L. J. Prodigal: prokaryotic gene recognition and translation initiation site identification. BMC Bioinformatics11, 119 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Chaumeil, P. A., Mussig, A. J., Hugenholtz, P. & Parks, D. H. GTDB-Tk: a toolkit to classify genomes with the Genome Taxonomy Database. Bioinformatics36, 1925–1927 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Nguyen, L. T., Schmidt, H. A., von Haeseler, A. & Minh, B. Q. IQ-TREE: a fast and effective stochastic algorithm for estimating maximum-likelihood phylogenies. Mol Biol Evol32, 268–274 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Letunic, I. & Bork, P. Interactive Tree Of Life (iTOL) v5: an online tool for phylogenetic tree display and annotation. Nucleic Acids Res49, W293–W296 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.NCBI Sequence Read Archivehttps://identifiers.org/ncbi/insdc.sra:SRP593620 (2025).
  • 23.NCBI Sequence Read Archivehttps://identifiers.org/ncbi/bioproject:PRJNA1291389 (2025).
  • 24.Baoshan, Z. Recovery of 178 metagenome-assembled genomes from sediments in subterranean estuary. figshare. Dataset 10.6084/m9.figshare.32213223.v1 (2026). [DOI] [PMC free article] [PubMed]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Citations

  1. NCBI Sequence Read Archivehttps://identifiers.org/ncbi/insdc.sra:SRP593620 (2025).
  2. NCBI Sequence Read Archivehttps://identifiers.org/ncbi/bioproject:PRJNA1291389 (2025).
  3. Baoshan, Z. Recovery of 178 metagenome-assembled genomes from sediments in subterranean estuary. figshare. Dataset 10.6084/m9.figshare.32213223.v1 (2026). [DOI] [PMC free article] [PubMed]

Supplementary Materials

Supplementary Table S1 (40.2KB, xlsx)

Data Availability Statement

The raw shotgun metagenome reads have been deposited in the NCBI database under BioProject accession number PRJNA1279602 and SRA accession SRP593620, with BioSample accession numbers SAMN49525810 to SAMN49525813. The 178 constructed MAGs are available under BioProject accession number PRJNA1291389 and BioSample accession numbers SAMN49962124 to SAMN49962301. Annotation files for all 178 MAGs have been deposited to Figshare (https://doi.org/10.6084/m9.figshare.32213223). All datasets are publicly accessible.

All versions of third-party software and scripts used in this study are described and referenced accordingly in the Methods section. The source code used for MAG generation is available at https://github.com/ZhangBaoshan668/binning/MAG_script.sh.


Articles from Scientific Data are provided here courtesy of Nature Publishing Group

RESOURCES