Bios and Abstracts
June 8-10, 2026 | Oak Island Resort, NS
Plenary Speaker Biographies
Sean Gibbons, PhD Institute for Systems Biology, Seattle, Washington
Sean Gibbons earned his PhD in biophysics from the University of Chicago in 2015. He completed his postdoctoral work at MIT in 2018. Sean is now an associate professor at the Institute for Systems Biology, in Seattle. His lab studies the ecology and evolution of microbial communities. In particular, Sean is interested in how host-associated bacterial communities influence the health and wellness of the host organism. His group designs computational and wet-lab tools for studying these complex systems. Ultimately, the Gibbons Lab aims to develop strategies for engineering the ecology of the gut microbiome to improve human health.
Isabelle Laforest-Lapointe, PhD Université de Sherbrooke, Sherbrooke, Quebec
Isabelle Laforest-Lapointe is an Associate Professor in the Department of Biology at the Université de Sherbrooke and holds a Canada Research Chair in Applied Microbial Ecology. Her research sits at the intersection of ecology, microbiology, genomics, and bioinformatics, with a focus on host-microbe interactions in both terrestrial and human ecosystems. Her work spans the phyllosphere microbiomes of forest trees, including their role in plant diversity and ecosystem function, to early-life dynamics of the human gut microbiome and their influence on immune development. She completed her PhD at the Université du Québec à Montréal and a postdoctoral fellowship at the University of Calgary. Her lab at Université de Sherbrooke is committed to open science and quantitative approaches to microbial ecology.
Steven W. Kembel, PhD Département des sciences biologiques, Université du Québec à Montréal
Dr. Steven W. Kembel is a Professor in the Département des sciences biologiques at Université du Québec à Montréal (UQAM) and Director of the CERMO-FC Genomics Platform. Since joining UQAM in 2012, following postdoctoral training at the University of California, Berkeley and the University of Oregon, his research program has focused on using quantitative approaches to study microbial ecology and evolution. He is particularly interested in applying methods from community and phylogenetic ecology to understand host-associated microbiomes and their importance for host health, ecosystem function, and responses to global environmental change.
Leila J. Hamdan, PhD University of Southern Mississippi, Hattiesburg, Mississippi
Dr. Leila J. Hamdan serves as Associate Vice President for Research at the University of Southern Mississippi (USM) supporting USM’s coastal and ocean research and is Professor of Marine Microbial Ecology in USM’s School of Ocean Science and Engineering. She began her career as a Postdoctoral Research Associate at the Naval Research Laboratory (NRL) and spent 10 years as a Research Microbial Ecologist in NRL’s Marine Biogeochemistry Section. Her current research centers on ocean exploration, and the discovery of natural and artificial reefs on the seabed using sub-sea robotics systems. In 2017, she received the NOPP Excellence in Partnering Award for leadership of a multidisciplinary study of impacts of the Deepwater Horizon Spill. She is a Past-President of the Coastal and Estuarine Research Federation (CERF), represents USM as principal investigator for the NOAA Ocean Exploration Cooperative Institute and is principal investigator on the NSF award for the operation of the future Regional Class Research Vessel Gilbert R. Mason. Hamdan served as a member of the National Academies’ 2025-2035 Decadal Survey of Ocean Sciences committee. She holds a BS in Biology from Rowan University of New Jersey, and MS and PhD in Environmental Science and Policy from George Mason University.
Mason Stothart, PhD Nova Scotia Department of Natural Resources, Kentville, Nova Scotia
Dr. Stothart studies wildlife–microbiome symbioses in natural, anthropogenic, and zoo-based captive breeding environments. In seeking to understand how microbes shape the ecology and evolutionary trajectory of plant and animal life, he melds perspectives from the fields of comparative physiology, metacommunity ecology, and quantitative genetics. As a scientist with the Government of Nova Scotia Department of Natural Resources and an adjunct professor at universities in the Canadian Maritimes, Mason is exploring ways to apply modern molecular methods and a microbiological lens to responsible natural resource management.
John Parkinson, PhD SickKids, Toronto, Ontario
Dr. Parkinson is a Senior Scientist at the Hospital for Sick Children, Toronto and Professor in the Departments of Biochemistry and Molecular Genetics at the University of Toronto. Combining computational genomics with systems biology, research in the Parkinson lab is focused on understanding the role of the microbiome and particularly eukaryotic microbes, in maternal and child health. Working with clinicians at SickKids and the Aga Khan University in Pakistan, Dr. Parkinson and his team are dissecting the microbial interactions that support healthy pregnancies and early infant development. Complementing these activities are animal studies aimed at developing synthetic consortia of microbes to support gut health.
Nezar Al-Hebshi, PhD Temple University, Philadelphia, Pennsylvania
Nezar Al-Hebshi, PhD, is an Associate Professor and Co-director of the Oral Microbiome Research Laboratory at Temple University Kornberg School of Dentistry (USA), and an Adjunct Associate Professor at the Center for Microbiology and Immunology, Temple University Lewis Katz School of Medicine. His research focuses on defining microbiome perturbations in health and disease through advanced sequencing and mechanistic approaches; modeling and modulating the oral microbiome to develop clinically relevant indices of dysbiosis for improved prediction and diagnosis of oral diseases; dissecting the role of high-risk Fusobacterium nucleatum isolates in oral carcinogenesis; and exploring the anticancer and immunomodulatory properties of health-associated oral bacteria.
Short Talk Abstracts
Adaptation, urbanization, and the microbiome: shifts in symbiont communities of a rapidly expanding soapberry bug
Colby College, Waterville, Maine
Anthropogenic global change is affecting all organisms, but these effects are disparate; while many species decliner in decline, others surge. One under-explored factor in these evolutionary fortunes is the relative distribution of commensal microbes. The red-shouldered soapberry bug (Jadera haematoloma) offers a unique model to study this question. Prior to the 1950s, these insects exploited native host plants with limited seed production, maintaining a dispersal-fecundity polyphenism. The introduction of ornamental Koelreuteria trees allowed the rapid evolution of a derived, highly fecund ecotype that is now expanding across human-dominated urban environments. To determine if the microbiome facilitates this rapid urban adaptation, we analyzed 204 microbiome samples from across the soapberry bug’s range in South Florida. Our preliminary analysis suggests significant differences in microbiome composition between the native and urban ecotypes. Of particular interest is the relative abundance of Wolbachia, a symbiont known to manipulate insect host fecundity. This study also contextualizes host-plant differences by examining seasonal effects on microbiome composition and defining the core microbiota through comparison with the sibling genus, Boisea.
Diversity, abundance, and prevalence of infant gut bifidobacteria across diverse populations and feeding practices
Wellesley College & Harvard T.H. Chan School of Public Health, Massachusetts
The early-life gut microbiome plays a critical role in shaping lifelong health, dynamically developing in both conjunction and response to environmental exposures during infancy. Among these, human milk oligosaccharides (HMOs) are key dietary substrates that drive early microbial assembly by selectively enriching taxa specialized in their utilization, particularly Bifidobacterium spp. Although multiple species, including Bifidobacterium longum subsp. infantis and Bifidobacterium breve, share the capacity to metabolize HMOs, the ecological relationships governing their coexistence and dominance across populations remain poorly understood. Here, we leverage cutting-edge metagenomic profiling to perform a cross-cohort, species-level analysis of Bifidobacterium community structure in over 1,500 infants from geographically and culturally diverse settings across the Americas, Europe, and Africa. Focusing on dominant HMO-utilizing species, we systematically evaluate patterns of co-occurrence and exclusion across cohorts. Across all studied populations, we observe a consistent pattern of mutual exclusion between B. infantis and B. breve, indicating that these species rarely co-dominate within the same microbial community. These observations suggest contrasting ecological distributions, with B. infantis enriched in lower-diversity community states and B. breve and related taxa spanning a broader range of community contexts. In contrast, B. longum and B. bifidum exhibit more permissive co-occurrence patterns, suggesting a spectrum of niche exclusivity within HMO-utilizing Bifidobacterium taxa. The characterized relationships persist across geographic contexts despite substantial variability in feeding practices and overall Bifidobacterium composition, suggesting that these configurations represent a conserved feature of early-life gut ecology. Together, these findings provide a foundation for understanding how key microbial taxa organize within early-life communities and establish a comparative framework for future work investigating the determinants and consequences of species-level dominance in the developing gut microbiome.
Characterizing metagenomic dark matter reveals distinct sequence signatures across marine microbial communities
1Department of Biology, Western University, London, Canada 2Department of Microbiology and Immunology, Schulich School of Medicine and Dentistry, Western University, London, Canada 3Department of Pathology & Laboratory Medicine, Western University, London, Canada 4Department of Computer Science, Western University, London, Canada
A large fraction of metagenomic sequences is derived from unknown taxa or have no known function or homology. To better understand the potential biological significance of this genetic "dark matter", we investigated whether it has characteristic sequence signatures. Here, we define dark matter as sequence reads that do not map to known or predicted protein-coding sequences (CDS) or sequences encoding rRNAs, tRNAs or small open reading frames (smORF). To examine the composition and distribution of dark matter across an environmental gradient, we analyzed 139 metagenomic datasets (over 22 billion reads) from the Tara Oceans project sampled from different ocean regions and depths. Reads were assembled with MEGAHIT and then screened for CDS (MetaGeneMark), rRNA (SortMeRNA), tRNA (Aragorn), non-coding RNA (Infernal) or smORFs (orfipy), with remaining contigs assigned to "dark matter". We tested for statistical associations between environmental covariates and k-mer (k=6) profiles or GC content (%GC) among categories of reads. Consistent with previous reports, %GC increased significantly with sampling depth but did not vary by region. Partitioning sequences into different categories revealed distinct compositional differences; for instance, %GC was significantly lower in dark matter sequences. Principal coordinates analysis of k-mer profiles also showed significant separation between dark matter, CDS and smORF sequences (PERMANOVA, p < 0.001) to a greater extent than sampling depth or region. k-mer frequencies among samples were significantly more variable in dark matter, indicating that specific k-mers were enriched in some samples relative to the sequence composition in coding regions. These results indicate that metagenomic dark matter exhibits structured sequence variation that is distinct from coding regions, consistent across marine environments. This distinctiveness suggests that dark matter captures non-random patterns that may reflect underlying biological processes. Future work will assess whether these patterns are conserved across other environments, including soil metagenomes.
An Integrated Data Analysis Toolset for Investigating Microbial Community Dynamics from Environmental DNA
Dalhousie University, Halifax, Nova Scotia
Microbial communities are large, complex, and their unseen dynamics influence virtually every natural and engineered environment on Earth. The analysis of high-throughput DNA sequencing data to identify biochemical processes of these communities is being increasingly leveraged by interdisciplinary collaborators across research and industry. Next Generation Sequencing (NGS) Pelorus (https://cwrs.shinyapps.io/NGS_Pelorus/) is a new curated data analysis toolset that facilitates this emerging integrated research to bridge microbial community monitoring with subject matter expertise, observation, and experimentation in diverse microbiome environments. Pelorus is a web-based dashboard interface that is continually updating to provide a ‘one-stop’ platform of curated methods appropriate for simplifying and interpreting microbiome data. We present its core application for data-driven extraction of community-wide features in relation to critical environmental metadata. Pelorus allows users to identify key microorganisms that explain diversity variation and divergent community compositions, described according to their putative metabolic functions. The suite of tools includes recently developed differential abundance diagnostics and biomarker selection methods to report informative microbial predictor candidates. Technical validation is demonstrated by employing model datasets from three diverse real-world microbiomes (microbial corrosion, marine algal blooms, and drinking water source monitoring studies) to assess 1) applicability, 2) benefits of pre-processing data and workflow integration, and 3) reproducibility and quality control options. Use of Pelorus is shown to uncover novel, robust, and interpretable insights for decision-making and advance method consensus and more accessible, cost-effective (time-saving) analyses in broad biological settings.
Beyond Community Assembly: Multi-Omics Analysis of 24-Month Spontaneous Beer Fermentations Reveals Metabolite Heterogeneity Driven by Early Batch and Seasonal Variations
University of Alberta, Edmonton, Alberta
Spontaneously fermented beer is the product of complex microbial interactions that are highly sensitive to initial environmental conditions. Although general community assembly in spontaneous fermentations is well-understood, factors such as pH, seasonality and batch to batch variation have not been explored in-depth. To address these unknowns this study incorporated shotgun-metagenomic sequencing, volatile metabolite profiling via GC×GC-TOFMS, and organic acid profiling via HPLC on 22 barrels from multiple batches and seasons at seven time points in a two year fermentation. Community assembly from our genomic dataset was consistent with known patterns, however metabolomic data displayed high levels of heterogeneity that was not clearly explained by microbial profiles. Fermentation age was a large driver of difference in microbial profiles, i.e., early timepoints were closely associated with batch and seasonality but converged as the fermentations reached maturity. Metabolite profiles, however, remained seasonally distinct. Wort acidity also appears to be important for fermentation success, as barrels with a pH range of 4.7-5.4 had extended enterobacterial phases and did not establish a population of lactic acid bacteria. These data suggest that starting conditions and early microbial communities impact metabolite production in late stage fermentations, regardless of later stage homogeneity. In order to elucidate the seasonal and batch variations from our datasets we are developing metabolic models of all fermentation timepoints; this will enable us to map how different genetic pathways are active in the same taxa across different barrels, batches and seasons. Additionally, we are leveraging over 100 high quality MAGs to determine the impact of strain diversity on fermentation outcomes. All sampled barrels in this study are first-fill, and/or have not been used for spontaneous beer fermentation, which challenges the need for a back-slopping effect as a necessary inoculant. Altogether, these data greatly advance our knowledge of microbial ecology in spontaneously fermented foods.
Giant viral diversity and discovery within Nova Scotian wetlands
Dalhousie University, Halifax, Nova Scotia
Giant viruses (GVs) exhibit exceedingly unusual biology and unheard-of genome complexity. Since their discovery a few decades prior our understanding of this group has dramatically expanded, in many parts owing to large-scale metagenomic surveys that allow us to observe these viruses independent of their hosts. While global surveys have been effective at uncovering much of the known diversity of this group, probing unique environments typically underrepresented in metagenomic data remains an important part of characterizing the interesting evolution and diversity of GVs. Here, we target local Nova Scotian wetlands using a long-read metagenomic approach to observe the GVs circulating within these environments. GV metagenomically assembled genomes (GVMAGs) recovered from an acidic bog and a non-tidal marsh add new species-level representatives to the phylum Nucleocytoviricota and the more recently identified GV phylum Mirusviricota. GVMAGs identified within the bog metagenome are primarily assigned to the order Asfuvirales, while MAGs recovered from a marsh GV-enriched fraction are more evenly distributed across known GV orders. GVMAGs display typical GV genomic features including encoded proteins from diverse cellular and viral origins, and proteins predicted to support viral replication via host metabolism modulation (ex. nucleotide biosynthesis, amino acid biosynthesis). A more unique functional landscape is also observed in GVs encoding bacterial derived antimicrobial resistance genes. Relying on the observation that extensive gene exchange occurs between GVs and their hosts, we built a network of shared protein homology between GVMAGs generated here and publicly available eukaryotic genomes. This network can be used to define a putative host range to inform subsequent host-GV culturing efforts from the sampled environments. The GV diversity uncovered here adds to the ever-expanding diversity of GVs observed globally, representing important drivers of their eukaryotic hosts' ecology and evolution.
MaAsLin 3: refining and extending generalized multivariable linear models for meta-omic association discovery
Harvard T.H. Chan School of Public Health, Boston, Massachusetts
Identifying microbes and functions within microbial communities associated with environmental or health phenotypes is a central goal of microbiome research. However, the unusual properties of microbial sequencing data, particularly sparsity, compositionality, and high dimensionality, continue to challenge accurate, reproducible, and comprehensive differential abundance (DA) testing. Here, we introduce MaAsLin 3 (Microbiome Multivariable Associations with Linear Models), a generalized modeling framework that robustly identifies microbiome feature-level associations across both taxonomic and functional data derived from microbiome sequencing in complex experimental designs. To do this, MaAsLin 3 models prevalence and abundance separately using logistic regression and log-linear regression respectively, then combines their effects into a unified per-feature association result, improving modeling within highly sparse data. To address compositionality, MaAsLin 3 employs a median coefficient comparison for relative abundance data or natively incorporates absolute abundance protocols, including spike-in normalization and total genomic abundance quantification via qPCR. Benchmarking on synthetic datasets demonstrated that MaAsLin 3 improved precision by up to 0.27 over MaAsLin 2 while maintaining comparable recall. Moreover, coefficient estimates from relative abundance data closely followed those from absolute quantification using qPCR or spike-in sequencing, matching ANCOM-BC2’s performance while natively supporting experimental spike-in data. Additionally, MaAsLin 3 maintained a low false discovery rate across 38 real datasets under metadata randomization. When applied to the HMP2 Inflammatory Bowel Disease Multi-omics Database (IBDMDB), MaAsLin 3 corroborated previously reported associations and extended our ability to detect complex ecological associations. MaAsLin 3 is freely available at https://huttenhower.sph.harvard.edu/maaslin3/.
From classification to confirmation: verifying taxonomic classifications by mapping metagenomic reads to reference genomes
Dalhousie University, Halifax, Nova Scotia
Obtaining high precision while maintaining high recall is an ongoing problem for metagenomic taxonomic classification in microbial ecology research. Parameter adjustments can achieve this in simulated samples, but in real samples –especially from environments like marine and soil– the proportion of classified reads drops sharply with precision increases. We therefore suggest verification of metagenomic taxonomic classifications obtained from a tool like Kraken by mapping their assigned reads to reference genomes to assess genomic coverage. In simulations, filtering the identified species to only those with >0.5% reference genome coverage removed 99.7% of false-positive taxa. Applying this method to samples from real datasets requires a more nuanced approach that considers sequencing depth, whether the samples are high- or low-microbial biomass, and database completeness with respect to the sampled environment. Nevertheless, we show that clinically relevant Kraken-identified taxa such as Helicobacter pylori identified in human stool samples lack any reads mapping to their reference genome and are likely false positives driven by contaminating phage sequences within reference genomes. Similarly, in human blood and lung tumour datasets, only 18 and 11 species, respectively, have >1% reference genome coverage and likely represent sample collection or sequencing contaminants. Marine and soil samples pose additional challenges due to lower representation in reference databases, leading to low nucleotide identity between sequenced reads and reference genomes and similarity only at higher taxonomic ranks. We recommend genome coverage checking to researchers in all fields of microbial ecology and provide an open-source pipeline on Github (GeCoCheck): https://github.com/R-Wright-1/GeCoCheck.
Investigating Host Urinary Microbiome-Uropathogenic E. Coli Interactions within a Glycosaminoglycan-integrated Mini Bioreactor Bladder Model
University of Ottawa, Ottawa, Ontario
Urinary tract infections (UTIs) are among the most common bacterial infections worldwide, predominantly caused by uropathogenic Escherichia coli (UPEC). The presence of a diverse urinary microenvironment (urobiome) can play a role in infection dynamics. Integral to this microenvironment is a glycosaminoglycans (GAGs) lining the urinary tract’s urothelium. To date, it is unknown whether GAGs provide a substrate used to promote UPEC persistence and whether residential urobiome species contribute synergistically to this process. The study aims to better understand how the microenvironment influences UPEC persistence within the bladder by examining microbial dynamics at the urine-GAG interface to determine whether UPEC exploits GAG-derived metabolites and naturally occurring metabolic processes as an adaptation strategy. We developed a bladder mini-bioreactor, which incorporates a GAG-hydrogel mimetic. maintained for six hours under microaerophilic conditions using fresh female urine inoculated with 7 log10 CFU/mL of UTI89, a reference UPEC strain. Four bioreactor conditions were tested to capture different microbial and infection activity at the urine-GAG interface: UPEC with resident urobiome and GAG, UPEC with GAG only, residential urobiome with GAG, and UPEC with resident urobiome but no GAG. After five developmental iterations of the urine-GAG mini-bioreactor model, we have begun preliminary experiments. We are applying a multi-omics approach to characterize UPEC interactions with the GAG layer and urinary commensals, integrating 16S rRNA gene sequencing, targeted metabolomics of GAG-catabolic pathways, and proteomics. Used alongside co-occurrence network analyses, this approach will investigate how interindividual urobiome differences influence UPEC metabolic and pathogenic activity at the urine-GAG interface and how metabolites produced by commensal bacteria may serve as nutrient sources and support UPEC growth. Establishing this experimental framework aims to shed light on microbial or metabolic factors underlying infection susceptibility in female urinary health. This approach will also be applied to UTI patient urine samples in a future experimental round.
Temporary discharge of combined untreated wastewater-stormwater and it's impact on the virome of a marine harbour
Dalhousie University, Halifax, Nova Scotia
Abstract Withheld
Unveiling protist-bacteria interactions through computational metabolic modeling
1Hospital for Sick Children, Program in Molecular Medicine, 2University of Toronto, Molecular Genetics Department, 3University of Toronto, Biochemistry Department
Tritrichomonas musculus (Tmu) is a common protist in the mouse gut. As many gut protists, Tmu is a pathobiont, capable of commensalism and pathogenesis. Previous studies suggest that gut bacteria could mediate protist’s switch to pathogens, at least in part, through metabolic interactions. To further explore the role of such interactions in Tmu pathogenesis we performed computational simulations of the mice gut. We reconstructed genome scale metabolic models (GEMs) of Tmu and 266 mouse gut bacteria. We applied these GEMs to simulate the impact of 45 different diets on Tmu growth, and its metabolic interplay with gut bacteria, using the BacArena modeling platform. For the diets we considered regular mouse chow with 1 of 45 different saccharides as carbon source. Among the tested carbon sources, we found that glucose, maltose and maltose oligomers increased Tmu numbers, while xylan reduced them. The amount of secreted succinate, a key modulator of host immunity, depended on the diet as well, with glucan resulting in the highest succinate levels, and pectin in the lowest. Finally, we also identified cross-feeding interactions that depended on diet. Under maltose, Tmu feeds alanine to bacteria, and bacteria feed Tmu with phenylalanine and riboflavin. In contrast, when grown with xylan, bacteria feed Tmu with riboflavin but stop receiving alanine from it. Our simulations revealed an important role of diet on Tmu growth, succinate secretion and Tmu-bacteria interactions. These diets will be validated in vivo to assess if we can promote commensalism in Tmu, and eventually other gut protists.
One Extraction, Two Microbiome Dimensions: An Integrated Method for the Joint Study of the Agricultural Soil Virome and Microbiome
Centre SÈVE, Department of Biology, Université de Sherbrooke; Saint-Jean-sur-Richelieu R&D Centre, Agriculture and Agri-Food Canada; Réseau québécois de recherche en agriculture durable (RQRAD)
The health of agricultural soils relies on complex microbial communities. However, the virome, the collection of soil viruses, remains largely unexplored, despite being recognized as a potentially major driver of ecological functions. This knowledge gap is primarily due to methodological challenges that prevent the simultaneous analysis of viruses and their microbial hosts without introducing spatial bias. To address this challenge, we developed a unified approach combining a novel extraction method with bioinformatics tools. First, we developed the SS-VIME (Single-Source Virome-Microbiome Extraction) protocol. This method allows for the sequential extraction of DNA (microbiome) and double-stranded RNA (dsRNA) to specifically target the active virome from a single sample. This effectively eliminates the sampling biases inherent in current methods that require separate extractions. Second, to process the generated data, we created MetaViralyst, a reproducible bioinformatics ecosystem. It combines an automated Snakemake pipeline, integrating a unified taxonomic database (merging viral sequences from NCBI and IMG/VR) for exhaustive identification, with an interactive web application (MetaViralyst Explorer). The latter facilitates downstream analysis through intuitive modules for exploring ecological diversity, differential abundance, and graphical visualization, without requiring programming skills. The performance of SS-VIME was rigorously validated using sterilized soils spiked with a standardized microbial community (Zymo) and synthetic viral dsRNA. Sequencing confirmed the accurate recovery of the theoretical bacterial and fungal profiles in the DNA fraction, as well as highly specific capture of the synthetic virus in the dsRNA fraction. Finally, the application of this approach to agricultural soil samples successfully reconstructed complex indigenous soil communities. This integrated platform offers precise and robust tools for biodiversity research and the study of microbiome modulation by the virome from a single extraction.
When Microbiome Function Is Not in the Individual Microbes: The Role of Interaction Structure in Community Function
Dalhousie University, Halifax, Nova Scotia
Microbiome research is largely built on the premise that community function can be understood from composition: who is present and in what abundance. Yet systems can defy this expectation. Communities with similar compositions can exhibit markedly different functional outcomes, while other communities can maintain similar functional states despite substantial compositional turnover. Furthermore, compositional shifts often fail to predict changes in health, stability, or productivity. I argue that these discrepancies are not simply due to missing variables or measurement noise, but reflect a deeper limitation of additive, composition-based models. In multi-species microbial systems, function can be constructed through configuration-dependent interactions; patterns of cooperation, division of labour, and threshold dynamics that depend on how organisms are arranged and interact, not just on their frequencies. I introduce a conceptual framework for identifying such systems, based on a simple diagnostic question: can any additive summary of community composition account for the observed functional outcome? When the answer is no, community function depends irreducibly on interaction structure. Collective functions that depend on an irreducible interactive structure are called synaptations. Synaptive systems can be understood as a distinct class of collective phenomena, a “fitness commons”, whose integrity depends on the underlying interaction structure rather than from individual taxa or their aggregate abundances. I illustrate this perspective using examples from biofilms and host-associated microbiomes, and discuss implications for interpreting eubiosis, dysbiosis and for designing interventions. I suggests that microbiome science could benefit from formal recognition of this distinct form of collective function.
Flash Talk Abstracts
#1OAE in a bottle: Insights into diatoms’ biomass increased and bacterial shifts – a 19-day mesocosm study
Dalhousie University, Halifax, Nova Scotia
Over the past six decades, the persistent increase in atmospheric CO2 has stimulated interest in marine Carbon Dioxide Removal (mCDR) strategies as a means to mitigate climate change. Among these, Ocean Alkalinity Enhancement (OAE) is notable for its substantial global CO2 removal potential and concomitant mitigation of ocean acidification. However, the localized ecological consequences of increased pH, particularly on microbial communities, remain insufficiently studied. Given their pivotal role in the marine ecosystem, perturbations to these communities could have far-reaching effects. This study, undertaken as part of the OAE Pelagic Impact Intercomparison Project (OAEPIIP), examines the impacts of NaOH and NaHCO3-based OAE on phytoplankton and microbial communities in a 19-day mesocosm experiment using Halifax Harbour surface seawater. Two treatments were implemented, each targeting an alkalinity increase of 500 µL.kg−1: an equilibrated treatment simulating post air-sea gas equilibration conditions, and an unequilibrated treatment focusing on pre-equilibration impacts. Key seawater parameters, including alkalinity, pH, temperature, nutrients, particulate organic carbon/nitrogen (POC/PON), biogenic silica, and chlorophyll a, were monitored throughout the experiment. The biological response of microbial communities was evaluated using flow cytometry, Microscopy imaging (FlowCam), and metabarcoding-based microbiome analysis. Results revealed no significant changes in biomass or community composition in the equilibrated treatment. In contrast, the unequilibrated treatment induced modest yet notable effects on phytoplankton biomass (evidenced by increases in POC/PON, chlorophyll a, biogenic silica, and natural fluorescence cell counts) and bacterial community composition (as indicated by beta-diversity and differential abundance analyses). Based on flow cytometry and microscopy cell counts, microbiome composition, POC/PON, biogenic silica, and chlorophyll a measurements, diatoms appeared more resilient to increased alkalinity, dominating the second bloom observed in the mesocosms. This initial assessment of OAE's impacts on seawater communities establishes a foundation for future in-depth studies and contributes to the development of policies for OAE-based climate change mitigation strategies.
#2An In Vitro Model of Tumour Hypoxia and Macrophage-Microbiome Dynamics
Dalhousie University, Halifax, Nova Scotia
Pancreatic ductal adenocarcinoma (PDAC) is a highly lethal malignancy characterized by a hypoxic tumour microenvironment that shapes immune function and disease progression. Microbial communities are active regulators of tumour biology. In particular, microbial signals are emerging as key modulators of tumour-associated macrophages (TAMs), influencing their polarization, function and spatial distribution within hypoxic tumour niches. Hypoxia and microbial exposure are critical determinants of macrophage differentiation; however, their combined effects remain poorly defined due to the lack of physiologically relevant experimental models. To address this gap, we developed a three-dimensional in vitro PDAC model that recapitulates tumour hypoxia and enables controlled interrogation of macrophage-microbiome interactions. Hypoxia reporter PDAC cells were seeded on coverslips, and after 24 h, a hypoxia chip was placed on top, creating a 500 μm gap that restricts media exchange and generates hypoxia through cellular oxygen consumption. Microspheroids (100 cells) were generated using the hanging-drop method, embedded in a collagen hydrogel, and overlaid onto a fibroblast monolayer to recapitulate stromal architecture. After 4 h, the hypoxia chip was applied on top to establish hypoxic conditions. Macrophages were pre-exposed to dsRED-labelled E. coli for 2 h to mimic tumour-associated microbial exposure prior to incorporation into the model. Spheroid size, cell death, macrophage migration, and infiltration were assessed. We confirmed the formation of spatially restricted hypoxic regions in the monolayer and microspheroid models. Microbial exposure did not compromise macrophage viability; instead, bacteria-primed macrophages retained migratory capacity and actively infiltrated collagen matrices toward tumour spheroids. This model establishes a tractable platform to dissect how microbial cues and hypoxia jointly regulate macrophage behaviour in PDAC. By enabling controlled manipulation of the tumour microbiome, this system provides a foundation to uncover mechanisms of microbiome-driven immunomodulation and their contribution to tumour progression.
#3Metagenome Profiling Choices Influence Diversity Estimates and Conclusions in Microbiome Studies
Université de Sherbrooke, Sherbrooke, Quebec
A common task in microbiome studies is estimating the taxonomic composition of metagenomes. These estimates often serve as the basis for diversity comparisons between groups of samples. Several tools can be used for taxonomic profiling, many of which have been benchmarked in recent years primarily using synthetic data; yet, no single tool consistently outperforms others, in part because they differ fundamentally in multiple respects, making direct comparisons challenging. Therefore, researchers face numerous interdependent choices: (1) the profiling approach (marker-gene alignments vs. k-mer matching to genomes), which constrains database options and dictates the type of relative abundance being estimated; and (2) sensitivity parameters, which directly influence the sensitivity/specificity tradeoff in taxon detection. None of these choices are trivial and each could be driven by research objectives or technical constraints. To this day, how these choices influence diversity estimates of real metagenomes, and the conclusions drawn from their analysis, remains unexplored. Here, we leverage 1,221 real metagenomes from eight published datasets to compare 14 tool-parameter-database combinations on the basis of the diversity estimates they produce. Our results show that methodological choices can substantially influence diversity estimates and analyses. Alpha diversity estimates were more robust to tool choice among marker‑gene approaches (MetaPhlAn and mOTUs) than k‑mer–based tools (Kraken2 and Sourmash), even though marker‑gene tools rely on largely different reference databases. Beta diversity, on the other hand, was mostly affected by the choice of reference database, with non-RefSeq methods showing the strongest pairwise correlations. Most importantly, our data shows that standard approaches to test differences in diversities can yield inconsistent effect sizes and p-values. Overall, this work demonstrates that taxonomy‑based diversity analyses of shotgun metagenomes are strongly influenced by choice of bioinformatic methods. These results motivate a better understanding of tools and emphasize the need for transparency and sensitivity analyses in comparative microbiome studies.
#4Evolutionary Dynamics of Nitrogen Fixation in Marine Thalassolituus
Dalhousie University, Halifax, Nova Scotia
The expansion of metabolic traits across microbial lineages reflects the combined influence of evolutionary forces acting on their genomes and the ecological contexts in which they occur. Biological dinitrogen (N2) fixation, catalyzed by the nitrogenase enzyme complex encoded by the nifH, nifD, and nifK genes, represents a major source of bioavailable nitrogen in marine ecosystems. Non-cyanobacterial diazotrophs (NCDs), particularly marine heterotrophic bacteria, are widespread and functionally important contributors to marine nitrogen input across ocean habitats. Despite their ecological importance, the evolutionary origins and spread of N2 fixation across many of these lineages remains unresolved, including within Thalassolituus, a marine hydrocarbon-degrading genus from the order Oceanospirillales, that includes a recently characterized heterotrophic diazotroph. To investigate the distribution and evolutionary history of nitrogen fixation in this group, we performed comparative genomic analyses of 74 high-quality Thalassolituus genomes within a broader dataset of 421 Oceanospirillaceae genomes, and evaluated whether their nif genes were inherited vertically or showed signatures of lateral gene transfer (LGT). We identified a total of 25 putative diazotrophs harboring complete nif gene clusters, revealing a patchy distribution of N2 fixation within the genus. Phylogenomic analysis based on core genes showed that these putative diazotrophic lineages are not monophyletic, but instead dispersed across multiple well-supported clades interspersed with non-diazotrophic taxa. Gene trees further revealed significant phylogenetic discordance relative to the species tree. While the core nifHDK region was conserved, flanking regions varied and included lineage-specific accessory genes and hypothetical proteins. Conflicting gene trees and disruptions in synteny suggest that these loci have evolutionary histories distinct from the rest of their genome. Collectively, these findings support LGT as a major force shaping the distribution and diversity of nitrogen fixation in Thalassolituus and related marine heterotrophs. Ongoing analyses aim to refine these evolutionary inferences and evaluate additional mechanisms contributing to nif gene diversification.
#5Characterizing the impact of River Enhanced Alkalinity (RAE) on the water and sediment microbial communities in a Nova Scotian river
Dalhousie University, Halifax, Nova Scotia
To avoid the consequences of climate change, we urgently need to remove CO2 from the atmosphere via carbon dioxide removal (CDR) strategies. One CDR strategy is river alkalinity enhancement (RAE), where crushed limestone is added to rivers via a “lime doser”. The limestone quickly dissolves, reacting with CO2 to form bicarbonate that is transported to oceans for long-term storage. RAE has known environmental benefits, such as increase in salmon populations that have been decimated by chronic acidification (acid rain). An important knowledge gap is the impacts of RAE on microbial systems. In this study, we collected both benthic and pelagic samples of microbes upstream and downstream of a newly installed lime doser. Prior to the lime doser being active, pretreatment samples of the river sediment and water were collected for microbiome analysis. DNA was extracted directly from the sediment and pelagic samples as well as a culture enrichment step, followed by DNA extraction from the pelagic samples. Amplicon sequencing of the bacterial 16S rRNA was completed for all DNA samples to identify the microbial taxa present. One round of post-treatment data has been collected since the doser has been active, with additional rounds to come. A variety of water chemistry parameters have been monitored continuously, beginning before pretreatment data collection. Together the microbiome and water chemistry data provide a robust dataset to investigate the impact RAE has on the microbial communities of both water and sediment within the river.
#6Exploring the microbiome of broiler chickens for antibiotic-alternatives to enteric disease management
1Molecular Medicine, SickKids Research Institute, 2Department of Agricultural, Food and Nutritional Science, University of Alberta, 3Department of Molecular Genetics, University of Toronto, 4SickKids Research Institute
Antibiotic resistance is becoming a global health challenge; it is imperative we explore alternatives to antibiotic treatments. This problem is emphasized within the chicken farming community where enteric diseases that affect both chicken and human health are primarily treated with antibiotics. The goal of this project is to understand the microbial communities which support a healthy chicken gastrointestinal system, and bacteria which may combat enteric disease. By doing so we hope to develop therapies for microbiome alterations that can be used in place of antibiotic treatments. We have access to cecum meta-genomic data from chickens raised in both intensive and extensive farming environments. The chickens raised in intensive environments were given antibiotics for combatting enteric disease. The chickens raised in extensive environments were not given antibiotics. To extract taxonomic information from the 16srRNA data, we developed a computational pipeline that included data cleaning, host filtering, alignment, binning, quantification, and taxonomy classification. Our preliminary results showed a larger number of taxa in extensive chickens than intensive chickens. While the cecum of extensive chickens were predominately dominated by the Bacteroidota phylum, intensive chickens were dominated by both Bacteroidota and the Bacillota A phylum, specifically the Clostridia class. The extensive chickens showed larger taxonomic differences between individual chickens than extensive chickens. We plan to identify taxa that: 1) show the largest difference between intensive and extensive chickens and 2) that may enable competitive exclusion of disease-causing bacteria. In future work we plan to use metabolic modelling to predict how changes in microbial communities of taxa from 1) and 2) impact enteric disease.
#7Gene prediction accuracy remains unsatisfactory for divergent eukaryote
1Department of Biochemistry & Molecular Biology, Faculty of Medicine, Dalhousie University; 2Faculty of Computer Science, Dalhousie University; 3Laboratory of Marine Microbiology, Division of Applied Biosciences, Graduate School of Agriculture, Kyoto University
Gene prediction is the process of identifying the location of genes: regions of DNA sequence in a genome that encode cellular products. Any inaccuracies will propagate through downstream analyses, limiting experimental molecular investigations, and hindering the development of effective diagnostics and treatments in the case of pathogens. Predicting genes accurately and consistently across eukaryotic diversity is very challenging because many aspects of gene structure (gene length, codon usage, intron number and size, etc.) vary greatly among lineages. This challenge is even more pronounced for divergent eukaryotes who have experienced high rates of sequence evolution, considering that homology information which can normally aid prediction, is generally not meaningful in their case. Traditional gene prediction algorithms employ Hidden Markov Models (HMMs) – Bayesian approaches that enforce rigid, limited gene structures, pre-determined by probabilities estimated based on observed genomic data (referred to as data henceforth). While they work well for model organisms with an abundance of data, they do not possess the flexibility to capture the variability in gene structure often proportional to the level of divergence, especially with a gross lack of data. With a recent surge of Deep Learning (DL) based gene prediction algorithms, this challenge may finally be properly addressed. Instead of being solely reliant on data, DL models can potentially exploit the intrinsic signals of DNA for prediction, independent of data. While promising, there has not been a convincing gene prediction benchmark which can demonstrate this – until now. Here, using a self-developed custom evaluation software, I compared the accuracy of standard (HMM-based) and emergent (DL-based) gene prediction algorithms on highly curated genomes in several divergent eukaryotic lineages. Preliminary results show that gene prediction for divergent eukaryotic genomes remain unsatisfactory, and current DL-based algorithms offer no discernable advantage over HMM-based models.
#8Modeling fecal-periurethral contamination to evaluate engineered probiotic E. coli Nissle 1917 for urinary tract infections in women
1School of Pharmaceutical Science, Faculty of Medicine, University of Ottawa 2Schulich School of Engineering, Department of Biomedical Engineering, University of Calgary
Background: Urinary tract infections (UTIs) are highly prevalent and predominantly caused by uropathogenic Escherichia coli (UPEC). Recurrence is common and is driven in part by UPEC persistence within the gut microbiome, which serves as a reservoir for reinfection via periurethral contamination. While antibiotics remain the first-line treatment, rising antimicrobial resistance and microbiome disruption highlight the need for targeted therapeutic alternatives. Objectives: This study aims to develop an in vitro bipartite model for fecal-periurethral contamination, and to evaluate the efficacy of synthetic antimicrobial therapeutic (synAMTs) candidates. Developed by genetically engineering Escherichia coli Nissle 1917 (EcN), a probiotic strain, to selectively target and eliminate gut and urine residing UPEC. Methods: A bipartite mini-bioreactor model was developed, including anaerobic batch fecal fermentation and microaerophilic urine reactors, simulating gut and urinary microbial environments. Bioreactors were seeded with fresh fecal and urine samples collected from female participants. The gut compartment is inoculated with 6log10 CFU/mL UTI-89 (UPEC strain) and sampled over time. Prior to synAMT testing, wild-type EcN will also be inoculated at a 1:1 ratio and tested using three different modalities. Translocation of UTI89 to the urinary compartment is simulated through dialysis tube. Results: Preliminary results indicate a viable design, capable of sustaining the microbial communities from both niches, and periurethral contamination bridging the two. Initial experiments with wild-type EcN, establishes a baseline for comparison with synAMT candidates currently under development by collaborators at uCalgary. Analyses include propidium monoazide (PMA)-qPCR to quantify viable UPEC, RT-qPCR, and 16S rRNA gene sequencing to evaluate microbial and functional interactions across compartments. Perspectives: Further work will characterize UPEC–microbial interactions and evaluate two different SynAMT candidates in this model. This study will lay the groundwork for alternative UTI therapy, targeting the gut as a UPEC reservoir, informing microbial interactions, and addressing the need for targeted alternatives to broad-spectrum antibiotics.
#9Profiling microbial communities in marine harbours in Nova Scotia using automated environmental DNA sampling
Dalhousie University and Dartmouth Ocean Technologies, Inc.
Environmental DNA (eDNA) uses the genetic material shed by organisms in the environment to provide a detailed view of biodiversity. eDNA is being used in applications including environmental risk assessment, biodiversity monitoring, and detection of invasive species. However, manual eDNA collection using (for example) Nalgene or Niskin bottles can be limited by site access and availability of trained personnel. Automated eDNA samplers offer the promise of consistent sampling in remote areas over long deployments, and can be combined with other sensors to collect complementary environmental data. Marine harbours provide a promising use case for automated eDNA sampling, allowing the identification of invasive species and microbial communities that reflect environmental impacts and can pose hazards to human health. These instruments must be able to withstand environmental challenges including low temperatures, high turbidity, and biofouling. We tested the performance of the Dartmouth Ocean Technologies, Inc. (DOT) eDNA sampler, which uses preservatives to stabilize eDNA and environmental RNA, in two marine harbours on the east coast of Nova Scotia: Lunenburg Harbour, an important tourist destination that has been impacted by untreated sewage in the past, and Halifax Harbour, which sees heavy industrial and recreational use. Microbial communities in Lunenburg Harbour were highly stable over a day of sampling, with most microbial taxa observed in all or nearly all eDNA and eRNA samples. We found no evidence of Enterococcus, which was present at high levels prior to improvements in wastewater treatment. Taxonomic and functional profiles from Halifax Harbour collected after a heavy rainfall were also highly stable, with microbial taxa frequently associated with the human microbiome such as Enterobacteriaceae and Lachnospiraceae constituting up to 15% of total bacterial diversity. Our results demonstrate the utility of the DOT sampler in marine applications, and highlight the utility of microbial profiling in environmental assessment.
#10Functional and dietary characterization of metagenome-assembled genomes in paediatric Crohn’s disease following nutritional therapy
1Department of Microbiology and Immunology, Dalhousie University, 2Department of Pharmacology, Dalhousie University, 3Integrated Microbiome Resource (IMR), Dalhousie University, 4Paediatric Gastroenterology and Nutrition Unit, Pediatric Inflammatory Bowel Disease (PIBD) Research Centre, Edith Wolfson Medical Centre, Holon, Israel 5Tytgat Institute for Liver and Intestinal Research-Amsterdam Gastroenterology Endocrinology and Metabolism, University of Amsterdam, 6Division of Gastroenterology, Hepatology and Nutrition, the Hospital for Sick Children, Toronto, Ontario, Canada 7Department of Paediatric Gastroenterology, Emma Children's Hospital, Amsterdam University Medical Centre, 8Amsterdam Gastroenterology Endocrinology Metabolism (AGEM) Research Institute, Amsterdam University Medical Centre
Crohn’s disease (CD) exclusion diet with partial enteral nutrition (CDED+PEN) and exclusive enteral nutrition (EEN) induce remission in mild-to-moderate pediatric CD. In nutrition-responsive patients, prolonged dietary therapy shifts the gut microbiome and metabolome toward a healthy state; however, dysbiosis often persists despite sustained treatment, reflecting excess abundance of Proteobacteria and reduced Firmicutes. Alterations in microbial amino acid metabolism, have been associated with disease activity and sustained remission following dietary therapy. Higher-resolution characterization of microbial metabolic capacity is therefore needed to elucidate microbial functions relevant to intestinal inflammation and response to dietary therapy. We aim to use metagenome-assembled genomes (MAGs) to characterize genome-level differences in microbial amino acid metabolism pathways and asses their association with nutritional therapy-induced remission and dietary adherence in paediatric CD. PacBio HiFi long-read and Illumina short-read shotgun metagenomic sequencing were performed on 42 longitudinal stool samples collected at weeks 0, 6, and 12 from treatment-naïve mild-to-moderate pediatric CD patients receiving either CDED+PEN (n=11) or EEN (n=7) from a prior randomized controlled trial (NCT01728870). Clinical remission is defined as a Pediatric Crohn’s Disease Activity Index ≤10 at week 6. Illumina data was taxonomically profiled using Kraken2 and Bracken, while PacBio HiFi data was processed using Anvi’o-based methods. Future functional annotation will focus on microbial pathways involved in amino acid metabolism and genes associated with adaptation to intestinal inflammation. MAGs were obtained from long-read assemblies (MEGAHIT) using an Anvi’o based workflow with binning performed by MetaBAT2, MaxBin2, and CONCOCT, and dereplication of bins completed using DAS Tool. MAG quality was assessed using CheckM2 and taxonomic assignment was performed with GTDB-Tk. Anvi’o analyses yielded 0-13 high-quality MAGs per sample, with Bacteroides and Phocaeicola among the most recovered genera in high-depth samples. This MAG-resolved framework will enable detailed characterization of microbial metabolic capacity and its relationship to diet-induced sustained remission.
Poster Abstracts
#11Characterizing viral diversity in the neovaginal microbiome using shotgun metagenomics
1Department of Microbiology and Immunology, Schulich School of Medicine and Dentistry, Western University; 2Department of Biology, Western University; 3Center for Advanced Microbiome Research and Innovation, Institute for Genome Sciences, University of Maryland School of Medicine; 4Department of Microbiology and Immunology, Institute for Genome Sciences, University of Maryland School of Medicine; 5Department of Epidemiology and Biostatistics, Schulich School of Medicine and Dentistry, Western University
The viral component of the neovaginal microbiome in transfeminine individuals with penile inversion vaginoplasty remains largely uncharacterized, limiting our understanding of viral diversity in this unique host-associated niche. Here, we used shotgun metagenomic sequencing of 44 neovaginal samples to characterize viral sequences and their relationship with the bacterial community. Following host read removal, quality control, and de novo assembly, viral sequences were identified using VirSorter2 and evaluated with CheckV. Taxonomic annotation was performed against the NCBI Viral database (2024). In parallel, metagenome-assembled genomes (MAGs) were reconstructed and taxonomically classified to provide a genome-resolved view of the bacterial community. Eight contigs corresponding to Human Papillomavirus (types 15, 39, 42, 43, 84, and 89) were detected across four participants, indicating the presence of eukaryotic viral signatures within neovaginal samples. In addition, 72 viral contigs were assigned as putative prophage, predominantly affiliated with Siphoviridae (n=31) and Microviridae (n=29). A total of 620 medium- to high-quality MAGs (completeness >70%, contamination <10%) were reconstructed, primarily representing anaerobic genera commonly associated with genital microbiomes, including Peptoniphilus, Prevotella, Porphyromonas, Mobiluncus, and Anaerococcus. The distribution of viral contigs broadly paralleled the dominant bacterial taxa recovered as MAGs, suggesting potential coupling between viral and bacterial community composition. Overall, these results provide a first genome-resolved characterization of the neovaginal virome and establish a foundation for future analyses of virus–host interactions in this anatomically and ecologically distinct niche.
#12Multiomics approaches to characterize host-microbe interactions in pediatric IBD patients with comorbid anxiety
University of Toronto, Toronto, Ontario
Inflammatory bowel diseases (IBDs) are a complex, multifactorial set of diseases, including Crohn’s Disease and Ulcerative colitis. IBD incidence is increasing globally, especially among children which represent the fastest growing patient subpopulation. Research has indicated that the bi-directional influence of the brain and gut microbiome is important in IBDs. Notably, anxiety, whose co-occurrence with IBDs increases disease severity and treatment resistance, is significantly more common among IBD patients than in the general population. The exact host-microbe interactions in IBD patients with anxiety remain poorly understood, especially in pediatric patients. To address this, we have recruited a cohort of pediatric patients co-diagnosed with IBDs and anxiety and performed deep read metagenomics, meta-transcriptomics, and metabolomics on stool samples. We will interrogate these data to assess differences between our anxiety and control groups to develop a comprehensive picture of how anxiety interacts with the IBD gut. Finally, we will use flux balance analysis to build comprehensive metabolic models for each patient and perform in silico screening to identify nutritional supplements that may modulate the effects of anxiety on the IBD gut microbiome. This research will help advance our understanding of the interactions between anxiety and IBDs and help uncover possible avenues for precision IBD treatments
#13The Bioinformatics Atlantic Network (BAtl): Connecting Trainees and Professionals in Atlantic Canada
1Canadian Bioinformatics Hub, Dalhousie University; 2Faculty of Computer Science, Dalhousie University; 3Departments of Computer Science and Biology, Memorial University of Newfoundland; 4Faculty of Agriculture, Dalhousie University; 5Biology, University of Prince Edward Island; 6Department of Mathematics and Computing Science, Saint Mary’s University; 7Department of Biology, Acadia University;
For over two decades, Bioinformatics User Groups (BUGs) have organized Bioinformatics, Computational Biology, and Data Science (BCBDS) community events across Canada to encourage local interaction, collaboration, and exchange of ideas. These established BUGs have successfully operated out of metropolitan areas. We sought to build on the success of existing BUGs and address the unique distribution of institutions across Atlantic Canada while simultaneously increasing accessibility to BCBDS resources within the region. We established BAtl – the Bioinformatics Atlantic Network in 2025. BAtl is the first dedicated BUG in Atlantic Canada, spanning ten institutions and serving students, researchers, educators and professionals. We designed BAtl programming to scale and reinforce bioinformatics community activities and expand to new local communities. We have implemented a turn-based model wherein the host institution rotates throughout the region to engage as many local participants as possible. Our hybrid meetings increase accessibility and continually grow our regional audience. Over our first five events we have engaged over 200 participants. Our model exemplifies that a distributed, turn-based seminar model is a viable strategy for engaging multiple institutions across a wider geographic region. Through this, we are building a thriving BCBDS community in Atlantic Canada by connecting local academic networks.
#14Seasonally increasing parasite loads drive microbiota dysbiosis in wild bumblebees
Colby College, Waterville, Maine
Wild bumblebees rely on a highly specialized core gut microbiota for health and disease resistance, yet its interaction with prevalent parasites like Crithidia bombi remains heavily contested in natural environments. To resolve this, we conducted a three-year field survey of 638 bumblebees across nine sympatric species, pairing 16S rRNA amplicon sequencing with qPCR pathogen quantification. We identified robust, host-specific bacterial lineages, but found that as C. bombi infections surged seasonally and spatially, they drove a dose-dependent microbiome dysbiosis. Crucially, high parasite loads consistently depleted core, protective microbial taxa, Apibacter and Gilliamella, while facilitating a corresponding expansion of opportunistic, environmentally derived microbes like Entomomonas. These findings demonstrate that while the core microbiome may not prevent initial pathogen transmission, parasite-driven dysbiosis serves as a critical biomarker of systemic health decline in wild pollinator populations.
#15Apple Genetic Variation Shapes Microbial Communities and Replant Disease Outcomes
Dalhousie University, Halifax, Nova Scotia
Apple growers frequently renew their orchards. During orchard renovation, trees are removed, and the pre-existing rows are replanted with newer trees. As a result, newly planted saplings are exposed to a buildup of pathogenic organisms that accumulated during the lifespan of the previous orchard, resulting in Apple Replant Disease (ARD). ARD commonly causes stunted growth in the early years which ultimately delays fruit bearing. Increasing evidence indicate that rootstocks from the Geneva series have a higher tolerance to ARD compared to those from the Malling series. We hypothesized that this tolerance is partly driven by plant-microbe interaction, unique to each cultivar. To test this hypothesis, rootstocks from the Geneva and Malling series more specifically G41, G935, M26 and M9 were planted in pasteurized and non-pasteurized ARD soils collected from old orchards. We evaluated the differences in microbiome structure of root associated microbes between these rootstock in different ARD soils. Rootstocks responded differently to soils pasteurization, revealing distinct level of ARD resistance, with G41 being the most resistant, G935 and M26 showing moderate resistance and M9 as the most susceptible. Among the bacteria, fungi and oomycete community investigated in this research, bacteria and fungi showed community structure specific to each rootstock. Resistant rootstocks benefited from a change in bacterial community to greater extent than susceptible rootstocks. For fungi the genus Nectria was inversely correlated with plant biomass and was found to not accumulate as much in resistant rootstocks. Inversely, species from the Fusarium genus were observed as antagonistic to Nectria and positively correlated with root biomass. Fusarium was also found to be significantly more present in the roots of resistant rootstock. Taken together these findings indicate that rootstock driven resistance in ARD soil may be driven by the plant’s ability to recruit and maintain beneficial relationships with the surrounding microbiome.
#16Mirusvirus Genomic Integration and Co-Evolution with Marine Protists
Dalhousie University, Halifax, Nova Scotia
Rapid and cheap genome sequencing has triggered a revolution in our understanding of how large DNA viruses integrate into the genomes of their eukaryotic hosts. The mirusviruses, a recently discovered group of chimeric marine viruses, are an essential piece of the eukaryotic virus puzzle. The thraustochytrid protist Aurantiochytrium limacinum is the first known host of mirusviruses with two viral genomic elements: one a circular episome and another integrated into one of the nuclear chromosomes. In A. limacinum, mirusvirus genomes are vertically inherited and persistently expressed with no obvious impact on the host. Since their discovery, mirusviruses have been detected in many eukaryotic genomes, including relatives of A. limacinum, but these datasets are too fragmented to distinguish endogenous viral elements from those associated with active infection. More generally, we do not know whether endogenized mirusviruses are feature of all thraustochytrids or whether it is restricted to certain genera and species. Here we used Nanopore sequencing to produce chromosome-scale genomic assemblies for seven thraustochytrids including several novel isolates taken from aquatic environments in Nova Scotia, Canada. These high quality thraustochytrid genomes are being analyzed to (i) search for the presence of mirusvirus genetic elements, (ii) determine whether they exist as episomal and/or endogenized forms and (iii) assess host-mirusvirus interactions. Our research will provide a foundation for elucidating mirusvirus distribution across the thraustochytrids and, by extension, how these enigmatic DNA viruses evolved with, and persistently infect, their hosts.
#17Metagenomic diversity of an enrichment culture containing a novel amoeba from coastal Nova Scotia
Dalhousie University, Halifax, Nova Scotia
Little is known about the microbial diversity, particularly of protozoa, in the marine environment of Nova Scotia, Canada. Aiming to expand the known diversity of microorganisms in the region, a sample was collected from a coastal site in Nova Scotia and cultured in nutrient media to enrich for eukaryotic microbes that can survive long-term under laboratory conditions. A non-axenic subculture containing an amoeboid isolate was established from single-cell picking, and bulk DNA was extracted from the subculture for sequencing on an Oxford Nanopore PromethION. The combined long-read data from two independent sequencing runs (140 Gb, ~14 million reads in total) were used to build metagenome-assembled contigs. Preliminary analysis of the contigs revealed the presence of diverse marine bacteria, primarily belonging to the Alphaproteobacteria. Importantly, the nuclear and mitochondrial genomes of the target amoeboid eukaryote were also sequenced (~39 Mb and ~52 kb, respectively). Initial morphological investigation of the eukaryote was performed using light and electron microscopy, and the 18S ribosomal RNA (rRNA) sequence for the eukaryote indicated that it is a novel amoebozoan with no closely related 18S rRNA sequence or complete genome in the public domain. Ongoing work aims to better understand the community dynamics of the subculture, further characterize the amoeba, and obtain a more polished draft genome of the amoeba for subsequent comparative genomic investigations.
#18Barrier-Mediated Niche Partitioning Maintains Distinct Plaque and Tissue Bacteriomes During Human Experimental Gingivitis
1Faculty of Dentistry, Department of Dental Clinical Sciences, Dalhousie University, Halifax, Nova Scotia, 2Mike Petryk School of Dentistry, Faculty of Medicine and Dentistry, University of Alberta, Edmonton, Alberta, 3Faculty of Health, Dalhousie University, Halifax, Nova Scotia, 4High School Youth Researcher Summer (HYRS) Program, University of Alberta, Edmonton, Alberta, 5Department of Periodontology, School of Dentistry. Indianapolis, Indiana
Background: Microbial communities at mucosal surfaces are shaped by barrier-mediated processes that constrain dispersal and filter community membership. The human gingival sulcus, where dental plaque directly abuts junctional epithelium, offers a tractable system to study cross-niche assembly during controlled inflammatory perturbation. Whether tissue-associated bacteriomes during early gingival inflammation represent passive spillover from adjacent plaque or reflect independent ecological assembly has not been systematically examined. Methods and Results: In a split-mouth experimental gingivitis study (HREB: Pro00112019) of 22 periodontally healthy adults, paired plaque and gingival tissue bacteriomes were characterized by 16S rRNA gene amplicon sequencing across a 21-day induction period. Plaque and tissue communities were compositionally distinct at every timepoint, with differences driven primarily by each niche harboring unique taxa rather than one being a subset of the other. Ecological modeling identified limited bacterial movement across the epithelial barrier as the dominant process maintaining this separation, with smaller contributions from environmental filtering and ecological drift. Tissue selectively recruited closely related anaerobic taxa, suggesting an active filtering process, whereas plaque assembly followed a more random pattern. The degree of separation between the two communities remained stable throughout the induction period and was unrelated to clinical signs of inflammation. Tissue was enriched for anaerobic, subgingival-associated genera (Segatella, Capnocytophaga, Treponema, Fusobacterium), while plaque was enriched for early colonizers (Streptococcus, Actinomyces). Conclusions: During early gingivitis, plaque and gingival tissue harbor ecologically independent microbial communities. The intact epithelial lining acts as the primary driver of this separation, limiting bacterial movement between compartments rather than allowing passive seeding from plaque into tissue. As periodontal disease progresses, breakdown of this barrier may allow the two communities to merge a shift that could serve as a measurable early marker of the transition from reversible gingivitis to destructive periodontitis, testable in future longitudinal studies.
#19Paramoeba holobiont genomics: from kinetoplastid and Gammaproteobacteria endosymbionts
Dalhousie University, Halifax, Nova Scotia
Within the realm of symbiosis, members of the amoebozoan genus Paramoeba (Neoparamoeba) are unique in possessing an obligate, non-photosynthetic eukaryotic endosymbiont of kinetoplastid ancestry. The endosymbiont—Perkinsela—retains canonical kinetoplastid features including polycistronic transcription, spliced leader (SL) trans-splicing, and mitochondrial RNA editing. For reasons unknown, Perkinsela lies near the nucleus of its host, and the fine ultrastructure, genomic organization, and degree of genetic and metabolic integration within the Paramoeba–Perkinsela system are poorly understood. We have used long-read sequencing to produce highly contiguous genome assemblies for Paramoeba atlantica, P. invadens, and P. pemaquidensis (60–100 Mb; ~11,000–15,000 genes), along with their kinetoplastid endosymbionts (~8–20 Mb; 4,500–5,700 genes). We also assembled the genome of a previously unidentified gammaproteobacterial endosymbiont in P. atlantica that shows strong signatures of reductive evolution. Cryo-FIB-SEM imaging revealed intricate 3D spatial interactions between Perkinsela and the Paramoeba nucleus, as well as the presence of bacterial endosymbionts and free-living. Comparative genomics reveals remarkable genomic plasticity: whereas the Perkinsela nuclear genomes are gene-dense and have unmethylated transposable-element (TE) regions, the host genomes hypermethylate TE hotspots and, in P. atlantica, genes occur in dense blocks separated from TE-rich tracts. Unexpectedly, Perkinsela-derived 19-nt SL motifs were identified in the amoeba host nuclear genomes, suggesting endosymbiont-derived SL-genes or retrogenes. The Gammaproteobacterial endosymbiont is affiliated with intracellular bacteria of the order Coxiellales, closely related to members recovered from subsea biofilms and sediments. Amplicon sequence data indicated the Gammaproteobacteria increased in relative abundance over time in the Paramoeba fraction, consistent with vertical transmission. These findings show Paramoeba as a mosaic eukaryote shaped by various endosymbiosis.
#20Inside Paramoeba: Genomic, Metabolic, and Cell Biological Foundations of a Tripartite Symbiosis
Department of Biochemistry & Molecular Biology and Institute for Comparative Genomics, Dalhousie University, Halifax, Nova Scotia
Paramoeba are free-living marine amoebae linked to recurrent mass mortality outbreaks in animal hosts across temperate coastal ecosystems. Beyond their ecological impact, these organisms represent a unique and remarkable model for symbiosis research: each Paramoeba cell harbors an obligate kinetoplastid endosymbiont, known as the Perkinsela-like organism (PLO), alongside a persistent intracellular community of bacteria. This system is the only known coevolving endosymbiotic relationship between two non-photosynthetic eukaryotes, yet its genomic and metabolic architecture remains poorly understood. To characterize these two coevolving eukaryotic genomes, we paired Oxford Nanopore long-read sequencing with a novel machine learning binning classifier, SymBiNet. This allowed us to produce fully resolved metagenomic assemblies for P. invadens, P. atlantica, and P. pemaquidensis, their respective PLOs, and their associated bacterial symbionts. A pangenomic investigation reveals a conserved genomic core of the Paramoeba-PLO association across all the three species, revealing species-specific adaptations that may distinguish pathogenic from non-pathogenic lineages. The PLO genomes show signatures of intense reductive evolution by retaining functions essential to its endosymbiotic role while losing the genomic toolkit required for independent existence. Reconstruction of metabolic networks reveals a system of interlocking biochemical dependencies, offering a genomic rationale for why this association has become irreversible. Strikingly, host nuclear genomes carry kinetoplastid-derived signatures, indicating that endosymbiont-to-host gene transfer has actively shaped the evolutionary trajectory of this relationship. These results highlight the Paramoeba tripartite symbiosis as a genomic model system, which reflects how obligate eukaryote-eukaryote endosymbiosis reshapes genomes, metabolism, and evolutionary potential.
#21Precision Metagenomics for River Water Surveillance: Assessing Computational Tools to Map Environmental Drivers in Agricultural Watersheds
1Faculty of Computer Science and Institute for Comparative Genomics, Dalhousie University; 2Canadian Food Inspection Agency, Charlottetown Lab; 3Charlottetown Research and Development Centre, Agriculture and Agri-Food Canada
Riverine ecosystems are critical biogeochemical hubs where microbial communities perform the vital work of nutrient cycling and ecosystem maintenance. Despite their importance, the factors governing microbial community assembly remain insufficiently characterized, particularly the impact of intensive agricultural practices on aquatic ecosystem health. This study uses a comprehensive shotgun metagenomic approach - analysing 80+ samples across multiple seasons and sites - to characterize environmental drivers of the Dunk River watershed in Prince Edward Island (PEI). By examining influences such as precipitation, runoff, and land use, this research establishes a robust framework for river water surveillance. Our results indicate that temporal dynamics dominate microbial community structure, with statistically significant seasonal shifts. While taxonomic profiles remained relatively uniform across spatial sites, functional profiling identified distinct signatures, particularly at the Maple Plains (MP) site, suggesting that mass effects from environmental propagation may override local selective pressures. Crucially, we identified noteworthy correlations between nutrient concentrations and "long-tail" rare microbial taxa. These rare taxa may serve as essential indicators of water quality and potential health risks associated with agricultural runoff - indicators often difficult to detect using traditional methods. Because rare biosphere detection is often limited by bioinformatic sensitivity, we are performing a computational assessment of taxonomic profiling tools to bridge this gap. This involves benchmarking 12 specialized profilers encompassing k-mer-based, marker-gene-based, and deep-learning-based architectures. To evaluate performance, they were tested against a simulated ground truth modelling river ecosystem complexity. This benchmarking methodology provides critical insights into which profilers are best suited to accurately capture low-abundance "long-tail" taxa. Ultimately, this work demonstrates how optimized bioinformatic pipelines enhance environmental monitoring and public health protection.
#22Gut microbiome profiling of individuals in an assisted-care facility
Dalhousie University, Halifax, Nova Scotia
Alterations in gut microbiome composition and metabolic activity have been reported to be closely associated with aging. Frailty, defined by a relative proportion of accumulated health deficits, is a critical aging-related health state. Older adults living in assisted-care facilities represent a highly frail population, and understanding the relationships between gut microbiome changes, aging, and frailty in this population may suggest new ways to assess and mitigate frailty. In this study, we collected up to five weekly fecal samples from a cohort of 47 residents in an assisted-care facility in Halifax, Nova Scotia, Canada, and profiled their microbial communities using both amplicon-based taxonomic profiling and metagenomic approaches. Despite their shared residence, substantial inter-individual variation was observed, while the majority of individuals exhibited relatively stable gut microbiome profiles over one month of sampling. In addition, we found no significant associations of frailty or age with gut microbiome taxonomic diversity. Metagenomic analysis identified dozens of functional pathways associated with age, with many related to core biosynthetic processes, including inosine-5'-phosphate, L-histidine, and peptidoglycan biosynthesis. No significant associations were found between functional pathways and either frailty or residence time at the facility. Age was also significantly associated with inter-individual differences in overall pathway composition. Finally, individuals with higher antimicrobial resistance (AMR) gene abundance showed increased aminocoumarin- and macrolide-related resistance and decreased tetracycline- and rifamycin-related resistance, distinct from AMR patterns in other individuals. Altogether, although no significant associations were found between frailty and microbiome features, the identification of several age-associated functional pathways and distinct AMR patterns in individuals with high AMR gene abundance provides insights into gut microbiome changes in aging populations.
#23Improving Divergent Antimicrobial Resistance (AMR) Gene Detection Using Gene Neighbourhood Information
Dalhousie University, Halifax, Nova Scotia
The increasing prevalence of drug resistance in bacteria is a global human-health crisis. The commensal microbiome serves as a reservoir for AMR genes that can facilitate their spread to pathogenic bacteria such as Pseudomonas aeruginosa via horizontal gene transfer (HGT). This exchange of genetic material may enable pathogens to acquire new resistance traits, leading to multidrug resistance. Methods for detecting novel and emerging AMR genes are key to tracking and clinically controlling their propagation. Current genomics detection methods typically rely on comparisons to reference AMR gene sequences, but these references are biased toward bacteria that have been intensively studied and sequenced due to their clinical relevance. While effective at detecting AMR genes from those well-characterized bacteria, current methods may fail to detect divergent threats. Consequently, novel and emerging AMR genes in less-studied members of the microbiome may go undetected. Examining the genes that flank AMR genes (the “neighborhood”) can augment the evidence gained through sequence comparison alone. In this study, we analysed the neighbourhoods of an aminoglycoside resistance-conferring gene, AAC(6′)-Ib, in a set of 13 Pseudomonas species, using 6,093 publicly available assembled genomes. Genes similar to the known AAC(6′)-Ib were predominantly detected in P. aeruginosa, with a few notable exceptions in two distantly related species, likely attributable to HGT. The genes flanking AAC(6′)-Ib showed moderate to high conservation even in the distantly related species. Our neighbourhood analysis revealed a potential divergent AAC(6′)-Ib gene with lower identity (<70%). The identified gene had matching neighbouring genes to those of known AAC(6′)-Ib flanking genes, despite not being detected by current methods. This pilot analysis supports the possibility of using gene neighbourhoods to detect divergent AMR genes, improving surveillance of emerging threats.
#24Soil biodiversity signatures for scalable agricultural assessments
1Department of Plant, Food, and Environmental Sciences, Faculty of Agriculture, Dalhousie University; 2Centre for Biodiversity Genomics, University of Guelph; 3Department of Integrative Biology, University of Guelph; 4McCain Foods Limited, Florenceville-Bristol, New Brunswick; 5Department of Botany, Jahangirnagar University
Soil bacteria, fungi, and invertebrates underpin soil fertility, nutrient cycling, crop productivity, and carbon storage, yet translating this biological complexity into actionable knowledge for agricultural decision-making remains a persistent challenge. Recent advances in DNA metabarcoding enable the monitoring of entire soil biological communities, offering new opportunities for soil health assessment. This study evaluated the application of metabarcoding at McCain’s Farm of the Future Canada. Using a multi-year soil dataset from this 500-acre commercial farm in New Brunswick managed under diverse regenerative practices, we show that DNA metabarcoding generates reproducible and scalable soil biodiversity signatures. Our results suggest that regenerative practices shape soil microbial dynamics and are associated with improved soil health indicators, including measurable gains in soil organic carbon. Together, these findings demonstrate that DNA metabarcoding can generate reproducible, farm-scale soil biodiversity signatures linked to regenerative practices and soil health outcomes in potato production with promise for broader applications in sustainable agriculture across crops and regions.
#25Evaluating Zooplankton community dynamics during the spring bloom in the Labrador Sea through multi-marker metabarcoding
Dalhousie University, Halifax, Nova Scotia
The Labrador Sea is an important region of the Northwest Atlantic where unique oceanographic processes create the biological carbon pump (BCP). Deep Winter convection, reaching depths of up to 2300 m, replenishes surface nutrients, while seasonal stratification following sea-ice melt initiates intense spring phytoplankton blooms. These blooms have been historically dominated by diatoms but are being increasingly influenced by Phaeocystis and other taxa, driving primary production and carbon export. As phytoplankton biomass accumulates and grazing pressure increases, zooplankton convert this production into energy-rich lipids and rapidly sinking particulate organic carbon, linking surface productivity to deep ocean carbon sequestration. Here, we characterize zooplankton, microzooplankton, and parasitic communities across the progression of the spring bloom using a multi-marker metabarcoding approach targeting both mitochondrial (COI) and eukaryotic ribosomal (18S rRNA) genes. We applied a metabarcoding framework to characterize zooplankton communities and quantify methodological biases. Zooplankton samples from 2022 and 2024 were size-fractionated, freeze-dried, homogenized, and DNA extractions were performed. A synthetic spike-in (1% of total copies) was incorporated into the 18s rRNA barcode to enable quantitative assessment of sequencing performance. Mock communities composed of equal DNA contributions from 18 representative taxa were constructed to evaluate amplification bias. Libraries targeting the COI and 18S rRNA V4 barcodes were prepared and sequenced on an Illumina NextSeq PE300 platform. This integrative approach enables high-resolution assessment of zooplankton community structure and methodological bias, providing improved understanding of community composition and carbon export potential in the Labrador Sea. Understanding these dynamics is essential for predicting how climate-driven shifts in plankton communities may alter the efficiency of the Northwest Atlantic biological carbon pump.
#26Assessing the use of quantitative transcriptomics to estimate growth, stress, and macromolecular allocation of an oceanic Synechococcus strain
Dalhousie University, Halifax, Nova Scotia
Marine phytoplankton are a group of microorganisms who contribute roughly 50% of the world’s annual net primary productivity. Oceanographers use a host of methods to categorize populations of phytoplankton and quantify their effects on global ecological and biogeochemical processes. Molecular techniques are increasingly being deployed to better determine phytoplankton community composition and cell physiology across oceanic regimes. Recently, quantitative metatranscriptomics, through the use of mRNA spike-ins, has been used as a taxa-specific measure of biomass over a transect in the eastern pacific. However, changing cell state could confound this relationship between transcripts and biomass. To assess this, we analyze the transcriptome of a cultured marine Synechococcus strain over varying light, temperature, and growth rates (at steady state) to assess the stability of its “transcript budget”. From this, we created a simple model which shows both growth rate and light and temperature stress have a positive impact on the total number of mRNA transcripts (per cell and per unit biomass) measured. KEGG annotations were also used to determine how specific transcripts associated with major cell macromolecular pools varied over these changing cell states.
#27Insights into body site-specific microbiomes and challenges in acquiring reliable profiles in low biomass samples
1Faculty of Medicine, Dalhousie University; 2Division of Thoracic Surgery, Queen Elizabeth II Health Sciences Centre, Dalhousie University
Background: Unique microbial profiles have also been described in tumours and other low-biomass samples however, recent controversy has highlighted the challenges of utilizing these microbiomes for cancer diagnosis and treatment. This work aims to address current challenges with low-biomass samples and identify microbial biomarkers that can be used to predict lung cancer. Methods: This is an observational prospective study of patients undergoing curative surgery for early-stage non-small cell lung cancer (NSCLC). Participants provided samples from multiple body sites (saliva, tumour, adjacent tissue, and blood). Extracted DNA underwent full length PCR amplification and PacBio sequencing of the 16S rRNA gene. Due to failed full length sequencing on low biomass samples, a second nested PCR amplification of the V6-V8 region and subsequent sequencing on the Illumina MiSeq was performed. A subset of samples was quantified using QIAcuity digital PCR and metagenomic sequencing on the NextSeq2000. Taxonomic profiles were analyzed in relation to clinical parameters including cancer stage, PDL1 status, and survival outcomes. Results: Our pilot data demonstrates that saliva and bronchoalveolar lavage fluid have over a hundred observed features while tumour, adjacent tissue, and blood have less than 10 features, similar to negative controls. Of concern is the lack of consistency and reproducibility with samples of low microbial biomass. On the other hand, microbially rich samples with robust profiles are associated with clinical variables. Implications: Current methodologies present challenges with samples of low microbial biomass, making it difficult to identify biomarkers that are distinct or above levels found in collection and processing environments.
#28Assessing dPCR for microbial absolute quantification in low microbial biomass samples and a cross platform human-derived positive control in a sequencing service framework
Integrated Microbiome Resource (IMR), Dalhousie University, Halifax, Nova Scotia, Canada
Rigorous and reproducible microbiome profiling remains limited by platform-dependent bias, host DNA interference, and the lack of biologically representative controls, particularly in applied sequencing settings. To improve analytical robustness within a microbiome sequencing service, a human-derived microbial positive control (MAK) was implemented across workflows, alongside the use of digital PCR (dPCR) to determine absolute bacterial copy numbers under controlled increases in host DNA background. A single human stool sample was homogenized to generate an in-house positive control for microbiome profiling. The MAK control was sequenced 164 times to assess consistency across multiple sequencing platforms and runs. Probe-based assays (Pan Bacteria 1 and 3; QIAGEN), and 16S variable-region primers were compared to quantify absolute bacterial copy number on a QIAcuity dPCR system (QIAGEN) across gut, oral, and controls (ZymoMock and MAK). Bacteria copy numbers were measured on a controlled spike‑in mixtures (0.0001–99.9%) of bacterial DNA (ZymoMock and MAK) to human DNA (HEK cells) representing host background. The dPCR-derived bacterial copy numbers from 4 samples, obtained using several variable-region primer assays showed high correlation with Pan Bacteria probe 1 (R² = 0.95) and probe 3 (R² = 0.97). In low-biomass samples (e.g., saliva), 100-fold dilutions improved precision (CI = 0.036), while 1000-fold dilutions performed better for high-diversity gut samples (CI = 0.027). The gradient experiments showed V3V4 primers exhibited substantial background amplification at low bacterial proportions, limiting reliable detection below a 10% bacteria-to-host ratio. Pan Bacteria probe 1 assay accurately detected incremental copy number changes down to 0.01% bacteria-to-host ratio. Initial comparison of the multi-run cross-platform MAK data showed concordance at the phylum-level among the bioinformatic workflows, however the genus-level was variable, while still showing Prevotella dominance in all cases. Together, these approaches enable improved standardization, workflow-specific bias detection, and accurate cross-study and cross-platform comparisons, especially for host-associated and low-biomass samples.
#29Full-length 16S ribosomal RNA gene sequencing reveals co-occurrence of tick-adapted and environmentally derived members of the microbiome of the black-legged tick, Ixodes scapularis in Nova Scotia, Canada
Dalhousie University, Halifax, Nova Scotia
Lyme disease is a tick-borne illness caused by the spirochaete bacterium Borrelia burgdorferi. The black-legged tick Ixodes scapularis, which transmits B. burgdorferi and several other important human pathogens, is endemic to the eastern United states and, due to climate change, is quickly moving north into regions of central and eastern Canada. The amplification and sequencing of bacterial DNA from I. scapularis is an increasingly common tool for monitoring the presence and abundance of B. burgdorferi and associated bacteria in ticks. However, the precise nature of the molecular data collected often varies within and between studies, which presents challenges for data analysis and interpretation. Here we use full-length Oxford Nanopore 16S ribosomal RNA gene amplicon sequencing to characterize the microbiome of I. scapularis, with an explicit focus on distinguishing between tick-adapted bacteria (endosymbionts and pathogens) and environmentally acquired taxa. We show that environmental dominance strength differs between these two ecological classes of bacteria, and that environmental dominance does not appear to represent stochastic background alone; environmental-derived taxa detected in tick microbiomes are not mere contaminants. Paired soil microbiome profiling from tick collection sites will be required to test whether environmental dominance and associated co-occurrence structure track with seasonal changes in exposure and environmental microbial populations.
#30Bacterial community structure on marine snow particles collected during two Phaeocystis-dominated spring blooms in the Labrador Sea
Dalhousie University, Halifax, Nova Scotia
The biological carbon pump describes a set of processes in which biologically produced carbon is exported to the deep ocean. This process is primarily driven by the sinking of marine snow, defined as organic particles larger than 0.5 mm. Microbial community composition regulates the production and remineralization of marine snow throughout the water column. Research expeditions conducted in the Central Labrador Sea in May-June 2022 and 2024 each captured a different phase of the Phaeocystis pouchetii-dominated spring bloom: bloom decline in 2022 and mid-bloom in 2024. During each expedition, we deployed marine snow catchers (MSCs) at distinct spatiotemporal locations to investigate the microbial communities associated with suspended, slow-sinking and fast-sinking particles at the base of the euphotic zone and in the mesopelagic zone. Here, we use eDNA metabarcoding of the V4-V5 hypervariable region of the 16S rRNA gene to investigate the drivers of bacterial community composition across environmental conditions, depth, and particle type. Beta diversity analyses revealed sampling year influenced bacterial community composition, and this influence increased with depth. When investigating the 2022 and 2024 samples separately, depth of the MSC deployment strongly influenced the bacterial community composition. Analyses within each part of the water column of each year revealed that communities just below the euphotic zone were influenced by spatiotemporal location, while particle type had a greater influence on community composition in the mesopelagic zone. Differential abundance testing revealed several taxa that were consistently enriched in the fast- and slow-sinking particles across MSC deployments including Ulvibacter, Paraglacieola, Fluviicola, Vicingus, Spongiibacteraceae and Colwellia. Our results reveal the shifts in bacterial community composition of particles as they sink in the water column and bacterial species associated with sinking particles across bloom stages and depths.
#31Tracking two exceptional nitrous oxide (N2O) peaks in Bedford Basin, Nova Scotia: nitrogen cycling and microbial community dynamics
Dalhousie University, Halifax, Nova Scotia
Nitrous oxide (N2O) is a highly potent greenhouse gas, around 200-300 times as potent as CO2, and the principal ozone-depleting agent. The oceans naturally contribute roughly one-third of global emissions, and marine N2O production is driven primarily by biogeochemical reactions of the nitrogen cycle performed by a wide array of microorganisms, including nitrification and denitrification. These processes are strongly influenced by oxygen, nutrient concentrations, and the presence of suitable microbial communities. Dissimilatory reduction of nitrate to ammonia (DNRA) is considered a recycling process that can fuel nitrification by producing ammonia. In recent years, N2O has been detected in the deep waters of Bedford Basin, a coastal fjord-like ecosystem located in Halifax, Nova Scotia (Canada). At this depth, the system undergoes annual nitrification and is periodically subjected to hypoxia, which is usually interrupted by mixing and oceanic intrusions. We present results from a 5-year time series with weekly resolution at 60 m in the basin, during which two unprecedented high-N2O accumulation events occurred at the lowest oxygen levels. We report nutrient dynamics and nitrogen cycle processes underlying the N2O peaks, along with other environmental variables. Using paired 16S rRNA amplicon sequencing, we describe the impact of the environment on the microbial community and the microorganisms associated with the N2O peak events. The N2O and oxygen concentrations, as well as some key microorganisms, during the N2O peaks in the basin, resemble those reported in Oxygen Minimum Zones (OMZs), the ocean regions with the highest N2O concentrations. In addition, we briefly discuss the use of metagenomics to identify the roles of key microorganisms during these N2O events.
#33Vaginal Microbiota Associated with Preterm Labour Proteolyze Uterine Receptors Involved in Labour Activation Pathways
Dalhousie University, Halifax, Nova Scotia
Preterm birth is the leading cause of neonatal mortality, affecting >15 million pregnancies annually. Approximately 60% of preterm labour (PTL) cases arise from intrauterine bacterial infection, often due to the vertical migration of dysbiotic vaginal microbiota including Prevotella bivia. Our previous work shows that P. bivia secretes proteases that degrade structural components of pregnancy tissues, mimicking the activity of human labour-activating proteases. Human proteases can also activate labour pathways via uterine proteinase-activated receptors (PARs), where cleavage of an activation ligand in the PAR extracellular domain initiates contractile and inflammatory signalling pathways. In this study, we examine PAR proteolysis by P. bivia secreted proteases as a mechanism of PTL initiation. PAR proteolysis was assessed with fluorometric protease assays using fluorophore-quenched (FQ) peptides corresponding to the extracellular domain sequences of PAR1 and PAR2. Cell-free supernatants from P. bivia proteolyzed PAR1 and PAR2 peptide pools, indicated by increasing fluorescence over time. Next, metallo-, cysteine, and serine protease inhibitors were incorporated into the assays. The metalloprotease inhibitor, 1,10-phenanthroline, abrogated proteolysis of the PAR1 and PAR2 peptide pools, confirming that secreted metalloproteases from P. bivia confer the PAR-proteolyzing activity. Prevotella bivia supernatants were incubated with the individual FQ peptides to discern the relative positionality of PAR1 and PAR2 proteolysis. Supernatants from P. bivia cleaved peptides corresponding to the activation ligand in PAR1 and PAR2. Cleavage also occurred in peptides corresponding to regions downstream of the activation ligands in PAR1 and PAR2, suggesting that P. bivia proteases also inhibit canonical PAR activation. Altogether, our findings reveal that secreted metalloproteases from P. bivia proteolyze peptides corresponding to the extracellular domains of PAR1 and PAR2, potentially leading to inappropriate activation or disarmament of uterine PARs. Bacterial targeting of uterine PARs would be a novel mechanism by which dysbiotic vaginal bacteria could contribute to premature labour initiation.
#34Secreted metalloprotease activity from vaginal Prevotella bivia dysregulates endothelin signaling in infection-induced preterm labour
Dalhousie University, Halifax, Nova Scotia
Bacterial vaginosis is a prevalent dysbiotic vaginal condition that increases the risk for health complications. In pregnancy, select BV-associated Prevotella species can vertically migrate to the uterus and have been implicated in triggering fetal membrane rupture and preterm labour. However, we lack mechanistic understanding of how Prevotella species contribute to adverse pregnancy outcomes. Vaginal Prevotella bivia encodes a metalloprotease called PepO that exhibits sequence and domain similarity to the human labour-activating protease ECE-1. Human ECE-1 contributes to labour onset by cleaving the precursor peptide big endothelin-1 (bET1) into bioactive endothelin-1 (ET1), which induces uterine contractions. We hypothesize that P. bivia PepO proteolyzes bET1 into bioactive ET1, thereby mimicking the activity of human ECE1 to initiate uterine contractile signalling and premature labour. Comparative sequence analysis of P. bivia PepO and human ECE-1 revealed conservation of the substrate-binding domain (VNAYY) and metalloprotease active site (HEXXH). Proteolysis was quantified from P. bivia cell-free supernatants and purified recombinant proteases using a fluorophore-quenched substrate specifically designed to measure ECE-1 activity. Purified P. bivia PepO proteolyzed the ECE-1 substrate, with a reaction rate that was 4-fold faster than the reaction rate for human ECE-1. Inclusion of the ECE-1 inhibitor phosphoramidon reduced PepO protease activity by 82% and site-direct mutagenesis of the P. bivia PepO active site (HAXXH) abolished protease activity. Cell-free supernatants from P. bivia were also found to proteolyze the ECE-1 substrate, which was partially inhibited by phosphoramidon. To assess bET-1 to ET-1 proteolytic conversion, cell suspensions of P. bivia were co-incubated with the bET-1 precursor over a time-course. SDS-PAGE analysis confirmed bET1 degradation by P. bivia, which also corresponded to detection of the bioactive breakdown product, ET1, via immunoassay. This work identifies a new mechanism through which P. bivia could induce or augment labour by inappropriately activating contractility signaling pathways through bET-1 proteolysis.
#35Tool assisted curation of gene predictions in eukaryotic genomes
Dalhousie University, Halifax, Nova Scotia
Prediction of protein coding genes in eukaryotic genomes has improved drastically in the last decades but remains a challenge in non-model organisms. Once a new genome has been sequenced, it is standard practice to apply gene prediction pipelines that integrate ab initio predictors with RNA-seq and protein data, followed by extensive but careful curation of the predicted genes. The curation process can be very laborious and time consuming. After applying the BRAKER2 gene prediction pipeline on the newly sequenced genome of the metamonad Ergobibamus cyprinoides, we noticed that many genes were predicted to have introns that were not supported by the RNA-seq data. False introns can lead to errors in gene models, such as truncated genes or artificial mergers of neighboring genes. To remedy this problem, we developed a python script, fix_genes_with_false_introns.py, that automatically identified such genes and replaced them with new gene models that only include supported introns. The new models were generally in line with how we would have manually curated the model. Although the script only fixes one type of gene prediction error, we found it drastically reduced the amount of time and work necessary to complete the curation process. The script is publicly available on our GitHub page github.com/Dalhousie-ICG/icg-shared-scripts
#36Dynamic microbial interactions of anaerobic protist Anaeramoeba
Dalhousie University, Halifax, Nova Scotia
Anaeramoeba flamelloides are free-living anaerobic protists that possess an extensive intracellular membrane structure called a symbiosome, where membrane invaginations connected to the extracellular environment encapsulate symbiotic sulfate-reducing bacteria, Desulfobacter. Multiple metabolic relationships are proposed for this symbiotic relationship, where the Desulfobacter symbiont consumes hydrogen, acetate, and propionate from the host hydrogenosomes for methylmalonyl-CoA metabolism, dissimilatory sulfate reduction, and Wood-Ljungdahl pathway to produce energy. This consumption by symbionts, in turn, relieves product inhibition of host energy metabolism. Furthermore, the symbiont encodes a near-complete anaerobic pathway for vitamin B12 biosynthesis, while the host genome of A. flamelloides encodes six vitamin B12-dependent enzymes, which is uniquely high for a eukaryotic organism. Surprisingly, in a survey of microbial composition among multiple A. flamelloides cultures originating from the same strain, several were found to have lost the original Desulfobacter symbiont. Instead, other bacteria, including Desulfovibrio, Halarcobacter, and Terasakiella, which are known to consume hydrogen for energy metabolism, were highly increased, suggesting potential replacement of the symbiont and alternative relationships among different members of the consortia. This study aims to identify the key species and the community-scale metabolic potential within Anaeramoeba flamelloides microbial consortia. To investigate the proposed system, selected A. flamelloides cultures will be comparatively analyzed for microbial composition and metagenomically sequenced. Metabolic complementarity among species will be predicted by constructing genome-scale metabolic networks from metagenomic data. In this presentation, I will share progress in comparative analyses of microbial community composition across A. flamelloides cultures, highlighting potential shifts in symbiotic associations.
#37Metagenomic reconstruction of a novel Izemoplasmatales lineage from Skoliomonas culture
Dalhousie University, Halifax, Nova Scotia
Bacterial communities present in xenic eukaryotic cultures form an important network. This network can provide various benefits to eukaryotic organism, most notably serving as a source of food. In addition, they can strongly influence growth by supplying different chemical compounds. The single celled haloalkaliphilic eukaryotic cultures of Skoliomonas litria and Skoliomonas sp. GEMRC harbor numerous bacterial associates. To better understand the culture dynamics and enhance Skoliomonas growth, metagenome sequencing was performed. In this study, we detected nearly 12 different taxa. One of these was identified as Mycoplasma-related; therefore, to investigate its potential host association, we reconstructed the complete genome of this uncultured bacterial lineage from a Skoliomonas culture. Phylogenetic analysis based on 120 marker genes placed our bacteria within the order Izemoplasmatales. Although it exhibits features commonly associated with obligate symbiotic or parasitic bacteria such as small genome size and low GC content, the absence of substantial metabolic reduction suggest that it is unlikely to have an obligate symbiotic lifestyle. However, we found that this lineage contains duplicated rRNA operons and multiple IS elements. In contrast, closely related genera such as Izemoplasma and Xianfuyuplasma do not show rRNA duplication or IS expansion. This suggests that these genome structural changes were likely acquired relatively recently after divergence at the genus level. These repetitive elements are known to promote genome rearrangements, and thus this lineage may currently be in a relatively dynamic phase of genome evolution. Overall, our study demonstrates that genome structural evolution can directly influence how a lineage is reconstructed from metagenomic data and highlights the importance of combining culture-based approaches with long-read sequencing to fully resolve such features.
#38Decoding microeukaryote microbiome interactions: Implications for marine ecosystems and biotechnological potential
Dalhousie University, Halifax, Nova Scotia
Microorganisms play a fundamental role in maintaining ecosystem balance and regulating key processes on Earth, including energy flow, biogeochemical cycles, and climate dynamics. Despite their importance, environmental microorganisms and the complex networks they form remain poorly understood. This work seeks to uncover and unravel the complex network of interactions between microeukaryotes and their microbiome, revealing how these relationships drive ecological dynamics and unlock biotechnological potential. This project explores the bacterial microbiome associated with the heterotrophic dinoflagellate Oxyrrhis marina. To address who makes up these microbiomes and how these associations function, we used a multidisciplinary and multi-omics approach. Microbiome and metagenomic analyses were employed to identify microbiome composition, while classical microbiology techniques combined with whole-genome sequencing enabled in-depth characterization of culturable microorganisms. Microbial interactions were investigated using chemotaxis assays, co-culture experiments, metabolic network predictions, and differential gene expression analyses based on RNA sequencing. Results revealed that the microbiome of O. marina comprises more than 100 bacterial species. Behaviorally, O. marina exhibited both positive and negative chemotactic responses depending on the bacterial partner. Its growth rate also varied in co-culture, in some cases being negatively affected, suggesting not only mutual dependence but also a finely balanced microbiome structure and composition. Metabolic predictions and gene expression analyses point to complex interactions driven by complementary metabolic pathways between organisms. Additionally, genomic analysis of culturable bacteria uncovered a wide range of potential applications, including the production of antifungal compounds, degradation of hydrocarbons and phthalates, reduction of heavy metals, biomineralization processes, and the synthesis of essential cofactors and vitamins such as cobalamin. This work suggests that microbiomes associated with microeukaryotes are far more complex than previously recognized. They represent rich reservoirs of microbial diversity with significant biotechnological potential and may play essential roles in the modulation, transformation, and resilience of marine ecosystems.
#39Microbiome Sequencing of Frozen Nasal Wash Samples from Adults Intranasally Inoculated with Bordetella pertussis
Dalhousie University, Canadian Centre for Vaccinology, Halifax, Nova Scotia
Bordetella pertussis is the causative agent of pertussis, a highly infectious respiratory disease. The nasal microbiome is understudied, particularly in disease states. The stability of the nasal microbiome during pertussis infection is of interest. This project investigated the nasal microbiome of healthy adults intranasally inoculated with B. pertussis in a Controlled Human Infection Model (CHIM) conducted by the Canadian Center for Vaccinology. Frozen, neat nasal wash samples collected during the CHIM were selected for this study based on vaccine priming status (whole-cell pertussis vs. acellular pertussis), clinical outcome (infected vs. non-infected), and the dose category to which they were inoculated. Per participant, samples from Day-1 (one day before participants were challenged), Day 3 (three days post challenge), Day 7, and Day 10 were included to study changes in microbiome composition after inoculation. Total nucleic acid (TNA) was extracted using the KingFisher Duo Prime magnetic particle processor and the MagMAX™ Prime Viral/Pathogen NA Isolation Kit. Full-length 16 Svedberg (16S) ribosomal RNA (rRNA) sequencing was performed using the PacBio Vega system. Bioinformatics analysis was performed using QIIME-2, following Microbiome Helper standard operating procedures. Microbiome analysis revealed that clinical outcome was associated with significant differences in β-diversity, while α-diversity remained unchanged. Dose-dependent effects were significant for α-diversity, in both infected and non-infected participants. These results suggest that inoculum level may influence baseline community structure more strongly than clinical outcome itself. Differential abundance analysis identified B. pertussis as significantly enriched in infected participants at Days 7 and 10 and Cutibacterium acnes as significantly enriched in non-infected participants at Day 10. These findings indicate that pertussis infection drives changes in the relative abundance of specific taxa without altering the alpha diversity within samples. Our findings provide a foundation for understanding microbiome-mediated factors that influence susceptibility to B. pertussis inoculation and infection.
#40Microbial Dominance Drives Apple Replant Disease Across Soil and Root-Associated Microbiomes
1Department of Plant, Food, and Environmental Sciences, Faculty of Agriculture, Dalhousie University; 2Agriculture and Agri-Food Canada, Kentville Research and Development Centre; 3United States Department of Agriculture (USDA), Agricultural Research Service, Grain Legume Genetics and Physiology Research Unit, Prosser, WA; 4Agriculture and Agri-Food Canada, Summerland Research and Development Centre, Summerland, BC; 5Agriculture and Agri-Food Canada, Fredericton, New Brunswick
Apple replant disease (ARD) is a persistent, biologically mediated soil syndrome that suppresses root development and limits orchard productivity, yet the microbial mechanisms underlying disease severity remain unresolved. We combined soil pasteurization assays with high throughput fungal community profiling to disentangle soil and root associated drivers of ARD across six apple orchards spanning a gradient of disease pressure. Soil pasteurization alleviated root growth suppression across all orchards, increasing mean root dry mass by nearly threefold (1.43 ± 0.06 vs. 0.50 ± 0.03 g; P < 0.05), confirming a strong biological basis for ARD. Disease severity, quantified as proportional root biomass reduction, ranged from 50.4% to 74.5% among orchards, indicating pronounced site specific microbial constraints. Despite these differences, beta diversity analyses revealed substantial overlap between high and low disease orchards. Community structure was shaped by orchard identity (soil R² = 0.56; roots R² = 0.41; P = 0.001), but not manifests large scale community turnover. Instead, ARD severity aligned with taxon specific shifts in dominance. Differential abundance analyses revealed that native field soils were enriched in multiple fungal taxa across functional guilds, including pathogenic (Nectria, −3.49), beneficial (Trichoderma, −1.28), and yeast genera (Apiotrichum, −3.7), while high disease soils showed selective enrichment or suppression of a small subset of dominant taxa. Notably, root associated fungal alpha diversity was decoupled from disease severity, whereas dominance sensitive metrics declined by 30–45 % in high disease orchards. Together, these results demonstrate that ARD arises from subtle but consequential re weighting of dominant fungal taxa within broadly conserved microbiomes. Our findings highlight fungal dominance dynamics, rather than diversity loss, as a central mechanism structuring disease expression in perennial agroecosystems, with implications for microbiome informed disease management.
#41A Streamlined Method for Bacterial RNA Extraction and Purification
Gladstone Institutes
Here we describe a streamlined bacterial RNA extraction and purification protocol that is faster and easier to perform. Previously, Stead et al. established a formamide-based extraction, RNAsnap, that provides a fast method for extracting RNA but still requires subsequent purification. In our version, we combine both the extraction and purification steps into one protocol that is comparable in time to doing a miniprep. Using our approach, we yield purified RNA concentrations in the thousands of ng/µL (from an input of ~10^8 CFUs). To evaluate our protocol’s versatility, we tested it across a variety of mono and di -derms, as well as mouse stool samples. Overall, our method offers a fast, accessible, and reliable alternative for bacterial RNA extraction and purification.
#42Developing an efficient and easy-to-use pipeline to identify genomes from metagenomic data
Université du Québec à Montréal, Montreal, Quebec
Metagenomics is a powerful approach studying the assemblage of genes and genomes from complex microbial communities and eDNA. It is vital to the understanding of rare, novel, and culture resistant species, and uncommonly studied environments. Bioinformatics tools for metagenome analysis enable the assembly, identification, and characterization of bacterial genomes, known as metagenome-assembled genomes (MAGs), from metagenomic shotgun sequencing. However, the continuous development of new technologies and algorithms is outpacing the capacity of current bioinformatic pipelines to stay up to date, highlighting the need for a new, more comprehensive pipeline for MAG assembly and analysis. We are developing a pipeline which will notably include: the capacity to input short and long reads (either alone or as a hybrid input), assembly algorithms designed to handle increasingly large samples with reduced computational demand, and a combination of the most commonly used and state-of-the-art AI-based binning algorithms. This pipeline will enable researchers in multiple fields to perform metagenomic analyses, to profit from novel technologies, and to easily adapt them to their needs, regardless of their experience in bioinformatics.