The field of human gut microbiome research has long grappled with the challenge of compiling comprehensive genomic databases that accurately reflect the diverse ecosystems within the gastrointestinal tract. Recently, a groundbreaking advancement has emerged with the introduction of the Human Reference Gut Microbiome version 2 (HRGM2), a vast and meticulously curated catalogue of near-complete metagenome-assembled genomes (MAGs) and isolate genomes. This resource represents an unprecedented leap in both the volume and quality of gut microbial genomic data, promising to redefine how we understand the microbial world inside us.
Historically, the construction of gut microbiome catalogues has been limited by geographic and genomic resolution constraints. Many existing datasets have tended to overrepresent certain populations, leading to skewed insights that fail to capture the full global diversity of the human gut microbiota. Furthermore, the majority of metagenome-assembled genomes previously available have been of medium quality, often missing significant portions of their genomic content—up to 50% in some cases. Such incomplete genomes hamper the accuracy of downstream analysis, including predictions of microbial functions and interactions.
The HRGM2 catalogue stands as a monumental upgrade, comprising 155,211 non-redundant genomes that meet stringent criteria for completeness and contamination. Each genome in the set exhibits at least 90% completeness while maintaining contamination levels below 5%, ensuring near-complete genomic integrity. Spanning an impressive 4,824 prokaryotic species sourced from 41 different countries, this collection not only increases the number of known gut-associated genomes by 66% but also boosts species diversity by 50% relative to the previous gold-standard Unified Human Gastrointestinal Genome catalogue.
This geographical richness is particularly noteworthy, as including samples from a wide array of global populations paves the way for more accurate, inclusive representations of gut microbiome diversity. As human gut microbiota composition is known to vary substantially due to diet, lifestyle, genetics and environment, HRGM2’s expanded geographic footprint promises insights that are more globally relevant. In turn, this can drive more effective, personalized interventions in microbiome-informed health and disease research worldwide.
Beyond sheer volume and diversity, HRGM2’s contribution to methodological advancement is compelling. The high-quality assemblies allow for more precise species profiling directly from DNA sequencing data. Researchers can now resolve strain-level heterogeneity within microbial communities with unprecedented clarity, an essential factor since subtle genetic differences between bacterial strains can translate to vastly different functional roles or clinical impacts.
The enhanced genomic resolution also facilitates a comprehensive survey of the human gut resistome—the collection of antibiotic resistance genes within the microbiome. Understanding how resistance genes are distributed and evolve across different populations is critical for combating the global threat of antimicrobial resistance. HRGM2’s extensive dataset allows for refined mapping of resistance mechanisms embedded within the gut flora, potentially informing more targeted and effective therapeutic strategies.
Crucially, the use of HRGM2 genomes dramatically improves the accuracy of functional annotation, particularly regarding metabolic potential. Prior reliance on fragmented or incomplete genomes between 50% to 70% completeness introduced uncertainties in predicting metabolic pathways and their contributions to host physiology. With near-complete genomes, microbial functional repertoires can now be inferred with high confidence, offering clearer windows into how gut bacteria synthesize vitamins, metabolize complex carbohydrates, or influence the host immune system.
This new genomic foundation has enabled the deployment of sophisticated, automated genome-scale metabolic models (GEMs) that simulate the entire gut microbiota’s metabolic interactions. These models integrate comprehensive genomic data to predict community-level metabolic fluxes and interspecies dependencies, unveiling the biochemical dialogues that modulate health and disease states. Researchers can explore how microbial metabolism varies between healthy individuals and those with diseases such as inflammatory bowel disease or metabolic syndrome, potentially revealing novel microbial therapeutic targets or biomarkers.
HRGM2 also promises to enhance strain-resolved longitudinal studies, empowering efforts to characterize the dynamic nature of microbial communities over time and under different environmental pressures including diet, antibiotics, or probiotics. Tracking the metabolic and genetic shifts at the strain level opens opportunities to intervene more precisely in microbiome-mediated diseases.
In addition to its scientific payoff, the HRGM2 resource embodies progress in computational microbiology. Constructing a catalogue of this scale with high-quality genomes required innovations in metagenomic assembly, contaminant filtering, and dereplication methods, highlighting the rapid evolution of bioinformatics tools in the microbiome field. The ability to integrate and harmonize datasets from numerous countries and sequencing studies also underscores the power of global scientific collaboration in microbiome research.
Moreover, this enriched catalogue provides an invaluable reference framework for future metagenome studies. The expanded database facilitates more sensitive and specific microbial read mapping, reducing noise and increasing accuracy in gut microbiome sequencing projects. Researchers examining novel cohorts can rely on HRGM2 for rapid taxonomic classification and functional inference, vastly improving the throughput and reliability of microbiome surveillance.
As microbiomes increasingly become targets for therapeutic modulation through fecal transplants, prebiotics, or engineered probiotics, having near-complete genomes of diverse gut bacteria is paramount. HRGM2 lays foundational work to better understand which microbial strains carry beneficial or harmful metabolic traits, guiding the design of microbiota-based precision medicine interventions.
This catalogue thus represents a watershed moment in gut microbiome science, enabling deep, mechanistic insights into microbial ecology and host interactions with far-reaching implications for human health. As studies continue to decipher the microbiome’s role in diseases ranging from cancer to neurological disorders, the HRGM2 resource will serve as a cornerstone on which more robust, globally representative, and functionally meaningful research can be built.
With its unprecedented scale, quality, and geographical breadth, the Human Reference Gut Microbiome version 2 marks a paradigm shift in microbial genomics. It moves the field beyond descriptive catalogs toward predictive and functional models of microbiota-host interplay. This powerful tool is set to accelerate discoveries into the molecular underpinnings of human gut health and disease, heralding a new era of data-driven, microbiome-centered biomedicine.
As microbial researchers around the world gain access to HRGM2, the landscape of gut microbiome analysis will be transformed. The resource’s ability to unravel the complexities of microbial communities with precision will facilitate breakthroughs, from identifying novel microbial therapeutics to understanding how lifestyle factors modulate metabolic networks in the gut. Ultimately, this catalogue will enable the scientific community to unlock the full diagnostic and therapeutic potential of the human microbiome, fueling innovations that may improve billions of lives worldwide.
Subject of Research: Human gut microbiome genomic catalogues and genome-scale metabolic modeling
Article Title: A human gut metagenome-assembled genome catalogue spanning 41 countries supports genome-scale metabolic models.
Article References:
Ma, J., Kim, N., Cha, J.H. et al. A human gut metagenome-assembled genome catalogue spanning 41 countries supports genome-scale metabolic models. Nat Microbiol (2025). https://doi.org/10.1038/s41564-025-02206-1
