In the rapidly evolving field of metagenomics, the quest to reconstruct complete microbial genomes from complex environmental samples has long been challenged by the intricacies of microbial communities and technological limitations. Traditional sequencing technologies often fell short in producing contiguous assemblies, particularly when faced with highly diverse populations containing closely related strains. However, a groundbreaking study published in Nature Biotechnology in 2026 unveils myloasm, a novel metagenome assembler tailored for the latest generation of long-read sequencing data, including PacBio HiFi and Oxford Nanopore Technologies (ONT) R10.4 reads. This new tool promises to revolutionize metagenomic assembly workflows by substantially improving the completeness and resolution of assembled genomes.
Long-read sequencing technologies like PacBio HiFi and ONT R10.4 have transformed genomics by generating extensive read lengths with high base accuracy, providing greater context to resolve repetitive regions and complex genomic architectures. Despite these advances, metagenome assembly remains a daunting task due to the inherent heterogeneity of microbial communities, where multiple strains coexist with varying abundance and sequence similarity. Conventional assemblers often struggle to disentangle these overlapping genomes, resulting in fragmented assemblies or incomplete reconstructions, a gap myloasm aims to fill with remarkable success.
At the core of myloasm’s innovation is its utilization of polymorphic k-mers to construct a high-resolution string graph that captures subtle sequence variations between closely related strains in metagenomic samples. Unlike typical k-mer approaches that collapse polymorphisms, myloasm exploits these differences to delineate strain-specific paths through the assembly graph. This strategy enables a more nuanced representation of genomic diversity, which is crucial for reconstructing individual bacterial genomes within highly similar populations. Consequently, myloasm enhances the accuracy and completeness of assemblies in complex microbiomes.
Another pivotal feature of myloasm is its novel graph simplification methodology based on differential abundance information. In metagenomes, bacterial species and strains exhibit diverse abundance profiles, offering a valuable cue to disentangle intersecting assembly paths. By leveraging these abundance gradients, myloasm prunes assembly graphs more intelligently, effectively separating genomes that share common sequences but occur at different frequencies. This abundance-aware graph processing marks a significant departure from traditional assemblers, which often rely on heuristic graph cleanup routines that may inadvertently merge or discard important strain-specific contigs.
The practical impact of myloasm’s approach is exemplified by its performance on real-world ONT metagenomes, where it assembled three times more complete circular contigs compared to the nearest competing assembler. Circular contigs represent complete bacterial chromosomes or plasmids, a gold standard for genome assembly quality. This leap in assembly completeness is particularly noteworthy given ONT’s historically higher error rates relative to HiFi reads, underscoring myloasm’s robustness and sophistication in handling noisier data while still producing high-fidelity genome reconstructions.
Myloasm’s ability to equate and even surpass PacBio HiFi assembly quality using ONT long reads holds transformative implications for metagenomics research, especially considering ONT’s relatively lower cost and faster turnaround times. A joint sequencing experiment of a gut microbiome illustrated this point vividly: myloasm applied to ONT data recovered more complete circular genomes than any assembler operating on HiFi data alone. This achievement shatters prior assumptions that PacBio HiFi is inherently superior for metagenome assembly and opens avenues for more accessible, cost-effective microbial genomics studies.
Beyond mere completeness, myloasm excels at recovering fine-scale within-species diversity, a crucial aspect for understanding microbial ecology and evolution. The tool successfully reconstructed six complete single-contig genomes of Prevotella copri, a prominent gut microbe implicated in both health and disease, from a single metagenomic sample. By distinguishing these closely related strains, myloasm enables unprecedented insights into strain-level dynamics, ecological niches, and potential functional differences that would otherwise be masked in aggregated assemblies.
Further illustrating its power, myloasm was applied to an oral microbiome dataset enriched for the elusive TM7 group, also known as Saccharibacteria. This group comprises reduced-genome bacterial species that have remained largely refractory to cultivation and high-quality assembly. Remarkably, myloasm recovered eight complete TM7 genomes with over 93% average nucleotide identity, highlighting its ability to capture previously inaccessible microbial dark matter. Such achievements are poised to fuel discoveries in microbiome research, unveiling hidden diversity and novel organisms.
The methodological breakthroughs embodied in myloasm also extend to its scalability and adaptability across diverse environments and sequencing technologies. Unlike assemblers optimized for specific datasets, myloasm’s polymorphic k-mer graph construction and abundance-based simplification strategies are broadly applicable, enabling robust performance in environments ranging from soil and marine to human-associated microbiomes. Researchers can now pursue comprehensive metagenomic investigations with improved confidence in assembly quality regardless of the underlying data platform.
Myloasm’s development reflects a paradigm shift favoring higher resolution and abundance metadata integration within assembly algorithms, advancing beyond traditional sequence-overlap frameworks. This paradigm is expected to gain traction as long-read throughput increases, and real-time metagenomic surveillance becomes more commonplace. Applying myloasm to clinical specimens, environmental monitoring, and industrial microbiomes could reveal hitherto uncharted strain diversity and evolutionary dynamics, informing therapeutic, ecological, and biotechnological applications.
The tool’s impact is accentuated by its open accessibility and user-friendly design. Myloasm is released as a comprehensive software package, allowing seamless integration into existing metagenomic analysis pipelines. Its compatibility with leading high-performance computing environments ensures that large datasets can be processed efficiently, democratizing advanced metagenome assembly for the global scientific community rather than confining it to specialized centers.
Looking ahead, the principles underlying myloasm set a foundation for further innovations. Integrating additional layers of information such as methylation patterns, Hi-C contact maps, or transcriptional profiles could refine strain-resolved assemblies even further. Coupling myloasm with functional annotation and comparative genomics platforms promises a holistic view of microbial community structure and function from long-read metagenomes, transforming raw data into actionable biological understanding.
In summary, the introduction of myloasm represents a monumental advance in metagenomic science, capitalizing on the strengths of modern long reads to deliver unparalleled assembly quality. By resolving the complexity of microbial populations with exceptional resolution and harnessing abundance cues for graph simplification, myloasm pushes the frontier of what is achievable in microbiome research. As the technology disseminates, it is poised to catalyze breakthroughs in microbiology, ecology, and human health, transforming the study of microbial life in unprecedented ways.
Subject of Research:
Metagenome assembly from modern long-read sequencing data
Article Title:
High-resolution metagenome assembly for modern long reads with myloasm
Article References:
Shaw, J., Marin, M.G. & Li, H. High-resolution metagenome assembly for modern long reads with myloasm. Nat Biotechnol (2026). https://doi.org/10.1038/s41587-026-03053-z
Image Credits: AI Generated
DOI: https://doi.org/10.1038/s41587-026-03053-z
Keywords:
Metagenomics, long-read sequencing, PacBio HiFi, Oxford Nanopore, metagenome assembly, strain-resolved genomes, microbial diversity, bioinformatics, polymorphic k-mers, abundance information, microbiome
