WYMM Tour: Houston

Structural variants (SVs) are key contributors to genetic diversity and disease, often residing in complex genomic regions. Oxford Nanopore sequencing, with its long-read capabilities, enables robust SV detection by resolving repetitive elements and large rearrangements. Our group has advanced Sniffles, a widely used SV caller, to not only detect germline variants from single samples to populations but also to identify somatic and mosaic SVs relevant to diseases. Addressing the challenge of SV prioritization for human disease research, we extended Sniffles with a population annotation feature that leverages 1,000 Nanopore genomes, facilitating automated ranking of 90–98% of SVs per genome.

This innovation significantly enhances pathogenic variant filtering. We demonstrated its utility in Mendelian disease trios from the GREGOR cohort, identifying causative variants with improved precision. Additionally, we applied this approach to neurological disorders and cancer genomes, showcasing its versatility and effectiveness.

Sniffles, combined with Oxford Nanopore sequencing, represents a transformative tool for large-scale, accurate SV characterization and prioritization, advancing both research and clinical applications.

Our study employs long-read sequencing technology on the Oxford Nanopore PromethION platform to identify genomic variants associated with male infertility, enhancing the detection capabilities beyond what short-read sequencing achieves. We sequenced the genomes of over 100 men with different infertility issues at coverages exceeding 40X, using the T2T-CHM13v2.0 assembly for precise Y-chromosome analysis. Our methodologies allowed us to identify clinically relevant variants potentially linked to observed infertility phenotypes. Key discoveries include several patients with pathogenic variants impacting hormone levels, sperm motility, spermiogenesis, and sperm-egg interaction—such as a novel homozygous variant in the LHB gene related to azoospermia and lack of puberty, and rare pathogenic variants in the DNAH2, TDRD6, TEX14, and PRSS55 genes linked to various sperm dysfunctions. These findings underscore the utility of long-read sequencing in revealing complex genetic factors underlying male infertility, offering new insights and enhancing genetic diagnosis and treatment approaches. This research advances our understanding of the genetic architecture of infertility and supports personalized medicine initiatives in reproductive health.

The rise of multidrug-resistant (MDR) bacteria presents a global challenge to public health, necessitating advanced methodologies for their rapid characterization and outbreak detection. Oxford Nanopore Technologies (ONT) long-read sequencing has emerged as a transformative tool for overcoming the limitations of traditional short-read sequencing, providing comprehensive insights into MDR bacteria pathophysiology. Recent advances, including the introduction of ONT V14 chemistries and improved Guppy super accurate basecalling models, have significantly enhanced the accuracy and utility of long-read sequencing. Over the past year, our studies have increasingly shifted toward long-read-only applications, where short-read error correction no longer provides the critical advantage it once did. These improvements have allowed us to simplify workflows and extract more reliable data directly from ONT sequencing runs. This talk will highlight how I have personally used ONT sequencing over time to resolve complex genomic features such as structural variants, mobile genetic elements (MGEs), and plasmid architectures. I will focus on particular MGEs such as the IS26 transposase and its role in disseminating beta-lactam resistance determinants and creating progressively more resistant pathogens. Additionally, I will discuss how we utilized ONT sequencing for bacterial pathogen outbreak detection in a low-resource setting. Finally, I will explore the potential future of ONT technologies at our institution. By leveraging the unique capabilities of long-read sequencing, we aim to further optimize our approach to addressing AMR and outbreak detection challenges.

Since 2000, the microbiome of the International Space Station (ISS) has been monitored to assess risk to both crewmembers and the spacecraft through onboard culture followed by ground-based analyses. While this approach has served to provide alerts to anomalies and increase confidence in operational controls, these data are biased toward microorganisms that grow on the singular media type and ISS-available growth conditions; this leads to the false depiction of an overall lack of biodiversity. As NASA moves to exploration beyond low-Earth orbit, where routine sample return will not be possible, it is critical to implement in situ methods for monitoring. A culture-independent, nanopore sequencing-based method has been implemented onboard the ISS and for returned swabs. A detailed statistical comparison of the 20+ years of culture-based data to the culture-independent nanopore data was performed. Data from both methods similarly describe a human-occupied environment. Both data sets depict a common core microbiome across time and location, but the nanopore data describe an expanded microbiome with increased diversity of human- and environmentally-associated microbes. Additionally, through detailed analysis of the nanopore-generated data, some ISS locations are emerging as unique ecological niches, potentially resulting from environmentally-driven microbial selection. The presence of some noted taxa within these unique locations has implications for crew health, planetary protection, and controls used in future spacecraft systems. In terms of data usability, an evaluation of the NASA risk assessment process found that, regardless of the data set utilized, the risk assessment was identical. Therefore, use of the nanopore data was not found to drive more frequent remediation or alter the number of remediation events. The ability to perform in situ molecular microbial profiling is transforming how NASA assesses risk and is a critical tool towards monitoring and maintain.

Background information: Biological threat agents (bioagents) are pathogens that can pose substantial harm to human health. We have a fundamental understanding of most existing bioagents and the ongoing development of novel therapeutics has substantially enabled effective interventions over time. However, these bioagents continue to present significant public health challenges. Early detection is critical in effectively containing, mitigating, and treating a biothreat. However, commonly utilized detection methods rely on a prior knowledge of the potential biothreat vectors. Sequencing-based approaches have the capability to detect any biothreats in a single assay agnostically. Previously, the wide utilization of sequencing technologies was limited by cost and infrastructure requirements. Currently, sequencing data poses challenges in discerning endemic pathogens from intentional releases. These challenges can be addressed by understanding the background bioaerosol makeup and developing novel algorithms to identify novel and unknown pathogens in complex backgrounds. Background bioaerosol data would serve as a common repository to enable the development of computational approaches to discriminate between background aerosols and hazardous aerosols.

Purpose: The Hazard Awareness & Characterization Technology Center within the Department of Homeland Security Science and Technology Directorate is supporting the development and standardization of approaches to isolate nucleic acids and perform field-capable metagenomic sequencing of air filters. Developed standardized approaches have been used to sequence and characterize the microbial composition of over 800 high-volume air filters from over 140 locations throughout the continental United States. Background bioaerosol characterization across the United States will enable the development of computational algorithms to identify hazardous aerosols.

Methods: Starting in January 2024, 3-micron fluoropore filters from environmental aerosol sample collectors located in major metropolitan areas and at select special events throughout the United States were gathered during each of the four seasons. A logistic operational protocol has been developed to preserve the microbial content of the air filters during transport. Commercially available DNA/RNA isolation kits have been evaluated. Additional optimization was implemented in utilizing the Oxford Nanopore Technologies sequencers. Captured bioaerosols from over 800 filters across the United States are being characterized by metagenomic and metatranscriptomic sequencing. Results: Environmental bioaerosol metagenomic data has been generated, analyzed, and will be made accessible to the community. The identified naturally occurring background aerosol composition will enable the development of computational approaches to detect/define potential anomalies and novel outbreaks. The developed standard operating procedures will allow the community to continually source additional background bioaerosol data.

Conclusions: The optimized methodology and publicly available databases describing the bioaerosol composition around the nation will provide rich spatial and temporal bioaerosol data. This information will enable the development of technologies to identify anomalies and potential bioagents in aerosols.

Long-read sequencing platforms, including Oxford Nanopore, have the potential to revolutionize personalized medicine through the ability to accurately assess all clinically relevant structural variants (SVs), repeats and rearrangements, often undetectable with short read technologies. True production-scale processing for clinical reporting requires consistent high-quality data, low DNA sample input requirements (1-3ug), platform and software stability, data standards as well as competitive cost models. The Baylor College of Medicine - Human Genome Sequencing Center (BCM-HGSC) has assessed long-read data from Oxford Nanopore PromethION long-read platform by the generation of >25x whole genome sequence data. Samples were selected from the NIH All of Us Research Program (AoURP) CDRv5 (98K) short read data set and to date we have completed 1098 genomes on Oxford Nanopore platform averaging ~34x coverage/PromethION flow cell. Specifically, library and sequencing kits (ONT LSK114/ R10.4.1 kits) were optimized to consistently deliver >80Gb/PromethION flow cell with an insert size >15Kb and optimizations targeting ~25Kb inserts. Optimized protocols for library, sequencing, and analysis pipelines were established, along with performance metrics to provide assessment of the platform production readiness. Laboratory optimizations included precision size cuts using Pippin HT, controlled shearing parameters, optimization of all library steps for maximum yield and titration of instrument loading amounts to maximize yield. Key production metrics were established including mean coverage, % genome coverage at 10x, HIFI yields, mean read length, contamination, and Q30 or Q10 mapped bases. Initial accuracy for SNVs was assessed using the standard NIST control (HG002) and 97.1% and SVs we observe similar high accuracies of 92.6%. This work was part of the operational development of platforms and data for the All of Us Research Program. These methods are now applied to projects such as the Somatic Mosaicism Across Human Tissues (SMaHT) program to create a large-scale systematic catalog of somatic variation.