WYMM Tour: Hamburg

Parkinson’s disease (PD) is the fastest growing neurological disorder. Accurate detection of pathogenic genetic variants is pertinent to enabling genetic diagnosis, prognosis, treatment and facilitating genetic counselling for patients with PD and related disorders. Our group utilizes long-read sequencing to detect structural variants, create a relevant PD panel, and repeat expansions. This talk will leverage specific examples to showcase the utility of long-read technology in these disorders.

Molecular brain tumor diagnosis is usually dependent on tissue biopsies or resections. This can pose several risks associated with anesthesia or neurosurgery, especially for lesions in the brain stem or other difficult-to-reach anatomical sites. Apart from initial diagnosis, tumor progression, recurrence, or the acquisition of novel genetic alterations can only be proven by re-biopsies. We employed nanopore sequencing on minimal amounts of cell-free DNA (cfDNA) from cerebrospinal fluid (CSF) and analyzed copy number variations (CNV) and global DNA methylation using a random forest classifier. We sequenced >400 CSF samples encompassing >25 brain tumor entities as well as non-neoplastic lesions. Results were compared to clinical diagnosis and molecular analysis of tumor tissue, if available. Among the pre-operative CSF samples with subsequent tissue biopsy, roughly half of them predicted the tissue biopsy result by CNV and/or methylation profiling, whereas the other half did not show detectable circulating tumor DNA (ctDNA) despite subsequent tumor proof. Notably, we did neither achieve any false-positive results nor did the classifier predict the wrong tumor entity in any case. Only a small minority of samples with detected ctDNA contained tumor cells detectable through microscopy. Our results suggest that nanopore sequencing data of cfDNA from CSF samples may be a promising approach for initial brain tumor diagnostics and an important tool for disease monitoring.

The World Health Organization has declared antibiotic resistance one of the ten most severe global health threats, with resistant infections leading to higher mortality and morbidity. Real-time genomic technology has the potential of speeding up pathogen and antibiotic resistance profiling in the clinical setting. This is especially relevant for infections with carbapenemase-producing Enterobacterales, which present a significant resistance burden due to their ability to hydrolyze a variety of antibiotics. Here, we directly compared pathogen and resistance predictions between current clinically established and real-time genomic approaches, using carbapenem-resistant Enterobacteriales infections at the Technical University of Munich Rechts der Isar hospital from 2023 as an example. Our preliminary results confirm high concurrence between both approaches while highlighting the advantages of real-time genomics in terms of speed and additional relevant infection control information through plasmid detection and annotation. We further use a specific infection case of a complex Klebsiella pneumoniae infection with resistance to the last-resort antibiotic CAZ-AVI as an example where genomic technology could have surpassed the clinically established approaches in terms of accuracy and sensitivity of antibiotic resistance prediction: In this case, originally low-frequency plasmids harboring CAZ-AVI resistance genes could not be detected by established approaches, leading to the administration of antibiotics which promoted selection of CAZ-AVI resistance; we show that genomic technology could have rapidly identified this “hidden” antibiotic resistance and directly informed clinical management in the case of real-time deployment in the hospital setting. In summary, we show the power of real-time genomic technology to accurately identify complex antibiotic resistance patterns, potentially impacting clinical decision-making and patient outcome, and discuss consequences for antibiotic stewardship.

Bacteriophages are highly abundant members of the microbial community in the human gut with crucial roles for community dynamics, human health, and a potential reservoir for undiscovered molecular tools. As most gut phages are predicted to be integrated into the genomes of their bacterial host, their dynamics and insights into phage-host relationship are of particular interest. However, studies based on short-read metagenomic sequencing often fail to fully assemble phage genomes or their host context. Here, we used long-read metagenomic sequencing on 12 longitudinal stool samples from 6 healthy individuals, spanning a 2-year timescale, to investigate prophage integration and phage-host dynamics. As expected, long-read sequencing results in the detection of more integrated prophages than short-read sequencing, thereby revealing their bacterial hosts. Using the longitudinal aspect of our data, we observe both active and inactive phages over time, stressing the importance of host context for the study of phage dynamics. For active phages, read mapping across timepoints can determine their genome boundaries with base-pair accuracy. Taken together, we demonstrate how long-read longitudinal metagenomic sequencing can be used to elucidate prophage integration, phage-host relationships, and phage dynamics in the human gut.

Critically ill pediatric samples often originate from individuals with underlying genetic conditions, necessitating rapid genomic analysis to inform urgent research decisions. Traditional genomic workflows generally take weeks and may involve multiple assays. Nanopore long-read genome sequencing (LR-GS) provides genome-wide results within days as a streamlined, comprehensive experimental workflow. As one of the first centers in Europe, we are implementing ultrarapid LR-GS for research on critically ill cases.

We enrolled 27 samples from a research cohort of critically ill individuals (mean age 3 months) in the intensive care unit to perform (ultra)rapid nanopore LR-GS alongside standard genomic research workflows. We compared result yield, turnaround time (TAT), and assessed the impact on research-driven decision-making. In 12/27 experiments, a relevant genetic variant was identified using (ultra)rapid LR-GS.

From DNA isolation, the average TAT was 4.4 days (range 1.3–7.2) for LR-GS compared to 19.0 days (range 2–42) for standard workflows. DNA methylation analysis derived from LR-GS data contributed to expedited variant detection in 3/27 samples. In 8/12 cases where a variant was identified, ultrarapid LR-GS enabled immediate research-informed changes in the care strategy, such as a medication switch or de-escalation of intervention.

Our findings demonstrate the potential utility of ultrarapid nanopore LR-GS for characterising genetic contributions in critically ill research samples, including the added value of integrated methylation analysis. These results highlight the promise of this approach and the challenges to be addressed for future integration into routine genomic research workflows.

Oxford Nanopore long-read sequencing technology is creating new opportunities for exploring the genome in greater detail. In this talk, I will present research insights from the implementation of whole genome sequencing and adaptive sampling workflows at the genomic sequencing core facility at Aarhus University Hospital. These approaches are being used to investigate complex genomic features across a range of conditions, including rare diseases and cancer. Our work aims to support a deeper understanding and evaluate the potential of long-read sequencing for future clinical applications.