NEW YORK – Oxford Nanopore Technologies and researchers at four German university medical centers have launched an initiative to evaluate the clinical implementation and utility of nanopore sequencing for rare disease testing.
Named the Clinical Long-read Genome Initiative (lonGER), the two-year study will analyze approximately 1,000 samples from clinical sites across Germany, potentially paving the way for the clinical adoption of nanopore sequencing on a national scale.
“We think that short-read [whole-genome sequencing] still leaves quite a lot of layers of information out,” said Tobias Haack, deputy director of the Institute of Medical Genetics and Applied Genomics at the University of Tübingen and one of the investigators of the project.
According to Haack, the goal of lonGER is to first explore the added diagnostic value of nanopore sequencing compared to short-read whole-genome sequencing. Subsequently, the project hopes to prove that the technology can be implemented in clinical settings with a desirable turnaround time and cost.
In addition to the University of Tübingen, Charité – Universitätsmedizin Berlin, Medical School Hannover, and University Hospital RWTH Aachen are also involved in the project.
Florian Kraft, a group leader for long-read sequencing at University Hospital RWTH Aachen who is involved with the project, said lonGER is partially funded by Oxford Nanopore, which is providing sequencing flow cells and reagents, with additional support coming from the four participating centers for personnel, analytical, and other costs.
Prior to the official project start in June, the collaborators conducted a pilot study where each of the four centers sequenced 20 test samples — including three Genome in a Bottle samples and 17 patient samples — on their Oxford Nanopore platforms, Kraft said. The results showed that nanopore sequencing was able to help detect difficult variants, such as repeat expansions, complex structural variants, and variants in duplicated genes, that were previously missed by short-read sequencing, he noted.
“Based on that, we then decided to go on to the full 1,000 samples that will be done in the study over two years,” said Stephan Ossowski, a professor of genome informatics at the University of Tübingen who is another collaborator in the study.
According to Ossowski, the samples to be analyzed in lonGER will include pediatric and adult patients with unsolved rare diseases, who will be recruited from clinical sites across Germany. There are two major enrollment criteria for the study: Participants need to consent to the data sharing scheme of the study, and samples need to have matching short-read whole-genome data for benchmarking and analysis.
The samples will then be divided evenly among the four participating centers, which will each analyze about 250 using the Oxford Nanopore PromethIon P2 or P24 devices.
The two-year lonGER program is expected to proceed in two stages. “In the first year, it's more about generating the data, putting all the pipelines in place, benchmarking, and preparing diagnostic translation,” Haack noted. During the second half of the project, the group will work on demonstrating the clinical utility of nanopore sequencing.
More specifically, the researchers initially plan to carry out a retrospective study of 100 samples from patients with neurodevelopmental disorders and their parents and about 50 samples from patients with neurological or imprinting diseases. They also seek to establish a standardized wet-lab workflow, which aims to sequence one sample per Oxford Nanopore R10 flow cell using the kit 14 chemistry.
During their earlier pilot run, the researchers achieved an average sequencing coverage of 40X to 50X for each sample across the four centers. “We think sequencing deep for diagnostics is crucial,” Ossowski said, adding that while nanopore sequencing’s accuracy for SNP calling is already at 99.9 percent with 30X genome coverage in their hands, they see a boost in indel-calling accuracy with every 10X coverage increase.
“For indels, we still have some way to go to get it to the same quality that we had with short reads,” Ossowski pointed out. “The best solution is these duplex reads … we will have to sequence deep enough to get enough of these higher-quality duplex reads.”
On the bioinformatics side, Kraft said the team hopes to establish “a gold standard pipeline, which is tailor-made for clinical usage." To achieve that, the researchers will benchmark the pipelines developed by different centers as well as one from Oxford Nanopore to compare their performance in terms of data quality, run time, and ability to process duplex reads.
Duplex reads “allow us from our first experience to get the error rate below half a percent, which is extremely important for diagnostics, because we need high accuracy [sequencing], especially for indels, for instance, but also to find the exact breakpoints of structural variants,” Ossowski noted.
During lonGER’s second phase, the researchers plan to analyze a prospective cohort of 150 trio samples from patients with neurological disorders as well as 200 from patients with other rare diseases to evaluate factors such as diagnostic yield, cost, and turnaround time.
Given that variant interpretation can be a foreseeable hurdle for implementing long-read sequencing clinically, the researchers also hope to establish a framework to enable clinicians to perform variant analysis using nanopore data.
“At the moment, the limitation [of nanopore-based diagnostics] is not whether we can generate the data, but the limitation is rather whether we can properly interpret the data we generate,” Haack noted. The clinicians “have to be specifically trained for long reads because they now can look into regions they have not looked into in the past.”
In the end, the team hopes to show that nanopore sequencing-based diagnostics can achieve comparable costs and turnaround times to short-read whole-genome sequencing across multiple sites while providing additional clinical insights, which would present a strong argument for insurance reimbursement.
According to Ossowski, Germany is slated to roll out a five-year pilot project under a new law that takes effect in January 2024, which “basically guarantees a rare disease patient or a cancer patient to get a genome sequence” and have public health insurance pay for it.
“Importantly, this law is technology-open, it does not say it must be this company or that company, but it says that the quality for the diagnostic level has to be reached,” he added.
With that in mind, another goal for lonGER is to receive clinical accreditation for their long-read nanopore diagnostic assay, Ossowski said, and to seize the insurance reimbursement opportunity once the pilot project starts.
While the quality of PacBio sequencing, the other commercially available long-read sequencing technology, is “still better” than that of nanopore sequencing at the moment, Ossowski said, it requires “a fairly expensive machine” that is not available at all the centers participating in the project.
In addition, an Oxford Nanopore P24 device can theoretically handle 2,500 genomes per year, he said, a throughput that PacBio sequencing currently “could not match."
“We want to roll [the test] out for routine diagnostics, so we simply need high throughput,” Ossowski said. “With ONT, we knew we can do that.”
As lonGER proceeds, the researchers also plan to release a series of publications sharing their protocols, benchmarking results, and data on the nanopore assay’s diagnostic value.
In addition, data from the project will be deposited to the German Human Genome-Phenome Archive (GHGA) and made available to other researchers in an access-controlled manner.
Furthermore, the group is happy to connect with similar initiatives worldwide to learn from each other and enlarge their background datasets.
“Overall, we have a vision that we provide a new, really transformative technology to the entire field of genetics and rare diseases, but also common diseases,” Haack said. “The vision for the next five to 10 years is to not only to generate the data backbone in Germany but also to connect to other European or [international] consortiums.”