Highlights from NCM Singapore
At NCM Singapore, we heard how researchers around the world are utilising nanopore sequencing technology in applications spanning from microbiology to clinical research. Here, we dive into three great presentations from a day packed with the latest research from the Nanopore Community.
Long cell-free DNA: the beginning of a new era in liquid biopsy
Stephanie Yu (The Chinese University of Hong Kong, China), began her talk by explaining that cell-free DNA (cfDNA) is released into the circulation upon cell death. During pregnancy or cancer, specific tissues release additional cfDNA molecules which can be used in non-invasive, liquid biopsy approaches to detect and monitor prenatal and cancer conditions.
Traditionally, short-read sequencing (SRS) has been used to analyse cfDNA; however, this method is limited due to both fragmentation and PCR procedures, leading Stephanie and her group to suggest that information from SRS ‘represents only the tip of the iceberg’. Initial experiments showed that ‘with short-read sequencing, we can hardly see any DNA molecules that are longer than 500 bp’. Using long sequencing reads, they reported cfDNA fragments up to 30 kb; Stephanie therefore suggested that short-read sequencing tools have become an ‘obstacle to the exploration of long cfDNA’. In contrast, nanopore sequencing of native DNA can be used to sequence any length of fragment.
‘Seeing the previously unseen’
Stephanie highlighted that one of the advantages of long sequencing reads is that ‘more genetic information’ can be obtained. Her group successfully detected long cfDNA of both foetal and maternal origin in prenatal research samples, and required fewer long cfDNA molecules to attain the same coverage of the foetal genome than when they used SRS. They are now using long sequencing reads to develop non-invasive techniques that could demonstrate potential to monitor for pregnancies affected by pre-eclampsia, which they found presented with reduced levels of long cfDNA.
Stephanie also highlighted that each long cfDNA molecule contains ‘more epigenetic information’ than short cfDNA molecules. Using SRS, epigenetic information is erased and is instead inferred using additional steps such as bisulfite treatment. Nanopore sequencing does not require PCR — the methylation status of CpG sites can be obtained from single plasma cfDNA molecules with no additional sample prep. The team developed an approach to determine tissue of origin by differentiating between foetal- and maternal-derived cfDNA based on methylation patterns, which could help in the development of non-invasive detection methods for maternally inherited mutations, such as Fragile X syndrome.
Stephanie and her team also assessed the potential of using cfDNA tissue-of-origin methylation patterns to detect cancer. Since the lengths of cfDNA molecules were comparable in length between hepatocellular carcinoma, hepatitis B, and healthy plasma research samples, the group instead developed a methylation scoring system which could potentially distinguish between research samples. Concluding her talk, Stephanie highlighted that nanopore sequencing provided a ‘2–4x higher number of fragments that are eligible for the tissue of origin analysis’ compared with an alternative long read sequencing technology and suggested that long cfDNA analysis ‘has potential clinical utilities in prenatal and cancer testing’.
Redefining telomere-to-telomere genome assembly strategy using the Oxford Nanopore platform
Jianjun Liu (Genome Institute of Singapore, Singapore) introduced his plenary talk by describing how human genome assembly has developed over time: from the release of the first draft of the human genome sequence in 2003 to the first telomere-to-telomere (T2T) assembly nearly two decades later, towards a pangenomics approach to represent the genetic diversity of human populations across the globe. He described how T2T assemblies require three ‘essential’ data types: high-quality long reads, ultra-long reads, and long-range information such as from chromatin conformation capture, or parental information via trio sequencing. In a pilot study, Jianjun and his team asked: can T2T assemblies be generated using only the Oxford Nanopore platform?
The team selected an Indian trio (I002) of genomic research samples from the Asian Reference Genome project, selecting the sample from the male child for assembly. They generated three types of nanopore sequencing data: 40x high single-molecule accuracy duplex reads (≥ 10 kb), 20x ultra-long reads (≥ 100 kb), and 30x Pore-C reads — a method combining chromatin conformation capture with long nanopore reads. Each dataset was downsampled from larger outputs. Assemblies were produced using two algorithms, Verkko and Hifiasm. Jianjun noted that the quality of the duplex reads was ‘very similar’ to that of high-fidelity reads from an alternative long-read technology.
'Haplotype-resolved high-quality telomere-to-telomere genomes'
Jianjun shared the results of each haplotype-phased assembly, noting that the genome lengths were similar to that of the CHM13 reference genome: Hifiasm produced slightly longer assemblies, spanning 2.95 Gb and 3.06 Gb for the two halpotypes. Comparing the algorithms, he noted that Verkko produced longer contigs, with contig N50s of 134 Mb and 139 Mb compared with 88 Mb and 111 Mb, but that Hifiasm was significantly faster, requiring 12 hours versus 191 hours, making it a good option for large-scale, pangenome-level analysis. Critically, he illustrated how ‘for many chromosomes, we’ve been able to achieve T2T’. He emphasised the importance of Pore-C data in phasing data into haplotypes, with parental data enabling a truly haplotype-resolved, chromosome-level assembly.
Summarising their work, Jianjun concluded that ‘constructing haplotype-resolved high-quality telomere-to-telomere genomes is achievable exclusively through ONT datasets’.
Application and history of nanopore sequencing in helping China's CDC system
Ji Wang (Institute of Virus Prevention and Control, China) began his plenary talk by explaining the CDC laboratory set-up across the country. There are over 3,000 CDC laboratories across China; approximately 30% have sequencing capabilities for coronaviruses. A survey across these laboratories highlighted the popularity of high-throughput sequencing and gene sequencing, as well as improved sequencing capabilities of influenza, SARS-CoV-2, and respiratory viruses.
Ji described how the ability to generate long reads in real time is of higher importance than targeted sequencing to the China CDC, demonstrating the potential to provide rapid sample to answer times. In 2019, two Chinese manganese minors died of unknown causes in South America. Whole-genome nanopore sequencing was used to sequence lung tissue clinical research samples and the data was analysed in real time using the What’s In My Pot? (WIMP) EPI2ME workflow, which characterised the pathogen Histoplasma capsulatum in high abundance across the samples, suggesting it as the cause of death. Ji noted that traditional targeted sequencing methods would not be appropriate for this application as the pathogen was not known in advance.
Ji went on to describe the importance of platform portability to the China CDC; in West Africa in 2015 during an Ebola outbreak, the China CDC experienced difficulty setting up a sequencing laboratory due to the large, heavy sequencing platform they were using, as well as limited space and unreliable power source. In contrast, a laboratory in the USA set up a sequencing facility in Sierra Leone at the same time using Oxford Nanopore MinION devices with minimal issues, due to the portability and low power consumption of the devices. Later, in 2020, the portability of the MinION enabled Ji to take the nanopore sequencing devices to CDC laboratories across the Xinjiang province in China during the COVID-19 outbreak, to teach others how to sequence and perform genome assembly to improve outbreak tracing through genomic epidemiology.
‘The whole picture’
Ji further highlighted the importance real-time sequencing for variant tracing during outbreaks, as well as to identify origin of transmission. Using nanopore sequencing, he was able to sequence SARS-Cov-2 research samples from two cold chain cargo handlers and identify that the source of the outbreak came from contaminated cargo, as the strains of viruses were of parent and child relation.
Ji concluded his talk by highlighting that generating long nanopore reads in real time help he and his team to ‘draw the whole picture of [an] outbreak’ by enabling the study of strain linkages and phylogenetics with whole-genome sequencing to identify dominant strains of a virus in an outbreak.
Attend the next Nanopore Community Meeting! Join us in Houston on 6th December to hear from a fantastic lineup of speakers who are using nanopore technology to gain new insights across a broad range of research areas.
Can’t make the event in person? The NCM Houston will also be streamed online if you’re not able to attend in person.