An extensively debated concern about CRISPR-Cas9 genome editing is that unspecific guide RNA (gRNA) binding may induce off-target mutations. However, accurate prediction of CRISPR-Cas9 off-target sites and activity is challenging. We have developed a new protocol for detection of gRNA binding and Cas9 cleavage, based on Oxford Nanopore sequencing (Nano-OTS). The Nano-OTS method was assessed using a human cell line, which was first re-sequenced using long reads to get a detailed view of all on- and off-target binding regions. We then applied Nano-OTS to investigate the specificity of three different gRNAs, resulting in a set of 55 high-confidence gRNA binding sites, many of which were not reported by off-target prediction software. We are currently investigating whether in vivo off-target editing can occur at these sites.

The legume Pisum sativum was the first model organism for genetics and remains the focus of modern research concerning plant-microbe interactions. The genome of the plant was assembled using only nanopore reads to high contiguity N50 = 2.5 Mbp., genome length ~4Gbp. Using the newly assembled plant genome, we investigated the transcriptional profiles of the root nodules and of the mycorrhizal roots of the plant using cDNA sequencing with nanopore. We were able to perform genome annotation and discern transcripts and isoforms, unique for the respective symbioses, improving the overall quality of genome annotation. Nanopore sequencing proved to be a potent method for rapid assembly of large, repeat-rich plant genomes and can also be used to perform transcriptome and isoform annotation.

African Swine Fever Virus (ASFV) is emerging as a serious threat to global pork production. The disease is characterised by haemorrhagic fever leading to up to 100% mortality in domestic pigs and wild boar. There is no treatment or vaccine, and the main strategy for controlling the disease is mass culling and biosecurity perimeters. It is estimated that last year a quarter of the world’s domestic pig population either died of ASFV or were culled to contain it. The virus is a ~180kb DNA virus. Porcine reproductive and respiratory syndrome virus (PRRSV) is a panzootic virus that causes one of the most economically important pig diseases worldwide. It is characterised by respiratory distress, inappetence, and fever, leading to up to 100% mortality in pre-weaned piglets and to death and mummification in utero or abortion in pregnant sows. Treatment is currently limited to stringent biosecurity measures as vaccination only confers partial protection. The virus is a 15kb positive-strand RNA virus, and with nanopore sequencing we can sequence the entire genome in a single read. This allows us not only to track how the virus is spreading but identify quasispecies and recombination events that are missed by methods focussed on the amplification of single open reading frames. Additionally, direct RNA sequencing allows us to characterise the profile of subgenomic RNA, produced during replication, across different viruses. The Philippines are the 8th largest producer of pork in the world, yet 64% of the ~12.5 million domestic animals are farmed in backyard farms. With differing practices between backyard farms and commercial farms, it is important to understand how pathogens spread in order to implement targeted interventions. PRRSV has a huge economic impact in farms around the world and improved surveillance of this virus could reduce the losses Filipino farmers experience and improve the welfare of their animals. ASFV was only introduced to the Philippines towards the end of 2019 and continues to spread throughout the country. If transmission routes are understood, this and potential future outbreaks may be contained better. By bringing the sequencer to the samples we can analyse these viruses locally in the Philippines. Here I will present the methods we have developed to amplify and sequence the genomes of these viruses in the Philippines, including a bioinformatics pipeline that can be run on a standard laptop, to produce highly accurate sequences.

Classification of hematopoietic malignancies is based on World Health Organisation (WHO) guidelines and requires assessment of genomic aberrations. Using whole genome sequencing with Oxford Nanopore technology we routinely create high-resolution copy number variation profiles based on a 24h run using one flow cell. A complementary transcriptome sequencing allows the analysis of fusion genes in a similar timeframe. This combined genome- and transcriptome-wide Oxford Nanopore-based sequencing workflow provides the opportunity to rapidly characterize hematologic malignancies at the molecular level, which is needed for improved genotype-based treatment strategies.

The hummingbird occupies a unique place in the vertebrate world. It has the highest known metabolic rate, needed to fuel incredible energetic demands of hovering flight the bird performs daily to collect nectar from flowers. Understanding the molecular basis of such extreme physiology will provide foundational knowledge to enable rational engineering of metabolic circuits in mammalian cells. To explore how the hummingbird is able to accomplish these incredible feats, we used a combination of nanopore, and Illumina sequencing methods set out to characterize the genome and transcriptome of fasted and fed hummingbirds.

The African Orphan Crops Consortium (AOCC) is a global partnership promoting strategic, genome-enabled improvement of under-researched crops for biodiversity-based nutritious food solutions in Africa. We present current status, opportunities and examples successes of AOCC. Orphan crops like Lablab (Lablab purpureus; ~452 Mbp), African Yam bean (Sphenostylis stenocarpa; ~800 Mbp) and Moringa (Moringa oleifera; ~315 Mbp), are often of high nutritive value and are climate resilient. The Oxford Nanopore MinION was used to generate long reads of all three crops, to compliment and generate contiguous draft genome assemblies. Work is under way to improve these assemblies to chromosome-scale using Hi-C mapping.

Alternative splicing generates various RNA isoforms that form the complexity of eukaryotic transcriptomes, and mis-splicing of RNA isoforms are involved in numerous diseases including cardiovascular disease. Here, using Oxford Nanopore sequencing, we developed high-throughput experimental and computational methods that accurately quantify and compare genome-wide full-length isoform expression with 97.6% accuracy estimated by synthetic RNA isoform controls. We identified 24.1% of the known protein coding isoforms in GENCODE comprehensive annotation. An average of 3.14 isoforms per gene was observed and 15.6% of total spliced isoforms are novel. We conducted, for the first time, differential expression tests of full-length isoforms iPSC-CMs containing RBM20 WT and mutations known to be associated with dilated cardiomyopathy, and pinpointed the full-length isoforms which are mis-spliced in RBM20 mutant. Our discovery that for the mis-spliced genes, only specific but not all isoforms are differentially expressed suggests that splicing is precisely regulated at the isoform level.

Variation in transcript structure via RNA splicing and differences in the 5’ and 3’ untranslated regions is a key feature of gene regulation. Disruption of transcript structure is one of the primary functional changes behind a large proportion of disease variants, both common and rare. These discoveries have so far been partly based on analysis of RNA-sequencing data of 50-100 bp reads, which cannot directly measure the real biological unit of the transcriptome – transcripts. The advent of long read technologies for RNA-seq has the potential to transform transcriptome analysis, since it can directly measure full-length isoforms. In this study, we created a large long-read transcriptome sequencing data set across diverse samples of the GTEx project using Oxford Nanopore technology and developed a new computational approach for investigating the effects of genetic variants on the transcriptome. We generated cDNA sequencing data using the PCR-based protocol from 88 samples across 15 different tissues, for which we also had access to Illumina RNA-seq data and 73 had whole genome sequencing data. We identified novel transcripts and showed how transcript quantifications vary across tissues. We developed pipelines to perform allele specific expression (ASE) and allele specific transcript structure (ASTS) analyses genome wide. In ASTS analysis we test for differences in transcripts originating from each haplotype of a sample by splitting the reads according to the haplotype of a heterozygous variant and determining to which transcript that read had been assigned to – an analysis that is not informative with short-read Illumina data. In order to do that we first devised a new mapping strategy to mediate the high reference bias observed in Oxford Nanopore data. Across all the samples, 14,212 genes showed allelic expression in at least one sample, of which 1,230 fulfilled the requirements for ASTS analysis. We found 187 genes that had significant allelic differences in their transcript distributions in at least one sample, only half of which also showed ASE. ASTS events capture splicing quantitative trait loci that have been mapped as part of the GTEx consortium, verifying that it reflects true genetic effects on splicing. We also discovered rare splice-disrupting variants using ASTS data. Finally, we knocked-down PTBP1, an RNA binding protein that mediates splicing, using siRNA in fibroblast cell lines generated from five donors. We demonstrate the successful ablation of ASTS in some cases, indicating that genetic effects on splicing can be modified by the cellular state. Altogether, our results provide evidence of the widespread nature of genetically driven allelic differences in transcript structure, and the power of long-read data and careful computational approaches to study it in human population samples.

Despite advances in clinical genetic testing, many patients with suspected genetic diseases remain undiagnosed. Sometimes, testing reveals complex structural changes that are difficult to evaluate using standard clinical methods. Other times, a single variant is identified in suspected recessive disorders, or no variants at all in suspected dominant or X-linked disorders. Thus, there is a need for better tools to more fully understand cases with suspected but undiagnosed genetic diagnoses. We tested whether targeted long-read sequencing using Read Until on the Oxford Nanopore platform could be used to evaluate cases such as these. Using Read Until, we identified known pathogenic structural variants, resolved complex structural changes, and identified missing disease-causing variants in patient DNA.

Genomics is used widely throughout animal agriculture, but the time between sampling and knowledge acquisition is a limiting factor in some environments. For example, in the extensive beef production environment of northern Australia, animals are often only handled once or twice each year. Nanopore sequencing offers a unique solution to this problem, because of the short turnaround times that are possible. Already nanopore data has been used to characterise economically important structural variants; but evidence suggests that it could one day become a portable solution to characterise the genetic merit of animals in situ.

With genome sequencing no longer being the limiting factor, there is clear need for efficient tools that can assemble, compare and analyze pan-genomes comprising hundreds of individuals at a speed matching data production. At KeyGene, we combine state-of-the-art sequence analysis software with entirely novel algorithms to tackle the complexities of plant genomes, both at the individual level (heterozygosity, polyploidy, and high repetitiveness) and at population scale (high nucleotide diversity and large-scale structural variation). In this presentation we highlight our breakthrough algorithmic innovations in this field and provide a future perspective on specialized algorithms we are developing to integrate novel Oxford Nanopore-based technologies such as Pore-C within our pan-genomics platform.

The burden and impact of healthcare-associated COVID-19 infections is unknown. We aimed to examine the utility of rapid sequencing of SARS-CoV-2 combined with detailed epidemiological analysis to investigate healthcare-associated COVID-19 infections and to inform infection control measures. We set up rapid viral sequencing of SARS-CoV-2 from PCR-positive diagnostic samples using nanopore sequencing, enabling sample-to-sequence in less than 24 hours. We established a rapid review and reporting system with integration of genomic and epidemiological data to investigate suspected cases of healthcare-associated COVID-19. Between 13 March and 24 April 2020, we collected clinical data and samples from 5191 COVID-19 patients in the East of England. We sequenced 1000 samples, producing 747 complete viral genomes. We conducted combined epidemiological and genomic analysis of 299 patients at our hospital and identified 26 genomic clusters involving 114 patients. 66 cases (57.9%) had a strong epidemiological link and 15 cases (13.2%) had a plausible epidemiological link. These results were fed back to clinical, infection control and hospital management teams, resulting in infection control interventions and informing patient safety reporting. We established real-time genomic surveillance of SARS-CoV-2 in a UK hospital and demonstrated the benefit of combined genomic and epidemiological analysis for the investigation of healthcare-associated COVID-19 infections. This approach enabled us to detect cryptic transmission events and identify opportunities to target infection control interventions to reduce further healthcare-associated infections.

The global spread of the SARS-CoV-2 virus epidemic over the last few months has generated unprecedented social and economic disruptions in developing countries compounded by weak healthcare facilities. Kenya, a lower-middle income country, lacks adequate resources for comprehensive preparedness to counter a large-scale epidemic similar to countries with large economies. To support the country’s epidemic preparations, we set up whole genome sequencing (WGS) of respiratory viruses using portable MinION technology to feed into public health control measures. We describe our approach in building capacity, understanding and informing public health agencies in tracking the spread of SARS-CoV-2 at the Kenyan Coast and the country in general.

Alongside his Ph.D. studies at Imperial College, Glen spent his spare time self-organising an expedition to retrace a 1932 route across Europe’s largest icecap, the Vatnajökull in Iceland, with a team of 3 friends in 2019. Inspired by the 1932 team, he wanted to push the bounds of expedition science. Packing an entire lab into the back of his sledge, Glen and the team performed the first example of entirely off-grid sequencing in a polar environment, running only on human and solar power. The team hope this work will be used in the future to studies organism is far-flung remote places.  

We present how downstream nanopore data analysis can be performed on embedded systems using a computational approach to run the popular mapping software minimap2 under limited memory (known as ‘minimap2-arm’), and ‘f5c’, a heavily optimised implementation of the signal to nucleotide alignment as employed by the popular variant calling and methylation calling tool Nanopolish. Minimap2-arm can perform read alignment on devices with just 2GB of RAM, without impacting the accuracy. F5c is capable of efficiently leveraging the computing power of lightweight embedded systems equipped with GPUs, such as NVIDIA Jetson devices, and performs ~3-5× faster than the original.

Fera Science Limited is the national reference laboratory for plant health in England and Wales, providing scientific advice and services to Defra, other government bodies, and commercial customers. It is responsible for providing diagnostic services to the Plant Health and Seed Inspectorate who work at UK ports and airports protecting UK agriculture and the natural environment from the introduction of invasive plant pests and diseases. For the last 10 years these services have included an increasing use of high throughput sequencing for environmental monitoring and disease diagnosis. Due to the high cost of the existing platforms, samples need to be batched (24 to over 100 samples per batch) to make the process viable leading to regular runs of particular samples types and waits of greater than a month for all but the highest priority samples. The cost and scalability of the MinION with the Flongle adapter opens up the potential to run single or small batches of samples as soon as they are requested. This concept is being tested in a number of case study trials. Much of our work involves correctly identifying insects from larvae intercepted on imported fruit. This is carried out using morphological examination and confirmed by DNA barcoding, with the sequencing being carried out by offsite commercial suppliers using capillary sequencers. In a trial with 8 samples, within an hour the Flongle had produced enough sequence to correctly and accurately identify the insect species present. A number of problematic samples with the target infected with a second insect (parasitoid or mite) were also tested. This approach has been extended and used to assess the invertebrate populations of sticky traps looking for invasive insect pests using metabarcoding. The Flongle has also been successfully used to identify novel plant viruses in import samples (direct RNA sequencing), identify the species of a disease-causing bacteria (whole genome sequencing) and confirm infections with phytoplasma, which are obligate bacterial parasites of plants (amplicon sequencing). Finally, using the Flongle we have also been able to show that we can determine the diet of spittle bugs (metabarcoding), possible vectors of Xyella fastidiosa, a bacterial pest currently causing serious damage across Europe. Knowledge of the vectors and their hosts will be important if this disease comes to the UK. The MinION with the option of the Flongle is showing great promise delivering rapid, scalable, useful data and will soon become a standard tool defending the UK from invasive plant pests and diseases.

COVID-19 is caused by the SARS-CoV-2 coronavirus and was first reported in central China in December 2019. Extensive molecular surveillance in Guangdong, China’s most populous province, during early 2020 resulted in 1,388 reported RNA positive cases from 1.6 million tests. In order to understand the molecular epidemiology and genetic diversity of SARS-CoV-2 in China, we generated 53 genomes from infected individuals in Guangdong using a combination of metagenomic sequencing and tiling amplicon approaches. Combined epidemiological and phylogenetic analyses indicate multiple independent introductions to Guangdong, although phylogenetic clustering is uncertain due to low virus genetic variation early in the pandemic. Our results illustrate how the timing, size and duration of putative local transmission chains were constrained by national travel restrictions and by the province’s large-scale intensive surveillance and intervention measures. Despite these successes, COVID-19 surveillance in Guangdong is still required as the number of cases imported from other countries is increasing.

The paradigm that single nucleotide polymorphisms are the primary metric to judge bacterial diversity is outdated. This is particularly true for the pathogen Bordetella pertussis, which judged solely by this metric, evolves at a ‘glacial’ pace. Our research found, however, that the repeat-rich genome of Bordetella pertussis can no longer be thought of as static but instead, as plastic. Exploiting both Illumina and Oxford Nanopore sequencing, we frequently found structural variants within global cohorts of isolates and single sequence samples, respectively. Our work champions the Oxford Nanopore platform to study structural variations both in Bordetella pertussis and in the wider bacterial kingdom.

To uncover the sources of SARS-CoV-2 introductions and patterns of spread within the U.S., we used the ARTIC Network pipeline on the Oxford Nanopore MinION to sequence nine viral genomes from first reported COVID-19 cases in Connecticut. Our phylogenetic analysis places most of these genomes with viruses sequenced from Washington state. Coupling our genomic data with domestic and international flight patterns, we show that early SARS-CoV-2 transmission in Connecticut was likely driven by domestic introductions. This study provides evidence for widespread, sustained transmission of SARS-CoV-2 within the U.S. and highlights the critical need for local surveillance.

The ability to identify and quantify the constituent pollen plant species that a pollinator forage for resources has essential applications in ecology, conservation, and agriculture. Particularly in fruit trees, where cross-pollination within cultivars is required to ensure fruit and seed development. Pollen metabarcoding protocols have opened the frontier to identify constituent plant pollen species. Regardless to this breakthrough, there are methodological reasons that inhibit identifying pollen grains to the cultivar level, quantify their relative abundances, and allow the detection of other organisms associated with the pollen sample. A four apple pollen cultivars library was developed, pollen cultivar mixes and pollen from foraging honey bees in an apple orchard were sequenced with MinION. The pollen cultivars were genotyped, and an approach to estimate their relative abundances was tested. In addition, a metagenome scan was performed to analyse the microbiome associated with pollen. This methodology could be applied for future crop adaptation, to monitor pollinator's effectiveness, and pathogen transmission by pollinators at agro-ecological systems.

This research aimed to detect Mycobacterium tuberculosis (M.tb) from sputum DNA using  nanopore sequencing for the prediction of strain lineage and drug susceptibility (DST). METHODS: DNA was isolated from 10 randomly selected sputum samples from pulmonary tuberculosis (TB) patients and sequenced using MinION. Sequence data was analysed using Guppy (v3.4.4), Mykrobe Predictor (v0.7.0), and WIMP. RESULTS: On average 100,000-18,000,000 reads were generated per sputum sample. Strain lineage and DST were predicted with less than 0.2% M.tb DNA in sputum. CONCLUSION: Nanopore sequencing of sputum DNA can be used for diagnosing pulmonary TB and for prediction of strain lineage and DST from sputum.

Due to its large genome size (4Gb) and highly repetitive nature, previous short-read assemblies of cultivated lentil (Lens culinaris) were highly collapsed and fragmented. Only 2.7Gb could be assembled (scaffold N50 54Kb), and 1.1Gb anchored, despite using a combination of long mate pair data, chromosome flow sorting, consensus genetic map, and an optical map. Long-read technology has resulted in dramatic improvements, with 3.7Gb (contig N50 1.4Mb) now assembled from 50-fold coverage of the genome. The updated genome assembly uncovers an additional 25% of the lentil genome with 121x increase in sequence contiguity. A HiC contact map provided final anchoring and orientation for 92% of the long-read contigs. Using the same process, we have also assembled de novo several wild relatives of interest to the lentil breeding program. Higher assembly contiguity allows us to fully characterize structural rearrangements previously seen only in interspecific genetic maps as well as identify novel structural variation relative to cultivated species. Improved assembly of highly repetitive regions and intact TE elements also allows for a closer look at genomic expansion within the genus Lens. 

Nanopore sequencers enable selective sequencing of single molecules in real time by individually reversing the voltage across specific nanopores. Thus, DNA molecules can be rejected and replaced with new molecules enabling targeted sequencing to enrich, deplete or achieve specific coverage in a set of reads to address a biological question. We previously demonstrated this method worked using dynamic time warping mapping signal to reference but required significant compute and did not scale to gigabase references. Using direct base calling with GPU we can now scale to gigabase references. We enrich for specific chromosomes, mapping against the human genome and we develop pipelines enriching low abundance organisms from mixed populations without prior knowledge of sample composition. Finally, we enrich panels including 25,600 exon targets from 10,000 human genes and 717 genes implicated in cancer. Using this approach, we identify PML-RARA fusions in the NB4 cell line in under 15 hours sequencing. These methods can be used to efficiently screen any target panel of genes without specialised sample preparation using a single computer and suitably powerful GPU.

Pathogenic hexanucleotide repeats (GGGGCC) in the C9orf72 gene are the leading cause of familial amyotrophic lateral sclerosis (ALS) and frontotemporal dementia (FTD) explaining 34.2% and 25.9% of cases, respectively. Quantification of the true length of the repeat has relied on Southern blotting and, clinically, a PCR assay is used to determine a patient's repeat status, but this cannot resolve the number of repeat units. CRISPR/Cas9 targeting and subsequent nanopore sequencing of C9orf72 enriched for reads that cover the repeat region allowing accurate quantification of the number of repeats. Simultaneously we were able to determine the methylation status of the locus in addition to clues about its sequence composition.

Oxford Nanopores sequencing technology has enabled us to generate high quality contig assemblies in the megabase contig N50 range. After polishing these genomes with the original nanopore reads using tools like Medaka and Racon, we are now employing different methods like Hi-C, pore-C and optical mapping to turn those assemblies into chromosome scale scaffolded assemblies. The combination of nanopore long-read assemblies and Hi-C even allowed us to generate haplotype phased chromosome assemblies. Using Oxford Nanopore Technologies new Chromosome conformation capture technique pore-C, we were also able to generate chromosome scale assemblies using nothing, but sequencing data generated on Oxford Nanopore-platforms even for larger genomes.

Whole-genome sequencing (WGS) techniques play a critical role in investigating structural variations in the genome to understand the genetic aspect of a disease. In autism spectrum disorders (ASD), studies typically concentrate on investigating pathogenic variations contributing to a proband’s condition in the protein-coding region, while a substantially larger non-coding region is neglected. Moreover, current short-read sequencing tools are incapable of detecting large and complex genomic variations. In this talk, I will present our results on the detection of structural variation in a family with triplets impacted by ASD. Comparing with other WGS technology, nanopore sequencing identified significantly large number of SVs in non-coding region of the genome.

Watch here

Novel immunotherapies directed against B-cell restricted surface antigens CD19 and CD22 are used successfully for the treatment of relapsed acute lymphoblastic leukemias (B-ALL), but acquired resistance is emerging as a major barrier. Our previous studies indicated that alternative splicing of CD19 could contribute to epitope loss and treatment failures. Here we utilized the Oxford Nanopore Technologies platform to determine complete exon structures of CD19 and CD22 transcripts in B-ALL cell lines and patient-derived xenografts. We discovered that many (but not all of them) encode functional proteins capable of conferring resistance to immunotherapy such as CAR T-cells.

Mammals and their microbiome constitute complex holobiont organisms. The interaction between their genomes and the environment modulate the host phenotype. There is evidence that the host genome exerts some control over the microbiota composition. This convoluted system poses a challenge in determining the causal source of certain phenotypes (e.g. diabetes, BMI, feed efficiency). Structural equation models (SEM) are useful to disentangle the causal effect of the hologenome on the phenotypic expression. Particularly, we apply SEM to investigate the causal genomic and metagenomic forces that contribute to the variability in methane emissions in cows to accomplish with the Climate Change’s Paris Agreement.

Grass pea (Lathyrus sativus) is a highly nutritious legume crop with outstanding resilience to both drought and flooding. The relative lack of genetic and genomic resources for this orphan crop has stifled efforts to develop safe and highly productive varieties. Using Illumina HiSeq and Oxford Nanopore PromethION data we have assembled drafts of the 6.3 Gbp grass pea genome — the largest legume genome sequenced to date. Long-read sequencing is helping us to overcome the challenge of this genome’s highly repetitive nature, particularly of large Ogre-type repeats, which limit the usefulness of short-read-based scaffolding. Our new genome and transcriptome sequences are opening the path to nutritionally improved, resilient grass peas for Sub-Saharan Africa and South Asia.

RNA modifications such as m6A have been found to contribute to molecular functions of RNAs. Third generation sequencing using the Oxford Nanopore technology enables the simultaneous profiling of transcript expression and modification through direct sequencing of native RNA. However, quantification of differences in RNA modifications has been challenging. Here, we present a computational method, xPore, to identify differential RNA modifications transcriptome-wide from direct RNA sequencing data. For each single site, we infer modification rates by fitting a multi-sample two-Gaussian mixture model, corresponding to unmodified and modified RNA. We demonstrate that RNA modifications can be quantitatively identified from direct RNA-sequencing data with high accuracy, opening many new opportunities for large-scale applications in genomics and biomedical research. xPore is available at https://github.com/GoekeLab/xpore.

Neuropsychiatric disorders have a strong genetic component and hundreds of risk genes have been identified. How these genes contribute to disease risk is not well understood. Nanopore amplicon sequencing was used to examine the entire coding regions of two high-confidence risk genes in seven regions of post-mortem human brain from five control individuals. Pilot sequencing also identified 78 novel exons in 13 risk genes. Three novel isoforms were shown to contribute to over 25% of MAPT expression. Isoform abundance differences between cerebellum and cortical regions were also identified. Our study demonstrates the power of nanopore sequencing to characterise full-length isoform diversity in human brain.

High-throughput amplicon sequencing is a powerful method for analysing variation in defined genetic regions. The method is therefore ideal for studying genetic populations with low abundant variants or high heterogeneity, such as cancer driver genes, virus populations and microbial communities, among others. Current short-read technologies are often employed for amplicon sequencing due to their high accuracy (~99.9%); but are incapable of generating amplicons over 500-2000 bp in length, which limits long-range information and assay resolution. Here, we describe a new approach to generate long amplicons with the Oxford Nanopore Technologies MinION that exceeds the accuracy of current short-read technologies while generating amplicons up to 10,000 bp in length. Our new method utilizes unique molecular identifiers (UMIs) annealed to each end of the template molecule prior to amplification and sequencing. A downstream bioinformatics approach was developed to bin reads based on their dual UMIs, filter chimeras, and generate polished consensus sequences. We demonstrated this new approach by generating over 10,000 amplicon consensus sequences from full-length ribosomal RNA (rRNA) operons of a mock microbial community (average 4500 bp in length) using both R9.4 and R10.3 flow cells. The average residual error rate in the amplicon sequences was 0.01%, with essentially no detectable chimeras. We anticipate the simplicity and low cost of this method to revolutionize virtually any amplicon sequencing application by allowing for full gene or operons to be accurately sequenced.

Effective analysis of bacterial communities requires a coordinated attack on two problems: DNA extraction and genome assembly. We demonstrate the effectiveness of the “three peaks” extraction method developed by Josh Quick, which combines chemical, enzymatic and physical lysis to achieve long-reads without loss of representation. We also introduce Reticulatus: a Snakemake-based assembly and polishing pipeline that attempts to codify current best practice for long-read metagenomics. We have applied our novel DNA extraction method to the previously characterised Zymo Mock Community Standard; a more realistic Gut Microbiome Mock and on a real faecal microbiota transplant. We use Reticulatus to assemble, polish and validate our long-read metagenomic assemblies and demonstrate the current state-of-the-art for long-read nanopore metagenomics.

The gut microbiome is an essential neuromodulator of brain-gut axis signaling, which can affect brain inflammation and pathology. Our central hypothesis is that gut microbiome of Alzheimer´s (AD) mice is linked to detrimental outcomes after brain damage. We utilized fecal microbiota transplants from AD mice and healthy young controls. We then relied on Oxford Nanopore MinION full-length 16S rRNA sequencing to analyze genus and species level dynamics in the fecal microbiome and analyze the associated neuropathology. While mice after fecal transplants had similar bacterial diversity as donors, the mice transplanted with young microbiota presented reduced inflammation and motor disfunction.

Cystic fibrosis (CF) is a life-shortening genetic disorder leading, particularly, to chronic lung disease. We studied airway microbiome at multiple timepoints in sputa of 6 patients subsequent to antibiotic intervention using both short (Illumina) and long (Oxford Nanopore) reads, and both WGS and 16S rRNA gene sequencing. Using sketchy, we were able to rapidly and accurately identify specific strains and resistant profiles of S. aureus infections from raw nanopore reads. By combining short and long-read sequencing data, we were able to generate high-quality assemblies of the most abundant species. This study demonstrates the potential for nanopore sequencing to be used in routine monitoring of infection dynamics for CF patients.

Nanopore technology allows mutation and methylation detection directly from native DNA, without the need for the bisulfite treatment and PCR. Combining a CRISPR-Cas9 system to nanopore technology, we simultaneously assessed the mutations and methylation status of the major glioma molecular markers, IDH1, IDH2, and MGMT promoter. We applied this method on well-characterized DNA standards, 4 cell lines and 4 brain tumor samples. In this presentation, we will show the results of mutation and methylation status assessment from both nanopore and conventional methods. These efforts are in line with improving precision medicine and can be applied to other cancer types.

DNA methylation plays an essential role in governing the architecture of gene expression across eukaryotes. In plants, DNA methylation not only protects the genome from transposable element (TE) jumping by repressing their expression, it also mediates environmental-specific expression responses. Here we use direct DNA methylation detection coupled to full-length cDNA sequencing across several Cannabis chemotypes (high and low THC/CBD) to understand the final step in the cannabinoid pathway. The final step of the cannabinoid pathway is the Tetrahydrocannabinol acid synthase (THCAS) and/or Cannabidiol acid synthase (CBDAS), which are intron-less genes that are nested in 50-80 kb TE units tandemly duplicated at several loci in the genome. Due to the complexity of the THCAS/CBDAS loci and the fact that they are highly similar, full length cDNA sequencing was required to resolve which synthases were expressed in different chemotype genomes. While 10-15 copies of the THCAS/CBDAS are found across high-quality genome assemblies, only one synthase is expressed in chemotype specific plants, and only one THCAS and one CBDAS is expressed in dual chemotype plants. The expression differences across the synthases cannot be explained by the promoter regions, which are completely conserved across tandem repeats. Once again due to the complexity of the synthase loci, short-read based bisulfite sequencing fails to distinguish the DNA methylation patterns at these loci. However, direct DNA methylation enables the characterization of DNA patterns across these complex loci, providing evidence that DNA methylation in part explains the observed expression patterns. These results have implications for the modulation of THC/CBD content and the application of full-length cDNA sequencing and direct DNA methylation detection in complex plant genomes.

The recent advent of third-generation sequencing technologies brings promise for better characterization of genomic structural variants by virtue of having longer reads. However, long-read applications are still constrained by their high sequencing error rates and low sequencing throughput. Here, we present NanoVar, an optimized structural variant caller utilizing low-depth (8X) whole-genome sequencing data generated by Oxford Nanopore Technologies. NanoVar exhibits higher structural variant calling accuracy when benchmarked against current tools using low-depth simulated datasets. In an acute myeloid leukemia patient cohort, we successfully validate structural variants characterized by NanoVar and uncover normal alternative sequences or alleles present in healthy individuals.

Disease-associated chromosomal rearrangements often have break points located within the disease gene or in its vicinity. In order to identify the disease gene, characterization of the breakpoints is a promising start. Balanced reciprocal translocations (BRTs) are difficult to map but recent next generation sequencing technology has emerged efficiently for accurate detection. Two BRT cases were characterized by both positional cloning and genome sequencing, but the long-read sequencing only helped in gene identification. This is the first BRT study delineated using nanopore sequencing. This technique can now be used for routine clinical investigations which is of great help in genotype-phenotype correlation. The applications and limitations of long-read sequencing of BRTs are discussed.

The development of long-read sequencing technologies has entirely changed the landscape of possibilities for sequencing and assembling complex plant genomes. While Pacific Biosciences SMRT sequencing has served admirably for a number of years now, Oxford Nanopore technology is highly portable and requires much less up-front investment. Concerns have arisen over nanopore error rates compared to PacBio or Illumina, but at least using current versions of flow cells, chemistry, and base-calling, we no longer find these misgivings tenable regarding the construction of a highly contiguous genome. Still, a combined approach is required, at minimum including polishing using low-error — but cheap — Illumina reads. Assemblies at the chromosome scale often require further efforts, such as HiC scaffolding — but the workflow is now democratized to the extent that any university lab should be able to generate a high-quality genome of its choice. We use case-by-case workflows for generating chromosome-scale assemblies of various-sized plant genomes. Unfortunately, no one assembly approach (e.g., De Bruijn graph or overlap-layout-consensus method) works best for all species, given the various and sundry nature of their heterozygositys, ploidy levels, and transposable element blooms (the latter two also in terms of their event ages). Despite not yet achieving a truly pipeline approach, we are satisfied with our ability to generate excellent de novo genomes on unprecedentedly low time and cost scales. We will describe several of our recent projects and the individual challenges encountered and how they were overcome.

The transcriptome is characterized by a diverse usage of alternative splicing, promoters, and poly-adenylation sites. However, the full complexity of transcriptional events remains difficult to study. Long-read RNA-Sequencing provides full-length reads of transcripts by directly sequencing RNA molecules, or cDNA molecules without amplification, potentially addressing challenges faced by short-read sequencing technology. Here we present the Singapore Nanopore-Expression project (SG-NEx), a comprehensive long-read RNA-Seq data resource of five commonly used human cancer cell lines generated using Oxford Nanopore technology for RNA, cDNA, and PCR-free cDNA sequencing. We benchmark the three protocols, compare the performance with short-read data for gene and transcript expression quantification, demonstrate how long-read RNA-Seq can be used to identify differential transcript usage, and discover novel genes and transcripts. In conclusion, we present a long-read RNA-Seq data resource that will support software development and benchmarking studies and provide important insights into the transcriptome of human cancer cell lines.

We applied on-site MinION sequencing of high-altitude permafrost in Qilian Mountain (4000m altitude) China. Long-read metagenomics is effective in revealing the microbial functions at genome-level resolution. A frame-shift correction tool FUNpore was developed. The post-correction long-reads showed encouraging precision and recall in functional prediction. The thawed permafrost soil showed higher transcriptional activity in denitrification which may lead to the discernible N2O release in summer. In contrast, methane release from permafrost thawing seems of less concern as very active aerobic methane oxidation by Methylomonas was observed in the topsoil and no methane can be detected in batch tests with elevated temperature.