WYMM Tour: Cambridge, UK

Understanding how genetic variants in the non-protein-coding genome impact function remains a fundamental challenge in the genetic study of disease. These variants can influence the binding of transcription factors, which may lead to large-scale changes in chromatin state and conformation. Each variant resides within a haplotype and interpretation of variant function requires an understanding of the sequence context of the defined by the haplotype and its associated regulatory activity. We therefore require a technology that can identify the simultaneous binding of multiple transcription factors and chromatin features resolved to each haplotype. Here, we utilise the multiomic capabilities of Oxford Nanopore sequencing to resolve genotype with phased DNA methylation and open chromatin simultaneously. We employ an exogenous DNA methyltransferase to convert accessible adenines (As) to 6mA, to map chromatin accessibility at basepair resolution. We use Nanopore to resolve 6mA and 5mCG modifications to a haplotype to identify Fiber-seq Inferred Regulatory Elements (FIREs) and calculate differential regulatory activity between haplotypes. We report on our progress studying genetic defects associated with diabetes using stem cell derived pancreatic islets. For example, in human embryonic stem cells, we identify haplotype-specific regulatory regions associated with chromatin conformation (CTCF and RAD21) and pluripotency (OCT4, SOX2, and NANOG), together with an enrichment of gene promoters subject to epigenetic repression. Our findings highlight the potential of Oxford Nanopore for understanding haplotype-specific chromatin organisation to gain insights in regulatory variants in disease.

Approximately 200 critically ill infants and children in New Zealand are in high-dependency care, many suspected of having genetic conditions, requiring scalable genomic testing. We adopted an acute care genomics protocol from an accredited laboratory and established a clinical pipeline using Oxford Nanopore Technologies PromethION 2 solo system and Fabric GEM™ software. Benchmarking of the pipeline was performed using Global Alliance for Genomics and Health benchmarking tools and Genome in a Bottle samples (HG002-HG007). Evaluation of single nucleotide variants resulted in a precision and recall of 0.997 and 0.992, respectively. Small indel identification approached a precision of 0.922 and recall of 0.838. Large genomic variations from Coriell Copy Number Variation Reference Panel 1 were reliably detected with ~2 M long reads. Finally, we present results obtained from fourteen trio samples, ten of which were processed in parallel with a clinically accredited short-read rapid genomic testing pipeline (Newborn Genomics Programme; NCT06081075; 2023-10-12).

Accurate splicing annotation and quantification are critical for understanding cellular diversity and disease mechanisms. Advances in long-read sequencing and droplet-based library preparation now allow researchers to capture full-length transcripts and resolve complex splicing events at single-cell resolution. However, downstream analysis presents challenges, including low read depth, high base error rates, and pervasive transcript truncation due to reverse transcription and PCR artifacts. These limitations have an amplified impact on the accuracy of splicing quantification at single-cell level, as many existing methods rely on uniquely mapping full-length reads. To overcome these challenges, we adapted an isoform annotation pipeline originally developed for bulk RNA-seq, generating a sample-specific annotation that was then used for improved per-cell quantification using Isosceles, a recently developed tool designed for single-cell isoform quantification. Using Oxford Nanopore Technologies, we profiled three iPSC-derived terminal lineages —neurons, microglia, and astrocytes, which are used for modelling neurodevelopmental disorders. In this study, we aimed to provide an insight into the role of alternative splicing in cell fate determination and neuronal function and improve isoform detection at the single-cell level.

Altos Labs' mission is to restore cellular health and resilience through cell rejuvenation, aiming to reverse disease, injury, and disabilities that can occur throughout life. As DNA methylation patterns regulate gene expression and other crucial biological processes, efficiently characterizing 5-methylcytosine (5mC) in the whole-genome context is of significant utility. This talk outlines a flexible approach leveraging Oxford Nanopore Technologies (ONT) sequencing to reliably deliver the required amount of data from as little as 60 ng of genomic DNA (gDNA). The methodology employed enables comprehensive methylation profiling while minimising sample input requirements, a valuable asset in the pursuit of advancing cell rejuvenation therapies.

Attaining insight into RNA modifications and base analogs is increasingly important in molecular biology and genetics. As a widely used chemotherapeutic drug for the treatment of solid cancer (colorectal, breast cancer), the efficacy of 5-Fluorouracil (5-FU) is partial and often associated with drug tolerance. Although 5-FU has been applied clinically for more than 60 years, nearly half of the patients are unable to respond to this therapy positively. Methods for identifying and quantifying 5-FU have often relied on time-consuming and labor-intensive experimental techniques but computational methods for detecting 5-FU are lacking. There is an urgent requirement for new tools to finely map the location of 5-FU within transcriptome to improve our understanding its role in cytotoxicity. Recent advances in direct RNA sequencing using Oxford Nanopore R10 platform provide a promising approach to discriminate and identify different RNA modifications in native RNA sequence. This next-generation tool introduces development challenges—such as version conflicts and analytical pipeline refinements—but enables deeper exploration of emerging biological questions. We developed RNAscent, which is, to our knowledge, the first algorithm capable of pinpointing 5-FU positions in individual RNA molecules using the latest ONT R10 platform. RNAscent combines robust data processing pipelines with a deep neural network to achieve high-sensitivity localization of 5-FU. Trained on in vitro samples with complete 5-FU incorporation, RNAscent achieved a sensitivity of 0.8858 and specificity of 0.8810, demonstrating robust performance even at lower incorporation rates. Overall, RNAscent offers a streamlined, user-friendly solution for researchers seeking to detect RNA modifications, including base analogs, at the single-molecule level. Notably, this is the first reported instance of a chemotherapy metabolite being successfully detected in nanopore-sequenced RNA. RNAscent represents a substantial leap forward in quantifying 5-FU at single-base resolution, promising new insights into RNA damage pathways, drug tolerance, and personalized therapeutic strategies.

Anorexia nervosa (AN) is a severe eating disorder defined by restrictive eating and low body weight, with the highest mortality rate among psychiatric conditions and a heritability estimate of approximately 50–60%. Binge eating disorder (BED), by contrast, is characterised by recurrent episodes of consuming unusually large quantities of food over a short period, coupled with a sense of loss of control, often resulting in significant distress and various adverse health outcomes (heritability ~40-50%). In order to identify key genetic contributors to both conditions, we are performing long-read sequencing on 2,000 extreme AN cases and 2,000 extreme BED cases as part of the NIHR BioResource Eating Disorders Genetic Initiative, which itself is nested within the broader NIHR BioResource Long Read Sequencing Initiative. We also incorporated data from the Genomics England 100K project, where we previously recruited 57 individuals with severe familial AN and 52 of their relatives, plus a further 97 AN cases from the wider 100K programme, underwent whole-genome sequencing using Illumina short-read technology (totalling 154 AN participants and relatives). To narrow the search space, we used results from a 2019 genome-wide association study that identified eight genome-wide risk loci, four of which were single-locus signals (CADM1, MGMT, FOXP1, and PTBP2we applied KGWAS, a deep-learning tool that uses a functional knowledge graph of variants and genes, to confirm FOXP1 and CADM1 as significant single-gene loci. Within the Genomics England Research Environment, we then employed bcftools and the Interactive Variant Analyser (IVA) to assess whether any of these 154 AN participants harboured damaging variants in FOXP1 or CADM1. We identified nine participants carrying potentially damaging variants in FOXP1, and 46 with damaging variants in CADM1. Currently, we are extending these analyses (which I will present) to the first ~400 eating disorder cases - both very severe AN and very severe BED - sequenced via Oxford Nanopore technology. I will present analyses of methylation status and genetic variation in the FOXP1 and CADM1 loci.