WYMM Tour: Hong Kong

This presentation highlights the innovative advancements of Oxford Nanopore Technologies in genetic and clinical research. Addressing the limitations of traditional sequencing methods, Oxford Nanopore’s real-time DNA/RNA sequencing platform explores poorly interrogated genome regions, providing comprehensive insights into genetic disorders, cancer, and infectious diseases. The technology’s flexibility and scalability, from personal to high-output sequencers, support diverse applications across health, agriculture, and biopharma. By enabling rapid and accurate genetic analysis, Oxford Nanopore aims to democratize genomic research, potentially enhancing diagnostic capabilities and driving advancements in personalized medicine and public health research.

In recent years, significant improvements in the accuracy, read length, and throughput of Nanopore sequencing have made it a powerful tool for studying microbiomes. Our goal is to enhance metagenomic analysis of the absolute abundance of genes and microorganisms within a microbiome using a standardized quantification method, taking antibiotic resistance genes (ARGs) as an example.

Existing diploid genome assemblers typically rely on two types of long reads: ultra-long ONT reads and HiFi reads. Ultra-long ONT reads are necessary due to their ability to span most repeat regions, while HiFi reads are crucial for accurately distinguishing between the two haplotypes of a diploid genome. However, the cost of sequencing both ultra-long ONT reads and HiFi reads can be prohibitive.

In this presentation, we explore the feasibility of assembling a diploid genome without the need for HiFi reads. We propose a method called hypo-assembler, which eliminates the reliance on HiFi reads. Instead, hypo-assembler utilizes ONT long reads and Illumina short reads. Initially, the hypo-assembler assembles a haploid draft genome using ONT long reads. Subsequently, through a process of polishing using both ultra-long ONT reads and Illumina short reads, the draft genome is refined into a diploid genome.

Direct RNA sequencing (RNA-seq) has revolutionized our understanding in RNA biology, but its potential is limited by high RNA input requirements and costs, further compounded by limited multiplexing capabilities. To address these limitations, we developed DecodeR, a novel machine-learning framework combining basecalling, classifying and computational analysis for multiplexed nanopore RNA sequencing.

Liquid biopsy using cell-free DNA (cfDNA) in plasma has provided a noninvasive approach for prenatal and cancer testing. However, the analyses of cfDNA have been focused on short DNA molecules (e.g., ≤ 600 bp). Using long-read sequencing technologies, our group revealed the presence of long cfDNA in plasma samples from healthy subjects, pregnant women and cancer patients. We characterized the fragmentomic and epigenetic features of long cfDNA. Taking advantage of the ability of long-read sequencing to directly analyze the methylation patterns of multiple CpG sites in each long cfDNA molecule, we developed an approach to determine the tissue-of-origin of individual long cfDNA molecules. Potential clinical applications of the analysis of long cfDNA, such as the detection of preeclampsia, the prenatal testing of monogenic diseases, and the detection of liver cancer, have been demonstrated in our recent proof-of-concept studies. These studies marked the beginning of a new era in liquid biopsy.

Genomic structural variants (SVs) are widespread in human genetic disease genomes. SVs can affect the cell’s growth, development and function by perturbing gene structures and expression regulation. The majority of SVs have not been identified by short-read sequencing because of the complexity and the long span. Nanopore long-read DNA sequencing technologies offer significantly longer read lengths, which will facilitate the detection and analysis of highly rearranged genome alteration. A comprehensive understanding of the structure and distribution of SVs in disease genome will empower the biomedical research community to reveal mechanisms that induce genome variation, identify prognostic signatures of disease and suggest potential targets for novel treatment strategies.

Genomic structural variants (SVs) including translocation, insertion and inversion are known to cause human infertility. However, the incidence is significantly underappreciated by conventional methods. In addition, the pathomechanism leading to human disease is still largely unknown.