Medhat Mahmoud

Video explaining the heterocateny model and summarizing the main findings of the paper.

Figure S1.1. Circos plots of human genomes. Figure S1.2. Circos plots of HPV genomes. Figure S1.3. PacBio HiFi read length distributions. Figure S1.4. Oxford Nanopore (ONT) read length distributions. Figure S2.1. Detection of circularized mitochondrial DNA by LR-seq reads. Figure S2.2. Detection of HPV concatemers by LR-seq reads. Figure S2.3. ONT reads show structural rearrangements in HPV concatemers. Figure S3.1. Absence of sequence variants at breakpoints. Figure S3.2. AmpliconArchitect predicts inaccurate HPV ecDNA structures. Figure S4.1. Evolution of structural variants at Chr. 22 of Tumor 2. Figure S4.2. Evidence of heterocateny in Tumor 5 and Tumor 3. Figure S5.1. HPV integration into icDNA in GUMC-395, HeLa, and HTEC. Figure S5.2. HPV integrants at Chr. 17 and Chr. X of VU147. Figure S5.3. Metaphase FISH identifies HPV-containing ecDNAs in cell lines. Figure S5.4. Circle-seq analysis of HeLa, GUMC-395, VU147, and HTEC cells. Figure S6. Transcription levels of HPV genes. Table S1.1. Depths of autosomal and mitochondrial genome sequencing coverage from WGS and LR-seq. Table S1.2. Depths of autosomal and HPV genome sequencing coverage from WGS and LR-seq. Table S2.1. Breakpoints and normal segment junctions in Tumor 4. Table S2.2. Definition of genomic segments flanking HPV integrants in Tumor 4. Table S3.1. Breakpoints and normal segment junctions in Tumor 2. Table S3.2. Definition of genomic segments flanking HPV integrants in Tumor 2. Table S3.3. Breakpoints and normal segment junctions in Tumor 5. Table S3.4. Definition of genomic segments flanking HPV integrants in Tumor 5. Table S3.5. Breakpoints and normal segment junctions in Tumor 3. Table S3.6. Definition of genomic segments flanking HPV integrants in Tumor 3. Table S4.1. Breakpoints and normal segment junctions in GUMC-395 cells. Table S4.2. Definition of genomic segments flanking HPV integrants in GUMC-395 cells. Table S5.1. Breakpoints and normal segment junctions in HeLa cells. Table S5.2. Definition of genomic segments flanking HPV integrants in HeLa cells. Table S5.3. Breakpoints and normal segment junctions in VU147 cells. Table S5.4. Definition of genomic segments flanking HPV integrants in VU147 cells. Table S5.5. Breakpoints and normal segment junctions in HTEC cells. Table S5.6. Definition of genomic segments flanking HPV integrants in HTEC cells.

The human papillomavirus (HPV) genome is integrated into host DNA in most HPV-positive cancers, but the consequences for chromosomal integrity are unknown. Continuous long-read sequencing of oropharyngeal cancers and cancer cell lines identified a previously undescribed form of structural variation, “heterocateny,” characterized by diverse, interrelated, and repetitive patterns of concatemerized virus and host DNA segments within a cancer. Unique breakpoints shared across structural variants facilitated stepwise reconstruction of their evolution from a common molecular ancestor. This analysis revealed that virus and virus–host concatemers are unstable and, upon insertion into and excision from chromosomes, facilitate capture, amplification, and recombination of host DNA and chromosomal rearrangements. Evidence of heterocateny was detected in extrachromosomal and intrachromosomal DNA. These findings indicate that heterocateny is driven by the dynamic, aberrant replication and recombination of an oncogenic DNA virus, thereby extending known consequences of HPV integration to include promotion of intratumoral heterogeneity and clonal evolution.Significance:Long-read sequencing of HPV-positive cancers revealed “heterocateny,” a previously unreported form of genomic structural variation characterized by heterogeneous, interrelated, and repetitive genomic rearrangements within a tumor. Heterocateny is driven by unstable concatemerized HPV genomes, which facilitate capture, rearrangement, and amplification of host DNA, and promotes intratumoral heterogeneity and clonal evolution.See related video: https://vimeo.com/845407469See related commentary by McBride and White, p. 814.This article is highlighted in the In This Issue feature, p. 799

Over the past decade, long-read sequencing has evolved into a pivotal technology for uncovering the hidden and complex regions of the genome. Significant cost efficiency, scalability, and accuracy advancements have driven this evolution. Concurrently, novel analytical methods have emerged to harness the full potential of long reads. These advancements have enabled milestones such as the first fully completed human genome, enhanced identification and understanding of complex genomic variants, and deeper insights into the interplay between epigenetics and genomic variation. This mini-review provides a comprehensive overview of the latest developments in long-read DNA sequencing analysis, encompassing reference-based and de novo assembly approaches. We explore the entire workflow, from initial data processing to variant calling and annotation, focusing on how these methods improve our ability to interpret a wide array of genomic variants. Additionally, we discuss the current challenges, limitations, and future directions in the field, offering a detailed examination of the state-of-the-art bioinformatics methods for long-read sequencing.

Chromosomal inversions (INVs) are particularly challenging to detect due to their copy-number neutral state and association with repetitive regions. Inversions represent about 1/20 of all balanced structural chromosome aberrations and can lead to disease by gene disruption or altering regulatory regions of dosage-sensitive genes incis. Short-read genome sequencing (srGS) can only resolve ∼70% of cytogenetically visible inversions referred to clinical diagnostic laboratories, likely due to breakpoints in repetitive regions. Here, we study 12 inversions by long-read genome sequencing (lrGS) (n= 9) or srGS (n= 3) and resolve nine of them. In four cases, the inversion breakpoint region was missing from at least one of the human reference genomes (GRCh37, GRCh38, T2T-CHM13) and a reference agnostic analysis was needed. One of these cases, an INV9 mappable only in de novo assembled lrGS data using T2T-CHM13 disruptsEHMT1consistent with a Mendelian diagnosis (Kleefstra syndrome 1; MIM#610253). Next, by pairwise comparison between T2T-CHM13, GRCh37, and GRCh38, as well as the chimpanzee and bonobo, we show that hundreds of megabases of sequence are missing from at least one human reference, highlighting that primate genomes contribute to genomic diversity. Aligning population genomic data to these regions indicated that these regions are variable between individuals. Our analysis emphasizes that T2T-CHM13 is necessary to maximize the value of lrGS for optimal inversion detection in clinical diagnostics. These results highlight the importance of leveraging diverse and comprehensive reference genomes to resolve unsolved molecular cases in rare diseases.

Background The goal of the Fifth Annual Baylor College of Medicine & DNAnexus Structural Variation Hackathon was to push forward the research on structural variants (SVs) by rapidly developing and deploying open-source software. The event took place in-person and virtually in August 2023, when 49 scientists from 14 countries and 8 U.S. states collaboratively worked on projects to address critical gaps in the field of genomics. The hackathon projects concentrated on developing bioinformatic workflows for the following challenges: RNA transcriptome comparison, simulation of mosaic variations, metagenomics, Mendelian variation, SVs in plant genomics, and assembly vs. mapping SV calling comparisons. Methods As a starting point we used publicly available data from state-of-the-art long- and short-read sequencing technologies. The workflows developed during the hackathon incorporated open-source software, as well as scripts written using Bash and Python. Moreover, we leveraged the advantages of Docker and Snakemake for workflow automation. Results The results of the hackathon consists of six prototype bioinformatic workflows that use open-source software for SV research. We made the workflows scalable and modular for usability and reproducibility. Furthermore, we tested the workflows on example public data to show that the workflows can work. The code and the data produced during the event have been made publicly available on GitHub (https://github.com/collaborativebioinformatics) to reproduce and built upon in the future. Conclusions The following sections describe the motivation, lessons learned, and software produced by teams during the hackathon. Here, we describe in detail the objectives, value propositions, implementation, and use cases for our workflows. In summary, the article reports the advancements in the development of software for SV detection made during the hackathon.

The assignment of variants across haplotypes, phasing, is crucial for predicting the consequences, interaction, and inheritance of mutations and is a key step in improving our understanding of phenotype and disease. However, phasing is limited by read length and stretches of homozygosity along the genome. To overcome this limitation, we designed MethPhaser, a method that utilizes methylation signals from Oxford Nanopore Technologies to extend Single Nucleotide Variation (SNV)-based phasing. We demonstrate that haplotype-specific methylations extensively exist in Human genomes and the advent of long-read technologies enabled direct report of methylation signals. For ONT R9 and R10 cell line data, we increase the phase length N50 by 78%-151% at a phasing accuracy of 83.4-98.7% To assess the impact of tissue purity and random methylation signals due to inactivation, we also applied MethPhaser on blood samples from 4 patients, still showing improvements over SNV-only phasing. MethPhaser further improves phasing across HLA and multiple other medically relevant genes, improving our understanding of how mutations interact across multiple phenotypes. The concept of MethPhaser can also be extended to non-human diploid genomes. MethPhaser is available at https://github.com/treangenlab/methphaser.

Comprehending the mechanism behind human diseases with an established heritable component represents the forefront of personalized medicine. Nevertheless, numerous medically important genes are inaccurately represented in short-read sequencing data analysis due to their complexity and repetitiveness or the so-called ‘dark regions’ of the human genome. The advent of PacBio as a long-read platform has provided new insights, yet HiFi whole-genome sequencing (WGS) cost remains frequently prohibitive. We introduce a targeted sequencing and analysis framework, Twist Alliance Dark Genes Panel (TADGP), designed to offer phased variants across 389 medically important yet complex autosomal genes. We highlight TADGP accuracy across eleven control samples and compare it to WGS. This demonstrates that TADGP achieves variant calling accuracy comparable to HiFi-WGS data, but at a fraction of the cost. Thus, enabling scalability and broad applicability for studying rare diseases or complementing previously sequenced samples to gain insights into these complex genes. TADGP revealed several candidate variants across all cases and provided insight into LPA diversity when tested on samples from rare disease and cardiovascular disease cohorts. In both cohorts, we identified novel variants affecting individual disease-associated genes (e.g., IKZF1, KCNE1). Nevertheless, the annotation of the variants across these 389 medically important genes remains challenging due to their underrepresentation in ClinVar and gnomAD. Consequently, we also offer an annotation resource to enhance the evaluation and prioritization of these variants. Overall, we can demonstrate that TADGP offers a cost-efficient and scalable approach to routinely assess the dark regions of the human genome with clinical relevance.

The All of Us (AoU) initiative aims to sequence the genomes of over one million Americans from diverse ethnic backgrounds to improve personalized medical care. In a recent technical pilot, we compare the performance of traditional short-read sequencing with long-read sequencing in a small cohort of samples from the HapMap project and two AoU control samples representing eight datasets. Our analysis reveals substantial differences in the ability of these technologies to accurately sequence complex medically relevant genes, particularly in terms of gene coverage and pathogenic variant identification. We also consider the advantages and challenges of using low coverage sequencing to increase sample numbers in large cohort analysis. Our results show that HiFi reads produce the most accurate results for both small and large variants. Further, we present a cloud-based pipeline to optimize SNV, indel and SV calling at scale for long-reads analysis. These results lead to widespread improvements across AoU.