About this Class

Cancer origination and progression is a complex process that can be viewed as a somatic evolutionary progression with clonal cellular expansion(s) driven by accumulation of survival-/evasion-beneficial genomic mutations, alongside constantly changing selective pressures. We are going to focus on two aspects of genomic abberations observed in cancers: large-scale somatic genomic copy number variations (CNV)  and extrachmosomal DNA (ecDNA) amplicons. CNVs amplify or delete one or both germline alleles of genomic segments, chromosome arms, or even entire chromosomes, while ecDNA represent novel circular centromere-less DNA molecules comprised of excised parts of linear chromosomes. Both of these genomic alterations can drive tumor heterogeneity, improve tumor microenvironment adaptation, and increase potential for drug treatment resistance. We present a computational workflow to infer clone- and haplotype-specific cancer CNV profiles and identify and assembly ecDNA in tumors by processing long nanopore reads obtained with high-throughput bulk sequencing of a tumor and matching normal samples. For CNV inference the workflow focuses on inferring heterozygous germline SNPs, phasing them, and then performing count-based inference of clone- and alelle-specific CNVs in tumor samples allowing for multi-clonal composition. For ecDNA analysis the workflow focuses on detecting focal amplification coverage regions most likely representing ecDNa comprising fragments, and performing de novo assembly with the reads originating from such coverage pileups. We evaluated the proposed workflow on a range of cancer cell lines with known CNV and ecDNA aberrations. We demonstrate that our approach can detect clone- and haplotype-specific CNVs in concordance with previously published bulk and single-cell analysis, with results being stable across tumor samples' sequencing coverage levels down to 40x, putting the proposed approach on par with the industry standard NGS-based experiments. We further observe the robust capability of the presented nanopore-based method to identify and assembly ecDNA amplicons, with results remaining stable for samples sequenced with <1x WGS coverage levels. We further demonstrate proof of concept multiplexing capabilities of the nanopore platform for multi-site tumor sampling and ecDNA analysis. Lastly, we showcase the of the unique ability of nanopore reads to retain single-molecule methylation signals, with the proposed workflow allowing us to identify differentially methylated regions both across intra-tumor multi-site samples, as well as in a tumor vs normal comparison, thus shedding light in acquisition/loss of DNA modifications in ecDNA and CNV regions. Overall, the presented results demonstrate how nanopore sequencing can be cost- and time-effective stand-alone platform used to resolve some of the complexity that characterizes structurally aberrant heterogeneous cancer samples, while also revealing the previously inaccessible dimension of allele-specific tumor methylation. For our next CDSL webinar we will have a guest lecture by Dr. Sergey Aganezov from the Genomics Applications group at Oxford Nanopore Technologies. Bio: Dr. Aganezov is a Bioinformatics Scientist in the Genomics Applications group at Oxford Nanopore Technologies. His main research focuses on structural genomics, cancer genomics, DNA methylation analysis, and programmable Nanopore platform applications.  Before joining ONT Dr. Aganezov was a Postdoctoral Research Fellow at Johns Hopkins University (prof. Schatz group) following a Postdoctoral Research Fellowship at Princeton University (prof. Raphael Group). During his postdoctoral research Dr. Aganezov focused on human and plant structural genomics, including areas of assembly, structural/copy number variation detection and integration, comparative genomics, including co-leading genome variation analysis in the Telomere-2-Telomere consortium. Dr. Aganezov holds a PhD from The George Washington University's department of Mathematics, and B.S from ITMO University in Computer Science and Applied Mathematics.