Debarshi Mustafi from the University of Washington spoke about using long-read approaches to learn about genomic and epigenomic basis of cancer. The session was part of an Oxford Nanopore Technologies (ONT) YouTube video I watched tonight. Mustafi said that “there can be a lot embedded in the genetic material that we are understanding over time.” However, Mustafi described the challenges of commercial cancer panels. With ONT, you can change a panel by modifying a BED file. With the ability to sequence longer stretches of DNA, you can uncover pieces of the puzzle that were overlooked before. Mustafi’s team extracts DNA from blood using bead-based extraction. They then sequence on a MinION running on a desktop workstation with two NVIDIA RTX A6000 GPUs. The team then uses deep learning tools to study genes that may contribute to cancer. Methylation signals can provide insight into epigenetic signatures. Mustafi spoke about the power of phasing and epigenetic data for learning about imprinting and its role in cancer development. Their approach can “allow for more informed decision making not only for the proband but also families to better assess risk of other siblings or more distant family members.” Mustafi concluded that the ability to phase a genomic variant with epigenetic signals is a novel method that can uncover new associations.
