In this talk, Dr. Aaron Wenger, describes how PacBio HiFi reads (15 kb – 25 kb, >99.9% accuracy) provide the most complete view of human genetic variation, including small variants in difficult-to-map regions and structural variants genome-wide. Further, PacBio sequencing simultaneously detects epigenetic modifications without requiring a specialized library preparation step like bisulfite conversion. This ability is commonly used to characterize epigenetic marks in bacterial genomes. Recent improvements in read length and data analysis have extended the ability to include the 5mC methylation that is present at CpG sites in human genomes. Using a deep learning model that integrates sequencing kinetics and base context, the accuracy of 5mC detection in humans for individual HiFi reads is around 80%. Combining multiple reads, the concordance to EMseq and bisulfite sequencing reaches around 90%. The single-molecule resolution of methylation, together with phasing from accurate long reads, allows the detection of allele-specific methylation patterns such as parental imprinting. This ability to detect bases and modifications allows HiFi sequencing to provide the most complete genome and epigenomes with a single technology and library preparation.
October 28, 2021 | Presentation