AGBT Day 4 Highlights: SMRT Sequencing for De Novo Genome Assemblies
Tuesday, February 26, 2013
Even as attendees’ energy was waning from three marathon days at AGBT, spirits were still high as we gathered for the final day’s Genomic Technologies session on Saturday morning. This session included two speakers presenting on applications using SMRT Sequencing: Eric Antoniou from Cold Spring Harbor Laboratory and Jonas Korlach, our CSO.
Antoniou, a research investigator and manager of the genome sequencing center at Cold Spring Harbor Laboratory, presented on “Increased Read Length and Sequence Quality with Pacific Biosciences Magbead Loading System and a New DNA Polymerase.” In it, he reported on the sequencing and assembly of the 470-Mb rice genome and other genome assemblies, using the PacBio RS’ new MagBead sample loading method as well as the recently released XL chemistry.
During his presentation, Antoniou noted that the XL chemistry has significantly improved read length, such that the smallest subreads from an XL run are longer than even the largest subreads obtained with the previous C2 chemistry. The newer chemistry is also responsible for the longest single read, 26,404 bases, Antoniou has observed to date. He said that careful sample preparation processes are important for getting the longest possible subreads from the sequencer for de novo assemblies, and that better cleanup steps and size selection before sequencing are helpful in that regard.
In the second talk, entitled “Automated, Non-Hybrid De Novo Genome Assemblies and Epigenomes of Bacterial Pathogens”, Korlach said that the community is seeing a revival of finished genomes now that long-read data are available, emphasizing that for certain organisms such as rapidly evolving microbes, having a de novo finished genome will be more useful than creating a draft based on a previous related reference genome. Korlach described two recent bioinformatic improvements from PacBio, a hierarchical genome assembly process (HGAP) and an improved consensus caller (Quiver), which are used to generate finished genomes from just long-read PacBio data, with final genome sequence accuracies over 99.999%. Korlach demonstrated the ability of PacBio-generated closed, high-quality de novo genome assemblies using HGAP and Quiver in the area of infectious disease on strains from Pertussis (whooping cough), as well as for several foodborne pathogen strains of Salmonella and Listeria.
He also noted that at the same time SMRT Sequencing generates information about the epigenomes of the bacteria sequenced, and showed a high epigenetic diversity in Salmonella and Listeria strains, with indications of potential new forms of DNA modifications.
Thanks to everyone who visited our suite, met with us, or attended our presentations at AGBT 2013!