AGBT 2017, Day 1: Infectious Disease and Improved Genome Assemblies
Wednesday, February 15, 2017
We’re thrilled to be at the AGBT conference this week, taking place this year in Hollywood, Fla. On the first full day of the meeting, everyone’s mandatory wristbands look shiny and new (we suspect by the end of the week they’ll be as wilted as us). And we’ve even been getting that work/life balance down thanks to some beach volleyball with our friends from BioNano Genomics and Swift Biosciences.
At the opening session on Monday, Eimear Kenny from the Icahn School of Medicine at Mount Sinai showed why it’s essential to fully understand natural genetic diversity in a fascinating talk about a deep analysis of hospital patients from across New York City. She offered one example of a variant that’s incredibly rare in most populations, but is found in about 2% of people of Puerto Rican descent and is likely pathogenic. We were delighted to hear the presentation, which fits nicely with the efforts of many PacBio users to generate population-specific reference genomes to help characterize the full breadth of natural genetic variation.
Tuesday’s talks included an infectious disease theme, with several speakers supporting the idea that the global incidence of viral outbreaks is rising. These talks made it clear that the optimal response to these health threats involves having complete genome assemblies, including accessory genomes, where genes associated with antimicrobial resistance are often found.
The infectious disease theme continued with a talk from the Broad Institute’s Daniel Neafsey, who presented results of an ongoing effort to produce a high-quality genome assembly for Aedes aegypti, the mosquito responsible for transmitting Zika and other viruses. As part of the Aedes Genome Working Group, Neafsey’s team is working to replace a 10-year-old Sanger assembly of this mosquito with a PacBio-based assembly. The project is not yet complete, but already represents a big step forward despite the organism’s high heterozygosity and highly repetitive content: the new assembly used FALCON-Unzip to reduce the number of contigs by at least 10-fold and boost the contig N50 to nearly 2 Mb. Most importantly, gene content is far more complete in the new assembly. Neafsey offered the example of a sex determination gene, Nix, which was absent from the 2007 assembly but was found in the new assembly and could be essential for CRISPR-based efforts to control mosquito populations. Neafsey also showed the results of several scaffolding technologies — including Dovetail Genomics, 10x Genomics, and BioNano Genomics — and noted that the final result should include data from all three approaches. In the next few weeks, the team will integrate all this information, remove homologous contigs, and complete annotation. Neafsey noted that he hopes this work inspires other research communities to improve older draft assemblies they’ve been working with for other organisms. If you’re attending AGBT, you can see more assembly details at poster #1105.
Speaking of major genome improvements, Jason Underwood from the University of Washington and PacBio spoke about long-read genome assemblies of primates, as well as a new approach to understanding transcripts. The standard pipeline for building better primate assemblies in his lab involves PacBio sequencing, FALCON assembly, and scaffolding with BioNano Genomics mapping. Structural variants are then called with the SMRT-SV protocol. This has resulted in drastic improvements, such as a 560-fold more complete and contiguous gorilla assembly. Underwood also spoke about projects designed to understand human-specific variation that can be identified with these improved resources. Using the Iso-Seq method, the team is sequencing full-length cDNAs; in one recent study, they used the Sequel System to generate 118,000 full-length reads from a single SMRT Cell. They also developed Iso-Cross, through which they compare transcripts from two closely related organisms to each other (such as human and chimpanzee); the ones that map better to the organism they came from are more likely to have functional and specific roles. One example Underwood showed was a 1.9 kb human-specific deletion that removes an exon found in our close primate relatives. He told attendees that their investigation has turned up 200 human-specific variants that seem likely to have functional importance.
We invite all AGBT attendees to visit us in suite 317, where our fun Lego station lets everyone build their own plastic doppelganger!