Expert Q&A: Smithsonian Scientists on Evaluating DNA Quality
Photo by Alex Proimos
With the rapid adoption of long-read sequencing and other technologies that require high molecular weight DNA input, there is rising demand for high-quality samples from biorepositories around the world. To accommodate this, scientists at the Smithsonian Institution’s National Museum of Natural History in Washington, DC, developed and published a rapid, inexpensive method for quantifying DNA quality. We spoke with co-authors Jonathan Coddington and Daniel Mulcahy to learn more about this new technique.
Q: Why is there interest in working with high molecular weight DNA?
JC: We work for the Smithsonian Institution, so we’re in the forever business. We want to build a library of the highest-quality genetic resources for biodiversity. There are about 1.9 million species described and many, many millions more out there that haven’t been described — and we’re facing the sixth extinction crisis. So it’s important to collect and preserve samples that would have the maximum utility to science for many decades into the future.
DM: We’re interested in high molecular weight DNA because now it’s very practical to sequence the entire genome of non-model organisms. Even though you shear the DNA up to make the libraries, it’s better to start with big chunks of complete DNA so you’re shearing randomly across the genome, not as it’s been sheared through wear and tear.
JC: We’re also anticipating better technology in the future. Long-read technology is upon us, so everything from what you do in the field until when you extract the DNA needs to be adapted for very high molecular weight DNA.
Q: Before your method was published, how were people assessing sample quality?
DM: A lot of people were putting it on a spectrophotometer and looking at the concentrations. Another common way was to just PCR amplify short fragments, but that doesn’t really tell you if it’s high molecular weight DNA. There are more expensive methods as well, but we were trying to come up with an inexpensive way to quickly assess a lot of tissues based on DNA size.
JC: The concern with DNA quality has been around ever since the beginning of sequencing, but because PCR was so dominant people only measured the quality of their sample based on the marker that they wanted to amplify, not whole genomic DNA. An initial baseline assessment of the size frequency distribution for whole genomic DNA is almost never done in biodiversity biorepositories. We need to move the community to thinking about whole genomes.
Q: Have you seen signs that people are going in that direction?
JC: The sequencing research community has really moved beyond PCR marker-based sequencing. They want to do whole genomes and they’re looking to us to supply that high-quality DNA.
Q: How will having DNA quality information change how genomic scientists work?
DM: Our method can tell you how much of your sample was greater than the band you chose. Ultimately we’ll get to a point where we can sequence an entire chromosome; ideally you would want to be able to run the DNA out to the level where you can see various-sized chromosomes if it’s completely intact DNA. That’s where the future of this is going.
JC: Our paper was designed to enable a really cheap, really fast way for our community to inform people who want to do experiments what sort of quality DNA we might have. It’s a gel electrophoresis-based method because everybody can run gels.
Q: How will these quality assessments affect biorepositories?
JC: Dan and I represent a rapidly growing community of museums and other kinds of biorepositories. The Smithsonian, for example, has publicly about 90,000 samples. We’re also supporting an international network of biorepositories using a common data standard to report the existence and quality of samples that we all hold. The global total is about 615,000 publicly discoverable genetic resources. When we’re loaning out really rare samples, we want to know what you want to do with it. If you need the high-end stuff for a really good reason, we’ll give it to you.
DM: But if they don’t need it, we’ll give them lower-quality DNA to save the better stuff for someone who wants to do whole-genome assemblies.
Q: In the paper, you also looked at preservation methods. What were your key findings?
DM: That was a very preliminary study. If you stick a piece of muscle in the freezer and freeze it, it’s probably as good as it can be. But when you thaw it out, those proteins start to break down and the enzymes start to chew up the DNA.
JC: The DNases can still be active on the thaw cycle. I think the community assumption was that fresh frozen in liquid nitrogen was as good as it gets. What we found that was surprising was that saturating it in some sort of preservation buffer prior to freezing produces results that are better than just freezing. That’s the intriguing implication.
DM: We’re conducting a much longer-term study now. We’d like to see other people test this with other organisms, other buffers, and different storage methods.
San Diego Is Calling: Are You Ready for PAG?
It’s the start of a bright, shiny new year. If you’ve been in the genomics field long enough, you know that can mean only one thing: we’re getting ready for PAG!
Officially the Plant and Animal Genome XXVI Conference, this meeting attracts more than 3,000 scientists from around the world. Sociologically speaking, it’s a remarkable event: attendees sort themselves by the organism they study, leading to get-togethers of “cow people,” “oat people,” and so on. Each group has its own traditions, and the whole experience is really something.
Held this year from January 13-17, PAG promises to do what it does best: serve as the gathering place where all these individual communities come together to share updates on the latest science and technology. We always enjoy geeking out on the wealth of posters and catching some of the high-profile plenary talks. This year, we’re particularly looking forward to hearing from the University of Washington’s Jay Shendure, who will be speaking about new genomic frontiers, and USDA’s Doreen Ware, who has accomplished real breakthroughs for the maize community recently and will be presenting on agricultural challenges and opportunities from big data. PAG also keeps up with cutting-edge science; new workshops this year focus on paleogenomics, managing big data, and farming insects as a food source, among others.
If you’ll be attending PAG, please take a moment to swing by booth #422 to say hello to the Sage Science team. We’ll be showcasing our portfolio of automated DNA size selection projects, which can help increase average read lengths for PacBio and Oxford Nanopore sequencers, and improve the sequencing efficiency for Illumina sequencers. We’d be happy to hear about your research and figure out whether automated sizing could help you generate better results.
We hope to see you in San Diego. Have a great PAG, everyone!
New Mosquito Assembly to Accelerate Science and Insect Control
There’s a lot to love about the new preprint announcing the first-ever truly high-quality reference genome assembly for the Aedes aegypti mosquito. The large consortium of scientists — the paper features a total of 72 co-authors — came together on its own, with researchers signing on in droves after a tweeted plea from mosquito researcher Leslie Vosshall. The project took two years and yielded a remarkably complete assembly with a contig N50 of nearly 12 Mb. Here at Sage Science, we’re also delighted to see that two of our automated DNA size selection instruments, the SageELF and BluePippin, were used to prepare libraries for sequencing.
The preprint, currently titled “Improved Aedes aegypti mosquito reference genome assembly enables biological discovery and vector control,” comes from lead authors Benjamin Matthews, Olga Dudchenko, and Sarah Kingan. Scientists on the project hailed from Rockefeller University, Baylor College of Medicine, NHGRI, and dozens of other institutions.
Launched during the height of the Zika virus outbreak, the consortium aimed to generate a more complete and contiguous assembly for Aedes aegypti than was previously available. Since the mosquito genome is repetitive and spans more than a gigabase, prior assemblies were marked by thousands of gaps, short contigs, and sequence that couldn’t be mapped to chromosomes. “We used long-read Pacific Biosciences sequencing and Hi-C scaffolding to produce a new reference genome (AaegL5) that is highly contiguous, representing a decrease of 93% in the number of contigs, and anchored end-to-end to the three Ae. aegypti chromosomes,” the scientists report. They also produced a complete assembly of the mitochondrial genome with zero gaps.
Deep dives into the final assembly prove that it is more accurate and complete than previous resources, with more robust gene predictions. “The high-quality genome assembly and annotation described here will enable major advances in mosquito biology and has already allowed us to carry out a number of experiments that were previously impossible,” the team writes. For instance, they identified new candidate loci that appear to be associated with competence as a vector of dengue virus. They also learned more about the sex-determining mechanism, which is important for efforts to shift mosquito populations toward harmless-to-human males.
All in all, the scientists conclude, this new assembly should allow the community to make major strides in understanding mosquito biology and controlling their impact on human health.
A Look Back at 2017
As we start to make lists of New Year’s resolutions (with bets on how long they’ll last), it’s the perfect time for a moment to absorb the themes and highlights of 2017. In our corner of the world, that means impressive advances around HMW DNA and long-read sequencing; novel biological insights about infectious disease, cancer detection, and more; plus new and improved sample prep methods for DNA sequencing.
For us, one of the breakthroughs of the year came in the form of CATCH, or Cas9-assisted targeting of chromosome segments. This method from Yuval Ebenstein and collaborators allows users to target large or complex regions of the genome for cost-effective sequencing. The innovation is based on CRISPR, taking advantage of the precise activity of Cas9 guide enzymes to snip out the region of interest. The original method, which relied on gel electrophoresis, was improved by swapping in the SageHLS instrument for a more streamlined and automated process with excellent recovery.
If you polled the Sage team about the best part of our jobs, it would be unanimous: getting to know our customers! This year, we had the honor of profiling great work from Anna Selmecki at Creighton University, who is using BluePippin to boost library recovery for investigations into genome instability in fungi, and the Broad Institute’s Michelle Cipicchio, who helps optimize methods before they are put into production. She has been using the PippinHT platform to get the best results from the 10x Genomics Chromium system.
Of course, we also spend plenty of time keeping up with the literature — especially the growing number of preprints. One of our favorite studies this year came from scientists in Brazil who reconstructed the transmission path of a recent chikungunya outbreak in their country. The team’s budget was tight, but their results show just how much can be accomplished with creativity and a little help. We were also particularly impressed by a preprint from UK scientists who demonstrated that size selection can significantly improve results from circulating tumor DNA studies, with implications for liquid biopsies in general. And the field of long-read sequencing continued to heat up, with lots of advances including this great comparison of PacBio and Oxford platforms for transcriptome analysis.
As always, we enjoyed hearing from luminaries in the genomics field through Mendelspod interviews this year. If you missed the podcasts with Mark Akeson, Deanna Church, Yuval Ebenstein, or Evan Eichler, we recommend carving out some time to listen.
From all of us at Sage Science, we wish you and yours a healthy and happy holiday season.
Preprint Roundup: Long-Read Sequencing, ddRAD-seq, and a New Look at Viruses
We always keep our eyes peeled for interesting new research from scientists using Sage Science automated DNA size selection instruments, and several recent preprints caught our attention. Here’s a look:
Nanopore Sequencing Reveals High-Resolution Structural Variation in the Cancer Genome
Authors: Liang Gong, Chee-Hong Wong, Wei-Chung Cheng, et al.
Scientists from The Jackson Laboratory for Genomic Medicine and China Medical University in Taiwan teamed up to detect structural variants in breast cancer genomes using a custom-built pipeline called Picky. They chose nanopore sequencing to generate long reads, identifying SVs with excellent sensitivity and specificity and finding that repetitive DNA was the primary source of cancer-related variation. This approach could prove useful in efforts to assess genome stability in a tumor over time. The team used BluePippin to size-select 12 Kb libraries prior to nanopore sequencing.
Authors: Jonas Korlach, Gregory Gedman, Sarah Kingan, et al.
In this work, scientists seeking to improve upon short-read genome assemblies for two birds deployed long-read PacBio sequencing to generate new diploid assemblies. The effort yielded assemblies with megabase-sized contigs, with a 150-fold improvement in the contiguity for the zebra finch genome and 200-fold improvement for Anna’s hummingbird. Since the birds are both models for vocal learning, the higher level of completeness, correction of previous misassemblies, and more accurate gene sequences will be important for many future studies. The team used BluePippin to size libraries for zebra finch and hummingbird prior to sequencing.
CIDER-Seq: unbiased virus enrichment and single-read, full length genome sequencing
Authors: Devang Mehta, Matthias Hirsch-Hoffmann, Andrea Patrignani, et al.
Scientists developed a new method for deeply sequencing viruses that can accurately represent populations with high levels of homology across genomes. They incorporated long-read sequencing with random circular amplification enrichment and a novel de-concatenation protocol, validating their results in a large population of geminiviruses. BluePippin was used for size selection prior to the enrichment step and again during library preparation for sequencing.
Authors: Derrick Thrasher, Bronwyn Butcher, Leonardo Campagna, et al.
Continuing with the avian theme, researchers used ddRAD-seq to analyze as many as 600 SNPs from up to 240 members of a population — validated in this case with a study of Malurus lamberti and other bird species. By comparing results to microsatellite markers, they determined that the ddRAD-seq method “results in substantially improved power to discriminate among potential relatives and considerably more precise estimates of relatedness coefficients,” they report. The pipeline they present, which relies on BluePippin for DNA fragment sizing, can be used with any other bird species, and other organisms as well.