Publication: Spike-In Standards for Managing Systematic Sequencing Errors
Here’s a paper worth checking out: “Synthetic Spike-in Standards Improve Run-Specific Systematic Error Analysis for DNA and RNA Sequencing” from lead author Justin Zook at the National Institute of Standards and Technology. Published in PLoS One last month, the paper describes a study of ways to better manage systematic errors in DNA or RNA sequencing.
Most sequencing work currently relies on algorithms that recalibrate base scores after calculating a correction factor using either a subset of the sequenced data set or a separate data set, the paper’s authors write. They propose using synthetic spike-in standards, in this demonstration using RNA spike-ins for sequencing human RNA. This is followed up with base recalibration with the Genome Analysis Toolkit (GATK from the Broad Institute) that more accurately adjusts based on the spike-in’s unique sequence signature.
“Compared to conventional GATK recalibration that uses reads mapped to the genome, spike-ins improve the accuracy of Illumina base quality scores by a mean of 5 Phred-scaled quality score units, and by as much as 13 units at CpG sites,” the authors write. “In addition, since the spike-in data used for recalibration are independent of the genome being sequenced, our method allows run-specific recalibration even for the many species without a comprehensive and accurate SNP database.”
In a paper that focuses on improving quality and uniformity, we were delighted to see that our Pippin platform was used for the cDNA size selection step with Illumina sequencing.
Congratulations to authors Justin Zook, Daniel Samarov, Jennifer McDaniel, Shurjo Sen, and Marc Salit!
For NRC Team, Pippin Platform Leads to Higher-Quality Assemblies
At the DNA Technologies Laboratory at the National Research Council of Canada, scientists are using the Pippin size selection platform to improve the quality of their genome assemblies.
Andrew Sharpe, Research Officer and Group Leader of the Saskatoon-based laboratory, got his first Pippin Prep last year. He added a second Pippin Prep as well as the longer-fragment Blue Pippin to his arsenal earlier this year.
In Sharpe’s lab, which also serves as a core facility for NRC and other Canadian government agencies, assembly projects tend to focus on large plant and fungal genomes. His team relies on Illumina and 454 sequencing, often adopting a hybrid assembly approach to take advantage of both platforms.
“The majority of libraries going through are the shorter, standard pair-end libraries of 200 to 400 bases,” Sharpe says, noting that those libraries run on all three Pippin machines. Longer mate libraries — usually in the range of 3kb to 10 kb — are also a good fit for the Pippin, he adds.
Sharpe and his colleagues use the Pippin platform to create multiple pair-end libraries for the same sample — constructing, for instance, three libraries with 200-base, 300-base, and 400-base inserts — and then assemble all of those sequences together, often using SOAPdenovo. “If you assemble one of the libraries, then you’ll end up with an assembly. But if you assemble all three together using three different lengths, you get quite a bit better product,” Sharpe says. “The nice thing with the Pippin Prep is being able to easily get those discrete size ranges.”
Before Sharpe had the Pippin, his team spent a lot of time on manual gel extractions. “Having the Pippin makes things quite a lot more efficient on the labor side,” he says. Now he’s looking to the Blue Pippin to take over for the field inversion gel electrophoresis his team runs for making larger 454 mate libraries. “The Blue Pippin offers the prospect of actually speeding up that process and hopefully getting away with less amounts of DNA,” Sharpe says. “You normally need a lot of DNA to operate on the standard FIGE gel, but with the Blue Pippin we should be able to get away with less.”
BluePippin:Collecting all DNA “Greater Than” 1kb
We have received several requests to use the Pippin to collect all remaining DNA above a programmed base pair value from a sample. Although Pippins currently have the capability to collect all DNA after a run time threshold (using the “Time” programming mode), there is no method to elute the entire sample after a programmed base pair value . Also, our protocol editor requires users to enter an ending base pair value (BP End) in the “Range” mode and will not accept values above 50,000 bp.
For the BluePippin, we have developed a protocol for this purpose, and named the cassette file “0.75% Greater Than – Marker S1”. This requires our 0.75% dye-free gel cassettes kit for lower ranges (BLF7510). With this protocol, users enter a 4 hour run time, and enter a 50,000 bp value into the “BP End” field using the “Range” programming mode. The 50,000 is a dummy value that tells the instrument to continue collecting until the end of the run.
At this time, the “0.75% Greater Than – Marker S1” cassette file is not available in the standard menu of cassette types, but we can provide it to you separately. Contact us if you are interested.
An example of “Greater Than” collections:
Broad Institute Teams with Sage Science for Automated Sizing
Don’t miss this great blog post from the Broad Institute (“A Sage partnership”) describing collaborative work between their genome sequencing team and Sage Science to design a better size selection process for the Broad’s sequencing pipeline.
Headed up by Sheila Fisher, assistant director of technology development for the Broad’s Genome Sequencing Platform, the goal was to replace error-prone, tedious manual gel extractions in the sample prep workflow. Working with Sage’s Pippin platform, Sheila and her team were able to automate the size selection step, improving accuracy and eliminating the chance for cross-sample contamination.
An added bonus was that Pippin sizing offered much higher yields than manual gel extraction had, allowing Fisher’s team to accept samples with just 100 nanograms of DNA, instead of the 3 or 4 micrograms the pipeline originally required. “This opened up a significant number of samples to the process that we couldn’t sequence before,” says Sheila in the blog post. “We were able to build a very strong partnership with Sage, and the result was a true co-development project.”
We couldn’t have said it better ourselves. It’s truly a pleasure to continue with our great collaboration with the Broadies!
Automated Size Selection Key for RADseq: PLoS One Publication
Congratulations to Sage Science customer Hopi Hoekstra and her team at Harvard University for their recent publication in PLoS One! Dr. Hoekstra, who works in the Organismic & Evolutionary Biology and the Molecular & Cellular Biology departments at Harvard, reports a full laboratory protocol for RADseq, or reduced-representation genome sequencing for use in population genotyping. The method allows for studying hundreds of thousands of markers across hundreds of individuals or more.
The paper, published on May 31 and entitled “Double Digest RADseq: An Inexpensive Method for De NovoSNP Discovery and Genotyping in Model and Non-Model Species,” can be found here. (link to paper)
The authors write: “Our method requires no prior genomic knowledge and achieves per-site and per-individual costs below that of current SNP chip technology, while requiring similar hands-on time investment, comparable amounts of input DNA, and downstream analysis times on the order of hours.”
As part of the library prep protocol, the Harvard team tested out manual gel extraction versus the Pippin Prep for size selection. The paper reports that manual gel excision did not perform as well as automated size selection, “likely because [gel excision] was imprecise or ‘leaky.’” The authors note that for manual gel electrophoresis, “careful practitioners can achieve roughly 50% of the precision and repeatability of automated DNA size selection.”
For more on the Hoekstra lab, click here. (link to her lab)