Digging information out of DNA costs less than it used to, but there’s no bargain-basement sequencing. Next-generation sequencing (NGS) delivers lots of data, but not for a low price. Plus, sequence generation is just one part of the process, and some parts like library preparation or data analysis cost more than others. Still, recent technologies help scientists sequence more for a better price.
According to the National Human Genome Research Institute, the cost of sequencing one million bases of DNA—just the raw data from a sequencer—dropped from about $1,000 in 2004 to about $0.10 in 2011 and is now closer to $0.01. That’s a significant decrease, but that raw data needs to be processed.
The most expensive part of sequencing depends on who is doing it, such as a company that provides the service or a researcher sequencing in the laboratory setting. One of the biggest costs—if not the highest—for a provider is sustaining the facility and keeping pace with technology updates. For users, explains Dibyendu Kumar, director of the genomics core facility at Rutgers University, library-preparation cost adds up fast for multiple samples, say 48 or more. On the other hand, sequencing cost is more, if users have only a few samples. For a new facility, Kumar points out that the highest cost is “owning the instruments and maintaining service contracts.”
When asked if reducing the cost of sequencing affects the results, Kumar replies: “Yes, but it depends on what is compromised—coverage/reads per sample, number of samples, or number of replicates.” Other experts agree. When asked the same question, Andy Felton, vice president of product management for clinical next-generation sequencing and oncology at Thermo Fisher Scientific, says, “Almost certainly.” He adds, “NGS requires time and care to both the wet side and especially the bioinformatics to get to the correct answer, and cutting corners on either end of this is likely to affect the results received.”
Image: Ion Chef is a library preparation instrument that automates that part of the Ion Torrent NGS workflow to reduce time and overall cost of sequencing. Image courtesy of Thermo Fisher Scientific.
Economics of analysis
Sequencing a nucleic-acid sample only makes up part of the process. The data must be analyzed, and the cost of analysis depends on the intended goals.
With human whole-genome sequencing (WGS), “basic analysis would cover the known variants of the 59 genes identified by the American College of Medical Genetics, where the reporting of incidental findings would likely have a medical benefit,” explains Shawn C. Baker, sequencing expert and creator of SanDiegOmics.com. Compared to other forms of analysis, this is relatively inexpensive. The reason, says Baker, is because “standard analysis just involves compute time and storage—the hard work of defining the variants and their likely impact has already been performed.”
When digging deeper with the analysis of human WGS, the range of possibilities gets quite broad. “At the simpler end, an orthogonal sequencing technology is used to confirm the original result—for example, PacBio for data generated with Illumina,” Baker says. “For idiopathic diseases with no obvious actionable result, the manual intervention of highly trained—and expensive—individuals may be required to track down the most likely ‘variants of unknown significance’.” He adds, “Depending on the scope of the investigation, it could easily involve $100,000 in salaries alone.”
Caution in the clinic
In discussing some of the more costly aspects of sequencing, Felton points out that bioinformatics can create the highest costs in a pathology lab. “It’s not the fundamental costs of the informatics, because that’s built into the system,” he explains, “but it’s the labor of reviewing the variants, reviewing the testing, noting false positives or false negatives, reviewing and coordinating the process to have all of the information that you need.”
Although some aspects of clinical sequencing can raise the costs, other features can lower the expense. “A clinical lab has routine numbers of samples that have to be run in a defined time,” Felton explains. “Therefore, automation and hands-on time and full-time equivalent costs are important.” As an example of automation, Felton mentions the Ion Torrent systems. “We reduced a lot of the hands-on time for sequencing, such as automating the creation of libraries.” Scientists at Thermo Fishers Scientific also automated much of the backend of the sequencing process, including automated variant calling and providing an online tool for reporting. “The pieces all work together seamlessly,” he says.
Overall cost cutting
At Oxford Nanopore Technologies, scientists focus on making sequencing possible in more labs by reducing the cost of platforms and consumables. This opens up a range of new applications through new technologies. For example, the MinION is a portable sequencer.
Initially at the Royal Botanic Gardens and later at the National Biofilms Innovation Centre, research fellow Joe Parker “used MinION in 2016 to do the first field-based genome-scale sequencing of a complex eukaryote genome to tell apart samples from the sister plants Arabidopsis thaliana and A. lyrate.” In 2017 and 2018, Parker says, he and his colleagues used the same platform “to do field-based sequencing to try and find the cause of Acute Oak Decline, a mysterious and devastating new disease of oak trees.”
Parker points out that this technology is low-cost, but produces long reads in real time. “In the lab, this means we can try out experiments to follow up on our hunches without risking thousands of pounds,” he says. “In the field or in the clinic, this means that—where we can define clear sequencing targets statistically a priori—we can look for clear indicative signals very rapidly, sometimes within an hour of collecting a sample.”
For higher throughput, scientists can use Oxford Nanopore Technologies’ PromethION. According to NGS scientist Alexander Wittenberg from KeyGene, “We have used this technology for a range of applications, often in order to obtain high-quality, contiguous reference genomes for crop species.” He adds, “The nanopore technology combined with KeyGene’s expertise on plant DNA extraction and advanced data analysis enables us to obtain genome insights accelerating plant breeding in an unprecedented way.” Plus, he mentions that producing the data at lower costs “now even enables us to generate genome insights for smaller breeding companies and in minor crops, as we have shown for raspberry.”
Image: PromethION and other Oxford Nanopore Technologies sequencing machines are revolutionizing plant breeding, even in minor crops and for small enterprises. Image courtesy of KeyGene.
Bringing a broader variety of options to scientists encourages more sequencing-based research. Plus, lower-cost sequencing translates to more applications. Some parts of sequencing, though, remain expensive, even where the overall technology and analysis costs less. Digging deep into DNA is just a big job in some cases.