A Practical Guide to Selecting a New Sequencer

Choosing a sequencer has become more complex as new chemistries, instrument designs, and expanding applications narrow the differences between platforms. Even with this convergence, instrument choice remains a significant decision because it affects budgets, workflows, and the kinds of experiments a lab can support. Researchers once framed the choice as a simple comparison of long-read versus short-read instruments, but now platform selection depends on a wider range of scientific and operational considerations.

Define the goal and the application

The first step is to identify the biological question driving the work. “Match the method to the question, not the instrument,” emphasized Pedro Echave, Ph.D., Senior Manager at Revvity. The initial goal is to understand what needs to be measured and which applications will answer that question with confidence. Determining whether the focus is variant detection, transcript structure, quantitative profiling, or structural resolution helps identify the appropriate techniques and required data type.

Clinical needs also influence technology decisions. “You need to pick technology based on what the customer that you service needs,” explained Bryan Dechairo, Ph.D., Chief Operating Officer at GeneDx. Projects requiring rapid results or long contiguous genomic information naturally point toward different methods. These needs help narrow the instrument choices early in planning and determine whether short-read or long-read sequencing is the best fit.

When to use short reads

“Short-read sequencing continues to be the workhorse for many researchers because it delivers reliable, high-quality data at a large scale and relatively lower cost per data point,” explained David Peoples, Chief Financial and Business Officer at Ultima Genomics. He added that infrastructure maturity is a major advantage. Short-read platforms offer standardized workflows, strong informatics support, and straightforward scaling. These features make them a practical option for projects where throughput, cost efficiency, accuracy, and flexibility across sample types are priorities.

In clinical care, this speed and reliability matter. Dechairo noted that GeneDx relies heavily on short-read sequencing in the NICU because quick answers guide interventions that need immediate attention. Short-read platforms meet that need while maintaining dependable performance under tight timelines.

Short reads are also well-suited for research applications where depth and accuracy outweigh the need for long-range information. Echave recommended them for fragmented or low-input samples such as FFPE, cfDNA, or single cells. He noted that they also support variant discovery in well-characterized genomes, quantitative assays (single-cell RNA-seq and ATAC-seq), targeted or amplicon-based panels, cfDNA, MRD, and microbial profiling by 16S or ITS, as well as functional genomics applications, including ChIP-seq and CRISPR screens.

The case for long reads

While early long-read approaches faced challenges with accuracy and strict sample requirements, the technology has since advanced and now plays an essential role in genomics. Long-read sequencing is important for researchers who need the big picture and to see parts of the genome that are more difficult for short-read technologies. Sequencing long fragments enables detection of structural variants, resolution of repeats and repeat expansions, haplotype phasing, de novo assembly, and full-length transcript characterization.

“An additional advantage is that long-read sequencing can directly detect epigenetic modifications without the need for chemical conversion,” explained Echave. This capability supports studies of imprinting disorders, cancer biology, and regulatory elements where methylation and base modifications matter. The technology is also a strong match for applications that depend on molecule-specific information since it directly sequences single molecules.

Long reads are also valuable when short-read platforms leave coverage gaps. Dechairo noted that some clinically important regions, including those involved in complex conditions like hearing loss, remain difficult to access with short reads. In these cases, long-read or medium-read technologies can fill in the missing regions or serve as the primary sequencing approach when coverage uniformity is essential.

Combining technologies

All three experts agreed that many projects benefit from integrating both approaches. Echave noted that long reads excel at resolving structure and phasing, while short reads provide economical depth, large cohort scalability, and sensitive detection of small variants. A common strategy is to generate a modest long-read dataset to define key structures and then rely on short reads for coverage and quantification.

Peoples added that short reads often serve as the first-line approach because of their cost efficiency and versatility. Long reads can then be applied selectively to fill gaps or clarify regions that short reads struggle to resolve. He emphasized that the ideal balance depends on the specific biological question and the type of information needed to answer it.

Additional considerations

Another critical factor when evaluating a technology is anticipating future needs, particularly scalability. Dechairo stressed that this is the most frequently overlooked challenge. Many labs operate with informal setups that work at small volumes but quickly become strained as demand increases. He cautioned that many newcomers recognize this only after growth occurs, and that scaling requires automation, robust informatics, and technology decisions aligned with sustained throughput.

Platform economics and application range also matter. Peoples highlighted that most expenses still originate from consumables, so researchers should evaluate cost structures carefully. In general, short-read systems remain the most flexible and cost-effective, supported by established workflows, while long-read technologies provide unique insights but often at a higher cost and with narrower applicability.

Computational requirements should be considered as well. Echave noted that short-read workflows rely on well-established pipelines, whereas long-read analysis often requires specialized expertise. But long-read sequencing can require less total computation because longer fragments reduce assembly needs and often span entire regions directly.

Sample characteristics are other important factors, Dechairo noted. Key considerations include the flexibility of available workflows, options for increasing throughput, compatibility with automation, and the impact of prep choices on overall cost. These choices also depend on the nature of the material being sequenced. High molecular weight DNA favors long reads because the technology relies on long fragments for contiguity, while low-input or degraded materials like FFPE samples are better suited to short reads because they accommodate short inserts efficiently.

Final advice

Taken together, long-read and short-read sequencing each have distinct strengths and work best as complementary tools. “Short reads deliver precision and scale, while long reads deliver accuracy, completeness, and context,” explained Echave. “The best sequencing strategy is the one that minimizes inference between the raw molecule and the biological insight you need.” He also encouraged researchers to assess nucleic acid quality early, think strategically about when the two data types should be paired, and plan for downstream integration and analysis needs, especially when comparing with public cohorts or using specialized pipelines.

“My advice is to think about sequencing as part of your broader data strategy, not just as a single experiment,” stated Peoples. “The real power comes from generating high-quality data at scale, because that’s what fuels better models and new discoveries.” As sequencing becomes quicker, less costly, and easier to access, Peoples explained that researchers will be free to explore broader scientific questions and include more samples and time points in their work.

“Imagine what your life needs to be like when you're really successful at what you're trying to do,” Dechairo emphasized. Instead of focusing only on current experiments, he advised planning with the future in mind and anticipating what will happen if demand increases rapidly. “Think about that at the beginning, and you're going to save yourself a lot of time in the long term.”