NGS Target Enrichment Panel Selection

Sequencing an entire genome can be expensive, time consuming, and require a lot of work and expertise to analyze. Although next-generation sequencing (NGS) costs have plummeted, it still often behooves researchers and clinicians to pare down what goes into the sequencer—not only for reasons of economy, but for added sensitivity as well. By shrinking the haystack, it’s easier to find the needles hidden therein.

Many vendors offer target enrichment panels, which allow the users to select and amplify regions they are interested in sequencing. These can be the 1–2% of the genome that makes up the exome, or a much more focused panel looking at all known cancer-relevant genes, for example, or even a more select collection for a specific research project.

While such panels can be purchased off-the-shelf, many (if not most) can be customized by adding to the selection of targeted genes. Or researchers can design their own bespoke panels from scratch, utilizing readily available online tools. Here we examine some pros and cons of each of these approaches.

Enriching the targets

There are a variety of ways to enrich targets for NGS. The two most prevalent are what are broadly termed hybridization capture-based and PCR amplicon-based.

A PCR amplicon-based enrichment is essentially a multiplex PCR reaction designed to make large numbers of copies of selected regions. Panels consist of one or more sets of primers. It is relatively fast, has a streamlined workflow, and requires a minimal amount of input DNA. It is only suitable for small regions of interest, and some understanding of the sequence being targeted is necessary to design the primers.

Hybridization capture-based enrichment uses (typically biotin-tagged) probes of about 50–150 nucleotides to grab any piece of input DNA containing a sequence complementary to the probe. It can be used to capture targets upstream and downstream without prior knowledge of their sequences.

There are many other factors, including number and extent of the regions of interest, the types of variants being investigated, and quality of input DNA, that go into deciding which of these target enrichment methods—or some variant or combination of them—should be used. These, as well as the details of the NGS library preparation, are beyond the scope of this article.

Customizing (or not)

If there is a pre-designed panel that works for given project, it may be best to use it, for a variety of reasons.

“Off-the-shelf panels are expertly curated by a team of scientists and fine-tuned and balanced for optimal performance,” says Eleonora Juarez, Business Unit Manager–NGS Commercial Product Management at IDT. “This results in an out-of-the-box ready and quality-controlled solution.”

Yet such panels may not meet some specific research needs. “It may not cover everything you’re looking for,” says Kathryn Sheldon, Research Project Manager at Penn State’s Institute for Personalized Medicine. Certain gene fusions, for example, “which are interesting events.”

It may often be possible to add targets to a pre-designed panel. This allows the researcher to take advantage of the expertise that went in to creating that panel, but customized to their needs.

On the other hand, designing a panel from scratch enables researchers “to save valuable sequencing resources by targeting the specific regions of interest and not wasting expensive reads,” notes Brian Sogoloff, Senior Support Scientist, Roche Sequencing & Life Science. At the same time, “developing and perfecting a design can be more time consuming and frustrating.”

Be wary, too, that “the upfront cost of the customized panel can make or break a project,” Sheldon warns. “You want to make sure that you have enough samples to justify the purchase of the panel.”

Design

While designing a custom or customized panel can be challenging, help is readily available. Free online tools can be accessed on most if not all panel providers’ websites, allowing for users to input their favorite genes and regions of interest (ROIs) and get a fully designed custom panel using the companies’ proprietary software. Similarly, the algorithms can determine how best to add new ROIs to an existing off-the-shelf panel.

Some customers prefer a more personal touch, or face more challenging variants of regions than they might feel comfortable handing to an algorithm. “We provide customers with the ability to … access a team of scientists that can provide white glove panel design” consultations free of charge, remarks Juarez. Many, if not most or all, providers offer similar services.

Some of the challenges researchers need to be aware of include high or low GC content of a region, repetitive or low complexity regions, and dealing with non-human genomes, notes Sogoloff. “Spanning fusions can be challenging. Roche expert designers have access to resources to overcome this. Roche has designs available for CNV [copy number variation] and MSI [microsattelite instability] targets.”

Best of both

Custom panels and off-the-shelf panels are not mutually exclusive,” says Juarez. “Using an off-the-shelf panel for proof-of-concept testing provides customers with an easy, quick, low barrier-to-entry option that can be customized later.”

A similar approach can be used in personalized medicine, for example. “They can start by sequencing a tumor to identify targets from that patient’s specific form of cancer. Twist will then design a customized panel for those targets that can be used to inform treatment options and monitor for residual disease,” points out Emily Leproust, CEO and Co-founder of Twist Bioscience. “These panels result in high quality output on par with pre-defined panels, with the significant benefit of custom content.”

A final word of advice: “You can get a lot of reads for anything you’ve added adapters to. But you want to make sure your enrichment is working,” says Sheldon. There are benchmarks and standards of performance—measured bioinformatically using tools such as those from GATK (Genome Analysis Toolkit)— that need to be met.