Next-generation sequencing (NGS) covers much ground and many applications. One traditional differentiator has been whole-genome versus targeted sequencing, corresponding roughly to “research” vs. “applied” investigations. Because it seeks to pick up distinct patterns from many millions of potential sequences, targeted sequencing almost always involves an enrichment step.
Users enjoy several choices, even within targeted sequencing. Hybrid capture methods use complementary probes to desired targets, while amplicon-based approaches rely on multiplexed polymerase chain reaction (PCR). Selective circularization (aka molecular inversion probes) offers a third alternative.
As its name implies, target enrichment improves signal to noise by amplifying signals from genes of interest relative to those from other genes in the sample.
Andrew Barry, senior manager for business development at New England Biolabs (NEB), explains that applications requiring enrichment are inherently demanding. “These applications typically involve finding the needle in the proverbial haystack, and require either very high coverage depth or selectivity of a specific sequence among some contaminating host DNA, which renders whole-genome sequencing intractable.”
Barry cites the example of somatic variant calling in cell-free tumor DNA, where investigators are looking for increasingly low allele fraction mutations in both samples harboring these variants at low frequency and samples swamped with “healthy” DNA.
“The enrichment process itself also introduces uncertanties. For amplicon-based methods amplification itself can introduce false-positives, while capture-based approaches subtract the targets of interest while exposing DNA to harsh temperatures and reagents that can cause similar issues.”
Developers of NGS sample-prep products have worked to overcome these challenges, for example through molecular counting strategies where, to resolve false positives, primary DNA samples are tagged with a molecular barcode before enrichment. “These have proved promising,” Barry says, “for both preserving native DNA integrity, and in correcting errors introduced downstream of sample preparation, in the sequencing reaction itself.”
NEB has focused its efforts to improve NGS workflows by introducing methods that preserve sample fidelity, while extracting the most information from samples in which targets exist in very low abundance, or which have been contaminated or otherwise compromised. “In some cases, it is possible to apply our NEBNext FFPE DNA Repair Mix, before target enrichment, to correct errors introduced during tissue fixation,” Barry explains. Another NEB product, the NEBNext Microbiome DNA Enrichment kit, exploits the fact that bacterial DNA is hypomethylated, and may be separated from human DNA on that basis. “We also have NEBNext Direct, which directly enriches genomic DNA before amplification, and uses a gentle, capture-based approach in combination with an enzymatic removal of off-target sequence to achieve very high specificity while preserving the primary sample integrity.”
Sources of variability
Emily Leproust, Ph.D., CEO and founder of Twist Bioscience, says that one of the main sources of variability in target enrichment is the inconsistency of enriched sequences. “Often, sequence capture is not uniform across the target region of the sample, resulting in underrepresentation of less common variants, which are often regions of interest. Consequently, more sequencing depth is required to capture that particular variation, which makes target enrichment less cost-effective. In addition, more sequencing of off-target regions leads to a deluge of unusable data, further increasing the cost of bioinformatics and data storage.”
Another issue of concern, she says, is in the time involved in customizing target-enrichment panels. Taking such panels from small-scale pilot to routine implementation of a consistent, well-defined panel can take several iterations over many months. “Due to the iteration time, scientists conclude that custom target enrichment is too expensive and time-consuming to be accessible, which causes many to default to fixed panels, or more commonly whole-exome sequencing.” However, as we expand our genomics understanding, we find key regulatory regions outside the exome. “Researchers want to be able to look at all of this information at one time. Creating customized panels to enrich for nonexonic content, as well as smaller custom panels, could allow for more rapid expansion into translational research applications.”
Twist launched its target-enrichment product line just last year, with the release of the Twist Human Core Exome Kit and Twist Custom Panels. In addition to the Twist Human Core Exome Kit, Twist offers library-preparation components and continues to expand its offering in fixed panels for both human and nonhuman species. The Twist Custom Panels enables rapid panel customization and optimization.
While target enrichment still poses challenges it is now possible, by leveraging efficient and precise DNA synthesis, to design “very effective probes coupled with exact synthesis to ensure that every probe works efficiently for the highest quality capture reaction,” LeProust says. Additionally, custom panels will continue to make target enrichment efficient and effective for targeted genomic analysis.
The highly variable nature of NGS samples, and the equally diverse set of end-results, suggest that target enrichment will continue as a top priority for sequencers, particularly toward developing actionable medical diagnostic tests.
“As a consequence, I don’t think there will ever be a one-sized-fits-all approach to target enrichment,” Barry adds. “Specific research goals have their own set of expectations from sequencing. These can vary significantly based on sample source and abundance, and ultimately the goal of the study with regard to detection sensitivity for samples with scientific impact.”
LeProust agrees that off-the-shelf enrichment strategies are not viable given current technology. “Researchers want the ability to quickly create custom panels, and iterate on those panels rapidly.” For example, Twist’s scalable, silicon-based oligonucleotide synthesis manufacturing platform generates high-performing probe panels for NGS target enrichment for pilot trials and at scale, and in significantly less time than a typical custom panel design, according to the company.