In genomics, researchers are often faced with a trade-off: Go wide with whole genome sequencing (WGS) for a panoramic view of the genome—but at a high cost. Or, go deep with whole exome sequencing (WES) to focus on coding regions—but risk missing important variations elsewhere in the genome.
But what if you could get the best of both worlds without blowing your budget?
Blended genome exome (BGE) sequencing is a novel approach that offers a way through that trade-off. By combining low-pass WGS with deep exome sequencing, BGE sequencing offers a cost-effective middle ground: broad variant detection and imputation power from genome-wide coverage, with the depth needed to confidently call rare and clinically-relevant variants.
Originally developed for large-scale biobank and population health studies, BGE is now finding wider clinical and translational relevance—from rare disease diagnostics to IVF screening. It’s a scalable, cost-efficient approach for projects that don’t require the full depth of WGS but demand broader insights than exome or array data can provide. Realizing these benefits, however, takes deep genomic expertise and a precisely tuned process.
Read on as we break down how BGE works, where it fits, and how short- and long-read sequencing technologies can be leveraged to support this powerful hybrid method.
Discuss a BGE Sequencing project with us today
What is Eremid®’s approach to BGE Sequencing?
BGE sequencing relies on the creation of two libraries from the same DNA sample: typically, a PCR-free whole genome library, plus a target-enriched exome library. These are then “blended” together and sequenced in the same run using high-throughput short-read sequencing1 to obtain a single dataset containing both low-pass genome-wide coverage and high-depth coverage of the exome.
The resulting data is aligned and processed through high-performance pipelines for variant calling (e.g., Illumina DRAGEN™), plus imputation and phasing (e.g., GLIMPSE2). However, DRAGEN has limitations in parsing exome and WGS content in blended datasets.
To overcome this, at Eremid® Genomic Services, we apply a custom hybrid pipeline that partitions genome and exome reads before performing region-specific QC, alignment, and variant calling—ensuring accurate detection in both low- and high-pass regions.
We then deliver fully annotated Variant Call Format (VCF) files optimized for downstream clinical interpretation or cohort-scale association studies.
As a result, Eremid’s BGE service offering combines the strengths of WGS and WES in a single cost-effective workflow, optimized to deliver a minimum for over 90% of the genome and high-resolution coverage for over 95% of the exome.
We have validated this workflow in our CLIA/CAP-certified laboratory from both saliva and blood specimens (Table 1). We generated a single dataset providing high-confidence detection for more than 97% of all ClinVar SNPs (version 4.2)—representing over 4 million genomic variants classified for diseases and drug responses—at a fraction of the standard 30x WGS cost.
Table 1: High-confidence germline detection of ClinVar pathogenic variants from a cohort of blood and buccal swab samples.
Figure 1: Blended Genome Exome (BGE) sequencing combines high-coverage whole exome sequencing (green) with low-pass whole genome sequencing (blue) from a single sample.
Advantages of BGE sequencing
Table 2: Key advantages of BGE sequencing.
Eremid’s BGE analysis pipeline offers enhanced quality control for BGE libraries and region-aware variant calling that improves both clinical and research-grade accuracy.
As described in Table 2, whether your goal is large-scale association studies or focused clinical screening, BGE delivers richer data than exome sequencing alone—and greater affordability than whole genome sequencing at scale.
Applications and use case: Where does BGE sequencing fit?
BGE sequencing was initially developed to serve the needs of large-scale population genomics—projects that required more variant resolution than arrays could provide, but couldn’t justify the cost of deep whole genomes.
This remains one of BGE’s core strengths: Ideal for large-scale studies, it enables high-quality imputation across the genome and rare variant detection, all from a single consolidated dataset. Since BGE avoids the marker bias inherent to SNP arrays, it has been found to be especially valuable for studies involving diverse or underrepresented populations, where array-based methods often struggle with imputation accuracy.
In a 2024 large-scale cohort study, BGE sequencing demonstrated >99% concordance with 30x whole genome sequencing and outperformed SNP arrays in variant detection across African, Latin American, and East Asian cohorts—underscoring its value as a more inclusive platform for population genomics2.
Potential of BGE sequencing for clinical applications
BGE is no longer limited to research settings. In rare disease diagnostics, the deep exome layer provides high-confidence detection of pathogenic coding variants, while the genome-wide coverage adds structural and regulatory context—critical for interpreting variants of uncertain significance1.
BGE is also increasingly being explored for pharmacogenomics and polygenic risk scoring, combining precision exome sequencing with genome-wide imputation to support more personalized approaches to risk prediction and treatment planning1.
Among other emerging clinical applications, BGE also has strong potential in reproductive genomics—supporting both parental carrier screening and future integration into embryo or prenatal screening workflows. Its blended design could enable simultaneous detection of inherited risk variants and broader genomic anomalies, although clinical implementation in these areas is still in development.
With its ability to adapt to both population-scale and individual-level analysis, BGE is rapidly becoming a go-to approach for projects that demand breadth and depth without the price tag of full WGS.
At Eremid, we offer research-grade and CLIA/CAP-validated BGE workflows for clinical projects, with customizable coverage and interpretation pipelines based on sample type, population background, and project goals.
Why long reads could be the next frontier for BGE
Today, most BGE workflows are built around short-read sequencing platforms such as Illumina, which offer high-throughput, cost-effective processing and broad support for exome and genome library prep. These systems are well-suited for workflows where throughput, accuracy and cost are the primary considerations.
Short-reads deliver reliable variant calling in well-mapped exonic regions, and when paired with large reference panels, low-pass genome data can be imputed with high accuracy across much of the genome. As a result, short-read BGE has already been successfully deployed in both research and clinical applications.
But as long-read sequencing platforms from companies like PacBio® and Oxford Nanopore® continue to mature, they offer exciting potential to enhance the BGE model even further.
At Eremid, our beta long-read (LR) BGE workflow uses the same blended data structure as short-read—mixed with PacBio® technology—to enable structural variant resolution, deep phasing, and enriched access to difficult-to-sequence genomic regions. The LR-BGE workflow can provide a greater resolution and broader detection of genomic variants, opening up new possibilities for both research and clinical applications.
LR-BGE sequencing brings several key advantages:
• Broader and more even coverage, including improved access to GC-rich and highly repetitive regions.
• Superior structural variant detection, including large insertions/deletions, inversions, and complex rearrangements often missed by short-reads.
• High-fidelity phasing and haplotype resolution, important for compound heterozygosity and cis/trans analysis.
• Greater accuracy in difficult-to-map regions, reducing alignment ambiguity and false positives.
While long-read BGE is still an emerging approach, its core benefits—depth, completeness and clarity—align directly with the goals of BGE sequencing. As costs continue to fall and throughput improves, long-read platforms offer a high-resolution, low-bias upgrade path for research and clinical teams already invested in the BGE model.
Table 3: Short-read versus long-read approaches for BGE sequencing.
Ready to run with BGE? So are we.
Blended genome exome sequencing is a smart, scalable approach that balances breadth, depth, and cost—making it ideally suited for everything from large-scale cohort studies to targeted rare disease investigations. It delivers the coverage you need without the overhead of deep WGS, and the flexibility to scale as your research evolves.
At Eremid® Genomic Services, we have the platforms, expertise and in-house capacity to take on BGE projects of any size. Our state-of-the-art Genomics Services Lab offers both short-read sequencing on Illumina’s X platform and long-read technologies, including PacBio’s Revio system, enabling us to tailor your project to exact requirements—whether you’re prioritizing imputation performance, structural variant detection, or rare variant resolution.
Our team brings unique expertise to create tailored BGE workflows, overcoming software limitations to create specific analysis pipelines for each project’s needs. Whether you’re working with challenging sample types, diverse populations, or clinical-grade requirements—we can help you build a custom BGE solution that maximizes output and minimizes compromise.
From project design to sequencing and bioinformatics, our experienced team works with you end-to-end to ensure data quality, reproducibility and scientific value at every stage.
Considering a BGE sequencing project? Talk to our team today and discover how Eremid can help you deliver deeper insights, faster.
References
1. DeFelice, M. et al. Blended Genome Exome (BGE) as a Cost Efficient Alternative to Deep Whole Genomes or Arrays. bioRxiv (2024) doi:10.1101/2024.04.03.587209.
2. Boltz, T. A. et al. A blended genome and exome sequencing method captures genetic variation in an unbiased, high-quality, and cost-effective manner. bioRxiv 2024.09.06.611689 (2024) doi:10.1101/2024.09.06.611689.
