Revolutionizing Genomics: Roche's Sequencing by Expansion (SBX) Technology and Data Analysis Advancements

The world of DNA sequencing is undergoing a transformation, driven by the need for faster, more efficient, and cost-effective methods. Roche is at the forefront of this revolution, developing innovative solutions that address the fundamental barriers hindering genomic discovery. This article explores Roche's Sequencing by Expansion (SBX) technology, its associated data analysis advancements, and the potential impact on research and clinical applications.

The Data Deluge in DNA Sequencing

DNA sequencing generates massive amounts of raw data. Billions of genetic fragments pile up as terabytes of raw data. The data analysis process is critical for transforming this chaos into organized, usable information. This includes capturing the raw signal, turning it into readable genetic code, analyzing, organizing, and storing those files, and sharing the data so scientists can use it in their research.

Current Challenges in DNA Sequencing Data Analysis

Genomics researchers face the challenge of balancing cost efficiency, flexibility, and turnaround time. Researchers often maximize the number of samples they can get from a sequencing run to drive cost efficiency, which can mean waiting until they have all of those samples and maximizing the amount of data generated. Once a sufficient number of samples are acquired, it can take one to two days just to generate the raw data, and the entire run needs to be complete before downstream analysis can begin.

Roche's Innovative Approach: Sequencing by Expansion (SBX)

Roche is working to improve DNA sequencing data analysis by developing sequencing by expansion (SBX), a new category of next-generation sequencing (NGS) that will enable faster and more flexible sequencing. SBX technology converts DNA information into a longer, “expanded” molecule, overcoming the spatial challenges of current nanopore technology and enabling higher signal-to-noise for improved accuracy.

How SBX Works

SBX technology utilizes a proprietary biochemical conversion process to expand and encode the sequence of a DNA template into an Xpandomer molecule. The building blocks of the Xpandomer are expandable nucleotide triphosphates, or X-NTPs. Each of the four, easily differentiated X-NTPs (one for each base) acts as the substrate for template-dependent, polymerase-based replication. XP Synthase, has been carefully engineered to incorporate large X-NTP monomers-enabling >99.3% mean raw read accuracy, uniform GC coverage, and longer read lengths. By stabilizing the extending molecule, PEMs play an important role in increasing read length beyond traditional short-read sequencing technologies. The Xpandomer molecule is then routed through a biological nanopore in a highly efficient and accurate manner. The highly differentiated reporter codes are easily measured during this translocation process via a scalable complementary metal oxide semiconductor (CMOS)-based array, which combines electrodes, detection circuits, and analog-to-digital conversion. Because the CMOS array contains roughly eight million microwells (each containing a nanopore), measurement occurs in a massively parallel, highly controlled manner without the convolution issues of traditional nanopore sequencing.

Read also: Your Guide to Roche Internships

Real-Time Data Analysis with SBX

SBX is designed to analyze data as it’s generated, where it’s generated. The approach leverages the inherent nature of SBX and centers on accelerated real-time downstream analysis, while offering flexibility to users through "data offramps." These offramps are changeable exit points within the data analysis workflow that let researchers choose exactly what they need. Researchers can use various offramps to fit their needs, regardless of batch size. Additional features of the approach allow researchers to more specifically identify data of interest, enabling labs to discard unwanted data at the source, which helps to mitigate downstream data transfer and storage challenges.

Working with an early access partner, Roche demonstrated at European Human Genetics Conference (ESHG) in May 2025 that the technology is able to go from library preparation to identifying genetic variants for a single whole genome sample in under five hours - a process that can often take one to two days.

SBX Optimized Open-Source (XOOS) Analysis Tools

Given the unique properties of SBX reads, the Roche team has developed a suite of permissive, free, SBX-optimised open source analysis tools (XOOS) to serve as the foundation upon which the SBX ecosystem will be built. The goal is to empower the NGS community and lower the barrier of adoption for SBX technology. In parallel, Zhao’s team has been working closely with the Google DeepVariant team to adapt DeepVariant for SBX, achieving strong performance on Whole Genome germline variant analysis.

Benefits of SBX and Real-Time Data Analysis

The combination of SBX technology and real-time data analysis offers several key benefits:

Faster Turnaround Time: Real-time data processing significantly reduces the time required to obtain sequencing results.
Flexibility: Data offramps provide researchers with the flexibility to access data at various stages of the analysis workflow.
Cost Efficiency: The ability to discard unwanted data at the source reduces downstream data transfer and storage costs.
Improved Data Management: SBX simplifies data management by allowing researchers to focus on the data of interest.

Applications of SBX Technology

The speed, flexibility, and efficiency of SBX technology make it well-suited for a wide range of applications, including:

Cancer Research: SBX can accelerate the identification of genetic variants in tumor samples, leading to more effective treatments.
Infectious Disease Surveillance: SBX can be used to rapidly sequence viral genomes, enabling the discovery of new mutations and the tracking of strain transmission.
Genomic Discovery: SBX can facilitate the discovery of new genes and genetic pathways, leading to a better understanding of human health and disease.

Expert Insights: Key People Behind SBX Development

Several experts are driving the development and commercialization of SBX technology and its associated data analysis tools.

Jaedon: Sr. Jaedon drives the commercial strategy and product roadmap for both sequencing platform management software, and SBX analysis tools, including the SBX Optimized Open-Source (XOOS) analysis tools. Prior to joining Roche in 2023, Jaedon held multiple roles at Illumina, contributing to several NGS product launches, most recently as a product manager supporting the commercialization of the NovaSeq X series.
Mahdi Golkaram: Mahdi Golkaram is the Director of Bioinformatics Data Science at Roche, where he leads algorithm development for the secondary analysis of Sequencing by Expansion (SBX) data. By integrating cutting-edge bioinformatics algorithms and data-driven approaches, including deep learning and artificial intelligence, XOOS enables SBX users to detect genetic alterations from sequencing data to support a wide range of applications.
John Mannion: John Mannion leads a global team of computational scientists and engineers, driving informatics research and development across Molecular Labs instrument product areas. Within sequencing, Mannion’s teams are responsible for development of analysis pipelines on accelerated hardware to process high throughput genomics data on edge compute, as close to the source of data generation as possible.
Andrew Carroll: Andrew Carroll is product lead for the genomics team in Google Research, where he coordinates the development of deep learning methods that operate on genomic data and in combination with other modalities such as imaging. Prior to joining Google, Carroll was chief scientific officer at DNAnexus. He led a team that developed and operated platforms powered by DNAnexus, including PrecisionFDA, and the St. Jude Pediatric Cancer Cloud.
Chen Zhao: Chen Zhao leads a global team of computational biologists and software engineers at Roche Diagnostics. During Zhao’s tenure at Roche, he and his team have helped drive the research and development efforts of Roche’s SBX sequencing platform, focusing on secondary analysis algorithms and application development. His team is also responsible for delivering production-grade, highly scalable, and accelerated open-source software. In addition to sequencing, Zhao and his team are also responsible for Roche’s Cobas x800 and Liat platforms when it comes to multiplexed PCR assay design and regulatory submission. Prior to joining Roche in 2022, Chen held multiple roles at Illumina where he led the oncology bioinformatics team to develop various tissue and liquid biopsy products (e.g., TST170, TSO500).

Additional NGS Tools: KAPA Target Enrichment Portfolio

Roche offers a range of NGS tools, including the KAPA Target Enrichment Portfolio. KAPA HyperPETE is a novel target enrichment technology that is designed to combine the performance of hybrid-capture with the ease-of-use and turnaround time of amplicon-based workflows.

Webinar on SBX-Duplex Data

To learn more about SBX-Duplex data and whole-genome germline small variant calling, Roche is hosting an in-depth webinar.

Read also: Leadership at La Roche

tags: #roche #sequencing #webinar