Skip to main content
Premium Trial:

Request an Annual Quote

Genomic Health Scales up Bioinformatics as it Adopts Next-Gen Sequencing for Biomarker Discovery

Premium

By Uduak Grace Thomas

Genomic Health is in the process of ramping up its bioinformatics capacity as part of a strategy to harness next-generation DNA and RNA sequencing in its early biomarker discovery programs.

Earlier this month, during a call to discuss the company's second-quarter financial results, CEO Kim Popovits said that during the quarter the company had "significantly expanded its information storage, bioinformatics, and analytical capabilities as our research team optimized its methods to generate large amounts of genomic information for our early biomarker discovery programs" using second-gen sequencing.

In a follow-up interview with BioInform this week, Genomic Health officials said that the company currently has between 50 and 60 employees in its information technology, bioinformatics, and biostatistics arms and plans to add staff as its next-gen sequence data analysis capabilities develop.

In addition, Joffre Baker, the company's chief scientific officer, said that the firm has increased its data storage capacity to handle "several scores of terabytes worth of data," and that it will continue to scale up its storage over time as needed.

In terms of analytical tools, Baker said that the company is evaluating several open source software packages and will also develop its own in-house tools to deal with the deluge of data it anticipates will arise from the expansion of its next-gen sequencing efforts.

So far, the company has an Illumina Genome Analyzer II and plans to add a HiSeq 2000. Officials said the firm is "looking at" sequencers from other vendors but did not elaborate.

To date, Genomic Health has used real-time PCR for its biomarker discovery efforts. For example, its flagship Oncotype Dx test for breast cancer recurrence, which it launched in 2004, was developed based on a review of published breast cancer biomarkers, which led to an initial set of 250 gene candidates that it then evaluated via RT-PCR to identify a set of 21 gene expression assays that make up the test.

Baker told BioInform this week that going forward, the company plans to move this analysis to next-gen sequencing platforms. As part of this effort, it plans to develop methods for full transcriptome profiling and target sequence capture using a blend of open source software and in-house tools.

Randy Scott, Genomic Health's executive chairman, said that while the company's long-term goal is to move its next-gen sequencing efforts toward the development of "a universal cancer assay" that combines whole-genome and transcriptome analysis, "ultimately the great value is going to be in the management of the information" and turning the information into a form that clinicians can use to make treatment decisions, making it one of the "fastest growing areas of the company for years to come."

Baker pointed out that there is "an enormous amount of activity" in the bioinformatics community to develop open source tools for evaluating genomic data.

Genomic Health is "evaluating algorithms that have been developed outside to tackle these questions," Baker said, though he noted that "the bioinformatics around these questions is far from optimized anywhere in the world."

Scott agreed, adding that one of Genomic Health's informatics goals is to make it easier to compare data from multiple large-scale clinical trials involving thousands of patient samples in a "rigorous clinical manner."

According to Baker, Genomic Health is currently exploring open source in tools in areas such as transcriptomics in "an agnostic mode."

For example, the firm is considering several tools commonly used by genomics researchers, such as TopHat, which maps splice junctions for RNA-seq reads and CuffLinks, which assembles RNA transcripts.

Scott explained. "[We are looking at] how you define regions of transcription that are outside the known exon/intron boundaries … and [determine] which ones are actually meaningful regulatory elements and which are noise within the transcriptome."

The firm also plans to delve into tools that can be used to discover de novo RNA splicing as well as to develop tools that can present genomic data in meaningful ways to clinicians.

"Part of that is determining how to normalize data from different patients to be able to correct for differences in the quality of the RNA," Baker said. "You want to separate signals that you are getting because of technical differences from the real biological signals."

In terms of in-house tool development, Scott said that the company would focus on filling gaps and would not "reinvent the wheel."

The company is also considering some commercial offerings, though Baker and Scott declined to provide specifics.

Scott added that the company has long-term plans to build a standalone data processing platform that would be made broadly available to clinical researchers and pharmaceutical companies, but did not elaborate.

The Scan

Booster for At-Risk

The New York Times reports that the US Food and Drug Administration has authorized a third dose of the Pfizer-BioNTech SARS-CoV-2 vaccine for people over 65 or at increased risk.

Preprints OK to Mention Again

Nature News reports the Australian Research Council has changed its new policy and now allows preprints to be cited in grant applications.

Hundreds of Millions More to Share

The US plans to purchase and donate 500 million additional SARS-CoV-2 vaccine doses, according to the Washington Post.

Nature Papers Examine Molecular Program Differences Influencing Neural Cells, Population History of Polynesia

In Nature this week: changes in molecular program during embryonic development leads to different neural cell types, and more.