Skip to main content
Premium Trial:

Request an Annual Quote

Canopy Biosciences Aims to Improve Rare Variant Detection With Error-Correction Sequencing Service


NEW YORK (GenomeWeb) – After receiving an exclusive license from the Washington University in St. Louis for its error-correction sequencing technology, startup Canopy Biosciences now plans to develop the bioinformatic tool further to identify and monitor ultra-rare gene variants in patient blood samples.

With the new license, Canopy will build upon its gene expression analysis portfolio by offering a streamlined research-use-only service for next-generating sequencing applications.

While next-generation sequencing costs have dramatically dropped over the past decade, researchers are still plagued by statistical errors cropping up during the sequencing process.

"When you're thinking about applications where you're looking for something very rare — which has a 1 percent error rate — you won't be able to distinguish that from the noise with the platform," Canopy Cofounder and Chief Operating Officer Crystal Winkeler explained. "What error-corrected sequencing allows you to do is distinguish true variance from sequencing errors."

Founded in 2016, Canopy initially began offering genetic engineering services and CRISPR products to customers in the research field. Winkeler and her colleagues built the firm around a management team that identified, developed, and rapidly commercialized technologies from academic institutions.

Winkeler explained that the firm's error-corrected sequencing platform begins by inserting unique molecular identifiers (UMIs) in the sample during the standard library prep process, prior to amplifying the desired targets.

After sequencing the material on an NGS platform such as Illumina's MiSeq, the team runs an error-correction algorithm that generates read families using the UMIs, computationally removes sequencing errors, and minimizes background noise. According to Winkeler, the extra step drops the error rate for NGS from 5 in 100 sequences to 1 in 10,000.

While Winkeler noted the turnaround time for the service is largely dependent on the number of samples users wish to sequence, she estimated that the average time from sample to result is between four and six weeks.

WUSTL associate professor of pediatrics Todd Druley initially developed the algorithm to detect minimal residual disease in pediatric acute myeloid leukemia (AML). He explained that he and his team wanted to use sequencing for leukemia mutations instead of flow cytometry, but they found that standard NGS had an inherent error rate of between 1 in 200 and 1 in 100.

"We [therefore] sought to develop a sequencing strategy that could perform comparably to flow cytometry, but also provide gene-specific information that could be used to not only quantify residual leukemia, but hopefully inform precision therapy," Druley explained. "We [also] needed to first quantify what a mutational spectrum would look like in order to identify abnormal [cases]."

Since then, Winkeler highlighted that customers have used the tool in several clinical situations. One application included identifying leukemia-associated mutations in banked pre-leukemia blood and bone marrow from patients with either therapy-related AML or therapy-related myelodysplastic syndrome.

Winkeler said that researchers can also use the tool to identify a wide variety of ultra-rare gene variants for drug discovery. Emphasizing that current technology enables researchers to examine both the RNA and protein levels, Winkeler argued that error-correction sequencing provides technology that will "allow them to look in-depth on the DNA level as well."

Winkeler highlighted that her team found that the error-correction technology works with Illumina's TrueSight chemistry, which searches for variants in 54 genes important in myeloid cancer.

In addition, Winkeler believes that researchers will eventually be able to use the technology in liquid biopsies to identify circulating DNA from solid tumors.

Druley explained that his team encountered several challenges regarding data analysis and input while developing the approach. WUSTL internal medicine resident Andrew Young and his colleagues at the Druley lab struggled to decide how much DNA input was required to identify a rare mutation. In addition, Druley said that the group needed time to identify reagents that optimized the approach's overall performance and develop bioinformatic strategies "to further mitigate systemic errors and optimize specificity."

While not a spinout of WUSTL, Winkeler explained that Canopy has closely partnered with the university to commercialize several of its research projects, including the TUNR genetic editing kits. In addition, several of Canopy's employees have previously been affiliated with the university and academic research projects. Winkeler said that the groups have signed a worldwide exclusive license to develop the error-correction technology for the research market and for analyte-specific reagents.

Canopy therefore plans to launch a service based on the error-correction technology within the next few months for users wishing to integrate the tool as part of their sequencing workflow. According to Winkeler, the firm aims to "marry the chemistry of this error correction with as many types of sample prep as possible."

Winkeler explained that Canopy will sequence liquid and tissue biopsy samples sent in by research groups, followed by a detailed report generated through data analysis.

"Our error-corrected sequencing service will offer researchers the ability to examine DNA mutations and small insertions/deletions as rare as 1 in 10,000, which is a sensitivity that's 100x greater than sequencing without a UMI-based approach," Winkeler said.

Canopy has not disclosed pricing for the service, but Winkler said the firm will release more information in the next couple of months. She believes the technology will eventually allow researchers to identify rare variants in cancer patients for applications such as recurrence monitoring.  

"For example, you have a patient and you've identified the disease-causing mutation," Winkeler noted. "If that patient is in remission, and you want to monitor their recurrence, you're going to need an extremely sensitive technology like error-corrected sequencing to catch the first sign of recurrence."

Druley believes that future research applications will be retrospective analysis of human clinical samples, quantifying clonal hematopoiesis in animal models, and searching for phenotype correlation. Like Winkeler, Druley envisions clinical applications expanding to circulating tumor or cell-free DNA profiling, in addition to solid tumor profiling and bone marrow transplantation surveillance.

"Given the clinical connections with clonal hematopoiesis and the growing arsenal of gene [and] mechanism-specific anti-cancer agents, the need for precise clonal profiling will only grow," he added.